1. food and drink
2. food
3. grains and pasta

DRAFT INTERNATIONAL STANDARD ISO/DIS 6498
ISO/TC 34/SC 10
Secretariat: NEN
Voting begins on:
2009-04-09
Voting terminates on:
2009-09-09
INTERNATIONAL ORGANIZATION FOR STANDARDIZATION • МЕЖДУНАРОДНАЯ ОРГАНИЗАЦИЯ ПО СТАНДАРТИЗАЦИИ • ORGANISATION INTERNATIONALE DE NORMALISATION
Animal feeding stuffs — Guidelines for sample preparation
Aliments des animaux — Lignes directrices pour la préparation d'échantillons
[Revision of second edition (ISO 6498:1998)]
ICS 65.120
ISO/CEN PARALLEL PROCESSING
This draft has been developed within the European Committee for Standardization (CEN), and processed
under the CEN-lead mode of collaboration as defined in the Vienna Agreement.
This draft is hereby submitted to the ISO member bodies and to the CEN member bodies for a parallel
five-month enquiry.
Should this draft be accepted, a final draft, established on the basis of comments received, will be
submitted to a parallel two-month approval vote in ISO and formal vote in CEN.
In accordance with the provisions of Council Resolution 15/1993 this document is circulated in
the English language only.
Conformément aux dispositions de la Résolution du Conseil 15/1993, ce document est distribué
en version anglaise seulement.
To expedite distribution, this document is circulated as received from the committee secretariat.
ISO Central Secretariat work of editing and text composition will be undertaken at publication
stage.
Pour accélérer la distribution, le présent document est distribué tel qu'il est parvenu du
secrétariat du comité. Le travail de rédaction et de composition de texte sera effectué au
Secrétariat central de l'ISO au stade de publication.
THIS DOCUMENT IS A DRAFT CIRCULATED FOR COMMENT AND APPROVAL. IT IS THEREFORE SUBJECT TO CHANGE AND MAY NOT BE
REFERRED TO AS AN INTERNATIONAL STANDARD UNTIL PUBLISHED AS SUCH.
IN ADDITION TO THEIR EVALUATION AS BEING ACCEPTABLE FOR INDUSTRIAL, TECHNOLOGICAL, COMMERCIAL AND USER PURPOSES, DRAFT
INTERNATIONAL STANDARDS MAY ON OCCASION HAVE TO BE CONSIDERED IN THE LIGHT OF THEIR POTENTIAL TO BECOME STANDARDS TO
WHICH REFERENCE MAY BE MADE IN NATIONAL REGULATIONS.
RECIPIENTS OF THIS DRAFT ARE INVITED TO SUBMIT, WITH THEIR COMMENTS, NOTIFICATION OF ANY RELEVANT PATENT RIGHTS OF WHICH
THEY ARE AWARE AND TO PROVIDE SUPPORTING DOCUMENTATION.
© International Organization for Standardization, 2009
ISO/DIS 6498
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but shall
not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In the
unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.
Copyright notice
This ISO document is a Draft International Standard and is copyright-protected by ISO. Except as permitted
under the applicable laws of the user's country, neither this ISO draft nor any extract from it may be
reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, photocopying,
recording or otherwise, without prior written permission being secured.
Requests for permission to reproduce should be addressed to either ISO at the address below or ISO's
member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail [email protected]
Web www.iso.org
Reproduction may be subject to royalty payments or a licensing agreement.
Violators may be prosecuted.
ii
© ISO 2009 – All rights reserved
ISO/DIS 6498
Contents
Page
Foreword ............................................................................................................................................................iv
1
Scope ......................................................................................................................................................1
2
Introduction............................................................................................................................................1
3
Definitions ..............................................................................................................................................1
4
Considerations to sample preparation errors ....................................................................................7
5
Principle................................................................................................................................................14
6
Safety precautions ..............................................................................................................................15
7
Equipment ............................................................................................................................................16
8
Procedure .............................................................................................................................................17
9
Performance tests (quality control)...................................................................................................28
10
Annexes: Categories of feeds – special remarks and flow charts.................................................31
Bibliography......................................................................................................................................................50
© ISO 2009 – All rights reserved
iii
ISO/DIS 6498
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 6498 was prepared by Technical Committee ISO/TC 34, Food products, Subcommittee SC 10, and by
Technical Committee CEN/TC 327, Animal feeding stuffs in collaboration.
This second/third/... edition cancels and replaces the first/second/... edition (), [clause(s) / subclause(s) /
table(s) / figure(s) / annex(es)] of which [has / have] been technically revised.
iv
© ISO 2009 – All rights reserved
DRAFT INTERNATIONAL STANDARD
ISO/DIS 6498
Animal feeding stuffs — Guidelines for sample preparation
1
Scope
This European Standard specifies guidelines for the preparation of test samples from laboratory samples of
animal feeding stuffs including pet foods mostly quoted from AAFCO guidelines [1]. The guidelines are
overruled by special instructions for sample preparation demanded by specific analysis methods for feeding
stuffs (e.g. ISO, CEN, IEC).
2
Introduction
The sample preparation standard describes the procedure for preparing a sample coming to a laboratory (in
general with minimum weight of 0,5 kg) to get a homogeneous test sample (with minimum weight of 100 g)
with the same constitution, with the same composition and without any contamination.
From a test portion (of 0,2 g up to 25 g and more) for weighing to feedstuff analysis representative results
should be achieved of the laboratory sample and finally of the whole lot from which the sample was drawn.
Therefore all the steps for sample preparation should be done rather quickly, under convenient and very clean
conditions so there could be no degradation of sensitive substances, no contaminations and no oxidation
process due to influences of too high temperatures, daylight or air or from residues of apparatus used or from
samples prepared before or simultaneously.
A loss or a change of moisture during sample preparation must be taken into account for reporting results to
origin moisture content for feedstuff control (or to dry mass of 100% or 88%).
3
Definitions
3.1 Definitions concerning 'Sample'
3.1.1
lot
a quantity of material that is assumed to be a single population for sampling purposes
3.1.2
laboratory sample
that portion of material sent to or received by the laboratory
3.1.3
test sample
prepared after subsampling or splitting from the laboratory sample, from which test portions are removed for
testing or for analysis. It may be the laboratory sample if no preparation is required
3.1.4
test portion
the quantity of material, of proper mass and volume for measurement of the analyte or other property of
© ISO 2009 – All rights reserved
1
ISO/DIS 6498
interest, removed from the test sample, taken from the laboratory sample directly if no preparation of the
laboratory sample is required (e.g. with liquids), but usually taken from the prepared test sample
3.1.5
reserve sample
in general left material from the laboratory sample where splitted / subsampled test samples are taken away
from and where no further particle size reduction is done. If mycotoxin- or GMO-analysis are done from the
whole laboratory sample, then the reserve sample is reduced to the corresponding particle sizes too. The
reserve sample should be stored under conditions maintaining integrity
3.2 Definitions concerning 'Substances'
3.2.1
substance
analytes or constituents for which the feeding stuff is to be analysed. Substituents could be classified to/as
nutrients (e.g. crude protein), feed additives (e.g. vitamins), undesirable substances (e.g. heavy metals) and
banned substances (e.g. proteins from animal origin). Substances are analysed by microscopic, (micro-)
biological- or chemical procedures
3.2.1.1
stable substances
analytes or constituents which are not influenced by handling in the relevant sample preparation steps and by
storing at room temperature over a longer time period
3.2.1.2
not-stable substances
analytes or constituents which are (1) volatile, degradable, heat-sensitive or (2) sensitive to light, enzymatic
degradation or chemical oxidation, such that they are largely affected by sample preparation steps (e.g. partial
drying, freeze drying, grinding) or sample storage. The whole sample preparation should be done quickly and
carefully under adequate conditions. Especially the heating of mills during a long grinding procedure should
be avoided. The corresponding (test) samples should be treated and stored under low temperatures
(refrigerator and/or freezer) and protected from intensive air- and daylight-influences e.g. by using brown
glass vessels
Table 1 — Classification of analytes to stable or not stable substances and reasons for degradation
Nutrients:
Feed
additives:
2
Stable substances:
Not-stable substances:
Reason(s) for
degradation / change:
(Crude) protein, fat,
ash, fibre
Moisture
Temperature (volatile)
Starch, sugar,
lactose
Ammonia
Temperature (volatile)
gas production,
enzyme soluble
organic substance
Organic acids (e.g. lactic acid, acetic acid,
butyric acid, citric acid, fumaric acid, formic
acid)
Temperature (volatile)
Minerals (e.g. Ca, P,
Mg, Na, K, Cl)
Fatty acids
Air oxidation (of double
bonds)
Trace elements (e.g.
Cu, Zn, Mn, Fe, Se,
Co)
Vitamins (e.g. vitamin A, D3, E)
Temperature, UV-light
(sensitive)
© ISO 2009 – All rights reserved
ISO/DIS 6498
Amino acids (e.g.
lysine, methionine,
tryptophan)
Enzymes (e.g.
phytases, not starch
polymerase
enyzmes)
Undesirable
substances:
Banned
substances:
NOTE
3.3
1,2-propandiol, glycol
Temperature (volatile)
Probiotics (e.g. Saccharomyces cerevisiae, Temperature (freezing),
Enterococcus faecium)
pressure (sensitive)
Heavy metals (e.g.
As, Pb, Cd, Hg)
Mycotoxins (e.g. aflatoxin B1,
deoxynivalenol, fumonisins, ochratoxin A,
T-2 and HT-2 toxin, zearalenone, ergot
alkaloids)
Mold growth and
change of mycotoxins
possible at room
temperature; UV-light
(sensitive –aflatoxin B1)
Dioxins and PCBlike Dioxins
Pesticides (e.g. PCBs, OCDs, other
pesticides)
Temperature (sensitive)
Hydrocyanic acid
Temperature (volatile)
Antibiotics
Temperature (sensitive)
Proteins of animal
origin
Too many microorganisms present in feeds can break down the organic compounds.
Definitions concerning `Animal feeding stuffs´
For identification and grouping a laboratory sample to the terms and annexes used within these guidelines
some specific definitions are given in this document.
NOTE
Definitions of animal feeding stuffs are given by legislation worldwide. As an example definitions of European
directives and for straight feeds in an alphabetical list from a German committee are mentioned within the bibliography [8],
[9], [10], [11], [12], [13].
© ISO 2009 – All rights reserved
3
ISO/DIS 6498
3.3.1
birdseed
grains and oilseeds that are fed to birds
3.3.2
whole cottonseed
the entire unprocessed cottonseed product, including the hulls, lint, and meat of the cottonseed
3.3.3
mineral mix
consist mainly of mineral ingredient in either granular, bead, or small pelleted form. Vitamins, in encapsulated
or beadlet form, may be incorporated into the mix. The entire mix is free flowing
3.3.3.1
mineral pellets
agglomerated feed formed by compacting and forcing through die openings by a mechanical process
3.3.4
dry feeds
a feed ingredient or a complete animal feed which typically contain not more than 15% moisture, 15% fat, or
15% sugar
3.3.4.1
pellets
agglomerated feed formed by compacting and forcing through die openings by a mechanical process
3.3.5
forages inclusively silage, hay, haylage, total mixed ration and by-products
3.3.5.1
forage
edible parts of plants, other than separated grain, that can provide feed for grazing animals or that can be
harvested for feeding, including browse, herbage, and mast. Usage: Generally, the term refers to more
digestible material (i.e. what is called pasturage, hay, silage, dehydrated and green chop) in contrast to lessdigestible plant material, known as roughage
3.3.5.2
silage
forage preserved in a succulent condition by organic acids produced by partial anaerobic fermentation of
sugars in the forage
3.3.5.3
roughage
fibrous, coarsely textured parts of plants, such as stovers, straws, hulls, cobs, and stalks
3.3.5.4
hay
the aerial portion of grass or herbage especially cut and cured for animal feeding
3.3.5.5
