Propionic acid (C₃H₆O₂; molecular weight 74.08 g/mol; pKa 4.87 at 25 °C in dilute aqueous solution) is a short-chain organic acid preservative whose antimicrobial mechanism is best described by one principle: pH matters more than dose. The active species is the undissociated acid, not the propionate anion. Dissociation is controlled by the Henderson-Hasselbalch equilibrium: undissociated fraction θ = 1 / (1 + 10^{(pH − pKa)}). Because microbial cytoplasmic pH is typically 7.0–7.4, undissociated propionic acid that diffuses across the plasma membrane dissociates inside the cell, releasing protons and propionate. Proton export through membrane-bound H⁺-ATPase increases the cell’s ATP demand, while the accumulating propionate anion disrupts intracellular osmotic balance and enzyme function. This weak-acid uncoupling mechanism makes external pH the primary determinant of antimicrobial flux. At pH 6.5, only 0.023 of total propionic acid remains undissociated; at pH 5.0, the fraction is 0.426. Increasing total dose at fixed pH produces linear changes in undissociated acid concentration, whereas moving pH across the pKa produces exponential changes. Therefore, pH dominates the dose-response relationship in practical preservation.

This principle does not make dose irrelevant. At a fixed pH, doubling total propionic acid doubles the undissociated acid concentration. However, the operating range for dose is limited by palatability, corrosivity, and regulatory permissions, while the operating range for pH is broader and biochemically more consequential. Preservation failures in feed and food systems are frequently caused by applying more propionate salt at an unfavorable pH rather than by using an inherently ineffective acid.

What is the undissociated fraction at a given pH?

The undissociated fraction θ is calculated as θ = 1 / (1 + 10^{(pH − pKa)}). At pH equal to pKa, θ = 0.50. At pH 4.0, θ is 0.881; at pH 4.5, 0.701; at pH 5.5, 0.190; at pH 6.0, 0.069; at pH 7.0, 0.0074. Total concentration sets the absolute amount of undissociated acid. A total propionic acid concentration of 3000 mg/L yields undissociated acid concentrations that decline from 2103 mg/L at pH 4.5 to 207 mg/L at pH 6.0 and 69 mg/L at pH 6.5.

These values are calculated for an unbuffered aqueous system at 25 °C. In feed, silage, or culture media, buffering by proteins, mineral ash, calcium carbonate, organic acid salts, and other components consumes protons and shifts the equilibrium toward the dissociated anion. The effective undissociated concentration in a real matrix is therefore typically lower than the ideal aqueous calculation. Because antimicrobial activity correlates with undissociated acid concentration, pH must be measured in the final equilibrated matrix, not inferred from the amount of acid added.

Dose escalation cannot compensate for low permeability

At pH 6.0, only 0.069 of total propionic acid is undissociated. If a preservation target is defined as the undissociated acid concentration produced by a 0.2% w/w propionic acid application at pH 5.0, the required total dose at pH 6.0 is 1.23% w/w, a 6.2-fold increase. This calculation is before additional acid neutralization by the matrix. At pH 6.5, the required dose becomes approximately 3.70% w/w to match the same active concentration; at pH 7.0, it exceeds 11.5% w/w. Such total acid levels are generally impractical and undesirable in feed and food because of palatability, corrosivity, and regulatory limits. Reducing matrix pH from 6.0 to 5.0 has the same effect on active acid concentration as a 6.2-fold increase in total dose at constant pH.

The apparent permeability coefficient of propionic acid across a lipid bilayer is several orders of magnitude higher than that of the propionate anion. The ionized propionate outside the cell does not contribute to transmembrane flux to any practical extent; increasing total propionate by adding sodium or calcium propionate at neutral pH is therefore ineffective until the matrix pH falls below approximately 5.5. This transport limitation is why formulation with acidulants, rather than simple dose escalation, is the standard corrective action in preservation systems operating at elevated pH.

In high-moisture corn stored at 18–22% moisture, the aqueous film on kernel surfaces is the site of mold germination. Commercial application equipment includes a continuous-flow paddle mixer or screw conveyor fitted with low-pressure spray nozzles and a positive-displacement metering pump. Propionic acid is commonly applied at 0.2–0.8% w/w, but efficacy depends less on the exact dose within this range than on the final surface pH. If the 1:10 aqueous suspension pH after application is 5.0, the undissociated fraction is 0.426. If the same dose leaves the suspension pH at 6.0, the undissociated fraction drops to 0.069. A fourfold increase in applied propionic acid from 0.2% w/w to 0.8% w/w at pH 6.0 yields an undissociated concentration of 0.055% w/w, still below the 0.085% w/w produced by 0.2% w/w at pH 5.0. Uniform distribution is monitored by sampling after mixing and measuring the 1:10 suspension pH with a calibrated electrode; consistent pH readings indicate that acid pooling has not occurred.

