Introduction

The toxicity of atmospheric aerosols is influenced by various factors, such as their size distribution, origin, and chemical composition1. Particularly concerning is exposure to PM\(_{10}\), as it can penetrate the lungs and enter the bloodstream, potentially resulting in a range of adverse health effects2. PM\(_{10}\) can contain highly toxic substances like heavy metals, elemental carbon, and organic carbon. Of the species found in the organic carbon fraction, PAHs and NPAHs are of particular interest owing to their carcinogenic and mutagenic effects upon prolonged exposure3. Even more, the toxic effects of several NPAHs are significantly more pronounced than those of their parent PAHs4, with carcinogenicity and direct-acting mutagenicity up to 10 to 1000 times higher despite NPAHs typically having atmospheric concentrations about 2-3 orders of magnitude lower5. The International Agency for Research on Cancer6 has classified certain NPAHs as possible or probable human carcinogens. Specifically, IARC has categorized carcinogens into five groups: carcinogenic to humans (1), probably carcinogenic to humans (2A), possibly carcinogenic to humans (2B), not classifiable (3), and probably not carcinogenic (4)7. Some NPAHs have been assigned Toxicity Equivalence Factor (TEF) values higher than benzo[a]pyrene (BaP)8.

Accurately determining the chemical composition of atmospheric particles is key to estimating reliable health risks. In particular, determining PAHs and NPAHs levels in ambient air particulate matter is an analytical challenge primarily due to their low concentration levels9,10. Various analytical techniques have been employed to address this drawback, including high-performance liquid chromatography with ultraviolet, diode array, fluorescence, chemiluminescence, or mass spectrometry (MS) detection, and gas chromatography (GC) with electron impact (EI) ionization, negative chemical ionization (NCI), or atmospheric pressure chemical ionization (APCI), coupled with electron capture (ECD), single MS, or MS/MS detectors4. The need for high-precision instrumentation has limited the development of a comprehensive understanding of the level of NPAHs across different conditions around the world.

PAHs found in atmospheric aerosols are released during the combustion of carbon-based materials such as fossil fuels, biomass, and waste. A fraction is emitted in the gas phase, while the other is bound to the emitted PM. PAHs are also released from the volatilization (typically uncontrolled) of petroleum and its byproducts in the gas and oil industry and any other sector that uses these products11,12. PAHs with low volatility coefficients will eventually condense onto the airborne PM. On the other hand, besides being directly emitted from combustion processes, NPAHs can form through secondary processes involving the heterogeneous oxidation of PAHs. Key atmospheric reactions responsible for NPAHs formation are initiated by •OH radicals generated from ozone and water photolysis or by NO\(_{3}\) radicals produced from the ozonolysis of NO\(_{2}\), which is produced by NO ozonolysis. Studies indicate that NPAHs can also form from the reaction of particle-bound PAHs with gaseous oxidants like O\(_{3}\), NO\(_{2}\), and N\(_{2}\)O\(_{5}\). Thus, gas-phase and heterogeneous transformations contribute to NPAHs formation in the atmosphere and will be relevant in urban centers with high levels of photochemical oxidants4,13,14.

Important efforts have been made to understand the spatial and temporal distribution of NPAHs worldwide. However, most research on PM-bound PAHs and NPAHs has been conducted in urban areas of developed nations, highlighting the need for studies in developing countries to gain a broader understanding of the relative distribution of particle-bound compounds within the PM bulk and their global impact on the atmosphere15. In Mexico, the Mexico City Metropolitan Area (MCMA) is the conurbation that has historically received the most attention in addressing levels of NPAHs in ambient air or the contribution from specific emission sources16,17,18,19,20. The MMA, the second largest conurbation in the country, represents a relevant experimental setup. It is arguably the most polluted city in Latin America in terms of the measured concentrations of PM\(_{10}\) ambient air. In contrast to the MCMA, the MMA is experiencing increasing trends in oxidant levels (nitrogen oxides and O\(_{3}\)) despite efforts to control emissions in the airshed21. This represents a critical condition as reactivity levels foster the production of some toxic species like NPAHs.

Air pollution has become a significant environmental and public health concern in northern Mexico. This issue is primarily due to unchecked urban and industrial expansion22. The MMA, in the State of Nuevo Leon, with a population of over 5.6 million23, experiences poor air quality because of rapid urbanization, high industrial activity, a large number of vehicles, and mountainous terrain that hinders the dispersion of pollutants. The MMA is considered one of Latin America’s most polluted conurbations in terms of particulate matter. In the State of Chihuahua, the cities of CHI and CDJ have the third-highest cancer death rates in Mexico, with 8.2 deaths per 10,000 inhabitants, largely attributed to PM\(_{10}\) pollution24. With 878,500 residents and 540,910 vehicles, CHI relies on mining and metallurgy activities for its economy. In contrast, CDJ, a border city with the United States, with 1.4 million residents and 640,900 vehicles, faces emissions from mining, the textile industry, and terrestrial commercial activity between Mexico and the US. The CDJ-El Paso (Texas) conurbation represents the third most populous international metro area of the US, after Detroit-Windsor and Sand Diego-Tijuana. Despite research efforts in this area (e.g.,25), there is a lack of studies quantifying airborne NPAHs in these three Mexican cities.