haylage
product resulting from ensiling forage with about 45% moisture in the absence of oxygen
4
© ISO 2009 – All rights reserved
ISO/DIS 6498
3.3.5.6
total mixed ration (TMR)
a single mixture of all feed ingredients (forages, grains, and supplements) that is supplied to an animal for a
24-hour period. In practice, the 24-hour allotment of the mixture may be offered in one or more feedings.
3.3.5.7
by-products
products which remaining during process-procedures (e.g. from fermentation like dried destillers grains with
solubles = DDGS) for the production of ingredients from plant material
3.3.6
oilseed
any seed from which oil is expressed (i.e. sunflower seeds)
3.3.7
large block feed and molasses block feeds
agglomerated feed compressed into a solid mass cohesive enough to hold its form and weighing over one kg,
generally weighing about 20 kg. It may be marketed as a mineral block or a `caramelized´ molasses drum,
containing various minerals and nutrients. Samples may be received in the lab as large chunks, cores or
`sticky clumps´
3.3.8
liquid feed
a feed product that contains sufficient moisture to flow readily. A liquid feed is generally, if not always, a
molasses based product
3.3.9
canned pet foods
a pet food which has been processed, packaged, sealed, and sterilized for preservation in cans or similar
containers
3.3.10
semi-moist pet food
a meat based pet food that has been partially dried to prevent microbial decomposition. The moisture content
may range from 15% – 40%. The product generally is in the form of strips or cubes and is designed to be
stored at room temperature
3.3.10.1
dog chews
also known as `rawhide bones´. Meat strips that have been completely dried to a leather-like consistency
3.3.11
premixtures
a uniform mixture of one or more micro-ingredients with diluent and/or carrier. Premixes are used to facilitate
uniform dispersion of the micro-ingredients (i.e. drugs, antibiotics, and/or vitamins) in a large mix
3.3.12
range cube and alfalfa hay cubes
an agglomerated feed formed by compacting and forcing the mix through die openings by a mechanical
process. This results in a pellet that is about 2 cm diameter and 5 cm long. They may contain molasses to
help hold them together. Usually they are fed on the ground. This procedure also applies alfalfa cubes that
3
consists of chopped alfalfa hay oppressed into cubes usually larger than 16 cm
3.3.13
texturized and sticky feed
a mix of assorted grains and commercial feed (generally pelleted) all of which has been treated with a coating
© ISO 2009 – All rights reserved
5
ISO/DIS 6498
of molasses. Some of the grains may have been steam heated and / or rolled prior to incorporating into the
texturized feed
3.3.14
aquatic feed
aquatic feed which has been mechanically processed in forms of pellets, flakes, crumble, encapsulated and
as powder, packaged sealed and that are fed to aquatic animals
3.4
Definitions concerning 'Sample preparation procedure'
3.4.1
homogeneity
the degree to which a property or a constituent is uniformly distributed throughout a quantity of material.
Homogeneity may be considered to have been achieved in a practical sense when the sampling error of the
processed portion is negligible compared to the total error of the measurement system. Since homogeneity
depends on the size of the units under consideration, a mixture of two materials may be inhomogeneous at
the molecular or atomic level, but homogenous at the particulate level. However, uniform visual appearance
does not ensure compositional homogeneity
3.4.2
partial drying ('pre-drying')
part of the sample preparation procedure for feedstuff samples with a high moisture content (dry mass < 85%,
e.g. silages) in which the sample is carefully dried to allow subsequent sample preparation procedures to be
applied e.g. particle size reduction by grinding with a mill. The partial drying procedure depends on the
feeding stuff (e.g. at temperatures below 55°C to 60°C for silages), and on the heat stability of the substance
(e.g. 70°C ± 10°C for trace elements and heavy metals)
3.4.3
freeze drying ('pre-drying')
careful drying process in the vacuum under heat supply, in order to sublimate the moisture
3.4.4
coarser grinding ('pre-grinding')
in cases when the laboratory sample contains large lumps or its particle size is over 6 mm a firstly coarse
grinding step of the whole sample is done before mass reduction. Finally it is a kind of particle size reduction
to ensure homogeneity of the laboratory sample
3.4.5
mass reduction
part of the sample preparation procedure to reduce the mass of a laboratory sample by splitting or
subsampling it by the use of (stationary or rotary) dividers or by fractional (alternate) shovelling
3.4.6
particle size reduction
part of the sample preparation procedure done by chopping, crushing, cutting, blending (homogenizing),
macerating, milling (grinding), pressing, pulverizing to obtain a homogenous test sample for further analysis.
In general the particle size reduction follows the mass reduction step of the sample preparation procedure
with different sieve-size-options to ensure integrity of the test sample(s)
6
© ISO 2009 – All rights reserved
ISO/DIS 6498
Figure 1 —Illustration of definitions concerning 'Sample', 'Substances', and 'Sample preparation
procedure'
"SAMPLING"
Choice from Lot
"SAMPLE PREPARATION"
Laboratory sample(s) (500 g) with:
Dry matter < 85%
Dry matter ≥ 85%
Lumps or particles > 6 mm
Partial or freeze drying ( `pre-drying´ )
Coarser grinding ( `pre-grinding´ )
Mass reduction (subsampling / splitting)
Reserve sample
Test sample(s) (100 g) for:
Particle size reduction
- Microscopy
with different sieve-size options
- Analysis of stable substances
(1,0 mm, 0,5 mm, < 0,5 mm, no grinding)
- Analysis of not-stable substances
"ANALYSIS"
4
Test portion(s) (0,2 to 25 g)
Considerations to sample preparation errors
Sample preparation steps have been shown to be some of the largest error of laboratory error. This error,
which is generally overlooked, may be much larger than the error in subsequent analytical procedures.
4.1 Subsampling and other errors
Sample heterogeneity may add to the total subsampling error (TSE) on two levels [4]:
4.1.1
Constitution heterogeneity
On a first level this is a result when not all of the particles of the laboratory sample have the same composition
(shape, size, density, etc). If a great overall composition-wise difference between the individual fragments
exists, the constitution heterogeneity is large, but if the fragments are more homogeneous constitution
heterogeneity is lower. The total contribution to heterogeneity is never zero, however, as that would be the
© ISO 2009 – All rights reserved
7
ISO/DIS 6498
case of all fragments being strictly identical. Mixing and blending does not change constitution heterogeneity.
The only way to alter the constitution heterogeneity of any given material would be by comminution (crushing /
cutting) or by other methods changing the physical properties of a sample. The reduction of the average grainsize is the dominating factor in reducing constitution heterogeneity by such means.
Therefore a firstly coarse grinding (`pre-grinding´) of the whole laboratory sample before subsampling /
splitting reduces constitution heterogeneity.
This fundamental subsampling error (FSE) can be controlled by the appropriate choice of the test sample
mass (see 4.2). Therefore collect enough mass to ensure that all of the particles of different composition are
contained in the subsample / split. The larger the particle size of a material, the larger the mass of the
subsample must be to minimize error.
4.1.2 Distributional heterogeneity
On a second level this is the non-random distribution of particles in the sample, results mainly from the forces
of gravity to particles of different densities, sizes and shapes which leads to a grouping and segregation of all
particles. Particles with large differences in size and / or density tend to segregate or stratify heavily, with the
smallest and / or densest particles at the bottom of the sample. For the sake of illustration, imagine a
laboratory sample consisting of black and white spheres and with significantly different grain-size distributions:
If all the black spheres are to be found at the bottom of the sample and the white spheres are more to the top,
the system displays a very high distributional heterogeneity. If on the other hand the spheres were to be well
mixed (homogenized), the system distributional heterogeneity would be significantly reduced.
To reduce this grouping and segregation error (GSE) mix / blend the sample before subsampling and collect
many increments at random from the laboratory sample (see 4.3).
Mixing is not adequate for many materials: for some materials and circumstances mixing may actually
increase segregation instead of reducing the grouping and segregation error. As long as gravity exists there
will be segregation. Many materials will always display an innate propensity for segregation, even immediately
after mixing, e.g. highly density-fractionated materials, suspensions. Such systems require constant
monitoring and treatment, but once this feature has been duly recognized it can always be dealt with
satisfactory.
Incrementing (= collection of many random increments from the laboratory sample to make up the subsample
or split) always works in reducing error from distributional heterogeneity and takes less time and equipment to
implement. 30 increments is generally adequate. More increments must be used for very heterogeneous
materials and, if it is known that little segregation exists, fewer increments can be used, but in any case no
fewer than 10 is recommended.
4.1.3 Other errors
Other errors that arise from sample preparation include losses and gains in analyte concentration from such
mechanisms as grinding, excessive heat, loss of fines, contamination, electrostatic separation, etc. These
errors can be large and are usually a result of carelessness or lack of knowledge.
4.2 Minimum mass
To properly represent a laboratory sample, the subsample or split must contain adequate mass ('minimum
mass').
The amount of mass required depends on the acceptable error in the subsample or split, on the density,
heterogeneity and concentration of the analyte particles and on the largest particle size (see examples below).
The fundamental subsampling error (FSE) is the error that remains when the subsampling procedure is rid of
incorrect errors and faults. This means that FSE is the minimum subsampling error that can be obtained in
practice and it is inherent only to the material heterogeneity. For this very reason it is, of cause, methodindependent. FSE can be calculated from a series of measurements as the difference between the estimate of
8
© ISO 2009 – All rights reserved
ISO/DIS 6498
the analyte concentration from measurements (aS) and the actual analyte particle concentration (aL) in the
laboratory sample from which the subsample is taken, i.e. FSE = (as – aL) / aL
The variance of the FSE can estimated to an order of magnitude from Gy’s formula [4], [7]:
σ 2 (FSE ) = cfgβ d 3 (1 / M s − 1 / M L )
(1)
where
c
is the constitutional parameter expressed in g/cm3 that accounts for the densities as well as the
proportions of the constituents,
f
is a ‘‘particle shape factor’’ (dimensionless) describing the deviation from the ideal shape of a cube.
A square will have f = 1, a sphere f = 0,52, spherical particles f = 0,5 and an almost flat disc f = 0,1,
g
is a ‘‘size distribution factor’’ (dimensionless) describing the span of particle sizes in the lot, where g
= 0,25 for a wide size distribution and g = 1 are for uniform particles,
β
is a ‘‘liberation factor’’ (dimensionless) describing the degree of liberation of the critical component
from the matrix. β = 1 for totally liberated particles and β = 0,03 for very small analyte particles
incorporated in large particles of the matrix. β can be equated by β = √(L/d), where L is the particle
size of the analyte particles “trapped” in the matrix particles with particle size d. For β = 0,03 with a
matrix particle size d = 0,01 cm this corresponds to an analyte particle size of ~10 microns
(molecular level),
d
is the ’’top particle size’’, defined as the square-mesh screen that retains 5% of the material
(dimension of length expressed in cm),
MS
is the mass of the subsample,
ML
is the mass of the laboratory sample from which the subsample is taken.
A more detailed description of the constitutional parameter, c is:
c=
(1 − a L ) 2
ρ c + (1 − a L ) ρ m , which can be reduced to c = ρc /aL for aL < 0,1
aL
(2)
where
aL
is the concentration of the analyte particles present in the laboratory sample,
ρc
is the density of the analyte particles,
ρm
is the density of the matrix.
Gy’s formula was derived from mineralogical samples, and works for binary mixtures of particulate material,
where the sought of particles (i.e. the analyte particles) are present as separate fragments and is therefore
very approximate for feeding stuffs but gives a high end estimate on the fundamental sampling error and also
an indication on the dependence on particle size and sample weight.
Three examples of the minimum sample weight needed for some set FSE (% RSD) and particle size values
are given below. It should be noted that using correct mass reducing devices such as riffle splitters, vario
