Buffer Capacity and Buffer Selection

Clinical blood gas analysis remains the most direct descendant of Van Slyke's own work: arterial blood pH is normally held at 7.40 ± 0.05 by the combined bicarbonate (p$K_{a1}'\approx6.1$ under physiological CO₂ partial pressure), phosphate, and protein buffer systems, and a physician diagnosing acidosis or alkalosis is implicitly reasoning about which buffer reservoir has been depleted — a direct clinical application of buffer capacity falling off away from p$K_a$. In pharmaceutical formulation, injectable drugs and eye drops are buffered not only to a target pH for stability and solubility, but with buffer capacity engineered to be just strong enough to resist pH drift from the container, the drug's own acid/base groups, or CO₂ absorption from air — but not so strong that injecting the formulation shocks the patient's local physiological buffering (a poorly buffered or over-buffered injectable can both cause tissue irritation). In HPLC and capillary electrophoresis, mobile-phase and running buffers are chosen with p$K_a$ near the operating pH specifically so that small variations in sample pH, injection volume, or dissolved CO₂ do not shift retention times or electroosmotic flow — a direct industrial-analytical use of the p$K_a$-matching principle from Example 2. In fermentation and bioprocessing, buffer capacity calculations size the base-addition (or acid-addition) capacity of a bioreactor's pH-control system: knowing $\beta$ at the target pH tells engineers how much titrant pump capacity and buffer salt inventory are needed to hold pH steady against the acid or base produced by cell metabolism over a multi-day fermentation run, exactly as computed in Example 2's kinase assay but scaled to thousands of liters. In environmental and geochemistry, the "acid neutralizing capacity" (ANC) of natural waters and soils is a direct macroscopic analog of $\beta$, integrated over the carbonate and mineral buffer systems present, and used to predict a lake's or soil's vulnerability to acid rain — low-ANC (low buffer capacity) waters show catastrophic pH crashes from the same acid load that a high-ANC system absorbs without effect, mirroring exactly the plasma-versus-water comparison that opened this article.

Ask the tutor: How does buffer capacity change when two overlapping buffer systems (like phosphate's pKa2 and pKa3, or bicarbonate and phosphate together) are mixed in the same solution?

Picture titrating a solution of a weak acid HA with strong base, plotting pH on the vertical axis against volume of titrant (or equivalently, moles of strong base added per liter, $C_b$) on the horizontal axis. Far from the equivalence point, the curve is nearly flat — a small addition of base produces almost no pH change. Near the equivalence point, the curve turns almost vertical — the same small addition of base swings the pH by several units. Buffer capacity is the reciprocal of the local slope of this curve: where the titration curve is flat, capacity is high; where it is steep, capacity is low. Chemically, the flat region exists because added OH⁻ is consumed by the reaction HA + OH⁻ → A⁻ + H₂O, converting acid to its conjugate base without much accumulation of free OH⁻ or consumption of H⁃buffered proton reservoir. The system can "absorb" the perturbation because there is a reservoir of both HA (to donate a proton when acid is added) and A⁻ (to accept a proton when base is added). When either reservoir runs low — near pH = p$K_a$ ± 2 or beyond — the system loses its absorptive power and behaves like unbuffered water. The curvature of the titration curve is literally the second derivative structure that buffer capacity is built to quantify: $\beta$ is defined as $dC_b/d(\mathrm{pH})$, the amount of strong base (mol/L) required per unit pH increase. Where the titration curve's slope $d(\mathrm{pH})/dC_b$ is small, $\beta$ is large, and vice versa — they are reciprocal measures of the same flatness.

The buffer capacity curve, plotted against pH, is a bell shape peaking exactly at pH = p$K_a$ — the point of half-neutralization, where $[\mathrm{HA}] = [\mathrm{A^-}]$. This symmetry is not approximate; it follows directly from the algebra of fractional composition, which reappears with more protonation states in Fractional Composition of Diprotic Systems.

