Effective Nuclear Charge and Shielding

A 3s electron in sodium and a 3p electron in the same atom feel different pulls from the nucleus. This is true even though both sit in the shell with principal quantum number 3. This single fact explains the entire shape of the periodic table. Energy in a many-electron atom depends on both $n$ and $l$ , not on $n$ alone. This is why transition metals exist as a distinct block. It is also why potassium's last electron goes into 4s rather than 3d. The culprit is shielding. Every electron partially screens the nuclear charge felt by every other electron. It does so unevenly. Electrons in different orbitals occupy different regions of space relative to the nucleus and to each other. The orbital approximation already tells us something important. A many-electron wavefunction is built, approximately, from one-electron hydrogen-like orbitals. But it leaves open a quantitative question: what nuclear charge does each orbital actually experience? The honest answer is that the true many-body Hamiltonian has no clean single answer. The electron–electron repulsion term $\sum_{i<j} e^2/4\pi\varepsilon_0 r_{ij}$ mixes everything together. The effective nuclear charge $Z_{\text{eff}}$ is the physicist's and chemist's workaround. It replaces the messy, fluctuating repulsion felt by one electron with a single averaged, radially-dependent screening constant $\sigma$ . So $Z_{\text{eff}} = Z - \sigma$ stands in for "how much bare nuclear charge actually reaches this electron once everyone else's negative charge is subtracted off." The physical mechanism behind why $\sigma$ depends on $l$ , and not just $n$ , is penetration. This is best understood through the radial distribution function $4\pi r^2 R_{nl}^2(r)$ . An $s$ orbital has a nonzero probability density at the nucleus. Its radial distribution function has an inner lobe that dives deep inside the electron cloud of the core. It spends real time very close to the bare nucleus, where shielding is weak. A $p$ orbital of the same $n$ is forced by its angular node structure to vanish at $r=0$ . It is on average kept further out and shielded more effectively. The result is an energy ordering $E_{ns} < E_{np} < E_{nd} < E_{nf}$ for the same $n$ . The degeneracy that hydrogen enjoys in $l$ is broken as soon as a second electron enters the picture. This article builds $Z_{\text{eff}}$ up from first principles. It covers the hydrogenic energy formula and the physical meaning of shielding and penetration. It includes a fully worked derivation connecting radial probability density to screening. It also presents three worked examples. These span a textbook calculation, a spectroscopic determination from real ionization energies, and a competition-style multi-electron ranking problem. It sets up, but does not itself derive, the bookkeeping shortcut known as Slater's Rules. It also sets up the ordering logic behind the Aufbau Principle and Hund's Rules.

Key takeaways

In one-electron hydrogen, energy depends only on $n$ : $E_n = -13.6\,\text{eV}\, Z^2/n^2$ , and all orbitals of a given $n$ are degenerate.
In many-electron atoms, electron–electron repulsion breaks this degeneracy: energy depends on both $n$ and $l$ , with $E_{ns} < E_{np} < E_{nd}$ for fixed $n$ .
Effective nuclear charge $Z_{\text{eff}} = Z - \sigma$ captures the average charge an electron actually feels after other electrons partially cancel the nucleus.
Penetration — the ability of an orbital's radial distribution to have density close to the nucleus, inside the shielding cloud of core electrons — controls $\sigma$ and hence the $n,l$ splitting.
$Z_{\text{eff}}$ can be estimated empirically from spectroscopic ionization energies via $Z_{\text{eff}} = n\sqrt{-E_n/13.6\,\text{eV}}$ , or approximated by rule-based schemes.
$Z_{\text{eff}}$ governs atomic size, ionization energy, electron affinity, and orbital filling order across the periodic table.
Key idea: electrons of the same $n$ but different $l$ penetrate the core to different extents, so they feel different effective nuclear charges, and this — not some new quantum number — is what splits hydrogenic degeneracy in real atoms.

