Differentiating Solvents and Extending the pH Scale

Dissolve perchloric acid and hydrochloric acid in water, and a pH meter reports them as equally strong. Both simply hand every last proton to water to make H₃O⁺. Dissolve the same two acids in glacial acetic acid instead, and they separate cleanly. HClO₄ conducts far better and titrates as the distinctly stronger acid. Nothing about the acids changed; the solvent did. This single experiment contains the whole idea of a differentiating solvent. It explains why chemists needed to build acidity scales that reach far beyond the 0–14 window that water imposes. Water is a superb solvent for everyday chemistry but a poor referee for strong acids and strong bases. Any acid whose intrinsic proton-donating ability exceeds that of H₃O⁺ is forced to react completely. This forcing happens by water's own basicity. This is the leveling effect, and it erases the very differences chemists often want to measure. The same problem occurs in reverse for very strong bases, which are all leveled to hydroxide. Water's amphoteric autoionization, $2\text{H}_2\text{O} \rightleftharpoons \text{H}_3\text{O}^+ + \text{OH}^-$ , fixes both a ceiling and a floor. This ceiling and floor limit what pH can distinguish. Non-aqueous solvents escape this trap because their own acid–base character is tunable. A solvent that is far less basic than water — glacial acetic acid, liquid SO₂, anhydrous HF — no longer accepts a proton from every strong acid equally eagerly. So acids of different intrinsic strength produce different, measurable concentrations of the solvent's own conjugate acid cation. Such solvents are called differentiating. They preserve, rather than erase, the strength differences among acids or bases. These are differences stronger than water can resolve. This behavior rests entirely on the same framework used to classify non-aqueous solvents by their autoionization constants and their own acidity/basicity. This article develops the physical chemistry of leveling and differentiation from the solvent autoionization equilibrium. It shows how acidity functions are constructed once pH itself becomes meaningless. It works through real numbers — autoprotolysis constants, Hammett acidity function values, and superacid data. These numbers let you predict, quantitatively, which solvent will resolve which acid pair. The endpoint of this reasoning is the acidity scale. This scale underlies the very definition of a superacid.

Key takeaways

Water levels all acids stronger than H₃O⁺ (and all bases stronger than OH⁻) to the same apparent strength.
A differentiating solvent is less basic (or less acidic) than water, so it does not force complete proton transfer, preserving intrinsic strength differences.
The window of measurable acid/base strength in any solvent is set by its autoprotolysis constant, $K_{ap}$ .
Solvents like glacial acetic acid, liquid SO₂, and anhydrous HF differentiate strong acids that water levels to "just H₃O⁺."
Beyond any solvent's own pH scale, acidity functions (Hammett $H_0$ ) measure acid strength in media too acidic for pH to apply.
Key idea: Leveling and differentiation are two faces of the same solvent autoionization equilibrium — a weakly basic solvent trades leveling power for resolving power among strong acids.

Where this lives today

Differentiating solvents and extended acidity scales are not historical curiosities; they underpin how modern chemists both measure and use extreme acidity. Non-aqueous titrations in glacial acetic acid or acetonitrile remain the pharmacopeial method for assaying weakly basic pharmaceutical compounds and for resolving mixtures of amine bases that would be indistinguishable by aqueous titration — many drug purity and salt-form assays (hydrochloride versus other salts of active pharmaceutical ingredients) rely on exactly this differentiation. Liquid SO₂ and liquid HF remain preparative solvents in industrial and specialty inorganic chemistry (electrochemical fluorination, for instance, is run in anhydrous HF), where their differentiating acid–base character controls which side reactions and protonation states are accessible. Superacid media built on Hammett-quantified $H_0$ scales — HF/SbF₅, FSO₃H/SbF₅, and related systems — are industrial workhorses for alkane isomerization and cracking catalysis (heterogeneous solid superacids such as sulfated zirconia are engineered analogs), and they are the standard laboratory tool for generating and directly observing long-lived carbocations by NMR, the foundational experimental program associated with George Olah's Nobel-winning chemistry. In biochemistry and enzymology, the same underlying logic — that "acid strength" is meaningless without specifying the reference base — reappears whenever enzyme active sites, membrane interiors, or ionic liquids are treated as non-aqueous or locally non-aqueous microenvironments whose effective acidity must be calibrated on its own terms rather than borrowed uncritically from aqueous pKa tables.

Ask the tutor: How exactly does the choice of indicator ladder in the Hammett method guarantee that $H_0$ values measured with different indicators stay consistent with each other across overlapping acidity ranges?