This research focuses on several contributions to environmental management and public health policy. It is organized into three main sections: Analysis of Concentrations and Sources (i), Health Risk Assessment (ii), and Policy Recommendations (iii). Each of these is detailed below.

  1. 1.

    This study examines the concentrations of 16 PAHs and 11 NPAHs in PM\(_{10}\) particles in MMA, CHI, and CDJ during the winter of 2020. By using DRs and PCA, we explore similarities and differences in probable main emission sources of these pollutants.

  2. 2.

    The study evaluates the long-term health impacts by calculating the BaPeq and the ECR for lifetime exposure to PAHs and NPAHs. This assessment is crucial for understanding the potential public health consequences of atmospheric pollution in these regions.

  3. 3.

    Based on the conclusions, this research offers recommendations grounded in scientific evidence for developing public policies and regulatory measures to reduce PAHs and NPAHs contamination in ambient air. These recommendations are designed to support the creation of effective regulatory frameworks to protect public health comprehensively and sustainably.

In summary, this research significantly advances the understanding of PAHs and NPAHs pollution in northern Mexico. It includes detailed analysis, source identification, and health risk assessments, providing valuable insights into the environmental and health impacts of these pollutants. Additionally, the study offers policy recommendations to reduce air pollution and protect public health, thereby contributing to the development of effective regulatory frameworks for managing urban air quality.

Materials and methods

Study sites and sampling

The MMA, CHI, and CDJ were chosen for PM\(_{10}\) particle sampling in the northern region of Mexico due to significant geographical differences and characteristic pollution levels. Figure 1 shows the location of the sampling sites. At the same time, Table 1 provides a statistical summary of the meteorological parameters, including temperature (T), relative humidity (RH), solar radiation (SR), atmospheric pressure (P), wind speed (WS), and wind direction (WD), as well as the concentrations of some criteria pollutants such as SO\(_{2}\) , PM\(_{2.5}\) , CO, O\(_{3}\) , NO\(_{2}\) , NO, and NO\(_{x}\) . Simultaneous 24-hour sampling occurred every six days between January and March 2020, with 12 samples collected at each study site. Dry-cold conditions were selected as past studies elsewhere have reported higher concentrations of PAHs and NPAHs in winter and lower levels in summer, often attributed to meteorological conditions that contribute to higher combustion activities26. Similar patterns have been observed in cities across China27,28,29, Taiwan30, Spain31, France32, Portugal, Italy, and Greece13.

High-volume PM\(_{10}\) samplers (TE-310, TISH model) with quartz filters (Whatman QM-A) served for sample collection. The filters underwent calcination at 450\(^\circ\)C in a furnace for six hours to eliminate organic residues before each sampling. Subsequently, they equilibrated in a controlled humidity chamber (less than 50%) and at a T between 20 and 25\(^\circ\)C for 24 hours, following the USEPA IO 2.1 guidelines33. After sampling, the filters were re-equilibrated, wrapped in aluminum foil, and stored at -20\(^\circ\)C until further analysis. The sample mass, measured using a Mettler Toledo XS205 analytical balance under controlled humidity, enabled the calculation of PM\(_{10}\) concentration by dividing the collected mass (g) by the total sampled air volume (m\(^3\))34. Specifically, for MMA, samples were gathered from the urban area, Santa Catarina (25°40’32.4”N 100°27’54.1”W), characterized by diverse industrial activities and heavy vehicular traffic. For CHI and CDJ, samples were collected near the Center for Advanced Materials Research (28°42’57.5”N 106°07’59.5”W) and the Delfhic-sec monitoring station (31°42’44”N 106°23’43”W), respectively. Both areas exhibit urban traits, moderate industrial activity, and light vehicular traffic, with CHI noted for metallurgical and mining industries and CDJ for textile and mining sectors.

Figure 1
figure 1

The map displays the exact locations where PM\(_{10}\) samples were collected in the MMA, CHI, and CDJ regions of Northern Mexico. These sites were selected based on their distinctive geographical and pollution profiles, with each site showing different levels of industrial and vehicular activity that affect the composition of particulate matter. The map was generated using QGIS version 3.40.0, available at https://qgis.org/.

