Amphiprotic Species and Intermediate pH

Dissolve pure sodium bicarbonate in water. The pH settles near 8.3. This holds almost regardless of whether the solution is 0.01 M or 0.5 M. That fact seems to defy everything learned about concentration-dependent equilibria. This near-constancy is the signature of an amphiprotic species. It is a substance like HCO₃⁻ that can act as both a Brønsted acid and base. It does this toward the same solvent. Bicarbonate, the dihydrogen phosphate ion H₂PO₄⁻, the hydrogen phthalate ion, and the glycine zwitterion all belong to this class. All obey the same remarkably compact formula for pH. This matters because amphiprotic intermediates are everywhere in real analytical chemistry. They are the working species in blood buffer chemistry (HCO₃⁻). They also appear in biochemical buffers (phosphate, Tris⁺-like systems). They dominate the equivalence points of essentially every diprotic and polyprotic titration curve. Any time a titration of a diprotic acid H₂A with base passes through the "half-neutralized-twice" point, the solution is dominated by HA⁻. Its pH is not found by a simple Henderson–Hasselbalch calculation. That method only works at the first half-equivalence point. Getting this wrong is one of the most common errors students make. It happens when they extend single-acid intuition to polyprotic systems. The big picture is this: an amphiprotic species sits at a saddle point. This point lies between two proton-transfer equilibria. It can lose a proton, characterized by $K_{a2}$ . It can also accept one, governed by the conjugate relationship to $K_{a1}$ . Building on Polyprotic Acid Equilibria and Stepwise Ka, this article derives the governing pH expression rigorously. The derivation uses mass and charge balance. It shows exactly when the beautifully simple $\sqrt{K_{a1}K_{a2}}$ shortcut is valid. It also shows when this shortcut fails. It also connects the result to the hydrolysis reasoning already familiar from Salt Solutions and Hydrolysis. By the end you will be able to write the exact quadratic pH expression for any amphiprotic salt. You will know precisely which terms can be dropped and why. You will compute pH for bicarbonate, phosphate, and phthalate systems with proper significant figures. You will also see how this single idea anchors the shape of every polyprotic titration curve. This foreshadows the fuller picture of Fractional Composition of Diprotic Systems.

Key takeaways

An amphiprotic species like HA⁻ can donate a proton ( $K_{a2}$ ) or accept one (reverse of $K_{a1}$ ), placing it at equilibrium between two acid–base systems simultaneously.
The exact proton balance gives a general pH expression; under common conditions it collapses to $\text{pH} \approx \tfrac{1}{2}(\text{p}K_{a1}+\text{p}K_{a2})$ .
This pH is approximately independent of concentration — the defining experimental fingerprint of an amphiprotic intermediate.
The approximation fails when $K_{a1}$ is not ≫ $C\cdot K_{a2}$ , i.e., at low concentration or when $K_{a1}$ is only weakly larger than $K_{w}/K_{a2}$ -type terms.
The full quadratic (or cubic, in the most rigorous form) is needed near titration equivalence points where concentrations are low or $K_a$ values are close together.
This same mathematics governs the pH at both intermediate equivalence points of any diprotic titration curve.
Key idea: the pH of an amphiprotic species is fixed by the geometric mean of its two flanking dissociation constants, almost independent of how much of it you dissolve.