dividers etc., will generally lead to lower values of the FSE than can be calculated from Gy’s formula, whereas
grab sampling can give rise to even higher values of FSE [4], [7]:
© ISO 2009 – All rights reserved
9
ISO/DIS 6498
Example 1: 0,4% methionine added to a mineral mixture (ρm = 1,5 g/cm3) as DL-Methionine sulfoxide
(ρc ~ 2 g/cm3). β = 1 since the analyte particles are completely liberated. g = 0,25 which corresponds to a wide
range of particle sizes. The sample mass, ML is 100 g and the minimum mass of the sub-sample, MS
depending on the tolerable FSE can be seen from table 2. C = cfgβ in the text below.
Example 2: 10 ppm Se added to a mineral premixture (ρm = 2 g/cm3) as Na2SeO3 (ρc = 3,1 g/cm3). β = 1 since
the analyte particles are completely liberated. g = 0,25 which corresponds to a wide range of particle sizes.
The sample mass, ML is 100 g and the minimum mass of the sub-sample, MS depending on the tolerable FSE
can be seen from table 3.
3
Example 3: 10 ppm Cu as a natural part of an organic feeding stuff (ρm = ρc = 0,8 g/cm ). β = 0,03 since the
analyte particles are completely built in the matrix particles (best case scenario). g = 0,25 which corresponds
to a wide range of particle sizes. The sample mass, ML is 100 g and the minimum mass of the sub-sample, MS
depending on the tolerable FSE can be seen from table 4.
Example 1
Example 2
Example 3
aL (%)
0,4
0,001
0,001
ρc (g/cm3)
2
3,1
0,8
ρm (g/cm )
1,5
1,5
0,8
Ms (g)
see table 2
see table 3
see table 4
ML (g)
100
100
100
C
498
309995
79999
f
0,5
0,5
0,5
g
0,25
0,25
0,25
β
1
1
0,03
C
62
38749
300
3
10
© ISO 2009 – All rights reserved
ISO/DIS 6498
Table 2 — Minimum mass of a subsample where methionine was added to a mineral mixture
Minimum sample (g)
d (mm) / FSE
(expected RSD)
20%
15%
10%
5%
2%
0,1
0,002
0,003
0,01
0,02
0,2
0,2
0,01
0,02
0,05
0,2
1
0,5
0,2
0,3
0,8
3
16
1
2
3
6
20
61
Table 3 — Minimum mass of a subsample where selenium was added to a mineral premixture
Minimum sample (g)
d (mm) /FSE
(expected RSD)
20%
15%
10%
5%
2%
0,1
1
2
4
13
49
0,2
7
12
24
55
89
0,5
55
68
83
95
99
1
91
95
97
99
100
Table 4 — Minimum mass of a subsample where copper is a natural part of an organic feeding stuff
Minimum sample (g)
d (mm) / FSE
(expected RSD)
20%
15%
10%
5%
2%
0,1
0,01
0,01
0,03
0,1
1
0,2
0,1
0,1
0,2
1
6
0,5
1
2
4
13
48
1
7
12
23
55
88
The equation below is only an approximation; therefore the variables do not have to be precisely measured. In
most cases estimation of variables is approbate.
Μ S = 10 ⋅ λ ⋅ d 3 / FSE 2
© ISO 2009 – All rights reserved
(3)
11
ISO/DIS 6498
where
MS
is the minimum mass of sample collected [g]
FSE
is the tolerable error (fundamental subsampling error) – expected RSD
d
is the size of the largest particle [cm]
λ
is the density of material [g/cm3]
3
Example: If the maximum particle size is 4 mm and the density is 0,8 g/cm and the tolerable error is 15%
RSD, the minimum mass to represent all the particle sizes for this error is 23 g.
The error of a particular mass can also be calculated:
FSE 2 = 10 ⋅ λ ⋅ d 3 / Μ S
(4)
The adequate largest particle size can also be calculated:
[
]
d = FSE 2 ⋅ Μ S / (10 ⋅ λ )
1/ 3
(5)
Example: A lab sample has a largest particle size of approximately 2 mm, a density of approximately 0,7
3
g/cm and weighs 1 kg. Without grinding and with a tolerable error of 5% at least a minimum mass MS of 22 g
would have to be analyzed.
(
)
Μ S = 10 ⋅ 0,7 ⋅ 0,2 3 / 0,05 2 = 22 [g ]
(6)
If the analytical procedure requires only 1 g, the error FSE is 24% without grinding.
FE 2 = 10 ⋅ 0,7 ⋅ 0,2 3 / 1 = 0,056, with FSE = 0,24
(7)
Therefore if the desired laboratory subsampling error is not to be more than 5% RSD, the sample would have
to be ground to a particle size of 0,07 cm = 0,7 mm (or less).
[
]
d = 0,05 2 ⋅ 1 / (10 ⋅ 0,7 )
1/ 3
= 0,07 [cm ] = 0,7 [mm ]
(8)
Table 5 — Minimum mass: expected relative standard deviation (RSD) from laboratory subsampling;
assumed density = 1 g/cm³.
FSE (expected RSD)
15%
10%
5%
2%
1%
Maximum particle size (d)
12
0,5 mm
0,06 g
0,13 g
0,5 g
3g
12,5 g
0,75 mm
0,2 g
0,4 g
2g
10,5 g
42 g
1 mm
0,4 g
1g
4g
25 g
100 g
2 mm
4g
8g
32 g
200 g
400 g
5 mm
56 g
125 g
500 g
3 130 g
12 500 g
© ISO 2009 – All rights reserved
ISO/DIS 6498
3
NOTE
For materials with densities other than 1 g/cm , this table can be modified by multiplying the entries by the
density of the material of interest. For example the subsampling of a material with largest particle size of 2 mm, a tolerable
3
subsampling RSD of 5% and a density of 0,5 g/cm would require 16 g.
4.3
Errors associated with splitting techniques
The data in table 6 below demonstrate the error associated with various splitting techniques for a model
mixture of sand particles. Figure 2 demonstrates the representativity (i.e sum of the sampling error related to
precision and accuracy) of 17 different mass reduction devices for a model mixture of 89,9% wheat, 10,0%
rape seed and 0,10% glass [4], [6]. The primary difference in the mass reduction methods is the number of
increments selected. For this to be true structurally correct use of the mass reduction devices is required (e.g.
equal probability for the selection of all particles, no loss of particles, centre of gravity rule obeyed, parallel
cuts) which is difficult or impossible to obtain with shovelling and grab sampling methods. Therefore mass
reduction methods based on grab sampling or shovelling methods can have substantial problems with
precision and accuracy on trace components present as separate particles, which may be due to selective
loss or poor sampling of smaller particles [4], [6]. From table 6 and figure 2 it can be concluded, that more
increments lead to improvements for the mass reduction in the laboratory by reducing the sampling error. In
general a rotational divider reaches about several hundred of increments, a stationary riffle splitter about 10 34 increments and cone & quartering only 2 increments. Therefore coning and quartering is not recommended
in the critical mass reduction step in the laboratory i.e. the mass reduction step with the largest contribution to
the total error. The preparation of the final test potion where the ratio between the mass of the laboratory
sample to the mass of the final test potion is 100-10 000 can usually be considered the critical step of the
mass reduction of the laboratory sample. Grab sampling is to be totally avoided for the critical mass reduction
step unless it is established that the sampling error is insignificant compared to the total analytical error.
Table 6 — Results from split of sand mixture. Test of a mix of 60% coarse sand with 40% fine sand
(P = 0,6). From: T.Allen and A.A.Kahn, Critical Evaluation of Powder Sampling Procedures,
The Chemical Engeneer, May, 1970.
Method A)
Number of
increments
Standard
deviation (sr) of
samples (%)
Variance (sr)
2
(%)
Coning & quartering
2
6,81
46,4
22,7
Stationary riffling
10 - 12
1,01
1,02
3,4
Rotary riffling
> 100
0,125
0,016
0,42
0,076
0,0058
0,25
Random variation
A
2
Estimated maximum
sample error (%)
Stationary rifflers with a higher number of increments and less subsampling error are available [6].
© ISO 2009 – All rights reserved
13
ISO/DIS 6498
Figure 2 — Pooled representativity, r2 (r2 = bias2 + precision2). Representativity should be as low as
possible. Higher sums thus mean lower reliability. RK # = Riffle splitter with # chutes. [6]
5
Principle
All the sample preparation steps depend on the different properties of the feedstuffs and on the substances to
be analysed. In every case the special instructions of the analysis methods concerning sample preparation
have to be considered.
In general the whole laboratory sample is reduced in mass and in particle size to obtain one or more test
samples for the analysis of stable and not-stable substances, for microscopy analysis and for reserve. If the
analysis protocol and the intended proceeding of the reserve sample permit it, the laboratory sample will
preferably “pre-grinded” completely at first to an adequate coarse particle size and then mass reduced to
ensure homogeneity of the test samples.
For feedstuffs with a higher moisture content (dry matter < 85%) partial drying or freeze drying before mass
reduction could be necessary.
For feedstuffs with lumps or particle sizes > 6 mm coarser grinding of the whole laboratory sample to a
particle size of < 6 mm before mass reduction / subsampling is absolutely necessary.
The samples have to be stored at every stage of the sample preparation under adequate conditions (e.g. at
room temperature, refrigerated, frozen, air-tight, light-protected / dark) to maintain their integrity.
14
© ISO 2009 – All rights reserved
ISO/DIS 6498
6
Safety precautions
The mills for crushing, cutting and grinding have sharp moving blades. Never put hands or fingers past the
edges of the introduction chamber. Never open the mills until they have completely stopped. Check to see that
satisfy interlocks on all equipment are operating properly.
Wear appropriate personnel protective equipment as required in the laboratory. Safety is of most importance
during the sample preparation phase of the analysis.
Operate the dust ventilation system during dust generation procedures. To minimize dust, use a vacuum
cleaner to clean the hood area, mills, and work area.
Check that all electrical equipment is properly grounded and maintained. Do not place metal item or
aluminium foil in the microwave oven when using it for drying samples.
© ISO 2009 – All rights reserved
15
ISO/DIS 6498
7
Equipment
7.1
Equipment for sample preparation in general
7.1.1
Brushes for cleaning grinders, etc.
7.1.2
Compressed air blower for cleaning.
7.1.3
Vacuum cleaner
7.2
Equipment for partial and freeze drying ('pre-drying')
7.2.1 Lyophilsation system, forced-air drying oven set at 55°C ± 5°C and / or microwave oven,
household type and / or vacuum oven.
7.2.2
Moisture dish (pan), made of plastic, aluminium, glass, e.g. with ≥ 50 mm diameter, ≤ 40 mm deep.
7.2.3
Analytical electronic balance, accurate to 0,1 mg.
7.3
Equipment for mass and particle size reduction of 'wet' feeds (e.g. forages)
7.3.1 Garden pruning clippers for cutting forages or a paper cutter for small sample volumes or a
laboratory forage chopper for large volumes.
7.3.2
Cutting mill, with 6 mm and 1 mm screens (e.g. Fritsch P15, Retsch SM100).
7.3.3
Shearing type mill with forage head and 1 mm screen (e.g. Udy or Tecator Cyclotec).
7.3.4 Riffle sample splitter; the minimum chute width is to be at least two times the diameter of the largest
particle + 5 mm.
7.4 Equipment for mass and particle size reduction of 'dry' feeds (e.g. cereals, mineral
mixtures, pelleted feeding stuffs)
7.4.1
Riffle splitter
7.4.2
Rotary splitter with vibratory feeder.
7.4.3 Shearing grinding mill (e.g. Fritsch P 14, Retsch ZM 100), equipped with 1,0 mm, 0,5 mm and
< 0,5 mm – sieves.
7.4.4
Cutting mill with 4-6 mm screen (e.g. Fritsch P 15, Retsch SM 100).
7.4.5
Shearing blending mill (e.g. household coffee mill or Tecator Knifetec mill).
7.5
Equipment for the storage of samples
7.5.1
Sample bottles with air-tight lids.
7.5.2
Wide mouth bottles (for mycotoxins), with screw cap, plastic.
7.5.3
Refrigerator
16
© ISO 2009 – All rights reserved
ISO/DIS 6498
7.5.4 Freezer
8
Procedure
After registration and a check of a laboratory sample (see 8.1) the homogenisation procedure consists in a
mass reduction step at which the laboratory sample of more than 0,5 kg is reduced to test samples of about
100 g (see 8.2).
In the second step the particles in the test samples are reduced to adequate sizes to minimise the
subsampling error that arises when the test portion is taken from the test sample. Particle size reduction
should be performed without deteriorating the integrity of the substance to be analysed (see 8.3).
For feedstuffs with higher moisture content (dry mass < 85%) firstly a partial drying below 55 – 60°C could be
necessary to grind a subsample by a mill to particle sizes of 1,0 mm for analysis of stable analytes
(see 8.4).
For feedstuffs containing lumps or consisting of particle sizes > 6 mm, firstly a grinding by a jaw crusher or a
chopping to particle sizes below 4 – 6 mm could be necessary before subsampling is possible (see 8.5).
For some fatty and/or sticky feedstuffs (e.g. oilseeds, pet foods, molasses block feed) special sample
preparation procedures are useful / necessary (see 8.6).
Finally the samples are stored (see 8.7).
NOTE
Samples taken for routine analysis by NIR should reflect the sample preparation carried out to derive the
calibration. By it’s very nature, NIR requires minimal or no sample preparation and is often used to analyse samples that
are either fresh or have been dried and only coarsely chopped. However, when building a calibration it should be
recognised that, because spectra may be collected and averaged over large samples it may be necessary to dry, finely
grind, and then reduce in mass using a splitter to obtain a subsample suitable for reference analysis. Although the spectra
represent an average of a larger sample than used to obtain the reference value this is acceptable practice.
8.1
Sample check
Firstly the laboratory sample will be registered and uniquely identified (e.g. with a unique code number).
Before starting the proper sample preparation procedure some laboratory sample checks have to be done:
8.1.1 Check of sample constitution
When arriving to the laboratory the sample should have no damages and still should be cooled or frozen if
necessary. Furthermore the sample protocol should be in accordance to the received sample and all
information concerning the sample should be available and complete.
Deficiencies (e.g. no information about the type of feeding stuff, container of the laboratory sample is opened,
the sample protocol does not fit to the sample container) should be documented and subsequently reported to
the principal. If possible, the deficiency shall be corrected. When this is not possible, and the observed
deficiency might affect the analytical result (e.g. when there is not enough sample mass or when there is
already mold in the laboratory sample because of too high moisture content or because the sample was not
sufficiently cooled during the transport to the laboratory), another sample from same lot is necessary.
8.1.2 Check of feeding stuff properties
The laboratory sample should be identified for grouping to the feeding stuffs cited within the definitions and
categories (see 3.3).
The moisture content of a laboratory sample determines the initial steps of sample preparation: `wet samples´
with higher moisture content (dry mass < 85%) should be prepared as soon as possible or stored at low
temperatures, otherwise deterioration starts.
For forages with moisture contents too high for direct grinding (dry mass < 85%) the whole laboratory sample
should be chopped to pieces of about 1 cm. If necessary, the laboratory sample is subsampled by alternate
© ISO 2009 – All rights reserved
17
ISO/DIS 6498
shovelling and subsequently partially dried. The above is recommended for stable analytes and for the whole
or at least remaining sample for mycotoxin analysis. For not-stable (volatile) analytes (e.g. organic acids,
ammonia, hydrocyanic acid, and for GMOs, organic residues) it is recommended to analyse a test sample as
such, without previous drying of the sample. Alternatively, vacuum drying at low temperatures or freeze-drying
could be performed when samples are to be analysed for non volatile components.
For 'dry feeding stuffs' composed of lumps or with particles larger than 6 mm firstly a coarser grinding of the
whole sample with a jaw crusher to particles sizes of 4 - 6 mm is recommended before mass reduction /
subsampling is initiated.
8.1.3 Check of substances to be analysed
The number of test samples depends on the analytes to be done.
For the analysis of stable substances, not-stable analytes and for microscopy analysis separate test samples
should be prepared. The remaining sample is used as reserve (backup).
This enables that for stable analytes the test sample could be reduced to adequate particle sizes at once and
subsequently stored at room temperature until further analysis.
Test sample for not-stable analytes should be stored at low temperatures and, in order to prevent degradation,
they should only be reduced to adequate particle size at the day of analysis (and not long in advance).
For testing the composition of feeding stuffs by microscopy and for analysis of probiotics it is important that no
particle size reduction by grinding (milling) is done. Test samples for analysis of probiotics should not be
frozen, only refrigerated.
For mycotoxins and analysis of GMOs by PCR if possible the whole laboratory sample or at least the greater
part of the remaining laboratory sample should be used for particle size reduction and then subsampled if
necessary.
After mass reduction, in general test samples for stable and not-stable analytes of the following particle size
classes should be prepared under adequate (temperature) conditions: see table 7 and table 8.
Table 7 — Recommended particle sizes of test sample(s) for stable analytes and for microscopy
1,0 mm
Nutrients (e.g. crude protein, crude fat, crude ash, crude fibre) if not grinded with a sieve
size of 0,5 mm and minerals, trace elements, heavy metals if not grinded with a sieve sieze
of 0,5 mm or 0,1 mm
0,5 mm
Starch, sugar, lactose, amino acids, methionine hydroxy analogue (MHA), crude protein
(Dumas)
0,1 mm
In mineral mixes for minerals, trace elements and heavy metals (see EN 15510 : 2007)
No grinding
Microscopy analysis (e.g. composition) or NIR/NIT analysis or NMR oil analysis
or phytase activity (if not grinded with a sieve size of 1 mm)
NOTE – Treatment and storage of samples at room temperatures is possible.
18
© ISO 2009 – All rights reserved
ISO/DIS 6498
Table 8 — Recommended particle sizes of test sample(s) for not-stable (degradable, volatile, heatsensitive) analytes
1,0 mm
In dry feed for moisture, vitamins, organic acids, propandiol, organic residues like PCBs,
OCDs, other pesticides, banned antibiotics, veterinary drugs and mycotoxins
0,5 mm
In dry feed for mycotoxins because of non-uniform distribution within a (laboratory / test)
sample if not grinded with a sieve size of 1,0 mm
No grinding but In dry feed (pellets) for probiotics
soft treating by
solving under
light pressure
No grinding
In mineral mixes and premixtures for vitamins, antibiotics, drugs when the particle size
distribution is sufficient, otherwise grind briefly to 1,0 mm to avoid heat generation
No grinding but In forages for analysis of the corresponding test sample with origin moisture content of
cutting to
organic acids, ammonia, hydrocyanic acid, carotene
pieces of 1 cm
No grinding but In forages for organic residues like PCBs, OCDs, other pesticides, banned antibiotics,
macerating with veterinary drugs
a thermo mixer
No grinding but Of the whole laboratory sample or at least the remainig rest of the laboratory sample for
macerating with GMOs by PCR
a thermo mixer
NOTE
Particle size reduction of the test samples should be done rather quickly and at the day of analysis if possible.
Heat generation during grinding procedure should be avoided. When the analysis does not start immediately after the
preparation of the test sample / test portion, the latter should be stored at low temperatures in a refrigerator. With the
exception of test samples to be analysed for probiotics, it is recommended to store test samples / test portions in a freezer,
when analysis is not commenced within 48 hours after particle size reduction.
Reserve samples are stored without particle size reduction, but if mycotoxin or GMO-analysis by PCR is
necessary the whole laboratory sample should be prepared to the corresponding particle sizes.
NOTE
Treatment and storage of the reserve sample should be done under conditions which maintain its integrity
over an adequate time period (e.g. until the guaranteed minimum durability of a sample is exceeded).
8.2 Mass reduction
Laboratory samples can be mass reduced by splitting devices or subsampling.
Mass reduction with the use of rotational dividers or riffle splitters is recommended and the techniques can be
used for the reduction of the 100 g test sample to < 1,0 g test portion without serious problems.
If it has been established that the mass reduction error is insignificant or if it is not possible to mass reduce
with correct mass reduction devices (i.e. rotary or riffle splitters), the mass reduction can be accomplished with
subsampling. With subsampling, anywhere from a single increment to as many as several hundred increments
are selected at random from the primary sample to form the subsample.
Unfortunately, it is a common practice to take only a few increments. If only a small number of increments are
selected, there can be very large subsampling errors due to sample heterogeneity. The number of increments
should not be determined by what is easy, but rather what is acceptable from an error point of view. Rotary
riffle splitting is the most accurate splitting method, coning & quartering is a very poor method and should
never be used.
If it is known that the material is not segregated, then fewer than 10 increments can be selected. If the
material is known or suspected to be heavily segregated, then more than 10 increments should be selected.
© ISO 2009 – All rights reserved
19
ISO/DIS 6498
Many materials have a wide range of particle sizes when they arrive to the lab and need more increments to
properly represent them (consider more than 10 increments).
During the sample preparation of grinding and sieving, the range of particle sizes is reduced and fewer
increments can be taken.
8.2.1
8.2.1.1
Mass reduction devices
Riffle splitters
Criteria for the design:
 Even number of chutes.
 Greater number of chutes is desirable.
 For non-gated rifflers the feeding scoop should be exactly the same width as all the riffles.
 Minimum chute width should be at least two times the diameter of the largest particle plus five
millimetres; be sure that chutes do not clog with particles. Pellets can clog the chutes if the chute with is
not of adequate size.
 Easy to clean.
 Rifflers must be from durable, inert materials (e.g. stainless steel).
 Rifflers with bent chutes or any defects must never be used.
Criteria for the proper use:
 The riffler should be on a firm, level plane.
 Do not feed the sample too fast (the chutes can fill up and overflow).
 For non-gated rifflers, do not feed the sample into the riffle splitter with the receiving pan (it is not the
correct width). Material in the feeding scoop should be spread out evenly prior to pouring into the splitter.
Material must be fed slowly in the centre of the chutes (to prevent overflow of material from the shallow
chutes into the deeper ones).
 For gated rifflers, slowly fed the material into the hopper in a back and forth motion. The material should
be evenly distributed in the hopper after feeding.
 Fine powders should be fed with care, as they can clog the chutes.
 Fines may stick to the splitter due to static electricity. If this occurs and measurement of the fines is
important, you may try grounding the riffler or using an anti-static mat.
Riffle splitters are very operator dependent. Carry out tests with material similar to the material to be split to
demonstrate their performance.
8.2.1.2
Rotational dividers
Criteria for the design:
 Should be made of inert material.
 Cutting edge should be radial from centre (pie shaped).
20
© ISO 2009 – All rights reserved
ISO/DIS 6498
 Maintain constant speed.
 Minimize drop from feeding chute to cutting edge to prevent dust formation.
 Fine powders should be fed with care as they can clog the holes.
 Easy to clean.
Criteria for the proper use:
 Use a vibratory feeder to feed the material into the rotary splitter. Hand feeding results in uneven feeding
rates and therefore non-uniform incremental splitting.
 Adjust the feeder rate so that the material flows through the feeder at a continuously slow even rate
without overflowing into the rotary splitter. Each split (bottle) should contain around 200 increments per
split. 50 increments is the recommended minimum of increments per split (bottle). A slower feeding rate
results in more increments per split (bottle) and therefore in a more representative subsample.
 After splitting each bottle should contain an equal volume of material. If the volumes are not equal, this
indicates that one or more of the splitting chutes became clogged. When the volumes are not equal,
then all of the material needs to be recombined together and resplit again.
If the material contains large particles, it may have to be coarsely ground through a 4-6 mm screen before
splitting (see 8.5).
8.2.2 Fractional (alternate) shovelling
It is a very simple splitting technique with following advantages:
 Can be implemented in the lab or field.
 Does not involve extra equipment (e.g. rifflers).
 Has minimal clean up and decontamination requirements.
 Any number of splits can be generated.
 Has very low sample splitting error.
The laboratory sample is split into the desired number of samples by collecting increments from the lab
sample. The increments from the lab sample are alternately placed into containers or piles to form the split
subsamples.
If a sample is to split into two equal subsamples, one split would contain the odd increments and the other
subsample would contain the even increments.
If a sample is to be split into three subsamples, the first split would contain increments 1, 4, 7, and so on, the
second split would contain increments 2, 5, 8, and so on, and the third split would contain increments 3, 6, 9,
and so on.
For larger numbers of split samples, the same pattern is followed.
To calculate the mass increment follow formula can be used for:
[
mass (increment ) = mass (lab sample ) / number (splits ) ⋅ number (increments )
]
(9)
The following precautions must be considered:
© ISO 2009 – All rights reserved
21
ISO/DIS 6498
 All increments must be of approximately the same size.
 Each split must have the same number of increments.
 Split samples must be selected at random.
 Each split should have at least 30 increments, where possible.
 All the material must be used.
NOTE
Towards the end of the splitting process there can be a small amount of fines. It is advisable to reduce the
increment size so that the fines are equally apportioned among the splits with at least 10 increments each. If this is not
done, it is possible that all fines will end up in only one of the splits.
8.3
Particle size reduction
8.3.1 General methods

