Selecting Indicators for Multi-Equivalence Titrations

Double-indicator carbonate/bicarbonate analysis remains a standard method in water-quality laboratories for reporting alkalinity as separate carbonate and bicarbonate contributions, directly using the phenolphthalein/methyl orange sequence worked through above. Pharmaceutical quality control for polyprotic active ingredients — citric acid formulations, amino-acid infusions, phosphate-buffered injectables — relies on the same equivalence-point-matching logic to validate potentiometric or visual titration methods against pharmacopeial specifications, often cross-checking the indicator endpoint against a pH-meter-recorded curve exactly as described under the companion topic of endpoint-versus-equivalence-point error. In biochemistry, the isoelectric-point calculation for diprotic and triprotic amino acids (glycine, histidine, glutamic acid) uses the identical "average of flanking $\mathrm{p}K_a$'s" formula derived here, applied to electrophoretic focusing rather than titrimetric analysis. Industrial process control for phosphoric acid production and fertilizer manufacture uses automated potentiometric titration to track the sharp second equivalence point of H₃PO₄ as a real-time purity and concentration check, precisely because the first and third points are known in advance to be unreliable visual (or even automated first-derivative) endpoints. Environmental and soil chemistry laboratories apply the same $\Delta\mathrm{p}K_a$ reasoning when selecting indicators or potentiometric endpoints for humic-acid and polyprotic organic-acid characterization, where overlapping, closely spaced $K_a$ values are the norm rather than the exception — a direct link to why Polyprotic Acid Titration Curves for natural samples often show broad, featureless buffer regions rather than crisp steps, and why analysts increasingly favor Gran-plot or first-derivative potentiometric methods over visual indicators in exactly the "$\Delta\mathrm{p}K_a$ too small" regime flagged throughout this article.

Ask the tutor: How do I decide, without doing the full derivation, whether a given pair of $K_a$ values is "close enough" to merge two equivalence points in practice?

In a monoprotic strong-acid/strong-base titration the curve has one dramatic vertical segment, and essentially any indicator with $\mathrm{p}K_{\mathrm{In}}$ between about 4 and 10 works, because the jump spans nearly the whole pH scale over a few hundredths of a milliliter (from Acid-Base Indicators: Theory and Transition Range). Polyprotic and multicomponent systems break this convenience. Each dissociation step behaves, locally, like a weak-acid/weak-base titration superimposed on the others, and the curve is really a sequence of buffer regions separated by inflections — some sharp, some so gentle they barely deserve the name "equivalence point" on a graph a human eye could read. The key quantity is the local slope of the titration curve at each stoichiometric point,

which depends on how much buffering capacity remains from the previous dissociation step and how much begins from the next one. If $K_{a1}$ and $K_{a2}$ are close together (say, within a factor of 100), the solution never runs out of the first conjugate pair before the second starts consuming base, the two buffer regions blend, and the pH rises almost linearly through both stoichiometric points rather than jumping at either one. If $K_{a1}/K_{a2} \geq 10^4$–$10^5$, the first acid is essentially fully consumed, and quantitatively "gone," before the second proton begins reacting in earnest; a genuine plateau (buffer region) followed by a genuine near-vertical jump appears at each point. A second, independent constraint is dilution. Even with well-separated $K_a$ values, if total analyte concentration is low (millimolar range), the height of each jump shrinks — fewer moles of strong base are needed to move the pH the same number of units when there is less buffer capacity to overcome, but also the absolute curvature (the second derivative) is smaller in magnitude, so the visual "snap" of the color change is less crisp. Indicator selection must therefore always be checked against both the $K_a$ spacing (chemistry) and the concentration regime (dilution) actually used.

The steepness at each candidate equivalence point can be read directly off the curve: the middle jump (second equivalence point of H₃PO₄) rises through roughly four pH units within a fraction of a milliliter, while the first "equivalence point" is really just a shallow shoulder buried inside overlapping buffer regions of H₂PO₄⁻/H₃PO₄ and H₂PO₄⁻/HPO₄²⁻.

For a diprotic acid H₂A with successive acid constants $K_{a1}$ and $K_{a2}$, the pH at the first equivalence point (formation of the amphiprotic species HA⁻) is well approximated, when $C \gg K_{a1}$ and $K_{a1} \gg K_{a2}$, by

where:

• $K_{a1}, K_{a2}$ = the first and second acid dissociation constants of the diprotic (or polyprotic) acid,
• $\mathrm{p}K_{a1} = -\log_{10}K_{a1}$, similarly for $\mathrm{p}K_{a2}$,
• the approximation assumes the amphiprotic species HA⁻ dominates and self-ionization/dilution corrections are negligible.