Where this lives today

Blood pH regulation depends critically on the bicarbonate/carbonic acid buffer, and the amphiprotic behavior of HCO₃⁻ (rather than a naive single-equilibrium treatment) explains why physiological buffering is more subtle than a textbook monoprotic buffer — clinical acid–base analysis (the Henderson–Hasselbalch equation applied to blood gases) implicitly relies on knowing where HCO₃⁻ sits relative to $K_{a1}$ and $K_{a2}$ of the carbonic acid system. Biochemical buffer design routinely exploits amphiprotic intermediates: phosphate buffers used in molecular biology (PBS) are prepared near the $\text{p}K_{a2}$ of H₂PO₄⁻/HPO₄²⁻, and understanding the amphiprotic pH of H₂PO₄⁻ itself is essential when formulating buffers at pH values between the two useful phosphate pKa's. Amino acids and peptides are amphiprotic at their isoelectric point in an entirely analogous sense — the zwitterionic form dominates and its pH-independent character (the isoelectric pH, pI) is calculated by the very same $\tfrac12(\text{p}K_{a1}+\text{p}K_{a2})$ formula, just relabeled for the amino and carboxyl $K_a$ values. In analytical titrimetry, every diprotic acid titration curve (e.g., titrating oxalic acid or phthalic acid with NaOH) passes through an intermediate equivalence point where the solution is essentially pure NaHA; recognizing this point as governed by the amphiprotic formula — rather than mistakenly treating it with simple buffer arithmetic — is essential for correctly predicting inflection points, choosing indicators, and interpreting potentiometric titration data. Industrially, buffer formulation for pharmaceuticals, fermentation broths, and water treatment (carbonate/bicarbonate alkalinity control) all depend on this same equilibrium logic, and it directly sets up the more complete picture developed in Fractional Composition of Diprotic Systems, where the amphiprotic point appears as the crossover of the $\alpha_{\text{HA}^-}$ fraction curve.

:::quiz[In the exact formula for [H+], what does the correction term (C + Ka1) in the denominator physically account for?]

The autoionization of water alone
The fact that some HA⁻ is converted back to H2A, depleting its equilibrium concentration below C ✓
Temperature dependence of Ka1
The ionic strength of the solution

Ask the tutor: How would the amphiprotic pH formula change if the parent acid were triprotic and we wanted the pH of the middle species HA²⁻ sitting between Ka2 and Ka3?

The physical picture

Consider the diprotic acid H₂A with successive dissociation constants $K_{a1}$ and $K_{a2}$ , where $K_{a1} \gg K_{a2}$ (true for essentially every real diprotic acid, since removing a second proton from an already-negative ion is always harder). The intermediate species HA⁻ is produced when H₂A loses one proton but two remain unremoved from the parent acid's total supply. HA⁻ is genuinely two-faced. It can act as an acid,

\text{HA}^- \rightleftharpoons \text{H}^+ + \text{A}^{2-}, \qquad K_{a2}

or it can act as a base, reversing the first dissociation step,

\text{HA}^- + \text{H}_2\text{O} \rightleftharpoons \text{H}_2\text{A} + \text{OH}^-, \qquad K_{b} = \frac{K_w}{K_{a1}}

Whether a solution of pure NaHA ends up acidic or basic depends on the competition between these two processes. If $K_{a2}$ (acid strength of HA⁻) exceeds $K_b = K_w/K_{a1}$ (base strength of HA⁻), the solution is net acidic; otherwise it is net basic. For bicarbonate, $K_{a1}(\text{H}_2\text{CO}_3) = 4.45\times10^{-7}$ and $K_{a2} = 4.69\times10^{-11}$ , so $K_b = K_w/K_{a1} = 2.25\times10^{-8}$ , which exceeds $K_{a2}$ — correctly predicting that bicarbonate solutions are mildly basic (pH ≈ 8.3), matching everyday experience with baking soda. The remarkable part is not just which way it tips but how little the concentration matters. In the derivation below, the proton condition forces essentially all concentration-dependence to cancel between numerator and denominator, leaving a pH set almost purely by the two equilibrium constants. This is why a physiological bicarbonate buffer and a dilute laboratory bicarbonate solution show nearly the same pH despite a hundred-fold difference in concentration — a fact with direct relevance to how the Salt Solutions and Hydrolysis framework must be extended for species with two roles.

Figure

HA⁻ sits between two proton-transfer equilibria: donating to give A²⁻, or accepting to regenerate H₂A

Figure

Bicarbonate pH is nearly flat across three decades of concentration — the experimental fingerprint of amphiprotism

The core ideas

The rigorous starting point is always mass balance plus charge balance (or, equivalently, the proton condition). For a solution of the salt NaHA at formal (analytical) concentration $C$ , the species present are H₂A, HA⁻, A²⁻, H⁺, OH⁻, and Na⁺.