Where this lives today

Effective nuclear charge is not a historical curiosity confined to introductory courses; it is the conceptual backbone of essentially every qualitative and semi-quantitative trend chemists use daily. Atomic and ionic radii, ionization energies, electron affinities, and electronegativities are all, at heart, statements about how strongly the outermost electron is bound, and that binding strength is set by $Z_{\text{eff}}/n$ . The characteristic shrinkage across the lanthanide series (the "lanthanide contraction") is a direct consequence of poorly-penetrating $4f$ electrons providing weak shielding for the $5s$ / $5p$ / $6s$ electrons that determine ionic radius, which is why elements just after the lanthanides (hafnium, tantalum) have radii anomalously close to their lighter congeners (zirconium, niobium) — a fact with real consequences for separating these metals industrially, since their nearly identical $Z_{\text{eff}}$ -controlled radii make hafnium and zirconium notoriously difficult to purify from each other. In transition-metal and organometallic chemistry, the near-degeneracy and configuration-dependent ordering of $(n-1)d$ and $ns$ orbitals — exactly the subtlety worked through in the iron example above — underlies crystal-field and ligand-field theory, catalytic cycles that shuttle metals between oxidation states, and the color and magnetism of coordination compounds, all of which hinge on how $Z_{\text{eff}}$ for $d$ orbitals shifts as ligands and oxidation state change. In spectroscopy and instrumentation, X-ray photoelectron spectroscopy (XPS) directly measures core-electron binding energies whose chemical shifts are interpreted through changes in effective nuclear charge felt by core electrons when the surrounding valence electron density (and hence screening) changes with oxidation state or bonding environment — a workhorse technique for surface and materials analysis. In computational chemistry, effective core potentials (pseudopotentials) used to make density functional theory and post-Hartree–Fock calculations tractable for heavy elements are, at their conceptual core, a numerically refined and quantum-mechanically rigorous version of exactly the same idea: replace the detailed dynamics of core electrons with an effective potential that reproduces the correct $Z_{\text{eff}}(r)$ felt by the valence electrons, without explicitly propagating every core electron's wavefunction.

Ask the tutor: How do Slater's empirical rules actually assign numerical values to the shielding constant σ for each type of electron, and why do they treat (n−1) shells differently from same-shell electrons?

The physical picture

Start from the hydrogen atom, the one exactly solvable case. Its orbital energies come purely from the balance between kinetic energy and the attraction to a single point charge $+Ze$ :

E_n = -\frac{13.6\,\text{eV}\, Z^2}{n^2}

This depends only on $n$ . A $2s$ and a $2p$ electron in hydrogen have exactly the same energy, a degeneracy of $2n^2$ states per shell (including spin) that is a special, almost accidental feature of the pure $1/r$ Coulomb potential — a consequence of a hidden dynamical symmetry, not something guaranteed for arbitrary central potentials. The moment a second electron is added, this accidental degeneracy is destroyed. Each electron now moves not in a bare Coulomb field $-Ze^2/4\pi\varepsilon_0 r$ but in the field of the nucleus plus the smeared-out, time-averaged charge distribution of every other electron. Because the other electrons are, on average, closer to the nucleus than the electron in question is not always true — they overlap in space — the net potential seen by one electron is no longer a clean $1/r$ potential. It becomes steeper than $1/r$ at short range (where nuclear charge dominates, unscreened) and shallower than $1/r$ at long range (where the other electrons have screened most of the nuclear charge, leaving an effective net charge as low as $+1e$ for a neutral atom). A potential that departs from pure $1/r$ splits states by $l$ at fixed $n$ , because the centrifugal barrier $\hbar^2 l(l+1)/2m_er^2$ pushes different $l$ orbitals to sample the deviation from $1/r$ differently. The physically transparent way to see this is penetration. Recall the radial distribution function $P_{nl}(r) = 4\pi r^2 |R_{nl}(r)|^2$ , which gives the probability of finding the electron in a thin spherical shell at radius $r$ . For $n=3$ in a many-electron atom:

The $3s$ radial distribution has three peaks: two small inner lobes very close to the nucleus (inside the region occupied by the $1s^2 2s^2 2p^6$ core) and one large outer lobe.
The $3p$ radial distribution has two peaks: one small inner lobe and one large outer lobe — less inner density than $3s$ .
The $3d$ radial distribution has only one peak, entirely outside the core, with essentially zero density at small $r$ .

Because $3s$ has real probability density deep inside the core — where the core electrons haven't yet finished screening the nucleus — the $3s$ electron feels, part of the time, close to the full unscreened nuclear charge $Z$ . The $3d$ electron almost never gets inside the core; it always sees the core electrons already between it and the nucleus, so it experiences a nearly fully screened charge close to $Z - (\text{number of core electrons})$ . The ordering of average screening is

\sigma_{3s} < \sigma_{3p} < \sigma_{3d}

which, combined with $Z_{\text{eff}} = Z-\sigma$ , gives $Z_{\text{eff},3s} > Z_{\text{eff},3p} > Z_{\text{eff},3d}$ and therefore, since energy scales as $-Z_{\text{eff}}^2/n^2$ , the ordering

E_{3s} < E_{3p} < E_{3d}

Figure

Radial distribution functions for n = 3 orbitals showing differential penetration into the atomic core

The consequence propagates across the whole periodic table. As $n$ and $l$ both grow, orbitals with high $l$ are pushed up in energy so much by poor penetration that they can cross entire shells — the reason 4s fills before 3d, and 6s and 4f interleave in the lanthanides. This crossing is a direct, quantitative statement about $Z_{\text{eff}}$ : it is not that $n=4$ is "closer" in energy to $n=3$ in some vague sense, it is that the $3d$ orbital's poor penetration leaves its $Z_{\text{eff}}$ so low that $-Z_{\text{eff}}^2/n^2$ for $3d$ ends up numerically close to, and in early transition metals higher than, that of $4s$ .

