Methods for the determination of ethylene oxide

Federation of the Employment Accidents Insurance Institutions of Germany (Hauptverband der Berufsgenossenschaften) Centre for Accident Prevention and Occupational Medicine Alte Heerstraße 111, 53757 Sankt Augustin Expert Committee Chemistry Carcinogenic substances Order number: ZH 1/120.27E Established methods Issue: November 1994 Methods for the determination of ethylene oxide Methods tested and recommended by the Berufsgenossenschaften for the determination of ethylene oxide in working areas after discontinuous sampling. For the assessment of working areas, both personal or stationary sampling are possible: 1 Sampling with a pump and adsorption on activated carbon, head space gas chromatography or gas chromatography after desorption. Operating method A: Head space gas chromatography. Ethylene oxide-1-hsgc Operating method B: Gas chromatography after desorption. Ethylene oxide-1-gc (Issue: May 1985) 2 Sampling with a pump and adsorption on activated carbon, gas chromatography after desorption and derivatization with hydrobromic acid. Ethylene oxide-3-gc (Issue: January 1989, revised November 1994) 3 Sampling with a pump and adsorption on the polymeric resin Amberlite XAD-4, gas chromatography after thermal desorption. Ethylene oxide-3-gc (Issue: May 1993) IUPAC name: CAS No: oxirane, ethylene oxide 75-21-8 AirMonitoringMethods,Vol.4

Analytical Methods AirMonitoringMethods,Vol.4 116 1 Sampling with a pump and adsorption on activated carbon, head space gas chromatography or gas chromatography after desorption Principle: Technical data: With a pump a measured air volume is drawn through activated carbon. The adsorbed ethylene oxide is desorbed with a nonvolatile solvent and detected by head space gas chromatography (Operating method A [1,2]) or determined by gas chromatography after desorption with carbon disulfide (Operating method B [1]). Quantification limit: Operating method A: Under the conditions encountered in practice the relative quantification limit is 0.5 ml/m 3 (ppm) ^= 0.9 mg/m 3 ethylene oxide in 10 L air sample. This corresponds to ^= 9 µg ethylene oxide per each activated carbon tube. Selectivity: Advantages: Disadvantages: Apparatus: Operating method B: Under the conditions encountered in practice the relative quantification limit is 0.5 ml/m 3 (ppm) ^= 0.9 mg/m 3 ethylene oxide in 10 L air sample. The absolute quantification limit is 4.5 ng ethylene oxide. The values may be too high as a result of interfering substances. However, interferences can be eliminated by selecting a different column. Personal sampling and specific measurements are possible. No detection of peak concentrations, very time-consuming. Pump equipped with gas meter or flow meter, activated carbon tubes, head space chromatograph or gas chromatograph equipped with flame ionization detector (FID).

117 AirMonitoringMethods,Vol.4 Ethylene oxide 1 Summary This method permits the determination of ethylene oxide concentrations in working areas averaged over the sampling time after personal or stationary sampling. With a personal or a stationary pump a measured air volume is drawn through a glass tube filled with activated carbon. The ethylene oxide is then desorbed with a non-volatile solvent and analysed by head space gas chromatography (HSGC) (Operating method A) or by gas chromatography after desorption with carbon disulfide (Operating method B). Under the conditions encountered in practice the relative quantification limit is 0.5 ml/m 3 (ppm). This corresponds to 0.9 mg/m 3 ethylene oxide for operating method A and a 10 L air sample or to 9 µg ethylene oxide for each activated carbon tube. Under the conditions encountered in practice the relative quantification limit of operating method B is 0.5 ml/m 3 (ppm). This corresponds to 0.9 mg/m 3 ethylene for 10 L air sample. The absolute quantification limit is 4.5 ng ethylene oxide. Detailed description of the method Contents Operating method A: Head space gas chromatography A 2 Equipment, chemicals and solutions A 2.1 Equipment A 2.2 Chemicals A 2.3 Solutions A 3 Sampling and sample preparation A 4 Operating conditions for gas chromatography A 5 Analytical determination A 6 Evaluation A 6.1 Calibration A 6.2 Calculation of the analytical result A 7 Reliability of the method A 7.1 Accuracy A 7.2 Quantification limit A 7.3 Selectivity A 7.4 Recovery A 8 Discussion A 9 Manufacturers A 10 References