Charge balance:

[\text{Na}^+] + [\text{H}^+] = [\text{HA}^-] + 2[\text{A}^{2-}] + [\text{OH}^-]

Mass balance: $[\text{Na}^+] = C$ and $[\text{H}_2\text{A}] + [\text{HA}^-] + [\text{A}^{2-}] = C$ . Combining these (subtract mass balance for A from the charge balance) yields the proton condition, the cleanest route to the result:

[\text{H}_2\text{A}] + [\text{H}^+] = [\text{A}^{2-}] + [\text{OH}^-]

where:

$[\text{H}_2\text{A}]$ — concentration of the fully protonated (parent) acid, formed when HA⁻ acts as a base
$[\text{H}^+]$ — free hydrogen ion concentration
$[\text{A}^{2-}]$ — concentration of the fully deprotonated conjugate base, formed when HA⁻ acts as an acid
$[\text{OH}^-]$ — hydroxide from water autoionization

This states plainly that every proton HA⁻ picks up (making H₂A) must be balanced by a proton it gives away (making A²⁻ or OH⁻) — the defining condition of an amphiprotic buffer. Substituting the equilibrium expressions

[\text{H}_2\text{A}] = \frac{[\text{H}^+][\text{HA}^-]}{K_{a1}}, \qquad [\text{A}^{2-}] = \frac{K_{a2}[\text{HA}^-]}{[\text{H}^+]}, \qquad [\text{OH}^-] = \frac{K_w}{[\text{H}^+]}

into the proton condition and solving for $[\text{H}^+]$ gives the general, exact result:

[\text{H}^+] = \sqrt{\dfrac{K_{a1}K_{a2}[\text{HA}^-] + K_{a1}K_w}{K_{a1} + [\text{HA}^-]}}

where:

$[\text{HA}^-]$ — the equilibrium concentration of the amphiprotic ion (not quite equal to $C$ , since a little is consumed/produced by the side reactions)
$K_{a1}, K_{a2}$ — successive acid dissociation constants of the parent diprotic acid
$K_w$ — ion product of water, $1.0\times10^{-14}$ at 25 °C

Reading this equation: the numerator combines the "acting as acid toward A²⁻" pathway ( $K_{a1}K_{a2}[\text{HA}^-]$ ) with the "generating OH⁻ via water" pathway ( $K_{a1}K_w$ ); the denominator corrects for how much HA⁻ has actually been consumed making H₂A. When $[\text{HA}^-] \approx C \gg K_{a1}$ and $C K_{a2} \gg K_w$ , both correction terms vanish and the whole expression collapses to the compact and famous:

[\text{H}^+] \approx \sqrt{K_{a1}K_{a2}} \quad\Longrightarrow\quad \text{pH} \approx \tfrac{1}{2}\left(\text{p}K_{a1} + \text{p}K_{a2}\right)

This is the celebrated result: the pH of an amphiprotic species is the arithmetic mean of the two flanking pKa values, and — crucially — it contains no $C$ at all, explaining the experimental flatness seen above.

Derivation AFrom proton condition to the sqrt(Ka1 Ka2) formula

Goal. Derive $[\text{H}^+] = \sqrt{K_{a1}K_{a2}[\text{HA}^-]+K_{a1}K_w \over K_{a1}+[\text{HA}^-]}$ from mass/charge balance, then show the standard simplification.

Step 1 — Write charge and mass balance. For NaHA at formal concentration $C$ : charge balance is $[\text{Na}^+]+[\text{H}^+] = [\text{HA}^-]+2[\text{A}^{2-}]+[\text{OH}^-]$ . Mass balance on sodium gives $[\text{Na}^+]=C$ ; mass balance on the acid gives $C = [\text{H}_2\text{A}]+[\text{HA}^-]+[\text{A}^{2-}]$ .

