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

The modern diffractometer is aligned for optimum resolution. Recommended changes to the diffractometer to increase the particle statistics will improve the accuracy and detection limits for both on-filter methods and bulk samples.

Modifications to the Diffractometer for Quantification

Accurate intensity information on samples being processed for quantification are degraded by a diffractometer aligned for high resolution because too few crystallites in the sample are allowed to diffract. Longer counting times will not improve the statistics unless other changes are made. The situation may be considerably improved by making changes to the diffractometer which are suggested by the analysis of crystallite statistics. First, a broad-focus X-ray tube should be employed, and the focal slit should be enlarged. Second, the solar slit should be replaced by the coarsest slit available or eliminated entirely. Third, a divergence slit should be used which allows the incident beam to cover the entire sample surface at the angle of measurement. Finally, an attachment should be constructed which will simultaneously rock the sample on the diffractometer axis while the sample is spinning. The rocking angle should be plus and minus one to two degrees. Care must be taken that the spinning and rocking motions are not synchronized, so that full coverage of the increased diffraction range is assured. The defocusing due to the rocking and the use of the broad-focus source will be covered by the larger focal slit. Longer counting times will now improve the intensity measurements by allowing time for the angular coverage of the sample by the rocking and spinning motions.

Because the limiting aspect of the diffraction pattern for the quantification of the small amounts required in silica analysis is the peak-to-background ratio, it is necessary to improve this ratio by all means possible. The above recommendations address increasing the peak intensities. Equivalent efforts should be focused toward decreasing the background level of the diffraction pattern. Two procedures should be incorporated in all measurements. Scattering from the atmosphere surrounding the sample should be eliminated by using a helium path for the X-ray beam. All sample holders should be designed to contribute no undesirable diffraction scattering to the experiment. This latter condition may be accomplished by constructing the sample support from a single crystal oriented not to cause any diffraction. Glass or other amorphous materials are unsatisfactory for this purpose. Ironically, single-crystal quartz is an excellent material for this function.

An additional modification to the diffractometer would be to employ a position sensitive detector rather than the conventional detector. Because resolution is not required, the PSD would cover a broader range of angles without requiring the diffractometer to be scanned during the sample rocking and spinning. If the diffractometer is used entirely for quantitative analysis, the suggested changes could be permanent. Only the interchange of the X-ray tubes, which requires realignment, is a time-consuming change or a change which incapacitates the diffractometer from serving as a multipurpose instrument.

Modifications to Sampling

The main goal of preparing a sample for X-ray quantification is to optimize the crystallite statistics. However, in the case of personnel samplers, the sample is limited to the number of particles accumulated during a working shift. Ideally, this value could be zero, then no problems would exist. The small size of the collected samples does provide a challenge to preparing the most effective sample. The goal is now to take full advantage of the sample area intercepted by the incident X-ray beam in the diffractometer with the particles that were collected. In the modern diffractometer with fixed or variable slits, the effective area is around 2.5 cm2. Spinning the sample increases the area to around 3.2 cm 2. The particles should be spread uniformly over this area. A 2 mg sample with an average particle size of 2.5 µm will just fill this area with a single layer of densely packed grains. It is evident that 2 mg of sample is the limit for effective use of thin-layer techniques unless the area is enlarged.

The main improvement to respirable dust measurements would be to increase the usable area in the diffractometer, so the full 37 mm filter could be used. The increase in size of the sample area would cause some defocusing of the diffracted X-ray beam, but the low resolution geometry already recommended allows for this defocusing. It is suggested that the use of the direct on-filter approach could be significantly improved by employing larger filters than the 25 mm size used in the currently recommended techniques providing longer sampling times or higher flow rates would add more particles to fully cover the filter. The improved crystallite statistics would probably provide better comparisons between the on-filter methods and the transfer methods and would certainly improve the comparisons between the X-ray diffraction methods and other methods such as infrared.

Other Recommendations

More calibration standards are needed for both quartz and cristobalite. These standards should be well characterized for particle size and size distributions as well as for effects such as extinction and the presence of the amorphous surface layers. Tridymite is not a significant industrial phase and probably does not need to be quantified. Opal and other amorphous silicas, on the other hand, probably do need to be quantified, and appropriate standards are needed.

The present NIOSH Method 7500 is a good method for quantification and has legal status, but it is time-consuming and not economical for large numbers of analyses. Direct on-filter methods are adequate for personnel monitoring purposes on a routine basis, and a direct method should be certified.

More studies are necessary on the quantification of cristobalite in respirable samples and in bulk samples. If the toxicity is indeed more than for quartz, the frequency of cristobalite in the workplace requires a better knowledge of its abundance in industrial atmospheres and dusts.

Quantification of opal and other amorphous silicas also requires more study. The techniques involving conversion of the amorphous forms to crystalline cristobalite show considerable promise, but interfering reactions and quantification of the conversion need to be evaluated.