N based methods for the

determination of protein
FOSS Laboratory Conference
September 16-17 2013
Hillerd Denmark
Jurgen.moller@bredband.net
1

Synopsis

Kjeldahl method
Dumas combustion method
NIR methods
Alternative methods (AAA & DBC)
Adulterations
Conclusions & Summary

Nitrogen-based methods do they

still have a future?
Johan Kjeldahl, 1883

Jean Baptiste Dumas,

1833

Kjeldahl - Reference

Kjeldahl from 1883

For Nitrogen / Protein

Dumas method 1980s

For Nitrogen / Protein

NIR/FTIR methods 1980s

For Crude Protein

Kjeldahl / Protein standards in OMA

920.53 (Hg)

930.33 (Cu)

920.70 (Hg)

932.08 (Hg)

920.87 (Hg)

939.02 (Hg)

920.109 (Cu)

940.25 (Hg)

920.123 (Cu)

941.06 (Cu)

920.155 (Hg)

945.01 (Hg)

920.176 (Hg)

945.18 (Hg)

925.31 (Hg)

945.23 (Hg)

928.08 (Hg)

945.48 (Cu)

930.01 (Hg)

950.09 (Hg)

930.02 (Hg)

950.10 (Hg)

930.25 (Hg)

950.48 (Hg)

930.29 (Cu)

955.04 (Hg)

962.10 (Hg)
969.37 (Hg)
976.05 (Hg)
977.02 (Hg)
978.04 (Hg)
979.09 (Hg)
981.10 (Hg)
984.13 (Cu)
988.05 (Cu/Ti)
991.20 (Cu)
2001.11 (Cu)

Recovery of Nitrogen
Lysine hydrochloride
16

100%
15

90-93%
% Protein

14

Hg
Se

13

Cu
Ti

12
11
10
10

20

30

40

50

digestion time (min)

60

70

Recovery in real samples

% recovery of protein
100
98
96
94

Se
Cu

92

Ti
90

Hg

88
86
84
Lysine

Dog food

Meat

Fishmeal

Wheat

EGN collaborative study

Method

% CP

ICC 105/2 (Kjeldahl, Cu)

12,46

ISO 5983-2 (Kjeldahl, Cu)

12,45

ISO 20483 (Kjeldahl, Cu/Ti)

12,39

ISO 16634 (Dumas/Combustion)

12,55

AAFCO PTS
Cu

Cu/TiO2
%
C
P sd %

Sample

% CP

sd %

AAFCO 0826

18,2

0,35

18,2

AAFCO 0827

12,9

0,36

AAFCO 0828

40,8

AAFCO 0829

Dumas

% CP

sd %

0,29

18,4

0,32

12,8

0,46

13,0

0,46

0,77

40,0

0,68

41,0

0,43

23,6

0,42

23,0

0,58

23,5

0,33

AAFCO 0830

18,2

0,27

18,2

0,24

18,5

0,32

AAFCO 0831

27,5

0,40

27,3

0,64

27,8

0,46

23,5

0,43

23,2

0,48

23,7

0,39

Average

Kjeltec from the 1970s

1970 introduction of block

digestion by FOSS Tecator
Since 1974 introduction of
direct steam distillation and
other improvements

Decreased use of chemicals

Improved digestion efficiency
No sample transfer
Alkali added in closed system
Distillation into boric acid
receiver, reducing the distillation
times and avoiding back titration

AOAC 2001.11

A method based on block digestion/ steam distillation /boric acid receiver

solution, having a wide scope of applicability fullfilled definitely a need of the
international laboratory society.

