Related Experiment Videos
Performance of laboratories measuring silica in the Proficiency Analytical Testing program.
S A Shulman1, J H Groff, M T Abell
1National Institute for Occupational Safety and Health, Division of Physical Sciences and Engineering, Cincinnati, OH 45226.
American Industrial Hygiene Association Journal
|January 1, 1992
Summary
Silica analysis methods show improved precision over time. While colorimetry consistently yields lower results, differences in bias and precision between analytical methods have diminished, with laboratory variation remaining significant.
Area of Science:
- Analytical Chemistry
- Environmental Science
- Materials Science
Background:
- The Proficiency Analytical Testing (PAT) program has monitored silica sample analysis since 1972.
- Participation in the PAT program grew from 5 to 130 laboratories over 101 rounds.
- Laboratories employ colorimetry, x-ray diffractometry, and infrared spectroscopy for silica analysis.
Purpose of the Study:
- To assess inter-method bias in silica analysis.
- To quantify method variability and identify influencing factors.
- To track changes in bias and precision over time.
Main Methods:
- Statistical analysis of silica sample results from 101 PAT program rounds.
- Comparison of colorimetry, x-ray diffractometry, and infrared spectroscopy performance.
- Evaluation of factors influencing analytical bias and precision.
Main Results:
- Colorimetry consistently produced lower silica results, especially at high concentrations.
- X-ray diffractometry showed higher bias than infrared spectroscopy in one period, potentially due to changes in sample preparation (smaller quartz particle size).
- Silica analysis precision improved across all three methods over time, with colorimetry showing the most improvement in precision.
Conclusions:
- Analytical methods for silica have become more precise.
- While inter-laboratory variability remains a significant factor, differences in bias and precision between methods have decreased.
- Changes in sample preparation, such as particle size, can influence analytical bias.