Full size image
Table 1 Statistical summary of meteorological parameters and criteria pollutants in Northern Mexico cities during winter 2020.
Full size table

Extraction, cleanup, and chromatographic analysis

Analysis of PAHs

During the entire sample treatment, the samples remained protected from light to prevent photochemical degradation of the analytes. For extracting PAHs from the PM\(_{10}\) samples, 30% of the filter area was selected, cut into small pieces for each sample, and placed in a 125 mL Erlenmeyer flask. Each flask received 30 mL of dichloromethane and 1 mL of a surrogate. The mixture underwent extraction using an ultrasonic bath for 15 minutes at room temperature. Following this, the supernatant liquid was separated from the filter fragments35,36.

Furthermore, to enhance extraction efficiency, the sonication process in the ultrasonic bath was repeated using 30 mL of dichloromethane. The extracts from both stages were combined and evaporated in a rotary evaporator at 50\(^\circ\)C and 160 rpm until the volume was reduced to less than 1 mL. The resulting extract was transferred to a 1 mL volumetric flask, and 20 \(\mu\)L of the internal standard was added. The solution was then adjusted to a final volume of 1 mL with dichloromethane. The extracted sample was purified using a 13 mm PTFE filter with a pore size of 0.45 \(\mu\)m. The filtered extract was directly transferred into a 2 mL vial, and all extracts were stored at -20\(^\circ\)C until chromatographic analysis.

The identification of 16 PAHs was achieved by employing an Agilent 7890B gas chromatograph coupled with an Agilent 5977A mass selective detector. Compound separation was conducted using an Agilent DB-UI 8270 D ultra-inert column (20 m x 180 \(\mu\)m x 0.36 cm). Helium was utilized as the carrier gas with a 1.5 mL/min flow rate. The equipment operated in splitless mode, with an injection temperature of 300 \(^\circ\)C. Regarding the detector conditions, the electron impact ionization source was maintained at 230 \(^\circ\)C, the quadrupole at 150 \(^\circ\)C, and the selected ion monitoring (SIM) mode was employed with a mass range of 50 to 500 m/z. The thermal ramp utilized consisted of an initial temperature of 40 \(^\circ\)C, a hold time of 0.5 min, a ramping at 40 \(^\circ\)C/min up to 260 \(^\circ\)C, an increase of 6 \(^\circ\)C/min up to 295 \(^\circ\)C, and a final stage with a ramp of 25 \(^\circ\)C/min up to 325 \(^\circ\)C. The injection volume was 1 \(\mu\)L. The calibration curves were generated using certified standards of PAHs from the AccuStandard brand (PAHs-Mix 16). This mixture includes compounds such as naphthalene (NP) acenaphthene (ACE), acenaphthylene (ACY), phenanthrene (PHE), anthracene (ANT), Fluorene (Fl), Fluoranthene (FLA), pyrene (PYR), benzo[a]anthracene (BaA), benzo[a]pyrene (BaP), benzo[b]fluoranthene (BbF), benzo[g,h,i]perylene (BgP), benzo[k]fluoranthene (BkF), chrysene (CRY), dibenzo[a,h]anthracene (DahA), and indeno[1,2,3-cd]pyrene (IP).

Quantification of NPAHs

In the cases of NPAHs, a portion equivalent to 30% of the filter was cut and extracted using 30 mL of high-purity dichloromethane (DCM, from Carlo Erba 99.9%) in an ultrasonic bath for 15 minutes at room temperature. This procedure was carried out in duplicate. Subsequently, the extracts were preconcentrated using a rotary evaporator (at 40\(^\circ\)C and 180 rpm) until nearly dry. Moreover, 5 mL of hexane was added to each sample, followed by a new sonication step for 15 minutes at the same temperature. A silica gel cartridge, previously conditioned with 5 mL of dichloromethane, was employed for purification. The NPAHs were eluted from the cartridge using 5 mL of dichloromethane and subsequently analyzed by gas chromatography-mass spectrometry (GC-MS) using an Agilent Technologies model 7890A coupled to an inert mass-selective detector (Agilent 5975C). An HP-5MS capillary column with dimensions of 30 m \(\times\) 250 \(\mu\)m \(\times\) 0.25 \(\mu\)m was employed37.

The injector was maintained at 350°C, injecting 1 \(\mu\)L of the sample in splitless mode. The MS was operated using a NCI source at 1324 V, with CH\(_{4}\) as the ionizing gas, maintaining the temperature at 350°C. The thermal ramp utilized included an initial temperature of 45°C, a hold time of 2 min, ramping at 45°C/min up to 150°C, followed by an increase of 5°C/min up to 300°C. The MS was operated at 350°C, and helium was used as the carrier gas at a 20 mL/min linear velocity. Calibration curves were generated using certified standards of NPAHs (AccuStandard), including 1-nitronaphthalene (1N-NAP), 2-nitrobiphenyl (2N-BIP), 5-nitroacenaphthene (5N-ACE), 2-nitrofluorene (2N-FL), 9-nitrophenanthrene (9N-PHE), 9-nitroanthracene (9N-ANT), 3-nitrofluoranthene (3N-FLA), 1-nitropyrene (1N-PYR), 2,7-dinitrofluorene (2,7DN-FL), 6-nitrochrysene (6N-CRY), and 7-nitrobenzo[a]anthracene (7N-BaA). All data were recorded and processed using Analytical Software Version B.05.

