Introduction

The general procedure for the collection and analysis of arsenic exposures is given in OSHA Method ID-105 (6.1.). This method was evaluated in 1981-2 and concerns arsenic on mixed-cellulose ester (MCE) filters. Collection or recovery of arsenic species using other types of sampling media can be found in references 6.2. and 6.3.

This back-up data report consists of the following sections:

(1) Determination of the precision and accuracy
(2) Determination of other metals compatible with the arsenic sample preparation procedure
(3) Determination of the compatibility of arsenic with the ICP digestion procedure
(4) Determination of detection limits
(5) Conclusions

1. Precision and Accuracy

Procedure: Quality control samples were independently prepared in the laboratory by using microliter spikes of arsenic-containing solutions on MCE filters. The solutions either were prepared from arsenic trioxide in weak solutions of nitric acid? or sodium hydroxide/sulfuric acid. These samples were analyzed by different chemists within the OSHA laboratory.

Results: Previous and recent quality control samples (6.4.) containing arsenic in the approximate range of 0.5 to 4 times the OSHA PEL (assuming 960-L air volumes), gave the following data:

	Sample Set #1	Sample Set #2
Bias	-0.024	+0.004
CV	0.097	0.10
Overall analytical error	±21.8%	±20.0%
Analysis period	2/1982-4/1982	1/1989-12/1990
N	78	100
Analytical technique	HGA/D2	HGA/LD2 (5%) HGA/ZL (95%)

Where:
HGA/D2	=	Heated Graphite Atomizer with deuterium arc background correction
HGA/LD2	=	Heated Graphite Atomizer/L'vov Platform with deuterium arc background correction.
HGA/ZL	=	Heated Graphite Atomizer with Zeeman/L'vov Platform

Approximately 95% of the samples from set #2 were analyzed using the HGA Zeeman/L'vov platform approach mentioned in this method. The remaining samples were analyzed with a HGA/L'vov platform and deuterium arc background correction only. A significant difference in results was not noted.

2. Other Analytes Compatible with the Arsenic Procedure

The potential for analyzing other metals beside arsenic using OSHA Method ID-105 was verified by the four experiments discussed below. For all elements analyzed below, the analytical parameters listed in Appendix II of the method (6.1.) were used.

2.1. Recovery of Cd, Cu, Fe, Zn using Sample Preparation for Arsenic (ID-105):

Procedure: Using spiked MCE filters, the above metals were analyzed at concentrations of 0.5, 1, and 2 times their respective PELs (the PELs were taken from the 29 CFR 1910.1000, Tables Z-1 and Z-2) based on a 500-L air volume and 25-mL dilution volume. Thirty-six samples of each metal were analyzed, twelve at each PEL level. Of the twelve, six were digested using the procedure listed in OSHA Method no. ID-121 (6.5.), and six using the arsenic procedure (with nickel spikes).

Recovery data from samples prepared using OSHA method no. ID-121 were averaged, and the average values were taken as the theoretical values for each level.

Results: The analytical method recovery (AMR), standard deviation (Std Dev) and coefficient of variation (CV₁) is presented in Table 1 for each set of six samples digested by the arsenic procedure.

2.2. Interference Effects of Cd, Cu, Fe, and Zn in the Analysis of Arsenic:

Procedure: From the above recovery study, 15-mL aliquots of the metal solutions at 2 times the PEL were used. To each aliquot a 100 µL spike of 100 µg/mL arsenic was added. The 10-µg arsenic spike was equivalent to 2 times the arsenic PEL (500-L air sample). Six 15-mL aliquots of diluting solution used in the arsenic procedure were also spiked, to be used as controls. For each set, a 15-mL aliquot of unspiked diluting solution was used as a blank.

All samples were then analyzed by AAS-HGA for arsenic. Two of the samples in each set were analyzed twice.

Results: The analytical method recovery (AMR), coefficient of variation (CV), and standard deviation (Std Dev) for arsenic in the presence of another metal is shown in Table 2.

