Slater's Rules for Shielding Constants

Slater's rules remain a staple of introductory and intermediate inorganic chemistry precisely because they let students compute $Z_\text{eff}$ rather than merely gesture at periodic trends. They rationalize, quantitatively, why atomic radii contract sharply across a period (increasing $Z_\text{eff}$ for the outermost shell as protons are added but same-shell screening stays weak), why the post-transition-metal and lanthanide contractions leave elements like hafnium and zirconium nearly indistinguishable in size and reactivity (poor $d$/$f$ shielding lets $Z_\text{eff}$ creep up steadily across the intervening block), and why successive ionization energies jump sharply once a full shell is breached (removing an electron from a lower, poorly shielded shell costs vastly more energy). In modern research and teaching, Slater-derived $Z_\text{eff}$ values are used as a quick diagnostic in rationalizing X-ray photoelectron spectroscopy (XPS) binding-energy trends across a period or down a group, in explaining crystal-field and ligand-field splitting magnitudes qualitatively (higher $Z_\text{eff}$ on the metal contracts $d$-orbitals and changes covalency), and in undergraduate-level rationalization of why certain oxidation states are stabilized (for example, the reluctance of late transition metals to lose $d$ electrons versus $s$ electrons, tracing back to the $Z_\text{eff}(nd) \gg Z_\text{eff}(ns)$ pattern demonstrated above for manganese). The underlying idea — that shielding is orbital-specific and governed by radial penetration — is the conceptual ancestor of the more mathematically refined Slater-Type Orbitals used as basis functions in quantum chemistry software, and of the systematic, quantitative treatment of Periodic Trends in Atomic Radii built on accurately computed effective nuclear charges from Hartree–Fock and density-functional methods. Even where the numerical accuracy of the original 1930 constants has been superseded, the grouping logic — same-shell weak shielding, inner-shell strong shielding, outer-shell zero shielding, with $d,f$ singled out for their poor penetration — remains the mental model every chemist reaches for first when asked "why does this trend go this way?"

Ask the tutor: How would Slater's rules change if applied to an excited-state configuration instead of the ground state — does the grouping order itself ever need to be reassigned?

Start from the one-electron picture that already works perfectly for hydrogen: an electron of charge $-e$ orbiting a nucleus of charge $+Ze$ feels the full nuclear pull, and its orbital energy is $E_n = -13.6\,\text{eV}\,(Z^2/n^2)$. Add a second electron, and each electron now feels two forces: the nuclear attraction, and the electron–electron repulsion from every other electron in the atom. Because electron–electron repulsion is (on average) directed outward from the nucleus, it partially cancels the nuclear pull. The net effect, averaged over the electron's motion, is as though the nucleus had a smaller effective charge:

The physically hard part is $S$, the shielding (or screening) constant, which depends on exactly how much time the shielding electrons spend between the nucleus and the electron in question. This is a statement about radial penetration. An electron in a $3s$ orbital has appreciable amplitude very close to the nucleus (a small "penetrating" hump inside the $1s$ and $2s$ shells) even though its most probable radius is large; an electron in a $3d$ orbital, by contrast, is kept away from the nucleus by a strong centrifugal barrier ($l=2$) and barely penetrates the core at all. This is the origin of the orbital energy ordering $E(ns) < E(np) < E(nd) < E(nf)$ within a given shell for many-electron atoms, in sharp contrast to hydrogen where all subshells of a given $n$ are degenerate. Slater's insight was to replace the true, continuous, orbital-dependent penetration behavior with a discrete grouping scheme. Electrons are sorted into shells ordered by increasing energy, roughly:

Note that $s$ and $p$ of the same $n$ are lumped together (they penetrate comparably), but $d$ and $f$ are split into their own separate groups, one shell "later" than the $s,p$ pair of the same $n$ — reflecting how poorly $d$ and $f$ orbitals penetrate compared to $s,p$ of the same principal quantum number. Within this grouping, three screening rules apply, calibrated once and used universally:

• An electron in the same group as the electron of interest contributes $0.35$ to $S$ (except that a $1s$ electron shielding another $1s$ electron contributes only $0.30$).
• An electron in a group one shell principal-quantum-number lower contributes $0.85$ to $S$ if the electron of interest is in an $s$ or $p$ orbital; if the electron of interest is in a $d$ or $f$ orbital, an electron in any lower group (not just one shell down) contributes the full $1.00$.
• An electron in a group two or more shells lower (for $s,p$ electrons) contributes the full $1.00$ — it behaves as if it were part of the nuclear charge itself, because it is almost entirely inside the electron of interest's orbital.
• Electrons in higher groups (further out) contribute zero — an outer electron cannot shield an inner one, because on average it is farther from the nucleus and does not sit between the nucleus and the inner electron.

