<< Back to Issues and Controversy: Measurement of Crystalline Silica

Disclaimer: This is not DOL or OSHA controlled material and is provided here for reference only. We take no responsibility for the views, content, or accuracy of this information.


X-ray diffraction techniques are appropriate for the identification and quantification of crystalline silica phases. In fact, they are the only positive way to identify which of the crystalline phases is present in environmental samples. At present the accuracy of quantification in respirable samples is around 4 % absolute and 1 % for bulk samples. This accuracy may be improved with a few modifications to the diffraction equipment used to collect the data. Detectability limits are around 3µg for respirable quartz on filter substrates and 0.03 weight percent in bulk samples where there are no interferences.

The methods for bulk samples may be improved significantly by preconcentrating the silica phases with chemical treatments prior to the diffraction measurements. Eliminating 90 % of the matrix in a bulk sample increases the silica level by an order of magnitude, and the 0.1 weight percent would then be 1 %. Even in the presence of interfering phases, modern diffractometry with digitized intensity information and new mathematical approaches to decomposing overlapped diffraction peaks will provide the necessary intensity values for phase quantification.

Certifying a single standard procedure for bulk analysis by X-ray diffraction is unwise and probably impossible. The types of samples and differences in matrixes are too varied to allow a single method to be employed. Although the matrix does not affect the diffraction from the silica phase per se, it does partially mask the phase by contributing to the absorption effect of the sample. Also, some diffraction peaks may interfere with the silica peaks in some cases. It is far more appropriate to develop a set of calibrated samples and criteria for certifying each individual procedure established by independent laboratories for specific sample types. The certification should require prescribed reproducibility and accuracy on test samples of comparable composition to the samples for which the procedure is intended.

Quantification is very sensitive to sample preparation methods primarily because of the tendency of the particles to orient crystallographically when packed in a sample. The small size of the respirable silica grains and the lack of a cleavage or other orienting influence minimizes this problem. Even in bulk samples whose grain size is small, this problem is minimal for the silica minerals. The bigger problem in bulk samples is often reducing the particle size small enough to satisfy particle statistics.

Matrix effects are not a problem in the thin samples used for respirable samples. In bulk samples, the matrix effects, which are primarily due to differential absorption effects, do not affect the diffraction patterns but do affect the intensity response. Intensity effects may be corrected for by absorption measurements or the use of internal or external standards. The matrix has no effect on the detection limits in respirable samples provided there are no peak interferences, but highly absorbing phases can mask the silica in bulk samples. When the particle sizes are larger than respirable, microabsorption also becomes a problem.

Particle size is also a problem when an amorphous layer forms on the surface of the silica particles which is well documented for quartz. As the size of the particles becomes smaller, the volume of this surface layer becomes significant, and the diffraction intensity response departs from linearity with respect to the amount of silica present. The most critical effect is when the particle size distribution of the analytes does not match the material used for calibration.

There are two certified standards produced by OSRD/NIST which are useful for calibration of respirable silica, but the particle size distribution does not always match the desired distribution of the samples under study. No certified standards are available for bulk calibration, and it will probably be difficult to create any because of the variability of samples to be analyzed. Several standard methods have been established by federal agencies for respirable silica. The NIOSH transfer method needs to be further evaluated to simplify the sample preparation, because the present method is not economic for the large number of samples that need analyzing. Individual companies and analytical service laboratories have set up and tested procedures for bulk analyses for specific types of samples. Although none of the methods are generally applicable, all the techniques are based on one of the three general methods of powder diffraction analysis: absorption measurement, internal or external standard.

Modern diffraction instrumentation is adequate for the measurements to be performed. Computerized instruments allow the data to be digitized for easy mathematical processing. Data collection parameters are flexible and usually defined by the amount of silica in the sample. Longer count times are needed for smaller amounts to maintain the statistical significance of the measurements. Step size and scan range are less important except when there are peak interferences which require more of the diffraction pattern to be measured. The instrumental measurements are not the time or cost limiting aspect of the use of diffraction methods; it is the sample preparation time that controls the numbers of samples that can be processed and the economics of making large numbers of measurements.

Several studies are warranted to improve the current status of diffraction analysis for the quantification of crystalline silica. Additional standards need to be certified in sufficient quantities to supply needs for many years. Both quartz and cristobalite are necessary. Some certified bulk samples should also be prepared, and a series of round-robin tests should be performed to test the individual methods in use in different industrial laboratories. Such a set of samples could be used to certify new laboratories. The use of modern mathematical methods of profile fitting and pattern matching for data analysis need to be evaluated for determining the intensities of the diffraction peaks compared to the usual methods of peak integration. The recommendations on defocusing the diffractometer to improve crystallite statistics need to be tested to determine how much improvement can be accomplished by the changes suggested.