Reading: the equivalence-point pH of an intermediate (amphiprotic) species is, to good approximation, the geometric mean of the two flanking acidity constants — independent of concentration, which is itself a useful diagnostic (if measured pH shifts strongly with dilution, the amphiprotic approximation is breaking down). For an $n$-protic acid, the same rule extends to each internal equivalence point:

where:

• $\mathrm{pH}_k$ = pH at the $k$-th equivalence point (after $k$ of $n$ protons removed),
• $\mathrm{p}K_{a,k}, \mathrm{p}K_{a,k+1}$ = the acidity constants immediately below and above that point,
• valid when successive equivalence points are well separated (the "independent monoprotic" limit).

The sharpness criterion — whether a given equivalence point is visually detectable — compares the separation of the pH at the equivalence point from its neighboring buffer regions to the indicator's transition width of roughly 2 pH units:

where:

• $\Delta\mathrm{p}K_a$ = spacing between successive acidity constants around the equivalence point in question,
• the threshold of $\gtrsim 4$ is empirical, derived from requiring the pH jump within $\pm 0.1\%$ of the equivalence volume to exceed about 2 pH units at typical (0.01–0.1 M) titrant/analyte concentrations.

Reading: this single inequality is the entire selection algorithm in miniature. Compute $\Delta\mathrm{p}K_a$ for every pair of adjacent constants (including the "virtual" constants $K_w$ and dilution limits at the very first and last points); wherever it clears about 4, that equivalence point is fair game for an indicator; wherever it does not, no indicator, however cleverly chosen, will show a clean break.

Example 1 — Foundational: locating and grading the equivalence points of oxalic acid, H₂C₂O₄. Oxalic acid has $K_{a1} = 5.9\times10^{-2}$ ($\mathrm{p}K_{a1} = 1.23$) and $K_{a2} = 6.4\times10^{-5}$ ($\mathrm{p}K_{a2} = 4.19$). Titrate 25.00 mL of 0.1000 M H₂C₂O₄ with 0.1000 M NaOH. Step 1: is the internal equivalence point sharp?

This is well below the $\gtrsim 4$ threshold — the two dissociation steps overlap substantially, and the first equivalence point (formation of HC₂O₄⁻) will show only a gentle inflection, not a sharp break. No single indicator reliably flags this point; in practice oxalic acid is titrated straight through to the second equivalence point. Step 2: locate and grade the second (final) equivalence point. Total volume of NaOH to reach the second equivalence point is twice the moles of acid:

At this point the solution contains C₂O₄²⁻ at concentration

This is a simple weak-base hydrolysis: $\mathrm{C_2O_4^{2-}} + \mathrm{H_2O} \rightleftharpoons \mathrm{HC_2O_4^-} + \mathrm{OH^-}$, with

Then

This point is bounded below by $K_{a2}$ ($\mathrm{p}K_{a2}=4.19$) and above by $K_w$; the "$\Delta\mathrm{p}K_a$" here is effectively $14 - 4.19 = 9.8$, easily sharp. Phenolphthalein ($\mathrm{p}K_{\mathrm{In}} \approx 9.4$, visible range 8.2–10.0) straddles pH 8.36 well and is the standard indicator for oxalic acid titrated to its total-neutralization endpoint. As a limiting-case check: if $K_{a2}$ were instead comparable to $K_{a1}$ (equal-strength diprotic acid, $\Delta\mathrm{p}K_a \to 0$), no internal inflection would exist at all, and the entire titration would resemble a single monoprotic-acid curve with total $n=2$ equivalents — consistent with the formula collapsing (there being no meaningful intermediate species to define $\mathrm{pH}_1$).

Example 2 — Applied: indicator selection across the three equivalence points of phosphoric acid. H₃PO₄ has $K_{a1} = 7.5\times10^{-3}$ ($\mathrm{p}K_{a1}=2.12$), $K_{a2} = 6.2\times10^{-8}$ ($\mathrm{p}K_{a2}=7.21$), $K_{a3} = 4.8\times10^{-13}$ ($\mathrm{p}K_{a3}=12.32$). Titrate 25.00 mL of 0.1000 M H₃PO₄ with 0.1000 M NaOH ($n_{\text{acid}} = 2.500\times10^{-3}$ mol). Gap 1 (first equivalence point, H₂PO₄⁻ forms):

This clears the $\gtrsim 4$ threshold — moderately sharp, though not dramatic. Estimated pH:

