Sulfur Oxoacids and Oxides

Sulfuric acid manufacture via the contact process (catalytic oxidation of SO₂ to SO₃ over V₂O₅, followed by absorption of SO₃ into 98% H₂SO₄ to give oleum, then dilution) remains the single largest-tonnage industrial chemical process on the planet, feeding phosphate fertilizer production (superphosphate and triple superphosphate manufacture consume the majority of world sulfuric acid output), petroleum alkylation units, and lead-acid battery electrolyte production. The Lewis-acid reactivity of SO₃ described here — its hunger for any nucleophile — is exploited directly in industrial sulfonation chemistry, where SO₃ or oleum reacts with aromatic rings to install –SO₃H groups, the key step in manufacturing synthetic detergents (linear alkylbenzene sulfonates) and many sulfa-drug intermediates. Sulfurous acid/sulfite chemistry underlies wine and dried-fruit preservation (SO₂ and bisulfite as antimicrobial, antioxidant additives, exploiting sulfite's mild reducing power to scavenge dissolved O₂ and inhibit oxidative browning enzymes) and the sulfite (Kraft-adjacent) and sulfite pulping processes in papermaking, where bisulfite/sulfite solutions delignify wood pulp. Atmospheric SO₂, largely from fossil-fuel combustion and metal smelting, oxidizes further to sulfate aerosol and falls as the sulfuric-acid component of acid rain — a direct environmental expression of the same SO₂ → SO₃ → H₂SO₄ oxidation pathway run industrially in the contact process, which is why flue-gas desulfurization (scrubbing SO₂ with lime or limestone slurries to precipitate calcium sulfite/sulfate) is mandated at large combustion sources worldwide. Thiosulfate chemistry is the basis of iodometric titration, the standard analytical method for quantifying oxidizing agents (dissolved oxygen via the Winkler method, active chlorine, peroxide content) by first generating a known amount of I₂ and then back-titrating with standardized Na₂S₂O₃ using starch as the endpoint indicator; the sharp, reproducible one-electron-pair chemistry at the terminal sulfur atom (the mechanism worked through above) is what makes this titration so precise. Thiosulfate's soft-donor affinity for Ag⁺, again traceable to the electron-rich terminal sulfur, is the chemical basis of traditional photographic film fixing (dissolving unexposed AgBr as soluble silver-thiosulfate complexes) and is now being explored as an environmentally gentler alternative to cyanide in gold and silver leaching from ores, precisely because thiosulfate forms strong, soluble complexes with soft metal cations without cyanide's acute toxicity. Related S–S-bonded sulfur oxoanions — dithionite (S₂O₄²⁻), a powerful, oxygen-sensitive reducing agent used in reductive vat dyeing and in some water-treatment and pulp-bleaching applications, and polythionates (S_nO₆²⁻ chains), relevant to microbial sulfur metabolism in geochemical and biogeochemical sulfur cycling — extend the same structural logic (variable sulfur oxidation states linked through S–S bonds) into industrial reduction chemistry and environmental microbiology.

Ask the tutor: Why does the second ionization constant of sulfuric acid (HSO₄⁻ ⇌ H⁺ + SO₄²⁻) fall so much closer to Pauling's predicted value than the first ionization does?

Start with the free molecules, because the acids are built from them. Sulfur dioxide, SO₂, has sulfur bonded to two oxygens with one lone pair on sulfur: three electron domains, so VSEPR predicts a bent, trigonal-planar-minus-one-vertex shape, and experiment confirms O–S–O = 119.5°, close to the ideal 120° of an AX₂E system, with the lone pair exerting only a modest extra compression. Each S–O bond length is 143.1 pm, shorter than a typical S–O single bond (~157 pm), signaling substantial π character. The bonding is best described not as two isolated S=O double bonds but as a σ framework (S sp² hybrids overlapping O sp² hybrids, plus the in-plane lone pair) topped by a 3-center-4-electron π system: one p orbital on S perpendicular to the molecular plane overlaps with one p orbital on each O, generating three π molecular orbitals (bonding, nonbonding, antibonding) that hold 4 electrons — bonding and nonbonding filled, antibonding empty. This delocalized picture, rather than a fixed pair of resonance structures S(=O)(=O) ↔ resonance hybrids with formal charges, is why both S–O bonds are identical in length and the net π bond order per S–O linkage is 0.5, on top of the σ bond — total bond order close to 1.5–2, consistent with the short, strong bond observed.

