Non-destructive determination of nitrogen in malting

Research article Received: 14 March 2017

Institute of Brewing & Distilling

Revised: 22 October 2017

Accepted: 4 December 2017

Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI 10.1002/jib.477

Non-destructive determination of nitrogen in malting barleys by instrumental photon activation analysis and its comparison with the Dumas method Ivana Krausová,1 Jiří Mizera,1,2 Pavel Dostálek3* and Zdeněk Řanda1 Instrumental photon activation analysis (IPAA) was used to measure the nitrogen content of malting barleys used in brewing and breeding in the Czech Republic. The study compares the fast and non-destructive IPAA method with the classical Dumas method requiring combustion of the sample. The IPAA procedure for nitrogen assay is based on measuring the positron–electron annihilation γ line at 511 keV from the photoactivation product 13N, which is interfered with by other positron emitting nuclides. Interference from 15O can be eliminated by counting after a sufficient decay time. Interferences from 34mCl and 38K can be corrected via measuring their specific γ lines at 146.4 and 2167.7 keV, respectively. The IPAA procedure has been verified by analysis of several biological reference materials and is suitable for nitrogen measurement in cereals and other plant materials with a detection limit down to 102 wt%. Nitrogen contents in malting barleys ranged from 1.4 to 2.2 wt% and were dependent on the cereal variety and growth conditions (fertilizer use and chemical protection). Copyright © 2018 The Institute of Brewing & Distilling Keywords: nitrogen; instrumental photon activation analysis; Dumas method; malting barley

Introduction Nitrogen is an important biogenic element, the basic component of amino acids and is present in all living organisms and plants. The nitrogen content as a measure of protein content is an important parameter in brewing and distilling, which is measured at various stages throughout the production process. Determination of total nitrogen is routinely carried out mostly by two classical chemical methods, the Kjeldahl method, introduced in 1883 (1), and the Dumas method, proposed by its author about 50 years earlier (2). The Kjeldahl method is based on digestion of the sample with sulphuric acid. The organic matter is decomposed by oxidation and organic nitrogen is reduced to ammonium sulphate, which is distilled with sodium hydroxide to liberate gaseous ammonia. The ammonia is determined by direct or back titration. In the case of direct titration, the ammonia gas is captured by boric acid, forming an ammonium borate complex, which is neutralized by a standardized mineral acid. In the case of back titration, the ammonia is captured by a standardized mineral acid and the excess acid is titrated with a standard alkaline base solution (3–5). Nitrogen assay by the Dumas combustion method requires combustion of sample in the presence of oxygen at 800–1000°C, which converts all forms of nitrogen to nitrogen oxides. The nitrogen oxides are catalytically reduced to nitrogen gas, which is measured by a thermal conductivity detector. The Kjeldahl method has been used for almost 100 years and is still is the leading standard reference method for the determination of total nitrogen in feeds and food. It has several disadvantages in comparison with the Dumas method. Inorganic forms of nitrogen such as nitrates and nitrites are not usually measured in the Kjeldahl method so that it gives slightly lower values than the Dumas method. Other disadvantages of the Kjeldahl method

J. Inst. Brew. 2018

are the time-consuming and multiple steps of the analytical procedure, together with the use of corrosive and toxic chemicals. Conversely, the Dumas method has a great potential for accuracy, speed of analysis and automation. Since 1990s, the Dumas method has been collaboratively tested and accepted as an alternative method to the Kjeldahl method for determination of total nitrogen in barley, malt and beer by the European Brewery Convention, the Institute of Brewing, and the American Society of Brewing Chemists (3–8). Using both conventional chemical methods entails a risk of nitrogen loss during decomposition, dissolution and chemical treatment of each sample to transform nitrogen into the desired detectable form. This risk can be avoided if the assay is based on non-destructive (in an instrumental mode) activation methods, allowing in vivo measurements in biological matrices. The nitrogen assay using instrumental neutron activation analysis (INAA) is based on irradiation with fast neutrons from neutron generators utilizing the reaction 14N(n, 2n)13N (9,10). Prompt γ-ray neutron

* Correspondence to: Pavel Dostálek, University of Chemistry and Technology, Prague, Faculty of Food and Biochemical Technology, Department of Biotechnology, Technická 5, 166 28 Prague 6-Dejvice, Czech Republic. E-mail: pavel. [email protected] 1