Step 2 — Eliminate C to get the proton condition. Substitute $[\text{Na}^+]=C=[\text{H}_2\text{A}]+[\text{HA}^-]+[\text{A}^{2-}]$ into the charge balance:

[\text{H}_2\text{A}]+[\text{HA}^-]+[\text{A}^{2-}]+[\text{H}^+] = [\text{HA}^-]+2[\text{A}^{2-}]+[\text{OH}^-]

The $[\text{HA}^-]$ terms cancel, and one $[\text{A}^{2-}]$ cancels against another, leaving the compact proton condition:

[\text{H}_2\text{A}]+[\text{H}^+] = [\text{A}^{2-}]+[\text{OH}^-]

This says the net "extra" protons on the acid side (relative to the reference level HA⁻) must equal the net "missing" protons on the base side — exactly balancing the two competing roles of HA⁻.

Step 3 — Insert equilibrium expressions. From $K_{a1} = [\text{H}^+][\text{HA}^-]/[\text{H}_2\text{A}]$ , solve $[\text{H}_2\text{A}] = [\text{H}^+][\text{HA}^-]/K_{a1}$ . From $K_{a2} = [\text{H}^+][\text{A}^{2-}]/[\text{HA}^-]$ , solve $[\text{A}^{2-}] = K_{a2}[\text{HA}^-]/[\text{H}^+]$ . Also $[\text{OH}^-] = K_w/[\text{H}^+]$ . Substitute all three into the proton condition:

\frac{[\text{H}^+][\text{HA}^-]}{K_{a1}} + [\text{H}^+] = \frac{K_{a2}[\text{HA}^-]}{[\text{H}^+]} + \frac{K_w}{[\text{H}^+]}

Step 4 — Clear denominators and solve for [H⁺]². Multiply both sides by $[\text{H}^+]K_{a1}$ :

[\text{H}^+]^2[\text{HA}^-] + [\text{H}^+]^2 K_{a1} = K_{a1}K_{a2}[\text{HA}^-] + K_{a1}K_w

Factor the left side:

[\text{H}^+]^2\left([\text{HA}^-]+K_{a1}\right) = K_{a1}K_{a2}[\text{HA}^-]+K_{a1}K_w

[\text{H}^+] = \sqrt{\frac{K_{a1}K_{a2}[\text{HA}^-]+K_{a1}K_w}{[\text{HA}^-]+K_{a1}}}

This is exact (given $[\text{HA}^-]$ is the true equilibrium concentration, essentially always well approximated by $C$ ).

Step 5 — Apply the two simplifications. If $[\text{HA}^-] \approx C \gg K_{a1}$ (true whenever $K_{a1}$ is a fairly weak acid and $C$ is not vanishingly dilute), the denominator $\to C$ . If additionally $C K_{a2} \gg K_w$ (true unless $K_{a2}$ is extremely small or $C$ extremely dilute), the $K_{a1}K_w$ term in the numerator is negligible next to $K_{a1}K_{a2}C$ . Then:

[\text{H}^+] \approx \sqrt{\frac{K_{a1}K_{a2}C}{C}} = \sqrt{K_{a1}K_{a2}}

Taking $-\log_{10}$ of both sides gives the final working formula:

\text{pH} \approx \frac{\text{p}K_{a1}+\text{p}K_{a2}}{2}

Go deeper · Derivation B · 1 minCharge-competition check: why bicarbonate is basic, not acidic

Derivation BCharge-competition check: why bicarbonate is basic, not acidic

Goal. Show independently, via the acid/base strength comparison, that HA⁻ is net basic exactly when $K_b = K_w/K_{a1}$ exceeds $K_{a2}$ — and connect this to the sign implied by the sqrt formula.

Step 1 — Frame the competition. HA⁻ reacts as an acid with intrinsic strength $K_{a2}$ (producing H⁺, pushing pH down) and as a base with intrinsic strength $K_b = K_w/K_{a1}$ (producing OH⁻, pushing pH up). Since both reactions draw from the same pool of HA⁻ and operate independently to leading order, the resulting pH is set by whichever process dominates — quantified precisely by comparing $K_{a2}$ to $K_b$ .

Step 2 — Relate this to pKa1 and pKa2 directly. The condition for a net-basic solution is $K_b > K_{a2}$ , i.e.

