Professional Documents
Culture Documents
2, 2002 309
AGRICULTURAL MATERIALS
Collaborators: S. Anderson; E. Bogren; G. Bolek; D. Budde; C. Ellis; S. Eriksson; G. Field; E. Frankenius; C. Henderson;
C. Henry; M. Kapphahn; L. Lundberg; H. Manson; J. Moller; M. Russell; J. Sefert-Schwind; M. Spann
A collaborative study was conducted to evaluate which titration begins automatically as soon as
the repeatability and reproducibility of an exten- distillation starts. The Study Directors recommend
sion of AOAC Official Method 991.20, Nitrogen that 1% H3BO3 as an optional alternative to 4% bo-
(Crude) in Milk, to animal feed, forage (plant tis- ric acid trapping solution be allowed for automatic
sue), grain, and oilseed materials. Test portions titrators that titrate throughout the distillation.
are digested in an aluminum block at 420°C in sul-
furic acid with potassium sulfate and a copper cat-
alyst. Digests are cooled and diluted, and concen-
he Danish chemist Johan Kjeldahl (1849–1900) devel-
trated sodium hydroxide is added to neutralize the
acid and make the digest basic; the liberated am-
monia is distilled by using steam distillation. The
liberated ammonia is trapped in a weak boric acid
T oped what today is well known as the Kjeldahl method
for determining nitrogen in organic substances. The
original method as presented by Kjeldahl has been continu-
solution and titrated with a stronger standardized ously improved. These developments have improved environ-
acid, hydrochloric acid; colorimetric endpoint de- mental and personal safety aspects, increased the speed and
tection is used. Fourteen blind samples were sent versatility of the method, and simplified the entire analytical
to 13 collaborators in the United States, Denmark, procedure.
Sweden, Germany, and the United Kingdom. Re- The Kjeldahl method has 3 different steps: digestion, distil-
coveries of nitrogen from lysine, tryptophan, and lation, and titration. Traditionally Kjeldahl flasks with a capac-
acetanilide were 86.8, 98.8, and 100.1%, respec- ity for 500–800 mL and gas or electric heating have been used
tively. The within-laboratory relative standard devi- for the digestion. For distillation, the addition of water and al-
ation (RSDr, repeatability) ranged from 0.40 to kali to the digested sample has been followed by the heating of
2.38% for crude protein. The among-laboratories the flask in order to distill >150 mL distillate. The distilled am-
(including within-) relative standard deviation monia has been captured in standardized acid, and back titration
(RSDR, reproducibility) ranged from 0.44 to 2.38%. with standardized sodium hydroxide has been used.
It is recommended that the method be adopted In 1970, the Swedish chemist Roger Mossberg introduced
First Action by AOAC INTERNATIONAL. A lower the modern concept of using a block digestor, which de-
concentration (1% H3BO3) of trapping solution was creased the use of chemicals and improved the efficiency of
compared with the concentration specified in the the digestion. The method was further improved in 1974 by
original protocol (4% H3BO3) and was found com- the introduction of direct steam distillation, which drastically
parable for use in an automatic titration system in decreased the time needed for distillation. Because the distil-
late was captured in a boric acid receiver, the need to standard-
ize 1 reagent was eliminated.
Submitted for publication September 2001. Today, this method is universally accepted and used in tens
The recommendation was approved by the Methods Committee on of thousands of laboratories throughout the world. However,
Feeds, Fertilizers, and Related Agricultural Topics as First Action. See
“Official Methods Program Actions,” (2001) Inside Laboratory when it comes to the official status of the method, the situation
Management, November/December issue. is somewhat complicated. In selected cases, official standard
Corresponding author’s e-mail: nancy_thiex@sdstate.edu. methods are available, describing the Kjeldahl method (Inter-
310 THIEX ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 2, 2002
national Organization for Standardization, Deutsches Institut Table 1. Results (%) of the comparability studies of the
fur Normung, American National Standards Institute, Ameri- copper/boric acid Kjeldahl method
can Society for Testing and Materials, and AOAC INTER- Proposed AOAC AOAC
NATIONAL). A review of the available AOAC procedures Sample method 976.06Ha 990.03b
for the Kjeldahl method shows that the laboratory that wants
Nitrogen
to follow an official procedure closely has a problem when
trying to use this method. Basically it is approved only for
dairy products and meat and meat products (1). For most other Lysine 14.99 15.266
sample types, traditional Kjeldahl flask digestion followed by Tryptophan 13.58 13.67
back titration of standard acid or block digestions followed by Nicotinic acid 2.37 3.04
colorimetric detection are still the only approved Kjeldahl Crude protein
methodologies (2–9). For most agricultural sample types, the
use of copper catalysts instead of mercury catalysts has been Protein block 39.74 39.44 40.53
A method based on block digestion/steam distillation/boric Chinchilla food 18.19 18.14 18.01
acid trapping solution having a wide scope of applicability to Corn silage green chop 7.26 7.17 6.95
agricultural products would definitely fulfill a need in the in- Legume hay 18.92 18.91 18.78
ternational laboratory community. Swine pellets 37.32 37.05 37.26
Sunflower seed 16.55 17.05 17.19
Comparability Study Fish meal 64.48 64.24 65.40
Avg. for crude protein 31.37 31.11 31.37
Results of comparability studies of the proposed method and
methods having AOAC Official status are provided in Table 1. a
Ref. 3.