Analytical Methods AirMonitoringMethods,Vol.4 118 A 2 Equipment, chemicals and solutions A 2.1 Equipment For sampling and preparation: Pump equipped with gas meter or flow meter Sample vial with polytetrafluoroethylene (PTFE) coated septum and aluminium cap Crimper Activated carbon tubes (standardized, consisting of two sections filled with about 700 and 400 mg activated carbon separated by a porous polymeric material) Caps for sealing the opened activated carbon tubes For analytical determination: Head space gas chromatograph equipped with flame ionization detector, automatic headspace sampler, if necessary with back-flush device Compensation recorder and/or integrator Gastight syringes, if sample charging is carried out manually, instead of using an autosampler A 2.2 Chemicals Ethylene oxide, purity >99% Dimethylphthalate (DMP) for GC Gases for gas chromatography: Nitrogen, hydrogen, purified air or synthetic air A 2.3 Solutions Ethylene oxide standard solution: Solution of about 9 mg ethylene oxide in 20 ml DMP. About 5 ml gaseous ethylene oxide is slowly injected with a gastight syringe through the septum into the liquid phase of a sample vial containing 20 ml DMP. At the tip of the needle ethylene oxide bubbles should not appear so that ethylene oxide does not escape when the needle is withdrawn through the septum. The weight of ethylene oxide is determined by weighing to the nearest ± 0.1 mg. Ethylene oxide calibration solution: Solution of about 4.5 mg ethylene oxide in 100 ml DMP. 10 ml of the ethylene oxide standard solution is transferred to a 100 ml volumetric flask containing 50 ml DMP and diluted to the mark with DMP (cf. Sect. A 8).

119 AirMonitoringMethods,Vol.4 Ethylene oxide A 3 Sampling and sample preparation An activated carbon tube is opened and connected to the pump. Pump and tube are either carried by a person during working hours or stationary sampling is performed. The air sample volume of 10 L per 1 g activated carbon must not be exceeded. The flow rate should not exceed 3 L/h. Each section of the loaded sampling tube is transferred to a 25 ml sample vial, 1 ml DMP is added and the sample vials are sealed tightly with septa and aluminium caps. A 4 Operating conditions for gas chromatography The method was characterized under the following experimental conditions. Apparatus: Head space gas chromatograph Perkin Elmer F 45 with flame ionization detector, autosampler and back-flush device Column: Material: Quartz capillary Length: 50 m Internal diameter: 0.3 mm Stationary phase: Carbowax 20 M Temperatures: Sample thermostat: 70 C Autosampler: 150 C Injector block: 150 C Column: Detector: 250 C Carrier gas: Nitrogen: 2 ml/min Detector gases: Hydrogen: 30 ml/min Synthetic air: 400 ml/min 25 C, isothermal (room temperature) A 5 Analytical determination A gas chromatogram of the contents of an unloaded sampling tube and 1 ml DMP is recorded according to the instructions for the analytical determination given below to ensure that the dimethylphthalate and the activated carbon are free from interfering impurities. The loaded sample vial containing activated carbon from a sampling tube and 1 ml DMP is placed in the sample thermostat at 70 C for 30 min. The same volume as used for calibration (cf. Sect. A 6.1) is then injected from the vapour space into the gas chromatograph and a gas chromatogram is recorded (operating conditions cf. Sect. A 4). Quantitative evaluation is carried out using the area or the height of the ethylene oxide peaks.

Analytical Methods AirMonitoringMethods,Vol.4 120 A 6 Evaluation A 6.1 Calibration In a sample vial the contents of an unloaded activated carbon tube are mixed with 1 ml of the ethylene oxide calibration solution and closed tightly. The sample vial is placed in the thermostat at 70 C for 30 min. A fixed volume is then injected into the gas chromatograph from the vapour space and a gas chromatogram is recorded (cf. Sect. A 4). Quantitative evaluation is carried out using the area or the height of the ethylene oxide peaks (cf. Sect. A 8). A 6.2 Calculation of the analytical result The ethylene oxide concentration in the air sample in mg/m 3 is calculated according to the equation (1): F. w1. 1000 cw F. V (1) 1 The concentration by volume c v in ml/m 3 is calculated from c w at 20 C and 1013 hpa: c v = 0.55 c w Legend: c w Ethylene oxide concentration by weight in the air sample in mg/m 3 c v Ethylene oxide concentration by volume in the air sample in ml/m 3 (ppm) F Peak area (peak height) of ethylene oxide in the sample F 1 Peak area (peak height) of ethylene oxide in the calibration solution w 1 Weight of ethylene oxide in 1 ml of the calibration solution V Air sample volume in L (2) A 7 Reliability of the method A 7.1 Accuracy For ethylene oxide concentrations in the range of 8 ml/m 3 to 33 ml/m 3 the method described yielded a standard deviation of 5.4% and a confidence interval of about 10% (n = 30, P = 95 %) compared to the values measured by a process gas chromatograph. The measuring error of the process gas chromatograph of about 3 % was taken into account.