The physical picture

Consider what "acid strength" really measures: the equilibrium position of proton transfer from an acid HA to some base. In water that base is always H₂O, so the practical acidity scale — pH — is really a measure of $\{\text{H}_3\text{O}^+\}$ activity, and it can never report an acid as "stronger than a fully dissociated strong acid," because once dissociation is complete, adding a stronger acid changes nothing further. Every acid with $K_a \gg 1$ in water reacts essentially quantitatively with water:

\text{HA} + \text{H}_2\text{O} \rightarrow \text{H}_3\text{O}^+ + \text{A}^-

and the solution's acidity is capped at whatever concentration of H₃O⁺ the total acid loading provides — HClO₄, HCl, HNO₃, and H₂SO₄ (its first proton) all look identical in water. This is leveling: the solvent's own conjugate acid (H₃O⁺) becomes the strongest acid that can exist in solution, because anything stronger simply protonates the solvent. The mirror-image statement holds for bases. NaNH₂, NaH, and NaOCH₃ are all leveled to NaOH-strength (i.e., to OH⁻) in water, because any base stronger than OH⁻ deprotonates water completely. The strongest base that can exist in a given solvent is the solvent's own conjugate base. Differentiation breaks this by choosing a solvent whose intrinsic basicity is much lower than water's. Glacial acetic acid (CH₃COOH) is a much weaker proton acceptor than water. When HClO₄ and HCl are dissolved in it, neither is forced to complete, quantitative ionization — instead each establishes its own partial equilibrium

\text{HA} + \text{CH}_3\text{COOH} \rightleftharpoons \text{CH}_3\text{COOH}_2^+ + \text{A}^-

with an equilibrium constant that depends on the intrinsic strength of HA. Because acetic acid is reluctant to accept the proton, only the genuinely strongest acids push this equilibrium far to the right; weaker "strong" acids sit further back. Conductivity, freezing-point depression, and potentiometric titration in glacial acetic acid all show the now-familiar order

\text{HClO}_4 > \text{HBr} > \text{H}_2\text{SO}_4 > \text{HCl} > \text{HNO}_3

a strength ordering that is completely invisible in water, where all five are "100% dissociated, pH = strong acid pH." The same logic, run for bases, explains why liquid ammonia differentiates bases that water levels. NH₃ is both less acidic and (for present purposes) a poorer proton acceptor than water in the relevant regime, so bases such as NaNH₂ and organolithiums, all leveled to "OH⁻ equivalent" in water, retain distinguishable strengths in liquid NH₃. Underlying every one of these solvents is its autoprotolysis (autoionization) equilibrium, exactly analogous to water's:

2\,\text{Solv} \rightleftharpoons \text{Solv-H}^+ + \text{Solv}^-

with an autoprotolysis constant $K_{ap} = a(\text{Solv-H}^+)\,a(\text{Solv}^-)$ . This constant plays the same structural role that $K_w = 1.0\times10^{-14}$ plays in water: it sets both the "neutral point" of the solvent's own pH-like scale and the total span of that scale (from pure Solv-H⁺ to pure Solv⁻). Crucially, $K_{ap}$ also controls leveling power — a solvent with a large $K_{ap}$ (meaning it self-ionizes extensively and both its acidic and basic conjugate forms are only weakly acidic/basic) tends to level strong species, while a solvent with tiny $K_{ap}$ leaves room to differentiate.

Figure

Leveling in water versus differentiation in acetic acid: the same three acids, two solvents

Once an acid is stronger than the strongest acid a given solvent can support (i.e., stronger than protonated solvent), pH-type scales — which are logarithms of an activity referenced to that solvent's autoprotolysis — become physically meaningless. There is no way to define "pH −15" as a concentration statement, because $[\text{H}_3\text{O}^+]$ cannot exceed the concentration set by complete protonation, and superacid media are not simply "very concentrated H₃O⁺." The Hammett acidity function $H_0$ was invented precisely to extend a thermodynamically meaningful acidity scale into this regime, using a family of weakly basic indicators rather than the solvent's own autoprotolysis as the reference.

