Determination of milk protein by the Kjeldahl method is the internationally recognized reference technique for quantifying protein in raw milk, dairy products, and milk-based ingredients. This article explains the underlying principle, analytical workflow, applicable standards such as ISO 8968, GB 5009.5, and AOAC Official Methods, and the practical considerations that ensure accurate, reproducible results in accredited third-party laboratories.
Technical Overview
The Kjeldahl method is the classical and most widely accepted reference procedure for determining the nitrogen content of dairy products, from which the protein content is calculated. First developed in 1883, the method remains the backbone of milk protein analysis in regulatory testing, quality control, and commercial milk payment systems worldwide.
The technique measures total nitrogen in a sample following three sequential steps: acid digestion, alkaline steam distillation, and acidimetric titration. Because dairy proteins contain approximately 15.67% nitrogen on average, the measured nitrogen value is multiplied by the specific conversion factor of 6.38 to obtain the milk protein content. It should be noted that Kjeldahl yields "crude protein," since it captures both protein nitrogen and non-protein nitrogen (NPN) such as urea, ammonium salts, and free amino acids. When true protein content is required, the non-protein nitrogen fraction is determined separately (e.g., following ISO 8968-4) and subtracted from the total.
The principal standards governing this analysis include:
- ISO 8968 series (Milk — Determination of nitrogen content), including Part 1 (macro Kjeldahl), Parts 2 and 3 (block digestion variants), and Part 4 (non-protein nitrogen)
- GB 5009.5 (National Food Safety Standard — Determination of Protein in Foods), which designates Kjeldahl as the first method
- GB 19301, GB 19645, and GB 25190, which set minimum protein limits for raw, pasteurized, and sterilized milk respectively
- AOAC Official Method 991.20 for total nitrogen in milk by block digestion
- ISO 17997 for casein nitrogen determination in milk
With modern block digesters and automated distillation-titration units, the method achieves repeatability (RSD) typically better than 1% for homogeneous dairy matrices, making it suitable for label claims, regulatory compliance, and dispute resolution.
Testing Samples
Liquid Dairy Products
Raw (farm and tanker) milk, pasteurized milk, UHT and sterilized milk, flavored milk drinks, cream, condensed milk, and reconstituted milk are routinely analyzed. Liquid samples are gently tempered (commonly to approximately 40 °C, then cooled to 20 °C) and homogenized before weighing to ensure representative fat and protein distribution.
Milk Powders and Protein Ingredients
Whole and skim milk powder, whey protein concentrate and isolate, milk protein concentrate, lactose-free powders, and dairy protein blends require dry, homogeneous subsampling. Powder samples are weighed directly into digestion tubes, typically at 0.1–0.3 g to keep the nitrogen load within the titration range.
Fermented Dairy Products
Yogurt, drinking yogurt, kefir, and fermented milk products are analyzed after thorough homogenization. For cheeses, the sample is grated or comminuted and mixed until uniform before subsampling.
Infant Formula and Dairy-Based Foods
Infant formula, follow-on formula, growing-up milk, and dairy-containing composite foods are tested for protein content against the applicable national and international composition standards, using the 6.38 factor where the formula is milk-based.
Testing Items
Typical test items reported under the Kjeldahl framework include:
- Total nitrogen content (g nitrogen per 100 g)
- Crude protein content (total nitrogen × 6.38)
- True protein content (total nitrogen minus NPN, × 6.38)
- Non-protein nitrogen (NPN) content
- Casein nitrogen and whey (non-casein) nitrogen content
- Protein content on a dry matter basis or as-is basis, per the client specification
Testing Methods
The analytical sequence comprises digestion, distillation, and titration, followed by calculation and quality control evaluation.
Digestion
A weighed sample (for example, approximately 1 g of liquid milk or 0.2 g of milk powder) is placed in a digestion tube with a potassium sulfate-based catalyst tablet (typically copper sulfate, with or without titanium dioxide), and 10–20 mL of concentrated sulfuric acid. The tube is heated in a block digester at approximately 400–420 °C. Organic matter is carbonized and then oxidized, and all organic nitrogen is converted to ammonium sulfate. Digestion continues until the digest is clear green or blue and remains so for an additional 30 minutes to ensure complete conversion. Mercury-free catalysts are strongly preferred for environmental and safety reasons.
Distillation
After cooling and dilution, concentrated sodium hydroxide solution (approximately 32–40%) is added to make the digest strongly alkaline, liberating ammonia. The ammonia is steam distilled and trapped in a receiving flask containing boric acid solution with a mixed indicator (bromocresol green–methyl red) or in excess standard acid. Distillation is continued until a defined distillate volume (commonly about 150 mL) has been collected or until a validated fixed-time protocol confirms quantitative recovery.
