Dumas – A well-established method for N/protein analysis
The determination of the total protein content is an essential tool for quality control and protein declaration according to international labeling laws in the food & feed industry and research facilities. Protein content can directly correspond to product properties and classifications e.g. dough properties, foam formation, the taste of beer, or the differentiation between starch and gluten-free starch. In all application areas highly precise, matrix-independent protein analyses are required.
In principle, protein quantification is achieved via a series of specific and non-specific physical and chemical reactions followed by suitable detection. Full method automation as well as accordance to industry standards is generally desirable in current laboratory operations, which has led to two different widely accepted primary methods for the determination of total protein content: the wet chemical method according to Kjeldahl and the high-temperature combustion method according to Dumas. As a secondary analysis method near infrared spectroscopy (NIRS) can be found, however, this method requires a primary method for calibration purposes.
For more than 100 years the Kjeldahl principle was the most commonly used method and described in the majority of standards for the determination of total protein content of food products. As a wet chemical analysis it is time consuming, labor-intensive and requires hazardous and toxic chemicals. All this is nowadays unwanted with regard to laboratory safety but also for economic reasons. This explains the trend in recent years that the Kjeldahl principle is more and more displaced by the Dumas principle.
Combustion versus wet chemistry
The Dumas principle relies on quantitative conversion of the sample into well-defined gaseous species at 950°C in an oxygen enriched environment. During the combustion phase, all nitrogen in the sample is converted to nitrogen oxides, which are reduced to nitrogen gas and quantified with a thermal conductivity detector. All other combustion gases, including excess oxygen, are trapped or absorbed prior to the quantification. The analysis time can be as fast as 4 minutes. Figure 1 shows the functional principle of Elementar’s next generation N/protein analyzer rapid N exceed, which uses the EAS REGAINER® technology for a significantly reduced cost per analysis. In the first step of the Kjeldahl principle the sample is digested using concentrated sulfuric acid and a catalyst at 420°C for 90 minutes. During this process the nitrogen in the sample is converted to ammonium sulfate. In a second step sodium hydroxide is added to the ammonium solution. The released ammonia is distilled off and trapped in a boric acid solution. The ammonium borate complex is finally titrated with sulfuric acid or hydrochloric acid. These two steps together require approximately 5 minutes. With the initial digestion step including the necessary heating and cooling time, a complete analysis will need at least 100 minutes.
Comparability of methods
Although the principles of the Kjeldahl and Dumas methods are different, both determine the total nitrogen content of the sample. For food samples it is known that nitrogen and protein content have a linear relationship. By using a suitable factor, the nitrogen concentration can therefore be converted to protein content. The factor varies depending on the relative amount of different proteins and their amino acid composition (see Table 1). Both methods do not differentiate between protein and non-protein nitrogen. In most cases, the results obtained by the Dumas method are slightly higher compared to the Kjeldahl method. This is due to a better nitrogen recovery of the Dumas method compared to Kjeldahl. There are numerous scientific studies published comparing the Kjeldahl and Dumas methods for a wide range of different sample types. Taking into account that the calculated protein content in both methods is only based on approximations, the careful selection of the calculation factor is of utmost importance. A conversion between Kjeldahl and Dumas values using appropriate coefficients is possible and suggested by Simonne et al.  for selected food groups. However, in many cases the difference between both methods is not statistically significant as was shown e.g. by Željko et al. 
Throughput and amount of work
For Dumas analyses, sample preparation is simple and fast. Samples are weighed into tin cups, wrapped and placed in the autosampler of the rapid N exceed. Or, in case of the rapid MAX N exceed, the samples are weighted into reusable stainless steel crucibles and placed in the autosampler. In both cases the sample preparation requires less than a minute before the first measurement can be started. With up to 90 sample positions in case of the rapid MAX N exceed, unattended runs of more than 7 hours are possible. Hence, up to 200 samples can be measured in a routine daily operation.
In contrast, the sample preparation for Kjeldahl analyses not only requires weighing of the sample, but also the addition of an exact amount of sulfuric acid solution and catalyst tablets. As already mentioned, the complete analysis will take at least 100 minutes. It is possible to increase the daily throughput of a Kjeldahl analyzer by batch processing. This approach has an upper limit of 20 samples per batch, because at this number the time necessary to perform the distillation/titration step for the whole batch equals the time for the digestion step. This sets the upper sample limit at 100 per day for a Kjeldahl analyzer (see Figure 1). Although both steps are automated, a manual sample transfer between the two modules is inevitable. Therefore, the unattended operation time is only 100 minutes.
Cost per analysis
Traditionally, Dumas analyzers use heated metals such as copper or tungsten to bind excess oxygen and to reduce formed nitrogen oxides to N2. The reduction metal is oxidized to inactive metal oxide typically after 200 - 300 samples when considering whole gas analysis. Therefore, the reduction metal together with the carrier gas are the main price drivers. Elementar’s proprietary EAS REGAINER technology increases the reduction metal lifetime significantly and lowers maintenance by a factor of up to five. Hence, the cost per analysis based on recommended end user prices amounts to approximately 0.25€, including all required reagents and gases in case of the rapid N exceed and 0.49€ for the rapid MAX N exceed.
The main price drivers in Kjeldahl analyses are the costs for chemicals and their disposal and manpower, resulting in an overall cost per analysis of approximately 6.00€. The incurred chemical waste from a typical Kjeldahl methodology for 2000 samples amounts to 560 liter compared to a negligible amount of only 430 g solids for the rapid N exceed, as presented in Table 2. The EAS REGAINER reagent is exhausted during the analysis and doesn’t form any solid waste. It has to be noted that the labor costs are not yet included in the calculations. Using the Dumas method, twice as much samples can be analyzed per day with just 2/3 of the working time required compared to Kjeldahl.
Our N/Protein Analyzers
Our N/protein analyzers use the high-temperature combustion method according to Dumas, which has clear advantages over Kjeldahl regarding laboratory safety, sample throughput, labor time, amount of chemical waste and thus cost-per-analysis. Our analyzers for the determination of nitrogen and protein are dedicated instruments serving today’s customer needs in regards to price-per-sample, throughput and sensitivity.
 Simonne, A.H., Simonne, E.H., Eitenmiller, R.R., Mills, H.A., Cresman III, C.P. 1997. Could the Dumas method replace the Kjeldahl digestion for nitrogen and crude protein determinations in foods? Journal of the Science of Food and Agriculture, 73:39-45.
 Željko, A.M., Sandra M.J., Nadežda, B.P., Željko, N.Ć., Milica, M.Ž., Comparison of the Kjeldahl method, Dumas method and NIR method for total nitrogen determination in meat and meat products Journal of Agroalimentary Processes and Technologies 2015, 21(4), 365-370.