The asymmetry between the $s,p$ rule and the $d,f$ rule for "one shell down" is the single most chemically important feature of the whole scheme: it is what makes $3d$ electrons shield the $4s$ electron much less effectively than a naive "one shell down → 0.85" rule would suggest, which is central to understanding the anomalous filling and ionization order of the transition metals.

The physical justification for treating outer electrons as contributing zero shielding follows from spherical charge-shell electrostatics: a spherically symmetric shell of charge exerts no net electric field on a point charge located inside it (this is the same theorem — Gauss's law applied to spherical shells — that says the Earth's crust exerts no net gravitational pull on someone standing at the center). Outer electrons, on average, orbit outside the radius of the inner electron in question, so by this shell theorem they cannot screen it at all. Electrons deep inside, conversely, look almost like an extra bit of nuclear charge smeared close to the center, hence their near-unit contribution.

The master relation, applicable to every electron in every atom, is:

where:

• $Z$ is the atomic number (number of protons),
• $S$ is the total shielding constant experienced by the electron of interest, summed over contributions from all other electrons in the atom according to the grouping rules,
• $Z_\text{eff}$ is the effective nuclear charge felt by that specific electron in that specific orbital (not a single number per atom — every subshell has its own $Z_\text{eff}$).

Reading: $Z_\text{eff}$ is orbital-specific. The same atom gives a different $Z_\text{eff}$ for its $3s$ electrons than for its $3p$ or $3d$ electrons, because the grouping rules and screening amounts differ subtly (same-group $s,p$ electrons shield each other identically, but a $3d$ electron sees a different environment than a $3s$ electron of the same atom, because the $(n{-}1)$ rule differs for $d$ versus $s,p$). The screening sum itself, written out explicitly for an electron in a shell with principal quantum number $n$ that is of type $s$ or $p$:

where:

• $N_\text{same}$ is the number of other electrons in the same $(n\ell)$ group (using the $s,p$-combined grouping),
• $N_{n-1}$ is the number of electrons in the shell with principal quantum number $n-1$,
• $N_{\le n-2}$ is the number of electrons in all shells with principal quantum number $n-2$ or lower,
• the special case of a $1s$ electron shielding another $1s$ electron uses $0.30$ instead of $0.35$.

For an electron in a $d$ or $f$ orbital, the rule changes to:

where $N_\text{inner}$ is the number of electrons in all groups lower in the Slater ordering than the $d$ or $f$ group in question (there is no intermediate $0.85$ tier for $d,f$ electrons — everything inward counts fully). Finally, the effective principal quantum number $n^*$ is sometimes introduced for use in a Slater-type radial wavefunction or in Slater's approximate orbital-energy formula:

where $n^*$ equals $n$ for $n = 1,2,3$, but is scaled down for higher shells ($n^*=3.7$ for $n=4$, $4.0$ for $n=5$, $4.2$ for $n=6$) to correct for the fact that Slater's simple orbitals overestimate orbital size at high $n$. This correction anticipates the more systematic radial functions used in Slater-Type Orbitals.

Example 1 — Foundational: $Z_\text{eff}$ for a valence electron in phosphorus (Z = 15). Phosphorus has configuration $1s^2\,2s^2 2p^6\,3s^2 3p^3$. Take one of the $3p$ electrons as the electron of interest. Group by Slater's convention: $(1s^2)(2s^2 2p^6)(3s^2 3p^3)$. Same-group ($3s,3p$) electrons other than the one of interest: there are $2+3-1 = 4$ others, each contributing $0.35$:

One-shell-down group ($2s,2p$, total 8 electrons), each contributing $0.85$ since the target is $p$ (an $s,p$-type electron):

Two-shells-down group ($1s$, 2 electrons), each contributing $1.00$:

Total shielding:

Compare this to the crude "core-only" guess a beginner might make — that all 10 core electrons ($1s^2 2s^2 2p^6$) shield completely and the 4 remaining valence electrons shield not at all — which would give $Z_\text{eff} = 15 - 10 = 5.0$. Slater's more careful accounting, which lets the valence electrons contribute a modest $1.40$ among themselves, brings the number down to $4.80$, roughly 4% lower and closer to spectroscopically fitted values. The take-home point: even "valence shields valence" matters, just weakly.

Limiting-case check embedded in this example: if phosphorus had zero $3p$ electrons besides the reference one (i.e., consider instead a hypothetical $3s^2 3p^0$ same-group count with only the $3s$ pair), the same-group contribution would shrink to just the other $3s$ electron, $0.35\times 1 = 0.35$, showing the formula scales sensibly and continuously with occupation — it is a simple linear count, not a threshold effect.