Method for the determination of Crude Protein in Animal Feed,

Forage, Grain, and Oilseed using Block Digestion, Copper
Catalyst and Steam Distillation into Boric Acid
Study director: Nancy J. Thiex, SD State University, Brookings, US
Study report: JAOAC, 85 (2), 2002, p 309 317
Summary: In Focus, 26 (2), 2002, p 10 - 12

Global Proteinstandard on
basis of AOAC 2001.11
EN ISO 5983-2:2005
Animal feeding stuffs Determination of nitrogen
content and calculation of crude protein content

Part 2: Block digestion/steam distillation method

ICS 65.120

EN ISO method on basis of AOAC 2001.11

Performance of EN ISO 5983-2 standard (AOAC 2001.11)

Range: 0,3 70 % protein

Samples:
1: protein block
2: swine pellets
3: corn silage
4: grass hay
5: fish meal
6: dog food
7: chinchilla feed
8: albumin
9: bird seed
10: meat and bone meal
11: milk replacer
12: soybeans
13: sunflower seed
14: legume hay

15: fish feed, small floating pellets

16: fish feed, large floating pellets
17: shrimp feed, crumble
18: shrimp feed, large sinking pellets
19: shrimp feed, small sinking pellets
20: larvae feed, flake
21: wheat grain

Validated by 24-26 international labs

Photometric and pH end point
Reference method also for Dumas
Excellent repeatability and reproducibility

Reapeatability and reproducibility

sdR and sdr vs % protein ISO 5983-2
1
0,9
0,8
0,7
sdr

0,6

sdR

0,5

Linear (sdR)

0,4

Linear (sdr)

0,3
0,2
0,1
0
0

20

40

60

80

100

EN ISO 20483:2006

Cereals and pulses

Determination of the nitrogen
content and calculation of the
crude protein content Kjeldahl
method
Block digestion with Cu/Ti
catalyst, steam distillation into
boric acid, automatic titration
with photometric (or pH) end
point determination
2 h digestion time, 20 ml acid

prEN ISO/DIS 20483:2006

Performance EN ISO 20483

1,6

1,4

1,2
y = 0,0129x + 0,061
1
sd %

sdr
sdR

0,8

Linear (sdR)
Linear (sdr)

0,6

0,4
y = 0,0058x + 0,0248
0,2

0
0

20

40

60

Protein %

80

100

ISO 89683:2004 and IDF

20-3 (2004)

Milk -- Determination of
nitrogen content -- Part
3: Block-digestion
method (Semi-micro
rapid routine method)
Joint development with
IDF and AOAC

Non-protein N and protein N

Nitrogen in milk
Total Nitrogen (AOAC 991.20 / ISO 8968-2/3 / IDF 20-2/3 )
Nonprotein Nitrogen (AOAC 991.21 / ISO 8968-4/ IDF 20-4)
Protein Nitrogen (AOAC 991.23/AOAC 991.24, ISO 8968-5, IDF 20-5 )
TCA precipitation of proteins allows a separation of non-protein Nitrogen
(Urea, Ammonia) in the filtrate from protein nitrogen (casein..) in the
filter cake

Dumas: EN ISO 16634 series

(2008)
Food products - Determination of the total

nitrogen content by combustion according to

the Dumas principle and calculation of the
crude protein content
Part 1: Oilseeds and animal feeding stuffs
Part 2: Cereals, pulses and milled cereal
products (TS)
Using the same factors as Kjeldahl
AOAC methods:
992.15 (meat)
992.23 (cereal grains & oilseeds)

Protein = Protein ?

Dumas determines total

Nitrogen, including
inorganic fractions like
NO2/NO3.
Kjeldahl determines
organic nitrogen plus
ammonia.
For many samples the
difference might be
negligible but you have
to check.

Can Dumas replace Kjeldahl ?

S. Seling et al., Max Rubner Institute, DE (2005)

Wheat harvest 2000-2004:

Some 2% of Dumas
protein is not determined
by Kjeldahl method
Kjeldahl protein =
0,959*Dumas + 0,258
Difference depends on
growing year, cultivar and
growing condition

Protein = Protein ?

Sample
French Bean
Summer Barley

Nitrate
8,9 / 6,9
0,1 / 150

Lettuce

33,2 / 9,0

Cucumber
Yam (dioscorea)
Cabbage
Spinach
Saw-dust

7,2 / 10,3
4,9 / 9,6
7,1 / 5,2
27,2 / 5,0
0,074/ 113%

Example:
33 000 mg/kg NO3 = 7,45
g N/kg = 0,75 % Nitrogen
0,75 x 6,25 = 4,7 % Protein
Conclusion:
Dumas is a routine
method
Applicability has to be
checked vs Kjeldahl

Trade conflicts ?