2.3. Recovery of Metallic Lead - Quality Control (QC) Data:

During the period from February to April 1982, 78 lead and arsenic QC samples were analyzed. These QC samples contained 20 to 40 µg of lead per sample. The mean recovery for the lead analysis was 0.984, the CV was 0.041, and overall analytical error was ±9.8%. [Note: Recent (6.4.) QC samples (from Jan. 1990 to Dec. 1990, n = 24) gave a mean recovery of 1.011, a CV of 0.091, and an overall analytical error of ±19.3%.]

2.4. Recovery of Pb from Lead Oxide (PbO₂), Lead Sulfate (PbSO₄), and Lead Sulfide (PbS):

Procedure: Eight samples of each compound were weighed out, approximately 10 mg each, into Phillips beakers. A 1,000 µg spike of arsenic from a stock solution was added to each sample. The samples were then digested according to the arsenic sample preparation procedure and diluted to 100 mL. The samples were analyzed for lead by atomic absorption using parameters listed in Appendix II of the method (6.1.).

Results: The recovered lead was compared to the theoretical lead content for the particular compound as a test (AMR) of the digestion procedure for these compounds. As shown, recoveries were adequate using this digestion procedure:

Lead
Compound	AMR	Std Dev	CV1
PbO2 PbSO4 PbS	1.072 1.069 1.018	0.021 0.019 0.008	0.019 0.018 0.008

3. Arsenic in Samples Prepared by OSHA Method ID-125 (6.6.)

3.1. The purpose of this experiment was to determine if the arsenic spiked onto MCE filters could be recovered after using the sample preparation suggested in OSHA method ID-125 (6.6.). This is an ICP method which used 4% H₂SO₄ and 4% HCl as the acid matrix for analysis [Note: The current ICP methods use 4%H₂SO₄/8% HCl or 4% HNO₃/32% HCl (6.7., 6.8.)].

3.2. Procedure: Six filters each were spiked at 0.1, 0.5, 1, and 2 times the OSHA PEL for arsenic. These calculations were based on a 480-L air sample. These filters were digested using the procedure mentioned in reference 6.6. and then diluted to a 50-mL solution volume. The solutions were analyzed using the arsenic standards prepared as mentioned in Section 6.5 of the method (6.1.). A 10-µL sample injection was used and an overlay (10-µL injection of 1,000 µg/mL Ni solution) as a matrix modifier was added to each sample or standard. A Model 5100 Zeeman HGA (Perkin-Elmer, Norwalk, CT) and the parameters mentioned in Appendix I of the method (6.1.) were used for the arsenic analysis.

3.3. Results: The recoveries for arsenic are shown in Table 3. Arsenic is not lost when using the OSHA method ID-125 digestion. In addition, arsenic in this matrix (4% H₂SO₄/4% HCl) can be analyzed using arsenic standards prepared in a 4% HNO₃/200 µg/mL Ni matrix and a Perkin-Elmer Model 5100 Zeeman HGA.

---

Note: After this study was performed, the ICP digestion was changed to facilitate solubility of large amounts of Pb. The digestion now uses 4% H₂SO₄/8% HCl (6.7.). Quality control samples analyzed using the new acid matrix did not appear to give significantly different results than those shown in Table 3. In addition, another acid matrix (4% HNO₃/32% HCl) is now used to digest any samples taken from solder operations (6.8.). This matrix has not been tested for suitability with arsenic.

---

4. Detection Limits (6.9.)

This study was undertaken to determine the detection limit for the analysis of arsenic by AAS-HGA using the analytical procedure and equipment mentioned in method ID-105 (6.1.). The detection limit determination was performed in 1990.

4.1. Dilute solutions of arsenic were prepared by serial dilution of stock solutions as mentioned below. Five different concentration levels were prepared and six standard solutions at each concentration were used. The concentrations were selected near the assumed detection limit. Three different analyses of the standards were performed. Two blank solutions prepared on different dates were analyzed three times each for each analysis.

