Air-Surface Exchange of Persistent Substances •by •Michael McLachlan •ITM, Stockholm University •michael.mclachlan@itm.su.se •for the summer school •The Advances and Trends in Environmental Chemistry and Ecotoxicology of Persistent, Toxic Substances •held in Brno, 11-17 July 2005 Phases in the Atmosphere •What phases do organic chemicals associate with in the atmosphere? • Phases in the Atmosphere •What phases do organic chemicals associate with in the atmosphere? •Gas phase •Particulate matter •Water •Ice/Snow Surfaces in the Environment •Which surfaces are important for exchange of organic chemicals with the atmosphere? • Surfaces in the Environment •Which surfaces are important for exchange of organic chemicals with the atmosphere? •Water •Soil •Vegetation •Snow/Ice Exchange Processes •What processes can transfer organic chemicals from the atmosphere to surfaces? Exchange Processes •What processes can transfer organic chemicals from the atmosphere to surfaces? •Deposition of water (wet deposition of dissolved chemical) •Deposition of ice/snow •Wet deposition of particulate matter •Dry deposition of particulate matter •Gaseous deposition Exchange Processes (II) •What processes can transfer organic chemicals from surfaces to the atmosphere? Exchange Processes (II) •What processes can transfer organic chemicals from surfaces to the atmosphere? •Aerosol generation •Volatilisation Phase Distribution in the Atmosphere •Between the gas phase and water Phase Distribution in the Atmosphere •Between the gas phase and water •according to the Henry’s Law constant (KAW=H/RT) •Dependent on temperature •Equilibrium generally reached, but perhaps not locally •Surface adsorption can contribute to levels in very small water droplets (fog) Phase Distribution in the Atmosphere (II) •Between the gas phase and particulate matter Phase Distribution in the Atmosphere (II) •Between the gas phase and particulate matter •Combination of dissolution, surface adsorption, and absorption in organic matter •Absorption believed to dominate for POPs, described by KPA = fOCKOC/KAW •Temperature dependent •Equilibration believed to be rapid, but not much experimental evidence Phase Distribution in the Atmosphere (III) •Between the gas phase and snow Phase Distribution in the Atmosphere (III) •Between the gas phase and snow •Surface adsorption •Dependent on surface area of snow •Little experimental evidence on magnitude and kinetics of partitioning Deposition of Rain and Snow •Dissolved/sorbed chemical Deposition of Rain and Snow •Dissolved/sorbed chemical •Surface independent (exception fog) •Flux is the product of concentration in rain/snow and the rain rate e.g. NRD (mol m-2 h-1) = vRCRD = vRCG/KAW •What is the relevant temperature? •Steady state may be a good assumption for compounds with high KAW Deposition of Rain and Snow •Chemical associated with particulate matter Deposition of Rain and Snow •Chemical associated with particulate matter •Surface independent •Flux is the product of concentration on PM, the rain rate, and the scavenging efficiency Q e.g. NRP (mol m-2 h-1) = vRQCP = vRQCGKPA •Q depends on particle size, cloud physics •Steady state not a good assumption! •”bulk” Q often used (20,000 – 200,000) Dry Deposition of Particulate Matter • Dry Deposition of Particulate Matter •Surface dependent! •Flux is the product of concentration on PM and a dry particle deposition velocity vP e.g. NP (mol m-2 h-1) = vPCP = vPCGKPA •vP depends on particle size, surface, turbulence •Mechanisms: diffusion, impaction, sedimentation •CP also depends on particle size •Steady state does not exist but may be a good approximation over long time scales Gaseous Deposition • • Gaseous Deposition •Surface dependent! •Thermodynamically governed phase distribution/partitioning •Flux can be calculated from Fick’s Law using the concentration gradient between the gas phase and surface and a dry gaseous deposition velocity vG e.g. NG (mol m-2 h-1) = vG(CG – CS/KSA) • Gaseous Deposition •Partition Coefficients Gaseous Deposition •Partition Coefficients •For water from Henrys Law Constant, affected by POM, DOM, temp., salinity •For