Current situation of polycyclic aromatic hydrocarbons (PAH) in PM2.5 in a receptor site in Mexico City and estimation of carcinogenic PAH by combining non-real-time and real-time measurement techniques

[Display omitted] •PAH in PM2.5 at a receptor site of Mexico City were determined.•A decrease around 40% in carcinogenic PAH was observed in 2016–2017, with respect to 2006. CO, NO and NO2 also showed a negative trend.•Non-real-time and real-time techniques were combined to estimate PAH in PM2.5 at low-cost.•A better air quality is probably attributed to strategies implemented by local and federal governments.•PAH diagnostic ratios are useful tools to identify the emission sources and the photo-oxidation processes. Air pollution is a public health concern. Polycyclic aromatic hydrocarbons (PAH) are ubiquitous atmospheric pollutants contained in the atmospheric aerosol. PAH in particulate matter with diameters ≤2.5 µm (PM2.5) represent a human health risk due to their toxic properties. In this study, PAH in PM2.5 at a receptor site of Mexico City during the dry cold season were determined. The most abundant PAH (median, 10–90th percentile, pg m−3) were benzo[ghi]perylene (467, 291–697), followed by pyrene (427, 218–642). A decrease around 40% in the carcinogenic PAH onto PM2.5 was calculated with respect to the same PAH measured a decade ago, at the same receptor site, despite of increase in vehicle fleet. The PAH decrease trend agrees with the decrease trend of CO, NO and NO2, released into the air by similar emission sources than PAH. Control emissions strategies implemented by local and federal authorities are discussed. PAH analyses were carried out by non-real-time and real-time methods. The PAH non-real-time method involved PM2.5 sampling, sample treatment and gas chromatography-mass spectrometry analysis. The PAH real-time method involved the use of a photoelectric aerosol sensor (PAS). The PAH determination by non-real time method was selective and efficient, with recoveries between 75 ± 14% and 98 ± 26%. By combining non-real-time and real-time methodologies, multivariate regression models were obtained based on PAS response, NO2 and wind speed to estimate PAH in PM2.5 at low-cost (r2 = 0.59 to r2 = 0.89). Fossil fuel combustion from vehicles was the major source around the sampling site. Diagnostic ratios (DR) based on retene, chrysene, and triphenylene, suggested biomass burning emission sources. Photo-oxidation in sunny months was observed based on benzo[a]pyrene, benzo[ghi]perylene, benz[a]anthracene, indeno[1,2,3-cd]pyrene and black carbon. The correlation analyses suggested transport of PM2.5, O3, BC and SO2 to the sampling site, and local em