Figure

Effective nuclear charge vs. atomic number for 2s/2p (top) and 3s/3p/3d (bottom) showing the l-splitting

The core ideas

The central working formula rewrites the hydrogenic energy expression with the true nuclear charge $Z$ replaced by an effective charge $Z_{\text{eff}}$ specific to each orbital:

E_{n} = -\left(13.6\,\text{eV}\right)\frac{Z_{\text{eff}}^2}{n^2}, \qquad Z_{\text{eff}} = Z - \sigma

where:

$E_n$ is the orbital energy of the electron of interest (negative, bound),
$13.6\,\text{eV}$ is the Rydberg energy, $R_H hc$ , the hydrogen ground-state binding energy,
$Z$ is the true (bare) nuclear charge, an integer equal to the atomic number,
$Z_{\text{eff}}$ is the effective nuclear charge felt by the electron in question,
$n$ is the principal quantum number of that electron's orbital,
$\sigma$ is the shielding (screening) constant, representing the average charge cancellation from all other electrons.

Reading: this equation says a many-electron orbital behaves, energetically, exactly like a hydrogenic orbital around a fictitious nucleus of charge $Z_{\text{eff}}$ instead of $Z$ . All of the complicated electron–electron repulsion is compressed into one number, $\sigma$ , that differs for each subshell because different subshells penetrate the core to different extents. The screening constant itself has no closed exact form (unlike $E_n$ for hydrogen, which is exact), but it obeys clear qualitative and semi-quantitative rules, summarized by the ordering of penetration:

\sigma_{ns} < \sigma_{np} < \sigma_{nd} < \sigma_{nf} \quad \text{for fixed } n

where the subscript denotes the subshell being screened. This inequality is the mathematical expression of "s penetrates most, f penetrates least," and it directly implies the energy ordering discussed above via $E \propto -Z_{\text{eff}}^2$ . A second core relationship links $Z_{\text{eff}}$ directly to experiment. Because ionization energy is (in the simplest single-electron picture) just $\text{IE} = -E_n$ , one can invert the energy formula to extract $Z_{\text{eff}}$ from a measured ionization energy:

Z_{\text{eff}} = n\sqrt{\dfrac{\text{IE}}{13.6\,\text{eV}}}

where:

$\text{IE}$ is the experimental ionization energy of the electron of interest, in eV,
$n$ is the principal quantum number of the orbital from which the electron is removed,
$13.6\,\text{eV}$ is again the Rydberg constant in energy units.

Reading: this equation turns $Z_{\text{eff}}$ from an abstract modeling parameter into a directly measurable quantity — every photoelectron spectrum or ionization-energy table is secretly a table of effective nuclear charges once you divide out $n$ and take a square root.

Derivation AFrom the central-field Schrödinger equation to Z_eff

Goal. Show explicitly how replacing the true many-electron potential with a central, effective potential naturally produces an effective nuclear charge that depends on $l$ through penetration, starting from the many-electron Hamiltonian.

Step 1 — Write the exact Hamiltonian. For an atom with $N$ electrons and nuclear charge $Z$ (in atomic units, energies in Hartree, distances in Bohr radii for compactness):

\hat{H} = \sum_{i=1}^{N}\left(-\frac{1}{2}\nabla_i^2 - \frac{Z}{r_i}\right) + \sum_{i<j}\frac{1}{r_{ij}}

The first sum is the kinetic energy plus bare nuclear attraction for each electron treated independently; the second sum is the full electron–electron repulsion, which couples every electron's coordinates to every other's and prevents exact separation into one-electron orbitals.

Step 2 — Apply the orbital (mean-field) approximation. Replace the instantaneous repulsion from all other electrons on electron $i$ by an averaged, spherically symmetric potential $V_i^{\text{eff}}(r_i)$ built from the charge densities of the other occupied orbitals:

V_i^{\text{eff}}(r_i) = -\frac{Z}{r_i} + \left\langle \sum_{j\neq i} \frac{1}{r_{ij}} \right\rangle_{\text{other electrons}}

This step is precisely the orbital approximation: each electron now moves independently in a common central potential, so the many-electron Schrödinger equation separates into $N$ one-electron radial equations of hydrogen-like form, each with its own effective potential rather than a pure $-Z/r$ .