Chopping: A material is mechanically cut into smaller parts.

Crushing: Applying pressure to fragment larger particles into smaller fragments; specially, variable jaw
crushers reduce large, hard samples to 1-15 mm diameter particle size.

Cutting: Cutting mills reduce soft to medium-hard and fibrous materials using rotating and stationary
cutting knives. Reduced size depends on sieves used in combination with mill.

Blending (homogenizing): Materials are broken into smaller parts and blended to make them more
uniform in texture and consistency.

Macerating: A soft material is torn, chopped, or cut into smaller pieces.

Milling / grinding: Grinding of materials to mechanically reduce particle size is accomplished by cutting,
shearing, impacting and attrition using various mills.

Pressing: Liquids from semisolid materials such as plants, fruit and meats are squeezed out for
additional analysis.

Pulverizing: Describe the action of various mills that further reduces small feed size material (< 10 mm)
to a final fineness usually below 75 µm.
8.3.2 Requirements for choosing size reduction equipment
Requirements for suitable size reduction methods differ widely and depend on the sample material.
The equipment should not corrupt the subsequent results of analysis (e.g. cause contamination with trace
elements or heavy metals like chrome or nickel from abrasion). Identical results must be achievable in the
same lengths of time when the same grinding tools are used.
Considerations for selecting size reduction equipment for a specific application include the following:

Type of material: How hard is the material? What are the physical and chemical properties? Is the size
reduction process affected by heat generation, moisture change or chemical reactions?

The initial maximum particle size (e.g. chunks, powder, ...etc.).

The final desired particle size (mm, µm) and the range of permissible particle sizes.
22
© ISO 2009 – All rights reserved
ISO/DIS 6498

Quantity of material to be ground and number of laboratory samples to be processed daily or weekly.

Amount of time available for size reduction in the overall sample processing.

Abrasion resistance of the grinding tools: Contamination due to wear of the grinding or cutting elements
by the grinding tools is a constant threat and must be absolutely avoided. It is important to select
suitable grinding elements that are constructed from materials that will not interfere with the analysis.
Typically, particle size reduction tools are made of stainless steel, tungsten carbide, agate, sintered
alumina, hard porcelain, and zirconium. A tool with harder surfaces than the lab sample material is
desirable, and it will minimize contamination.

Versatility of grinding equipment: Due to the nature of some sample materials, wet grinding may be
necessary or the sample must have to be cooled or embrittled during size reduction. Some materials
must be ground in an inert atmosphere, with liquid nitrogen, or in a vacuum.

Requirements for operator time and cleaning equipment. It is impossible to grind laboratory samples
without loosing minute amounts of sample because some adheres to the grinding surface. This material
is lost during the cleaning.
8.3.3
Types of particle size reduction equipment
There is no industry standard for categorizing particle size reduction equipment. Following is an attempt of
'AAFCO sample preparation working group' [1] to describe and group equipment currently on the market and
potentially useful for feed laboratories (see table 9).
NOTE
Products of suppliers mentionned in these guidelines serve only as examples and not as proposals to illustrate
the classification of reduction equipments existing on the markets. Other products not mentionned here could also be
appropiate for reduction purposes.
8.3.3.1
Crushers
Reduce particle size by crushing the material. They are generally used to reduce a very large particle size
(diameters as large as 150 mm) to 0,5 – 1 mm fragments. Other types of mills can then be used to further
reduce particle size.