b
Ref. 10.
Collaborative Study
The Study Director with the cooperation of 9 additional Study materials were chosen to be representative of differ-
collaborating laboratories conducted the collaborative study. ent animal feeds, pet foods, feed ingredients, forage, cereal
These laboratories represented a variety of laboratory types, grains, and oilseeds. The categories that each material repre-
including research laboratories, commercial laboratories, sents are provided in Table 2. The samples were also chosen to
manufacturers/industry laboratories, and state regulatory lab-
be robust with respect to carbohydrate and fiber concentra-
oratories. Four laboratories were from outside the United
tions, with respect to species, and with respect to ingredients.
States (2 Swedish, 1 Danish, and 1 Scottish). Participants re-
All materials used for the study were natural or “real world”;
ceived no compensation.
none were spiked. The nitrogen content of the samples ranged
Collaborating laboratories provided the digestion block,
from 1 to about 15% (7% crude protein to about 80% equiva-
digestion block tubes, distillation unit, burette/titrator, sulfuric
lent crude protein). Samples were coded at random so that
acid, and sodium hydroxide. Collaborators were provided
there was no preselection from the order of presentation. Ap-
with sufficient portions of the test samples, Kjeldahl catalyst,
boric acid trapping solution, standardized hydrochloric acid proximately 20 g of each sample material was supplied, which
titrant, and weighing paper. was more than the amount needed to complete the study.
Collaborating laboratories were asked to analyze 14 animal The chinchilla food, birdseed, and swine pellets were
feed, pet food, and feed ingredient materials (including for- ground to pass through a 0.75 mm screen in a Retsch ZM 100
ages, grain, and oilseed materials) as blind duplicate pairs. A Ultra Centrifugal mill (Retsch, Haan, Germany) with a
blank material (weighing paper) and 3 recovery materials flow-through attachment. The soybeans, dog food, sunflower
were also shipped, for a total of 32 samples. Familiarization seed, and meat and bone meal were ground in a Foss Tecator
samples were sent to each collaborator to be analyzed before Knifetec mill (Foss North America, Eden Prairie, MN). The
the study samples to acquaint collaborators with the methods grass hay, alfalfa hay, and corn silage green chop were ground
and to ensure that each laboratory was capable of handling the to pass through a 1.0 mm screen in a Tecator Cyclotec mill.
study samples. The milk replacer, albumin, and fish meal were split without
THIEX ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 2, 2002 311
Table 2. Categories represented by each collaborative See Tables 2001.11A and B for the results of the
study material and respective RSD values obtained for interlaboratory study, expressed on a protein basis (N × 6.25),
the determination of crude protein in duplicate analyses supporting acceptance of the method.
of 3 random bags of each study material
A. Principle
Category
The material is digested in H2SO4 to convert the protein N
Animal
Material feed Forage Grain Oilseed RSD, % to (NH4)2SO4 at a boiling point elevated by the addition of
K2SO4 with a Cu catalyst to enhance the reaction rate. Ammo-
nia is liberated by alkaline steam distillation and quantified
Meat and bone meal X 0.25
titrimetrically with standardized acid. Aluminum block heat-
Dog food X X 0.24
ers increase the efficiency of the digestion.
Chinchilla food X X X 0.20 The digest must contain residual H2SO4 to retain the NH3.