Step 3 — Define the shielding function. Because $V_i^{\text{eff}}(r)$ is not simply $-Z_{\text{eff}}/r$ at every radius (the screening from the other electrons' charge density $\rho(r')$ is itself radius-dependent — an electron at small $r$ is inside less of the other electrons' charge cloud than one at large $r$ ), define an effective, radius-dependent nuclear charge through

V^{\text{eff}}(r) \equiv -\frac{Z_{\text{eff}}(r)}{r}, \qquad Z_{\text{eff}}(r) = Z - \int_0^r 4\pi r'^2 \rho_{\text{other}}(r')\, dr'

At $r \to 0$ , essentially none of the other electrons' charge lies inside radius $r$ , so $Z_{\text{eff}}(r) \to Z$ (nearly bare nucleus). At $r \to \infty$ , essentially all $N-1$ other electrons' charge lies inside, so $Z_{\text{eff}}(r) \to Z - (N-1)$ (for a neutral atom, this is $+1$ , matching the familiar fact that an electron very far from a neutral atom sees a net charge of $+1e$ , exactly as in a Rydberg orbital).

Step 4 — Fold in the angular (centrifugal) term to see why l matters. The full radial equation for orbital $nl$ is

-\frac{1}{2}\frac{d^2u_{nl}}{dr^2} + \left[V^{\text{eff}}(r) + \frac{l(l+1)}{2r^2}\right]u_{nl} = E_{nl}\, u_{nl}

The centrifugal term $l(l+1)/2r^2$ grows with $l$ and diverges as $r\to 0$ ; it acts as a repulsive barrier that suppresses the radial wavefunction's amplitude at small $r$ more strongly as $l$ increases. Physically, this is exactly the statement that high- $l$ orbitals cannot penetrate the core: the angular momentum barrier keeps them out of the region of small $r$ where $Z_{\text{eff}}(r)$ is largest.

Step 5 — Extract a single effective charge per subshell. Because $Z_{\text{eff}}(r)$ varies continuously with $r$ while textbook formulas want one number per orbital, define an orbital-averaged effective charge as the density-weighted average:

Z_{\text{eff}}^{(nl)} = \int_0^\infty Z_{\text{eff}}(r)\, P_{nl}(r)\, dr

where $P_{nl}(r) = 4\pi r^2 |R_{nl}(r)|^2$ is the radial distribution function of the orbital in question. Because $s$ orbitals have larger $P_{nl}(r)$ at small $r$ (where $Z_{\text{eff}}(r)$ is large) than $p$ , $d$ , or $f$ orbitals of the same $n$ , the weighted average $Z_{\text{eff}}^{(nl)}$ is necessarily larger for $s$ than for $p$ , larger for $p$ than $d$ , and so on. This closes the loop: the same physical fact (penetration, controlled by the centrifugal barrier) that appears in the exact radial equation reappears as the empirical ordering $Z_{\text{eff},ns} > Z_{\text{eff},np} > Z_{\text{eff},nd}$ used throughout chemistry.

Go deeper · Derivation B · 1 minRecovering Z_eff from measured ionization energy — the spectroscopic route

Derivation BRecovering Z_eff from measured ionization energy — the spectroscopic route

Goal. Derive the working formula that lets a chemist extract $Z_{\text{eff}}$ directly from an experimental ionization energy, and show its limiting behavior.

Step 1 — Start from the hydrogenic energy expression applied to the outermost (valence) electron. Treat the removal of one electron from an atom as approximately hydrogen-like: the departing electron sees a nucleus of effective charge $Z_{\text{eff}}$ , so its bound-state energy is

E_n = -13.6\,\text{eV} \cdot \frac{Z_{\text{eff}}^2}{n^2}

This is an approximation — real many-electron ionization involves relaxation of the remaining electron cloud after removal, and $Z_{\text{eff}}$ extracted this way is best thought of as an effective orbital energy parameter, not literally a snapshot of instantaneous charge.

Step 2 — Connect energy to ionization energy. By definition, the first ionization energy is the energy required to remove the electron from its bound orbital to $r=\infty$ with zero kinetic energy, i.e., $\text{IE} = 0 - E_n = -E_n$ . Substituting,

\text{IE} = 13.6\,\text{eV}\cdot\frac{Z_{\text{eff}}^2}{n^2}

Step 3 — Solve algebraically for $Z_{\text{eff}}$ . Rearranging,

Z_{\text{eff}}^2 = \frac{n^2\, \text{IE}}{13.6\,\text{eV}}

Z_{\text{eff}} = n\sqrt{\frac{\text{IE}}{13.6\,\text{eV}}}

Since $Z_{\text{eff}} > 0$ by definition (a positive effective attraction), only the positive root is physical.

