Artefacts in the sampling of nitrate studied in the INTERCOMP campaigns of EUROTRAC-AEROSOL

Sampling of aerosol-nitrate can be problematic because of evaporative loss of the semi-volatile ammonium nitrate or adsorption of nitric acid gas. Such artefacts, which depend on filter type and ambient conditions, are not well documented for the filters in use in Europe and this was the reason to study these in a series of intercomparison trials. The trials were performed within the INTERCOMP programme of the AEROSOL subproject of EUROTRAC-2. The major effort was a 2-week field campaign at the rural site of Melpitz, a village near Leipzig in eastern Germany (INTERCOMP2000). Samplers were used containing the most common filter types in use in Europe, i.e. quartz, Teflon, (mixed) cellulose ester and cellulose. The concentration of nitrate in PM2.5, mainly present as ammonium nitrate, was on average 3.3 mugm(-3). The variability in the concentrations stemming from the samplers appeared to be rather constant: +/- 0.5 mugm(-3) from the average of all samplers. The reason for the constant (but random) variability remains unexplained. Thus, the concentrations stemming from the samplers agreed very well at the average level with relative differences of 15% and less for higher concentrations. This is evidence that the influence of the mentioned artefacts was negligible. The absence is explained by extrapolation of results of tests on the artefacts in a laboratory setting (INTERCOMP99). It was found there that the loss of ammonium nitrate from Teflon and quartz filters is only substantial when temperatures are much higher than those during the field campaign. Cellulose and cellulose-acetate filters quantitatively collected both ammonium nitrate and nitric acid in the laboratory study, but in Melpitz measured nitric acid concentrations were too low to identify its adsorption. Possible artefacts due to adsorption of nitrous acid were negligible. We also used the laboratory information to evaluate the results of a further intercomparison (INTERCOMP98) in the Po-Valley, performed at much higher temperatures than at Melpitz. We found evidence of adsorption of nitric acid by cellulose filters and evaporational loss of aerosol-nitrate from quartz filters. For the conditions encountered during the campaign we parameterised the evaporational loss in a general way as a function of temperature, as follows. There is complete evaporation at temperatures exceeding 25degreesC and full retention at temperatures less than 20degreesC. At temperatures between 20 and 25degreesC the retention is on average 50%, but with high variability. A main conclusion from this study is that under central European conditions quartz is a suitable filter material for sampling nitrate as long as the temperature does not exceed 20degreesC during sampling. Cellulose-type filters quantitatively collect aerosol nitrate and nitric acid, but negligible amounts of nitrous acid. Teflon filters were more vulnerable for evaporation losses than quartz. Indications for losses from Teflon below 10degreesC (at Melpitz) were not obtained.(C) 2004 Elsevier Ltd. All rights reserved.