2.1 Equipment

2.1.1 Calibrated personal sampling pumps capable of sampling within ±5% of the recommended flow rate of 0.5 L/min are used.

2.1.2 Tygon or other flexible tubing for connecting pumps to samples.

2.1.3 Sampling media:

Impregnated glass fiber filters (IGFFs) are used for sample collection and are obtained from OSHA-SLTC for OSHA use. The filters can also be prepared as explained below:

3.1 Safety Precautions

3.1.1 Review appropriate IC instrument manuals, UV-VIS detector or spectral array detector maintenance manual, and the Standard Operating Procedure (SOP) for proper instrument operation (Ref. 5.14).

3.1.2 Observe laboratory safety regulations and practices.

3.1.3. Review any MSDSs provided with reagents and samples. Observe all precautions. Many chemicals are hazardous. Use appropriate personal protective equipment such as safety glasses, goggles, face shields, gloves, and lab coat when handling these chemicals.

3.2 Equipment

3.2.1 Ion chromatograph (Model 4000i or 4500i Dionex, Sunnyvale, CA) equipped with a UV-VIS detector (Linear UVIS-206, Multiple wavelength detector, Linear Instruments Corporation, Reno, NV) or a conductivity detector.

3.2.2 Automatic sampler (Dionex Model AS-1) and 0.5-mL sample vials/caps (Dionex part no. 38011).

3.2.3 Laboratory automation system: Ion chromatograph interfaced with a data reduction system (AI450, Dionex).

3.2.4 Separator and guard columns, anion (Model HPIC-AS9 and AG9, Dionex).

3.2.5 Forceps.

3.2.6 Disposable beakers (10 and 50 mL).

3.2.7 Cassette opener (SKC E-Z Opener, Cat. No. 225-13-5, SKC) or similar tool such as a coin or a screwdriver.

3.2.8 Disposable syringes (1 mL).

3.2.9 Syringe prefilters, 0.5-µm pore size (part no. SLSR 025 NS, Millipore Corp.,Bedford, MA).

3.2.10 Miscellaneous volumetric glassware: Pipettes, volumetric flasks, Erlenmeyer flasks, graduated cylinders, and beakers.

3.2.11 Equipment for eIuent degassing (vacuum pump, ultrasonic bath).

3.2.12 Analytical balance (0.01 mg).

3.2.13 Scintillation vials, 20 mL, with polypropylene- or Teflon-lined caps.

3.2.14 Treated glass fiber filters (IGFFs from Section 2.1.3) for spiking or matrix matching (if necessary).

3.3 Reagents - All chemicals should be at least reagent grade.

3.3.1 Principal reagents:

Sodium carbonate (Na₂CO₃), 99%
Sodium bicarbonate (NaHCO₃), 99%
Sodium nitrate (NaNO₃), 99.9%
Deionized water (DI H₂O)

3.3.2 Eluent (1.0 mM Na₂CO₃ + 1.0 mM NaHCO₃):

Dissolve 0.424 g Na₂CO₃ and 0.336 g NaHCO₃ in 4.0 L DI H₂O. Sonicate this solution and degas under vacuum for 15 min.

Nitrate (NO₃¯) stock standard (1,000 µg/mL):

Dissolve and dilute 1.3710 g of NaN0₃ to 1.0 L with DI H₂0. Prepare every 6 months.

3.3.4 Nitrate (N0₃¯) standard solutions, 100, 10, and 1 µg/mL: Pipette appropriate volumes of the 1,000 µg/mL as NO₃¯ stock standard into volumetric flasks and dilute to the mark with eluent. Prepare monthly.

3.4 Working Standard Preparation - Prepare fresh prior to beginning the analysis.

3.4.1 Prepare NO₃¯ working standards in eluent. A suggested scheme for preparing a series of working standards using 10-mL final solution volumes is shown below:

working std (µg/mL)	std solution (µg/mL)	aliquot (mL)	eluent added (mL)
0.5   1.0 * 2.0 5.0 10.0	1.0   1.0 10.0 10.0 10.0	5.0 - 2.0 5.0 -	5.0 - 8.0 5.0 -
* Already prepared in Section 3.3.4

3.4.2 To prepare each working standard (Working Std) listed above, transfer an appropriate amount of the Std Solution to a disposable beaker, pipette an appropriate aliquot (Aliquot) of the specified standard solution (prepared in Section 3.3.4) from the disposable beaker to an appropriate container (scintillation vial, Erlenmeyer flask, etc.). Add the specified amount of eluent (Eluent Added).

3.4.3 As an alternative, pipette each aliquot into a 10-mL volumetric flask and dilute to volume with eluent.

3.5 Sample Preparation

3.5.1 Carefully open each cassette with a cassette opener (or similar tool, such as a coin or a screwdriver), remove each IGFF and transfer each filter using a clean forceps into separate 20-mL scintillation vials.

3.5.2 Pipette 5.0 mL of DI H₂O into each vial. Make sure the filter is wetted. Cap the vials using polyethylene-lined plastic caps.

3.5.3 Allow the samples to sit for at least 15 min. Occasionally swirl each solution.

3.5.4 If the sample solutions contain particulate, remove the particles using a prefilter and syringe.

3.6 Analysis

It is imperative that the large nitrite peak (from the sampling media) is adequately separated from the nitrate peak. This can be assured by desorbing an IGFF (Section 3.2.14) with eluent, spiking the solution with a known amount of nitrate working standard, and analyzing this solution prior to analysis. The chromatogram shown below (Comparison of a Standard and a Sample) demonstrates the peaks obtainable from a sample and a standard without any matrix-matching. Peak characteristics of the nitrate in the standard and sample are similar, retention times appear very close, and there is adequate separation of nitrite and nitrate. If a comparison of a spiked sample and a nitrate standard indicates poor separation or significantly different NO₃¯ retention times, matrix-matching or a change in analytical conditions should occur. A new column could be used or the eluent strength may be changed to facilitate separation. If matrix-matching of standards and samples is the only alternative, standards should be prepared with treated filters in the same fashion as samples.