Standard Preparation

Concn (µg/mL As)	Method of Preparation
Blank	Diluting Solution
0.0025	5 mL of 0.05 µg/mL As diluted to 100 mL
0.005	5 mL of 0.1 µg/mL As diluted to 100 mL
0.01	10 mL of 0.1 µg/mL As diluted to 100 mL
0.02	20 mL of 0.1 µg/mL As diluted to 100 mL
0.05	5 mL of 1 µg/mL As diluted to 100 mL

The arsenic source was 1,000 µg/mL stock solution [RICCA Chemical Company, Arlington, TX) lot #E264, expiration date 1/24/91]. The 0.05 to 1 µg/mL standards were prepared using serial dilution of the stock solution. All standards were diluted with diluting solution [Section 6.3.7. of the method (6.1.)].

4.2. Results: Using a least-squares regression program (linear fit and not forcing zero), the slope for the data from each analysis was determined. Because the solution containing the 0.0025 µg/mL As standard was not detectable under the conditions used, its readings were omitted from the calculations. For calculational purposes, all absorbance readings were multiplied by 100. The qualitative (Qual) and quantitative (Quant) detection limits (DL) were calculated using the International Union of Pure and Applied Chemistry (IUPAC) method (6.10.).

Calculated Values

	1	2	3
Slope	945.2991	1536.5812	1553.9398
Blank Std Dev	1.3291601	1.3662601	1.5055453
Qual DL	0.004 µg/mL	0.003 µg/mL	0.003 µg/mL
Quant DL	0.014 µg/mL	0.0089 µg/mL	0.0097 µg/mL

Detection limits are calculated using the following equation:

DL =	(Blank Std Dev)(k) Slope

Where:	k	=	3 for the Qual DL
	k	=	10 for the Quant DL

The average detection limits of the three analyses using the current analytical parameters are:

qualitative:	0.003 µg/mL
quantitative:	0.01 µg/mL

5. Conclusions

The evaluation shows analytical method ID-105 to be precise and accurate, and has sufficient sensitivity for arsenic. The experiments in Section 3. display the capability of using OSHA ID-105 for the digestion and analysis of Cd, Cu, Fe, Pb, and Zn along with arsenic. Arsenic is also compatible and analyzable with other elements when using the ICP digestion procedure found in reference 6.7.

6. References

6.1. Occupational Safety and Health Administration Technical Center: Arsenic in Workplace Atmospheres (USDOL/OSHA-SLTC Method No. ID-105). Salt Lake City, UT. Revised 1991.

6.2. Costello, R.J., P.M. Eller, and R.D. Hull: Measurement of Multiple Inorganic Arsenic Species. Am. Ind. Hyg. Assoc. J. 44(1): 21-28 (1983).

6.3. Occupational Safety and Health Administration Analytical Laboratory: As on FWSB filters by ICP digest by C. Merrell. Salt Lake City, UT. 1989 (unpublished).

6.4. Occupational Safety and Health Administration Technical Center: OSHA Laboratory Quality Control Division Data by B. Babcock. Salt Lake City, UT. 1991 (unpublished).

6.5. Occupational Safety and Health Administration Technical Center: Metal and Metalloid Particulate in Workplace Atmospheres (Atomic Absorption) (USDOL/OSHA-SLTC Method No. ID-121). Salt Lake City, UT. Revised 1991.

6.6. Occupational Safety and Health Administration Analytical Laboratory: OSHA Analytical Methods Manual (USDOL/OSHA-SLTC Method No. ID-125). Cincinnati, OH: American Conference of Governmental Industrial Hygienists (Publication. No. ISBN: 0-936712-66-X). 1985.

6.7. Occupational Safety and Health Administration Technical Center: Metal and Metalloid Particulate in Workplace Atmospheres (ICP) by J. Septon (USDOL/OSHA-SLCAL Method No. ID-125G). Salt Lake City, UT. Revised 1991.

6.8. Occupational Safety and Health Administration Technical Center: ICP analysis of metal/metalloid particulates from solder operations by D.C. Cook (USDOL/OSHA-SLTC Method No. ID-206). Salt Lake City, UT. 1991.

6.9. Occupational Safety and Health Administration Technical Center: As Detection Limit Study by J. Rima. Salt Lake City, UT. 1991 (unpublished).

6.10. Long, G.L. and J.D. Winefordner: Limit of Detection -- A Closer Look at the IUPAC Definition. Anal. Chem. 55: 712A-724A (1983).