The investigation of harmful semi volatile organic compounds (SVOC) in the indoor environment is important because on average people spend over 90% of their time indoors. Lindane, an SVOC which was widely used in the UK until 2004, adsorbs to house dust. House dust acts as a reservoir for such contaminants which are remitted by desorption into the air over time. A method for measuring Lindane air concentrations in a vial using SPME without the use of water or any solvent was developed in order to carry out Lindane adsorption and desorption tests. Dynamic tests were carried out to determine adsorption and desorption coefficients as well as equilibrium time. Adsorption and desorption constants (k1 and k2 respectively) were determined by fitting results from the dynamic adsorption tests to an existing two compartment model described in chapter 2, using the statistical analysis software SPSS (vs16). These dynamic tests were carried out for two size fractions (<20µm and >45µm<63µm) and whole dust samples to determine the effect of size fraction on adsorption. For the >45µm to <63µm, k1 = 0.568h-1 and k2 = 0.047h-1, (standard error 0.119 and 0.030 respectively), for the <20µm fraction, k1 = 1.686h-1, k2 = 0.125h-1 (standard error 1.888 and 0.324 respectively), and the whole dust k1= 2.587 h-1, k2= 0.288 h-1 (standard error 0.514 and 0.113 respectively). Static tests were carried out at equilibrium to establish an adsorption isotherm and obtain partition coefficients for different size fractions. The adsorption constants Ka were 4.2 x 10-4mh-1, 7.67 x 10-5mh-1, and 3.03 x 10-3mh-1 respectively. The desorption constants Kd were 0.125h-1, 0.047h-1, 0.288h-1. The partition coefficients Kp were 4.8 x 101µgm-2, 4.08 x 101µgm-2, 1.05 x 102µgm-2, for the <20µm, >45µm<63µm, and whole dust respectively. The higher Kp value for the smaller <20µm fraction compared to the >45µm<63µm fraction, suggests that Lindane adsorbs more strongly to smaller dust size particles. This is significant because it means that the inhalable dust fractions which fall within the <20µm fraction, will have higher concentrations and therefore could potentially be more harmful as the get into the lungs. A possible explanation for the higher Ka value for the whole dust fraction over the two other smaller fractions could be because whole dust is a more complex mixture containing more fibrous substances that may have stronger affinities for Lindane than dust e.g. carpet fibres.

A three-compartment model is extended to estimate the fate and transport of DEHP in a realistic residential environment. The model considered eight environmental media (i.e. air, particulate matter with six size fractions, vinyl flooring, carpet, furniture, dust, wall and ceiling). Particle movement (deposition and resuspension), dust removal (vacuuming), indoor cooking, and adsorption/absorption on indoor surfaces are included. The predicted airborne DEHP concentrations at steady state are within 0.1 [microgram]/m³ to 0.6 [microgram]/m³, which are similar to those measured in field studies. After vinyl flooring (the primary source) is removed, it takes 2 years for the indoor airborne DEHP level to reduce 0.01 [microgram]/m³, and the time increases significantly when carpet present. The results indicate that carpets as well as other interior surfaces may be important phthalate sinks and if the only removal mechanism is ventilation, strongly sorbing phthalate may persist for years. Phthalate amount in dust is strongly influenced by the deposition surface. The concentration of DEHP presents 10 times higher in dust on the source (vinyl flooring) than on the sink (furniture), and it takes more than a year for DEHP to reach equilibrium between bulk air and dust. The domestic activity of cooking is then included in the model and it shows that suspended particle concentration has a substantial impact on gas-phase DEHP level indoors, while the influence of ventilation is only to some extent. Three other SVOCs (DMP, BBP and DiDP) are also investigated and their environmental fates show that chemical’s vapour pressure and octanol/air partition coefficient have substantial influences on sorbing mechanisms and the gas phase and airborne concentrations. / text

If sufficiently persistent, semi-volatile organic contaminants (SVOCs) can travel long distances through the atmosphere from their points of release and become concentrated in cold, remote regions. As air is sampled for SVOCs to establish both their presence and the success of emission reduction efforts, it becomes helpful to determine sampling site proximity to sources and the origin of the sampled air masses. Comparing three increasingly sophisticated methods for quantifying source proximity of sampling locations, it was judged necessary to account for the actual history of the sampled air through construction of an airshed, especially if wind is highly directional and population distribution is very non-uniform. The airshed concept was improved upon by introducing a ‘geodesic’ grid of equally spaced cells, rather than a simple latitude/longitude grid, to avoid distortion near Earth’s poles and to allow for the comparison of airshed shapes. Assuming that a perfectly round airshed reveals no information about sources allows the significance of each cell of an airshed to be judged based on its departure from roundness. Combining air-mass histories with a 2 year-long series of SVOC air concentrations at Little Fox Lake in Canada’s Yukon Territory did not identify distinct source regions for most analytes, although γ-hexachlorocyclohexane appears to originate broadly in north-eastern Russia and/or Alaska. Based on this remoteness from sources, the site is judged to be well suited to monitor changes in the hemispheric background concentrations of SVOCs. A model-based exploration revealed wet-gaseous deposition as the dominant process responsible for cold-trapping SVOCs in mountain soils. Such cold trapping is particularly effective if precipitation rate increases with altitude and if temperature differences along the mountain are large. Considerable sensitivity of the modeled extent of cold-trapping to parameters as diverse as scale, mean temperature, atmospheric particle concentration and time relative to emission maxima is consistent with the wide variety of observed enrichment behaviour. Concentration gradients of polycyclic aromatic hydrocarbons and polychlorinated biphenyls in air and soil measured on four Western Canadian mountains with variable distance from sources revealed source proximity as the main driver of concentrations at both the whole-mountain scale and along individual mountain transects.

