Penetration and Shielding

Penetration and shielding are not a historical curiosity; they underpin working explanations across chemistry. In spectroscopy, X-ray photoelectron spectroscopy (XPS) resolves core-level binding energies that shift systematically with oxidation state precisely because changing the number of valence electrons changes the shielding felt by core electrons — the "chemical shift" in XPS is a direct, quantitative signature of penetration/shielding physics, used routinely to determine oxidation states of metals in catalysts and battery electrode materials. In coordination chemistry, the fact that $3d$ orbitals barely penetrate the argon-like core (as shown numerically above) is why they are strongly perturbed by ligand fields yet remain compact and localized — the entire crystal-field and ligand-field treatment of transition-metal color and magnetism assumes exactly this contracted, non-penetrating character of $d$ orbitals. Lanthanide contraction — the anomalously small ionic radii of the third-row transition metals following the lanthanides — is a direct penetration/shielding effect: poorly-shielding $4f$ electrons let the nucleus pull inward on the $5d$/$6s$ shells more than expected, with consequences for gold's chemistry, the density of tungsten, and the near-identical radii of zirconium and hafnium that famously complicates their separation in nuclear reactor materials. In drug design and bioinorganic chemistry, the effective charge felt by metal-bound electrons (understood via penetration arguments) governs Lewis acidity trends used to predict how zinc, iron, and copper centers in metalloenzymes activate substrates. Finally, every periodic-table trend taught at the introductory level — ionization energy, atomic radius, electron affinity, electronegativity — is, underneath the memorized arrows, a story about how penetration-dependent shielding reshapes $Z_{\text{eff}}$ across rows and down columns, the systematic version of which is developed under Effective Nuclear Charge.

Ask the tutor: Why does the 3d orbital drop below 4s in energy for transition-metal cations even though Slater's rules treat the core screening of 3d as essentially unchanged from the neutral atom?

Start from the one-electron atom, where the Schrödinger equation gives energies depending only on $n$:

For hydrogen, $2s$ and $2p$ have identical energy. Add a second electron — helium, or better, think of sodium's valence electron moving in the field of a $\mathrm{Na}^{+}$ core — and the potential the outer electron feels is no longer a clean $-Ze^2/r$. Close to the nucleus, where inner-shell electron density is sparse, the outer electron sees nearly the full nuclear charge $Z$. Far from the nucleus, the inner electrons' charge cloud sits between the outer electron and the nucleus, and by Gauss's law that inner charge cancels part of the nuclear attraction, so the outer electron sees an effective charge close to $Z - (\text{number of inner electrons})$. The crucial question is: how much time does an orbital spend in the "sees full $Z$" region versus the "screened" region? That is exactly what the radial distribution function $4\pi r^2 R_{n l}^2(r)$ answers. Recall that $ns$ orbitals have $n-1$ radial nodes, $np$ have $n-2$, $nd$ have $n-3$, and so on — for fixed $n$, lower $l$ means more radial nodes and (crucially) a larger-amplitude inner lobe close to $r=0$. An electron in $3s$ has real, non-negligible probability density very near the nucleus (inside the $1s$ and $2s,2p$ shells), while a $3p$ electron has a smaller inner lobe, and a $3d$ electron — with no radial nodes at all for $n=3$, just the pure $r^{l}$ falloff at small $r$ — has essentially none. This is penetration made quantitative: it is literally area under the radial distribution curve at small $r$, inside the core.

The figure makes the essential point visually: all three orbitals ($3s$, $3p$, $3d$) have their main, outermost lobe at roughly similar radius (they belong to the same shell), but only $3s$ and $3p$ have extra inner lobes that dip deep into the core region shaded on the left. The $3s$ curve has two small inner humps (from its two inner radial nodes) plus the outer lobe; $3p$ has one inner hump; $3d$ has none. That inner probability is what "penetration" means physically, and it directly controls how much nuclear charge is screened away. Now connect this to energy. Define, for an electron in orbital $\psi_{nl}$, the effective nuclear charge

where $\sigma_{nl}$ is the shielding constant, the amount of nuclear charge cancelled by the other electrons as experienced by this particular orbital. Because $\sigma_{nl}$ depends on how much the electron penetrates the core, $\sigma_{3s} < \sigma_{3p} < \sigma_{3d}$, so $Z_{\text{eff}}$ decreases in that same order, and by the hydrogen-like energy formula (now with $Z\to Z_{\text{eff}}$),

a larger $Z_{\text{eff}}$ means a more negative (more stable) energy. Hence within the same $n$: $E_{ns} < E_{np} < E_{nd} < E_{nf}$. This is the $l$-splitting the hydrogen atom does not show, and it is entirely a penetration/shielding effect.

Note this is purely a shielding/penetration effect within one shell; it says nothing yet about cross-shell competition such as $4s$ vs. $3d$, which involves comparing different $n$ and is treated with real numbers in the worked examples below and formally in Orbital Energy Ordering in Multielectron Atoms.