Jaw Crushers (e.g. Fritsch P-1, Retsch BB100, BB200, BB3, BB51, BB300): Provide a first step in a
sequential reduction of coarse materials. They operate by compressing material in a chamber between 2
strong breaking jaws – one stationary jaw and one moving jaw. The jaws are located between thick
panels that form a duct, which tapers down toward the adjustable discharge gap.
8.3.3.2
Mills
Mills can be grouped into cutting mills, grinding mills, combination cutting/grinding mills, impact mills, and air
jet mills.
8.3.3.2.1
Cutting (shearing) mills
Utilize blades or rotors to shear or cut the material. Can be classified as whether material is reduced by
rotating blades causing a reduction against fix blades, or by rotating blades throwing the material against a
sieve or an abrasive grinding ring. Can be purchased as either a floor model with feed sizes of typically 60100 mm and the final fineness ranges from 0,25 to 20 mm depending on the material (e.g. Thomas Wiley Mill,
Gilson 124, Fritsch P-15 and P-25, Retsch SM 100 and SM 2000, Romer Mill) or a bench model, where the
fineness of the grind is determined by sieves (e.g. Retsch ZM 100, Fritsch P-14, Udy Mill, Foss Cyclotec, Foss
Knifetec 1095).
© ISO 2009 – All rights reserved
23
ISO/DIS 6498
8.3.3.2.2
Grinding mills
Can be grouped into grinding action (impact, friction, shearing, attrition, etc.) and the corresponding initial and
final particle size of the material processed. These include ball mills, centrifugal mills, disk mills, planetary
mills, pulverizers, mortar mills, and vibrating cup or ring and puck mills.
 Ball mills (e.g. Spex Wig-L-Bug, US Stoneware Ball Mills, Retsch S 1 000, Fritsch Pulverisette 5). They
have hard balls inside an enclosed grinding jar or bowl that pulverize by impact and mix soft-hard, brittle,
and fibrous materials. Material is placed in a bowl (or jar) with grinding balls and rotated. Throughput and
milling efficiency can be affected by sizes and shapes of grinding jars, the rotation speeds, and the
number, weight and size of balls added to the grinding jar. Material particle size can generally be
reduced from 5-10 mm to < 10 µm. Ball mills can perform wet or dry grinding.
 Centrifugal mills (= centrifugal ball mills and centrifugal shearing mills). In the case of the ball mills it is
referring to the rotation of the grinding vessel. In the case of the shearing mill it is referring to the rotation
of the rotor.
 Disk or burr mills (e.g. Fritsch P-13, Straub 4E, Retsch RS1). They pulverise soft to medium-hard and
fibrous material in either a continuous or a batch mode. The material is pulverized by feeding between
stationary and slowly rotating grinding disks with radial teeth. Material is gravity fed to the centre of the
stationary disk and is progressively ground finer as it moves with the sloped grinding teeth until
discharged at the outer edge of the disks. Final particle size is set by adjustment of a gap setting.
Generally, these mills reduce materials from about 20 mm down to about 0,1 mm.
 Impact mills. A typical impact mill is a hammermill. These mills have a fast-moving part which collides
with a stationary part, compressing and fracturing material. Rotary hammers include the swing hammer
which is designated to break up relatively large pieces. In these mills further reduction occurs with
subsequent impact with casing or screen.
 Planetary mills (e.g. Fritsch P-5, P-6, P-7; Gilson LC-105, 107, 122; Retsch PM 400). They develop high
grinding energy via planetary actions. This type of mill employs a two-way planetary action to grind
rapidly by both impact and friction, resulting in very narrow particle size range. Material to be reduced is
placed in a bowl (or jar) with grinding balls and placed on a rotating platform. In planetary action, bowls
rotate opposite to direction of the bowl platform rotation and centrifugal forces alternately add or subtract.
Grinding balls roll halfway around bowls, then are thrown across the bowls to impact opposite walls at
high speed. Grinding is intensified by interaction of balls. High-energy planetary action gives a narrow
particle size range in shorter grinding times than conventional ball mills with gravity tumbling. This type
of mill can be used for dry or wet grinding of soft to hard/brittle materials or for mixing, homogenizing and
emulsifying of suspensions and pastes. Generally, particle sizes can be reduced from 10 mm to < 1 µm.
 Pulverizers. Term used for grinding mills that reduce material of an initial particle size of 4-6 mm down to
about 75-250 µm. There is no common mechanism or mode of action, the only common feature of these
mills is the very fine end product.
 Mortar grinders or mortar mills (e.g. Retsch RM 100 and MS; Fritsch P-2 and P-0). The mortar mill is an
automated version of the traditional mortar and pestle. A graded pestle is connected to a variable speed
overhead motor, and the material is crushed by pressure and friction between the grinding bowl and the
grinding arm (or pestle). Mortar mills can be used for both wet and dry grinding. The longer the grinding,
the smaller the final particle size. Particle size can generally be reduced from about 8 mm to 10-50 µm.
 Vibrating cup mill or ring and puck mill (e.g. Fritsch P-9). They utilize high friction and impact energy to
reduce samples. Inside a grinding vessel, a disc or disc/ring set are vibrated and accelerated by
centrifugal force. These are for extremely fast high-energy dry or wet grinding.
8.3.3.2.3
Combination grinder / cutting (cross beater) mills (e.g. Retsch SK 100)
Utilize both cutting (shearing) and grinding actions to reduce the particle size of a material. Materials are fed
through a chamber where samples are reduced repeatedly until it is small enough to fall through a sieve.
24
© ISO 2009 – All rights reserved
ISO/DIS 6498
8.3.3.2.4
Air jet mills (e.g. Gilson LC-600)
A high speed stream of air is created in the grinding chamber by introducing air supply through nozzles. The
material is fed into the steam at controlled rates with a feeder and pulled into the grinding chamber where it
undergoes a series of high velocity collisions resulting in pulverization of the particles. As the stream enters
the classifier, properly sized product is trapped by the exciting flow and conveyed to the collector. Oversized
particles remain in the stream until sufficient reduced. Regulation of particle size is achieved by controlling the
velocity gradient. The air supply can be a source of clean, compressed air or inert, compressed, bottled gas
such as nitrogen. Jet mills are used for materials that are abrasive, sensitive to contamination, sensitive to
heat, or volatile. Since the sample itself is the grinding medium, sample purity can remain very high. Final
particle size is in the 0,5 mm to 45 µm range.
Table 9 — Guide for representative types of grinders, pulverizers and cutting mills (adapted from
GILSON company, inc. catalog)
Applications
Temp. sensitive
95 x 95 mm
205 kg/h
1 - 15 mm
++
++
+
++
-
-
+
-
Friction
8 mm
150 ml
10 - 20 µm
-
++
+
++
-
+
++
++
Planetary Mill
Impact / Friction Force
10 mm
4 x 225 ml
< 1 µm
++
++
++
++
+
+
+
++
Soil Mill
Shearing Force
30 mm
2 kg
< 2 µm
-
++
-
-
-
-
-
-
Vibrating Cup
Impact Force
10 mm
250 ml
10 - 20 µm
++
++
+
++
-
+
-
+
Disk Mill
Shearing Force
20 mm
150 kg/h
0,1 - 12 mm
++
++
+
++
-
+
-
+
Cutting - Bench Mill
Shearing Force
10 mm
5 kg/h
0,08 - 6 mm
-
+
++
+
+
+
+
-
Cutting - Floor Mill
Cutting Shearing
60 - 100 mm 50-60 kg/h 0,25 - 20 mm
-
-
++
+
++
+
-
-
Ball Mill
Impact
< 8 mm
< 10 µm
+
+
+
+
-
-
-
-
4 - 6 mm
75 - 250 µm
+
+
+
Pulverizers
Moist
Pressure
Mortar Mill
Fibrious
Jaw Crusher
Tough
Final fineness
Brittle
Max.
capacity
Soft
Max. feed
particle size
Med.-hard
Working Principle
Hard
Equipment
Hard (abrasive): Ferro alloys, granite, iron ore.
Medium hard: Glass, quartz, calcite, ash, agricultural & horticultural soils.
Soft: Salts, talc, animal feed, foliage herbage, vegetation, pigments, spices, dragees.
Brittle: Salts, tablets; also after embrittlement with liquid nitrogen.
Tough: Leather, hides.
Fibrous: Wool, wood, cellulose, plant roots.
Temperature-sensitive: Pharmaceuticals and detergents.
Moist: Soil, grass, hay and leaves.
8.3.4 Maintaining integrity of the laboratory sample
To minimize moisture loss and preserve the integrity of materials that contain thermally labile or volatile
components, heating during the grinding process must be minimized.
Dry ice sometimes can be added directly to a mortar or ball mill to keep samples cool during grinding (dry ice
should be prepared from CO2 that is free from impurities that could contaminate the sample).
© ISO 2009 – All rights reserved
25
ISO/DIS 6498
Some mills can be fitted with a cooling block to permit the circulation of cool liquid during grinding.
Pulverizing the material under liquid nitrogen can be performed in a cryogenic mill if lower temperatures are
necessary to solidify a material.
When using cooling agents, it is necessary to avoid condensation of moisture on the material to preserve the
integrity of the sample.
8.3.5 Mixing techniques
Mixing is used by many in an attempt to 'homogenize' the sample. Once the sample is 'homogenized' any
increment or grab from the sample is then deemed representative of the sample without further consideration.
Nothing could be further from the truth for particulate matter.
While it may be possible to mix in some cases in other cases mixing actually promotes segregation. This is
especially true for material with divergent particle sizes and densities.
Always use caution when mixing particulate matter. While mixing may be advisable for many minerals, the
errors previously discussed still exist and must be addressed.
One type of mixing technique that is not recommended is putting a spatula in the top of a sample container
and stirring a few times.
A second unacceptable technique for dry, ground material is shaking a sample container if the jar is full or
almost full. The material at the bottom is not adequately mixed in and the stirring may have actually promoted
segregation making the mass reduction error even greater.
If mass reduction is performed with proper alternate shovelling or rotary riffling then mixing is not an issue and
does not have to be performed.
There are many other mixing techniques available that are superior to stirring (e.g. Drum mixer, stirring vane
mixer, V-shaped double cylinder mixer).
Mixing is a very effective technique to improve mass reduction accuracy for liquids and semi-solid materials
(e.g. canned pet food).
8.4 Partial drying
For 'wet' feeds with less than 85% dry matter (e.g. forages, total mixed rations, but not for liquids), it is
necessary to partially dry them prior to fine grinding to analyse stable substances; for not-stable substances a
partial drying is not possible.
Partial drying can be done by using either a forced-air oven or a microwave oven. The goal is to dry the
feeding stuff while keeping sample temperature below 55 – 60°C so that chemical composition is minimally
affected. Drying at higher temperature (greater than 60°C) causes chemical changes in the feeding stuff (e.g.
protein degradation). The dried feeding stuff should be equilibrated at room temperature for about 15 minutes
before measuring partial dry matter to minimize the potential change in moisture that can occur during
grinding and storage. Drying at low temperature (less than 60°C) does not remove all of the water from the
feeding stuff; therefore (initial) partial drying does not represent the total dry matter of the feeding stuff.
Following drying, the subsample is ground and analysed for (final) laboratory dry matter (the remaining 3 –
15% moisture) when other chemical constituents are determined.
Therefore, a two-step procedure for determining dry matter is recommended: First determine partial dry matter
(if less than 85% dry matter), then determine laboratory dry matter on ground test sample and multiply partial
dry matter times laboratory dry matter to determine total dry matter.
26
© ISO 2009 – All rights reserved
ISO/DIS 6498
Example: Procedure for partial drying of forages
Discard any roots from plants (if necessary) and brush off dirt particles. Chop entire laboratory sample to
pieces of 1 cm using either hand clippers, paper cutter, or the lab forage chopper. Include any ears attached
to corn plants. Silages do not need chopping.
-
Partial drying using draft oven:
The entire (chopped) laboratory sample or a subsample of 50 g (± 10 mg) are weighed in a tarred vessel (e.g.
aluminium-box 20 x 12 cm2 with edges of 4 cm) and dried at 55 – 60°C for 16 – 24 hours till the moisture
content is about 8% to 12%.
-
Partial drying using microwave oven:
The entire (chopped) laboratory sample or a subsample are weighed in a tarred dry paper boat and dried.
Drying times and power setting vary depending upon forage type and moisture content.
NOTE – 'Hot spots' can occur during drying using a microwave oven. A fire may start in a hot spot even though the rest of
the forage may still be wet. It is recommended to never dry forages at more than a 50% power setting. A high fire danger
is indicated while mixing the forages between drying cycles if the forage is too hot to touch, or if it is smoking or begins to
smell charred.
After drying the box or paper boat is cooled down for equilibrating to room temperature and then weighed for
determination of the partial moisture content of the pre-dried sample due to following formula:
Partial dry matter [%] = (W 3 − W1 )× 100 / (W 2 − W1 )
(10)
where
W 1 is the empty weight of container [g]
W 2 is the wet weight of forage and container [g]
W 3 is the dry weight of forage and container [g]
Together with the moisture content of the grinded test sample the total dry mass is calculated for reporting the
results of the stable analytes to origin mass or to 100% or 88% dry mass.
For analysis of mycotoxins and GMOs by PCR if possible the whole chopped laboratory sample is to be dried
at 55 – 60°C and grinded to adequate particle sizes.
Formula for calculation the total dry mass and for calculation analytes (A) from partial dry matter to total and
origin dry matter (DM):
DM total [%] = DM partial [%]⋅ DM test sample [%] / 100
(11)
Α total dry matter [%] = Α partial dry matter [%]⋅ 100 / DM test sample [%]
(12)
Α origin dry matter [%] = Α total dry matter [%]⋅ DM total [%]
(13)
8.5 Coarser grinding
When a 'dry feed' is composed of lumps or its particle sizes are more than 6 mm the whole laboratory sample
