California Department of Pesticide Regulation

Environmental Monitoring And Pest Management Branch

1020 N Street, Room 161

Sacramento, California 95814-3624









PROTOCOL 162: EVALUATION OF METHYL BROMIDE MONITORING METHODS



December 9, 1997







I. Introduction



Air monitoring methods generally require more extensive method development work than other media. The main reason for this is the interdependence of sampling and analysis procedures. In media like water and soil, sampling and analysis are almost completely decoupled. How we take a well water sample, for example, does not vary depending on the chemical we try to detect. In air, however, sampling procedures are tailored not only to the chemical of interest but also to the analysis method used by the laboratory. This makes it mandatory that the sampling procedure is fully included in the method development and method performance evaluations, unlike water and soil analyses where the sampling procedure is not considered part of the analysis method. There are just two aspects that air analysis shares with water and soil. Like water, air consists of two components: gaseous species and suspended particulate matter. Particulates are generally collected on filter surfaces, true gases are traditionally sampled by capture on an adsorbent material. This is where the soil aspect comes in: the laboratory actually analyzes for a compound bound to the adsorbent, not for a compound in air.



To check the performance of our methyl bromide monitoring method using operating conditions that are representative of field sampling procedures, a gas mixing and delivery system was set up in our West Sacramento laboratory. This protocol describes a set of tests that will try to quantify the effects of environmental conditions like temperature, humidity and concentration on method performance.



Air samples containing a known fraction of methyl bromide are taken over a period of six to twelve hours using charcoal tubes to trap the methyl bromide. The charcoal tubes are then analyzed by the laboratory using extraction with ethyl acetate detection by GC/ECD. A second method under evaluation will sample this test air in a Summa canister, followed by laboratory analysis using cryofocusing and GC/MS. A third method using charcoal tubes and headspace analysis may also be included in some tests.







II. Objectives



A) Determine the behavior of our current sampling and analysis procedures under varying ambient conditions.



B) Evaluate other sampling and/or analysis methods.





III. Personnel



This study will be conducted by personnel from the Environmental Hazards Assessment Program in the Environmental Monitoring and Pest Management Branch of the Department of Pesticide Regulation. Study personnel include:

Project Leader: Heinz Biermann

Senior Scientist: Bruce Johnson

Statistics: Terri Barry

Chemist: Paul Lee

Lab Liaison: Nancy Miller

Agency and Public Contact: Pat Dunn



All questions concerning this study should be directed to Pat Dunn at (916) 324-4077.





IV. Study Plan / Experimental Design



Charcoal tube samples will be taken using the same equipment and operating conditions that are used in the field, i.e., SKC sampling tubes with SKC pumps operating at a flow of about 15 ml/min for six or twelve hours. Summa canisters will collect whole air samples simultaneous with the SKC samplers. Other sampling/analysis methods may be added later on.



Sample Preparation



All samples will be prepared at our West Sacramento facility using the gas mixing system shown in Figure 1. A commercial compressed gas tank containing a certified mixture of methyl bromide in air and commercial tank of compressed air are the gas sources. The flows from these two tanks are regulated and monitored by mass flow controllers. The flow from the compressed air tank is split with one part of the flow bubbling through water to add moisture to the air. The amount of moisture in the final mix can be varied using a valve that determines the ratio of the air going through and bypassing the bulb with water. The moist air is then combined with the flow from the methyl bromide/air mixture in a second bulb. The outflow from this bulb passes through a manifold with eight sampling ports. The effluent from this manifold can be routed through a glass cell attached to an FTIR system. This allows for an independent confirmation of the methyl bromide concentration. Because of the high FTIR detection limit, this check will be employed only at levels above 500 ppb.

Figure 1. Schematic of gas mixing system.

Near the end of the exhaust line, inside a ventilated hood, a relative humidity probe is inserted into the flow to measure the moisture content of the gas mixture. The charcoal tubes will be sent to the CDFA laboratory for analysis; the Summa canisters will be shipped to Quanterra. Because the sampling equipment operates at flows of about 15 ml/min, total flow through the system will be about 100 to 200 ml/min. The total volume of the gas mixing system is about 1500 ml. These low flows require an adjustment period of about one hour between runs.