soil from KOC/KAW for neutral species, KOC can be predicted from KOW, affected by quality of OM, temp., relative humidity, concentration (non-linearity!) •For vegetation few data, non-linear function of KOA, function of species, age, temp. Gaseous Deposition •Deposition Velocity Gaseous Deposition •Deposition Velocity •Two film resistance model •Air-side resistance depends on meteorology, surface roughness, diffusion coefficient in air •Surface-side resistance to major reservoir - water: surface film, turbulence, diff coeff. - soil: diffusion in soil air and soil water, vertical sorbed phase transport -vegetation: diffusion through cuticle, stomata •Dominant resistance depends on partition coeff. Part II: An Example using Field Data •Continuous monitoring of atmospheric deposition over 12 months •Sampling with glass jars, collection every month •One set of jars in a beech/oak forest, one in a clearing Air Sampler Continuous sampling: Coniferous canopy: 2 yr Deciduous canopy: 1 yr Sampling Characteristics of the Sampler •Do the glass jars collect the different forms of deposition in the clearing a representative manner? Sampling Characteristics of the Sampler •Do the glass jars collect the different forms of deposition in the clearing a representative manner? •Deposition of rain/snow: Yes for both dissolved and particle associated •Dry deposition of particulate matter: Yes for sedimentation, no for impaction and diffusion •Dry gaseous deposition: No Sampling Characteristics of the Sampler II •Do the glass jars collect the different forms of deposition in the forest a representative manner? Sampling Characteristics of the Sampler II •Do the glass jars collect the different forms of deposition in the forest a representative manner? •Deposition of rain/snow: Yes for both dissolved and particle associated •Dry deposition of particulate matter: Yes for sedimentation, yes for impaction and diffusion to the canopy, no for impaction and diffusion to the forest floor •Dry gaseous deposition: Yes for the canopy, no for the forest floor Sampling Characteristics of the Sampler III •In summary, the difference between the samplers in the forest and in the clearing represents the gaseous deposition and the dry particle bound deposition via diffusion and impaction/interception (i.e. of small particles) to the canopy. •We call this difference the “interception”. Comparison of the Bulk Deposition of PCDD/Fs to a Clearing and Two Forest Canopies during 1 Year Estimate Dry Particle Bound Deposition Velocity of Chemical to Canopy vP Divide canopy interception (I) by average particle bound concentration in the atmosphere and exposure time (6 months). vP = NP/CP = IP/CPΔt Similar values for most involatile PAHs (vP = 0.0073 m/s) Higher values for more volatile PAHs indicates other deposition mechanism (I > IP) Estimate Gaseous Deposition Velocity of Chemical to Canopy vG Divide canopy gaseous interception (IG = I - IP) by average gaseous conc. in the atmosphere and exposure time. vG = NG/CG = IG/CGΔt Similar values for PCDDs, involatile PCBs vG = 0.035 m/s (compare to <0.001 to soil) Higher values for more volatile PCBs indicates deposition limited (equilibrium reached) Estimate Partition Coefficients (KVA) of the Chemicals between Canopy and Atmosphere Divide canopy gaseous interception by gaseous conc. in autumn and specific canopy volume (VV). KVA = CV/CG = IG/VVCG Yields a correlation between KVA and KOA KVA = 14 KOA0.76 >> > Forests are efficient traps of gaseous organic chemicals. Forest Filter Effect plotted as a Function of the Octanol-Air (KOA) and Air-Water Partition (KAW) coefficient of the Chemical Horstmann and McLachlan (1998) Atmos. Environ. 32, 1799-1809 McLachlan and Horstmann (1998) Environ. Sci. Technol. 32, 413-420 Applying Process Understanding in Monitoring •How would you monitor atmospheric deposition of PCBs and PCDD/Fs? •For higher chlorinated PCDD/Fs, wet particle bound deposition dominates (a bulk deposition sampler in a field may be adequate) •For PCBs, gaseous deposition must be quantified (a bulk deposition sampler in a clearing will not work. Perhaps in a forest ...)