Quality control

For the analysis of PAHs, two calibration curves were generated with working ranges of 20 to 100 \(\mu\)g/L and 80 to 400 \(\mu\)g/L, using certified standards, while for NPAHs, curves in the range of 0.25 to 10 ppb were used. The correlation coefficients obtained in the calibration curves for both PAHs and NPAHs were greater than 0.998. All samples and blanks (reagents and filters) were analyzed in duplicate. Precision, expressed in terms of repeatability, was determined by analyzing 10 times the standards of the compounds of interest, resulting in a coefficient of variation values less than 10%. In order to determine the accuracy of the method, the samples were fortified by adding a known amount of PAHs and NPAHs to the filters with the collected material. The recovery percentages obtained ranged between 70 and 130% for the PAHs and between 85 and 112% for the NPAHs. The detection limits were: NP 9.36 ng/mL, ACY 10.4 ng/mL, ACE 9.07 ng/mL, Fl 9.18 ng/mL, PHE 8.01 ng/mL, ANT 11.35 ng/mL, FLA 8.9 ng/mL, PYR 8.87 ng/mL, CRY 6.22 ng/mL, BaA 9.07 ng/ml mL, BbF 7.18 ng/mL, BkF 10.26 ng/mL, BaP 6.54 ng/mL, IP 7.31 ng/mL, DahA 7.47 ng/mL, and BgP 7.85 ng/mL. In the case of NPAHs, the detection limits ranged between 0.04 and 0.37 ng/mL.

Preliminary source contributions

DRs have been widely documented in the literature as effective tools for qualitatively identifying the emission sources of PAHs and NPAHs27,38,39. This methodological approach allows researchers to distinguish between various sources, such as PAHs and NPAHs derived from pyrogenic and petrogenic origins and coal combustion, biomass burning, wood burning, and vehicular emissions from gasoline and diesel combustion. However, these ratios with caution are essential, as PAHs and their derivatives can be emitted from multiple sources, particularly combustion-related ones, and their profiles can vary due to their reactivity40. For PAHs, the calculated DRs included IP/(IP + BgP), FLA/(FLA + PYR), ANT/(ANT + PHE), FL/(FL + PYR), BaP/BgP, and BbF/BkF, while for NPAHs, the ratios used were 1N-PYR/PYR, 9N-ANT/PYR, and 3N-FLA/1N-PYR. Table 2 presents the reference values and the significance of each ratio used. In addition to the DRs, a PCA was conducted to better understand the potential sources and the distribution of PAHs and NPAHs in the environment. The statistical analysis, including PCA, was performed using Python 3, leveraging libraries such as NumPy41, Pandas42, and scikit-learn (43. All computations were carried out in Google Colab (https://colab.research.google.com/).

Table 2 Molecular DRs of PAHs for different source categories.
Full size table

Evaluation of the excess of cancer risk

An evaluation of the ECR associated with the inhalation of PAHs and NPAHs found in PM\(_{10}\) was carried out in accordance with the guidelines set forth by the Office of Environmental Health Hazard Assessment (OEHHA) of the California Environmental Protection Agency (CalEPA)51. To begin, the total equivalent concentration, denoted as [BaP]\(_{eq}\), was calculated using Equation 1:

$$\begin{aligned} {[}BaP]_{eq} = \varvec{\sum } (C_i \times TEF_i) \end{aligned}$$
(1)

In this context, C\(_i\) refers to the mass concentration of each PAHs or NPAHs in the samples, measured in nanograms per cubic meter (ng/m\(^3\)). The subscript i indicates the specific compound being referenced. Additionally, TEF\(_i\) denotes the toxicity equivalent factor for species i, as documented in the literature52. The ECR (Estimated Cancer Risk) was calculated using Equation 2.

$$\begin{aligned} ECR = [BaP]_{eq} \times UR_{BaP} \end{aligned}$$
(2)

In the above Equation, UR\(_{BaP}\) represents the inhalation cancer unit risk factor for BaP. This factor is defined as the number of people at risk of developing cancer from inhaling a concentration of 1 ng/m\(^3\) of BaP over a lifetime of 70 years (13,53. According to the World Health Organization (WHO), the UR\(_{BaP}\) value is \(8.7 \times 10^{-5}\), while the value provided by CalEPA is \(1.1 \times 10^{-6}\)54,55.