The Czech Academy of Sciences, Nuclear Physics Institute, Hlavní 130, 250 68, Husinec-Řež, Czech Republic

2

The Czech Academy of Sciences, Institute of Rock Structure and Mechanics, V Holešovičkách 41, 182 09, Prague 8, Czech Republic

3

Faculty of Food and Biochemical Technology, Department of Biotechnology, University of Chemistry and Technology, Prague, Technická 5, 166 28, Prague 6-Dejvice, Czech Republic

Copyright © 2018 The Institute of Brewing & Distilling

I. Krausová et al.


activation analysis utilizing the reaction 14N(n, γ)15N has often been used (11–14). Although another possibility is activation with charged particles (15–18). Photon activation analysis, in the present work used in an instrumental mode (IPAA), was one of the first nuclear radioanalytical methods used for nitrogen assay (19–23). The IPAA determination of nitrogen utilizes the reaction 14N(γ, n)13N. The reaction product 13N, the same analytical radionuclide as in the case of INAA determination, is a pure positron emitter, which is detectable only via the 511 keV γ rays from the electron–positron annihilation. The annihilation radiation is nonspecific and can be interfered with by radiation emitted simultaneously from a number of other positron-emitting radionuclides produced besides 13N upon bremsstrahlung irradiation of a sample. Characteristic nuclear parameters of 13N and the radionuclides potentially interfering with its detection in the nitrogen assay by IPAA are presented in Table 1. The aim of this study was to design and optimize an IPAA procedure for assaying nitrogen in biological materials, especially cereals, using the MT-25 microtron for sample irradiation. For testing experimental parameters such as maximum beam energy, irradiation – decay – counting times and interference corrections, assessing detection limits and uncertainties of the assay. The biological reference materials that were used in this work included RM 8433 Corn Bran, SRM 1547 Peach Leaves, RM 8414 Bovine Muscle Powder, and SRM 1577b Bovine Liver from the National Institute of Standards and Technology (NIST). The procedure was then applied in the analysis of malting barleys used in brewing and breeding in the Czech Republic, with RM 8433 and SRM 1547 used for quality control.

Experimental Preparation of malting and new barley varieties Samples 14 varieties of malting barley of in two groups grown under different field conditions, were obtained for analysis. Differences between the two sets identified as T and UT (treated and untreated, respectively) were in the application of fertilizer and chemical protection during growth. Nitrogen content in the barleys was determined previously by the Dumas method; however, the previous analysis by the Dumas method and the present analysis by IPAA analysis were not carried out with identical samples (i.e. taken from one batch of grains after milling and homogenization), but with two separate batches taken as whole kernels from one crop. For the IPAA procedure, kernels (ca. 15 g) were ground in an electric grinder with a stainless steel blade, then pulverized in an agate ball-mill, and finally homogenized in a polyethylene bottle rotating along two axes (Pulverizette 5, Fritsch, Germany). The pulverized samples (ca. 1.5 g) were pelletized (pellet diameter 27 mm, thickness 3 mm), and heat sealed into disc-shaped capsules made from acid-cleaned, high-purity polyethylene foil.

Standards and monitors Calibration standards for nitrogen calibration and for interference corrections were prepared from suitable amounts of stoichiometrically defined compounds (NH4SCN, KI, NaCl) corresponding to 60–100 mg of a respective element mixed with starch (ca. 1.2–1.4 g). Copper foils were used as photon dose monitors.