121 AirMonitoringMethods,Vol.4 Ethylene oxide A 7.2 Quantification limit Among other things the quantification limit depends on the type of gas chromatograph and its quality. Under the conditions encountered in practice, amounts as low as 9 µg ethylene oxide per 1 g activated carbon can be detected. This corresponds to ethylene oxide concentrations of 0.5 ml/m 3 (ppm) or 0.9 mg/m 3 for an air sample volume of 10 L. A 7.3 Selectivity The selectivity of the method depends on the type of gas chromatography column. In practice the capillary column described has proved suitable. If interfering substances are present the separation phase must be changed. A 7.4 Recovery Recovery was >90% under the experimental conditions described in Sect. A 7.1. A 8 Discussion Instead of dimethyl phthalate, chlorobenzene, or in the absence of acids and bases, water can be used as a desorbent with no change in the operation conditions. When injecting from the head space, less-volatile compounds, depending on their vapour pressure, especially DMP, are also injected into the gas chromatograph. If the gas chromatograph is not equipped with a back-flush device, interfering peaks or baseline fluctuations may result. Sealed with caps, the loaded activated carbon tubes can be stored for one day at room temperature. However, it is recommended that they are kept cool. If possible, the peaks of the sample and the calibration solutions should be of equal size. In general, deviation by a factor of ± 5 (above and below) is permissible without repeating the gas chromatographic determination with adapted concentrations. Calibration must be carried out at regular intervals. Acid and bases may lead to errors in the measured values and must be taken into account in the complete method. Continuous determination of ethylene oxide using a process gas chromatograph is also possible; however, extensive data obtained with this method is not yet available.

Analytical Methods AirMonitoringMethods,Vol.4 122 A 9 Manufacturers Pump: e.g. Compur Electronic GmbH, München, Du Pont Instruments, Supplier in Germany: DEHA-Haan & Wittmer GmbH, Friolzheim Activated carbon tubes: Type Qazi & Ketcham (SKC Cat. No. 226-236), e. g. Compur Electronic GmbH, München, Fleischhacker KG, Schwerte/Ruhr, Günter Karl OHG, Geisenheim, Chrompack Deutschland, Müllheim (Cat. No. 14965) MSA (Charcoal Collection Tube for Organic Hydrocarbons, part. No. 463 072), e.g. Auergesellschaft GmbH, Berlin Dräger, Type B, e. g. Drägerwerk AG, Lübeck Head space gas chromatograph: e. g. Bodenseewerk Perkin Elmer & Co GmbH, Überlingen, Carlo Erba, Supplier in Germany: Erba Science, Hofheim/Ts., Siemens AG, Karlsruhe A 10 References [1] Hachenberg H, Schmidt AP (1979) Gas Chromatographic Head Space Analysis. Heyden and Sons, London, Philadelphia, Rheine [2] Qazi, Ketcham (1977) A new method for monitoring personal exposure to ethylene oxide in the occupational environment. Am. Ind. Hyg. Ass. J. 38: 635 647.

123 AirMonitoringMethods,Vol.4 Ethylene oxide Detailed description of the method Contents Operating method B: Gas chromatography after desorption B 2 Equipment, chemicals and solutions B 2.1 Equipment B 2.2 Chemicals B 2.3 Solutions B 3 Sampling and sample preparation B 4 Operating conditions for gas chromatography B 5 Analytical determination B 6 Evaluation B 6.1 Calibration curve B 6.2 Calculation of the analytical result B 7 Reliability of the method B 7.1 Accuracy B 7.2 Quantification limit B 7.3 Selectivity B 8 Discussion B 9 Manufacturers B 10 References B 2 Equipment chemicals and solutions B 2.1 Equipment For sampling and preparation: Pump equipped with gas meter or flow meter Sample vial with polytetrafluoroethylene (PTFE) coated septum and aluminium cap Crimper Activated carbon tubes, (standardized, consisting of two sections filled with about 700 and 400 mg activated carbon separated by a porous polymeric material) Caps for sealing the opened activated carbon tubes For analytical determination: Gas chromatograph equipped with flame ionization detector Compensation recorder and/or integrator