Step 4 — Sanity-check against the one case with no shielding: hydrogen itself. For the hydrogen atom, $n=1$ and $\text{IE} = 13.6\,\text{eV}$ exactly, so

Z_{\text{eff}} = 1\times\sqrt{\frac{13.6}{13.6}} = 1

which correctly returns $\sigma = Z - Z_{\text{eff}} = 1 - 1 = 0$ : no shielding, because there is only one electron. This is the essential limiting case that any formula for $Z_{\text{eff}}$ must reproduce, and it is built into the derivation by construction since Step 1 assumed a purely hydrogenic form.

Step 5 — Extract the shielding constant. Once $Z_{\text{eff}}$ is known from experiment, the shielding constant follows immediately from its definition,

\sigma = Z - Z_{\text{eff}}

so that a full spectroscopic determination of $Z_{\text{eff}}$ automatically yields an empirical $\sigma$ for that orbital in that specific atom or ion — the same $\sigma$ that approximate schemes like Slater's rules are built to predict without needing the experimental number in hand.

Worked examples

Example 1 — Foundational: effective nuclear charge on the 3s electron of sodium from its ionization energy. Sodium's first ionization energy is $\text{IE}_1 = 5.14\,\text{eV}$ , removing the single $3s$ valence electron ( $n=3$ ). Using the spectroscopic formula derived above,

Z_{\text{eff}} = n\sqrt{\frac{\text{IE}}{13.6\,\text{eV}}} = 3\sqrt{\frac{5.14}{13.6}}

First evaluate the ratio inside the root:

\frac{5.14}{13.6} = 0.3779

Then take the square root:

\sqrt{0.3779} = 0.6148

Multiply by $n=3$ :

Z_{\text{eff}} = 3\times 0.6148 = 1.844

So the $3s$ electron in sodium behaves as though it orbits a nucleus of charge $+1.84$ , not the bare $Z=11$ . The shielding constant is

\sigma = Z - Z_{\text{eff}} = 11 - 1.844 = 9.16

This is a strikingly clean sanity check: sodium has exactly 10 core electrons ( $1s^2 2s^2 2p^6$ ) shielding the lone $3s$ valence electron. A shielding constant of $9.16$ — close to, but sensibly less than, 10 — says the core electrons screen almost one full unit of charge each, but not perfectly, because the $3s$ electron does penetrate the core somewhat (recall its inner radial lobes), reducing $\sigma$ slightly below the naive value of 10 that a "perfect point-charge core" model would predict.

Limiting-case check: if the $3s$ electron penetrated the core completely (spent zero probability outside it, the opposite of reality), $\sigma$ would equal the full nuclear charge minus 1 corresponding to feeling nearly the bare nucleus, i.e., $Z_{\text{eff}}\to 11$ ; if it penetrated not at all, $\sigma \to 10$ exactly and $Z_{\text{eff}}\to 1$ . The measured value $Z_{\text{eff}}=1.84$ sits just above the "perfect shielding" limit of 1, confirming modest but real penetration — physically sensible and quantitatively consistent with the radial distribution argument above.

Example 2 — Applied: comparing Z_eff for 3s vs. 3p across the same atom using ionization data, and predicting the energy gap. Consider magnesium, $Z=12$ , electron configuration $[\text{Ne}]3s^2$ . Its first ionization energy (removing a $3s$ electron) is $\text{IE}(3s) = 7.65\,\text{eV}$ . Now compare this to phosphorus, $Z=15$ , configuration $[\text{Ne}]3s^2 3p^3$ , whose first ionization energy (removing a $3p$ electron) is $\text{IE}(3p) = 10.49\,\text{eV}$ . These are different elements, so a direct comparison needs care — but the point of this example is to show how the same formula, applied within a row, reveals the $s$ -vs- $p$ split even though $Z$ itself is changing. For magnesium's $3s$ electron:

Z_{\text{eff}}(3s,\text{Mg}) = 3\sqrt{\frac{7.65}{13.6}} = 3\sqrt{0.5625} = 3\times 0.75 = 2.25

(Note $0.5625$ is a perfect square of $0.75$ , chosen here because Mg's real IE happens to land there — a pleasant numerical coincidence, not a rounding trick; $\sqrt{0.5625}=0.75$ exactly.)