should be grinded by a jaw crusher or a cutting mill or chopped to particle sizes of 4 – 6 mm before mass
reduction / subsampling is initiated.
© ISO 2009 – All rights reserved
27
ISO/DIS 6498
Of course a "pre-grinding" (coarser grinding) of the whole laboratory sample to an adequate uniform low(er)
particle size before mass reduction is initiated is recommended to ensure homogeneity. But this is a very time
consuming procedure and aspects for non stable substances are to consider (e.g. heat generation, grinding at
the day of analysis) otherwise deterioration starts.
8.6
Special sample preparation procedures
For samples consisting of high fat, gelatine, and / or molasses some special sample preparation procedures
helps to get a representative subsample and fine grinded test samples:
 Place the entire laboratory sample in a freezer overnight and prepare the sample in a frozen or chilled
state.
 Use dry ice during splitting and grinding to keep the sample cold enough to prevent clumping or melting.
 Use of blending type mills, blend in intervals of 30 s.
 First coarse grind the whole laboratory sample to pass a 6 mm sieve.
All this steps enables to prepare fatty and / or sticky feeding stuffs whilst subsampling and / or fine grinding to
get representative test samples (see 100 - Annexes).
8.7 Storage
Once a representative test sample has been prepared from the laboratory sample, it is essential to maintain
its integrity throughout its duration in the laboratory, including all analytical process, reporting of data and the
ultimate disposal of any remaining material.
Proper storage may include storage at reduced temperature (refrigeration or freezing), protection from
moisture gain or loss, protection from UV light, and so on to avoid the harmful effect of too many
microorganisms which can break down the organic compounds.
The proper storage conditions will vary among the various types of feed materials and substances which are
to analyse.
In deciding upon the proper storage conditions for each material and analyte combination, the laboratories
need to consider the effect that composition, matrix interactions, chemical and/or enzymatic activity have on
the analyte(s).
Storage and disposal policy should be established and documented within the laboratory to address these
issues.
9
Performance tests (quality control)
Performance tests are described to estimate the sample preparation error, which vary with the procedure, the
material, the analytes and the analyst.
Performance tests can be used to evaluate new equipment by comparing results from previous equipment.
The basis is to test each preparation step using two or more very dissimilar materials that can easily be
separated so that the amount of error can be easily measured.
When choosing materials, maximize the concentration range of the analyte of interest in the materials. For
example, a mixture of sugar and salt can vary between 100% sugar + 0% salt and vice versa. A mixture of 9%
protein corn and 14% protein cattle pellets will have far less detectable protein error caused by segregation
due to its limited maximum protein range. Feed containing low concentrations vitamin A supplied in 0,5 mm
beads of 650.000 IE/g vitamin A presents a far greater preparation problem than simple grain mixtures.
28
© ISO 2009 – All rights reserved
ISO/DIS 6498
Consequently, preparation methods adequate for one analyte (e.g. protein) may be inadequate for another
analyte (e.g. vitamin A).
Room temperature, humidity, and air exchange filtration may affect sample preparation procedure quality,
especially in climates with major changes in humidity and/or temperature.
Use the equipment normally used for routine samples. It is essential that the performance tests should not be
restricted to the newest screens or best mills, as these will be give better results that are not so representative
for the performance of the equipment. The objective is to insure all equipment and analyst's techniques are
within requirements.
As an illustration, some performance test procedures are given below. Adopt these procedures to meet your
laboratories needs.
9.1 Performance test for mass reduction (splitting)
The test is to evaluate the splitter using a heterogeneous mixture.
Example: Use cleaned maize retained on a sieve with a particle size of > 5 mm and weigh 400,0 g in a jar.
Then add 40,0 g of oats retained on a sieve with a particle size of 4 – 5 mm, discarding the material retained
on the sieve with a particle size of > 5mm and passing through the sieve with a particle size of 4 – 5 mm.
Then add 4,0 g of alfalfa seed cleaned using a sieve with a particle size < 4 mm. Mix briefly by tumbling.
Split using your normal procedure. Separate the components of each split using sieves with particle sizes of <
4 mm and of > 5 mm and weigh.
Repeat above procedure a total of five times.
Calculate the average and standard deviation of the total split in the right and left splits. Does the splitter give
a 50:50 split?
Calculate average and standard deviation of the corn/oat/alfalfa recovery for the right and left splits separately.
Is there a bias between the right and left splits? What is the standard deviation (sr / Vr)?
Maintaining the ratio of components is critical during splitting. Calculate the percent recovery in each split for
each component as a percentage of the theoretical recovery.
Example: Left maize recovery [%] = 100 x [(actual left maize weight) / (actual weight left maize + oats +
alfalfa)] / [200 g / (200 g + 20 g + 2 g)]
Accumulate sufficient data to determine an acceptable variation for each splitting device.
For rotational dividers number the jars and compare all number 1 positions, and so forth.
9.2 Performance test for particle size reduction (grinding)
9.2.1
Grind quality and recovery
Clean the mill thoroughly. Record the empty weights of a 20, 30, 40 mesh sieve (see table 10) and pan.
Weigh 200,00 g of shelled maize. Grind, weigh recovered ground material. Sieve using a 20, 30, and 40 mesh
sieves. Calculate the weight retained on each sieve (total sieve weight – tare sieve weight) and the percent on
the 20, 30, 40 and percent passing through. Calculate percent recovery using total material recovered vs.
initial weight.
Clean mill and repeat using 100 g of a large granular form of sodium chloride, such as rock salt. Save ground
salt for the carryover test.
Percent recovery, particle size and carryover will vary with different matrices.
© ISO 2009 – All rights reserved
29
ISO/DIS 6498
Some materials may not grind properly without special treatment (e.g. materials with molasses). In some
mixtures, too fine a grind may result in excessive heating and destruction of analytes such as vitamin A. A mill
is definitely too hot when water, fat or urea condense on the neck of the inlet.
Table 10 — Conversion of mesh values (from: U.S. Bureau of Standards)
mesh
mm
mesh
mm
2,5
8,0
50
0,30
3
6,73
60
0,25
5
4,0
80
0,18
8
2,38
100
0,149
10
2,0
140
0,105
14
1,41
170
0,088
18
1,0
200
0,074
20
0,84
270
0,053
30
0,59
325
0,044
40
0,42
400
0,037
9.2.2 Carryover
Prepare a solution of 0,5% iodine in aqueous 5% potassium iodide. Store in dark / actinic bottle.
Detect the carryover cornstarch in the ground salt from above experiment by developing the purple-black
colour using iodine-iodate-solution. Weigh 10 g salt into a test tube, add 20 ml water, mix, then add two drops
of starch-iodine reagent. Compare intensity of colour with known spikes of the ground maize in pure salt.
Carryover with routine clean should generally not exceed 50 mg maize in 10 g of salt or 0,5 g carryover from
200 g ground.
It is important that the tests reinforce the need to process samples with a procedure which will minimize
carryover, such as processing medicated premixes and concentrates separately after low-level materials of
the same analyte.
9.3 Performance test for mixing
Grinding is often a source of segregation, since harder to grind particles may not pass the sieve as quickly as
softer particles. Improper mixing may only accentuate this problem.
The mixing container should never be more than two-thirds full, preferably one-half full.
Using your normal storage container, add one-fourth of the container volume of ground wheat or similar lightcoloured material. Layer on top one-forth volume of a ground coloured mineral mix. Mix using a rolling motion
with the centre of axis at 45 degrees to the bottle. Record the time needed to mix the contents until they are
visually blended. This unmixed layers method can be used to test any mixing method in the feed lab.
30
© ISO 2009 – All rights reserved
ISO/DIS 6498
Many mixing techniques may begin to have difficulty when the sample volume is over half of the total
container volume. Serious mixing problems generally occur if the container is over three-fourth full.
10 Annexes: Categories of feeds – special remarks and flow charts
This chapter gives examples of the most animal feeding stuffs with some important remarks to the listed
categories and a corresponding flow chart to illustrate sample preparation.
In general all equipments must be kept clean to avoid contaminating one sample with another, especially
when handling drug and antibiotic samples and particularly when handling materials that have a vitamin A
claim of more than a million IE/kg or have a drug claim in units of g/kg or mg/kg. When necessary, equipment
should be washed between samples. Particular care should be taken with premixtures.
© ISO 2009 – All rights reserved
31
ISO/DIS 6498
Flowchart – Overview for sample preparation in general
Laboratory sample
Dry feedstuffs
Wet feedstuffs
Other feedstuffs
(trade feedstuff)
(forages)
(e.g. liquids, molasses blocks)
> 85% dry matter
< 85% dry matter
Coarse grinding of
Chopping of
whole sample to 4 - 6 mm
Special sample preparation
whole sample to approx. 1 cm
if necessary
if necessary
Predrying possible
No predrying
Mass reduction
by riffle splitters, rotational dividers or alternate shovelling
Test Sample(s)
Reserve Sample
Particle size reduction
No particle size reduction
with different sieve size options
(Exception: Analysis of mycotoxins, GMOs)
< 0,5 mm
Trace minerals
0,5 mm
Amino acids,
starch, lactose, sugar
Store:
32
at room temperature,
1,0 mm
Nutrients,
not stable analytes
refrigerated
No grinding
Microscopy,
probiotics,
NIR, NIT
or frozen
© ISO 2009 – All rights reserved
ISO/DIS 6498
10.1 Birdseed
The ingredients of birdseed have the tendency to separate due to differences in density and particle size of
the assorted grains and oilseeds. Therefore, a rotary splitter or a gated riffle splitter is used to subsample
birdseed. Save a reserve sample for microscopy and/or for an examination to detect the presence of noxious
weed seeds.
Grind test samples to a fine particle size to achieve uniform dispersal of all the ingredients. Oilseeds grind
best in a blending type mill or using cooling techniques such as dry ice.
© ISO 2009 – All rights reserved
33
ISO/DIS 6498
10.2 Whole cottonseed
Whole cottonseed is received unprocessed (e.g. with the hulls and lint surrounding the meat). Due to
difficulties in riffling caused by the non-uniformity of the lint-bearing seed, the entire laboratory sample is
ground through a 6 mm screen. The remaining lint then is removed by hand by passing the meat through a 6
mm sieve.
The separated cottonseed meat is ground to a fine particle size. If mycotoxins are to be determined, an
analytical subsample is stored frozen to prevent mold growth. A reserve of unground sample may be saved in
case of a question concerning subsampling or grinding or a dispute in results of analysis.
Nutritional analysis are done on both the lint fraction and the meat fraction. The analytical results are
combined mathematically based upon the weights of the total lint fraction and the total meat fraction:
NW = (NL × WL ) / [(WL + WM ) + (NM × WM ) / (WL + WM )]
(14)
where
NW =
nutritional result for the whole cottonseed
NL =
nutritional result for the lint fraction
NM =
nutritional result for the meat fraction
WL =
weight of the lint fraction
WM =
weight of the meat fraction
Precaution must be taken to keep all equipment clean to avoid contaminating one sample with another. When
necessary, equipment should be washed between samples.
34
© ISO 2009 – All rights reserved
ISO/DIS 6498
10.3 Mineral mix
Since differences in density and particle size may cause the components of mineral mixes to separate, it is
advised to do the mass reduction using rotary or riffle splitters. The sample is mass reduced until the mass of
the test sample is adequate. Save ungrounded reserve and test samples for microscopy and for drug and
vitamin analysis at low temperatures, these analyses require large sample sizes.
Grind the test sample to a fine particle size for mineral analysis. Do not grind the vitamin and antibiotic test
sample(s) if the particle size distribution is adequate, otherwise grind the test sample(s) at the day of analysis
to a particle size of 1,0 mm. Rupturing, mixing, and interaction of the ingredients with the heat of grinding can
cause deterioration of vitamins. Grinding also introduces air into the sample, causing oxidation. Refrigerate or
freeze the vitamin test sample to prevent degradation.
Splitting or grinding a feed sample immediately after splitting or grinding a mineral mix creates a potential for
contamination. Grouping samples by analyte concentration levels to avoid contamination is a good practice.
Process the feeds and lower level mixes first. When necessary, wash equipment thoroughly whenever the
potential for contamination exists.
© ISO 2009 – All rights reserved
35
ISO/DIS 6498
10.4 Dry feeds
Since differences in density and particle size may cause the components of dry feed to separate, use a riffle
splitter or a rotational divider to divide the sample into equal fractions. Save reserve ungrounded sample for
microscopy and/or subsampling if required by laboratory policy. Save a separate container of ungrounded
laboratory sample for vitamin and/or drug analysis since this analysis requires large sample sizes.
Grind one of the test samples to a fine particle size for proximate and mineral analysis. Test samples for
vitamins and certain antibiotics are not ground until the day of analysis. In the interval, to prevent degradation
refrigerate or freeze vitamin and drug test samples. Rupturing, mixing, and interaction of the ingredients with
the heat of grinding expedites deterioration. Grinding also introduces air into the sample, causing oxidation.