Comparison of a Standard and a Sample

Text Version: This figure shows a sample chromatogram superimposed over a standard chromatogram. The sample chromatogram shows the chromatographic separation of the nitrite peak from the nitrate peak in a sample.  The retention times for the nitrate peak in the sample and the nitrate peak in the standard are almost identical and show that the separation of nitrite and nitrate is adequate.

3.6.1 Pipette or pour a 0.5- to 0.6-mL portion of each standard or sample into separate automatic sampler vials. Place a filtercap into each vial. The large filter portion of the cap should face the solution.

3.6.2 Load the automatic sampler with labeled samples, standards, and blanks.

3.6.3 Set up the ion chromatograph in accordance with the SOP (Ref. 5.14). Typical operating conditions for a Dionex 4000i or 4500i with a UV-VIS detector (Spectral Array detector) and an automated sampler are listed below:

Ion chromatograph with UV detector * at 200 nm wavelength

Eluent:	1.0 mM NaCO3/1.0 mM NaHCO3
Column temperature:	ambient
Anion precolumn:	AG9
Anion separator column:	AS9
Output range:	2 absorbance units full scale (AUFS)
Rise time:	5 sec
Sample injection loop:	50 uL
Pump
Pump pressure:	~900 psi
Flow rate:	2 mL/min
Chromatogram
Run time:	5 min
Peak retention time:	~3.00 min for NO3¯

* For detection using a conductivity detector, output range and rise time are not used. A sensitivity setting on the conductivity detector of 0.1 µS is used. All other settings are similar.

Soluble nitrate compounds can interfere when using either UV or conductivity detector. Response to nitrate using either detector is similar and appears to be dependent on column conditions, eluent strength, and sensitivity settings.

3.6.4 Analyze samples, standards, and blanks according to SOP (Ref. 5.14).

3.7 Calculations

3.7.1 After the analysis is completed, retrieve the peak areas or heights. Obtain hard copies of chromatograms from a printer. A chromatogram of a sample collected at an ozone concentration of approximately 2 times the PEL for 180 min is shown below:

ret time	component name	concentration	height	area
3.22	nitrate	6.142 µg/mL	218782	2202892

Text Version: This figure shows a sample chromatogram superimposed over a blank sample chromatogram. The sample chromatogram represents nitrate from an ozone concentration of approximately two times the PEL for a 180 min sample.  The nitrate normally occurring in a blank sample is shown for illustrative purposes.  Detector response and column retention times were obtained using equipment and analytical conditions specified in this method.

* Relative absorbance units using a UV-VIS detector

3.7.2 Prepare a concentration-response curve by plotting the peak areas or peak heights versus the concentration of the NO₃¯ standards in µg/mL.

3.7.3 Determine total µg for each sample and blank. Perform a blank correction for each IGFF. Subtract the total µg blank value from each total µg sample value.
 

Ab = (µg/mL NO3¯)b × (Sol Vol)b × (CF) As = (µg/mL NO3¯)s × (Sol Vol)s × (CF) A = As - Ab

Then calculate the air concentration of O₃ (in ppm) for each air sample:
 

ppm O3 =

A × (Mol Vol)   AV × (Mol Wt)

where:
Ab	=	Total µg O3 in blank
As	=	Total µg O3 in sample
A	=	µg O3 after blank correction
(µg/mL NO3¯)b	=	Amount found (from calibration curve) in blank
(µg/mL NO3¯)s	=	Amount found (from calibration curve) in sample
(Sol Vol)b	=	Blank solution volume (mL) from Section 3.5.2 (normally 5 mL)
(SoI Vol)s	=	Sample solution volume (mL) from Section 3.5.2 (normally 5 mL)
CF	=	Conversion factor = O3/NO3¯ = 0.7742
Mol Vol	=	Molar volume (L/mol) = 24.45 (25°C and 760 mmHg)
AV	=	Air volume (L)
Mol Wt	=	Molecular weight for O3 = 47.997 (g/mol)

3.8 Add the results of the first and second filters to give one final O₃ concentration. If a significant amount of analyte (>25 % of first filter) is found on the back-up (second) filter, breakthrough may have occurred. Report possible breakthrough as a note on the report form.

3.9 Report results to the industrial hygienist as ppm O₃.

1. An analysis of 20 spiked samples (7 samples each at 1 and 2 times, and 6 samples at 0.5 times the TWA PEL) to evaluate analytical recovery as desorption efficiency (DE).
    
2. A sampling and analysis of 22 samples (7 samples each at 1 and 2 times, and 8 samples at 0.5 times the TWA PEL) collected from dynamically generated test atmospheres at 50% RH to determine bias and overall error. Samples at a concentration near the STEL (0.3 ppm) were also taken.
    
3. A determination of the sampling medium collection efficiency at approximately 2 times the TWA PEL.
    
4. A determination of potential breakthrough.
    
5. An evaluation of storage stability at room (20-25°C temperatures for 26 collected samples.
    
6. A determination of any significant humidity effects during sampling.
    
7. A determination of the qualitative and quantitative detection limits.
    
8. Comparison of sampling methods - impinger vs. treated filter vs. passive monitor (AKI vs. IGFF vs. OPS).
    
9. Interface study.
    
10. Shelf-life of the IGFFs.
     
11. Summary.

OEi% = ±(|biasi| + 2CVi) X 100% (at the 95% confidence level) Where i is the respective sample pool being examined. Block Diagram of the Laboratory Generation System

Figure 2

Text Version: Lab air passes through an Air Purifier and then enters the Flow-Temp-Humidity Control System. Lab Water passes through an Ionic Exchange Column, where it is purified, and then enters the Flow-Temp-Humidity Control System. The Flow-Temp-Humidity Control System generates properly conditioned dilution air that enters the Mixing Chamber. Ozone is produced by the Ozone Generator System and it enters the Mixing Chamber to be completely mixed with dilution air to produce the ozone test atmosphere. The ozone test atmosphere enters the Active Sampling Manifold. The Active Sampling Manifold provides a means to sample the ozone atmosphere such as described in OSHA Method ID 214. The ozone test atmosphere next enters an Ozone Passive Exposure Chamber where diffusive samplers can be exposed to the ozone test atmosphere. The ozone test atmosphere next passes through a Dry Test Meter where the total air/ozone volume is measured. Finally, the ozone test atmosphere enters the laboratory Exhaust system and goes to waste.