Irradiation in microtron and activity counting Photon irradiation was carried out in the microtron MT-25 of the Nuclear Physics Institute ASCR. The microtron is a radio frequency cyclic electron accelerator. The high-energy photon radiation (bremsstrahlung) is obtained by braking accelerated electrons on a tungsten target (25). Two to four samples (including reference materials for quality control), calibration (and correction) standards and the copper monitors placed between every two samples or standards, stacked as a ‘sandwich’ into a PE container, were exposed to bremsstrahlung radiation positioned in the beam axis. Sample distance from the tungsten converter ranged from 5 to 11 cm. The irradiation time of 10 min was chosen as optimum based on half-lives of both the an13 alytical radionuclide N and the interfering radionuclides. The electron beam energy (i.e. the maximum photon energy) was set at 17 MeV based on optimization carried out for fluorine assay (26), which utilizes a reaction with a similar threshold energy (cf. in Table 1). The beam current was held at 10 μA, the maximum attainable under the given microtron conditions and settings. To eliminate the presence and activation of the atmospheric nitrogen, the sample container was placed in a chamber evacuated to -4 2–5 × 10 Pa using a membrane vacuum pump. For counting, a coaxial HPGe detector (23% relative efficiency, FWHM 60 resolution of 1.8 keV at 1332.5 keV line of Co) coupled to a Canberra Genie 2000 γ-spectrometric system was used. Samples were counted for 10– 20 min (depending on decay time) in the same counting geometry, with the first count starting 15–20 min after the end of irradiation to ensure 15 sufficient decay of interfering O. The NH4SCN calibration standard was counted for 10 min immediately after the second sample. The correction standards (necessary only for potassium and chlorine) were irradiated and counted only in the design-optimization stage to determine the correction factor. The KI standard was counted for 5 min before counting samples, and the NaCl standard for 10 min after counting samples.

Results and discussion It could be expected that the highest levels of interference in nitrogen determination in biological samples will be caused by the presence of carbon and oxygen, which are major constituents and their photoactivation products, radionuclides 11C and 15 O are, just as 13N, pure positron emitters. However, production of these radionuclides begins at significantly higher threshold energies than 13N production. In the case of 11C, irradiation below the threshold energy ensures complete elimination of

Table 1. Radionuclides and photonuclear reactions considered in nitrogen assay by instrumental photon activation analysis (IPAA) and their characteristic nuclear parameters (24) Radionuclide 13

N C 15 O 18 F 34m Cl 38 K 45 Ti 53 Fe 11

wileyonlinelibrary.com/journal/jib

Half-life

Nuclear reaction

Threshold (MeV)

Main photons (keV)

9.97 m 20.4 m 2.04 m 1.83 h 32.2 m 7.64 m 3.08 h 8.51 m

14

10.55 18.72 15.67 10.43 12.78 13.07 13.19 13.62

511 511 511 511 511; 146.4 511; 2167.7 511; 719.6 511; 377.9

13

N (γ, n) N C (γ, n) 11C 16 O (γ, n) 15O 19 F (γ, n) 18F 35 Cl (γ, n) 34mCl 39 K (γ, n) 38K 46 Ti (γ, n) 45Ti 54 Fe (γ, n) 53Fe 12


J. Inst. Brew. 2018

Non-destructive determination of nitrogen in malting barleys the interference. In the case of 15O, elimination is only partial because a beam energy >16 MeV is required for a higher yield of 13 N. However, if activity counting is performed after a sufficiently long decay time, the significantly shorter half-life of 15O compared with 13N ensures adequate elimination of the interference. Another pure positronic emitter, 18F, has its production threshold energy very close to 13N, but owing to a significantly longer halflife, it has a significantly lower production rate. Moreover, fluorine in biological materials is usually present at much lower levels than nitrogen. For the same reasons, interference from 45 Ti is negligible. Owing to high contents of potassium, and also relatively high contents of chlorine (and possibly iron) in biological matrices, with respect to nuclear parameters of their radioisotopes 38K, 34mCl and 53Fe produced by photoactivation, their possible interference should also be considered. However, these radionuclides are not pure positron emitters and corrections can be done based on activities measured for their specific γ-lines. See Table 1 for nuclear parameters of the above-discussed interfering radionuclides. Correction of the above interferences consists of calculation of a proportion of 13N in the 511 keV peak area by subtracting from the


total 511 keV peak area counted in an analysed sample all significant interference contributions: P511;13 N ¼ P511; tot ∑CF i Pspecific; i

(1)

i

The correction factor CF for interference from a given interfering nuclide is obtained as a ratio of the 511 keV peak area (P511) to the specific peak area (Pspecific), both obtained from counting a respective correction standard with the same detector and counting geometry as the nitrogen calibration standard and analysed samples: CF ¼ P511 =Pspecific

(2)