Example 2 — Applied: comparing $Z_\text{eff}(4s)$ and $Z_\text{eff}(3d)$ in the manganese atom (Z = 25), and why 4s ionizes before 3d. Manganese: $1s^2\,2s^2 2p^6\,3s^2 3p^6\,3d^5\,4s^2$. Grouping: $(1s^2)(2s^22p^6)(3s^23p^6)(3d^5)(4s^2)$. First, the $4s$ electron. Same group ($4s$, one other electron): $0.35 \times 1 = 0.35$. One-shell-down for an $s,p$ target is the $n=3$ shell combined, i.e. $3s,3p,3d$ all count as "$n-1$" region — but Slater's rule specifically says: for an $s,p$ target, the $(n-1)$ shell contributes $0.85$ per electron, and this includes the $3d^5$ electrons too, since they share $n=3$. That gives $3s^2+3p^6+3d^5 = 14$ electrons at $0.85$ each:

Two-or-more-shells-down ($n=1,2$: total $2+8=10$ electrons) at $1.00$ each:

Now the $3d$ electron. Same group ($3d^5$, four other electrons): $0.35 \times 4 = 1.40$. For a $d$-target, everything inward (the $1s,2s2p,3s3p$ groups — note $4s$ is outer and contributes zero) counts fully at $1.00$: that's $2+8+8 = 18$ electrons:

Here is the chemically loaded result: $Z_\text{eff}(3d) = 5.60$ is markedly larger than $Z_\text{eff}(4s) = 2.75$. A $3d$ electron in neutral Mn feels roughly twice the effective nuclear charge that a $4s$ electron feels. This seems to contradict the familiar aufbau statement "$4s$ fills before $3d$" — but it does not, once orbital energy is properly distinguished from $Z_\text{eff}$ alone. Because $n^*(4s)=3$ while $n^*(3d)=3$ as well in the simple scheme (both use $n^*=n$ for $n\le3$... more carefully, Slater's $n^*$ correction only kicks in from $n=4$ onward for the principal quantum number of the orbital itself, so $3d$ uses $n^*=3$ and $4s$ uses $n^*=4$), the energy comparison uses:

This crude ratio overshoots the true near-degeneracy of $4s$ and $3d$ in neutral transition metals (Slater's simple orbital-energy formula is known to be unreliable exactly in this regime — a caveat discussed below), but the qualitative direction it captures is the experimentally central fact: once an electron is removed to form $\ce{Mn+}$, the 4s orbital, being large and diffuse with a much smaller $Z_\text{eff}$, is easier to empty first, which is exactly why transition-metal cations are $[Ar]3d^n$ rather than $[Ar]3d^{n-2}4s^2$ — the textbook "4s fills first, but 3d empties first" rule falls directly out of the relative sizes and $Z_\text{eff}$ values, not from a naive reading of orbital-filling order alone.

Example 3 — Exam/competition level: shielding constant and ionization energy of the outermost electron in $\ce{Cu+}$, and comparison with experiment. Copper(I), $\ce{Cu+}$, has configuration $[Ar]3d^{10}$ (25 + 1 = wait, Cu has Z = 29; $\ce{Cu+}$ has 28 electrons): $1s^2 2s^2 2p^6\,3s^2 3p^6\,3d^{10}$, with no $4s$ electrons — the $3d$ subshell is now full and outermost. Take a $3d$ electron. Same group ($3d^{10}$, nine other electrons): $0.35\times 9 = 3.15$. Inner groups, all fully counted at $1.00$ ($1s^2, 2s^22p^6, 3s^23p^6$, total $2+8+8=18$):

Using $n^*=3$ (since $n=3\le3$ needs no correction):

This is the Slater estimate for the energy needed to remove a $3d$ electron from $\ce{Cu+}$ (i.e., roughly related to the second ionization energy region of copper, since removing the first electron from neutral Cu — $[Ar]3d^{10}4s^1$ — actually comes from the $4s$ orbital, not $3d$). The experimental second ionization energy of copper (removing an electron from $\ce{Cu+}$, $3d^{10}\to 3d^9$) is about $20.29\ \text{eV}$ — dramatically smaller than the $93.2\ \text{eV}$ Slater estimate. This large discrepancy is instructive rather than embarrassing: it is a well-known limitation of Slater's rules specifically for compact, highly-charged, filled $d^{10}$ configurations, where electron correlation and relaxation effects (the remaining electrons contract and reorganize substantially once one is removed, badly violating the frozen-orbital assumption of Step 2 in the second derivation above) are large, and where the crude $0.35$/$1.00$ weighting systematically underestimates shielding for a filled $d^{10}$ shell of this compactness. The lesson for exam purposes: Slater's rules give the right qualitative orbital ordering and trend direction almost always, but for $d^{10}$ and $f^{14}$ closed subshells in cations, the quantitative ionization-energy prediction can be off by a large factor, and one should quote the $Z_\text{eff}$ number with the caveat that it is a comparative, not absolute, tool in this regime.