Bookmark and Share

From: Food Quality & Safety magazine, August/September 2008

Protein Determination in Cereals and Seeds

New combustion methods and technologies eliminate sample preparation

by Liliana Krotz, PhD; Elena Ciceri, PhD; and Guido Giazzi, PhD

Cereals and seeds are significant components of the human diet and the principal part of feeding stock for domestic animals. One of the most important nutrients present in these is protein. Functioning as enzymes, hormones, and antibodies, as well as transport and structural components, proteins are required for the body’s structure and proper function.1 In addition to its dietary importance, protein content has become a guideline for some cereal trade transactions as well as a means to assess quality. The monitoring of protein content through the measurement of nitrogen release must be accurate in order to determine the nutritional quality of produce.

The two most prevalent protein determination methods rely on the release of nitrogen from the amine groups found in the peptide bonds of protein’s polypeptide chains. Regulations for appropriate monitoring methods for grain and seed analysis vary throughout the world. In the United States, the Association of Official Analytical Chemists (AOAC) International Method 992.23 and the American Association of Cereal Chemists (AACC) Method 46-30 (1999) indicate that the suitable fineness of grind must be determined for each different material analyzed to achieve precision that gives a residual standard deviation (RSD) of �2% for 10 successive determinations of nitrogen.

The traditional Kjeldahl method relies on oxidation to release nitrogen, while the Dumas combustion method, as its name implies, breaks down the bonds in the peptide chains, permitting the release of nitrogen through complete combustion of the sample.

Food analysis is a growing and important market that requires superior accuracy in order to ensure optimum productivity. The dominance of the traditional Kjeldahl method for nitrogen measurement has recently come under threat by the challenge of safer, cleaner instruments employing the Dumas combustion principle.2 The combustion method offers a faster, safer, and more reliable method for the analysis of foodstuffs.

The Kjeldahl Method

The Kjeldahl method originates from the determination of the protein content of grains used in the brewing industry and relies on quantitative determination of nitrogen. This method can be broken down into three basic steps.

  • Digestion of the sample: The most time-consuming step in the analysis, this is designed to break down the bonds that hold the polypeptides together and convert them to simpler chemicals such as water, carbon dioxide, and ammonia. Adding strong sulfuric acid and heating the mixture to about 370ºC to 400ºC for 60 to 90 minutes oxidizes the organic material and releases ammonium ions.
  • Distillation: This separates the ammonia from the digestion mixture by raising the pH with sodium hydroxide, which changes the ammonium ions into ammonia gas. The ammonia is collected through boiling and distillation of the gas into a trapping solution of hydrochloric acid.
  • Titration: As the ammonia dissolves into the trapping solution, it is back-titrated so that the quantity of distilled-off ammonia can be calculated and the amount of nitrogen in the protein determined.

Because the initial digestion of the sample is so slow, scientists have tried to speed up the process using many different catalysts, including mercury and selenium. The addition of a neutral substance, such as potassium sulfate, raises the boiling point of the digesting acid and the temperature of the reaction, promoting acceleration. Despite these measures, the reaction remains time-consuming, inefficient, and costly on a large scale.

The Combustion Method

The combustion method involves burning a sample in an oxygen-rich atmosphere at high temperatures and analyzing the resulting gases. This process has three stages.

  • Combustion: Once the sample is weighed and purged of any atmospheric gases, it is heated in a high-temperature furnace and rapidly combusted in the presence of pure oxygen at about 1,000ºC. Cupric oxide may be used to complete the oxidation.
  • Reduction and adsorption: The combustion products—mainly carbon dioxide, water, nitrogen dioxide, and nitrogen gas—are collected and allowed to equilibrate. An aliquot of the gas mixture is passed over hot copper to remove any oxygen and catalytically convert nitrogen dioxide to nitrogen. The sample is then passed through a trap that removes carbon dioxide and water.
  • Quantitation: The total nitrogen is measured by thermal conductivity.

Historically, the original Dumas combustion method has proven time-consuming and laborious, with incomplete combustion giving nitrogen oxides and air contamination.3

A Comparison

The two methods achieve similar repeatability in a variety of feed samples, with the standard deviation for nitrogen content staying within AOAC and AACC guidelines.4 The Dumas combustion method was developed several years before the Kjeldahl technique but suffered from imprecise and inaccurate results. However, new higher quality instruments have dramatically improved the accuracy and reliability of this test.5 Analyzers based on the dynamic flash combustion of the sample, including the Thermo Scientific Flash 4000 N/Protein Analyzer, are now able to handle a wide array of laboratory requirements, such as accuracy, day-to-day reproducibility, and high sample throughput.