If sufficiently persistent, semi-volatile organic contaminants (SVOCs) can travel long distances through the atmosphere from their points of release and become concentrated in cold, remote regions. As air is sampled for SVOCs to establish both their presence and the success of emission reduction efforts, it becomes helpful to determine sampling site proximity to sources and the origin of the sampled air masses. Comparing three increasingly sophisticated methods for quantifying source proximity of sampling locations, it was judged necessary to account for the actual history of the sampled air through construction of an airshed, especially if wind is highly directional and population distribution is very non-uniform. The airshed concept was improved upon by introducing a ‘geodesic’ grid of equally spaced cells, rather than a simple latitude/longitude grid, to avoid distortion near Earth’s poles and to allow for the comparison of airshed shapes. Assuming that a perfectly round airshed reveals no information about sources allows the significance of each cell of an airshed to be judged based on its departure from roundness. Combining air-mass histories with a 2 year-long series of SVOC air concentrations at Little Fox Lake in Canada’s Yukon Territory did not identify distinct source regions for most analytes, although γ-hexachlorocyclohexane appears to originate broadly in north-eastern Russia and/or Alaska. Based on this remoteness from sources, the site is judged to be well suited to monitor changes in the hemispheric background concentrations of SVOCs. A model-based exploration revealed wet-gaseous deposition as the dominant process responsible for cold-trapping SVOCs in mountain soils. Such cold trapping is particularly effective if precipitation rate increases with altitude and if temperature differences along the mountain are large. Considerable sensitivity of the modeled extent of cold-trapping to parameters as diverse as scale, mean temperature, atmospheric particle concentration and time relative to emission maxima is consistent with the wide variety of observed enrichment behaviour. Concentration gradients of polycyclic aromatic hydrocarbons and polychlorinated biphenyls in air and soil measured on four Western Canadian mountains with variable distance from sources revealed source proximity as the main driver of concentrations at both the whole-mountain scale and along individual mountain transects.

La miniaturisation et la complexification croissante des composants microélectroniques induit une sensibilisation et une fragilisation accrue des composants vis-à-vis des contaminations présentes dans les zones de productions appelées “salles blanches”. Dans ces espaces, le contrôle actuel de la contamination organique n’est pas suffisant puisqu’il ne permet pas d’éviter la contamination de surface des plaquettes de silicium et des optiques des robots de production utilisés pour la photolithographie. Un contrôle accru des concentrations des contaminants organiques dans les atmosphères des salles blanches devient donc nécessaire et de nouvelles méthodes analytiques doivent être développées et validées. Dans le cadre de ce travail, des méthodes d’analyse ont été développées et validées afin de disposer d’une gamme d’outils permettant un suivi rigoureux des contaminations. Ces outils permettent d’identifier et de quantifier les contaminations surfaciques des plaquettes de silicium par des composés organiques semi-volatils (phtalates et organophosphorés) mais aussi de déterminer les concentrations de composés organiques volatils présents dans l’atmosphère des salles blanches. Ces méthodes font appel aux technologies du WOS/ATD-GC-MS « Wafer Outgassing System/Automated Thermal Desorber–Gas Chromatography–Mass Spectrometry » et de la DART-ToF-MS « Direct Analysis in Real Time-Time of Flight–Mass Spectrometry » pour les analyses de surfaces et au PTR-ToF-MS « Proton Transfer Reaction – Time of Flight - Mass Spectrometry » pour l’analyse de l’atmosphère. / The recent advances in the miniaturization and complexification of microelectronic components induce an increase in the sensitivity of these components regarding the organic contamination present in the production zone called “clean room”. Although, the control of organic contamination in the clean room is very rigorous it does not avoid the contamination of silicon wafer surfaces and robot lenses used in the photolithography process. The later implies that new analytical methodologies should be developed and validated. In this work, analytical methods were developed and validated in order to have a panel of tools which allows careful monitoring of organic contaminants. These tools allow the identification and quantitation of the contamination of silicon wafer surface by semi-volatiles organic compounds (phthalates and organophosphates) and the determination of volatile organic compounds concentrations in the clean room atmosphere. These methods uses the WOS/ATD-GC-MS « Wafer Outgassing System/Automated Thermal Desorber–Gas Chromatography–Mass Spectrometry » technology and the DART-ToF-MS « Direct Analysis in Real Time-Time of Flight–Mass Spectrometry » technology for wafer surface analysis and the PTR-ToF-MS « Proton Transfer Reaction – Time of Flight - Mass Spectrometry » technology for gas-phase analysis.