The interference corrections included 34mCl and 38K, for which the specific peaks at 146.4 and 2167.7, respectively, were detected in most samples. Interference from 53Fe was ignored as the specific line at 377.9 keV was not detected in any of the reference materials or samples analysed. The IPAA method was verified by analysis of NIST biological reference materials (SRM 1547 Peach Leaves, RM 8433 Corn Bran, RM 8414 Bovine Muscle Powder, SRM 1577b Bovine Liver),

Table 2. Determination of total nitrogen in biological reference materials by IPAA Reference material

IPAA value ± 2uca (wt%)

Reference value ± ub (wt%)

NIST RM 8433 Corn Bran NIST SRM 1547 Peach Leaves NIST RM 8414 Bovine Muscle Powder NIST SRM 1577b Bovine Liver

0.871 ± 0.020 2.94 ± 0.13 12.32 ± 0.70 10.13 ± 0.55

0.882 ± 0.027 2.94 ± 0.12 13.75 ± 0.32 10.60 (non-certified value)

a

Expanded combined standard uncertainty combining counting statistics with marginal sources of uncertainty such as weighing and fluctuations in irradiation or counting geometries. b Uncertainty defined in the Certificate of Analysis.

Figure 1. Comparison of nitrogen assay in malting barleys grown in neighbouring fields under different conditions (T, treated; UT, untreated) by instrumental photon activation analysis (IPAA) and the Dumas method. The error bars for the IPAA data represent an expanded (k = 2) combined standard uncertainty and for the Dumas data a typical reproducibility [0.116 wt% according to Johansson (6)]. [Colour figure can be viewed at wileyonlinelibrary.com]

J. Inst. Brew. 2018



I. Krausová et al.


repeated three times. The results, presented in Table 2, agreed with the certified values within their uncertainties except for RM 8414 where IPAA provided a result lower by ~10 rel.%. The detection limits for the nitrogen assay by IPAA was assessed using the Currie MDA (minimum detectable activity) algorithm (27) to be 10-2 wt% in plant tissues and 101 wt% in animal tissues. Results of the IPAA determination of nitrogen in samples of new varieties of malting barley, grown under different field conditions (treated and untreated), are illustrated in Fig. 1. For comparison, results obtained using the Dumas method are presented. The comparison of both methods is illustrated also by a correlation scatter plot and a Bland–Altman plot (28) presented in Fig. 2. The mass fractions of nitrogen in malting barleys, as determined by IPAA, ranged from 1.4 to 2.2 wt% in the treated group and from 1.4 to 2.0 wt% in the untreated group. The nitrogen mass fractions obtained by the reference Dumas method ranged from 1.7 to 2.1 wt% in the treated group and from 1.6 to 2.0 wt% in the untreated group. Nitrogen contents determined by IPAA were significantly lower than those obtained by the Dumas

method. The difference was 0.16 wt% on average, and the largest difference observed was 0.44 wt%. The difference was statistically highly significant (p = 0.000044 at the paired t-test). The correlation scatter plot in Fig. 2 shows quite a strong positive correlation (r = 0.66) between the datasets from both methods. The negative deviation of the IPAA data from the line of equality is pronounced, particularly for lower values of the measured total nitrogen range, which is evident also from the difference (Bland–Altman) plot in Fig. 2. With IPAA, quality control was provided by analysis of the NIST reference materials SRM 1547 Peach Leaves and RM 8433 Corn Bran; alternatively one of the reference materials was irradiated and analysed in each irradiation series, resulting in a satisfactory agreement with the reference values. For the results obtained by the Dumas method, information on quality control as well as on the uncertainty was not provided by the analytical laboratory. The observed difference, however, should be ascribed to different moisture contents rather than to a systematic error in determinations by one of the methods, because the samples analysed by INAA were not additionally dried after milling, contrary to the samples analysed by the Dumas method. Moisture contents were not provided with samples, but an average value of 10 wt% is common for whole kernels dried and stored at ambient temperatures and relative humidities (8). Further, sample suites analysed by both methods were not completely identical, i.e. both analytical laboratories obtained two batches of whole kernels sampled from one crop. The differences in results obtained by IPAA and the Dumas method, which would remain after accounting for different moisture contents, could result from nitrogen variation in single kernels (8). Nitrogen fertilization can affect growth, development, dry matter, and accumulation and partitioning of nitrogen in barley, which has an effect on the grain yield. In production of malting barley, however, fertilization strategies must be optimized to balance between maximizing grain yield and achieving low nitrogen content in grain (29,30). In the present study, systematically slightly higher nitrogen contents in treated barleys were indicated by both analytical methods. Nitrogen content in treated varieties of malting barley determined by IPAA and the Dumas method were on average 0.091 and 0.066 wt% higher, respectively, than in untreated barleys. The differences, however, are comparable with the assay uncertainty. Differences in nitrogen content among individual samples reflecting different barley variety and growing conditions were observed by both analytical methods with generally similar trends. Generally, more pronounced differences among samples were observed by IPAA.