Alternatively, the Kjeldahl method’s universality, precision, and reproducibility had made it the internationally recognized method for estimating the protein content in foods and the standard method against which all others were judged. Unfortunately, for this method, different correction factors are needed for individual proteins to account for each protein’s unique amino acid sequence. Additional disadvantages, such as the need to use concentrated sulfuric acid at high temperature and the relatively long testing time, compare unfavorably with the combustion method for measuring crude protein content.6 With relatively recent technological advances, the combustion method now offers significant advantages over the Kjeldahl method, including a shorter analysis time—minutes as opposed to hours—along with lower costs and environmental suitability.7

Advances in Combustion Technology

New generation combustion methods are capable of taking a “sample-in, results-out” approach, operating in real time and requiring minimum sample preparation. The large sample size capacity of these instruments ensures optimum homogeneous analysis and results precision, reducing sample handling time and minimizing matrix effects. Thermal conductivity detectors feature an extensive working range to cover nitrogen/protein applications from low parts per million to high percentage concentrations. Contemporary software such as Thermo Scientific’s Eager Xperience allows 24/7 unattended operation, ensuring maximum output. And automatic carbon dioxide adsorber regenerating technology ensures that the self-cleaning filters never need to be changed, saving time and money.

A relative standard deviation of less than or equal to two for 10 successive determinations of nitrogen—as per AOAC and AACC regulations—must be achieved in order to substantiate the new combustion method tools offered.

For our experiment, we chose a variety of cereals and seeds to validate the accuracy and reproducibility of Flash 4000 N/Protein Analyzer according to AOAC and AACC guidelines with the pre-treatment of the sample. The instrument conditions were as follows:

  • Temperature left reactor 950ºC
  • Temperature right reactor 840ºC
  • Temperature oven 50ºC
  • Carrier flow (Helium) 300 ml/min
  • Reference flow (Helium) 300 ml/min
  • Standard 500 mg
  • Ethylenediaminetetraacetic acid (9.59% Nitrogen)
  • Sample weight 600 mg to 1.6 g

The Eager Xperience software automatically calculates the protein content using the default protein factor of 6.25 (5.70 for rice). The protein factor can be changed in accordance with the food type.

We determined nitrogen and protein levels in cereals samples first (see Table 1, p. 38). We homogenized barley and rice to particle sizes of 1 mm and 2 mm. The data obtained from 10 consecutive determinations showed excellent reproducibility. In all cases, the relative standard deviation was less than 2% according to official methods. We observed no memory effect when changing sample type, indicating complete detection of the nitrogen present in the sample. We observed no significant differences in the results when changing the sample particle size from 2 to 1 mm.

We then determined nitrogen and protein levels for sunflower seeds (see Table 2, below). Oil seeds, due to the sample nature of high fat content, require a proper optimization of the oxygen amount needed for combustion to obtain accurate data (see Figure 1, p. 38). We homogenized the sample at a particle size of 2 mm. The data were found to be reproducible, with an RSD percentage as per official method requirements. The left side of the table shows the reproducibility of 10 determinations using a weight of around 500 mg, while the right side of the table indicates the reproducibility of 10 determinations in a range from 700 to 1000 mg of the same sunflower sample. We observed no memory effect when changing sample weight.

The new combustion methods, combined with newer technologies like the Flash 4000, eliminate the need for sample preparation and are capable of protein determination that falls within the required accuracy determined by the AOAC and AACC guidelines. In addition, consumable costs associated with sample preparation are minimized and safety in the laboratory is increased. Improved accuracy, day-to-day reproducibility, and high sample throughput make the combustion method faster, safer, and more reliable than the Kjeldahl method.

Dr. Krotz is a product specialist, Dr. Cicerci is in analytical product development, and Dr. Giazzi is a product manager with Thermo Fisher Scientific in Milan, Italy. Reach them at For more information, call (800) 532-4752 or e-mail


  1. Kreutler P, Czajka-Narins D. Protein. In: Nutrition in Perspective. Upper Saddle River, N.J.: Prentice Hall; 1987:121–162.
  2. Mugford DC, Fox GP, Blakeney AB. Nitrogen/protein measurement by Dumas or Kjeldahl methods for Australian grains. Paper presented at: AACC Annual Meeting; October 31-November 3, 1999; Seattle.
  3. Sella A. Classic kit: Kjeldahl flask. Chem World. 2008;5(5). Available at: Accessed August 21, 2008.
  4. Sader APO, Oliveira SG, Berchielli TT. Application of Kjeldahl and Dumas combustion methods for nitrogen analysis. Arch Vet Sci. 2004;9(2):73-79.
  5. Nollet LML. Handbook of Food Analysis. 2nd ed. New York: CRC Press; 2004.
  6. McClements DJ. Analysis of proteins [lecture on University of Massachusetts Web site]. Available at: Accessed August 21, 2008.
  7. Marco A, Rubio R, Compano R, et al. Comparison of the Kjeldahl method and a combustion method for total nitrogen determination in animal feed. Talanta. 2002;57(5):1019-1026.



Current Issue

Current Issue

April/May 2014

Site Search

Site Navigation