Information on phase distribution of airborne semivolatile organic compounds (SVOCs) is important in occupational health because of the difference in the intake and uptake mechanisms of the two phases. Particle deposition in the lungs is a function of aerodynamic diameter, whereas gas deposition is a function of tissue solubility, which is related to the air-lung partitioning ratio.⁽¹⁾ In the case of semi-volatile chemicals, mass distributes between the two phases according to a gas-particle partitioning ratio.

Many chemicals tend to be distributed between the particle and vapor phases in workplaces. Since 2000, the American Conference of Governmental Industrial Hygienists (ACGIH^®) has added an IFV (inhalable fraction and vapor) notation for chemicals that need to be evaluated for both inhalable fraction and vapor due to their volatility. The Threshold Limit Value (TLV^®) list for 2012 includes 55 chemicals having this notation not including the 4 chemicals on the Notice of Intended Change. Most of them can be classified as semivolatile chemicals and a few as volatile chemicals, according to their vapor pressures. Pesticides make up over 70% of TLVs having an IFV notation.⁽²⁾ Control methods for each phase including personal protection equipment are fundamentally different. Filtration for particles and adsorption for vapor molecules has been commonly considered as an economic control method, but it has been less than successful in controlling SVOCs.

Agricultural workers are at risk from exposure to pesticides when working in different crop categories. The Department of Labor estimates that approximately 1.8 million workers perform hired agricultural crop work in the United States.⁽³⁾The National Center for Food and Agricultural Policy (NCFAP) reported pesticide use in U.S. crop production for 1997 by their active ingredients.⁽⁴⁾The amount of semivolatile insecticides (based on the vapor pressures of active ingredients) used in the United Sates was estimated to be ca.50×10⁶ pounds of active ingredients and over 95 percent of the semivolatile insecticides had TLVs. The upper and lower cut-off of vapor pressures for SVOC were 10⁻⁴atm and 10⁻¹¹atm, respectively.⁽⁵⁾ This report has not been updated since 1997.

Almost all pesticides having Occupational Safety and Health Administration (OSHA) Permissible Exposure Limits (PELs) are covered by National Institute for Occupational Safety and Health (NIOSH) Sampling and Analytical Methods #5600 for Organophosphorus Pesticides and # 5601 for Organonitrogen Pesticides, which list 19 and 15 chemicals, respectively, under each method. All of those chemicals, except one solid pesticide, belong to SVOCs according to their vapor pressures reported in the methods or in their Material Safety Data Sheets (MSDSs).^(6,7) Most TLVs or PELs for organophosphorus pesticides are lower than 1.0 mg/m³, indicating a potentially high toxicity.

Inhalation exposure of pesticides is important even considering that other routes of exposure are often considered to contribute up to 80% of total pesticide exposure.⁽⁸⁾ Durham and Wolfe⁽⁹⁾ suggested the “more rapid and more complete absorption of toxicant drawn into the lungs as compared with that deposited on the skin.” In many studies the absorption of the respiratory dose has been considered as 100%, while that of the dermal dose has been considered 10% or lower.^(9–12) In some special environments such as the inside of greenhouses or storage buildings where mixing and loading happen, inhalation exposure can be more important due to greater confinement compared to the outdoor situation. Aprea et al.⁽¹³⁾ reported that depending on their tasks the respiratory dose of workers during industrial formulation was several times higher than skin contamination. Dowling and Seiber⁽⁸⁾ emphasized the importance of respiratory exposure to pesticides among agricultural populations. Their study identified some high risk activities of application of high volatility soil fumigants, where respiratory exposures are important, including handling volatile liquid pesticides for mixers and loaders, elevated track boom application, and greenhouse applications. In the National Agricultural Workers Survey,⁽¹⁴⁾ depending on the size of farms, 4.5–25.1% of farmworkers whose primary occupation is working in the field on tasks like harvesting and weeding have experienced loading, mixing, or application of pesticides in the last 12 months.

Few devices are available to evaluate the phase distributions of semivolatile chemicals. Several different types of sampling devices are available to evaluate concentrations of aerosols. Popular sampling methods for chemicals in the particle phase include impingers, filters, electrostatic precipitators (ESPs), impactors, and virtual impactors. Many of these methods have been shown to be subject to evaporative losses when used to collect SVOCs.^(1,15,16) Depending on the principles of the sampling mechanism, some methods are more subject to evaporative loss than others. Several different types of sampling devices are available to evaluate concentrations of chemicals in vapor form. Sorbent tubes are most widely used to collect chemical vapors and will also sample some, but not all, particles. Devices sampling only one phase are therefore not appropriate for measuring semivolatile aerosols due to the characteristics of SVOCs described previously. Portability of sampling instruments is also a consideration for personal exposure assessment.