Figure

Extending the scale: pH's window versus the Hammett H0 continuum

The core ideas

The autoprotolysis (self-ionization) equilibrium of any protic solvent HSolv is

2\,\text{HSolv} \rightleftharpoons \text{H}_2\text{Solv}^+ + \text{Solv}^-

K_{ap} = a(\text{H}_2\text{Solv}^+)\, a(\text{Solv}^-)

where: $K_{ap}$ is the autoprotolysis constant, and $a(\cdot)$ denotes activity (approximated by concentration in dilute, ideal solutions). Reading: exactly as $K_w$ sets water's neutral pH at 7.00 (25 °C) and its 0–14 span, $K_{ap}$ sets the analogous "neutral point" $pK_{ap}/2$ and the full pH-like span $pK_{ap}$ of any solvent.

\text{pH-analog range in solvent} = pK_{ap} = -\log_{10}K_{ap}

where: a larger $pK_{ap}$ (smaller $K_{ap}$ , i.e. the solvent barely self-ionizes) corresponds to a wider window over which acid/base strengths can, in principle, be resolved before hitting the ceiling (pure H₂Solv⁺) or floor (pure Solv⁻). Reading: this is why solvents like glacial acetic acid ( $pK_{ap} \approx 14.5$ ) or liquid ammonia ( $pK_{ap} \approx 29$ near its boiling point) offer generous room to differentiate strong acids or strong bases respectively, compared to water's comparatively modest $pK_{ap} = 14.00$ . For an acid HA dissolving in solvent HSolv, the controlling equilibrium is

\text{HA} + \text{HSolv} \rightleftharpoons \text{H}_2\text{Solv}^+ + \text{A}^-, \qquad K = \frac{a(\text{H}_2\text{Solv}^+)\,a(\text{A}^-)}{a(\text{HA})}

where: $K$ is the effective acidity constant of HA in that solvent, playing the role $K_a$ plays in water. Reading: HA is "leveled" in this solvent if $K \gg 1$ so that the equilibrium runs essentially to completion for every acid under comparison (all HA behave identically, as pure H₂Solv⁺ sources); HA is "differentiated" if $K$ remains modest and different for different acids, so relative amounts of undissociated HA versus H₂Solv⁺/A⁻ vary measurably with the intrinsic strength of HA. Once acids become too strong even for the most weakly basic ordinary solvents (fluorosulfonic acid, oleum, HF/SbF₅ mixtures), a genuine equilibrium constant referenced to a specific solvent autoprotolysis is no longer the natural language; the community instead reports a thermodynamic acidity function. The Hammett acidity function is defined via a weak-base indicator B that undergoes

\text{B} + \text{H}^+ \rightleftharpoons \text{BH}^+

H_0 = pK_{a}(\text{BH}^+) - \log_{10}\frac{[\text{BH}^+]}{[\text{B}]}

where: $pK_a(\text{BH}^+)$ is the (aqueous, dilute-limit) dissociation constant of the protonated indicator, and $[\text{BH}^+]/[\text{B}]$ is measured spectrophotometrically (protonated and free indicator have different UV-vis absorptions) in the actual superacid medium. Reading: $H_0$ reduces to ordinary pH in dilute aqueous solution (where activity coefficient ratios cancel and $[\text{H}^+]\approx a(\text{H}_3\text{O}^+)$ ), but remains well-defined and physically meaningful even when the medium is so acidic that "pH" as a concentration statement has lost meaning — this is exactly the extension the title promises, and it is the formal gateway to the superacid definition ( $H_0 < -12$ , more acidic than pure H₂SO₄).

Derivation AWhy a solvent's autoprotolysis constant sets its leveling/differentiating power

Goal: show, from the autoprotolysis equilibrium alone, why a solvent with small $K_{ap}$ differentiates strong acids that a solvent with large $K_{ap}$ (or, more precisely, high solvent basicity) levels.

Step 1 — Define the reference acid. In any solvent HSolv, the strongest acid that can exist in solution is the protonated solvent H₂Solv⁺ (any stronger acid transfers its proton to HSolv and becomes H₂Solv⁺ plus its own conjugate base). Symmetrically, the strongest base that can exist is Solv⁻. This is a direct structural consequence of proton-transfer thermodynamics: if HA is a stronger acid than H₂Solv⁺ (meaning HA more readily gives up its proton), the reaction HA + HSolv → H₂Solv⁺ + A⁻ is exergonic and proceeds forward.

Step 2 — Compare two acids of different intrinsic strength. Let HA₁ be intrinsically much stronger than HA₂ (e.g. HClO₄ versus HNO₃). In water, both easily out-compete H₂O for the proton — their reactions with water are both strongly exergonic, driving both essentially to 100% completion. Because water is a moderately good proton acceptor (basicity is high enough that even "medium-strong" acids exceed the H₃O⁺ reference), the distinguishing free-energy gap between HA₁ and HA₂ (which might be tens of kJ/mol) is swamped: both land past the "cliff edge" into full ionization, and the system cannot resolve which one fell further past the edge.