The FTIR sampling volume adds about 5500 ml to the total volume, but is located past the other air sampling ports. To make sure that the FTIR cell contains the same amount of methyl bromide that is present at the sampling ports, equilibration times of up to three hours are required.

The general steps involved in the sample preparation procedure are:

a) Set flows and humidity.

b) Equilibrate for at least one hour.

c) Sample for six hours.



Four sets of experiments are planned to study the effects of storage, concentration, moisture and temperature. Detailed descriptions of each test are given in the following sections.



Storage Stability Tests



Because of the six-hour long sampling periods, sample sets can only be prepared over a number of days. If a whole set is then submitted to each laboratory, samples within each set will have different storage times. It has to be checked first that varying holding times between one day and two weeks will not influence the analysis results. Quanterra provided documentation that methyl bromide inside a Summa canister can be stored for up to 30 days at ambient temperatures. Therefore, only a few check samples will be prepared using Summa canisters.



Because there is a possibility that the methyl bromide level and the moisture content may affect method performance, storage stability will be tested for two levels (70 ppb and 700 ppb) and two relative humidities (25 % and 70 %). A constant level of methyl bromide will be set to flow through the system for one week. During this week, two charcoal tube samples each will be taken in the morning at the low humidity level and two samples in the evening at the high humidity level. At the end of the week, one tube from each set will be sent to the laboratory for immediate analysis. The second sample from each set will be sent one week later, yielding storage stability data covering two weeks. This procedure will be repeated with the lower methyl bromide concentration. An analysis of variance will be performed on the data, and common functions describing decay processes will be fitted to the data if a significant effect is found.



Effects of moisture and concentration



The range of relative humidities and concentrations are based on the range of values seen in our previous field studies. For moisture content, relative humidities of 20 %, 50 % and 80 % will be targeted. Concentration levels will be 2 ppm, 200 ppb and 20 ppb. Three replicate samples each will be taken for every level, but only one sample set (nine samples) at a time will be submitted for analysis, spaced about three days apart (if there are no problems with storage times). This will provide some information about between-days variability in the analysis.



The data will be analyzed to check for significant differences in the recoveries at the various levels of concentration and moisture. If there are significant differences, these data will not be sufficient to derive an equation describing the observed behavior as this most likely involves nonlinear dependencies.





Effects of variation in moisture content and concentration



The tests in this section will only be done if the previous results indicate that the effects of moisture and/or concentration are of such magnitude that they cannot be neglected.



Because ambient humidity and the methyl bromide concentration itself usually vary during the long sampling periods, it needs to be checked if variations in these factors affect the analysis result. For example, do the analysis results vary if moisture was low during the initial hours, but high during the later hours of sampling, and vice versa? For moisture, two tests each are planned at a methyl bromide concentration of 200 ppb: two hours of 20 % RH, followed by two hours at 50 %, then two hours at 80 %, and the reverse sequence. Note that there will be no instantaneous change in relative humidity. Rather, at low flows, it may take about one hour to reach the new value.



Two additional tests each will be run at a constant humidity of about 50 % with varying concentrations of methyl bromide. Levels will be changed hourly in a linear sequence between 600 ppb and 100 ppb, and vice versa. Even though methyl bromide concentrations after an application generally decrease over tine, this last set with increasing methyl bromide concentrations will be run to check conditions where changes in wind direction can cause a low level to be followed by a higher concentration.



Effects of sunlight/temperature



A single, initial test indicated that analysis results are lower for samples taken during daylight hours than for samples taken at night. At least four additional tests will be made with the gas mixing system located outdoors, trying to catch two sunny and two overcast days. Each test will cover a 24-hour period broken down into four six-hour sampling intervals (two day, two night) at a constant methyl bromide level of 200 ppb and a constant relative humidity of 50 %. Duplicate samples will be taken during each test.



Considering that we are close to the winter solstice, sunlight intensity and outdoor temperatures tend to be more mild during the next few months. Data obtained now will not be representative for meteorological conditions in the summer.



Other tests



If modifications to sampling and/or analysis methods are suggested that promise improved performance, reevaluation of the new procedures may be necessary. Based on the results from the above measurements, the best sampling/analysis approach will be selected. Additional tests will be needed to fully validate the selected method.