Investigation of errors associated with the use of cascade impactors

The magnitude of some errors associated with the use of cascade impactors for atmospheric aerosol sampling has been estimated by a number of experiments in the field. Three types of impactors namely the Andersen Ambient, the Battelle and the Lundgren have been compared. As impaction surfaces Whatman 41 cellulose filter and Fresh-Pack polyethylene were applied, with and without adhesive coating. The collection efficiencies were evaluated on the basis of concentration data obtained by INAA for 10 elements. It appears that for sampling times of at least 95 hours the wall-losses and the bounce-off effects in the Andersen are less influenced than in the Battelle by changing the impaction surface or by coating it with Apiezon. The Andersen yields very reproducible mass-size functions. This impactor is also very insensitive to its positioning upright or downwards or to the addition of a specially modified inlet. It was reassuring to note that the Andersen and Battelle impactor collect, within 10%, the same mass and yield very similar MMD values. With the Lundgren impactor slightly deviating results were obtained.