Theory & Concepts

A buffer is a solution that resists changes in pH when small amounts of acid or base are added — and even when it is diluted. Living systems depend on this: human blood is held near pH 7.4, and an enzyme assay or a cell culture can be ruined by a drift of just a few tenths of a pH unit. A buffer is the chemist's way of pinning the pH where the biology needs it.

Every buffer is built from a conjugate acid–base pair: a weak acid HA and the conjugate base A⁻ it turns into when it gives up a proton. The two interconvert through a single equilibrium:

Because the acid is weak, this reaction does not run to completion — it sits at an equilibrium where appreciable amounts of both HA and A⁻ are present at once. That coexistence is the whole point. The weak acid HA is a reservoir that can donate protons, while the conjugate base A⁻ is a reservoir that can soak up protons:

• Add a strong acid (extra H⁺)? The base form mops it up:  A⁻ + H⁺ → HA.
• Add a strong base (extra OH⁻)? The acid form neutralises it:  HA + OH⁻ → A⁻ + H₂O.

Either way the added H⁺ or OH⁻ is mostly converted, not left free in solution, so the pH barely moves. This is why a buffer is often called a chemical shock absorber: just as a car's suspension absorbs the energy of a bump so the cabin stays level, a buffer absorbs incoming acid or base so the pH stays level. The shock absorber only works while it has travel left in both directions — and a buffer only works while it still holds a meaningful amount of both HA and A⁻. Run out of one form and the cushioning is gone.

How far the equilibrium lies is captured by the acid dissociation constant Ka:

A larger Ka (smaller pKa) means a stronger acid that gives up its proton more readily. We almost always work with pKa rather than Ka because the numbers are friendly: acetate's Ka of 1.8 × 10⁻⁵ is much easier to remember as pKa = 4.76. As you will see in the next section, the pKa is also the single number that tells you the pH a buffer naturally settles at — and therefore where it works best.

The Henderson–Hasselbalch equation is the workhorse of buffer design. It connects the pH of a buffer to its pKa and the ratio of the two forms present:

It is just the Ka expression rearranged. Start from the equilibrium, take −log₁₀ of both sides:

Reading the equation term by term:

• pH — what you measure and want to control.
• pKa — a fixed property of the buffer system (at a given temperature). It anchors the equation.
• [A⁻] / [HA] — the ratio of conjugate base to weak acid, the one thing you adjust when you make the buffer. Note it is a ratio: only the proportion matters for pH, not the absolute amounts.

The most important consequence falls straight out of the maths. When the two forms are present in equal amounts, [A⁻] = [HA], so their ratio is 1, and log₁₀(1) = 0. The equation collapses to:

So a buffer made with equal parts acid and base form sits exactly at its pKa. Want a pH one unit above the pKa? You need ten times as much base as acid (log₁₀ 10 = 1). One unit below? Ten times as much acid. Each pH unit away from the pKa costs you a tenfold imbalance — which, as the next section explains, is why buffers weaken as you stray from the pKa. The equation is a teaching idealisation (it assumes activity equals concentration and ignores ionic strength), but it is accurate to a few hundredths of a pH unit and is exactly how recipes are calculated.

Try it: Buffer Calculator →

A buffer holds pH steady, but not infinitely — add enough acid or base and you will eventually overwhelm it. Buffer capacity (β) measures how much it can take. Formally, β is the moles of strong acid or base needed to shift one litre of buffer by one pH unit:

A high β means the pH barely flinches when you challenge it; a low β means it lurches. Two factors control β, and both are intuitive once you picture the reservoirs of HA and A⁻.

Capacity is greatest at pH = pKa. At the pKa the two forms are equal, so the buffer has the most “room” to move in both directions at once — plenty of A⁻ to absorb incoming acid and plenty of HA to absorb incoming base. Move away from the pKa and one reservoir starts to run dry, so the buffer becomes lop-sided and easier to push around. The capacity at the pKa is, for a monoprotic buffer of total concentration C:

Capacity rises with concentration. Double the total amount of buffer and you double β: a 0.2 M buffer absorbs twice the insult of a 0.1 M one before the pH gives way. That is why “make it more concentrated” is a real fix for a buffer that drifts under load.

