Determination of Preservative Content in Dairy Products: Methods, Standards, and Compliance Testing

Determination of preservative content in dairy products is a critical food safety and regulatory compliance task that quantifies sorbates, benzoates, natamycin, nisin, and other regulated additives in milk, yogurt, cheese, and milk powder. Accurate preservative testing protects consumer health, substantiates label claims, and verifies conformity with GB 2760, the Codex General Standard for Food Additives, and EU food additive regulations. This technical article reviews the sample types, target analytes, validated analytical methods, and practical considerations for accredited third-party laboratories performing preservative analysis in dairy matrices.

Technical Overview

Dairy products are nutrient-rich but highly perishable matrices with high water activity and near-neutral pH, making them susceptible to spoilage by yeasts, molds, and bacteria. To extend shelf life, manufacturers may add chemical or biological preservatives such as sorbic acid, benzoic acid, natamycin, nisin, or propionates to specific dairy categories. Because these additives are strictly regulated, quantitative determination of preservative residues is an essential element of food safety surveillance, factory quality control, and international trade conformity assessment.

The regulatory framework for preservatives in dairy varies by jurisdiction but is harmonized in principle. In China, GB 2760 defines which additives are permitted in each dairy category and sets maximum use levels or maximum residue limits. Internationally, the Codex Alimentarius General Standard for Food Additives (CXS 192), EU Regulation (EC) No 1333/2008, and the relevant sections of 21 CFR in the United States perform an equivalent function. Notably, plain pasteurized and sterilized milk are generally not permitted to contain added preservatives, while flavored fermented milk, processed cheese, and surface-treated cheese may carry defined limits. A further analytical challenge is that fermented dairy can contain low levels of naturally formed benzoic acid produced by lactic acid bacteria, which must be distinguished from illegal additive use.

• Primary analytical platforms: HPLC-UV/DAD, LC-MS/MS, GC-FID, and microbiological bioassays.
• Key reference standards: GB 5009.28 (benzoic and sorbic acids in foods), EN 12855 (natamycin in cheese rind), GB 5009.121 (dehydroacetic acid), and ISO 707 for sampling of milk and milk products.
• Typical reporting basis: mass fraction in g/kg (mg/kg for natamycin), expressed as the free acid equivalent.

Testing Samples

Liquid Milk and Milk Beverages

This category includes pasteurized milk, extended shelf-life (ESL) milk, UHT milk, reconstituted milk, and flavored or modified milk beverages. Although added preservatives are prohibited in plain liquid milk in most jurisdictions, these samples are routinely tested to detect illegal addition, cross-contamination from flavored production lines, and carry-over from reconstituted ingredients. Liquid matrices are relatively simple to prepare, requiring mainly dilution, protein precipitation, and filtration before instrumental analysis.

Fermented Dairy Products

Yogurt, drinking yogurt, kefir, cultured buttermilk, sour cream, and other fermented milks constitute the most frequently tested group, because sorbates and, in some markets, benzoates are permitted in flavored fermented milk at defined maximum levels. Interpretation of benzoic acid results in these products requires care, since microbial metabolism during fermentation and storage can generate several milligrams per kilogram of benzoic acid endogenously. Accredited laboratories therefore report quantitative values together with the applicable regulatory limit and contextual guidance.

Cheese and Processed Cheese

Natural cheese may be surface-treated with natamycin to suppress mold growth, while processed cheese and cheese analogues may contain sorbates, nisin, or propionates. Sampling strategy is decisive: natamycin is applied to the rind, so standards such as EN 12855 specify a defined rind depth (commonly a few millimeters) for sampling, and results must be clearly identified as surface-layer concentrations rather than whole-body averages.

Milk Powder, Cream, Whey, and Dairy Desserts

Milk powder, whey powder and concentrates, cream, butter, dairy-based desserts, and ice cream mixes are tested both as finished goods and as ingredients. In powders, preservative levels must be converted between as-is and dry-matter bases for correct comparison against limits. Ingredient testing also supports downstream manufacturers who must account for carry-over when verifying compliance of their own formulated products.

Testing Items

The scope of preservative determination in dairy typically covers the following regulated analytes, each quantified against its category-specific maximum limit:

• Sorbic acid and potassium sorbate (E200/E202): permitted in flavored fermented milk and certain cheese and fat-spread categories; typical maximum use levels are on the order of 1.0 g/kg expressed as sorbic acid.
• Benzoic acid and sodium benzoate (E210/E211): permitted in selected flavored dairy and dessert categories; endogenous formation in fermented milk must be considered during interpretation.
• Natamycin (E235): permitted for surface treatment of cheese, with maximum residue limits commonly set around 10 mg/kg in the rind layer, and penetration depth often additionally restricted.
• Nisin (E234): a bacteriocin permitted in processed cheese and certain dairy desserts in several jurisdictions.
• Propionic acid, sodium and calcium propionates (E280–E283): used primarily in cheese and cheese products for mold inhibition.
• Dehydroacetic acid (E266): monitored as a non-permitted residue screening item in dairy matrices under GB 5009.121.

Results are reported as the free acid or free base equivalent where required by regulation, together with measurement uncertainty, which is essential when a value lies close to a statutory maximum limit.

