Summary. The Chemical Section of OMH1-2 have embarked on the preparation of wheat sample series as CRMs in 1992. The certification process are cerried out accordingly to the recommendations of ISO. Since then we have developed a series of flour samples as well. The investigations of the long-term stability and the application of wheat and flour CRMs are continouos.
Wheat is one of the most widely grown crops in Hungary and it is one of the major determinative factor in the economy. Its uniform and objective qualification is of great importance. There are well equipped laboratories available with sufficient experience, however, as the proficiency testing regularly showed, certified samples need to be used to achieve exact and uniform measuring results.
Choosing the Sample
The certified reference wheat series consists of three, the flour serie of two samples, they are produced – in the required quantity – annually, from the new harvest. For the wheat samples clean-bred, sound material is used from various habitats.
The most important conditions of the selection are:
– in case of wheat the difference of protein and wet gluten content
– in case of flour the difference in grindings, the difference of wet gluten content and that of the value of water absorption must be at least three times greater than the permissible error of measures between various laboratories.
Sample Preparation
The wheat samples
– are cleaned, the broken and the other grains and the dust are removed
– the homogenisation is carried out with TURBULA Shaker
The prepared samples are filled in the necessary quantity into moisture-proof triplex foil bags and vacuum-closed. The packed material is sterilised by 5 KGy radiation dose.
Testing for Homogenity by protein content measurement results
12 containers are selected at random after dividing the sample bulk into the containers that will be distributed to the participants. The content of each of the 12 selected containers are homogenised separately and from each of them two test portions are taken.
The sampling variance (ss) are estimated from the results of randomly analysed each second test portions.
The analytical variance (sa) are estimated from all results by one-way analysis of variance without exclusion of outliers. There are no significant differences between the samples on the basis of F-test.
So the material is sufficiently homogenous for the purpose of preparation CRM.
Certification Campaign
For testing the samples are sent to 15 accredited laboratories which are suitably experienced in testing cereals. They are to carry out the tests with great care according to the currently standard procedure. Some measurement method are specified:
– measurement of the moisture content by drying the samples at 130-133 ℃ for 2 hours;
– determination of the protein content by measuring the Kjeldhal nitrogen after digestion; and
– determination of wet gluten by manual or mechanical gluten washing.
The organisation of the interlaboratory test and the evaluation of the results are based on ISO GUIDE 353 (Certification of Reference Materials – General and Statistical Principles). Cohran’s Test is used for analysing the deviation in the laboratory’s replicate measurement results and for discard the outlier values: Grubb’s Test helps us to find the outlier results on inter- and intralaboratory basis, and to discard them from the evaluation.
Based on the results the consensus value is certified and its uncertainty is defined.
The Long-term Stability of the Certified Values and the Storability of the Samples
In 1995. the certified samples – prepared with the method above – were kept at room temperature. During their storage they were tested six monthly at 18 laboratories regarding their certified and other parameters.
The evaluation of the results shows that during the storage the samples’ internal properties changed within the uncertainly of the certified values. These results prove that the packaging and storage we apply is suitable for wheat and flour CRMs.
Using Wheat and Flour Samples
We have used a certificate compaign the same samples for proficiency testing. The material is sufficiently homogenous for the purpose of the proficiency testing.
In the test thirteen laboratories took part, all of them differs and are independent of the above mentioned laboratories. The organisation of test and the evaluation of the data was based on the ISO REMCO N280 19934 (PROFICIENCY TESTING FOR CHEMICAL ANALYTICAL LABORATORIES). Based on the resulting “z-score” values the laboratories are classified. The results of several laboratories are significantly different from each other and from the average of the measurement, they exceed the alloved target value for standard deviation. The calculated z-score of laboratory measurement is clearly shows the direction and the extent of the deviation of laboratory result from the consensus value. If the z-score value is 3 of an measurement the laboratory has to carry out improving activity for which our samples are efficiently used.
Conclusion.The test described above as well as the need of laboratories, producers of instruments and services for our samples prove that our goal has been achieved and we have succeeded in developing a practically usable CRMs series.
The proposed application of the CRMs series are
– assesment of different measurements
– adjustment of the calibrate measuring instrument
– validation of different measurements.
The results of our examinations demonstrate that our wheat and flour samples are suitable for checking and detection of mistakes and for the unifying of measuring results.
dr. Judit BUDAI and Judit FÜKŐ National Office of Measures (OMH) Budapest, 1124 Németvölgyi u. 37-39, Hungary