Results and discussion

This section presents the main findings regarding the concentrations of PAHs and NPAHs in PM\(_{10}\) at the selected sites. It includes an analysis of concentration levels, spatial distribution, and correlations with meteorological parameters. The contributions of different sources are examined using DRs and PCA, while health risks are assessed through BaPeq values and the ECR. The results emphasize the necessity for informed public health policies and effective environmental management strategies to reduce pollution exposure and safeguard community health.

Concentration levels of PAHs and NPAHs

Variations in PM\(_{10}\) concentrations among the cities have been previously reported and discussed in34. Table 3 shows the average concentrations of individual PAHs and NPAHs in PM\(_{10}\) samples collected from three urban locations, and Fig. 2 presents a spatial visualization of pollutant distribution, emphasizing the intensity and spread of PAHs and NPAHs across the cities.

Regarding PAHs, the total concentrations were 108.89 ± 99.90 ng/m\(^3\) for MMA, 100.69 ± 122.60 ng/m\(^3\) for CHI and 73.26 ± 90.85 ng/m\(^3\) for CDJ. As shown in Fig. 2, high molecular weight (HMW) PAHs were the most abundant in all cities, suggesting significant contributions from gasoline-powered vehicle emissions44,48. In contrast, low molecular weight (LMW) PAHs were detected at lower levels across sites, with compounds such as NP, ACE, FL, PHE, ANT, FLA, and PYR absent in CDJ. This lower concentration of LMW PAHs likely results from gas-particle partitioning, which depends on the vapor pressures of PAHs, meteorological conditions (T and RH), and photochemical transformations, along with seasonal variation effects.46.

Figure 3 shows the proportional contribution of individual PAHs to the overall PAHs concentration across cities. It is observable that the BbF was the most abundant PAHs, accounting for 19-25% of the total PAHs, followed by BgP (15%), IP (13%), and BaP (12%). CDJ consistently showed the lowest PAHs concentrations for all compounds analyzed except for DaA. BaP, one of the most toxic PAHs and classified as a Group 1 carcinogen by the IARC, exhibited its highest concentrations in MMA and CHI. Statistical analysis using the Mann-Whitney Test (\(p< 0.05\)) showed no significant differences in BaP concentrations between these cities. For IP, BaP, and BgP, MMA and CHI showed relatively similar levels, approximately 1.3 to 1.5 times higher than in CDJ. In contrast, CRY and BaA levels in CHI and CDJ were comparable, measuring 0.63 to 0.76 times lower than in MMA. The elevated concentrations of highly toxic PAHs, such as BaP, BaA, and IP, at these sites, underscore the need for effective public policies to reduce the health and environmental impacts of these pollutants.

Levels of NPAHs exhibited variability across different sites. Five of the 11 compounds analyzed were quantified in both MMA and CDJ, while only three were detected in CHI, with total NPAH concentrations significantly lower than those of PAHs. Notable differences in NPAH levels were observed among the cities, with the exception of 1N-PYR. The total concentrations of NPAHs in MMA, CHI, and CDJ were 470.32 pg/m\(^3\), 247.26 pg/m\(^3\), and 193.20 pg/m\(^3\), respectively. Specifically, 9N-ANT, 3N-FLA, and 1N-PYR were detected at all sites, while 1N-NAP was unique to MMA, and 6N-CHR was exclusive to CDJ. Figure 3 illustrates 3N-FLA as the most predominant NPAH, accounting for 56% (263.39 pg/m\(^3\)) in MMA, 55% (105.19 pg/m\(^3\)) in CHI, and 39% (95.19 pg/m\(^3\)) in CDJ. The strong correlation among common NPAHs in MMA and CHI (r² = 0.997) suggests similar emission profiles in these cities. The second most abundant NPAH, 9N-ANT, serves as an indicator of diesel exhaust and biomass combustion56 and forms through heterogeneous reactions involving anthracene and nitrating agents57. Its strong correlation with ANT (r² = 0.976) further supports this association. The concentration of 1N-PYR in MMA and CDJ was roughly double that found in CHI. 1N-PYR is the dominant NPAH in diesel exhaust, with a possible minor contribution from gasoline engine emissions58,59,60. These findings indicate that vehicular traffic and direct emissions may have a more pronounced impact on MMA and CDJ compared to CHI, likely due to the greater number of vehicles in the former and the influence of commercial border crossings in the latter. Regarding toxicity, compounds such as 1N-PYR and 6N-CRY, detected at notable levels in CDJ, are classified by the IARC as Group 2A carcinogens. This underscores the urgent need for strategies to reduce their emissions61.