4.1 Analytical Recovery

Ozone oxidizes sodium nitrite to sodium nitrate on the filter. To test the relative analytical capability of this method, sodium nitrate was used as the analytical spike. Twenty samples were prepared by adding known amounts of NO₃ (as NaNO₃) stock solution to the IGFFs to determine desorption efficiencies (DEs) for the analytical portion of the method.

4.1.1 Procedure: Each IGFF was spiked using a 25- or 50-µL syringe (Hamilton Microliter/Gastight Syringe, Hamilton Co., Reno, NV). The IGFF samples were inside cassettes when spiked with aqueous solutions. Spikes were either 11.5, 23.0, and 46.0 µg NO₃¯. These levels correspond to approximately 0.5, 1, and 2 times the TWA PEL for a 90-L air sample at a 0.5-L/min flow rate. The cassettes were allowed to sit overnight and then analyzed.

4.1.2 Results: Desorption efficiencies are presented in Table 1. As shown, the average DE is very close to 1.0. No DE corrections are necessary for O₃ collection using IGFFs.
 

Table 1
Ozone (as NO₃¯) Analysis - Desorption Efficiency (DE)

level (¯)	N	mean DE	SD	CV1
0.5 × PEL 1 × PEL 2 × PEL all levels	6 7 7 20	0.979 1.025 0.977 0.994	0.055 0.029 0.47 -	0.056 0.028 0.048 0.045 *
*CV1 (pooled)

4.2 Sampling and Analysis

To determine the precision and accuracy of the method, known concentrations of O₃ were generated, samples were collected, and analyzed.

4.2.1 Procedure:

1. Test atmospheres of O₃ were generated using two ozone generators (Model 565, ThermoElectron Instruments, Hopkinton, MA) simultaneously to achieve as high O₃ concentrations as possible. The O₃ gas was diluted with filtered, humidified air using the system shown in Figure 2 and discussed below. A glass mixing chamber was used to facilitate blending of ozone with the diluent air.
    
2. Dynamic generation system
   
   A Miller-Nelson Research Inc. flow, temperature, and humidity control system (Model HCS-301, Monterey, CA) was used to control and condition the dilution airstream. All generation system fittings and connections were Teflon. The O₃ concentrations were varied by adjusting the dilution airstream Volume. The dilution airstream was adjusted using the mass flow controller of the Miller-Nelson system. For this experiment, the system was set to generate test atmospheres at 50% RH and 25°C. Test atmosphere concentrations were approximately 0.5, 1, and 2 times the OSHA TWA PEL and at the OSHA STEL.
    
3. The total flow rate of the generation system was measured using a dry test meter.
    
4. IGFF/cassette samples were attached to the Teflon sampling manifold using Gilian Gil-Air SC pumps (Gilian Instrument Corp., W. Caldwell, NJ) to draw the O₃ test atmosphere through the IGFF samples. Pump flow rates were approximately 0.5 and 1.5 L/min and sampling times were 180 and 15 min for TWA and STEL experiments, respectively.

4.2.2 Results: The results are shown in Tables 2a and 2b. The spiked sample (Table 1) and test atmosphere sample (Table 2a) results each passed the Bartlett's test and were pooled to determine a CVT for the TWA sampling and analytical method.
 

Table 2a
Ozone sampling and Analysis - TWA PEL Determinations

level (¯)	N	ave recovery	SD	CV2	OE2 (±%)
0.5 1 2 all levels	8 7 7 22	1.032 1.071 0.937 1.014	0.060 0.023 0.028 -	0.059 0.022 0.030   0.041*	14.9 11.5 12.3      9.7 **
* CV2 (pooled) - ** OE2 (pooled)

The total pooled coefficients of variation (CV_T), bias, and total overall error (OE_T) are as follows:

CVT (pooled) = 0.045

bias = + 0.014

OET = 10.4%

(Note: The CV_T and OE_T values include data from Section 4.1 and are calculated using equations specified in Refs. 5.16-5.17.)
 

Table 2b
Ozone Sampling and Analysis - STEL PEL Determination
(Known O₃ Concentration = 0.33 ppm)

level(¯)	N	mean ppm found	SD	CV	recovery	OE
STEL	5	0.325	0.018	0.054	98.5%	±12.3%

4.3 Collection Efficiency

Procedure: Seven IGFF/cassettes were used to collect a concentration of approximately 2 times the OSHA TWA PEL for 180 min at 0.5 L/min (50% RH and 25°C. The amounts of O₃ gas collected on the first and second IGFFs were determined. The collection efficiency (CE) was calculated by dividing the amount of O₃ collected in the first filter by the total amount of O₃ collected in the first and second IGFFs.

Results: The results in Table 3 show a CE of 100%. No O₃ was found in the second IGFF for the CE experiment and indicates the IGFFs have adequate collection of O₃ near the PEL.
 