When pH rises above 5.5 in high-moisture grain

Grain or feed matrices that contain limestone, urea-based buffering agents, or mineral ash can remain above pH 5.5 after the total propionic acid dose reaches 0.8% w/w. At pH 5.5, the undissociated fraction is 0.190. This may slow mold growth but is often insufficient to prevent mycotoxin-forming Aspergillus and Penicillium species during long-term storage. Published strain-specific MIC data for mycotoxin-producing fungi across pH gradients are limited, but preservative challenge tests that compare the same total acid concentration at pH 5.0 and pH 6.5 consistently show higher minimum inhibitory concentrations at the higher pH. Because medium pH is controlled during the assay, the dose-response relationship should be compared through undissociated acid concentration rather than total acid concentration.

Preservative challenge testing according to ISO 20976-1:2019 is appropriate for quantifying this effect because it compares growth curves at the same total acid concentration but different pH values. Where MIC values are generated by broth microdilution under CLSI M38-A2, the pH of the medium must be reported; comparison of total MIC values across studies without pH data is of limited value. The concentration gradient of undissociated acid across the hyphal membrane is the driving force for entry. If matrix pH remains above 5.5, the gradient is too small to deliver inhibitory intracellular levels before the applied acid is neutralized by buffering reactions.

Silage, bakery, and feed preservation contexts with pH constraints

Silage fermentation usually lowers pH to 4.0–4.2. At pH 4.0, the undissociated propionic acid fraction is 0.881, making even low addition rates active against aerobic spoilage organisms at feed-out. In bakery products, calcium propionate or sodium propionate is used at flour weight levels of 0.1–0.3%. Dough pH commonly ranges from 5.0 to 5.7, providing undissociated fractions between 0.426 and 0.129. When formulation changes raise dough pH above 5.7, mold inhibition weakens; doubling the propionate dose may not compensate unless the formulation also reduces buffering or uses acidulants.

Propionic acid in drinking water for livestock requires pH lowering. At water pH 7.0, only 0.0074 of the total acid is undissociated. Acidifying drinking water to pH 4.5 raises the active fraction to 0.701. Organic acid blends therefore include acidulants such as formic, citric, or lactic acid to lower pH and shift propionic acid toward the undissociated form. The salts sodium propionate and calcium propionate do not lower pH directly; they release propionic acid only when the environment provides protons. At neutral pH, they remain predominantly as the propionate anion and exhibit little antimicrobial activity until acidification occurs.

Current good manufacturing practice means that the substance may be used at levels not exceeding those required to accomplish the intended effect. The regulatory permission does not override the pH-dependent efficacy boundary; a GMP dose at neutral pH can still be microbiologically inadequate.

Buffer capacity is the practical input that determines whether a given propionic acid dose will produce the target pH. It is quantified as milliequivalents of acid per kilogram of matrix required to lower pH to 5.0. Feed matrices containing soybean meal, distillers grains, or limestone can require substantially more acid than low-ash cereal grain before the pH moves across the pKa. A batch-specific titration should be conducted before setting the preservative dose. Without that measurement, a defined propionic acid addition may not reach the pH zone where its antimicrobial mechanism becomes effective.

The Fickian flux of undissociated acid across a microbial membrane is approximately proportional to the concentration difference of undissociated acid between the external environment and the cytoplasm. Because the external undissociated concentration is fixed by pH and total dose, the flux drops sharply as pH rises. This model explains why weak-acid preservatives have pH windows rather than simple concentration windows. Propionic acid does not reduce water activity; at water activity 0.90–0.95, mold control depends on acid concentration and pH, not on water activity suppression. Process control should therefore specify the equilibrated matrix pH, buffer capacity, and active acid concentration as separate release parameters.

Material compatibility and safety boundaries also limit dose escalation. Propionic acid is corrosive to carbon steel and aluminum; storage and transfer lines are typically specified in 316L stainless steel or high-density polyethylene according to the acid supplier’s chemical compatibility data. Concentrated propionic acid produces irritating vapor at ambient temperature, so unloading and spray mixing are conducted in ventilated areas. These operational limits create a dose ceiling that is reached before pH-neutral matrices can be compensated by additional acid. The antimicrobial activity of propionic acid therefore remains governed by the pH-dependent equilibrium between the membrane-permeable undissociated acid and the largely inactive propionate anion.