The third piece of the physical picture is the water background. At extreme pH (below 2 or above 12), buffering by the weak acid/base pair becomes negligible, but water autoionization itself starts to contribute non-trivial capacity, since strongly acidic or basic solutions resist pH change partly because $[\mathrm{H^+}]$ or $[\mathrm{OH^-}]$ is itself large and any titrant addition is a small fractional perturbation. This term, $2.303([\mathrm{H^+}]+[\mathrm{OH^-}])$, is usually negligible near neutral pH but dominates outside pH 3–11.

The formal definition of buffer capacity, due to Van Slyke (1922), is:

where:

• $C_b$ = moles per liter of strong base added (cumulative, from an initial reference state),
• $C_a$ = moles per liter of strong acid added,
• pH = $-\log_{10}[\mathrm{H^+}]$,
• $\beta$ has units of mol L⁻¹ per pH unit — always positive, since adding base always raises pH.

Reading: $\beta$ is a local slope, evaluated at a particular pH, not a single number for a whole solution. It answers: "at this pH, how many moles of strong base per liter would I need to add to raise the pH by exactly one unit," in the limit of an infinitesimal addition (the finite-addition version is only approximately equal to $\beta$ unless the addition is small). For a monoprotic weak acid HA of total analytical concentration $C = [\mathrm{HA}] + [\mathrm{A^-}]$ (constant, since dilution during titrant addition is assumed negligible or corrected for), the closed form is:

where:

• $\alpha_{HA} = \dfrac{[\mathrm{H^+}]}{[\mathrm{H^+}]+K_a}$ and $\alpha_{A^-} = \dfrac{K_a}{[\mathrm{H^+}]+K_a}$ are the fractions of total buffer present as acid and conjugate base,
• $2.303 = \ln(10)$, converting natural-log derivatives to base-10 pH derivatives,
• the $[\mathrm{H^+}]$ and $[\mathrm{OH^-}]$ terms are the "water contribution," significant only outside roughly pH 3–11.

Reading: the buffer term $C\,\alpha_{HA}\alpha_{A^-}$ is a product of two fractions that each range from 0 to 1 and sum to 1 — a classic "$x(1-x)$" shape, maximized when $\alpha_{HA}=\alpha_{A^-}=1/2$, i.e., when pH = p$K_a$. This immediately explains the bell curve in the figure above. At the maximum (pH = p$K_a$, ignoring the small water term):

Reading: the peak buffer capacity is just over half the total buffer concentration — a convenient rule of thumb. A 0.10 M acetic acid/acetate buffer at pH 4.76 can absorb about 0.0576 mol of strong base per liter for each unit of pH rise, right at its peak.

Example 1 (Foundational): Buffer capacity of an acetate buffer at three pH values. Consider an acetic acid/acetate buffer, total concentration $C = 0.200$ M, p$K_a = 4.76$ ($K_a = 1.74\times10^{-5}$). Compute $\beta$ at pH 3.76, pH 4.76, and pH 5.76, and compare to the p$K_a\pm1$ rule. At pH 4.76 (= p$K_a$): $[\mathrm{H^+}] = 10^{-4.76} = 1.74\times10^{-5}$ M, equal to $K_a$, so $\alpha_{HA}=\alpha_{A^-}=0.500$.

The water term contributes only 0.03% here — completely negligible, as expected near neutral pH. This matches $\beta_{max}=0.576\times0.200 = 0.1152$ M/pH exactly, confirming the formula. At pH 3.76 (= p$K_a - 1$): $[\mathrm{H^+}] = 10^{-3.76} = 1.738\times10^{-4}$ M.