\sigma(3s,\text{Mg}) = 12 - 2.25 = 9.75

For phosphorus's $3p$ electron:

Z_{\text{eff}}(3p,\text{P}) = 3\sqrt{\frac{10.49}{13.6}} = 3\sqrt{0.7713}

Evaluating the square root:

\sqrt{0.7713} = 0.8783

Z_{\text{eff}}(3p,\text{P}) = 3\times 0.8783 = 2.635

\sigma(3p,\text{P}) = 15 - 2.635 = 12.365

Both $Z_{\text{eff}}$ values increase steadily left to right across period 3 (a general trend, since nuclear charge grows faster than shielding within a shell), but the interesting comparison is the increment per proton: going from Mg to P adds 3 protons ( $12\to15$ ) and $Z_{\text{eff}}$ rises by $2.635-2.25=0.385$ , meaning roughly $0.13$ of effective charge is gained per added proton — each added proton is almost entirely (about 87%) screened by the added electron in the same shell, consistent with same-shell electrons being relatively poor (but not zero) mutual shielders. This "same-shell screening is imperfect but substantial" behavior is exactly the empirical foundation later formalized by Slater's rules for electrons in the same group.

Limiting-case check: if same-shell electrons shielded each other perfectly (shielding constant contribution of exactly 1.00 per added electron), $Z_{\text{eff}}$ would stay essentially flat across a period, contradicting the observed steady rise; the actually observed slow-but-real rise in $Z_{\text{eff}}$ across period 3 is the experimental signature that same-shell shielding is real but incomplete — typically quantified around $0.35$ per electron in Slater-type schemes, in the right ballpark as the $0.13$ – $0.39$ range one extracts crudely from IE data with different orbital types mixed in.

Example 3 — Competition-level: ranking Z_eff and predicting an orbital energy crossover. A common competition-style question: rank the following four one-electron removal energies from a hypothetical 3d transition metal ion in order of the effective nuclear charge felt by the removed electron, and predict whether $3d$ or $4s$ empties first upon ionization for a first-row transition metal such as iron. Consider neutral iron, $Z=26$ , ground-state configuration $[\text{Ar}]3d^6 4s^2$ . Using a Slater-type estimate (previewed here, not derived, since the full bookkeeping belongs to Slater's rules) one obtains approximate screening constants:

\sigma_{4s} \approx 22.25, \qquad \sigma_{3d} \approx 20.65

giving

Z_{\text{eff}}(4s) = 26 - 22.25 = 3.75

Z_{\text{eff}}(3d) = 26 - 20.65 = 5.35

At first glance this seems to say the $3d$ electron, feeling the larger $Z_{\text{eff}} = 5.35$ , should be more tightly bound than a bare formula $E\propto -Z_{\text{eff}}^2/n^2$ would suggest for $4s$ . Check the orbital energies explicitly:

E_{4s} = -13.6\,\text{eV}\times\frac{(3.75)^2}{4^2} = -13.6\times\frac{14.06}{16} = -13.6\times 0.8789 = -11.95\,\text{eV}

E_{3d} = -13.6\,\text{eV}\times\frac{(5.35)^2}{3^2} = -13.6\times\frac{28.62}{9} = -13.6\times 3.180 = -43.25\,\text{eV}

This says the $3d$ orbital, once occupied, sits lower in energy than $4s$ — which looks at first like it contradicts the aufbau filling order (4s fills first). The resolution is the well-known subtlety: the filling order (which orbital gets electrons first as $Z$ increases across the neutral-atom series) is governed by the total-energy-minimizing configuration for a neutral atom being built up, which is not the same question as which orbital is lower in energy for a fixed, already-formed configuration like $\text{Fe}^0$ or $\text{Fe}^{2+}$ . Once $3d$ is substantially populated, the strong core-like screening by other $3d$ electrons drops $\sigma_{4s}$ 's effectiveness while $3d$ – $3d$ mutual shielding is comparatively weak (same-shell, same- $l$ shielding is typically less complete than filled-core shielding), so in the ion $3d$ ends up markedly more stable than $4s$ — which is exactly why transition metal cations lose their $4s$ electrons before their $3d$ electrons upon ionization (e.g., $\text{Fe}\to\text{Fe}^{2+}$ removes both $4s$ electrons, not two $3d$ electrons), even though $4s$ filled first in the neutral building-up process.

Limiting-case check: for a hypothetical atom with $Z_{\text{eff}}(3d) = Z_{\text{eff}}(4s)$ exactly, the $E\propto -Z_{\text{eff}}^2/n^2$ formula would force $3d$ ( $n=3$ ) to sit strictly below $4s$ ( $n=4$ ) purely from the $1/n^2$ factor, since a smaller $n$ with equal $Z_{\text{eff}}$ always binds more tightly — showing that the celebrated "4s below 3d" ordering in neutral atoms only ever survives because $Z_{\text{eff}}(4s)$ is pushed low enough (by 3d's poor same-shell shielding of 4s) to overcome this $1/n^2$ penalty, and even then only barely, and only in the neutral atom build-up context, not in the ionized species.

