Page:NIOSH Manual of Analytical Methods - 1024.pdf/6

 1,3-BUTADIENE: METHOD 1024, Issue 2, dated 15 August 1994 - Page 6 of 8

0.054, 0.091, 0.126, and 0.056; the respective 95% confidence intervals for the mean recoveries were 0.986 ± 0.032, 0.961 ± 0.014, 0.994 ± 0.008, 1.029 ± 0.015, 1.064 ± 0.021, and 1.074 ± 0.021. Prior to linear regression of the recoveries versus the amounts spiked and/or days stored, three results, two high and one low, were determined to be outliers by application of one-sided Grubbs tests [4] at the 2.5% significance level and were dropped from the data set. Linear regression of percent recovery on days stored for the data segregated by level resulted in respective slopes and 95% confidence intervals of 0.060 ± 0.080, 0.005 ± 0.128, –0.003 ± 0.092, 0.060 ± 0.179, 0.249 ± 0.188, and 0.018 ± 0.247 percent per day. Thus, the only statistically significant correlation between recovery and days stored was at the next to highest level, for a gain rather than loss over time. Over all levels, the slopes and 95% confidence intervals for recovery versus amounts spiked and days stored were 0.017 ± 0.009 percent per µg and 0.045 ± 0.051 percent per day, respectively. Thus, according to the latter model: the recovery for the blind spikes increased at a rate corresponding to approximately 11% over the range prepared; as stored, the blind spikes appeared to be stable—the 95% confidence interval of the slope over time indicated a maximum gain of 5.7% or loss of 0.4% during the average 59-day storage period. REFERENCES: [1]	 NIOSH Manual of Analytical Methods, 2nd. ed., V. 2, S91, U.S. Department of Health Education, and Welfare, Publ. (NIOSH) 77-157-B (1977). [2]	 NIOSH Current Intelligence Bulletin 41, “1,3-Butadiene,” U.S. Department of Health and Human Services, Publ. (NIOSH) 84-105 (1984). [3]	 NIOSH Current Intelligence Bulletin 46, “Methylene Chloride,” U.S. Department of Health and Human Services, Publ. (NIOSH) 86-114 (1986). [4]	 Grubbs, F. E. “Procedures for Detecting Outlying Observations in Samples,” Technometrics, 11(1), 1–21, (February, 1969). [5]	 MacCallum, R. N., and J. J. McKetta. “Low-Pressure Zs of C4 Hydrocarbons,” Hydrocarbon Process. Petrol. Refiner, 42(5), 191–194 (1963). METHOD WRITTEN BY: R. Alan Lunsford, Ph.D., Yvonne T. Gagnon, NIOSH/DPSE, and John Palassis, NIOSH/DTMD. APPENDIX A. GAS CHROMATOGRAPH COLUMN SELECTION, INSTALLATION, AND OPERATION: Any column which separates 1,3-butadiene from the other substances present, and which otherwise provides satisfactory chromatographic performance, is acceptable. The column specified in NIOSH Method S91 [1] is 6 m × 3-mm OD stainless steel, packed with 10% FFAP on 80/100 mesh Chromosorb W AW-DMCS. It provides a convenient separation of 1,3-butadiene from the desorbing solvent. However, if other C4 to C6 hydrocarbons are present, interferences are likely. For the development of this method, a 50 m × 0.32-mm ID fused-silica porous-layer open-tubular (PLOT) column coated with Al2O3/ KCl (Cat. # 7515, Chrompack, Bridgewater, NJ) was chosen as the analytical column because it provides a very efficient separation at temperatures above ambient. However, water from the samples deactivates the aluminum oxide, reducing retention times, and high-boiling or polar substances may accumulate on the column and irreversibly degrade the separation. The degradation was eliminated by using a backflushable pre-column, i.e., 10 m × 0.5-mm ID fused-silica CP Wax 57 CB (Cat. # 7648, Chrompack, Bridgewater, NJ). The pre-column allows light hydrocarbons to pass through, but water, methylene chloride, and polar or high boiling components are retained and can be backflushed. Eliminating the solvent peak significantly reduces the time required to complete the analysis. Figures 1 and 2 schematically illustrate the installation and operation of the recommended columns in a Hewlett-Packard 5880A gas chromatograph with split-splitless capillary inlet systems installed in the “B” and “C” injector positions. The only change to the “B” system involves the normally closed (NC) port of NIOSH Manual of Analytical Methods (NMAM), Fourth Edition