Comparison of measurement methods for benzene and toluene

Diffusive sampling and active (pumped) sampling (tubes filled with Tenax TA or Carbopack B) were compared with an automatic BTX instrument (Chrompack, GC/FID) for measurements of benzene and toluene. The measurements were made during differing pollution levels and different weather conditions at a roof-top site and in a densely trafficked street canyon in Stockholm, Sweden. The BTX instrument was used as the reference method for comparison with the other methods. Considering all data the Perkin–Elmer diffusive samplers, containing Tenax TA and assuming a constant uptake rate of 0.406 cm 3 min −1 , showed about 30% higher benzene values compared to the BTX instrument. This discrepancy may be explained by a dose-dependent uptake rate with higher uptake rates at lower dose as suggested by laboratory experiments presented in the literature. After correction by applying the relationship between uptake rate and dose as suggested by Roche et al. (Atmos. Environ. 33 (1999) 1905), the two methods agreed almost perfectly. For toluene there was much better agreement between the two methods. No sign of a dose-dependent uptake could be seen. The mean concentrations and 95% confidence intervals of all toluene measurements (67 values) were (10.80±1.6) μg m −3 for diffusive sampling and (11.3±1.6) μg m −3 for the BTX instrument, respectively. The overall ratio between the concentrations obtained using diffusive sampling and the BTX instrument was 0.91±0.07 (95% confidence interval). Tenax TA was found to be equal to Carbopack B for measuring benzene and toluene in this concentration range, although it has been proposed not to be optimal for benzene. There was also good agreement between the active samplers and the BTX instrument.