Table 3
Collection Efficiency (CE)
2 × PEL - 25°C - 50% RH

	ppm O3
sample no.	1st IGFF	2nd IGFF	CE, %
1 2 3 4 5 6 7	0.209 0.220 0.203 0.216 0.211 0.204 0.206	ND ND ND ND ND ND ND	100.0 100.0 100.0 100.0 100.0 100.0 100.0

Notes:	(a)	Sampled at 0.5 L/min for 180 min.
	(b)	Samples desorbed using a sample solution volume of 5.0 mL
	(c)	ND = None detectable (< 0.008 ppm O3)

4.4 Breakthrough

(Note: Breakthrough is defined as > 5 % loss of analyte from the first IGFF to a backup IGFF at 50% RH)

Procedure: The same procedure as the CE experiment (Section 4.3) was used with two exceptions: In addition to the 2 × concentration, the generation concentration was increased to a level approximately 4 times the TWA PEL, and samples were taken at approximately 0.5 L/min for 240 min. Another test was conducted for 6 times the TWA PEL using a sampling rate of approximately 0.25 L/min for 240 min. Due to limitations on the O₃ generators and the generation system, larger O₃ concentrations could not be achieved.

The amount of breakthrough for each sampling cassette was calculated by dividing the amount collected in the second IGFF by the total amount of O₃ collected in the first and second IGFFs.

Results: For measurements near the TWA PEL, no breakthrough of O₃ into the second section was found at an approximate concentration of 0.2 ppm O₃ (Table 4a), and indicates the first IGFF has adequate retention of O₃ at 2 times TWA PEL. However, the average breakthrough was 7.5% at an approximate concentration of 0.4 ppm O₃ (Table 4b) for 240 min at 0.5 L/min flow rate. No break-through was found at the approximate concentration of 0.6 ppm O₃ (Table 4c) when using a lower flow rate of 0.25 L/min. For the STEL, no breakthrough was found at approximate concentration of 0.3 ppm O₃ (Table 4d) for 15 min at 1.5 L/min sample collection flow rate.

Table 4a
Breakthrough Study - 0.5 L/min
2 × PEL - 25°C - 50% RH

	ppm O3
sample no.	1st IGFF	2nd IGFF	Breakthrough, %
1 2 3 4 5 6	0.242 0.281 0.190 0.227 0.238 0.215	ND ND ND ND ND ND	0 0 0 0 0 0

Notes:	(a)	Sampled at - 0.5 L/min for 240 min
	(b)	Due to the larger sampling period and thus larger mass collected, the first IGFF was desorbed using larger sample solution volumes of 10.0 mL.
	(c)	ND = None detectable (< 0.008 ppm O3)

Table 4b
Breakthrough Study - 0.5 L/min
4 × PEL - 25°C - 50% RH

	ppm O3
sample no.	1st IGFF	2nd IGFF	CE, %
1 2 3 4 5 6	0.425 0.385 0.395 0.363 0.383 0.342	ND ND ND ND ND ND	0 0 0 0 0 0

Notes:	(a)	Sampled at - 0.5 L/min for 240 min
	(b)	Due to the larger sampling period and thus larger mass collected, the first IGFF was desorbed using larger sample solution volumes of 15.0 mL.
	(c)	Statistical analysis - N = 8; mean = 7.5; SD = 1.5; CV = 0.20

Table 4c
Breakthrough Study - 0.25 L/min
4 × PEL - 25°C - 50% RH

	ppm O3
sample no.	1st IGFF	2nd IGFF	Breakthrough, %
1 2 3 4 5 6	0.563 0.600 0.586 0.661 0.566 0.558	ND ND ND ND ND ND	0 0 0 0 0 0

Notes:	(a)	Sampled at - 0.25 L/min for 240 min
	(b)	Due to the larger sampling period and thus larger mass collected, the first IGFF was desorbed using larger sample solution volumes of 10.0 mL.
	(c)	ND = None detectable (<0.008 ppm O3)

Table 4d
Breakthrough Study - 1.5 L/min
1 × PEL - 25°C - 50% RH

	ppm O3
sample no.	1st IGFF	2nd IGFF	Breakthrough, %
1 2 3 4 5 6	0.440 0.308 0.333 0.346 0.306 0.334	ND ND ND ND ND ND	0 0 0 0 0 0

Notes:	(a)	Sampled at - 1.5 L/min for 15 min
	(b)	Samples desorbed using a sample solution volume of 5.0 mL
	(c)	ND = None detectable (< 0.032 ppm O3)

4.5 Storage Stability

Procedure: A study was conducted to assess the stability of the NO₂¯ + O₃ reaction product, NO₃¯ on the IGFFs. A room temperature storage stability study using 26 samples taken near the OSHA TWA PEL of 0.1 ppm was performed. All samples were stored under normal laboratory conditions (20-25°C) in a plastic bag in a drawer. Seven samples were initially desorbed and analyzed; seven more samples were desorbed and analyzed after 5 days, followed by six samples at 15, and 30 days, respectively.

Results: The mean of samples analyzed after 30 days was within 10% of the mean of samples analyzed the first day, as shown in Table 5 and Figure 3 below.

Table 5
Storage Stability - Ozone
(25°C, and 50% RH)
(Known O₃ Concentration = 0.123 ppm)

Day	N	Mean O3Found	SD	CV	Recovery (%)
1 5 15 30	7 7 6 6	0.122 0.120 0.135 0.116	0.005 0.004 0.002 0.006	0.038 0.036 0.015 0.052	99.2 97.6 109.8 94.3

Storage Stability Figure 3

Text Version: Figure 3 is a graph of the storage stability data shown in Table 5. The graph shows percent ozone recovery-plotted on the y-axis against days of storage time-plotted on the x-axis.  The storage stability is excellent, with approximately 94% of the ozone recovered after 30 days of ambient storage.

4.6 Humidity Study

Procedure: A study was conducted to determine any effect on recovery results when samples are collected at different humidities. Samples were taken using the generation system and procedure described in Section 4.2. Test atmospheres were generated at 25°C and at approximately 0.5, 1, and 2 times the OSHA TWA PEL. Relative humidities of 30%, 50%, and 80% were used at each concentration level tested.