Kim and Raynor⁽¹⁷⁾ developed a Semivolatile Aerosol Dichotomous Sampler (SADS)). The SADS works as a preseparator similar to cyclones. The SADS is a round nozzle virtual impactor having an inverted flow ratio, which changes the airflow through the sampler and the trajectories of some particles relative to the traditional virtual impactor flow settings. The 50% cutsize of the SADS was optimized in previous experiments at 0.25 μm aerodynamic equivalent diameter when the total inlet flow rate was set to 2.0 l.min⁻¹ and 0.2 l.min⁻¹ for the vapor flow (flow fraction 0.1).⁽¹⁸⁾Any collection media designed for active air movement can be utilized to sample the SVOCs of interest. SADS was tested in laboratory settings for oil mists and worked well.⁽¹⁹⁾Separation of particles from aerosols using SADS minimizes evaporative loss during sampling and the apparatus is less affected by temporal variation of the phase distributions. The SADS was designed for personal sampling and optimized to accommodate the pressure drops typically encountered with sampling media. The SADS also does not require electrical power to separate particles and can be easily utilized for personal sampling. This feature also lowers the cost for fabrication.

Many semivolatile pesticides can be sampled with OSHA versatile sampler (OVS) tubes.^(20,21) OVS tubes are not subject to evaporative loss during sampling because of the presence of sorbent media after the filter. Thus the structure of the OVS tube allows for simultaneous collection of aerosol and vapors. However, even if the media were analyzed separately, it is unlikely that accurate phase information can be obtained using this method and thus the media is not separated for analysis. Note that while OSHA requires glass fiber filters, NIOSH requires quartz filters in their similar procedure. A thorough evaluation of the particle size-selection efficiency of the OVS tube has not been published.

Table I is a summary of semivolatile pesticides exposure studies performed in greenhouses. Few studies have obtained data on phase distribution. Siebers and Mattusch ⁽²³⁾ did some additional sampling using glass fiber filters in front of the Tenax tubes. While parathion was mainly detected in the Tenax tubes (>95%), pirimicarb was found at an average of 65% in the filter and 35% in the Tenax tube. It seems the vapor pressures of two active ingredients, 5.0×10⁻⁸atm@20°C and 3.0 × 10⁻⁸atm@20°C, respectively, cannot explain these results thoroughly.

The overall objective of this pilot study was to assess exposures to an airborne semivolatile pesticide (chlorpyrifos) in greenhouses according to its phase profile. The specific aim of this study was to compare available methods for selection of an appropriate method that would determine phase information to evaluate phase distributions of airborne chlorpyrifos in greenhouses. A secondary aim, to survey the factors affecting phase distributions in greenhouses, will require further investigations, but some discussion of these factors is included.

Static area sampling was performed inside of greenhouses. Sampling sites were two greenhouses in research facilities located in Minnesota and Maryland, which are noted as MN and MD in Table II. The square footage areas of MN and MD were 750 ft² and 2,990 ft², respectively. Air samples were collected during summer 2010 and 2011. Plants were growing in pots and located two feet above the ground in site MN and on the ground in site MD. The height of plants was on average 4 feet and 6 feet in site MN and site MD, respectively. The distance between plant pots was approximately 1 foot.

For each site, samples were collected for three time slots: T1, the first 1 or 2 hr after the commencement of application; T2, a 6-hr period immediately following T1; and T3, a 6-hr period after the required re-entry interval (24 hr for chlorpyrifos). T1 was a duration of 1 hr in 2010. After finding the amounts of chlorpyrifos collected were below limit of detection (LOD), T1 was increased to 2 hr in 2011. The duration of T2 and T3 was 6 hr in all visits (Figure 1).