Sulfur trioxide, SO₃, in the gas phase is a single trigonal-planar molecule, D₃ₕ symmetry, S–O = 142 pm, O–S–O = 120° exactly by symmetry. Now there are three σ bonds and no lone pair on sulfur (AX₃E₀), and the π system extends over four atoms: one S p-orbital perpendicular to the plane plus one p-orbital on each of three oxygens gives a 4-center system holding 6 π electrons (3 O lone pairs feeding in, sulfur contributing its own p-orbital as the acceptor/participant), producing one fully bonding MO, two degenerate ones of intermediate character, and one antibonding MO — net π bonding delocalized symmetrically over all three S–O linkages. This is why all three S–O bonds are equal and short (142 pm, even shorter than in SO₂), each carrying roughly 1.33 bond order from π delocalization on top of the σ skeleton. Solid SO₃ actually exists as trimers (S₃O₉ rings, "γ-SO₃") or as polymeric chains (β-SO₃, –O–SO₂–O–SO₂–O–) built from corner-sharing SO₄ tetrahedra, because in the condensed phase sulfur prefers four-coordination; the monomeric trigonal-planar form is the gas-phase and dilute-solution species and the one relevant to Lewis-acid reactivity toward water. The Lewis-acidity of SO₃ is the single most important fact for everything that follows: sulfur in SO₃ is electron-poor (formal oxidation state +6, three strongly electron-withdrawing oxo groups), so SO₃ reacts with any nucleophile with startling ease. With water it gives sulfuric acid instantly and exothermically (which is why oleum, not water, is used to absorb SO₃ industrially — direct water absorption is so violent it creates an unmanageable acid mist). With sulfate ion it gives pyrosulfate (disulfate), S₂O₇²⁻, in which SO₃ has inserted into an S–O–H or S–O⁻ bond exactly as it does with water. Once you see SO₃ as "a naked S⁶⁺ core wrapped in three oxo ligands, hungry for a fourth donor," essentially all of oleum and pyrosulfate chemistry, and the mechanism of sulfuric acid formation, follow immediately. Sulfuric acid itself, H₂SO₄, has sulfur at the center of a distorted tetrahedron: two –OH groups and two terminal (non-hydroxyl) O atoms. The S–O(H) bonds are long, about 154 pm, essentially single bonds, while the S=O terminal bonds are short, about 142 pm, again showing pronounced π character analogous to the S–O bonds in SO₃. VSEPR is straightforward — AX₄, four bonding domains, ideal tetrahedral angles, and the real O–S–O angles cluster near 101–124° depending on which pair (terminal–terminal angles open toward ~124° because those bonds are shorter and more repulsive at close range, hydroxyl–hydroxyl angles compress toward ~101°). Sulfurous acid, H₂SO₃, would be the AX₃E tetrahedral-minus-a-vertex analog — sulfur bonded to two OH groups, one terminal O, and one lone pair — but this discrete molecule is not isolable in the solid or liquid state; "sulfurous acid" in aqueous solution is predominantly hydrated, hydrogen-bonded SO₂, with only a small equilibrium population of true H₂SO₃ or its tautomer HOS(=O)OH.

Now overlay the acid-base story onto these shapes. Deprotonating H₂SO₄ once gives HSO₄⁻ (bisulfate), still AX₄ at sulfur, now with one terminal OH and three effectively equivalent O centers (the negative charge delocalizes over the three non-protonated oxygens by resonance/π-conjugation, shortening those S–O bonds relative to the remaining S–OH). Removing the second proton gives sulfate, SO₄²⁻, which is a perfect Td ion — all four S–O bonds equal at 149 pm (intermediate between the 142 pm terminal and 154 pm hydroxyl values of the parent acid, exactly as symmetry demands once all four oxygens are equivalent) with the negative charge spread evenly by symmetry. The same logic applies one rung down in oxidation state: HSO₃⁻ (bisulfite) and SO₃²⁻ (sulfite) are AX₃E pyramidal ions at sulfur, and thiosulfate S₂O₃²⁻ is literally sulfate with one O replaced by S, keeping the Td-like AX₄ shape at the central sulfur but with one S–S bond (201 pm) noticeably longer than the three S–O bonds (147 pm).

The quantitative backbone of this entire family is Pauling's rule for oxoacid strength, applied to a generic oxoacid written as (HO)ₘXOₙ, with $n$ counting the number of non-hydroxyl (terminal, doubly-bonded-character) oxygens on the central atom:

where:

• $n$ is the number of terminal O (or, by extension, S) atoms not bearing an ionizable proton,
• successive p$K_a$ values increase by about 5 units per deprotonation ($K_{a,i+1}/K_{a,i}\approx 10^{-5}$), reflecting the growing electrostatic cost of removing a proton from an increasingly negative species,
• the rule is empirical but tracks the inductive/resonance withdrawal of electron density from the O–H bond by the terminal oxo groups.