\frac{K_w}{K_{a1}} > K_{a2} \quad\Longleftrightarrow\quad K_w > K_{a1}K_{a2}

Taking $-\log_{10}$ (which reverses the inequality):

\text{p}K_w < \text{p}K_{a1}+\text{p}K_{a2} \quad\Longleftrightarrow\quad 7.00 \times 2 < \text{p}K_{a1}+\text{p}K_{a2} \text{ (at 25°C, using p}K_w=14.00\text{)}

which simplifies to $\text{p}K_{a1}+\text{p}K_{a2} > 14.00$ .

Step 3 — Connect to the sqrt formula's prediction. From Step 5 of the first derivation, $\text{pH} = \tfrac12(\text{p}K_{a1}+\text{p}K_{a2})$ . The solution is basic (pH > 7) exactly when $\text{p}K_{a1}+\text{p}K_{a2} > 14.00$ — precisely the same condition derived independently from the $K_b$ vs. $K_{a2}$ competition in Step 2. The two routes agree perfectly, confirming the formula's internal consistency.

Step 4 — Apply numerically to carbonic acid. $\text{p}K_{a1} = -\log(4.45\times10^{-7}) = 6.351$ ; $\text{p}K_{a2} = -\log(4.69\times10^{-11}) = 10.329$ . Sum: $6.351+10.329 = 16.680$ , which exceeds 14.00 — confirming bicarbonate must be basic, and predicting $\text{pH} = 16.680/2 = 8.34$ , matching the well-known value for sodium bicarbonate solutions.

Worked examples

Example 1 — Foundational: sodium bicarbonate, checking the approximation's validity. Find the pH of 0.100 M NaHCO₃ at 25 °C, given $K_{a1}(\text{H}_2\text{CO}_3) = 4.45\times10^{-7}$ and $K_{a2}(\text{HCO}_3^-) = 4.69\times10^{-11}$ . First check the two conditions for the simplified formula. Condition 1: is $C \gg K_{a1}$ ? Compare $0.100$ to $4.45\times10^{-7}$ — yes, by six orders of magnitude, so $[\text{HA}^-] \approx C$ is excellent. Condition 2: is $C K_{a2} \gg K_w$ ? Compute $C K_{a2} = (0.100)(4.69\times10^{-11}) = 4.69\times10^{-12}$ , and compare to $K_w = 1.0\times10^{-14}$ — yes, by nearly three orders of magnitude, so the simplified formula applies cleanly.

\text{pH} = \frac{\text{p}K_{a1}+\text{p}K_{a2}}{2} = \frac{6.351 + 10.329}{2} = \frac{16.680}{2} = 8.340

To three significant figures, pH = 8.34, independent of the exact value of $C$ as long as $C \geq$ roughly $10^{-3}$ M (checked below). This matches the textbook fact that baking soda solutions sit reliably near pH 8.3.

Example 2 — Applied: dihydrogen phosphate buffer component, using the exact formula. A buffer contains 0.0500 M NaH₂PO₄ (treating H₂PO₄⁻ as the amphiprotic species, flanked by $K_{a1}$ of H₃PO₄ and $K_{a2}$ of H₂PO₄⁻). Given $K_{a1} = 7.11\times10^{-3}$ , $K_{a2} = 6.32\times10^{-8}$ , find the pH — and check whether the simplified formula is trustworthy here, since $K_{a1}$ is unusually large. Check condition 1: is $C \gg K_{a1}$ ? Here $C = 0.0500$ and $K_{a1} = 7.11\times10^{-3}$ — the ratio is only about 7:1, not overwhelmingly large. This is a warning sign; the simplified formula may be inaccurate. Use the exact expression instead, with $[\text{HA}^-] \approx C$ as a first pass (self-consistent since HA⁻ consumption is what we are testing):

[\text{H}^+] = \sqrt{\frac{K_{a1}K_{a2}C + K_{a1}K_w}{C+K_{a1}}}

Numerator: $K_{a1}K_{a2}C = (7.11\times10^{-3})(6.32\times10^{-8})(0.0500) = 2.247\times10^{-11}$ ; the $K_{a1}K_w$ term is $(7.11\times10^{-3})(1.0\times10^{-14}) = 7.11\times10^{-17}$ , utterly negligible in comparison. Numerator $\approx 2.247\times10^{-11}$ . Denominator: $C + K_{a1} = 0.0500 + 0.00711 = 0.05711$ .