Conclusions

Figure 2. Comparison of IPAA and the Dumas method used for nitrogen assay in malting barleys with a correlation scatter plot (top) and a Bland–Altman plot (bottom). [Colour figure can be viewed at wileyonlinelibrary.com]


The results of the determination of nitrogen in biological reference materials and malting barley using IPAA with short-time irradiation demonstrate the applicability of this method for nitrogen assay in cereals and other biological materials. There is no doubt that IPAA cannot compete with standard techniques such as the Dumas and Kjeldahl methods in terms of speed, cost and precision. However, as an independent, non-destructive technique based on a completely different (nuclear vs chemical) analytical principle, it may have an important role in, for example, the quality assurance process, collaborative trials or certification of new reference materials, similar to other nuclear methods such as prompt γ-ray neutron activation analysis.


J. Inst. Brew. 2018

Non-destructive determination of nitrogen in malting barleys

Acknowledgements This work was supported by the project P108/12/G108 of the Czech Science Foundation (Grantová Agentura České republiky) and TE02000177 ‘Centre for Innovative Use and Strengthening of Competitiveness of Czech Brewery Raw Materials and Products’ of the Technology Agency of the Czech Republic (Technologická Agentura České republiky).

References 1. Kjeldahl, J. (1883) A new method for the determination of nitrogen in organic matter, Z. Anal. Chem. 22, 366–382. 2. Dumas, J. B. A. (1831) Lettre de M. Dumas à M. Gay-Lussac sur les Procédés de l’analyse organique, Ann. Chim. Phys. 47, 198–213. 3. Jung, S., Rickert, D. A., Deak, N. A., Aldin, E. D., Recknor, J., Johnson, L. A., and Murphy, P. A. (2003) Comparison of Kjeldahl and Dumas methods for determining protein contents of soybean products, J. Am. Oil Chem. Soc. 80, 1169–1173. 4. Thompson, M., Owen, L., Wilkinson, K., Wood, R., and Damant, A. (2002) A comparison of the Kjeldahl and Dumas methods for the determination of protein in foods, using data from a proficiency testing scheme, Analyst 127, 1666–1668. 5. Brandstreet, R. B. (1965) The Kjeldahl Method for Organic Nitrogen, Academic Press, New York. 6. Johansson, C. G. (1996) Determination of total nitrogen in barley and malt by combustion method – Collaborative trial, J. Inst. Brew. 102, 249–250. 7. Buckee, G. K. (1994) Determination of total nitrogen in barley, malt and beer by Kjeldahl procedures and the Dumas combustion method – Collaborative trial, J. Inst. Brew. 100, 57–64. 8. Angelino, S. A. G. F., van Laarhoven, H. P. M., van Westerop, J. J. M., Broekhuijse, B. M., and Mocking, H. C. M. (1997) Total nitrogen content in single kernel malting barley samples, J. Inst. Brew. 103, 41–46. 9. Nagai, T., Fugii, I., Muto, H., and Inouye, T. (1969) Total-body nitrogen and protein determined by in vivo fast-neutron activation analysis, J. Nucl. Med. 10, 192–196. 10. Brune, D., and Arroyo, A. (1971) Determination of the nitrogen content of grains by means of fast neutron activation analysis, Anal. Chim. Acta 56, 473–476. 11. Tiwari, P. N., Bergman, R., and Larsson, B. (1971) Determination of nitrogen in organic materials by prompt (n, γ) technique, Int. J. Appl. Radiat. Isot. 22, 587–588. 12. Beddoe, A. H., Zuidmeer, H., and Hill, G. L. (1984) A prompt gamma in vivo neutron activation analysis facility for measurement of total body nitrogen in the critically ill, Phys. Med. Biol. 29, 371–383. 13. Stamatelatos, I. E., Dilmanian, F. A., Ma, R., Lidofsky, L. J., Weber, D. A., Pierson, R. N. Jr., Kamen, Y., and Yasumura, S. (1993) Calibration for measuring total body nitrogen with a newly upgraded prompt gamma neutron activation facility, Phys. Med. Biol. 38, 615–626.