Step 3 — Reduce the solvent's proton affinity (basicity). Now transfer both acids to a solvent HSolv that is a much weaker proton acceptor than water — glacial acetic acid is the standard case, since the conjugate acid CH₃COOH₂⁺ is itself a reasonably strong acid, meaning acetic acid does not want the proton nearly as badly as water does. The reaction HA + CH₃COOH ⇌ CH₃COOH₂⁺ + A⁻ is now only weakly exergonic (or even mildly endergonic) for both acids, so neither reaction runs to full completion. The equilibrium constants $K_1$ (for HA₁) and $K_2$ (for HA₂) are now both modest, of order 1–100 rather than $10^{6}$ – $10^{10}$ , and — critically — they retain the same relative ordering and gap set by the acids' intrinsic gas-phase/solution proton affinities. Because the reaction no longer saturates at "100% ionized" for either acid, the difference between $K_1$ and $K_2$ shows up directly in measurable quantities: solution conductivity (which tracks free ion concentration), freezing-point depression (colligative particle count), and potentiometric titration curves (sharpness and equivalence-point potential).

Step 4 — State the general rule. A solvent differentiates a pair of acids exactly when both acids' effective ionization constants in that solvent are of comparable, sub-saturating magnitude — neither pinned at the ceiling (complete ionization) nor at the floor (negligible ionization). This happens generically when the solvent's basicity (its proton affinity, closely tracked by the acidity of its own conjugate acid H₂Solv⁺, i.e. by $pK_{ap}$ combined with the intrinsic acid strength of H₂Solv⁺) is low enough that even the strongest acids of interest do not exceed the reference "cliff." Water's basicity happens to sit below essentially all mineral acids' strengths, so it levels; acetic acid's basicity sits below (or comparable to) the mineral acids' strengths in a way that spreads them out, so it differentiates. The same argument, run for base strength using Solv⁻ as the reference conjugate base, explains why liquid ammonia differentiates strong bases that water levels to hydroxide.

Go deeper · Derivation B · 2 minExtending the acidity scale: constructing

H_0

and reading off superacid strength

Derivation BExtending the acidity scale: constructing

H_0

and reading off superacid strength

Goal: derive the Hammett acidity function from first principles of indicator protonation equilibria, and show explicitly why it remains valid outside the region where pH has meaning.

Step 1 — Start from the indicator equilibrium. Choose a weak, uncharged, colored base B (a substituted nitroaniline is the classical choice) that protonates reversibly:

\text{B} + \text{H}^+ \rightleftharpoons \text{BH}^+

Its thermodynamic acidity constant, referenced in the usual way to the protonated form's dissociation, is

K_{BH^+} = \frac{a(\text{H}^+)\,a(\text{B})}{a(\text{BH}^+)}

so that

pK_{a}(\text{BH}^+) = -\log_{10}a(\text{H}^+) - \log_{10}\frac{a(\text{B})}{a(\text{BH}^+)}

Step 2 — Recognize that $a(\text{H}^+)$ is not directly accessible in strongly acidic, non-ideal media. In dilute aqueous solution, $-\log_{10}a(\text{H}^+)$ is exactly pH by definition, and it can be measured directly (e.g. glass electrode). In concentrated sulfuric acid or in HF/SbF₅ mixtures, activity coefficients diverge wildly from unity, ionic strength is enormous, and there is no meaningful "concentration of free H⁺" to speak of — the solvent medium itself is saturated with proton-donating/accepting species. What can still be measured is the ratio $[\text{BH}^+]/[\text{B}]$ , spectrophotometrically, because protonation changes the indicator's chromophore (loss of the amine lone pair into the ring, for a nitroaniline, shifts the absorption sharply).

Step 3 — Define $H_0$ to absorb exactly the terms that resist direct measurement. Hammett's move is to define a function that equals $-\log_{10}a(\text{H}^+)$ whenever that quantity is meaningful, but that remains computable from the measurable ratio even when it is not:

H_0 \equiv pK_a(\text{BH}^+) - \log_{10}\frac{[\text{BH}^+]}{[\text{B}]}

Because $pK_a(\text{BH}^+)$ is measured once, in dilute aqueous solution (where it does equal $-\log_{10}a(\text{H}^+) + \log_{10}([\text{BH}^+]/[\text{B}])$ rearranged, i.e. is a genuine thermodynamic constant extrapolated to infinite dilution), the same numerical constant can be carried into any medium. In the new medium only the indicator ratio needs to be re-measured.

