(dieteticmilkpowder).Theresults indicatethat thismethodcan
determine,with sufficient precision, from53.6 to18.5mg/kg.
(
j
)
Horwitz equation
.—Comparison of the Horwitz
functionwiththatobtainedfrom45results (i.e., nineelements
in five matrixes) of the collaborative study highlights
excellent precision as shown in Figure 1. This graph shows
that the trend of data from collaborative study follows the
theoretical Horwitz function, as the slope is similar. The
intercept is significantly different, and hence, this consistent
and constant deviation from predicted HorRat value can be
explainedby theexcellent experienceand trainingof analysts
(microwave digestion handling/ICP practice) in laboratories
that were also informed of the maximum concentration of
each element in the fivematrixes tested.
Conclusions
MethodPerformance
The MDO and MDC procedures used for SLV and RT
wereshown tobeeffectivewithacceptableagreement toward
certified and in-house reference values in terms of recovery
for most samples covering the nine sectors of the food
triangle.
This ICP-AESanalyticalmethodwasproven tobe simple,
selective, accurate, and reliable for the determination of Ca,
Cu, Fe, K, Mg, Mn, P, Na, and Zn in most food matrixes.
Recommended and alternate analytical lines corrected by
appropriate IS and Cs ion buffer concentration ranging
between 0.1 and 1% (w/v) were validated according the
statistic treatment of selectivity, sensitivity, linearity,
accuracy, and precision through SLV using ICP-AES
equipment with axially viewing plasma after MDC, and
through ruggedness tests using ICP-AES equipment with
radial anddual viewings of the plasma afterMDCandMDO
procedures.
Performance characteristics reported for 13 certified and
in-housereferencematerialscoveringthetrianglefoodsectors
fulfilled AOAC criteria and recommendations in terms of
accuracy (trueness, recovery, and
z
-scores) and precision
(repeatability and reproducibility RSDs, HorRat values)
regardingSLVand collaborative study.
The determination of nine nutritionalminerals in fortified
foodproductsbyICP-AESisfit-for-purposeaccordingtoISO
17025normandAOACacceptability criteria.
Improvement of AOACOfficialMethod
984.27
Main improvements compared to the AOAC Method
984.27
are: (
1
)Theuseofmicrowavedigestionsystemswitha
singleacid(nitricacid) foranoptimizedsamplepreparationin
order to improve element recoveries fromdifficult matrixes
and to increase sample throughput, favoring safety
precautions and time-savings for operators in laboratories. It
is a significant improvement regarding long, time-consuming
aciddigestionwithunsafeacidhandling(HNO
3
/HClO
4
)used
for AOAC Method
984.27
. (
2
) The use of appropriate
analytical wavelengths for each element of interest and the
automatic addition of a solution of appropriate IS and
ionization buffer to correct for physical and chemical
interferences (i.e., to compensate for matrix effects induced
bythecomplexityof thefoodsamples).Theaimis toimprove
short-term accuracy (repeatability) and long-term stability
(reproducibility and calibration curve validity in a long
analytical batch). Neither internal standardization nor ion
bufferareusedforAOACMethod
984.27
. (
3
)Theapplication
for all food matrixes covering all of the nine AOAC food
triangle sectors, including infant formula, which were the
uniquetypesofmatrixesvalidatedforAOACMethod
984.27
.
This fully validated multielemental method is
cost-efficient, time-saving, accurate, and fit-for-purpose. It is
proposed as an improved version of AOACOfficialMethod
984.27
for fortified foodproducts, including infant formula.
P
OITEVINETAL.
: J
OURNALOF
AOACI
NTERNATIONAL
V
OL
. 92,N
O
. 5, 2009
1517
Figure 1. Ring trial: Comparisonof Horwitz equationandexperimental equation found for the ring trial.
MTE-01
FORWORKINGGROUP/ERPUSEONLY
DONOTDISTRIBUTE