Results: Results of the humidity tests are listed in Table 6. An F test was used to determine if any significant effect occurred when sampling at different RHs. As shown, at the 99% confidence level, the calculated F values are much smaller than critical F values (Ref. 5.16) for all the concentrations tested; therefore, no significant difference in results occurred across the RH ranges tested.

Table 6
Humidity Test - Ozone
25°C

Level	RH, %	N	Mean O3 Found	SD	CV	Taken	Recovery, %	Fcrit	Fcalc
0.5 × PEL	30 50 80	7 8 7	0.073 0.072 0.060	0.008 0.004 0.001	0.107 0.059 0.024	0.070 0.070 0.058	104 103 103	5.93	0.02
1 × PEL	30 50 80	6 7 7	0.119 0.118 0.101	0.007 0.003 0.002	0.059 0.022 0.022	0.115 0.110 0.098	103 107 103	6.11	2.62
2 × PEL	30 50 80	7 7 7	0.174 0.222 0.231	0.005 0.006 0.006	0.030 0.028 0.027	0.172 0.224 0.237	101 99.1 97.5	6.01	2.71

4.7 Qualitative and Quantitative Detection Limit Study

A modification of the National Institute for Occupational Safety and Health (NIOSH) detection limit calculations (Refs. 5.18-5.19) was used to calculate detection limits.

Procedure: Low concentration samples were prepared by spiking aqueous standards prepared from NaNO₃ (Section 3.3.4) at five different concentrations on the IGFFs. Samples were analyzed using a 50-µL sample injection loop and a UV-VIS detector setting of 2 AUFS.

Results: The IGFF spiked sample results are shown in Table 7 for qualitative and quantitative detection limits, respectively. The qualitative detection limit is 0.37 µg/mL as NO₃¯ at the 99.8% confidence level. The quantitative detection limit is 1.25 µg/mL as NO₃¯. Using a 90-L air volume and a 5-mL sample solution volume, the qualitative and quantitative detection limits are 0.008 ppm and 0.03 ppm, respectively, as O₃.

Table 7
Qualitative and Quantitative Detection Limits (NIOSH Method)

	O3 (as NO3¯) Level
Sample No.	Blank PA	0.1 µg/mL PA	0.2 µg/mL PA	0.5 µg/mL PA	1.0 µg/mL PA
1 2 3 4 5 6	2.05 1.98 2.02 2.03 2.02 1.74*	2.73 2.60 1.81* 2.60 2.68 2.69	2.25 3.15 3.15 3.23 4.55* 3.79	3.17 4.15 3.21 4.09 4.12 3.24	5.87 4.99 4.98 5.76 5.81 5.81
* Outlier PA - Integrated Peak Area (NO3¯)/100,000

The average responses of the low-level calibration samples were plotted to obtain the linear regression equation (Y = mX + b), and the predicted responses (_i) at each X.

Using the equations:

Q1 =   3Sy m

Q2 = (3.33)Q1

Text Version: First subtract each obtained response from its predicted response, and then square that difference.  Sum all these values.  Divide the sum by the number of data points minus two.  Take the square root of that value.  The result is the standard error of the regression line.

Therefore,	Q1	=	(3Sy)/m = 0.37 µg/mL as NO3¯
			1.85 µg as NO3¯ (5-mL sample volume)
			0.008 ppm O3 (90-L air volume)
	Q2	=	(3.33)Q1
			0.027 ppm O3 (90-L air volume)

where:
B	=	mean blank response
b	=	intercept of the regression
m	=	analytical sensitivity or slope as calculated by linear regression
SY	=	standard error of the regression = 0.21667
N	=	number of data points
Q1	=	qualitative detection limit
Q2	=	quantitative detection limit

The Correlation Coefficient (r) and Coefficient of Determination (r²) for the above data were r = 0.986 and r² = 0. 972.

4.8 Comparison of Sampling Methods

This method was compared with the classical AKI approach and a passive monitor method. The Ogawa passive sampler (OPS), developed by the Harvard School of Public Health (HSPH), was originally designed to sample for nitrogen oxides in the environment. Modifications allowed its use to monitor ambient environmental ozone. The reaction principle and analysis are similar to this lGFF method; however, the impregnating solution is slightly different (and proprietary for the passive system), and the samples are analyzed by IC for nitrate ion using conductivity detection instead of UV-VIS Prior to using the OPS method for this comparison, the sampling rate was examined. Due to face velocity dependence, sampling rates are critical to the performance of the passive monitor during this comparison. The determination of sampling rate is detailed in the Appendix.

Procedure: In order to compare performance, the IGFF/cassettes (this study), AKI samples, and OPSs were collected side by side from the generation system at approximately 0.5, 1 and 2 times the PEL. The IGFF/cassettes and OPSs were analyzed by IC. The AKI samples were analyzed by a colorimetric procedure further described in Ref. 5.2. The average sampling rate as determined by SLTC for the face velocity achieved, 21.93 cm³/min, was used for OPSs.

Results: Table 8 shows the results of the comparison study. As shown, the IGFF/cassettes, the AKI samples, and OPSs are in good agreement except that OPSs are slightly higher for 1 times and lower for 2 times PEL.

Table 8
Comparison of Methods (Summary)
(25°C and 50% RH)

Set #	Method	O3 Found, ppm	N	SD	CV
1	AKI IGFF OPS	0.070 0.072 0.073	3 8 4	0.006 0.004 0.003	0.086 0.056 0.041
2	AKI IGFF OPS	0.110 0.118 0.129	3 7 7	0.002 0.003 0.017	0.018 0.025 0.132
3	AKI IGFF OPS	0.224 0.210 0.187	3 7 7	0.008 0.006 0.012	0.036 0.029 0.064

where:	AKI Alkaline potassium iodide (Ref. 5.2)
	IGFF Impregnated glass fiber filter (this study)
	OPS Ogawa passive sample for ozone (Refs. 5.10 and 5.20)

4.9 Interference Study

As previously discussed in Section 1., oxidizing gases have interfered with the determination of O₃ in previous methods (Refs. 5.1-5.2, 5.4). Several tests were conducted to evaluate any possible interference from NO₂ or SO₂.