Table 3 Average concentrations of the individual PAHs and NPAHs in PM\(_{10}\) samples collected in three cities in Mexico during the winter season.
Full size table

The concentration of 9N-PHE was found to be 2.6 times higher in MMA compared to CDJ, and it was not detected in CHI. This compound is primarily known to originate from biomass combustion rather than from vehicular emissions62. Therefore, this indicates a greater influence of biomass combustion in MMA than CDJ. When comparing total NPAHs concentrations with other urban regions, MMA, CHI, and CDJ showed higher values than those reported in Ciudad Real and Madrid, Spain31,63, and Medellín, Colombia64. However, NPAHs levels in MMA were comparable to those observed in northern Mexico City16, while CHI and CDJ values were lower. The metropolitan area of Porto Alegre, Brazil, presented significantly higher levels than those in this study65.

MMA showed the highest concentrations of both PAHs and NPAHs, suggesting a potentially higher risk for the exposed population. These findings are crucial for developing and implementing effective policies and management strategies to improve air quality. By identifying specific compounds and understanding their distribution across urban areas, targeted mitigation efforts can be designed to address the unique emission sources of each region.

Figure 2
figure 2

Average concentrations of PAHs and NPAHs in PM\(_{10}\) across MMA, CHI, and CDJ. (a) shows the average concentrations of 15 PAHs compounds, highlighting significant differences among the cities. (b) displays the average concentrations of 5 NPAHs compounds, indicating the predominance of certain compounds in specific urban areas.

Full size image
Figure 3
figure 3

Quantitative distribution of PAHs (a) and NPAHs (b) compounds in PM\(_{10}\) across sampling sites. It highlights the specific proportions of various PAHs and NPAHs present in PM\(_{10}\), allowing for a comparative analysis of the pollutant profiles associated with each examined urban environment.

Full size image

Correlation between PAHs, NPAHs, meteorological parameters, and criteria pollutants

Figure 4 shows the Spearman correlation matrix for the MMA, depicting correlations between PAHs, NPAHs, meteorological parameters, and criteria pollutants. Most PAHs in MMA exhibited significant positive correlations with RH, indicating that elevated RH favors PAHs condensation onto particles, aligning with previous findings37. Conversely, significant negative correlations were observed between these compounds and SR and T, suggesting the degradation of these compounds as both parameters increase. Similar results have been reported by other researchers66. HMW PAHs, such as BaP, BbF, BkF, and IP, as well as DahA, showed high correlations with each other, suggesting that these compounds, which are more persistent in the environment and more prone to particle adsorption, may be emitted from similar sources such as incomplete combustion of fossil fuels and biomass burning. In the case of NPAHs, 3N-FLA, the most predominant NPAHs in PM\(_{10}\), showed a significant positive correlation with SO\(_{2}\), suggesting a primary origin possibly associated with vehicular emissions (diesel) and petrochemical activities at the East outskirts of the MMA where one of the six operational refineries of the country is established (Cadereyta Refinery) which is a well-known major SO\(_{2}\) and PM point emissions source. Several studies have shown that diesel engines are a significant source of 3N-FLA67,68. Some NPAHs, such as 1N-PYR and 9N-ANT, showed negative correlations with O\(_{3}\), which evidences the susceptibility of these compounds to the action of highly oxidizing species. The inverse relationship between PAHs and WS suggests that higher WS leads to greater dispersion of these compounds.

Figure 4
figure 4

The spearman correlation matrix for the MMA illustrates the relationships between PAHs, NPAHs, meteorological parameters, and pollutants. The matrix shows strong positive correlations between HMW PAHs and RH and negative correlations with T and SR. The positive correlation between 3N-FLA and SO\(_{2}\) is notable, suggesting sources linked to vehicular emissions.

Full size image

For CHI and CDJ, the correlation matrices are shown in Figs. 5 and 6, respectively. In CHI (Fig. 5), a significant negative correlation was observed between PAHs and T and O\(_{3}\). Although significant positive correlations were found between PAHs with both particle sizes (PM\(_{10}\) and PM\(_{2.5}\)), the correlation was notably higher with PM\(_{10}\), especially with HMW PAHs such as BgP, BbF, BkF, BaP, and IP. It suggests a common origin associated with direct emissions, such as the incomplete combustion of fossil fuels and/or biomass burning. In contrast, in the MMA, only a moderate positive correlation between PM\(_{10}\) and \(\sum\) PAHs was obtained, while the correlation was negative for PM\(_{2.5}\). The correlation between the PM fractions and \(\sum\) PAHs was low in both cities, though higher in CHI. In the MMA, a considerable fraction of the PM\(_{2.5}\) is secondary in origin, possibly explaining the negative correlation between this pollutant and the \(\sum\) PAHs, which are primary in origin. On the other hand, meteorological parameters such as P and WS influenced the observed behavior of the pollutants; for example, a significant positive correlation was observed between atmospheric pressure and PM\(_{10}\), PM\(_{2.5}\), and heavier PAHs, which could be attributed to high-pressure conditions usually being related to a lower dispersion of pollutants in the air. Moreover, the average WS was considerably lower in CHI (1.63 m/s) compared to other sites (Table 1 and Fig. 7), which contributed to a lower dispersion of pollutants. Regarding NPAHs, a significant positive correlation was found between 1N-PYR and NOx, suggesting that vehicle emissions, especially those from diesel engines, represent a significant source of this atmospheric pollutant in the region65. A high positive correlation was observed between 1N-PYR and 3N-FLA, indicating that both pollutants could originate from the same source.