Step 4 — Check consistency: $H_0 \to$ pH in the dilute aqueous limit. As the medium approaches dilute aqueous solution, activity coefficients of neutral species B and BH⁺ approach those used in defining $pK_a(\text{BH}^+)$ , and

H_0 = pK_a(\text{BH}^+) - \log_{10}\frac{[\text{BH}^+]}{[\text{B}]} \longrightarrow -\log_{10}a(\text{H}_3\text{O}^+) = \text{pH}

recovering the familiar scale exactly where both are valid — this is the required limiting-case check: $H_0$ is not a new, incompatible quantity, but pH's proper generalization.

Step 5 — Chain indicators to cover an enormous range. No single indicator's $[\text{BH}^+]/[\text{B}]$ ratio can be measured accurately over more than about 2 units of acidity (below some ratio the protonated form is undetectable; above it the free base is undetectable). Hammett's method uses a ladder of indicators with overlapping, successively lower $pK_a(\text{BH}^+)$ values (each next indicator calibrated by an acidity region where the previous indicator's ratio is still measurable), so that $H_0$ can be walked down continuously from dilute aqueous acid to $H_0 = -12$ (pure H₂SO₄) and on to $H_0 \approx -21$ to $-23$ (HF·SbF₅, "magic acid"/fluoroantimonic acid systems) — thirteen or more orders of magnitude past where pH itself stopped meaning anything. This ladder construction is exactly how the operational definition of a superacid ( $H_0$ more negative than that of pure sulfuric acid) becomes a measured number rather than a qualitative label.

Worked examples

Example 1 — Foundational: water's leveling window from $K_w$ , and why HClO₄/HCl are indistinguishable. Water's autoprotolysis constant at 25 °C is

K_w = [\text{H}_3\text{O}^+][\text{OH}^-] = 1.0\times10^{-14}

so $pK_w = 14.00$ and water's pH-type window runs from pH 0 (essentially pure H₃O⁺, e.g. 1 M strong acid) to pH 14 (essentially pure OH⁻, e.g. 1 M strong base), with neutrality at pH 7.00. Take 0.10 M HClO₄ and 0.10 M HCl, both nominally "strong" acids ( $K_a \gg 1$ in water for each). Both react with water essentially completely:

\text{HClO}_4 + \text{H}_2\text{O} \rightarrow \text{H}_3\text{O}^+ + \text{ClO}_4^-, \qquad K \sim 10^{9}\text{–}10^{10}

\text{HCl} + \text{H}_2\text{O} \rightarrow \text{H}_3\text{O}^+ + \text{Cl}^-, \qquad K \sim 10^{6}\text{–}10^{7}

Even though the underlying equilibrium constants differ by 2–3 orders of magnitude — a real, substantial difference in intrinsic acid strength — both reactions are so far to completion that essentially all of the 0.10 M acid ends up as H₃O⁺ in either case:

[\text{H}_3\text{O}^+] \approx 0.10\ \text{M in both cases}, \qquad \text{pH} \approx 1.00 \text{ for both}

This is leveling in numbers: a 100–1000-fold difference in intrinsic $K$ produces a pH difference smaller than $0.001$ units, far below any experimentally resolvable margin. Limiting-case check: as the intrinsic $K$ for either acid is pushed even higher (say from $10^9$ to $10^{15}$ ), $[\text{H}_3\text{O}^+]$ does not change at all — it stays pinned at 0.10 M — confirming that once $K \gg 1/C_0$ (with $C_0$ the total acid concentration), further increases in intrinsic strength are invisible to a water-based pH measurement.