Birdseed X X 1.11 Water is added manually or automatically to the digest to
Table 2001.11A. Interlaboratory study results for the determination of crude protein by block digestion with a copper
catalyst and distillation into 4% boric acid
ID No. of labsa Mean, % RSDr, % RSDR, % HORRAT
Table 2001.11B. Interlaboratory study results for the recovery of nitrogen from standard compounds by block
digestion with a copper catalyst and distillation into boric acid
Compound No. of labsa Theoretical yield, % N Avg. found, % N Avg. rec., % RSDR, % HORRAT
mixture of corn grain and soybeans (2 + 1). Fineness required Table 2001.11C. Standards
to achieve this precision must be used for all dry mixed feeds Approximate Theoretical
and other nonuniform materials. Mix liquids to uniformity. Standard weight, g yield, % N
Protein Swine Corn Grass Fish Dog Chinchilla Meat and Milk Sunflower Legume
block pellets silage hay meal food food Albumin Birdseed bone meal replacer Soybeans seed hay
Lab A(1) B(13) A(2) B(10) A(3) B(23) A(4) B(24) A(5) B(9) A(6) B(25) A(7) B(26) A(8) B(27) A(11) B(22) A(12) B(17) A(14) B(21) A(15) B(18) A(16) B(19) A(20) B(28)
2 41.25b 39.80b 36.86b 38.72b 7.09 7.22 7.15 6.98 63.79 63.73 24.05 24.58 17.75b 17.60b 79.65 79.58 13.71 13.67 50.01 48.22 20.54 20.93 38.90b 40.26b 18.21 17.08 18.56 18.94
3 40.51 40.70 37.17 36.88 6.93 6.95 6.97 7.09 65.40 65.14 24.36 23.96 17.91 18.07 79.28 79.57 13.43 13.42 50.27 50.06 20.69 21.03 38.87 39.07 17.05 17.03 18.84 18.58
4 39.62 39.96 36.83 36.90 7.33 7.03 7.15 7.07 65.65 64.04 24.48 24.44 18.05 18.15 79.10 79.32 13.42 13.48 49.42 49.86 20.70 20.52 38.87 38.69 17.48 17.51 19.14 18.90
5 40.33 40.22 36.62 36.61 6.88 6.95 7.24 6.88 65.01 65.24 24.77 24.69 18.16 17.95 80.58c 80.84c 13.11d 13.07d 51.17 48.85 21.07 20.17 38.01 39.36 17.58 17.11 18.34 18.80
6 40.12 39.93 36.92 37.38 7.42 7.34 7.36 7.21 63.36 64.02 24.51 24.67 17.71 18.28 79.14 78.31 13.68 13.27 49.47 50.65 20.70 21.19 38.92 39.16 17.49 17.48 18.83 19.12
7 39.65 40.03 37.16 37.11 7.14 6.93 7.01 7.19 65.07 64.90 24.41 24.58 18.06 18.13 78.81 78.99 13.55 13.43 49.26 50.34 21.08 20.91 38.74 38.70 17.91 17.44 18.63 19.00
8 40.22 40.40 37.22 37.04 7.22 6.92 7.22 7.13 64.55 64.92 24.48 24.49 18.16 18.20 78.96 78.90 13.46 13.60 49.03 51.72 20.77 20.80 38.70 38.74 17.19 17.46 18.76 18.94
8 40.24e 40.28e 37.07e 36.92e 7.09e 7.08e 7.04e 7.14e 63.69e 64.09e 24.67e 24.57e 18.11e 18.18e 78.48e 78.89e 13.50e 13.56e 50.33e 51.02e 20.89e 20.92e 38.49e 39.06e 17.33e 17.29e 18.85e 19.11e
9 34.71f 35.62f 32.59f 32.08f 6.89f 6.46f 6.11f 6.46f 56.08f 56.30f 21.44f 22.09f 15.98f 16.33f 68.73f 70.93f 12.28f 12.41f 44.04f 43.34f 18.24f 18.77f 33.22f 34.34f 15.44f 9.21f 16.75f 17.15f
10 40.45 40.14 37.16 37.18 7.06 7.10 7.24 6.83 65.07 65.37 24.44 24.93 17.71c 18.01c 79.34 79.07 13.80 13.53 49.56 49.90 20.56 20.99 38.91 38.22 17.01 18.36 19.44 18.63
11 40.63 40.58 37.00 37.48 7.21 7.17 7.14 7.11 64.53 65.15 24.63 24.51 17.99 18.03 79.37 78.56 13.45 13.44 51.37 50.30 20.96 20.63 38.49 38.55 17.41 17.11 18.61 19.11
12 39.99 40.16 37.06 36.95 7.06 6.99 7.05 7.07 63.64 64.71 24.20 24.63 17.98 18.16 79.30 78.91 13.44 13.41 50.50 50.32 20.94 20.65 38.68 38.83 17.11 17.36 18.87 18.81
a
A and B are blind duplicates; sample numbers are shown in parentheses.