Analytical Methods AirMonitoringMethods,Vol.4 124 B 2.2 Chemicals Ethylene oxide, purity >99% Carbon disulfide for GC, analytical grade Gases for gas chromatography: Nitrogen, hydrogen, purified air or synthetic air B 2.3 Solutions Ethylene oxide standard solution: Solution of about 9 mg ethylene oxide in 20 ml carbon disulfide. About 5 ml gaseous ethylene oxide is slowly injected through the septum with a gastight syringe into the liquid phase of a sample vial containing 20 ml carbon disulfide. Ethylene oxide bubbles should not appear at the tip of the needle, to avoid escape of ethylene oxide when the needle is withdrawn through the septum. The weight of ethylene oxide is determined by weighing to the nearest ±0.1 mg. Ethylene oxide calibration solutions: Solutions of about 0.45, 0.9, 2.25, and 4.50 mg ethylene oxide in 20 ml carbon disulfide. 1.0, 2.0, 5.0 and 10.0 ml of the ethylene oxide standard solution are each transferred to a 20 ml volumetric flask and diluted to the mark with carbon disulfide. B 3 Sampling and sample preparation An activated carbon tube is opened and connected to the pump. Pump and tube are either carried by a person during working hours or stationary sampling is performed. The air sample volume of 10 L per 1 g activated carbon must not be exceeded. The flow rate should not exceed 3 L/h. Each section of the loaded sampling tube is transferred to a 5 ml sample vial, 2 ml carbon disulfide is added and the sample vials are sealed tightly with septum and aluminium cap. The vials are allowed to stand for 15 min and are occasionally shaken (sample solutions). B 4 Operating conditions for gas chromatography The method was characterized under the following experimental conditions: Apparatus: Carlo Erba, Type 2300 equipped with flame ionization detector Column: Material: Stainless steel Length: 2 m Internal diameter: 3 mm Stationary phase: Porapak P

125 AirMonitoringMethods,Vol.4 Ethylene oxide Temperatures: Injector block: 150 C Column: 120 C, isothermal Detector: 150 C Carrier gas: Nitrogen: 50 ml/min B 5 Analytical determination A gas chromatogram of the contents of an unloaded sampling tube and 2 ml carbon disulfide is recorded according to the instructions for the analytical determination given below to ensure that the carbon disulfide and the activated carbon are free from interfering impurities. 1 µl of each of the sample solutions (according to Sect. B 3) is injected into the gas chromatograph and a gas chromatogram is recorded. Quantitative evaluation is carried out using the peak areas or the peak heights obtained. B 6 Evaluation B 6.1 Calibration curve 1 µl of each of the calibration solutions is injected into the gas chromatograph. To obtain the calibration curve, the peak areas (peak heights) determined are plotted against the ethylene oxide weights in µg in 2 ml of each calibration solution. B 6.2 Calculation of the analytical result The ethylene weight in µg in 2 ml of the sample solution is taken from the calibration curve using the peak areas (peak heights). The ethylene oxide concentration in the air sample in mg/m 3 is calculated according to the equation (1): c w w V The concentration by volume c v in ml/m 3 is calculated from c w at 20 C and 1013 hpa: (1) c v = 0.55 c w (2) Legend: c w Ethylene oxide concentration by weight in the air sample in mg/m 3 c v Ethylene oxide concentration by volume in the air sample in ml/m 3 (ppm)

Analytical Methods AirMonitoringMethods,Vol.4 126 w V Weight of ethylene oxide in µg in the sample solution taken from the calibration curve Air sample volume in L B 7 Reliability of the method B 7.1 Accuracy For a gas mixture containing 10 ml/m 3 ethylene oxide prepared in the laboratory a relative standard deviation of ± 6% for the complete analytical method was determined from 10 individual measurements. Experience has shown that the accuracy of operating methods A and B are comparable. B 7.2 Quantification limit Under the experimental conditions described the relative quantification limit for ethylene oxide is 0.5 ml/m 3 (ppm) ^= 0.9 mg/m 3 for an air sample volume of 10 L, 2 ml of the desorption solution and a 1 µl injection volume. The absolute quantification limit is 4.5 ng ethylene oxide. This corresponds to 9 µg ethylene oxide for each activated carbon tube. B 7.3 Selectivity The selectivity of the method depends on the type of gas chromatography column. In practice the packed column described has proved suitable. The column packing must be changed if interfering substances are present. B 8 Discussion The loaded activated carbon tubes sealed with caps can be stored at room temperature for one day. However, it is recommended to store them cool. Acids and bases may lead to errors in the measuring values and must be taken into account in the complete method.