Airborne chlorpyrifos concentrations were measured by three different methods: SADS, electrostatic precipitator (ESP), and NIOSH standard method (OSHA versatile sampler (OVS) tube containing Amberlite XAD-2 sorbent resin in combination with polyurethane foam and a pre-filter). Five optimized SADS were machined from aluminum and used throughout this study. XAD-2 OVS sorbent tubes (226–58A, SKC Inc., Eighty Four, Pa) were attached to the two outlets of the SADS to act as collection media for the vapor flow and the particle flow. The SADS mounted with sampling media was connected to a vacuum pump (Gast, Benton Harbor, Mich.). The flow rates for the vapor and particle outlets were 0.2 and 1.8 l.min⁻¹, respectively, to give a total flow rate of 2.0 l.min⁻¹ and a vapor flow fraction of 0.1 of the total, per previous studies.^(18,19) Flow rates were maintained by two mass flow controllers (GFC17, Aalborg Instruments & Controls, Inc., Orangeburg, N.Y.).

Another sampling train was the ESP with a XAD-2 OVS tube. An ESP (Aerosol Associates, Chapel Hill, N.C.) developed for personal sampling, which is reported to have less evaporative loss than filtration methods, was used to sample both phases.⁽¹⁶⁾ A rectangular sheet of aluminum foil measuring 4 cm by 4 cm, was coiled inside the aluminum tube provided by the manufacturer to act as the substrate for particulate collection. A XAD-2 OVS tube was connected in series to the outlet of the ESP to collect vapor and any uncollected particulate. This sampling train was connected to the same vacuum pump to which the SADS was connected. The flow rate for the ESP was 1.0 l.min⁻¹ and was controlled by a mass flow controller (GFC17, Aalborg Instruments & Controls, Inc., Orangeburg, N.Y.).

XAD-2 OVS tubes (226–58A, SKC Inc., Eighty Four, Pa.) also were used as a control to sample the total pesticide (vapor plus aerosol) independently. The sampling flow rate was 1.0 l.min⁻¹ and was controlled by a mass flow controller (GFC17, Aalborg Instruments & Controls, Inc., Orangeburg, N.Y.).

An Aerosol Particle Sizer (Model 3321, TSI Inc., Shoreview, Minn.) was used to monitor the number concentration and the size distribution of particles. The APS provides real-time aerodynamic measurements of particles between 0.5 μm and 20 μm aerodynamic equivalent diameter and in this range the size distribution was evaluated as log-normal. Temperature and humidity inside greenhouses were recorded using a Temperature Humidity USB Data Logger (THL1, Universal Enterprise Inc., Beaverton, Ore.) during sampling. The APS data and temperature/humidity were automatically logged per minute.

Mass flow controllers for three samplers and the APS were placed inside a purpose-built chamber sitting on top of a cart to prevent the surface of the devices from being contaminated by pesticides during sampling. The chamber was built from plastic panels and an aluminum frame. The sampling inlet of the APS was open to the top of the sampling cart and the exhaust air was vented to the outside of the chamber. The vacuum pump was outside the chamber. The temperature in the chamber was monitored using the same temperature/humidity logger. Three samplers were mounted on a horizontal rig located next to the sampling cart and connected to the mass flow controllers using Tygon tubing and quick disconnect couplings.

Both ends of the XAD-2 OVS tubes were capped and the tubes were shipped to the NIOSH contract laboratory for analysis. The aluminum substrates mounted in the ESP were immediately separated from the ESP after each sampling period and were placed into 8.75 mm × 50 mm test tubes (with screw caps) containing extraction solution (90% toluene and 10% acetone). After capping tightly, the test tubes were agitated vertically by hand for 20 sec and were shipped along with XAD-2 OVS tubes to the contract laboratory for analysis.

The phase distribution and the concentrations of each phase of semi-volatile organic compounds (SVOCs) are important in establishing strategies of exposure assessment, hazard control, and worker protection. The mechanisms of the worker protection method are different for each phase and different phases can influence the respiratory toxicity of certain chemicals (e.g., upper respiratory versus lower respiratory effects). This study attempted to compare two sampling methods capable of determining phase information to evaluate methods for sampling chlorpyrifos in greenhouses that might provide more useful information than the typical collection of all phases without separation. The electrostatic precipitator (ESP) appeared to be affected by formation of chemical artifacts and was excluded as a potential candidate method for phase profile sampling in this situation. Although the SADS with XAD-2 OVS tube sample collection underestimated the overall concentrations compared to the independent XAD-2 OVS tubes (NIOSH #5600 method), it separated vapors and particles and enabled estimation of the phase distribution of the semivolatile chlorpyrifos, even though the accuracy of phase separation and recovery could not be verified in the absence of a standardized procedure. Particle size distribution and temperature / humidity are believed to be factors significantly affecting the phase distributions of semivolatile aerosols, but this could not be proven in this pilot study.