Example 2 — Applied: differentiating HClO₄ from HCl in glacial acetic acid using conductivity-derived ionization constants. Glacial acetic acid has autoprotolysis

2\,\text{CH}_3\text{COOH} \rightleftharpoons \text{CH}_3\text{COOH}_2^+ + \text{CH}_3\text{COO}^-, \qquad pK_{ap} \approx 14.5

Because acetic acid is a much weaker base than water (its conjugate acid CH₃COOH₂⁺ is a considerably stronger acid than H₃O⁺, meaning acetic acid "wants" the proton far less), the reaction

\text{HA} + \text{CH}_3\text{COOH} \rightleftharpoons \text{CH}_3\text{COOH}_2^+ + \text{A}^-

no longer saturates for typical mineral acids. Reported ionization constants (from conductometric and cryoscopic measurements in glacial acetic acid) place HClO₄ roughly two orders of magnitude more ionized than HCl under comparable conditions; representative literature values give effective ionization constants of order

K(\text{HClO}_4) \approx 4\times10^{-5}, \qquad K(\text{HCl}) \approx 5\times10^{-7}

(values vary modestly with the exact experimental protocol, but the ordering and rough two-order-of-magnitude gap are robust). Take an asymmetric loading, $C_0(\text{HClO}_4) = 0.050$ M and $C_0(\text{HCl}) = 0.20$ M, and estimate the ionized fraction $\alpha$ for each from the weak-acid-style relation

K \approx \frac{(\alpha C_0)^2}{C_0(1-\alpha)} \approx \alpha^2 C_0 \quad (\alpha \ll 1)

For HClO₄:

\alpha^2 = \frac{K}{C_0} = \frac{4\times10^{-5}}{0.050} = 8.0\times10^{-4} \implies \alpha \approx 0.028

For HCl:

\alpha^2 = \frac{K}{C_0} = \frac{5\times10^{-7}}{0.20} = 2.5\times10^{-6} \implies \alpha \approx 1.6\times10^{-3}

So even though HCl is present at four times the total concentration of HClO₄, its ionized (conducting) concentration is

[\text{CH}_3\text{COOH}_2^+]_{\text{from HCl}} = \alpha C_0 = 1.6\times10^{-3}\times0.20 = 3.2\times10^{-4}\ \text{M}

while HClO₄ gives

[\text{CH}_3\text{COOH}_2^+]_{\text{from HClO}_4} = 0.028\times0.050 = 1.4\times10^{-3}\ \text{M}

roughly 4–5 times more free ions from the more dilute HClO₄ solution — a clearly measurable, order-unity conductivity difference, in stark contrast to water where both would look identical. Limiting-case check: if the acetic acid basicity were pushed up toward water's (i.e. $K$ values both driven toward $\alpha \to 1$ ), this differentiation collapses back to leveling, exactly consistent with the mechanism derived above.

Example 3 — Exam/competition level: locating fluorosulfonic acid on the Hammett scale and testing whether it qualifies as a superacid. Fluorosulfonic acid, HSO₃F, is used industrially and in superacid chemistry (often mixed with SbF₅ to make "magic acid," FSO₃H·SbF₅). Its measured Hammett acidity function is

H_0(\text{HSO}_3\text{F}) \approx -15.1

compared with pure sulfuric acid,

H_0(\text{H}_2\text{SO}_4, 100\%) \approx -12.0

The operational superacid criterion is $H_0$ more negative than $-12$ (i.e., more acidic than pure H₂SO₄). Since

-15.1 < -12.0

HSO₃F satisfies the criterion by a margin of 3.1 $H_0$ units. Because $H_0$ is defined through

H_0 = pK_a(\text{BH}^+) - \log_{10}\frac{[\text{BH}^+]}{[\text{B}]}

each unit of $H_0$ corresponds to a factor-of-10 change in the protonated/free indicator ratio at fixed indicator, so a 3.1-unit difference means the same indicator, in HSO₃F relative to pure H₂SO₄, sits at a protonated-to-free ratio larger by

10^{3.1} \approx 1260

— roughly a thousand-fold greater degree of protonation of the same weak base, using the same logic that lets one compare acidities once ordinary pH has failed. This is exactly the quantitative handle used to rank superacid media (fluoroantimonic acid, HF–SbF₅ mixtures at various mole ratios, reach even further, to $H_0 \approx -21$ to $-23$ , corresponding to protonation ratios roughly $10^{9}$ – $10^{11}$ times greater than in pure H₂SO₄) and is the direct numerical bridge into the formal treatment of superacids.