127 AirMonitoringMethods,Vol.4 Ethylene oxide B 9 Manufacturers Pump: e. g. Compur Electronic GmbH, München, Du Pont Instruments, Supplier in Germany: DEHA-Haan & Wittmer GmbH, Friolzheim Activated carbon tubes: Type Qazi & Ketcham (SKC Cat. No. 226-236), e. g. Compur Electronic GmbH, München, Fleischhacker KG, Schwerte/Ruhr, Günter Karl OHG, Geisenheim, Chrompack Deutschland, Müllheim (Cat. No. 14965) MSA (Charcoal Collection Tube for Organic Hydrocarbons, part. No. 463 072), e. g. Auergesellschaft GmbH, Berlin Dräger, Type B, e.g. Drägerwerk AG, Lübeck Gas chromatograph: e. g. Bodenseewerk Perkin Elmer & Co GmbH, Überlingen, Carlo Erba, Supplier in Germany: Erba Science, Hofheim/Ts., Siemens AG, Karlsruhe, Varian GmbH, Darmstadt B 10 References [1] Qazi, Ketcham (1977) A new method for monitoring personal exposure to ethylene oxide in the occupational environment. Am. Ind. Hyg. Ass. J. 38: 635 647.

Analytical Methods AirMonitoringMethods,Vol.4 128 2 Sampling with a pump and adsorption on activated carbon, gas chromatography after desorption and derivatization with hydrobromic acid This method permits the determination of ethylene oxide concentrations in working areas averaged over the sampling time after personal or stationary sampling. Principle: Technical data: With a pump a measured air volume is drawn through an activated carbon tube. The adsorbed ethylene oxide is desorbed with a mixture of toluene and carbon disulfide, reacted with hydrobromic acid to form bromoethanol and analysed by gas chromatography. Quantification limit: absolute: 0.71 ng bromoethanol ^= 0.25 ng ethylene oxide, relative: 0.1 mg/m 3 ethylene oxide for 12 L air sample, 5 ml desorption solution and 1 µg injection volume. Selectivity: Advantages: Disadvantages: Apparatus: The selectivity must be checked in each individual case. Personal and selective measurements are possible. Peak concentrations are not detected. Pump equipped with gas meter or flow meter, activated carbon tubes, gas chromatograph equipped with electron capture detector (ECD).

129 AirMonitoringMethods,Vol.4 Ethylene oxide Detailed description of the method Contents 1 Equipment, chemicals and solutions 1.1 1.2 Equipment Chemicals 1.3 2 Solutions Sampling 3 Analytical determination 3.1 Sample preparation and analysis 3.2 Operating conditions for gas chromatography 4 Evaluation 4.1 Calibration 4.2 Calculation of the analytical result 5 5.1 Reliability of the method Accuracy 5.2 5.3 Quantification limit Selectivity 5.4 6 Recovery Discussion 1 Equipment, chemicals and solutions 1.1 Equipment For sampling: Pump equipped with flow meter or gas meter and holder for the activated carbon tubes (e. g. Compur 4903 from Hartmann & Braun AG, Compur Monitors, München) Activated carbon adsorption tubes (standardized, consisting of two sections filled with about 400 and 200 mg activated carbon separated by a porous polymeric material) (e.g activated carbon tubes from MSA, Pittsburgh, supplier in Germany: Auer, Berlin) For sample preparation and analytical determination: Flat-bed shaker 10 ml and 100 ml Volumetric flasks 10 ml Beaded rim glass Sample vials with polytetrafluoroethylene (PTFE) coated septum and aluminium cap Glass cutter 1 ml Pipette Glass capillaries for dosing the hydrobromic acid Gas chromatograph equipped with ECD Recording and/or computing unit: Compensation recorder and/or computing integrator