Titration and Calculation
The absorbed ammonia is titrated with standardized hydrochloric or sulfuric acid (commonly 0.05–0.25 mol/L depending on sample loading) to a colorimetric or potentiometric endpoint, with a reagent blank run in parallel. The protein content is calculated as follows:
Protein (g/100 g) = (Vs − V0) × c × 1.4007 × 6.38 / m, where Vs and V0 are the titration volumes (mL) of sample and blank, c is the exact molarity of the standard acid (mol/L), and m is the sample mass (g). Results are typically reported to 0.01 g protein per 100 g of product.
Non-Protein Nitrogen and Protein Fractionation
For true protein, proteins are precipitated (for example with acidified conditions or trichloroacetic acid following ISO 8968-4), the precipitate is filtered off, and the Kjeldahl procedure is performed on the filtrate to quantify NPN. True protein equals total nitrogen minus NPN, multiplied by 6.38. Casein nitrogen can be determined separately by isoelectric precipitation at pH 4.6 in accordance with ISO 17997.
Quality Control and Validation
Each analytical batch should include reagent blanks, duplicate samples, a certified or in-house reference milk powder, and recovery checks with a nitrogen reference compound such as ammonium sulfate, tryptophan, or acetanilide. Method performance is verified against the precision limits stated in ISO 8968 or GB 5009.5, with repeatability typically below 1% RSD. The Dumas combustion method (ISO 14891) may be used as a complementary rapid technique, provided equivalence with the Kjeldahl reference is demonstrated.
Applications
Kjeldahl-based milk protein determination supports a broad range of commercial and regulatory purposes:
- Nutrition labeling and protein content claims on retail dairy packaging
- Verification of compliance with minimum protein limits for raw milk (≥ 2.8 g/100 g) and pasteurized or sterilized milk (≥ 2.9 g/100 g) under the applicable national standards
- Raw milk quality grading and payment schemes at dairy collection centers
- Process control in cheese and fermented product manufacture, where casein content influences yield
- Incoming inspection of milk powders, whey proteins, and dairy ingredients
- Infant formula compliance testing against statutory compositional requirements
- Screening for economically motivated adulteration, in combination with NPN fractionation and targeted chromatographic methods for specific nitrogen-rich adulterants
- Nutritional and product development research in accredited third-party laboratories and industry R&D facilities
FAQ
Why is the factor 6.38 used for milk protein?
Milk proteins contain on average 15.67% nitrogen by mass. The conversion factor is the reciprocal of this fraction (100 ÷ 15.67 ≈ 6.38), so multiplying nitrogen content by 6.38 estimates protein content. For many non-dairy foods the general factor 6.25 is used instead, which is why the matrix-specific factor must be specified in reports.
What is the difference between crude protein and true protein?
Crude protein is calculated directly from total nitrogen and therefore includes non-protein nitrogen such as urea and free amino acids. True protein is obtained by subtracting the measured NPN fraction from total nitrogen before applying the 6.38 factor. For fresh raw milk, NPN typically accounts for roughly 5–6% of total nitrogen, so the distinction matters for accurate product characterization.
Can the Kjeldahl method detect protein adulteration?
Not on its own. Because the method measures total nitrogen indiscriminately, nitrogen-rich adulterants can falsely inflate apparent protein, as demonstrated by the 2008 melamine incident. Effective adulteration control therefore combines Kjeldahl with NPN fractionation and confirmatory instrumental methods such as LC-MS/MS for specific adulterants.
How long does a Kjeldahl protein determination take?
Batch digestion typically takes 60–120 minutes for a rack of 20 tubes, followed by approximately 5–8 minutes per sample for automated distillation and titration. Including weighing, cooling, and calculation, routine throughput of 40–100 samples per day is achievable with modern block digestion and automatic distillation equipment.
What are the main safety considerations?
The method involves hot concentrated sulfuric acid, strong sodium hydroxide, and corrosive acid titrants. Laboratories must operate digestion under efficient fume extraction with acid-resistant manifolds, use mercury-free catalysts, provide appropriate personal protective equipment, and neutralize acid–alkaline waste before disposal in accordance with environmental regulations.
Which standards should the testing laboratory follow?
Testing should be performed to ISO 8968 (including Part 4 when NPN is required), GB 5009.5, or AOAC Official Method 991.20 as appropriate to the market and regulatory context. Accredited third-party laboratories operating under ISO/IEC 17025 quality management systems provide results that are widely accepted for regulatory submissions, labeling, and trade disputes.