Reading: each non-hydroxyl oxygen pulls electron density away from the remaining O–H bonds inductively and stabilizes the developing negative charge on the conjugate base by delocalization, so more terminal oxygens means a stronger acid. Sulfurous acid, written (HO)₂S=O, has $n=1$, predicting p$K_{a1}\approx 3$ (experiment: p$K_{a1}=1.85$, p$K_{a2}=7.2$ — order-of-magnitude agreement, with sulfurous acid's real p$K_{a1}$ somewhat lower than the naive rule because the "acid" is really solvated SO₂ and the effective acidity constant folds in the SO₂(aq) ⇌ H₂SO₃ pre-equilibrium). Sulfuric acid, written (HO)₂SO₂, has $n=2$, predicting p$K_{a1}\approx -2$, i.e. a strong acid fully dissociated in water — which is exactly what's observed (p$K_{a1} < 0$; only the second ionization, HSO₄⁻ ⇌ H⁺ + SO₄²⁻, is measurably weak, p$K_{a2} = 1.99$, consistent with the $+5$ increment rule applied to the second step). The second organizing relationship is the formal oxidation-state bookkeeping that tells you which redox chemistry to expect. For a sulfur center bonded to $x$ oxygens (each counted as −2) and any S–S bonds (which do not change oxidation state, since like atoms split the bonding electrons evenly), the oxidation state of a given sulfur atom is

or more usably, for a mononuclear SₓOᵧ²⁻ fragment with all oxygens as −2 and overall ionic charge $z$:

where:

• $x$ = number of sulfur atoms assumed chemically equivalent (only valid when they truly are, e.g. all four S in S₄O₆²⁻ are NOT equivalent, so this simple form applies cleanly only to mononuclear ions like SO₃²⁻, SO₄²⁻),
• $y$ = number of oxygen atoms (each taken as −2, the almost-always-safe assumption for terminal/bridging oxo groups not part of a peroxo O–O link),
• $z$ = the overall ionic charge.

Reading: this equation is just charge conservation with oxygen fixed at −2; it correctly gives +4 for SO₂/SO₃²⁻ and +6 for SO₃/SO₄²⁻, but it silently breaks down for any species with an S–S bond joining inequivalent sulfurs (thiosulfate, polythionates, dithionite), because those molecules must be solved atom-by-atom using a full Lewis structure, not by a single global equation. That breakdown is not a flaw in the rule — it is the rule correctly telling you that "oxidation state of sulfur in S₂O₃²⁻" is an ill-posed question unless you specify which sulfur.

Example 1 (foundational): predicting the shape and π-bond order pattern of SO₂ from first principles. Sulfur dioxide has 6 (S) + 6 + 6 (two O) = 18 valence electrons, 9 pairs. Draw the Lewis structure: S is central, bonded to two O; distributing electrons to satisfy octets with minimal formal charge gives one S=O double bond and one S=O double bond is not possible with a single resonance structure without leaving a lone pair count mismatch — resolve this by resonance: structure A has S=O(left), S–O(right) with a lone pair on the right O and a negative formal charge there, +1 on S; structure B is the mirror image. The true structure is the symmetric average, both S–O bonds identical, length 143 pm, intermediate between typical S=O (~143–149 pm, e.g., in H₂SO₄ terminal O, 142 pm) and S–O single bond (~157 pm, as in the hydroxyl of H₂SO₄). Electron domain count at S: two bonding domains (treating each S–O multiple bond as one domain) plus one lone pair = 3 domains, AX₂E, predicted bond angle slightly less than 120° (lone pair compresses it), observed 119.5° — excellent agreement, confirming the VSEPR count. The π system: sulfur contributes one 3p orbital and one electron (after σ-bonding accounting) to a 3-center π array spanning S–O–O is wrong geometrically (the two oxygens are not bonded to each other) — correct picture is S(pπ)–O(pπ) and S(pπ)–O(pπ) combining through the shared sulfur p-orbital into a 3-center-4-electron π system (one π bonding MO delocalized over both S–O linkages, one nonbonding MO, one π* antibonding, filled with 4 electrons in bonding + nonbonding), giving each S–O bond a π-bond-order contribution of 0.5 on top of the σ bond, total bond order ≈1.5, consistent with the intermediate 143 pm bond length.