b
Excluded by the Cochran test; P = 2.5% (1-tailed), both when 1% boric acid results are included and excluded.
c
Excluded by the Grubbs test; P = 2.5% (2-tailed); P = 1.25% (1-tailed), both when 1% boric acid results are included and excluded.
d
Excluded by the Grubbs test; P = 2.5% (2-tailed); P = 1.25% (1-tailed), only when 1% boric acid results are included.
e
Value obtained by using 1% boric acid trapping solution in place of 4% boric acid trapping solution.
f
Excluded by the laboratory ranking test described by Youden and Steiner (11).
THIEX ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 85, NO. 2, 2002 315
Table 4. Results of the laboratory ranking tests F factors are 5.70 for wheat, 6.38 for dairy products, and
Laboratory ranking score
6.25 for other feed materials.
Ref.: J. AOAC Int. 85, 311–315(2002)
Excluding data using Including data using
Laboratory 1% boric acid 1% boric acid
Results and Discussion
1 70 76
Study materials were shipped December 11, 2000. Results
2 81 88 were received from 12 laboratories (Table 3) over a period of
3 96 105 3 months, with the last set received on March 13, 2001. The
4 85 92 study materials were sent to a 13th laboratory, but it could not
5 103 112 provide data because of an in-house miscommunication.
Weighing paper, which had been shipped with the collabora-
6 73 80
tive study materials, was reported by some collaborators to
7 80 88
a
NR = not reported.
b
Value obtained by using 1% boric acid trapping solution (all other data were obtained by using 4% boric acid trapping solution).
Table 6. Interlaboratory study results for the determination of crude protein by block digestion with a copper
catalyst and distillation into 4% boric acid and including one set of 1% boric acid results
Product No. of labsa Mean, % RSDr, % RSDR, % HORRAT
Marsha Kapphahn, Department of Animal and Range Sci- (3) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
ences, North Dakota State University, Fargo, ND TERNATIONAL, Gaithersburg, MD, Method 976.06
Lisa Lundberg, Plantedirektoratet, Lyngby, Denmark (4) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
Harold Manson, Olson Biochemistry Laboratories, South TERNATIONAL, Gaithersburg, MD, Method 990.02
Dakota State University, Brookings, SD (5) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
Jurgen Moller, Elisabet Frankenius, and Eva Bogren, Foss TERNATIONAL, Gaithersburg, MD, Method 984.13
Tecator AB, Hoganas, Sweden (6) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
Marcy Russell and Shirley Anderson, Foss North America, TERNATIONAL, Gaithersburg, MD, Method 977.02
Eden Prairie, MN (7) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
Jamie Sefert-Schwind, Adell Corp., Adell, WI TERNATIONAL, Gaithersburg, MD, Method 979.09
Misti Spann, Kansas Department of Agriculture Labora- (8) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
tory, Division of Laboratories, Topeka, KS TERNATIONAL, Gaithersburg, MD, Method 920.87
(9) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
References TERNATIONAL, Gaithersburg, MD, Method 945.39
(10) Official Methods of Analysis (2000) 17th Ed., AOAC IN-
(1) Official Methods of Analysis (2000) 17th Ed., AOAC IN- TERNATIONAL, Gaithersburg, MD, Method 990.03
TERNATIONAL, Gaithersburg, MD, Method 991.20 (11) Youden, W.J., & Steiner, E.H. (1975) Statistical Manual of
(2) Official Methods of Analysis (2000) 17th Ed., AOAC IN- the AOAC, AOAC, Arlington, VA
TERNATIONAL, Gaithersburg, MD, Method 998.05 (12) Kane, P.F. (1987) J. AOAC Int. 70, 907–915