Modifications of the transition flow reactor theory for the more accurate measurement of nitric acid and nitrate and ammonium ions.

Sixty five samples were taken with a Transition Flow Reactor (TFR) for the measurement of nitric acid (HNO3) and nitrate particulates (NO 3 (-) ), in a central Athens street having a heavy traffic density, between February 1989 and February 1990. It was established that the TFR theory overestimated HNO3 and underestimated NO 3 (-) . This is caused by the retention of a fraction of the NO2 and a fraction of the HNO2 as NO 2 (-) on the sampler's nylon linear. NO 2 (-) is in turn oxidized to NO 3 (-) and this is determined as HNO3. Two modifications of the TFR theory are proposed for a better estimation of the HNO3 and NO 3 (-) . With the first modification HNO3 is estimated as being equal to the smaller of two values: HNO3 TFR and the value calculated if it is considered that 91% of HNO3 is retained by the nylon filter. The second modification was based on the quantification of the effect of NO2 on the concentrations of HNO3 and NO 3 (-) . This modification is an improvement on the first one and enables the calculation of a lower limit for the concentration of HNO2. Both modifications improve the accuracy of the TFR sampler in measuring HNO3 and NO 3 (-) , without altering it structurally. It was also found that the nylon filter partially retains some NH3, causing a negative error in the measurement of NH 4 (+) with the TFR. The extent of NH3 retention on the nylon filter is proportional to the HNO3 concentration. This error must be taken into account whenever NH 4 (+) is being determined.