[\text{H}^+] = \sqrt{\frac{2.247\times10^{-11}}{0.05711}} = \sqrt{3.935\times10^{-10}} = 1.984\times10^{-5}\ \text{M}

\text{pH} = -\log(1.984\times10^{-5}) = 4.70

Compare to the naive simplified formula: $\text{p}K_{a1}=2.148$ , $\text{p}K_{a2}=7.199$ , giving $\text{pH} = (2.148+7.199)/2 = 4.67$ . The exact and approximate answers differ by 0.03 pH units here — small but non-negligible, illustrating exactly why the denominator correction $(C+K_{a1})$ rather than bare $C$ matters when $K_{a1}$ is only moderately smaller than $C$ . pH = 4.70 is the trustworthy answer.

Example 3 — Exam/competition level: dilute amphiprotic solution where the approximation fails outright. A student prepares $1.00\times10^{-4}$ M sodium hydrogen phthalate, NaHP, where phthalic acid has $K_{a1} = 1.12\times10^{-3}$ and $K_{a2} = 3.91\times10^{-6}$ . Compute the pH exactly, and quantify the error the naive $\sqrt{K_{a1}K_{a2}}$ shortcut would introduce. Check condition 1: is $C \gg K_{a1}$ ? $C = 1.00\times10^{-4}$ , $K_{a1}=1.12\times10^{-3}$ — here $C$ is actually smaller than $K_{a1}$ , a clear violation. The simplified formula is expected to fail badly. Use the exact formula, being careful that at this dilution $[\text{HA}^-]$ may itself deviate somewhat from $C$ ; as a good working approximation for a first solve, take $[\text{HA}^-]\approx C$ and verify self-consistency afterward. Numerator: $K_{a1}K_{a2}C = (1.12\times10^{-3})(3.91\times10^{-6})(1.00\times10^{-4}) = 4.379\times10^{-13}$ . $K_{a1}K_w$ term: $(1.12\times10^{-3})(1.0\times10^{-14}) = 1.12\times10^{-17}$ — negligible next to $4.379\times10^{-13}$ . Numerator $\approx 4.379\times10^{-13}$ . Denominator: $C + K_{a1} = 1.00\times10^{-4} + 1.12\times10^{-3} = 1.220\times10^{-3}$ .

[\text{H}^+] = \sqrt{\frac{4.379\times10^{-13}}{1.220\times10^{-3}}} = \sqrt{3.589\times10^{-10}} = 1.895\times10^{-5}\ \text{M}

\text{pH} = -\log(1.895\times10^{-5}) = 4.72

Now compare to the naive shortcut: $\text{p}K_{a1} = -\log(1.12\times10^{-3}) = 2.951$ ; $\text{p}K_{a2} = -\log(3.91\times10^{-6}) = 5.408$ ; naive pH $= (2.951+5.408)/2 = 4.18$ . The naive shortcut gives 4.18, while the exact calculation gives 4.72 — a difference of over half a pH unit, a serious error. The reason: at this dilution the amount of HA⁻ converting back to H₂A (governed by the relatively large $K_{a1}$ , meaning HA⁻ is a comparatively strong base here) is no longer negligible relative to the total supply, and the denominator correction $(C+K_{a1})$ dominates over bare $C$ . This example demonstrates concretely why blind application of $\sqrt{K_{a1}K_{a2}}$ without checking $C \gg K_{a1}$ is a serious analytical error, particularly relevant near equivalence points of titrations run at low concentration.