The governing relation is the modified hydrogenic energy formula:

where:

• $Z_{\text{eff}}(n,l) = Z - \sigma_{nl}$ is the effective nuclear charge for that specific orbital,
• $n^{*}$ is the effective principal quantum number (equal to $n$ for $n \le 3$; reduced for $n = 4,5,6$ in Slater's scheme),
• $\sigma_{nl}$ is built from other electrons' contributions, weighted by how much they shield this orbital.

Reading: the formula says an electron's binding energy scales with the square of the charge it actually experiences, not the bare nuclear charge — this is why even a modest change in shielding (a change in $\sigma$ of order 1) produces a large, chemically significant change in energy, since the effect is squared. The shielding constant itself, in Slater's rules, is built additively:

where:

• the sum runs over all other electrons in the atom,
• $k_i$ is the individual screening contribution of electron $i$, which depends on whether $i$ is in the same group, the shell just inside, or further inside, and on whether the orbital being screened is $s/p$-type or $d/f$-type.

Reading: this equation is simply "shielding is additive over the other electrons," but the values of $k_i$ encode all the penetration physics — an inner $d$ electron screens an outer $s$ electron much more effectively (nearly full $k=1$) than an outer $d$ electron screens another $d$ electron in the same subshell (partial $k \approx 0.35$), because $s$ orbitals penetrate deep inside $d$ shells while same-subshell $d$–$d$ interactions are mutually poorly screening. Finally, effective nuclear charge connects to observable ionization energy through

where:

• $IE_1$ is the first ionization energy (removing the outermost, highest-$n^*$ electron),
• the same $Z_{\text{eff}}, n^{*}$ as defined above apply to the electron being removed.

Reading: this is the empirical bridge between the abstract shielding constant and something measured in a mass spectrometer or photoelectron spectrometer — every trend in ionization energy across a period or down a group is, at bottom, a story about how $Z_{\text{eff}}$ for the outermost electron changes.

Example 1 — Foundational: Z_eff for the sodium valence electron. Sodium, $Z=11$, configuration $1s^2\,2s^2 2p^6\,3s^1$. We want $Z_{\text{eff}}$ felt by the lone $3s$ electron. Group structure: $(1s^2)(2s^2 2p^6)(3s^1)$. Screeners of the $3s$ electron:

• Same group ($3s,3p$): no other electrons here (the group has only this one electron), contributes $0$.
• $n=2$ shell ($2s^2 2p^6$, 8 electrons), one shell in: each contributes $0.85$:

• $n=1$ shell ($1s^2$, 2 electrons), two shells in: each contributes $1.00$:

Total shielding:

Compare with the naive (wrong) picture that ten core electrons cancel ten units of charge, giving $Z_{\text{eff}}=1$: Slater's rules show the real number is over twice that, because the $1s$ electrons screen almost perfectly but the $n=2$ shell screens only 85% as effectively — some $3s$ density penetrates inside the $n=2$ shell. Using $n^*=3$ (Slater's rule: $n^*=n$ for $n\le 3$):

Converting, $7.32\ \text{eV} \times 96.485\ \text{kJ mol}^{-1}\text{eV}^{-1} = 706\ \text{kJ/mol}$. The experimental value is 496 kJ/mol. Slater's rules overestimate here (a known limitation for alkali metals, where the simple $k=0.85$ tier undershoots how much the $n=2$ shell actually screens an outer $ns$ electron), but the qualitative conclusion — $Z_{\text{eff}}$ is small (around 2, not 11 or 1) — is exactly the physically correct statement, and the order-of-magnitude drop in ionization energy compared to hydrogen's 1312 kJ/mol is captured. This is the essence of the article's opening claim: sodium ionizes easily not because its nucleus is weak, but because ten inner electrons collectively (and imperfectly) cancel nine-elevenths of the nuclear charge as seen by the valence electron.

Example 2 — Applied: comparing 3s, 3p ionization within chlorine, and rationalizing electron affinity trends. Chlorine, $Z=17$, configuration $1s^2\,2s^2 2p^6\,3s^2 3p^5$. Compute $Z_{\text{eff}}$ for a $3p$ electron. Group structure: $(1s^2)(2s^2 2p^6)(3s^2 3p^5)$. The electron of interest is one of the five $3p$ electrons. Screeners:

• Same group ($3s,3p$): total electrons in group $=2+5=7$; screening electron itself excluded, so $6$ others, each contributing $0.35$:

• $n=2$ shell (8 electrons), each contributing $0.85$:

• $n=1$ shell (2 electrons), each contributing $1.00$:

Now do the same for a $3s$ electron in chlorine (same configuration, screening a $3s$ instead of $3p$ — Slater's rules treat $3s,3p$ identically as screeners and as the screened electron, so the same-group count is identical: $6$ other electrons in the combined $3s,3p$ group):