Analytical Methods AirMonitoringMethods,Vol.4 130 1.2 Chemicals Carbon disulfide, analytical grade Toluene, analytical grade Ethylene oxide, purity > 99 % Hydrobromic acid, 47%, analytical grade Potassium carbonate, analytical grade 2-Bromoethanol, about 95 % (GC) (e. g. Aldrich, Steinheim) 1.3 Solutions Desorption solution: Mixture of toluene and carbon disulfide in a 99:1 volume ratio. 2-Bromoethanol stock solution: Solution of 3.53 mg/ml bromoethanol in elution solution. 200 µl ( ^= 353 mg) 2-bromoethanol is transferred to a 100 ml volumetric flask and diluted to the mark with the desorption solution and occasionally shaken. 2-Bromoethanol calibration solutions: 2-Bromoethanol solutions of 0.71, 1.77, 3.53, 17.65 and 35.3 µg/ml. 2, 5, 10, 50, and 100 µl of the 2-bromoethanol stock solution are transferred to five separate 10 ml volumetric flasks and diluted to the mark with the desorption solution. Using these solutions and with a sample volume of 12 L a concentration range of 0.1 5.2 mg/m 3 ethylene oxide is covered. 2 Sampling An activated carbon tube is opened and connected to the pump. Pump and sampling tube are either carried by a person during working hours or they are used in a stationary position. The flow rate must not exceed 2.4 L/h. The total air sample volume must not exceed 12 L. 3 Analytical determination 3.1 Sample preparation and analysis The contents of the loaded sampling tube are transferred to a sample vial. For desorption of the ethylene oxide, 5 ml of the desorption solution is added. The vial is occasionally shaken. If necessary, the activated carbon sections are prepared separately.

131 AirMonitoringMethods,Vol.4 Ethylene oxide After 1 hour of desorption, 2 ml of the liquid phase is removed with a pipette, transferred to a sample vial and 30 µl hydrobromic acid is added. The vial is shaken on a flat-bed shaker for 30 min. The surplus hydrobromic acid is then neutralized with 75 mg potassium carbonate. The supernatant solution (sample solution) is transferred to a sample vial and the bromoethanol formed in the derivatization of ethylene oxide with hydrobromic acid is determined using gas chromatography (cf. Sect. 3.2). For this purpose 1 µl of the sample solution is injected. Quantitative determination is carried out according to the external standard method using the peak areas or peak heights. Derivatization must not be carried out in the presence activated carbon. Otherwise the formation of bromoethanol is not quantitative. 3.2 Operating conditions for gas chromatography The method was characterized under the following instrumental conditions: Apparatus: Varian 3400 equipped with ECD and autosampler Varian 8100 Column: Material: Quartz capillary Length: 25 m Internal diameter: 0.25 mm Stationary phase: FFAP (modified polyethylene glycol) Film thickness: 1 µm Detector: Electron capture detector Temperatures: Injector block: 210 C Detector: 250 C Column temperature programme: Starting temperature: 80 C Heating rate: 10 C/min Final temperature: 180 C, 5 min isothermal Injection: Splitless Carrier gas: Helium 4.6 Make-up gas: Nitrogen 4 Evaluation 4.1 Calibration 1 µl of each of the calibration solutions (cf. Sect. 1.3) is injected into the gas chromatograph. The calibration curve is obtained by plotting the peak areas (peak heights) against the corresponding weights of ethylene oxide obtained after the conversion of the 2-bromoethanol weights (conversion factor 0.352).

Analytical Methods AirMonitoringMethods,Vol.4 132 4.2 Calculation of the analytical result The ethylene oxide concentration in the air sample in mg/m 3 is calculated according to the equation (1): w c w V (1) To calculate the concentration by volume c v in ml/m 3 at 20 C and 1013 hpa: c v = 0.55 c w Legend: c w Ethylene oxide concentration by weight in the air sample in mg/m 3 c v Ethylene oxide concentration by volume in the air sample in ml/m 3 (ppm) w Weight of ethylene oxide in the sample solution in µg V Air sample volume in L (2) 5 Reliability of the method 5.1 Accuracy To determine the accuracy of the method 2, 5, 10, 25 and 35 µl of a solution (1.8 mg ethylene oxide per ml in the desorption solution) were each injected into a glass tube, six times. Laboratory air was drawn through each glass tube connected serially with two activated carbon tubes with a flow rate of 2.4 L/h for 5 hours. For an air sample volume of 12 L, the quantities injected correspond to average concentrations of 0.3 5.25 mg/m 3. The relative standard deviations were between 6 and 12%. 5.2 Quantification limit The absolute quantification limit is 0.71 ng bromoethanol. This corresponds to 0.25 ng ethylene oxide. For 12 L air sample, 5 ml of the sample solution and 1 µl injection volume the relative quantification limit is 0.1 mg/m 3. This corresponds to 0.05 ml/m 3 (ppm) of ethylene oxide. 5.3 Selectivity The selectivity must be checked in each individual case. Compared to that of the ethylene oxide method with flame ionization detection, the selectivity is significantly increased by the derivatization of the ethylene oxide and the use of an electron capture detector.