Figure 5
figure 5

The Spearman correlation matrix for CHI highlights significant negative correlations between PAHs and T and O\(_{3}\). Strong positive correlations between PAHs and PM\(_{10}\) suggest primary emissions from combustion sources.

Full size image

Instrumental limitations prevented the inclusion of SR, WD, and PM\(_{2.5}\) in the CDJ matrix (see Fig. 6). In the CDJ, there was a strong internal correlation among PAHs (correlation coefficient > 0.86), but no significant correlation was observed with NPAHs. This might indicate differences in sources or behavior between these groups. Some atypical air quality responses in the CDJ-El Paso region have been attributed to industrial and commercial activities along the US/Mexico border. Unlike in MMA and CHI, most PAHs in CDJ did not show significant correlations with T and RH, possibly due to the low variability of these parameters during the study period. This has been similarly observed in studies on PM\(_{2.5}\) in Xinxiang, China66. Additionally, a slightly positive correlation was noted between the heavier PAHs (BbF, BkF, IP, BgP) and PM\(_{10}\), revealing a greater adsorption capacity on particles, which could have significant implications in terms of the transport and deposition of these pollutants in the environment. Regarding NPAHs, the significant positive correlation between 3N-FLA, 6N-CRY, and 9N-ANT suggests a common emission source. In contrast to MMA and CHI, \(\sum\)PAHs and \(\sum\)NPAHs positively correlated with WS in CDJ. This could indicate primary and secondary components being transported from the binational basin.

Figure 6
figure 6

Spearman correlation matrix for CDJ shows strong internal PAH correlations and significant correlations with WS, suggesting pollutant transport. Differences in correlations between PAHs and NPAHs indicate distinct source behaviors in the region.

Full size image
Figure 7
figure 7

Temporal variation of meteorological parameters (T, RH, WS, and SR) in MMA, CHI, and CDJ during the study period.

Full size image

Source contribution of PAHs and NPAHs

Diagnostic relations

Analysis of PAHs

The DRs calculated in this study are shown in Table 4 for both PAHs and NPAHs. The values in this table can be found elsewhere44,48. In general, the average results from molecular DRs indicate the presence of pyrogenic and petrogenic PAHs with a clear dominance of pyrogenic sources given the results for the LMW/HMW ratio \(<1\)39, that is common in urban cities like CDJ, CHI, and MMA in which the industrial processes and vehicular emissions are predominant. The average ratios of IP/(IP + BgP) are quite uniform suggesting the petrogenic origin of the PAHs48, such as engine oil and diesel combustion39,47, in all sites; other studies suggest electro-static precipitator from cement plants69, that could be associated with the active cement industry in MMA. The average FLA/(FLA + PYR) ratios range from 0.38 to 0.50, indicating a mix of emission sources across the different sampling sites. For MMA, the ratio suggests influences from biomass burning (0.43-0.58) and gasoline vehicles (\(<0.50\)); here, the vehicle fleet is comprised of 75% of motor gasoline vehicles (INEGI, 2023). In CHI, the sources include emissions from both heavy-duty and light-duty vehicles (0.37-0.40), while in CDJ, the dominant sources appear to be fossil fuel combustion from catalyzed cars (0.40-0.50). These findings are supported by research from47,69,38,39. The average ratios of BaA/(BaA + CRY) for CHI and CDJ were quite similar (0.477 and 0.469), which are in the range of coal combustion and vehicular emissions but a little bit lower that of biomass burning (\(<0.50\)), whereas for MMA was the opposite, the average ratio is the range of biomass burning but a little bit higher than that for coal combustion46. The average ratios of ANT/(ANT + PHE) for MMA and CHI exceeded 0.1, suggesting a pyrogenic origin of PAHs, possibly from coal combustion, as detailed by46 and supported by48. Additionally, this ratio may be linked to atmospheric processes, as observed in the case of CDJ, where no ANT concentrations were detected. This absence could be due to the quicker gas-phase reactivity of ANT with OH\(^{\cdot }\) radicals compared to PHE. This difference also suggests a potential seasonal variation44,70.