Splitting or grinding a feed sample immediately after splitting or grinding a mineral mix, vitamin or drug premix
creates a potential for contamination. Grouping samples by analyte concentration levels to avoid
contamination is a good practice. Finished feed should be processed before premixes or mineral mixes. When
necessary, wash equipment thoroughly whenever the potential for contamination exists.
Store samples in air-tight containers to prevent a change of moisture content.
36
© ISO 2009 – All rights reserved
ISO/DIS 6498
10.5 Forages inclusively silage, hay, haylage, TMR and by-products
Due to the diverse nature of forage materials, take care to obtain test samples of the laboratory sample.
Coarsely grind and mix the entire sample. Save a reserve portion of the coarsely ground material in case of a
question concerning sample reduction or analysis. Store dry reserve samples at room temperature; refrigerate
or freeze wet reserve samples. For some microorganisms (e.g. yeasts) a freezing of the sample for storage
could lead to degradation when the sample is defrosted.
Most forage materials received at a laboratory fall into one of the following categories:
 those dry enough to grind and analyse immediately (dry matter ≥ 90% , e.g. grass or native hays, alfalfa
pellets)
 those dry enough to be coarsely ground to pass a 4 mm to 6 mm sieve, but too wet to be finely ground
(dry matter ≥ 85%, e.g. legume hays)
 those samples which need to be partially dried before the sample can be coarsely ground (dry matter
< 85%, e.g. silages, fresh plants).
The “wetness” at which a forage material can be ground will be determined by the forage material itself, the
type of grinder to be used, and the fineness of the grind. Most forages at
© ISO 2009 – All rights reserved
37
ISO/DIS 6498
> 85% dry matter can be ground to pass a 4 mm to 6 mm sieve using a cutting mill without problems (sticking
in the mill, moisture loss, etc.). However, when using a mill to grind forage material to pass a 1 mm sieve,
most forages materials need to be > 90% dry matter to grind properly.
Forage materials at approximately ≥ 85% dry matter that are too large to grind in their entirety to the fineness
required for analysis are first ground through a cutting mill to pass a 4 mm to 6 mm sieve. Reduce the coarse
ground sample in a gated riffle splitter or by alternate shovelling. When necessary, the partially dried and
reduced analytical sample is ground to the fineness desired for analysis.
For carotene analysis split out a subsample after grinding through a 6 mm screen and regrind the subsample
using a 1 mm screen, place in air tight container and freeze.
For mycotoxin analyses use the entire laboratory sample when no other nutritional analysis are required.
If both analyses are required split out a nutritional test sample after coarse grinding and use the reminder of
the material for mycotoxin analysis. Weigh the entire laboratory sample or subsample into a dry, tarred pan(s).
Dry in forced draft oven. Equilibrate to room temperature and weigh. Grind to pass a 1 mm screen. Transfer
the entire ground sample to a large wide mouth bottle and mix by tumbling for at least 5 minutes. Bottle must
be less than 2/3 full to allow for mixing.
Forage materials < 85% dry matter need to be partially dried prior to grinding. The wet material is chopped, if
needed, to facilitate drying and subsampling. The entire wet laboratory sample is dried, then coarsely ground
to pass a 6 mm screen and then subsampled.
Certain forage materials may require special handling to avoid loss of the analyte during sample preparation:
Cyanide (hydrocyanic acid) is rapidly released from chopped sorghum / sudangrasses. Place a representative
portion of the chopped wet laboratory sample in a plastic bag, seal, and immediately freeze. If whole plants
were submitted from a grazing situation, sample only the portion which would be grazed.
Ammonia and organic acids (lactic acid, acetic acid, butyric acid) may be volatilized from ammoniated silages
during oven drying. In this case, split out a representative portion of the undried laboratory sample, place in a
plastic bag, seal and freeze, needed to be saved for protein, total nitrogen, non-protein nitrogen or ammonia
or organic acid determination.
38
© ISO 2009 – All rights reserved
ISO/DIS 6498
© ISO 2009 – All rights reserved
39
ISO/DIS 6498
10.6 Oilseeds and high fat feeds
Oilseeds are easily subsampled using either a riffle splitter or a rotational divider. Due to the oil content, it may
be difficult to grind oilseeds in mills routinely used for dry feeds. Blending type mills work pretty well to grind
oilseeds. Keeping the grinding chamber cold (by circulating coolant around it) or by grinding with dry ice helps
to prevent the melting and/or the oxidation of fat during grinding.
Subsampling hard fat and suet is extremely difficult since the fat particles cling to one another and to the
equipment. Alternate shovelling is probably the best way to obtain a subsample. Blend hard fat (or suet) in a
frozen state to prevent the melting and/or oxidation of the fat. Use precautions during freeze blending to
prevent atmospheric moisture from condensing on the sample which may alter the results.
40
© ISO 2009 – All rights reserved
ISO/DIS 6498
10.7 Large block and molasses block feeds
Accurately subsampling from this sticky material is difficult. A recommended approach is to place the entire
laboratory sample in a freezer overnight. Use a mallet and break it into pieces small enough to be ground in a
cutting mill. Using plenty of dry ice, grind the entire laboratory sample through the cutting mill using a 4 mm to
6 mm screen. Split the laboratory sample using a riffle splitter. Use generous amounts of dry ice during
grinding and splitting to keep the laboratory sample cold enough to prevent clumping. Place the reserve
sample in a plastic bag and store in a freezer. Regrind the analytical samples using a shearing mill or a
blending mill and dry ice.
When using frozen laboratory samples, perform the entire process as quickly as possible to avoid clumping
and condensation of atmospheric moisture onto the sample which will alter the analytical results.
Process large feed and mineral blocks in a similar manner except freezing and dry ice are not needed.
Follow correct procedures for splitting and grinding so that the ground test samples are representative
portions of the laboratory sample received. Take special care to transfer the correct labels to sample
containers so that sample numbers are accurate.
© ISO 2009 – All rights reserved
41
ISO/DIS 6498
10.8 Liquid feeds
The ingredients of a feed have a tendency to separate due to differences in density and particle size.
Therefore, a liquid feed sample is shaken, mixed, or blended thoroughly before obtaining the test sample. The
container needs to have sufficient space to allow mixing. Subsample during mixing or immediately following
completion of mixing. All liquids feed samples are refrigerated to prevent degradation. For some
microorganisms (e.g. yeasts) a freezing of the sample for storage could lead to degradation when the sample
is defrosted.
42
© ISO 2009 – All rights reserved
ISO/DIS 6498
10.9 Canned pet food
Take care when chopping and mixing the canned pet food material to include all of the liquid, fat, gelatine, and
any other ingredients. Conduct sample preparation expeditiously to minimize the loss of moisture due to
evaporation. Grind samples by blending in a food processor. The fat has a tendency to adhere to the walls of
the bowl especially with a prolonged blending time. Therefore, blend in intervals of 30 seconds each and wipe
down the walls between each interval. Blend high fat containing samples in a chilled state to prevent
separation of fat.
© ISO 2009 – All rights reserved
43
ISO/DIS 6498
10.10 Semi-moist pet food and dog chews
This tough leathery product grinds best in a mill that uses sharp knives such as a cutting mill or a blender.
The drier the material the easier it is to grind. Moist material needs to be ground in a frozen state.
Subsampling is best done using either a gated riffle splitter or by alternate shovelling.
Fine grind the analytical sample by either freeze blending or using a cutting mill. Moist materials grind better
by freeze blending while dry materials can be ground in a cutting mill.
44
© ISO 2009 – All rights reserved
ISO/DIS 6498
10.11 Premixtures
Since differences in density and particle size may cause the components of premixes to separate it is advised
to do the mass reduction using rotational or riffle dividers. The sample is mass reduced until the mass of the
test sample is adequate. Save a reserve sample if required by laboratory policy.
The premixes generally are of a particle sizes that do not require grinding. If grinding is necessary, grind for a
minimum amount of time to avoid heat generation. Rupturing, mixing, and interaction of the ingredients with
the heat of grinding expedite deterioration. Grinding also introduces air into the sample, causing oxidation.
Refrigerate or freeze the sample to prevent degradation.
Splitting or grinding a feed sample immediately after splitting or grinding a premix creates a potential for
contamination. Grouping samples by analyte concentration levels to avoid contamination is a good practice.
Process feeds and the lower level mixes first. When necessary, wash equipment thoroughly whenever the
potential for contamination exists.
© ISO 2009 – All rights reserved
45
ISO/DIS 6498
10.12 Range and alfalfa hay cubes
To obtain a representative subsample from a bag containing large cubes and small particles (fines) is
challenging. An unacceptable practice would be to randomly select 3 cubes – 6 cubes, ignoring fines, for
grinding from the 15 cubes – 20 cubes submitted to the laboratory.
Grind the entire laboratory sample including the fines through the cutting mill using a 4 mm to 6 mm screen.
Use of a 4 mm to 6 mm screen reduces the chance of heating and speeds the process of grinding. Collect all
the ground material, mix, and split the sample into two or more representative portions using a riffle splitter.
Return the reserve portion of the ground sample to the original plastic bag and if it contains heat-sensitive /
degradable analytes, store in a freezer. Otherwise, store it at room temperature in an approbate place. Fine
grind the different test samples for stable and not-stable analytes.
46
© ISO 2009 – All rights reserved
ISO/DIS 6498
10.13 Texturized and sticky feed
Their sticky nature makes texturized feeds difficult to split and grind. The grinding of a sticky material in a cold
state helps to prevent the material from clumping together and reduces the potential for particles to adhere to
the grinding equipment.
To obtain representative subsample of a sticky feed, first coarse grind the feed to pass a 6 mm screen.
Subsample feeds that are too sticky for a riffle splitter by the fractional shovelling technique.
Grind test samples to a fine particle size to achieve uniform dispersal of all ingredients. Feeds that are too
sticky to grind in mills routinely used for dry feeds can be ground by blending type mills.
Store test and reserve samples in air-tight containers to prevent a change in moisture content. Use
precautions during freeze blending to prevent atmospheric moisture from condensing on the sample which
may alter the results.
© ISO 2009 – All rights reserved
47
ISO/DIS 6498
10.14 Aquatic feeds
Grind one of the test samples to a fine particle size for stable analytes e.g. nutrional and mineral analysis.
Test samples for vitamins and certain antibiotics are not ground until the day of analyse, these analyses
require large test sample sizes. In the interval, to prevent degradation refrigerate or freeze vitamin and drug
samples. Rupturing, mixing, and interactions of the ingredients with the heat of grinding expedite deterioration.
Grinding also introduces air into the sample, causing oxidation.
Store samples in air-tight containers to prevent a change of moisture contents.
48
© ISO 2009 – All rights reserved
ISO/DIS 6498
© ISO 2009 – All rights reserved
49
ISO/DIS 6498
Bibliography
[1]
Association of American Feed Control Officials Incorporated (AAFCO), Guidelines for Preparing
Laboratory Samples, prepared by : AAFCO Laboratory Methods and Service Committee – Sample
Preparation Working Group (Nancy Thiex, Lawrence Novotny, Charles Ramsey, George Latimer,
Laszlo Torma, Robert Beine), Second Edition March 2003.
[2]
ISO 6498:1998, Animal feeding stuffs – Preparation of test samples.
[3]
Verband Deutscher Landwirtschaftlicher Untersuchungs- und Forschungsanstalten (VDLUFA),
Methodenbuch Band III, Nr. 2 (mit 6 Ergänzungslieferungen 1983, 1988, 1993, 1997, 2004, 2006).
[4]
Lars Petersen, Pentii Minkkinen, Kim H. Esbensen: Representative sampling for reliable data analysis.
Chemometrics and Intelligent Laboratory Systems, 77 (2005), 261-277.
[5]
T. Allen and A. A. Kahn: Critical Evaluation of Powder Sampling Procedures. The Chemical Engineer,
May (1970).
[6]
Lars Petersen, Casper K. Dahl, Kim H. Esbensen: Representative mass reduction in sampling - a
critical survey of techniques and hardware. Chemometrics and Intelligent Laboratory Systems 74
(2004), 95– 114.
[7]
Pentti Minkkinen: Practical applications of sampling theory. Chemometrics and Intelligent Laboratory
Systems 74 (2004), 85-94.
[8]
Council Directive 79/373/EEC of 2 April 1979 on the marketing of compound feedingstuffs.
[9]
Council Directive 93/74/EEC of 13 September 1993 on feedingstuffs intended for particular nutritional
purposes.
[10]
Council Directive 96/25/EC of 29 April 1996 on the circulation and use of feed materials, amending
Directives 70/524/EEC, 74/63/EEC, 82/471/EEC and 93/74/EEC and repealing Directive 77/101/EEC.
[11]
Regulation (EC) No 178/2002 of the European Parliament and of the Council of 28 January 2002
laying down the general principles and requirements of food law, establishing the European Food
Safety Authority and laying down procedures in matters of food safety.
[12]
Commission regulation (EC) No 1831/2003 of the European Parliament and of the Council of 22
September 2003 on additives for use in animal nutrition.
[13]
Central Committee of the German Agriculture, Standards Commission for Straight Feedingstuffs (Berlin
2007):
Positive
List
for
Straight
Feedingstuffs
(Feed
Materials),
6th
Edition
(http://www.futtermittel.net/pdf/positivliste_6e.pdf).
50
© ISO 2009 – All rights reserved