133 AirMonitoringMethods,Vol.4 Ethylene oxide 5.4 Recovery For an air sample volume of 12 L/h and a flow rate of 2.4 L/h recovery was >95% under the conditions described in Sect. 5.1. Recovery was not influenced by a relative humidity of the calibration gas of up to 75 %. 6 Discussion The loaded activated carbon tubes should be analysed immediately after sampling. The shelf life of pure ethylene oxide in the adsorbed state is 7 days.

Analytical Methods AirMonitoringMethods,Vol.4 134 3 Sampling with a pump and adsorption on the polymeric resin Amberlite XAD-4, gas chromatography after thermal desorption This method permits the determination of ethylene oxide concentrations in working areas averaged over the sampling time after personal or stationary sampling. Principle: Technical data: With a pump a measured air volume is drawn through a metal tube filled with XAD-4. The adsorbed ethylene oxide is desorbed by heating in a thermal desorber and analysed by gas chromatography. Quantification limit: absolute: 20 ng ethylene oxide, relative: 0.1 mg/m 3 ethylene oxide in 200 ml air sample. Selectivity: Advantages: Disadvantages: Apparatus: The selectivity must be checked in each individual case. Personal and selective measurements are possible. Peak concentrations are not detected, generally only one measurement for each adsorption tube possible. Pump for 1 10 ml/min, flow meter for 1 10 ml/min, XAD-4 adsorption tubes, thermal desorber, gas chromatograph equipped with flame ionization detector (FID), Permeation oven and permeation tube filled with ethylene oxide for the preparation of calibration gases (e.g. MCZ, Rosbach).

135 AirMonitoringMethods,Vol.4 Ethylene oxide Detailed description of the method Contents 1 Equipment and chemicals 1.1 Equipment 1.2 Chemicals 2 Sampling 3 Analytical determination 3.1 Sample preparation and analysis 3.2 Instrumental operating conditions 4 Evaluation 4.1 Calibration 4.2 Calculation of the analytical result 5 Reliability of the method 5.1 Accuracy 5.2 Quantification limit 5.3 Selectivity 6 Discussion 7 References 1 Equipment and chemicals 1.1 Equipment For sampling: Pump, flow rate 1 10 ml/min (e.g. SKC-PCEx224 from Analyt, Müllheim) Adsorption tubes with XAD-4 as collection phase The adsorption tubes are metal tubes filled with XAD-4. The external diameter and length are adapted to the thermal desorber used. The length of the adsorption section which is enclosed by metal grids depends on the heating zone of the desorption apparatus. After the tube is filled it has to be prepurified in a stream of helium several times in the desorber at 150 C. It is then tightly sealed. Caps for sealing the adsorption tubes For sample preparation and analytical determination: Thermal desorber Gas chromatograph equipped with flame ionization detector Recording and/or computing unit: Compensation recorder and/or computing integrator Permeation furnace and permeation tubes filled with ethylene oxide for the preparation of calibration gases (e. g. MCZ, Rosbach)

Analytical Methods AirMonitoringMethods,Vol.4 136 1.2 Chemicals Polymer resin, Amberlite XAD-4, 60-80 mesh (e.g Serva, Heidelberg) Tenax TA, 60-80 mesh (e.g. Chrompack, Frankfurt) Gases for gas chromatography: Helium (4.6), purity > 99.996% Hydrogen (5.0), purity > 99.999% Purified or synthetic air 2 Sampling Prior to sampling the prepurified adsorption tube has to be heated again in the thermal desorber for 15 min at 100 C in a stream of helium and closed tightly. For sampling the adsorption tube is opened and connected to the pump. During sampling the pump and tube are carried by a person or they are used in a stationary position. The flow rate and air sample volume must be adjusted to ensure that 2/3 of the breakthrough volume is not exceeded [1]. These conditions are fulfilled for a tube content of 450 mg XAD-4 at 20 C and 500 ml air sample volume. After sampling the tube is again sealed tightly. 3 Analytical determination 3.1 Sample preparation and analysis The loaded adsorption tube is opened and placed into the thermal desorber. The adsorbed ethylene oxide is desorbed by heating in a stream of helium, trapped in a cooling trap packed with Tenax TA, transferred to the analytical column of the gas chromatograph by flash-heating and then chromatographed. Quantitative evaluation is performed according to the external standard method using the peak areas or the peak heights. 3.2 Instrumental operating conditions The method was characterized under the following instrumental conditions: Apparatus: Thermal desorber ATD-400 and gaschromatograph 2000 equipped with FID from Perkin Elmer GmbH, Überlingen Column: Material: Quartz capillary Length: 25 m Internal diameter: Stationary phase: 0.32 mm Poraplot Q from Chrompack, Frankfurt