Example 2 (applied): rank the acid strengths of H₂SO₃, HSO₃⁻, H₂SO₄, and HSO₄⁻ using Pauling's rule and check against real p$K_a$ values. Write each in (HO)ₘXOₙ form and count $n$, the terminal (non-OH) oxygens:

• H₂SO₃ = (HO)₂S=O: $n=1$, predicted p$K_{a1} = 8-5(1)=3$; observed p$K_{a1}=1.85$.
• HSO₃⁻ (as an acid, losing its remaining proton) = (HO)SO₂⁻: treat the anionic oxygens as "terminal-like" for the purpose of the second ionization step using the +5-per-step increment rather than recomputing $n$ from scratch: predicted p$K_{a2} = pK_{a1}+5 = 1.85+5=6.85$; observed p$K_{a2}=7.2$ — good agreement using the increment form of the rule.
• H₂SO₄ = (HO)₂SO₂: $n=2$, predicted p$K_{a1}=8-5(2)=-2$; observed: fully dissociated in water, p$K_{a1}$ estimated (from non-aqueous/Hammett measurements) around −3, i.e. a strong acid, consistent with the prediction of "very negative."
• HSO₄⁻ = (HO)SO₃⁻ (second ionization of sulfuric acid): predicted p$K_{a2} = pK_{a1}+5 \approx -2+5=3$; observed p$K_{a2}=1.99$ — same order of magnitude, rule slightly overestimates but correctly places it as a weak acid, much weaker than the parent H₂SO₄.

The ranking predicted (strongest to weakest): H₂SO₄ ≫ HSO₄⁻ > H₂SO₃ > HSO₃⁻, and this matches the experimental ordering exactly (p$K_a$: <0, 1.99, 1.85, 7.2 respectively) apart from the very close H₂SO₃/HSO₄⁻ pair, where the rule's ±1 pK unit accuracy is simply at its expected resolution limit.

Example 3 (exam/competition level): compute the effective Nernst-shifted potential for the S(VI)/S(IV) couple in 18.0 M H₂SO₄ at 358 K (a typical "hot concentrated" digestion temperature), and decide whether it can, in principle, oxidize solid silver ($E^\circ_{\text{Ag}^+/\text{Ag}} = +0.80\text{ V}$). Use the half-reaction $\text{SO}_4^{2-} + 4\text{H}^+ + 2e^- \rightarrow \text{SO}_2(g) + 2\text{H}_2\text{O}$, $E^\circ = 0.17\text{ V}$ at 298 K, and treat $E^\circ$ as approximately temperature-independent over this modest range (a standard simplifying assumption for an estimate). The Nernst equation:

Take $[\text{H}^+]\approx[\text{SO}_4^{2-}]^{1/2}$-scale approximations aside and simply use the bulk acid concentration as a proxy activity for both $[\text{H}^+]$ and $[\text{SO}_4^{2-}]$ at 18.0 M (a crude but standard estimate for concentrated-acid Nernst shifts), and take $P_{\text{SO}_2}\approx 1$ atm (gas escaping at its partial pressure in a well-ventilated system, a reasonable reference state). Then

Compute $18.0^5$: $18.0^2 = 324$, $18.0^3 = 5832$, $18.0^4 = 104{,}976$, $18.0^5 = 1{,}889{,}568$. So the ratio is $5.29\times10^{-7}$, and

At $T=358\text{ K}$, $\frac{RT}{F} = \frac{(8.314)(358)}{96485} = \frac{2976.4}{96485} = 0.03084\text{ V}$, so $\frac{RT}{2F}=0.01542\text{ V}$. Then

So the effective potential of the S(VI)/S(IV) couple under these hot, concentrated conditions rises to about 0.39 V — enough to have oxidized Cu (0.34 V, matching the qualitative fact that hot concentrated sulfuric acid dissolves copper), but still short of the +0.80 V needed to oxidize metallic silver. This correctly reproduces the well-known fact that hot concentrated H₂SO₄ dissolves copper and other moderately reducing metals but does not dissolve silver — silver is attacked instead by nitric acid, whose NO₃⁻/NO couple sits at a substantially higher standard potential (+0.96 V), or by hot concentrated sulfuric acid only under more forcing conditions than this estimate captures (and even then the reaction with Ag is sluggish and incomplete, consistent with the potentials being close but unfavorable). The limiting case is instructive: setting the acid concentration back to 1 M recovers $E = E^\circ = 0.17\text{ V}$ exactly, since $\ln(1/1^5)=\ln 1=0$, confirming the shift vanishes at standard-state concentration as it must.