J. Inst. Brew. 2018


14. Im, H. J., Cho, H. J., Song, B. C., Park, Y. J., Chung, Y. S., and Kim, W. H. (2006) Analytical capability of an explosives detection by a prompt gamma-ray neutron activation analysis, Nucl. Instrum. Meth. A 566, 442–447. 15. Johansson, A., Larsson, B., Tibell, G., and Ehrenberg, L. (1969) Possibilities of nitrogen determination in seeds by direct nuclear reactions, new approaches to breding for improved plant protein, Gustaf Werner Inst., Uppsala. 16. Sundqvist, B., Gönczi, L., Bergman, R., and Lindh, U. (1974) A fast method for nitrogen determination in single seeds, Int. J. Appl. Radiat. Isot. 25, 277–281. 17. Gönczi, L., Berggren, H., Didriksson, R., and Sundqvist, B. (1978) The use 14 15 of the N(d, p0) N reaction for studies of nitrogen depth distribution in grains of barley and in samples of chromium steel, Nucl. Instrum. Meth. 149, 337–340. 18. Strijckmans, K., Mortier, R., Vandecasteele, C., and Hoste, J. (1982) The determination of nitrogen in zirconium by charged-particle activation analysis, Microchim. Acta 78, 321–331. 19. Baker, C. A. (1967) Gamma-activation analysis, The Analyst 92, 601–610. 20. Chapyzhnikov, B. A., Evzhanov, K. N., Malikova, E. D., Kunin, L. L., and Samosyuk, V. N. (1972) Determination of nitrogen in lithium and sodium by photon activation, Radiochem. Radioanal. Lett. 11, 269–273. 21. Brune, D., Düring, L., Larsson, B., and Lundqvist, H. (1973) Analysis of nitrogen in grain by photonuclear activation analysis, Int. J. Appl. Radiat. Isot. 24, 1–6. 22. Segebade, C., Weise, H. P., and Lutz, G. J. (1988) Photon Activation Analysis, Walter de Gruyter, Berlin. 23. Görner, W., Berger, A., and Jost, P. (1998) Precise determination of macro contents of nitrogen in titanium carbonitride by instrumental photon activation analysis, J. Radioanal. Nucl. Chem. 234, 91–94. 24. Řanda, Z., and Kreisinger, F. (1983) Tables of nuclear constants for gamma activation analysis, J. Radioanal. Chem. 77, 279–495. 25. Krist, P., Horák, Z., Mizera, J., Chvátil, D., Vognar, M., and Řanda, Z. (2015) Innovation at the MT-25 microtron aimed at applications in photon activation analysis, J. Radioanal. Nucl. Chem. 304, 183–188. 26. Krausová, I., Mizera, J., Řanda, Z., Chvátil, D., and Krist, P. (2015) Nondestructive assay of fluorine in geochemical and others materials by instrumental photon activation analysis with a microtron, Nucl. Instrum. Meth. B 342, 82–86. 27. Currie, L. A. (1968) Limits for qualitative detection and quantitative determination. Application to radiochemistry, Anal. Chem. 40, 586–593. 28. Bland, J. M., and Altman, D. G. (1999) Measuring agreement in method comparison studies, Stat. Meth. Med. Res. 8, 135–160. 29. Baethgen, W. E., Christianson, C. B., and Lamothe, A. G. (1996) Nitrogen fertilizer effects on growth, grain yield, and yield components of malting barley, Field Crops Res. 43, 87–99. 30. Dordas, D. (2012) Variation in dry matter and nitrogen accumulation and remobilization in barley as affected by fertilization, cultivar, and source–sink relations, Eur. J. Agron. 37, 31–42.