Slater's rules predict $Z_{\text{eff}}(3s) = Z_{\text{eff}}(3p) = 6.10$ for chlorine — this is the well-known blind spot of Slater's original scheme: by lumping $ns$ and $np$ into one group, it cannot resolve the $s$–$p$ splitting that penetration actually produces. Physically, the $3s$ electron does penetrate more than $3p$ (as shown in the first derivation) and so should have a somewhat larger true $Z_{\text{eff}}$ — more refined parameter sets (Clementi–Raimondi, from self-consistent-field calculations) give $Z_{\text{eff}}(3s)\approx 6.12$ vs. $Z_{\text{eff}}(3p)\approx 6.76$ for chlorine, restoring the physically correct order $Z_{\text{eff}}(3p) < Z_{\text{eff}}(3s)$. Chemically, this matters for chlorine's very negative electron affinity ($-349$ kJ/mol, the most exothermic of any element): the incoming electron enters a $3p$ orbital that is reasonably well shielded from other $3p$ electrons individually (each same-subshell partner only screens at $k=0.35$) yet the whole atom still exerts a respectably large $Z_{\text{eff}} \approx 6.1$ on it, giving strong net attraction with only modest electron–electron repulsion penalty — exactly the balance that makes halogens such effective electron acceptors compared with, say, oxygen or noble gases (full shell, extra electron forced into next $n$, tiny $Z_{\text{eff}}$).

Example 3 — Exam/competition level: why 4s empties before 3d fills — Ca vs. Ca²⁺ vs. Sc, using explicit numbers. This is the classic "which orbital is lower, $4s$ or $3d$?" puzzle, and penetration/shielding resolves the apparent contradiction between neutral-atom filling order and ion configurations.

Neutral calcium, $Z=20$: $1s^2\,2s^2 2p^6\,3s^2 3p^6\,4s^2$ (not $3d^2$). Compute $Z_{\text{eff}}$ for the two competing orbitals as if populated, in the same underlying core $1s^2 2s^2 2p^6 3s^2 3p^6$ (18 electrons): For a hypothetical $3d$ electron added to this core (Slater group rules: $n^*=3$ for $3d$, and using the $d/f$ tier where every electron in a lower group screens fully):

For a $4s$ electron added to this same core:

• Same group ($4s,4p$): only this one electron, contributes $0$.
• $n=3$ shell ($3s^23p^6$, 8 electrons), one shell in, $k=0.85$:

• $n\le2$ shells ($1s^22s^22p^6$, 10 electrons), two-or-more shells in, $k=1.00$:

Slater's rule uses $n^{*}=3.7$ for $n=4$ (the empirical Slater correction for $n\ge4$: $n^*=3.7$ at $n=4$, $4.0$ at $n=5$, $4.2$ at $n=6$), so:

Comparing the two proportional energies (more negative = lower energy = more stable, so we want the larger magnitude of $(Z_{\text{eff}}/n^*)^2$ since $E=-(Z_{\text{eff}}/n^*)^2\times$Ry):

The $4s$ orbital is predicted (correctly) to be lower in energy than $3d$ for building up neutral calcium — even though $4s$ has a smaller $Z_{\text{eff}}$ (3.20 vs. 2.00), its much smaller effective quantum number penalty ($n^*=3.7$, close to $3d$'s $n^*=3$) is not enough to overcome $4s$'s superior penetration advantage relative to how poorly-shielded it manages to stay. The $n^2$ denominator working against $4s$ is outweighed by $3d$'s far smaller $Z_{\text{eff}}$ (since every one of the 18 core electrons fully shields $3d$, versus only the outer eight doing so at 85% for $4s$) — this is penetration in action: $4s$ has small inner lobes that let it "peek" past part of the core, while $3d$ has no such lobes and is shielded almost completely.

Now ionize: for $\mathrm{Ca}^{2+}$ (remove the two $4s$ electrons, giving argon-core configuration $1s^22s^22p^63s^23p^6$), and ask which orbital an added electron would prefer — this models $\mathrm{Sc}^{2+}\to\mathrm{Sc}^+$ or explains why, once $3d$ starts filling in the transition series, $3d$ becomes lower than $4s$ (transition-metal cations have configurations like $[\mathrm{Ar}]3d^n$, never $[\mathrm{Ar}]3d^{n-2}4s^2$-for-the-ion). With the $4s$ electrons already gone, the $3d$ orbital's screening environment hasn't changed (still 18 electrons all fully screening it in this Slater approximation, since $3d$'s formula doesn't distinguish "core" further), but critically, once even one electron occupies $3d$ in the neutral transition metals, additional $3d$ electrons screen each other only at $k=0.35$ (poor, same-subshell screening) while continuing to be almost perfectly shielded from outside by the core — so as $Z$ increases across the first transition series, $Z_{\text{eff}}(3d)$ rises faster than $Z_{\text{eff}}(4s)$, and by scandium the $3d$ orbital has already dropped below $4s$ in orbital energy for the ion, even though the neutral atom is still built by filling $4s$ first due to interelectronic repulsion subtleties beyond simple Slater accounting. This is precisely the resolved paradox behind the famous "$4s$ fills first, empties first" rule, and it is treated in full in Orbital Energy Ordering in Multielectron Atoms.