Ratio to maximum: $0.03848/0.1152 = 0.334$ — matching the predicted 33.1% (small discrepancy from the tiny water contribution at this pH, and rounding). At pH 5.76 (= p$K_a + 1$), by the symmetric calculation, $\alpha_{A^-}=0.9091$, $\alpha_{HA}=0.0909$, and the water term ($[\mathrm{OH^-}]$ now) is again negligible:

The slight asymmetry between 0.03848 and 0.03807 (about 1%) comes from the $[\mathrm{H^+}]$ vs. $[\mathrm{OH^-}]$ terms, entirely negligible at these pH values — confirming the bell curve is, for practical purposes, perfectly symmetric on the pH axis near neutral pH.

Limiting case check: as pH moves far from p$K_a$ (say pH = p$K_a + 3$), $\alpha_{A^-}\to 0.999$, $\alpha_{HA}\to0.001$, so $f\to0.000999$, and $\beta_{buffer}\to 2.303\times0.200\times0.000999 = 0.00046$ M/pH — buffering has essentially vanished, and the solution behaves like a solution of the conjugate base alone with little resistance to further pH change from that component (though the water/strong-base term may start to matter at very high pH).

Example 2 (Applied): Selecting a buffer for an enzyme assay. A biochemist runs a kinase assay that must be held at pH 7.40 and expects the reaction to release up to $0.0150$ mol/L of H⁺ over the course of the assay (from ATP hydrolysis byproducts) in a total reaction volume of 10.0 mL. The lab has two buffer stocks available: (a) acetate, p$K_a = 4.76$, and (b) HEPES, p$K_a = 7.48$. Determine which is suitable, and find the minimum total HEPES concentration $C$ needed to keep the pH drift under 0.05 units when the 0.0150 mol/L of acid is generated. Step 1 — Screen by p$K_a$. Target pH 7.40 is 2.64 units from acetate's p$K_a$ (4.76) — far outside the useful p$K_a\pm1$ window; acetate is disqualified regardless of concentration, since at pH 7.40 acetic acid is almost entirely (>99.9%) deprotonated and there is essentially no HA reservoir left to buffer against further base, nor is there meaningful buffering against the acid produced (since $\alpha_{HA}\alpha_{A^-}$ is minuscule). HEPES, with p$K_a = 7.48$, is only 0.08 units away — squarely in the optimal zone. Step 2 — Use the buffer capacity relation to size $C$. We want the pH to drop by no more than $\Delta\mathrm{pH} = 0.05$ when $\Delta C_a = 0.0150$ mol/L of acid is added (equivalent to $-\Delta C_b = 0.0150$). Approximate $\beta \approx |\Delta C_a|/|\Delta \mathrm{pH}|$ for small changes (valid since 0.05 pH units is a small perturbation relative to the curve's local curvature):

Step 3 — Evaluate $\beta$ at pH 7.40 as a function of $C$, using HEPES's p$K_a=7.48$. First find $\alpha_{HA}$ and $\alpha_{A^-}$ at pH 7.40: $[\mathrm{H^+}]=10^{-7.40}=3.981\times10^{-8}$ M; $K_a = 10^{-7.48}=3.311\times10^{-8}$ M.

So $\beta = 2.303\,C\,(0.2479) = 0.5709\,C$ (water term negligible at pH 7.4). Step 4 — Solve for $C$.

This is unusually concentrated for a lab buffer — a signal to sanity-check the assumption. In practice, biochemists commonly build in a safety factor and would recognize that 0.0150 mol/L of acid released is a substantial load; a more typical target might allow $\Delta\mathrm{pH}=0.10$ instead of 0.05, halving the requirement to $C \approx 0.263$ M, closer to a realistic 0.20–0.30 M HEPES stock. This example shows explicitly how buffer capacity converts a chemical process specification (acid released) into a concrete concentration requirement — not a guess.