Limiting cases, scope & common traps

Regime	Condition	Behavior of pH formula
Standard case (most textbook salts)	$C \gg K_{a1}$ and $CK_{a2}\gg K_w$	$\text{pH}\approx\tfrac12(\text{p}K_{a1}+\text{p}K_{a2})$ , independent of $C$
Moderately dilute / $K_{a1}$ not small	$C$ only somewhat $> K_{a1}$	Must use $[\text{H}^+]=\sqrt{(K_{a1}K_{a2}C+K_{a1}K_w)/(C+K_{a1})}$
Very dilute solution	$C \lesssim K_{a1}$	Simplified formula fails badly (Example 3); full exact treatment needed
Very weak second dissociation	$CK_{a2} \not\gg K_w$	$K_w$ term in numerator can no longer be dropped; water autoionization competes
$K_{a1}\approx K_{a2}$ (rare, unusual acid)	poor separation of steps	Simple diprotic model itself questionable; overlapping equilibria complicate mass balance
As $C\to\infty$	—	pH → exactly $\tfrac12(\text{p}K_{a1}+\text{p}K_{a2})$ , the true asymptotic limit
As $C\to0$	—	pH → 7.00 (pure water dominates; HA⁻ too dilute to buffer anything)

The chemist behind it

The mathematics of amphiprotic equilibrium is inseparable from the broader Brønsted–Lowry framework, but its quantitative treatment in analytical chemistry owes much to Niels Bjerrum, the Danish chemist whose systematic studies of stepwise dissociation constants in polybasic acids and complex ions established the conventions ( $K_{a1}, K_{a2}, \ldots$ ) still used today. Bjerrum's insight — that successive ionization constants of a polyprotic acid could be measured, tabulated, and combined predictively — is exactly the machinery this article exploits. Before Bjerrum's systematic approach, the idea that a single ion like HCO₃⁻ could be rigorously treated as sitting "between" two independent equilibria, each with its own measurable constant, was not standard practice.

Origins · 1 minNiels Bjerrum and how this was discovered

Niels Bjerrum

1879–1958 · Danish physical chemist, University of Copenhagen

Bjerrum trained under Svante Arrhenius and Wilhelm Ostwald, absorbing the ionic theory of solutions at its source, then returned to Copenhagen to build a research program on stepwise equilibria — in polyprotic acids, in metal-ammine complex formation, and in the thermodynamics of strong electrolytes. His 1923 work on complex-ion stability constants introduced formalism directly analogous to successive $K_a$ values, and his broader emphasis on rigorous mass-action treatment of multi-step equilibria shaped how generations of analytical chemists — including those who later formalized amphiprotic pH theory — approached polyprotic and polydentate systems quantitatively.

History

- 1887 — Arrhenius proposes the ionic dissociation theory, providing the conceptual basis for equilibrium constants in solution.

- 1909 — Sørensen introduces the pH scale, giving polyprotic and amphiprotic chemistry a compact language for reporting acidity.

- 1921 — Bjerrum publishes systematic studies establishing stepwise dissociation-constant notation for polybasic acids.

- 1923 — Brønsted and Lowry independently formulate the proton-donor/acceptor definition of acids and bases, formally legitimizing the idea of a species being simultaneously acid and base.

- 1938–1950s — Physical and analytical chemistry textbooks (Kolthoff, Laitinen, later Skoog) codify the

\sqrt{K_{a1}K_{a2}}

amphiprotic pH formula as standard undergraduate material.

Why is the pH of an amphiprotic salt almost independent of concentration?

Because in the standard regime ( $C \gg K_{a1}$ and $CK_{a2}\gg K_w$ ), the concentration $C$ appears identically in both the numerator and denominator of the exact expression for $[\text{H}^+]$ and cancels exactly, leaving $[\text{H}^+]=\sqrt{K_{a1}K_{a2}}$ , a ratio of pure equilibrium constants with no concentration dependence at all. Physically, diluting the solution reduces both the "acid pathway" (via A²⁻ production) and the "base pathway" (via H₂A production) by the same proportion, so their competitive balance — and hence the resulting pH — stays fixed. This cancellation is not automatic at very low concentration, where the correction terms $K_{a1}$ (in the denominator) and $K_w$ (in the numerator) become comparable to $C$ and its products, breaking the cancellation.

How is this different from a normal Henderson–Hasselbalch buffer calculation?

