SILICA, CR YST ALLINE , by XRD: M ETH OD 7500, Issue 4, dated 1 5 Ma rch 2003 - Page 6 of 9 NOTE: The re is an alternative to scanning an area air sample, settled dust sample, or ground bulk sam ple to prove lack of contamination. A slow scan of the three main peak s of quartz (also cristobalite and tridym ite if their ab sence has not been pre viously confirm ed) on a personal air sample, with verification that their intensity ratios are within 15% of pure quartz, is sufficient evidence that other materials are not interfering in the silica determination. 12. Perform the following for each sam ple, working standard, and blank filter: a. Mount the reference spe cim en. Determ ine the net intensity, I r, of the reference specimen before and after each filter is scanned. Use a diffraction peak of high intensity that can be rapidly but reproducibly (S r <0.01) measured. b. Mount the sam ple, work ing standard, or blank filter. Measure the diffraction peak area for each silica po lymo rph. Scan times m ust be long, e.g., 15 min (longer scan times will lower the limit of detection). c. Mea sure the background on each side of the peak for one-half the time used for peak scanning. The sum of these two counts is the average background. Determine the position of the background for each sample. d. Ca lculate the net intensity, Ix, (the difference between the peak integrated count and the total back ground c ount). e. Ca lculate and reco rd the norm alized intensity, î x, for each peak:
NOTE: Select a convenient normalization scale factor, N, which is approximately equivalent to the net count for the reference specimen peak, and use this value of N for all analyses. Normalizing to the referenc e sp ecim en inte nsity compensates for long-term drift in X-ray tube inte nsity. If intensity measurem ents are stable, the reference specimen m ay be run less frequently and the net intensities should be normalized to the most recently-measured reference inten sity. f. Determine the norm alized coun t, ÎAg, of an interference-free silver peak on the sample filter following the same procedure. Use a short scan time for the silver peak (e.g., 5% of scan time for analyte peaks) throughout the method. g. Fie ld blanks may be analyzed by scanning the 2-theta range used for the analyte and silver peaks to verify that contam ination of the filters has not occu rred. The analyte peak should be absent. The normalized intens ity of the silver pea k shou ld m atch that of the m edia blank. Each laboratory sh ou ld dete rmine the specifics of field blank use for its application. W hen contamination does occur, the reason should be investigated and appropriate action taken. In practice, contamination of field blank s is extrem ely rare and usually is not consistent across filters. The analysis of blanks may be abbreviated if experience indicates that contamination is not likely with current field and laboratory operations ; how ever, occ asional co nfirm ation o f non -contam ination is pruden t.
CALCULATIONS: 13. Ca lculate the conc entra tion of crystalline silica, C (m g/m 3), in the air volume sa m pled, V (L):
îx b m f(t)
= = = =
normalized intensity for sample peak intercept of calibration graph (Î ox vs. :g) slope of calibration graph, counts/:g !R ln T/(1 ! T R) = absorption correction factor (Table 1)
NIOSH Manual of Analytical Methods (NMAM), Fourth Edition