Example 3 (Exam/competition level): Buffer capacity of a diprotic-adjacent mixture and the "20 mM rule." A student proposes buffering a solution at pH 6.00 using a phosphate buffer built from NaH₂PO₄/Na₂HPO₄ (p$K_{a2} = 7.21$ for H₂PO₄⁻/HPO₄²⁻), at a modest total phosphate concentration of $C = 0.0500$ M. (a) Estimate $\beta$ at pH 6.00 using the monoprotic formula (treating the H₂PO₄⁻/HPO₄²⁻ couple in isolation, ignoring the more distant p$K_{a1}=2.15$ and p$K_{a3}=12.32$, which is justified since pH 6.00 is far from both). (b) Compare this to the capacity the same total concentration would provide if instead used at its optimal pH (7.21), and hence quantify the "penalty" of choosing a mismatched buffer. (c) Determine what total concentration $C'$ of phosphate at pH 6.00 would be needed to match the buffering power of 0.0200 M MES (p$K_a = 6.10$) at its near-optimal pH — illustrating why practitioners prefer a buffer with p$K_a$ close to target pH even if it means using a less familiar reagent. (a) Compute $\beta$ at pH 6.00 for phosphate, p$K_{a2}=7.21$.

(b) Compute $\beta$ for the same $C$ at pH = p$K_{a2}=7.21$ (optimal).

Penalty ratio:

So using phosphate 1.21 pH units off its p$K_a$ costs a factor of 4.6 in buffering power (only 21.9% of what the same total concentration could deliver at the optimal pH) — a substantial, quantifiable penalty consistent with the earlier derivation that predicted roughly 33% capacity at exactly 1 unit off and even less at 1.21 units off. (c) Compute $\beta$ for 0.0200 M MES at pH close to its p$K_a$ = 6.10 (use pH 6.00, 0.10 units off). $[\mathrm{H^+}]=10^{-6.00}=1.000\times10^{-6}$; $K_a = 10^{-6.10}=7.943\times10^{-7}$.

Now solve for the phosphate concentration $C'$ (still at pH 6.00, so $f = 0.05473$ from part a) that matches this:

So matching MES's modest 0.0200 M capacity at pH 6.00 with mismatched phosphate would require $C' = 0.0902$ M — over 4.5 times as much total buffer salt. This is the quantitative content of the "20 mM rule" heuristic: choosing a buffer with p$K_a$ close to the target pH lets you use dramatically lower (and cheaper, less ionic-strength-perturbing) concentrations for the same protective power, which is why practicing analysts keep a shelf of buffers spanning the whole pH range (phosphate, acetate, MES, HEPES, Tris, borate, carbonate) rather than relying on one system diluted or concentrated to force it to work far from its native pH.

The formal, exact treatment of buffer capacity is due to Donald Dexter Van Slyke, a Dutch-American biochemist working at the Rockefeller Institute for Medical Research in New York in the early twentieth century. Van Slyke was, above all, a builder of quantitative physiological chemistry: he developed manometric gas analysis apparatus still recognizable in clinical laboratories, precise methods for blood gas and electrolyte measurement, and — in a 1922 paper in the Journal of Biological Chemistry titled "On the Measurement of Buffer Values and On the Relationship of Buffer Value to the Dissociation Constant of the Buffer and the Concentration and Reaction of the Buffer Solution" — he gave buffer capacity its rigorous mathematical definition, coined the term "buffer value," and derived the very bell-curve relationship worked through above. His motivation was intensely practical: understanding how blood, a bicarbonate/phosphate/protein buffer system, resists the pH swings that would otherwise accompany metabolic acid production, and how clinicians could measure and interpret that resistance in patients with respiratory or metabolic acid-base disorders. Van Slyke's broader career exemplifies the analytical chemist's habit of turning a clinical or biological puzzle into a precise, reproducible number. He built apparatus (the Van Slyke manometric blood gas apparatus) that clinical laboratories used for decades to measure blood CO₂ content, and he insisted throughout his career on expressing physiological quantities — buffer capacity among them — with the same rigor as a titration in a beaker. That insistence is why buffer capacity today is not a vague qualitative notion ("this buffer is strong") but a defined, calculable, testable quantity with units of mol L⁻¹ per pH unit.