Limiting cases, scope & common traps

Regime	What happens	Governing quantity
Acid intrinsically weaker than H₂Solv⁺	Ordinary partial ionization; solvent's own scale applies normally (this is just "weak acid in water" territory)	$K \ll 1$ relative to solvent autoprotolysis reference
Acid intrinsically much stronger than H₂Solv⁺, solvent strongly basic (e.g. water for mineral acids)	Complete ionization; acid strength leveled to strength of H₂Solv⁺ (H₃O⁺ in water)	$K \gg 1$ , saturates regardless of exact value
Acid strength spans a range near the solvent's own basicity threshold, solvent weakly basic (e.g. acetic acid, liquid SO₂)	Differentiation: distinguishable partial ionization for each acid	$K$ values remain sub-saturating and spread over orders of magnitude
Acid stronger than the strongest species any ordinary solvent's conjugate acid can represent	pH/pK-type language breaks down; must use an acidity function	Hammett $H_0$ (or related scales)
As solvent basicity → water's basicity (from below)	Differentiation smoothly collapses back into leveling	Transition tracked continuously by effective $K$ approaching saturation
As indicator's own $pK_a(\text{BH}^+)$ regime overlaps dilute aqueous acid	$H_0 \to$ pH exactly	Consistency check built into the $H_0$ definition

Common trapMisconception: "a stronger solvent acid always means a wider leveling window, so H₂SO₄ should level everything"

It is tempting to think that because pure sulfuric acid is a strongly acidic medium, it should level acids the way water levels bases — erasing all differences among strong acids by forcing them all to full protonation. The opposite is closer to the truth for the acid side: sulfuric acid is such a weak base (poor proton acceptor, since H₃SO₄⁺ is a very strong, unstable acid) that it does not level strong acids at all — instead it differentiates them, and in fact many "strong" mineral acids barely ionize in it. What sulfuric acid strongly levels is bases: any base at all stronger than HSO₄⁻ is leveled to that reference. Leveling and differentiating behavior must always be assigned separately for the acid side and the base side of a solvent, judged by comparing to that solvent's own conjugate acid and conjugate base respectively.

Common trapMisconception: "pH is just a universal number for acidity, so 'pH = −20' for a superacid is a legitimate extrapolation"

Because pH is defined as $-\log_{10}a(\text{H}_3\text{O}^+)$ specifically referenced to water's autoprotolysis and to dilute, ideal-solution activity conventions, it is not meaningful to write "pH = −20" for fluoroantimonic acid: there is no aqueous H₃O⁺ activity in that medium at all, and the very act of dissolving such an acid in water would immediately react it away or dilute it into an entirely different chemical system. The correct quantity is the Hammett acidity function $H_0$ , built from indicator equilibria that remain interpretable in strongly non-ideal, non-aqueous, superacidic media — a genuinely different (though numerically continuous) scale, not merely "pH extrapolated below zero."

The chemist behind it

Origins · 1 minLouis Plack Hammett and how this was discovered

Louis Plack Hammett

1894–1987 · Physical organic chemistry, Columbia University, 1920s–1950s

Hammett trained as a physical chemist but became the central architect of quantitative physical organic chemistry in the United States. Working at Columbia University from the 1920s onward, he confronted a very practical problem that arose repeatedly in industrial and mechanistic organic chemistry: reaction media (concentrated sulfuric acid used for nitrations, sulfonations, and Friedel–Crafts-type chemistry) were far too acidic for pH electrodes or aqueous acid-strength language to describe meaningfully, yet chemists needed a reliable, reproducible number to correlate reaction rates and mechanisms across different strongly acidic solvents. Hammett's solution — publishing the acidity function concept in the early 1930s together with coworker A. J. Deyrup, using a ladder of substituted nitroaniline and related weak-base indicators calibrated by ultraviolet-visible spectrophotometry — turned a qualitative sense of "very strong acid" into a rigorously defined, experimentally anchored thermodynamic quantity. The same conceptual toolkit (linear free-energy relationships, substituent effects, acidity functions) that Hammett built around this problem became foundational to physical organic chemistry generally, most famously in the Hammett equation relating substituent constants to reaction rates and equilibria — but the acidity function, $H_0$ , remains his most direct and lasting contribution to inorganic and physical acid–base chemistry, providing the operational backbone later used to define and rank superacids.

History

- 1884–1887 — Arrhenius and Ostwald establish the electrolytic theory of acids and bases, giving the first quantitative basis for aqueous acid strength.

- 1923 — Brønsted and Lowry (independently) define acids and bases by proton donation/acceptance, freeing acid–base chemistry from water-specific definitions and enabling non-aqueous acid–base theory.

- 1932–1934 — Hammett and Deyrup develop the acidity function

H_0

, using indicator ladders to quantify acidity in concentrated sulfuric acid media beyond the reach of pH.

- 1927–1940s — Systematic conductometric and cryoscopic studies (notably by Kolthoff, Bruckenstein, and others) establish the differentiating behavior of glacial acetic acid, liquid ammonia, and liquid SO₂ toward mineral acids and bases.