Argentine supplier of soymeal claims protein content of 47,2

%
Malaysian importer of soymeal claims protein content of 44,9
%
Reason: Seller uses Dumas method, buyer applies Kjeldahl
method

Can Dumas replace Kjeldahl ?

European Commission confirms the Kjeldahl method

as the community method for official controls
(Commission Regulation (EC) No 152/2009)

Kjeldahl factors

Most common: 6,25 = 16% N

Established on basis of the respective amino acid profile and the composition

Tryptophan

204.225 g/mol - 13.72% N factor 7.3

Protein factors
Amino acid

Formula

MW (g/mol)

%N

Factor

Alanine

C3H7NO2

89,09

15,71

6,37

Arginine

C6H14N4O2

174,2

32,15

3,11

Asparagine

C4H8N2O3

132,12

21,19

4,72

Aspartic acid

C4H7NO4

133,1

10,52

9,51

C3H7NO2S

121,16

11,55

8,66

C5H9NO4

147,13

9,52

10,50

C5H10N2O3

146,14

19,16

5,22

Glycine

C2H5NO2

75,07

18,65

5,36

Histidine

C6H9N3O2

155,15

27,07

3,69

Leucine

C6H13NO2

131,17

10,67

9,37

Lysine

C6H14N2O2

146,19

19,15

5,22

Methionine

C6H11NO2S

149,21

9,38

10,66

Phenylalanine

C9H11NO2

165,19

8,48

11,79

Proline

C5H9NO2

115,13

12,16

8,22

Serine

C3H7NO3

105,09

13,32

7,51

Threonine

C4H9NO3

119,12

11,75

8,51

C11H12N2O2

204,23

13,71

7,29

Tyrosine

C9H11NO3

181,19

7,73

12,94

Valine

C5H11NO2

117,15

11,95

8,37

Cysteine
Glutamic acid
Glutamine

Tryptophane

Comparison of Protein contents

Sample

Type

Kjeldahl

Dumas

Amino Acid

AAFCO 200921

Chicken

17,29 (0,15)

17,64 (0,33)

14,22 (0,17)

AAFCO 200922

Pig starter

23,94 (0,33)

24,51 (0,39)

19,73 (1,18)

AAFCO 200923

Chow

12,3 (0,52)

12,51 (0,65)

7,16 (0,19)

Sum of 18 reported AA: Alanine, Arginine, Aspartic Acid, Cystein. Glutamic

acid, Glycine, Histidine, Iso-Leucine, Leucine, Methionine, Phenylalanine,
Proline, Serine, Threonine, Tryptophane, Tyrosine, Valine
Significant differences. Sources:
Non-Protein Nitrogen (Ammonia, Urea) can be excluded.
Incomplete recovery in AAA ?

N-based methods
Method

cost/analysis
(USD)

thruput
(samples/day)

Accuracy

Applicability

Kjeldahl

1,50 - 2,50

100-200

+++

+++

Dumas

1-2

100-200

++

+++

NIR

0,1

400-500

Most widely used for (crude) protein analysis

Simple to use, with adequate accuracy and wide applicability
Probably > 50,000 installed units
Probably > 30 Mio analyses / year
NIR calibrations based on Kjeldahl or Dumas
Susceptible to adulterations

Dye Binding Method

Epsilon N

Dye

A negatively charged dye

binds to the positively
charged amine groups
Alpha-amino-N

Protein

DBC calibrated against Kjeldahl

Dye Binding Capacity
Absorbance of the
Original dye
Absorbance of dye
After filtration

DBC detects irregularities

Like high protein contents
When there is no DBC

Adulterations

Positive (allowed), e.g. urea, biuret

Negative (prohibited), e.g. melamin
At contamination levels (ppm, ppb)
At intended adulteration levels ( > 0,2-0,4 % CP)

Melamine
66% Nitrogen

Protein content = >400%

Solubility in water: 3,2 g/l

Examples for N-fractionation

Nitrogen fraction in wort and beer:

Nitrogen in malts (ale, lager, distilling)

Total nitrogen
Heat coagulable protein
HMW protein (MgSO4 precipitation)
Total Nitrogen
Soluble Nitrogen (hot water extract)

Nitrogen in animal feed / forage

Crude protein/ Urea/Biuret/NH4

ADIP / ADIN and NDIP / NDIN

.. more examples

Nitrogen in milk
TN, NPN and protein N (AOAC/ ISO / IDF )
(after TCA precipitation)

Nitrogen in eggs
Nitrogen in eggs (AOAC 925.31)
Water-soluble N and Crude Albumin (AOAC 932.08)
(pI precipation at pH 4; salting out with NaCl / EtOH)

Nitrogen in soymeal

Crude protein
Protein dispersibility index
NPN (after TCA precipitation)

Adulterants have different mid-IR signatures

Melamine dissolved in milk - FT6000

0.1

absorbance difference

absorbance difference

Urea dissolved in milk - FT6000

0 ppm
250 ppm
500 ppm
1000 ppm
1500 ppm
2000 ppm
2500 ppm
3000 ppm

0.05

1000

1100

1200
1300
wavenumber

1400

1500

absorbance difference

Ammonium sulphate dissolved in milk - FT6000

0 ppm
500 ppm
1000 ppm
2000 ppm
3000 ppm
4000 ppm
5000 ppm
6000 ppm

0.1

0.05

1000

1100

1200
1300
wavenumber

1400

1500

0.1

0.05

1600

1600

0 ppm
250 ppm
500 ppm
1000 ppm
1500 ppm
2000 ppm
2500 ppm
3000 ppm

1000

1100

1200
1300
wavenumber

1400

1500

1600

Adulterants have different spectral

signatures, making it easy to
distinguish even related
compounds
Adulterants show different
absorption at similar
concentrations detection limits
are therefore dependent on the
adulterant

FTIR: Spectral integrity / Principle Component Analysis (PCA)

0.15

Normal milk
Adulterated milk

Cyanuric acid

Score PC5

0.1

Melamine+water
(diluted samples)

0.05

Urea

Melamine

-0.05

Ammonium sulphate
-0.1

-0.5

-0.4

-0.3

-0.2

0
-0.1
Score PC2

0.1

0.2

0.3

0.4

NIR spectra of pure Melamine and skim milk powder

Melamine
Skim milk powder

NIR example: PCA plot incl

adulterated samples
Cal set
Test set 1

Test set 2

Spectral integrity

Only spectral information from normal samples is needed

No reference measurements needed

Samples, deviating from normal samples, can be identified

The LoD for melamine as unknown adulterant is 500 ppm

NIR: Validation of NIR calibration

Values in % melamine

Test set 2, 144 samples, SEP 0.05

EN ISO 12099:2010 Guidelines for the application of

near infrared spectrometry

Focuses on the validation of calibration models with

independant test sets
Defines statistics for performance measurements, i.e. SEP,
Bias, Slope, confidence limits for bias and unexplained
error
Covers also
Standardization of instruments if operated in a network
Validation on local samples before use
Instrument diagnostics at regular intervals
Checking instrument stability Control sample/chart
Running performance check of calibration during use
Control charts
Jrgen Mller, May 8, 2012

42

How good is NIRS? AAFCO - Protein

Jrgen Mller, May 8, 2012

43

Conclusions

N based methods for the determination of

protein will still be in use in the foreseeable
future (at least a couple of decades)
There are simply no alternatives that could
replace N based methods with regard to
accuracy, precision, sample throughput and cost
per analysis
NIR (and FTIR) as fast and cost effective N based
methods that can screen for adulterants
Jrgen Mller, February 7, 2012

44

Summary

Thank you for your attention.

Kjeldahl is still the most

important reference and
routine method for the
determination of crude
protein / protein fractions
New standards reflect the
instrumental and
methodological progress
Fractionation schemes to
trace adulterations
FTIR and NIR methods
have a potential as fast
screening for adulterants
Accreditation of NIR
methods using EN ISO
12099 possible