Limiting cases, scope & common traps

Regime	Behavior of $Z_{\text{eff}}$	Why
One-electron atom/ion ( $\text{H}$ , $\text{He}^+$ , $\text{Li}^{2+}$ )	$Z_{\text{eff}} = Z$ exactly, $\sigma=0$	No other electrons to screen; hydrogenic formula is exact, not approximate
Core electrons ( $1s$ in a heavy atom)	$Z_{\text{eff}} \approx Z$ , only slightly reduced	Little or nothing lies between them and the nucleus to screen
Valence electron in a neutral atom	$Z_{\text{eff}} \ll Z$ , often $1$ – $3$	Nearly the full core cancels the nuclear charge, leaving a small net attraction
Same $n$ , increasing $l$ (fixed atom)	$Z_{\text{eff}}$ decreases: $s>p>d>f$	Higher- $l$ orbitals are pushed out by the centrifugal barrier, penetrate the core less
Across a period (fixed $n$ , adding protons+electrons)	$Z_{\text{eff}}$ rises slowly	Same-shell shielding is real but incomplete, so added protons are not fully canceled
Down a group (fixed valence subshell type, increasing $n$ )	$Z_{\text{eff}}$ roughly constant, slight increase	New shells add both a proton and a screening electron, roughly balanced
Highly excited Rydberg orbitals ( $n$ large, one electron far out)	$Z_{\text{eff}}\to 1$ for a neutral atom	Electron sees essentially a point charge of $+1e$ from the completely-enclosed core

Common trapTwo frequent misconceptions

First, $Z_{\text{eff}}$ is not a literal, instantaneous physical charge sitting at the nucleus — it is an orbital-averaged bookkeeping device that summarizes a genuinely dynamic, correlated many-body interaction into one number per orbital. Different definitions (spectroscopic, Slater, Clementi–Raimondi self-consistent field values) give somewhat different numbers for the "same" $Z_{\text{eff}}$ , and none of them is more "true" than another; they are different averaging conventions. Second, "4s fills before 3d, so 4s must be higher in energy than 3d, so ionization should remove 3d first" is a trap. Filling order for building up neutral atoms and removal order upon ionizing an already-formed atom are governed by different total-energy comparisons, precisely because populating 3d changes the screening felt by 4s. Iron loses its two 4s electrons before any 3d electron — this is not a contradiction of the aufbau order but a demonstration that $Z_{\text{eff}}$ , and hence orbital energy ordering, is configuration-dependent, not a fixed, universal ladder.

The chemist behind it

The modern quantitative language of shielding and effective nuclear charge grew out of the marriage between spectroscopy and the new quantum mechanics in the 1920s, but its most influential codification for chemists came from John C. Slater, an American physicist who built the empirical rule set that lets $\sigma$ be estimated by hand from electron configuration alone — the very rules this article's next natural companion piece takes up in full.

Origins · 1 minJohn Clarke Slater and how this was discovered

John Clarke Slater

1900–1976 · American theoretical physicist, Harvard and MIT, central figure in early quantum theory of solids and atoms

Slater trained at Harvard and studied briefly in Europe during the ferment of the mid-1920s, absorbing the new matrix and wave mechanics directly from its inventors. He returned to build a distinctly American school of quantum theory applied to real materials and atoms rather than abstract formalism alone. His 1930 paper on atomic shielding constants distilled the complicated physics of orbital penetration — the same physics worked through in this article's derivations — into a short empirical rule chemists could apply with pencil and paper, without solving any Schrödinger equation. He also pioneered the Slater determinant, the antisymmetrized many-electron wavefunction that underlies essentially all of modern quantum chemistry, and later became a major architect of solid-state and band-structure theory at MIT.

History

1913

— Bohr's model introduces quantized one-electron orbits, giving the first

-13.6\,\text{eV}/n^2

energy law, with no shielding concept yet needed.

1926

— Schrödinger's wave equation and its exact solution for hydrogen make the

n,l,m_l

orbital picture rigorous, and immediately raise the question of what happens with more than one electron.

1928–1930

— Hartree's self-consistent field method computes numerically the effective radial potential each electron feels in real many-electron atoms, giving the first quantitative

Z_{\text{eff}}(r)

curves.

1930

— Slater publishes his empirical shielding rules, turning Hartree-style physics into a rule-based tool usable across the whole periodic table.

1963

— Clementi and Raimondi refine effective nuclear charges using accurate self-consistent-field wavefunctions, producing the precise

Z_{\text{eff}}

tables still quoted in modern textbooks.