- 1966 — Olah and coworkers popularize the term "superacid" and develop HF–SbF₅ (and related fluoroantimonic) systems, extending measured

H_0

values well past −20 and founding modern superacid chemistry.

Why can't we just make a "more sensitive" pH meter to resolve leveled acids in water?

The limitation is not instrumental resolution but chemistry: once an acid reacts essentially completely with water, there is simply no remaining undissociated HA left in solution to distinguish one acid from another, and the H3O+ concentration is capped by the total acid loading, not by the acid's intrinsic strength. A more sensitive electrode would report the same [H3O+] more precisely, but that number itself no longer contains information about which acid was stronger. The only way to recover that information is to move to a solvent whose own basicity is low enough that the reaction with the acid does not go fully to completion, which is a chemical change of reference frame, not an instrumental upgrade.

Is glacial acetic acid differentiating for both acids and bases, or just acids?

Glacial acetic acid is well known as a differentiating solvent for strong acids, because it is a weak base and its conjugate acid, protonated acetic acid, is itself a fairly strong acid, so it does not force complete protonation of mineral acids. For bases, acetic acid behaves more like a leveling solvent, because it is itself a moderately strong Brønsted acid (pKa around 4.76 in water, but acting neat it readily donates protons), so essentially any base stronger than acetate ion gets leveled to acetate-strength. This asymmetry — differentiating on one side, leveling on the other — is completely general and must be checked separately for acids and bases in any solvent.

How is the autoprotolysis constant of a non-aqueous solvent actually measured?

Conductometric methods are the classic approach: pure, rigorously dried solvent still shows a small residual conductivity attributable to the tiny concentration of self-ionized ionic species, and this conductivity, combined with known ionic mobilities (often estimated from related salts dissolved in the same solvent), gives the ion product. Cryoscopic (freezing-point depression) and spectroscopic methods provide independent cross-checks in some solvents. For solvents like liquid ammonia or liquid HF, these measurements must be done under carefully controlled temperature and pressure since the solvents are volatile at room temperature, adding significant experimental difficulty compared with water.

Does the Hammett H0 scale apply only to acids, or is there an analogous function for very strong bases?

There is a directly analogous function, usually denoted H_ (H-minus), constructed the same way but using indicators that are weak acids losing a proton rather than weak bases gaining one. H_ is used to quantify extremely basic media — molten alkali hydroxide mixtures, superbase systems, and certain amide/hydride-containing non-aqueous solutions — exactly mirroring how H0 extends acidity measurement past the aqueous pH floor of 0, H_ extends basicity measurement past the aqueous pH ceiling of 14.

Why does the same acid, HCl, ionize essentially completely in water but only partially in glacial acetic acid?

The degree of ionization is set by the free-energy change of the proton-transfer reaction with the specific solvent acting as base, not by any fixed, solvent-independent property of HCl. Water is a considerably stronger proton acceptor (more basic in this specific sense) than acetic acid, largely because the resulting H3O+ is a comparatively weak, well-stabilized acid, whereas protonated acetic acid (CH3COOH2+) is itself a much stronger, less stable acid. The reaction HCl + H2O → H3O+ + Cl- is therefore far more exergonic than HCl + CH3COOH ⇌ CH3COOH2+ + Cl-, and equilibrium position tracks that free-energy difference directly.

Is it correct to say "superacids don't have a pH"?

Essentially yes, in the strict operational sense: pH specifically denotes $-\log_{10}$ of the activity of H3O+ in a dilute, aqueous, close-to-ideal solution, and superacid media are neither aqueous nor dilute nor close-to-ideal, so there is no H3O+ activity to report. What superacids do have is a well-defined, experimentally measured Hammett acidity function H0, built from indicator protonation equilibria rather than from any assumption about a specific solvated proton species, and this is the physically meaningful quantity to cite instead of an extrapolated "very negative pH."

How does this connect to the definition of a superacid?

The entire differentiating-solvent framework is what makes the superacid definition rigorous rather than qualitative. Because pH fails once acidity exceeds what water (or any single reference solvent) can express, superacids are defined using the solvent-independent Hammett function: any medium with $H_0$ more negative than that of pure sulfuric acid ( $H_0 \approx -12$ ) is classified as a superacid. This threshold is not arbitrary — it is chosen because 100% sulfuric acid is the strongest single, "ordinary" Brønsted acid readily available, so anything measurably more acidic by the same indicator-based yardstick genuinely represents a new acidity regime.