137 AirMonitoringMethods,Vol.4 Ethylene oxide Temperatures: Thermal desorber: Desorption furnace: 100 C Transfer line: 70 C Column: 150 C isothermal Detector: 200 C Cooling trap: Filling: 40 mg Tenax TA (60-80 mesh) Temperatures: 30 C (adsorption) 300 C (injection) Desorption time: 5 min Carrier gas: Helium 200 kpa Desorption rate: 45 ml/min Split: 10 ml/min Detector gases: Hydrogen: about 30 ml/min Synthetic air: about 300 ml/min 4 Evaluation 4.1 Calibration A calibration gas is prepared according to the method of permeation [3]. For this purpose a permeation tube filled with ethylene oxide is kept at 35 C in the permeation furnace. With a permeation tube of 5 cm in length a permeation rate of 6 µg/min is obtained. Concentrations of 0.5 12 mg/m 3 are prepared by dilution with synthetic air. From this calibration gas known volumes are drawn onto adsorption tubes as described in Sect. 3 and analysed [1]. The calibration curve is obtained by plotting the measured peak areas or peak heights against the ethylene oxide weights in ng. The curve is linear in the range of 0.1 20 mg/m 3. 4.2 Calculation of the analytical result The ethylene oxide weight corresponding to the peak area or peak height is taken from the calibration curve. The ethylene oxide concentration in the air sample in mg/m 3 is calculated according to the equation (1): c w w V Legend: c w Ethylene oxide concentration by weight in the air sample in mg/m 3 w V Weight of ethylene oxide in the sample in ng Air sample volume in ml (1)

Analytical Methods AirMonitoringMethods,Vol.4 138 To calculate the concentration by volume c v in ml/m 3 at 20 C and 1013 hpa: c v = 0.55 c w (2) 5 Reliability of the method 5.1 Accuracy The standard deviation of the method was checked in the range of 0.5 12 mg/m 3 for three calibration gas concentrations (preparation according to [3]) with a relative air humidity of 40 60%. Table 1. Characteristic data for the determination of ethylene oxide (n = 6). Concentration mg/m 3 Standard deviation mg/m 3 Standard deviation (rel.) s % 12 0.173 1.34 2.1 0.56 0.019 0.026 0.94 4.39 5.2 Quantification limit The quantification limit is 20 ng ethylene oxide per adsorption tube. This corresponds to a relative quantification limit of 0.1 mg/m 3 in a 200 ml air sample. 5.3 Selectivity The selectivity of the method depends on the analytical column used. In practice the column described has proved suitable. A different column must be selected if interfering components are present. 6 Discussion This method was tested using XAD-4 tubes but other adsorbents may also be used. If other adsorbents or tubes of different dimensions are used the breakthrough volume has to be determined again. If stored for longer periods, the adsorption tubes must be heated immediately before sampling in the thermal desorber as interfering components may be transferred to the adsorption phase from the material of the cap washers. The water vapour in humid air (>90% relative humidity) hardly displaces ethylene oxide from the surface of the adsorption phase.

139 AirMonitoringMethods,Vol.4 Ethylene oxide In the temperature range of 20 to 25 C no losses from adsorption tubes loaded with ethylene oxide were found one day after sampling. This was established by determination with a calibration gas of 50% relative humidity at a concentration of 2.1 mg/m 3. Determination of 1,2-epoxypropane is also possible using this method. Determination of ethylene oxide in alveolar air is also possible with two serially connected adsorption tubes using the respiratory gas sampler BSC-1 (Perkin Elmer GmbH, Überlingen) equipped with a holder for adsorption tubes [2]. 7 References [1] Tschickardt M (1990) Routineeinsatz des Thermodesorbers ATD-50 in der Gefahrstoffanalytik. In: Perkin Elmer GmbH (Ed) Angewandte Chromatographie, Vol 51. [2] Tschickardt M, Petersen R (1990) Bestimmung organischer Luftschadstoffe in Alveolarluft. In: Perkin Elmer GmbH (Ed) Angewandte Chromatographie, Vol 51. [3] Association of Engineers (VDI) (1980) VDI Guidline 3490, Sheet 9, Measurement of gases Preparation of calibration gas mixtures by permeation procedures. Beuth Verlag, Berlin.