Sections 1 - 4 of ANSI/SEI 104-1998 are as follows:

1. Purpose, Practice, Rationale and Scope
2. Determination of Standard Compliance
3. References
4. Definitions

Sections 5 and 6 describe the test method and procedures of validation. Descriptions and related data follow.

Stock standard gas was created by static dilution from 100% analyte, mixed volumetrically with input air pumped at a pre-set flow rate through an inert polypropylene chamber containing Diffusive Samplers under test. Flow was verified by in-line rotameter and analyte concentrations were verified by charcoal tube samples continuously drawn from locations in the chamber bracketing the Samplers under test.

Analyte recovery and de-sorption efficiency determined by analysis (Method AT541) of charcoal wafers "spiked" from standard analyte solutions in carbon disulfide. Samplers were tested at several "spike" levels cor-responding to levels expected for 8-hr Sampler exposures at 0.5-2.0 times the OSHA PEL. Results in Table 6.2.

Table 6.2 %Recovery

(De-Sorption Efficiency)

Table 6.2 (a) De-Sorption Method = Forward

(b) De-Sorption Solvent = 97% Carbon disulfide + 3% Benzyl alcohol

(c) De-Sorption Volume = 2.0 ml

(d) Media = Assay Technology Monitor AT541

Samplers were subject to chamber exposures as described in Section 5. then analyzed by Method AT541. Exposures were applied to Samplers in the range 1-8 hours and 0.1-2.0 times the OSHA PEL. Results in Table 6.3.

Samplers were subject to Exposure Pulse (> OSHA PEL) with a duration less than 50% of the Recommended Sampling Time (RST) followed by a Zero Exposure Period (ZEP) for the duration of the RST. The recovery of analyte from Samplers analyzed immediately following Exposure Pulse was compared with analyte recovery from identically-exposed Samplers analyzed at the end of the RST (i.e. following the Zero Exposure Period). The difference between these two recoveries is taken as the extent of Reverse Diffusion (i.e. evaporative loss as % of Sample) from the Sampler under the experimental conditions chosen.

In practice, Bias Due to Reverse Diffusion will depend on the extent and duration of actual Exposure Pulses in the environment being monitored which cannot be exactly predicted in a lab test. For this evaluation, Bias Due to Reverse Diffusion was estimated as the extent of Reverse Diffusion (evaporative loss as % of Sample) when an Exposure Pulse at 1.0 times the PEL is applied for 12% of the duration of the RST followed by a Zero Exposure Period of 100% of the RST. Results in Table 6.4.

After Zero Exposure Interval

(% Loss = "Reverse Diffusion")

Unexposed Samplers analyzed by Method AT541 to determine background Analyte levels (if any) on the Sampler prior to sampling.

Samplers exposed to atmospheres of benzene, toluene, and xylene for 2-4 hrs at 1-2 times the OSHA PEL (see Section 5) in a Chamber with 3 zones of different cross-sectional areas such that linear velocities of 15, 50, and 150 cm/sec, respectively, were generated. Samplers were placed in each zone with 50% of samplers placed normal to and 50% of Samplers perpendicular to the flow direction. When data were compared from the six locations (representing normal air velocity and orientation variation in workplaces), no significant differences were found among the six groups indicating the absence of an effect of Air Velocity & Orientation on Sampling Rate in the range 15-150 cm/sec. This result is applicable to other organic vapors when the same Sampler is used.

Samplers were exposed to atmospheres of benzene, toluene, and xylene for 2-4 hrs at 1-2 times the OSHA PEL (as per Section 5.) in several Chamber runs in which nearly identical exposures were applied with variations in temperature and humidity as follows: 22^oC/50%RH, 10^oC/50%RH, 30^oC/30%RH, 30^oC/70% RH. Data from the four conditions (representing normal temperature & humidity variation) showed no significant differences among the groups indicating the absence of an effect of Temperature & Humidity on Sampling Rate in the range 10-30^oC and 30-70% RH. This result is applicable to other organic vapors when the same Sampler is used.

Two identical sets of Samples exposed (see Section 5) to Humidity (> 80%RH) overnight prior to exposure to Analyte concentrations at the OSHA PEL for at least 50% of the RST at 20-25^oC. One set analyzed immediately, and the 2^nd set after storage at prescribed intervals. Results in Table 6.8.

(after sampling)

Ethylene Oxide Samplers (Monitor 502) in sealed packaging exposed to >10 ppm ethylene oxide for >2 hours, then analyzed as directed in the Instructions for Use. Results from analysis were not significantly different from results for un-exposed Samplers (blank values) demonstrating the integrity of Sampler packaging. This result with ethylene oxide (which has highest permeability through plastics and pinholes of all analytes tested) is applicable to all Samplers manufactured by Assay Technology and packaged in its standard aluminum foil pouch.

Monitors 541 and 546 incorporate a collection wafer made from coconut charcoal demonstrated to collect upwards of 200 volatile organic compounds. The likelihood of a Sampler's collecting interfering substances is addressed by an analytical method (capillary gas chromatography similar to OSHA 7) which can separate and analyze 100's of VOCs. Method AT541 features co-injection of analytical sample onto dual,high-resolution capillary columns (60 m x 0.32mm) providing identification of each analyte from its characteristic emergence time on two analytical columns and quantitation of analytes. A list of VOCs analyzed by this method are included in Table 6.10. This Table applicable to the VOCs listed when analysis is performed using Method AT 541.

Table 6.10 VOCs Sampled on Monitor 541/546/548

Analyzed by Dual-Column GC

Two groups of Samplers (one group freshly manufactured and one group manufactured 16 months previously) were subject to three exposure tests (see Section 5.) each including benzene, toluene, and xylene for 2-4 hrs at 1-2 times the PEL. The two groups were compared in each of the three exposures. No significant differences found between the two groups indicating the absence of any effect of Sampler stability when stored at room temperature for up to 16 months. This result applicable to other volatile organic analytes sampled on Monitors 541 and 546.

Three groups of Samplers from separate manufacturing Lots subject to three exposure tests (see Section 5.) including benzene, toluene, and xylene for 2-4 hrs at 1-2 times the PEL. When data from the three Lots were compared in each of the three exposures, no significant differences were found among the groups indicating the absence of differences among different Lots of Samplers. This result applicable to other volatile organic analytes sampled on Monitors 541 and 546.

Samplers 541 and 546 have been evaluated for sampling benzene, toluene ,and xylenes together on a single sampler. The overall accuracies expressed as Maximum Total Error (95% confidence) are as follows. Benzene = +13%; Toluene = +17%; Xylene = +14%

Based on estimated Sampler-to-Sampler variation of +5%, Laboratory variation of +3%, and Exposure Chamber Variation of Error +8%, less than 5% or the Maximum Total Error is attributed to Bias (i.e. systematic error). We have estimated the Bias Due to Reverse Diffusion as +0%.

It is recommended that Samplers 541 and 546 be used within the envelope of conditions specified, but, in general, minor excursions outside these limits would be expected to have only minor effects. Due to the detection limit of Benzene, lower levels or shorter sampling times could not be accommodated. Due to lack of detectable Reverse Diffusion for 8 hr zero exposure interval, increases in Concentration, Sampling Time, or Humidity above the limits described here could probably to tolerated with minimal.

Prepared by:  CR Manning, PhD, May 2000