Procedure: Possible interferences from NO₂ and SO₂ were tested using several sets of IGFF/cassette samples. A test was conducted by taking four samples at approximately 6 ppm NO₂ and compared to four samples without NO₂ which served as "control" samples. Several tests were conducted to evaluate any SO₂ interference by comparing results of six samples with SO₂ to another four to six samples without SO₂, present. These tests included two different SO₂ concentrations and use of oxidizer tubes for removal of SO₂ from the sampled air prior to O₃ reaction with the treated filters.

Two different kinds of oxidizer tubes were evaluated. Both were manufactured by SKC Inc. (Eighty Four, PA) and are used to convert nitric oxide (NO) to nitrogen dioxide (NO₂) during sampling for NO. The two types of oxidizer tubes are:

Tube Label	Substrate	Abbrev.
Oxidizer Special	Chromate impregnated sand	OS
Misc-Spec	Chromate impregnated material (Composition of substrate unknown)	MS

Both tube labels are designations given by SKC. Chromate impregnated sorbent has been shown to effectively remove SO₂ during ozone sampling (Ref. 5.21). All samples were taken at a flow rate of about 0.5 L/min for 180 min. The generation system concentration was approximately 1.5 times the TWA PEL for ozone.

Results: Table 9 shows the results of the IGFF/cassette sample sets:

Sample Set No.	Description
1)	O3 with and without NO2
2)	O3 with and without SO2 (3.41 ppm)
3)	O3 with and without SO2 (1.06 ppm)
4)	O3 with and without SO2 (0.35 ppm)
5)	O3 + SO2 with and without OS oxidizer
6)	O3 + SO2 with OS oxidizer before and after conditioning
7)	O3 + SO2 with and without MS oxidizer
8)	O3 + SO2 with MS oxidizer before and after conditioning
9)	Comparison study between 50% and 80% RH for O3 +SO2 with MS oxidizer after conditioning.

Notes:	(a)	NA = Not applicable
	(b)	ND = None detectable (< 0.008 ppm 03)
	(c)	Flow Rate = 0.5 L/min
	(d)	Sample Solution Volume for Desorption = 5.0 mL
	(e)	All oxidizers were conditioned for 4 h at a concentration of approximately 0.1 ppm O3

As shown in Sample Set #1, 6.38 ppm NO₂ caused no interference when sampling at 1.5 times TWA PEL ozone. When SO₂ is present along with ozone, a negative interference equal to 100% of an equimolar concentration of ozone is noted as shown in Sample Sets #2, #3 and #4. Sample Sets #5 and #7 show no interference occurs when using the oxidizer tubes. Sample Sets #6 and #8 show the difference in recovery when using conditioned and unconditioned oxidizer tubes. As shown, the oxidizer gave results about 23% lower when it was not conditioned (0.108 vs. 0.141 ppm O₃ when conditioned). Although. the recoveries improved for the MS oxidizer without conditioning (0.141 vs. 0.153 ppm when conditioned), they were still low and it is recommended to passivate either type of oxidizer tube. Sample Set #9 shows no significant difference in O₃ recovery when SO₂ is present at 50% and 80% RH.

An additional test was conducted to determine if the passive monitor would be adversely affected by SO₂ in a similar fashion as the active sampler. Side-by-side active and passive samples were taken while varying the amount of SO₂. Both passive and active samples were prepared using the procedure stated in this method for IGFFs. (Section 2.1) Additional passive samplers were also purchased from Ogawa; the procedure, type, and amount of chemicals used in their treatment preparation is unknown.

As shown in Table 10, the passive monitor, regardless of treatment in-house or from Ogawa, appears to display the same SO₂ interference as the active sampler. Detection limits are similar to what is stated earlier for both active and passive samplers.

Table 10
Active vs. Passive Sampler - SO₂ Interference

Sample Set #	Active or Passive	Interferant, SO2 Concn (ppm)	N #	Mean O3, ppm	SD O3, ppm	CV %
1 1 1 1 2 2 2 2	Active Active Passive Passive Active Active Passive Passive	0 1.89 0 1.89 0 1.89 0 1.89	4 3 6 6 2 3 6 6	0.164 ND 0.167 ND 0.132 ND 0.130 ND	0.006 - 0.017 - 0.007 - 0.012 -	3.5 - 10.1 - 0.5 - 9.0 -

Note:	N = number of samples taken.
	Sample Set #1 represents passive samplers prepared using 13-mm glass fiber filters prepared as stated in Section 2.1.
	Sample Set #2 represents passive samplers purchased from Ogawa.
	Sets 1 and 2 used identical Ogawa sample holders.

4.10 Shelf Life of the IGFFs

Thirty-nine IGFFs were prepared according to the procedure described in Section 2.1.3 to determine the potential shelf-life of the nitrite-impregnated filters. Previous reports indicate the Ogawa passive monitors have a conservative shelf-life due to aging of four weeks. The manufacturer indicates an 8-week life-span can be used if necessary and appropriate blank corrections are performed. The aging, or eventual conversion to nitrate appears to be facilitated by oxygen and small amounts of ozone in the atmosphere. The passive monitors use a reaction principle similar to the active sampling filters in this method. For this active sampling method, the extent of nitrite conversion to nitrate on stored filters was used to indicate stability and was measured over a period of up to 58 days.