Combine the two facts and you get the practical rule for the buffer range: a buffer is useful only within about pKa ± 1 pH unit. At pKa ± 1 the ratio of the forms is 10:1 and β has already fallen to roughly a third of its peak; beyond that the minority form is too scarce to cushion anything. Aim for your target pH to sit within ±1 of the pKa, and ideally within ±0.5, where the buffer is at its strongest.

There is a trade-off, though, which sets the upper limit on concentration. Cranking up C raises β, but it also raises the ionic strength and osmolality of the solution. Too much salt can stress cells, inhibit enzymes, alter protein solubility and shift activities enough to move the “true” pH. Biological buffers are therefore usually a compromise — typically 10–100 mM — chosen to be strong enough to hold pH but mild enough for the organism.

Try it: Capacity Explorer →   Try it: Titration Simulator →

With dozens of buffers available, picking the right one is a matter of working through a short checklist. The first criterion outranks all the others.

• pKa match (the deciding factor). Choose a buffer whose pKa is within ±1 of your target pH — ideally within ±0.5. This is non-negotiable: outside that window the buffer simply has too little capacity (see §3). To run an assay at pH 7.4 you would reach for HEPES (pKa 7.48) or phosphate (pKa₂ 7.20), not acetate (pKa 4.76).
• Temperature behaviour. Some buffers shift pKa sharply with temperature (Tris is the notorious example). If you will prepare a buffer at the bench but use it in a cold room or an incubator, pick a temperature-stable system or set the pH at the use temperature (see §6).
• Ionic strength & osmolality. A buffer that needs a high concentration to do its job adds a lot of salt. For cell work you want a buffer that holds pH at modest concentration.
• Biological interference. The buffer must not react with your system. Common pitfalls:
  • Phosphate precipitates with Ca²⁺ and Mg²⁺ and inhibits many enzymes (kinases, polymerases) that use those ions or that phosphate mimics a substrate of.
  • Citrate is a powerful metal chelator — it strips Ca²⁺/Mg²⁺/Fe out of solution, which is sometimes the goal (anticoagulant) but disastrous for metal-dependent enzymes.
  • HEPES can generate hydrogen-peroxide and other radicals when exposed to light, damaging light-sensitive cells and oxidisable reagents — keep it dark.
• UV transparency. If you measure absorbance in the UV, the buffer must not absorb where you read. HEPES and other aromatic/ring buffers absorb strongly below ~230 nm, masking protein and nucleic-acid signals; phosphate is UV-transparent and often preferred for spectroscopy.
• Cost, purity and availability. Acetate and phosphate are cheap and bulk-friendly; the zwitterionic Good's buffers cost more but earn it in biological compatibility.

Several of the best biological buffers come from one landmark study. In 1966 Norman Good and colleagues set out to design buffers fit for biology and published a now-classic set of criteria. A good biological buffer should have a pKa between roughly 6 and 8 (the physiological range), be highly water-soluble but membrane-impermeant (so it stays outside cells), show minimal binding of metal ions, be chemically stable and non-toxic, resist enzymatic and other interference, and have a pKa that is relatively insensitive to temperature, concentration and ionic composition. The Good's buffers — MES, MOPS, PIPES, HEPES, TES, Tricine and relatives — are the zwitterionic compounds that meet these targets, and they dominate modern cell biology and biochemistry.

The table below lists the buffers used throughout this site, with the pKa relevant to each buffering region (at 25 °C), the approximate useful pH range (≈ pKa ± 1), and the practical note you should keep in mind. For polyprotic acids such as phosphate and citrate, the listed pKa is the one for the region we model; the others are given for context.

A buffer's pKa is not truly constant — it drifts with temperature, and the size of that drift varies enormously between systems. The temperature coefficient (ΔpKa/°C) tells you how much. Two contrasting cases illustrate why this matters in practice.

Tris is the cautionary tale. Its ΔpKa/°C ≈ −0.028, one of the largest in common use. The minus sign means the pKa rises as the solution gets colder, dragging the pH up with it. A Tris buffer adjusted to pH 8.0 on the bench at 25 °C will read about:

That is a swing of nearly a full pH unit between the fridge and the incubator — more than enough to denature a fussy protein or kill an enzyme reaction. (You can estimate any of these with ΔpH ≈ ΔpKa/°C × Δtemperature.)

Phosphate is the reassuring counter-example. Its ΔpKa/°C ≈ −0.0028 — ten times smaller — so a phosphate buffer is almost flat across the same range, shifting only a couple of hundredths of a pH unit. This temperature stability, plus UV transparency, is a big part of why phosphate is so popular.