An indicator with $\mathrm{p}K_{\mathrm{In}} \approx 4.7$, such as methyl orange ($\mathrm{p}K_{\mathrm{In}}\approx3.7$, range 3.1–4.4) or bromocresol green ($\mathrm{p}K_{\mathrm{In}}\approx 4.7$, range 3.8–5.4), is a plausible match on paper. In practice, though, this jump — while formally "sharp" by the $\Delta\mathrm{p}K_a$ criterion — spans barely 1.5 pH units in the useful titration-error window and is often skipped in routine analysis in favor of the far more distinct second point; textbooks typically flag it as "usable with care," consistent with $\Delta\mathrm{p}K_a = 5.09$ sitting in the 4–6 "marginal" band identified in the sharpness derivation. Gap 2 (second equivalence point, HPO₄²⁻ forms):

Volume to reach this point:

Phenolphthalein (8.2–10.0) is an excellent match — this is the textbook-standard indicator for phosphoric acid titrated to its second, most usable equivalence point, and matches the pH curve sketched in the figure above. Gap 3 (third equivalence point, PO₄³⁻ forms): bounded above only by $K_w$, giving an effective separation of $14.00 - 12.32 = 1.68$ — far below threshold, and moreover the required pH (roughly 12.7–13, from hydrolysis of PO₄³⁻) approaches the region where water's own autoionization and the practical difficulty of maintaining/measuring such high pH accurately erode any sharp break. No indicator is used for this point; it is simply not accessible as a clean visual endpoint in aqueous titration, consistent with the earlier claim that phosphoric acid effectively offers only one indicator-grade equivalence point in routine practice.

Example 3 — Exam/competition level: choosing indicators for a Na₂CO₃/NaHCO₃ mixture analyzed by sequential double-indicator titration. A 25.00 mL sample contains both Na₂CO�3 and NaHCO₃, titrated with 0.1000 M HCl. Relevant constants for carbonic acid: $K_{a1} = 4.3\times10^{-7}$ ($\mathrm{p}K_{a1}=6.37$, for H₂CO₃), $K_{a2} = 4.8\times10^{-11}$ ($\mathrm{p}K_{a2}=10.32$). Titrating with acid, the reactions proceed as

Locating the first equivalence point pH: this is the amphiprotic species HCO₃⁻ forming from CO₃²⁻, so

just under the sharp-break threshold — this is why the first "phenolphthalein endpoint" in carbonate analysis is famously described as soft: the color fade is gradual rather than a snap transition, exactly consistent with $\Delta\mathrm{p}K_a \approx 4$ sitting right at the marginal boundary. Phenolphthalein (8.2–10.0) is nonetheless the conventional choice, chosen deliberately on the high side of pH 8.35 because the fading endpoint is easier to catch by eye than a sudden appearance would be at this borderline sharpness. Locating the second equivalence point pH: here the product is H₂CO₃, which decomposes toward dissolved CO₂; the relevant equilibrium is simple weak-acid ionization of H₂CO₃ at whatever concentration remains after two decomposition steps and substantial dilution — practically, the measured endpoint pH for the bicarbonate/CO₂ system falls near 3.8–4.0 (lower than a naive $K_{a1}$-only estimate because of the CO₂ escaping and the compounded dilution). Methyl orange ($\mathrm{p}K_{\mathrm{In}}\approx3.7$, range 3.1–4.4) is the standard choice, and the transition (yellow → orange-red) is sharp because the effective boundary here is set by $\mathrm{p}K_{a1}=6.37$ against essentially "excess strong acid," giving an effective $\Delta\mathrm{p}K_a$ against the next relevant reference (the strong-acid limit) that is large. Quantitative volumes, to tie the two endpoints together: suppose titration to the phenolphthalein fade requires $V_1 = 15.00$ mL and titration onward to the methyl orange color change requires an additional $V_2 = 22.50$ mL (total 37.50 mL). Because the first equivalence point converts only the carbonate's first proton equivalent while the second converts both carbonate's second proton and all the original bicarbonate,

Converting to concentrations in the original 25.00 mL:

This is the classic Winkler/double-indicator carbonate-bicarbonate analysis, and it depends entirely on recognizing that two different indicators, each matched to a different equivalence point of the same diprotic system, are required in sequence — a direct extension of the single-acid logic in Examples 1 and 2 to a mixture.

The systematic language of stepwise dissociation constants and their use in predicting titration-curve shape owes much to the Danish chemist Søren Sørensen, who introduced the pH scale in 1909 while working on protein and enzyme buffer systems at the Carlsberg Laboratory, and to the broader early-20th-century development of quantitative acid–base equilibrium theory that followed Brønsted and Lowry's proton-transfer definition of acids and bases. The specific practice of choosing indicators against computed equivalence-point pH values — rather than by trial and error — became standard laboratory doctrine through the mid-20th-century quantitative analytical chemistry texts (Kolthoff, Sandell, and later Skoog and West), which tabulated indicator transition ranges alongside systematic $pK_a$ compilations precisely so students could perform the kind of matching worked through in this article.