Procedure: Four tests were conducted to assess IGFF shelf life:

Set 1)	The first test was performed using 15 IGFFs which were stored in a clean and sealed plastic bag after preparation. Five IGFFs were initially taken and served as "control" IGFFs, desorbed with DI H2O and analyzed for total nitrite using peak area; then six IGFFs were desorbed and analyzed after 22 days; finally, the remaining four IGFFs were desorbed and analyzed after a 45-day storage.
Set 2)	A second test was conducted with ten more filters; six were analyzed after 6 days, and four filters analyzed after 28 days.
Set 3)	A third test was performed using 11 IGFFs which were placed in cassettes. The cassettes were then sealed with gel bands and plastic plugs, and stored in a clean and sealed plastic bag after preparation.
Set 4)	This set of four filters was prepared similar to the third set; however, this set was used to assess ability to collect samples after storage. Three of the IGFF/cassettes were used to collect O3 vapor (0.15 ppm O3) after 58 days of storage.

Results: Results are listed in Table 11 and further discussed below:

Set 1)	The conversion of nitrite to nitrate does not significantly occur under the storage conditions specified above for a period of approximately 20-30 days. After 45 days, conversion appears evident. The mean peak area of the IGFFs analyzed after 22 days was only a 9% increase over the Day 0 value and almost a 50%increase after a 45-day storage.
Set 2)	The mean of the IGFFs analyzed after 28 days was only a 2% increase over the value of Day 6.
Set 3)	The mean of the IGFFs analyzed after 57 days was a 23 % increase over the value of Day 0.
Set 4)	After blank correction and 58-day storage, the mean recovery of the O3 collected was 95.5%. Mean O3 found was 0.143 ppm after blank IGFF correction, and 4.0% CV.

Table 11
Shelf-Life Test of IGFF

Sample Set #	Day i	N #	Mean * ×105	SD ×105	CV %	Ratio Xi/X0
1      2     3**	0 22 45 6 28 0 57	5 6 4 6 4 6 5	2.02 2.20 2.93 2.33 2.37 4.71 5.30	0.025 0.080 0.100 0.200 0.091 0.740 1.180	1.3 3.7 3.5 8.6 3.8 15.7 22.2	1.00 1.09 1.45 1.00 1.02 1.00 1.23
*	Peak area.
Xi/Xo	Ratio of IGFFs (mean peak area of Day 1 compared to that of mean Day 0).
**	IGFFs were placed and stored in cassettes, scaled with scaling bands and plastic plugs.

4.11 Summary

The validation results indicate the method meets both the OSHA criteria for accuracy and precision ( Ref. 5.17). The performance during collection efficiency, storage stability, and humidity tests is adequate. For the breakthrough study, it appears that 7.5% breakthrough occurs onto a second IGFF at a concentration of 0.4 ppm O₃ at 0.5 L/min for 240 min. Although the second filter effectively captures the analyte at 0.4 ppm, precautions should be taken at higher concentrations. For O₃ concentrations above 0.4 ppm, a flow rate of 0.25 L/min can be used. Breakthrough is not evident at lower concentrations; however, the second IGFF should always be analyzed to assure capture of all analyte. Experiments above approximately 0.6 ppm using a sample collection rate of 0.25 L/min were not performed due to limitations in the test atmosphere generation system. Detection limits (as NO₃¯) are adequate when samples are taken for 180 min at 0.5 L/min. The conversion of nitrite on the IGFFs appears limited up to 28 days after impregnating if the 2 treated filters are stored in a clean, sealed plastic bag.

The mechanism of the SO₂/O₃ interference which diminishes the O₃ conversion of nitrite to nitrate is unknown. Using the AED-030 (semiconductor sensor) direct-reading instrument side-by-side with the IGFFs while sampling an SO₂/O₃ atmosphere, a corresponding loss of O₃ was not noted. The ability of glass fiber filters to capture and convert SO₂, due primarily to their slightly basic nature, was previously noted in OSHA Method ID-200 for sulfur dioxide. It has been reported in the literature (Ref. 5.22) that the chemistry of SO₂ in ambient air and on surfaces is complex. Fortunately, an oxidizer tube appears to completely remove SO₂ from the sampled stream. Presumably the SO₂ can react with any ozone or oxygen in the presence of nitrite (and possibly glass fiber filters) to form sulfite and eventually sulfate. No significant increase in the sulfate content over background amounts was noted in the chromatograms of IGFF samples taken after using oxidizer tubes to sample an SO₂/O₃ atmosphere. For samples taken in the SO₂/O₃ atmosphere without oxidizer tubes, a significant increase in sulfate content was noted from the resultant oxidation of SO₂. The SO₂ interference appears to be a sampling phenomena occurring at the surface of the IGFFs and is not dependent on analysis. Other environmental pollutants which could potentially adversely affect this ozone sampling method have been considered in the literature. For example, nitric acid vapor, if present, could be collected on the IGFFs during sampling. However, under typical ambient conditions this positive interference probably represents less than 5% of the nitrate formed during the nitrite/ozone reaction (Ref. 5.23). Further study may be needed to determine other oxidized or reduced compounds which may coexist with O₃ and cause either positive or negative interferences, such as peroxyacetyl nitrate (PAN), a strong oxidant, which could oxidize nitrite to nitrate. Since ambient concentrations of PAN are typically 10-20 times smaller than ozone concentrations, significant interference in most locations is not expected (Ref. 5.24).

This method was validated using a UV-VIS detector. A conductivity detector was used to assess potential interference byproducts such as sulfite/sulfate concentrations. Prior to completion of the method another chemist was given approximately 25 field samples to analyze and indicate any problems that may occur during routine analysis. Sample concentrations covered a wide range and were analyzed both by UV and conductivity detection. A difference in ozone results was not noted between the two detectors. Either detector should have adequate sensitivity and capability. The IC conductivity detector has been used for nitrate determination since its inception over 15 years ago. The UV detection technique may be less prone to interferences because of the greater selectivity (wavelength specificity) for each analyte. More crucial to analysis is the ability to separate the nitrite and nitrate peaks using appropriate columns. Precautions should be taken to assure adequate separation prior to sample analysis regardless of which detector is used.