The practical rule follows directly: always set the pH at the temperature where the buffer will be used. If your experiment runs at 4 °C, chill the buffer and adjust the pH cold; if it runs at 37 °C, adjust it warm. Setting pH at room temperature and assuming it holds is one of the most common — and most invisible — sources of failed experiments.

Two further effects are worth flagging. Dilution shifts pH: diluting a buffer changes the ionic strength and therefore the ion activities, so a concentrated stock will not have exactly the same pH once diluted to working strength — prepare and pH-adjust at the final concentration, or re-check after dilution. And ionic strength itself alters the gap between activity and concentration, which is the main reason the Henderson–Hasselbalch prediction and the meter reading can disagree by a few hundredths. For teaching we ignore these activity corrections; in a real protocol you set the final pH empirically with a calibrated meter.

Try it: Prep Bench →

A pH meter does not measure pH directly — it measures a voltage and converts it. The sensing element is the glass electrode: a thin, ion-selective glass membrane that develops a small electrical potential proportional to the activity of H⁺ ions in the solution it touches. That potential is compared against a stable reference electrode (often combined into one body), and the difference — a few hundred millivolts — is what the meter reads.

The relationship between voltage and pH is given by the Nernst equation. At 25 °C an ideal electrode changes its potential by a fixed amount per pH unit:

This Nernstian slope is temperature-dependent (it is proportional to absolute temperature, ≈ 0.1984 × T mV/pH), which is why good meters include automatic temperature compensation (ATC) — a probe that corrects the slope for the sample temperature. Note this corrects the electrode response, not the buffer's own temperature drift from §6; those are two separate effects.

Because no real electrode is perfect, you must calibrate before measuring. Calibration uses standard buffers of certified pH — typically the trio 4.01, 7.00 and 10.01 — to find two numbers:

• Slope — the actual mV/pH response, reported as a percentage of the ideal Nernstian value. A healthy electrode reads 95–105 %. A slope drifting below ~95 % signals an aging, dirty or dehydrated electrode that should be cleaned or replaced.
• Offset — the small voltage error at pH 7 (the “zero” point). It should sit near 0 mV; a large offset points to a contaminated junction or a tired reference.

Always bracket your sample: calibrate with standards that lie on either side of the pH you expect (e.g. 7.00 and 10.01 for a pH 8 buffer) so you are interpolating between known points, not extrapolating beyond them. And treat the electrode kindly — store it in KCl storage solution (or pH 4 buffer), never in deionised water. Pure water leaches ions out of the glass membrane and the reference junction, irreversibly degrading the slope. A dry or DI-soaked electrode is the single most common reason a “broken” meter ends up in the bin.

Try it: pH-Meter Bench →

Knowing the chemistry is only half the job; preparing a buffer reproducibly is the other half. Good Laboratory Practice (GLP) turns a one-off success into a procedure anyone can repeat. The standard preparation workflow runs in a fixed order for good reasons:

1. Calculate the masses from the recipe (Henderson–Hasselbalch + molar masses).
2. Weigh each component on an analytical balance, recording the actual mass.
3. Dissolve in ~80 % of the final volume — never the full volume. You need headroom, because adjusting the pH and bringing to volume will add liquid.
4. Adjust the pH with small additions of strong acid or base, using a calibrated meter, at the intended use temperature.
5. Bring to final volume in a Class A volumetric flask. Topping up only now is what makes the final concentration exact.
6. Re-check the pH after making to volume — adding the last water can shift it slightly.
7. Label and store: name, concentration, pH, date, preparer and any storage conditions.

Get the order wrong — for instance, adjusting pH only after going to volume, or dissolving in the full volume — and you end up with the wrong concentration or having to start over. Try the Prep Bench to run this sequence yourself.

Before a buffer is released for use it should pass a short panel of quality-control checks:

Finally, document everything. A written, version-controlled SOP ensures the buffer is made the same way every time; a batch record with lot numbers, balance readouts, meter calibration data and the preparer's initials provides traceability — if a result looks odd weeks later, you can trace it back to exactly which batch and reagents were used. Reproducibility is not an accident; it is the product of disciplined records.

Try it: QC Troubleshooting →   Test yourself: Knowledge Quiz →