A Henderson–Hasselbalch buffer involves two independently chosen concentrations of a conjugate acid–base pair, and its pH depends on their ratio: $\text{pH}=\text{p}K_a+\log([\text{base}]/[\text{acid}])$ . A pure amphiprotic salt solution has only one formal species (HA⁻) added to the beaker; the small quantities of H₂A and A²⁻ present are generated internally by HA⁻'s own dual reactivity and turn out to be equal to each other, not freely adjustable. There is no external ratio to set — the pH is entirely determined by $K_{a1}$ and $K_{a2}$ themselves, which is precisely why the result reduces to a fixed number rather than a log-ratio expression.

Does the sqrt(Ka1 Ka2) formula apply to triprotic acids, like phosphoric acid?

Yes, with care about which two constants flank the species in question. For H₂PO₄⁻, the flanking constants are $K_{a1}$ (of H₃PO₄) and $K_{a2}$ (of H₂PO₄⁻), giving $\text{pH}\approx\tfrac12(2.15+7.20)=4.68$ . For HPO₄²⁻, the flanking constants are $K_{a2}$ and $K_{a3}$ (of HPO₄²⁻), giving $\text{pH}\approx\tfrac12(7.20+12.35)=9.78$ . Each intermediate amphiprotic species in a triprotic (or higher) system uses its own local pair of neighboring dissociation constants — the formula generalizes species-by-species rather than needing all constants simultaneously.

Why does the exact formula include a Kw term at all if water's autoionization seems irrelevant here?

The $K_w$ term accounts for the possibility that hydroxide production from plain water autoionization becomes comparable to hydroxide-equivalent production from HA⁻ acting as a base. This matters specifically when $K_{a2}$ is extremely small (HA⁻ is a very weak acid) or when $C$ is very dilute, so that the "A²⁻ pathway" numerator term $K_{a1}K_{a2}C$ shrinks enough that the $K_{a1}K_w$ term is no longer negligible by comparison. In most everyday amphiprotic salts at moderate concentration this term is many orders of magnitude smaller and safely dropped, but rigorous treatment always includes it for completeness and to correctly handle edge cases.

How does this connect to the equivalence points of a diprotic titration curve?

When titrating H₂A with strong base, the first equivalence point (one equivalent of base added) corresponds to a solution that is, to a very good approximation, pure NaHA — exactly the amphiprotic species treated here. The pH at that inflection point is therefore given by the same $\sqrt{K_{a1}K_{a2}}$ formula (subject to the same validity checks), which is why titration curves of diprotic acids show a a predictable, calculable inflection pH at the first equivalence point rather than requiring a separate ad hoc calculation. This is one of the most direct practical payoffs of the amphiprotic pH theory in real titrimetric analysis.

Can amphiprotic behavior occur in species that are not simple polyprotic acid anions?

Yes. Zwitterionic amino acids are a classic example: at the isoelectric point, the neutral zwitterion form (analogous to HA⁻) can either donate a proton from its ammonium-like group or accept one at its carboxylate-like group, and its pH-independent isoelectric point pI is calculated with the identical $\tfrac12(\text{p}K_{a1}+\text{p}K_{a2})$ formula, using the amino acid's two relevant pKa values. Many biological buffer components (bicarbonate, phosphate, several zwitterionic Good's buffers) share this same amphiprotic mathematical structure despite looking chemically quite different.

What is the single most important check before applying the simplified formula in an exam problem?

Verify numerically that $C$ is at least roughly two orders of magnitude greater than $K_{a1}$ , and that $C\cdot K_{a2}$ is at least roughly two orders of magnitude greater than $K_w$ . If either check fails — as it does for phthalate at $10^{-4}$ M or for phosphate's $H_2PO_4^-$ where $K_{a1}$ is unusually large — fall back to the exact quadratic expression $[\text{H}^+]=\sqrt{(K_{a1}K_{a2}C+K_{a1}K_w)/(C+K_{a1})}$ rather than trusting the shortcut, since errors of several tenths of a pH unit are common when the checks are skipped.