Quantitative X-ray diffraction analysis of quartz in bulk samples: outcomes of an international interlaboratory test...

The quantification of crystalline silica (quartz) in materials helps communicate the potential risk of exposure when the materials are then included in a work process that could generate an aerosol of particulate. Although many analytical laboratories have developed procedures for quantification of quartz in bulk samples, these procedures are not standardized and are generally limited to their individual needs. Therefore, the International Standards Organization (ISO) working group for silica measurement (ISO/TC146/SC2/WG7 Silica) developed an international standard that describes 3 analytical methods for the determination of quartz in bulk samples by X-ray diffraction (XRD): the external standard method for samples prepared on filters, the internal standard method, and the spiking method. This paper describes the results of an interlaboratory measurement comparison in which the 3 XRD standardized methods were evaluated to verify if the 3 methods produce results that agree with each other. Simple analytical systems, represented by 4 binary mixtures of quartz added to another mineral, were prepared. Sub-samples of the 4 test powders were delivered to the 7 participating laboratories from 6 countries that agreed to participate in this study. The precision of each measurement method, calculated as a percentage of quartz weight within the range 10 to 16 wt%, was similar for the 3 XRD methods. The repeatability SDs, sr, ranged between 0.4 and 0.8 (relative SD, RSDr, between 4.3% and 8.7%) for thin specimens on filter, 0.4 to 1.1 (RSDr between 4.2% and 10.6%) for the internal standard method, and 0.3 to 1.2 (RSDr between 3.7% and 11.9%) for the spiking method. The reproducibility SDs, sR, of the results in the studied wt% range were between 1.1 and 2.1 (RSDR between 12.7% and 21.5%) for the method of thin specimen on filter, 1.1 to 1.4 (RSDR between 7.5% and 14.1%) for the internal standard method and 1.1 to 2.6 (RSDR between 9.5% and 23.7%) for the spiking method. These methods obtained equivalent results at 2 concentration levels of 10% and 16% in these binary mixtures. A further step of interlaboratory testing should be implemented to investigate the performance of the 3 methods at the lower quartz concentration values.