Non-matrix Matched Glass Disk Calibration Standards Improve XRF Micronutrient Analysis of Wheat Grain across Five Laboratories in India

Abstract

Within the HarvestPlus program there are many collaborators currently using X-Ray Fluorescence (XRF) spectroscopy to measure Fe and Zn in their target crops. In India, five HarvestPlus wheat collaborators have laboratories that conduct this analysis and their throughput has increased significantly. The benefits of using XRF are its ease of use, minimal sample preparation and high throughput analysis. The lack of commercially available calibration standards has led to a need for alternative calibration arrangements for many of the instruments. Consequently, the majority of instruments have either been installed with an electronic transfer of an original grain calibration set developed by a preferred lab, or a locally supplied calibration. Unfortunately, neither of these methods has been entirely successful. The electronic transfer is unable to account for small variations between the instruments, whereas the use of a locally provided calibration set is heavily reliant on the accuracy of the reference analysis method, which is particularly difficult to achieve when analyzing low levels of micronutrient. Consequently, we have developed a calibration method that uses non-matrix matched glass disks. Here we present the validation of this method and show this calibration approach can improve the reproducibility and accuracy of whole grain wheat analysis on 5 different XRF instruments across the HarvestPlus breeding program.

An energy-dispersive X-ray fluorescence method for analysing Fe and Zn in common bean, maize and cowpea biofortification programs

“…Unfortunately, from the data shown here it is evident that the calibration equations are not similar enough to enable this without compromising accuracy for XRF screening. We have previously shown the benefits of employing a non-matrix matched calibration for calibrating XRF instruments for the analysis of wheat, rice and pearl millet (Guild and Stangoulis 2016). Using glass beads to calibrate the instrument this eliminates the need to carry a plant material internationally when calibrating a new instrument, hereby reducing quarantine difficulties.…”

“…We have employed this same approach for the crops studied here. The glass calibration was recently developed (Guild and Stangoulis 2016) with 10 glasses each analysed with the appropriate flour calibration method five times, and the average of these results averaged to determine a Bmatrix adjusted^calibration set. The All units are presented as mg kg −1 , apart from RSD (%) and r analysis of the validation set with the developed glass calibration correlates strongly with the flour calibration XRF analysis (Fig.…”

“…Analyses were conducted in supplied sample cups prepared as reported previously (Paltridge et al 2012a, b;Guild and Stangoulis 2016), with 4 μm Poly-4 XRF sample film used to seal one end of the cup. Sample cups were cleaned and prepared prior to each analysis to minimise cross-contamination between samples.…”

Progress update: Crop development of biofortified staple food crops under HarvestPlus

“…For high-throughput screening, HarvestPlus and its partners have developed X-ray fluorescence (XRF) spectrometry calibrations and glass standards. This new technology has proven cost-and time-efficient for mineral screening of a wide range of crops including wheat, rice, pearl millet, beans, sorghum, lentil, cowpea, and Irish potato [18,19,20]. To date, HarvestPlus has implemented 25 state-of-the-art micronutrient analytical laboratories at nine CGIAR centers, 12 NARS, and four universities, and trained more than 100 laboratory staff in 13 countries in field sampling, sample preparation, equipment calibration, and operation.…”

Identification of High-Yielding Iron-Biofortified Open-Pollinated Varieties of Pearl Millet in West Africa

“…Pearl millet grain micronutrient estimation was carried out using the energy dispersive X-ray fluorescence (EDXRF) ( Paltridge et al, 2012a , b ; Guild and Stangoulis, 2016 ) method. The EDXRF analysis was carried out using Oxford X-Supreme 8000 (Oxford Instruments plc, United Kingdom) with a 10-sample carousel.…”