What exactly is the difference between "shielding" and "penetration"?

Penetration is a property of a single orbital: how much of its radial probability density lies inside the region occupied by other (core) electrons, close to the bare nucleus. Shielding is the resulting reduction in nuclear charge felt by that electron because of everyone else's charge. They are cause and effect: an orbital that penetrates strongly (like $ns$ ) is shielded weakly, because it spends real time inside the screening cloud rather than always outside it. An orbital that penetrates weakly (like $nd$ or $nf$ ) is shielded strongly, because it almost always finds the full core charge already canceling much of the nucleus before it gets there. You can think of penetration as the geometric cause and shielding as the energetic consequence.

Does Z_eff have one single correct numerical value for a given orbital?

No, and this is a common source of confusion. $Z_{\text{eff}}$ is a modeling construct, not a directly observable quantum number, so its numerical value depends on the definition used to extract it. Slater's rules give one set of values from simple counting rules; Clementi–Raimondi values come from fitting accurate self-consistent-field (Hartree–Fock) wavefunctions; spectroscopic values come from inverting measured ionization energies through the hydrogenic formula. These generally agree in trend (same orderings, same qualitative behavior) but can differ by several tenths in absolute magnitude, because each method averages over the many-body physics slightly differently.

Why does the centrifugal term in the radial equation matter so much for shielding?

The centrifugal term $\hbar^2 l(l+1)/2m_e r^2$ in the radial Schrödinger equation acts like a repulsive potential barrier that grows without bound as $r\to 0$ , and grows more steeply for larger $l$ . This forces the radial wavefunction, and hence the probability density, toward zero at small $r$ for high- $l$ orbitals — physically, angular momentum "throws" the electron outward, away from the nucleus. Since penetration is exactly the ability of an orbital to have appreciable density at small $r$ (inside the core), the centrifugal barrier is the mathematical origin of why high- $l$ orbitals penetrate poorly and are therefore shielded more effectively than low- $l$ orbitals of the same $n$ .

Is Z_eff the same thing as the oxidation state or formal charge on an atom?

No, these are unrelated concepts that happen to share the word "charge." Oxidation state and formal charge are bookkeeping devices for electron counting in bonds, assigned by rules about electronegativity or arbitrary equal-sharing conventions; they are integers (or occasionally simple fractions) and say nothing about orbital energetics. $Z_{\text{eff}}$ is a continuous, orbital-specific quantity describing the net attractive pull an electron experiences from the nucleus after accounting for shielding by other electrons in the same atom or ion. An atom can have an oxidation state of $+2$ while its various orbitals have several quite different $Z_{\text{eff}}$ values simultaneously.

Why does Z_eff increase across a period but the increase is much smaller than the increase in Z itself?

Moving across a period adds one proton to the nucleus and one electron to the same outer shell. If that added electron shielded the added proton perfectly, $Z_{\text{eff}}$ would stay exactly flat across the whole period. In reality, electrons within the same shell are relatively poor shields for each other — they occupy similar regions of space and don't sit neatly "between" each other and the nucleus the way an inner-shell electron does — so each added proton is only partially canceled. The net result is that $Z_{\text{eff}}$ creeps upward across a period, but much more slowly than $Z$ itself, which is exactly why atomic radii shrink and ionization energies rise gradually, rather than dramatically, from left to right.

How does Z_eff explain the anomalously low reactivity gap between elements like zirconium and hafnium?

Hafnium sits directly below the lanthanide series, meaning fourteen $4f$ electrons have been added to the core before hafnium's own valence electrons are considered. The $4f$ orbitals penetrate very poorly (high $l=3$ ), so they shield the outer $6s$ / $5d$ electrons much less effectively than one might naively expect from simply adding fourteen electrons. The net effect nearly cancels the size increase normally expected from adding a whole new principal shell, leaving hafnium's $Z_{\text{eff}}$ , and hence its atomic and ionic radius, almost identical to zirconium's, one row above. This is the lanthanide contraction, and it is a direct, load-bearing consequence of $f$ -orbital shielding being unusually weak.

Can Z_eff ever exceed the true nuclear charge Z?

No, not in the physically meaningful sense used here, because $\sigma \geq 0$ always — other electrons can only reduce, never reverse-augment, the net attraction felt by a given electron (electron–electron interaction is always repulsive, which always partially cancels, never adds to, nuclear attraction). The only case where $Z_{\text{eff}} = Z$ exactly is the one-electron limit, where there is no other electron present to provide any shielding at all, so $\sigma = 0$ and $Z_{\text{eff}}=Z$ exactly, the natural upper bound.