5.1 U.S. Environmental Protection Agency: Evaluation of I Percent Neutral Buffered Potassium Iodide Procedure for Calibration of Ozone Monitors by M. E. Beard, J. H. Margeson, and E. C. Ellis (EPA-600/4-77-005). Environmental Monitoring Series. Research Triangle Park. N.C., 1977.

5.2 National Institute for Occupational Safety and Health: Documentation of the NIOSH Validation Tests by D. Taylor, R. Kupel and ". Bryant (DHEW-NIOSH publication No. 77-185). NIOSH Analytical Methods for Standard Completion Program (Method No. S8 - Ozone). Washington, D.C.: U.S. Government Printing Office, 1977.

5.3 Hekmat, M., P. Fung, and R. Smith: Instability of Ozone Samples Collected in Alkaline Potassium Iodide Solution. Am Ind. Hyg. Assoc. J. 53: 672 (1992).

5.4 Occupational Safety and Health Administration Salt Lake Technical Center: Ozone (KIBRT) in Workplace Atmospheres (USDOL/OSHA-SLCAL Method No. ID-150). Salt Lake City, UT: Occupational Safety and Health Administration Salt Lake Technical Center, 1984.

5.5 Occupational Safety and Health Administration Salt Lake Technical Center: Ozone (Stilbene) in Workplace Atmospheres (USDOL/OSHA-SLCAL Method No. ID-209). Salt Lake City, UT: Occupational Safety and Health Administration Salt Lake Technical Center, 1990 unpublished.

5.6 Sawatari, K.: Personal Dosimeter for Ozone using the Ozonolysis of Trans-stilbene. Industrial Health 22: 117-126 (1984).

5.7 Ku, J.C.: Private Notes, 1989.

5.8 Analytical Instrumental Development Inc.: Ozone Portable Analyzer. Model 560, Technical Document. Avondale, PA: Analytical Instrumental Development Inc., 1980.

5.9 In USA Inc.: Ozone Hunter, AET - 030. Technical Document. Newtonville, MA: In USA Inc., 1992.

5.10 Koutrakis, P., J.M. Wolfson, A. Bunyaviroch, S.E. Froehlich, K. Hirano and J.D. Mulik: Measurement of Ambient Ozone Using a Nitrite-Coated Filter. Anal. Chem. 65: 209-214(1993).

5.11 Hawley, G.G.: The Condensed Chemical Dictionary. I 11th ed. New York: Van Nostrand Reinhold Co., 1987.

5.12 American Society for Testing and Materials (ASTM): Standard Practice for Safety and Health Requirements Relating to Occupational Exposure to Ozone (E591-80). In 1988 Annual Book of ASTM Standards, Section H, Volume 11.03, Philadelphia, PA: ASTM, 1988. pp. 411-439.

5.13 "Ozone" Federal Register 54:12 (19 Jan. 1989). pp. 2519-2520.

5.14 Occupational Safety and Health Administration Salt Lake Technical Center: Ion Chromatography Standard Operating Procedure (Ion Chromatographic Committee). Salt Lake City, UT: Occupational Safety and Health Administration Salt Lake Technical Center, in progress.

5.15 Mandel, J.: Accuracy and Precision, Evaluation and Interpretation of Analytical Results, The Treatment of Outliers. In Treatise On Analytical Chemistry. 2nd ed., Vol. 1, edited by 1. M. Kolthoff and P. J. Elving. New York, NY: John Wiley and Sons, 1978. pp. 282-285.

5.16 National Institute for Occupational Safety and Health: Documentation of the NIOSH Validation Tests by D. Taylor, R. Kupel, and J. Bryant (DHEW/NIOSH Pub. No. 77-185). Cincinnati, OH: National Institute for Occupational Safety and Health, 1977. pp. 1-12.

5.17 Occupational Safety and Health Administration Salt Lake Technical Center: Evaluation Guidelines of the Inorganic Methods Branch. In OSHA Analytical Methods Manual. 2nd ed. Cincinnati, OH: American Conference of Governmental Industrial Hygienists, 1991. pp. I18.

5.18 Burkart, J.A.: General Procedures for Limit of Detection Calculations in the Industrial Hygiene Chemistry Laboratory. Appl. Ind. Hyg. 1: 153-155(1986).

5.19 National Institute for Occupational Safety and Health: Standard Operating Procedures for Industrial Hygiene Sampling and Chemical Analysis, SOP 018, Cincinnati, OH: National Institute for Occupational Safety and Health, Revised Sept., 1992.

5.20 Ogawa & Co., USA, Inc.: The Ogawa Passive Sampler for Ozone Operation Manual. Pompano Beach, FL: Ogawa & Co., USA, Inc., 1993.

5.21 Committee on Medical and Biologic Effects of Environmental Pollutants: Ozone and Other Photochemical Oxidants. National Academy of Sciences, Washington, D.C., 1977. p. 264.

5.22 Chang, D.P.Y.: Sulfur Compounds ' in Ambient Environments and Their Simulation in the Laboratory. Generation of Aerosols and Facilities for Exposure Experiments edited by K.Willeke. Ann Arbor, MI: Ann Arbor Science Publishers, Inc., 1980. p. 302-304.

5.23 Koutrakis, P., P. Mueller: Paper No. 89-71.4. In Proceedings of the 82nd Annual Meeting of the Air and Waste Management Association, 1989.

5.24 Finlayson-Pitts, B. and J.N. Pitts: Atmospheric Chemistry, New York, NY: John Wiley &Sons, 1986.