Testing Methods

Sampling and Sample Preparation

Sampling follows ISO 707 guidance for milk and milk products, with representative subdivision of bulk lots and cold-chain transport where required. Laboratory preparation involves thorough homogenization, followed by matrix-appropriate cleanup: proteins are precipitated using Carrez reagents (potassium ferrocyanide and zinc acetate) or acidified acetonitrile, fat is removed by centrifugation for high-fat products such as cheese and cream, and extracts are filtered prior to injection. For natamycin in cheese, the rind is separated at a standardized depth and extracted with aqueous methanol in accordance with the principles of EN 12855.

HPLC-UV/DAD Quantification

Reversed-phase HPLC on a C18 column with methanol–ammonium acetate mobile phase and diode-array detection is the workhorse method for sorbic and benzoic acids, following GB 5009.28. Detection is performed at the characteristic absorption maxima of each analyte, approximately 223–230 nm for benzoic acid and 255–260 nm for sorbic acid, with DAD spectral confirmation of peak purity. External calibration with matrix-matched or standard-addition curves compensates for residual matrix effects in fermented and high-protein samples. Typical limits of quantification are in the low mg/kg range, comfortably below the regulatory maxima.

LC-MS/MS Confirmation and Multi-Analyte Screening

Liquid chromatography–tandem mass spectrometry provides definitive confirmation of identity at low residue levels, simultaneous determination of multiple preservatives in a single run, and reliable quantification in difficult matrices such as colored or fruit-preparation-containing dairy products. LC-MS/MS is the preferred technique for dispute resolution, regulatory enforcement actions, and detection of non-permitted additive use in plain milk.

Natamycin Determination

Natamycin is quantified by HPLC with UV/DAD detection at approximately 303 nm following methanol-water extraction of the cheese rind, using an isocratic or shallow-gradient C18 separation. Because the analyte is light-sensitive and surface-localized, samples must be protected from intense illumination and the sampling depth documented in the test report.

Nisin Bioassay and Alternative Techniques

Nisin is classically determined by microbiological agar-diffusion (cylinder plate) bioassays against a standard curve of a reference nisin preparation, reflecting its activity-based unit system. Increasingly, LC-MS/MS methods for intact nisin A quantification and immunochemical screening kits are used to improve throughput and specificity. Propionic acid and other short-chain organic acids are alternatively determined by GC-FID or ion chromatography after acidified extraction from cheese.

Method Performance and Quality Assurance

Validated methods in accredited third-party laboratories demonstrate recovery typically within 85–110%, repeatability RSD below 5–10%, and linearity covering the regulatory decision range. Quality assurance includes procedural blanks, spiked recovery controls, certified or in-house reference materials where available, duplicate analysis of homogenized subsamples, and participation in proficiency testing schemes for dairy additives.

Applications

Preservative content determination in dairy supports a broad range of quality, safety, and commercial objectives across the supply chain:

• Regulatory compliance testing for market surveillance authorities and import inspections against GB 2760, Codex, EU, and destination-market requirements.
• Factory quality control, including verification of dosing accuracy in flavored fermented milk and processed cheese production lines.
• Verification of "no added preservative" and clean-label claims for plain milk, organic dairy, and premium product lines.
• Export certification support for dairy shipments requiring analytical evidence of additive conformity.
• Investigation of consumer complaints, spoilage incidents, and suspected illegal additive use.
• Research and development support for shelf-life studies, preservative system optimization, and evaluation of naturally occurring benzoic acid formation during fermentation and storage.
• Supplier and incoming ingredient qualification for milk powders, whey products, and fruit preparations used in dairy formulations.

FAQ

Why can yogurt contain benzoic acid even when no preservative was added?

Lactic acid bacteria and other fermentation microorganisms can biosynthesize small amounts of benzoic acid during yogurt production and storage, typically at levels of a few milligrams per kilogram. Regulatory frameworks in many jurisdictions recognize this endogenous occurrence, and experienced laboratories interpret results in this context rather than treating every detection as additive misuse.

Are preservatives allowed in plain milk?

No. Under GB 2760 and equivalent regulations in most markets, plain pasteurized and sterilized milk must not contain added preservatives. Detection of sorbic or benzoic acid in plain milk above trace background levels indicates non-compliance and warrants confirmatory LC-MS/MS testing.

What detection limits can be achieved?

Routine HPLC-UV/DAD methods achieve limits of quantification of approximately 5–10 mg/kg for sorbic and benzoic acids in dairy, while LC-MS/MS can reach sub-mg/kg levels for most analytes. These limits are well below typical regulatory maxima, providing ample margin for confident compliance decisions.

How is natamycin on cheese correctly sampled and reported?

Because natamycin is applied to the cheese surface, the rind is sampled at a defined depth, commonly a few millimeters, following the approach set out in EN 12855. Test reports must state whether results refer to the rind layer or the whole cheese, since whole-body dilution can substantially understate surface concentrations.

How should results be expressed for milk powder products?

For milk powders and whey powders, results should be reported both on an as-received basis and on a dry-matter basis, with moisture content documented, so that measured levels can be compared correctly with regulatory limits defined for reconstituted or finished-product categories.

What is the typical turnaround time for preservative testing?

Routine single-analyte HPLC determinations are generally completed within five to seven working days after sample receipt, while multi-analyte LC-MS/MS panels, confirmatory testing, or dispute-related retesting may require additional time for extraction optimization and quality review.