The FL was not detected in CHI and CDJ. As mentioned above, this could be associated with a seasonal effect that allowed the presence of FL only in MMA and with the fact that FL predominates in the gas phase and PYR in the gas-particle phase47. Therefore, the FL/(FL+PYR) ratio was only quantified in MMA, suggesting gasoline emissions47. The average ratios of BaP/BgP range from 0.77 to 0.86, indicating a traffic emission origin of PAHs39,71 that is in line with populated regions where more vehicles are used. The average ratios of BbF/BkF range from 2.3 to 3.5, suggesting different emission sources for all sites. For CDJ, the ratios align with diesel emissions, typically greater than 0.647, and are indicative of aluminum smelter emissions, with specific values between 2.5 and 2.948,66. For MMA and CHI, the ratios indicate various industrial emissions, including steel production, forging, and the extraction and refining of metals.

Table 4 DRs of PAHs and NPAHs for all sampling sites.
Full size table

Analysis of NPAHs

Some molecular DRs have been used to trace the NPAHs sources, and their results for this study are listed in Table 4. The 1N-PYR/PYR ratio usually indicates vehicle exhaust and coal combustion with values of 0.36 and 0.001, respectively38,49. The average ratios of 1N-PYR/PYR are uniform (\(<0.05\)), suggesting an important contribution from coal combustion. Another ratio to assess the contributions of biomass burning and motor vehicle exhaust is 9N-ANT/1N-PYR. Values greater than 10 indicate wood combustion, while values less than 10 suggest emissions from motor vehicle exhaust38,50. On the one hand, the average ratios of 9N-ANT/1N-PYR for MMA and CHI were approximately 4.2, suggesting that coal combustion and vehicle exhaust are likely contributors to NPAHs, with a possible minor contribution from biomass burning. This ratio was 85% lower in CDJ, indicating a significantly higher predominance of coal combustion and vehicle exhaust at this site. The average ratios of 3N-FLA/1N-PYR suggest different sources of NPAHs at the analyzed sites. Ratios greater than 5 for MMA and CHI indicate that the NPAHs likely originate from atmospheric reactions. Conversely, in CDJ, ratios less than 5 suggest that the NPAHs primarily come from direct combustion emissions.

The results showed specific differences among these cities. Firstly, the MMA exhibits a clear presence of biomass-burning sources compared to CDJ and CHI. Secondly, despite all cities showing the contribution of traffic emissions, the MMA site has a predominance of diesel emissions in comparison with CDJ and CHI, which is in line with the fact that MMA has 1.72 times more transport bus fleet and similar heavy trucks fleet72. Finally, CDJ and MMA seem more affected by atmospheric processes and seasonal effects, whereas CHI seems to have a more stable atmosphere during this study.

Statistics analysis (PCA)

A PCA aimed to identify potential emission sources for PAH and NPAH levels across these cities. The findings from this analysis supported several conclusions previously derived through DRs. Figure 8 illustrates the loadings and variances for each component identified at the three study sites. Due to the limited detection of NPAHs in the samples, the PCA analysis concentrated exclusively on PAH concentrations, which enabled a more extensive data set. In the analysis of MMA, two main components emerged. The first component (PC1) accounted for 61% of the total variance and comprised PHE, FLA, PYR, BaP, IP, DahA, and BgP. The contrasting signs of the loadings suggest that this component arises from a mixture of sources. Specifically, PHE, FLA, and PYR have been linked in the literature to biomass burning67,73,74. In contrast, BaP, IP, DahA, and BgP are often associated with vehicular emissions, primarily from gasoline combustion13,75. Notably, iso-propyl (IP) and benzopyrene (BgP) are frequently identified as tracers for gasoline vehicles by various authors47,76. The outcomes for PC1 align with the high traffic density and charcoal burning prevalent in the MMA, corroborating the DRs for this city. Principal Component 2 (PC2) explains 26.1% of the variance and features moderate positive loadings of acenaphthylene (ACY) and fluorene (FL), alongside moderate negative loadings of chrysene (CRY), benzanthracene (BaA), benzo[b]fluoranthene (BbF), and benzo[k]fluoranthene (BkF). Compounds such as CRY, BaA, BbF, and BkF are typically associated with high-temperature processes, including industrial oil burning77. In contrast, lighter polycyclic aromatic hydrocarbons (PAHs) like ACY and FL generally originate from low-temperature combustion processes.

In CHI, two components emerged. PC1 accounted for 72.8% of the variance, including CRY, BaA, BbF, BaP, BkF, IP, and BgP. This component aligns with high-temperature industrial emissions linked to the steel and iron industry, as well as metal extraction and refining processes78,79, consistent with the strong influence of metallurgical and mining industries in the region. Vehicular emissions from gasoline combustion could also contribute to these PAHs in the atmosphere. The second component (PC2) explained 20.1% of the variance. It featured moderate loadings for NP, FLA, and PYR, suggesting an association with biomass burning, where similar findings were reported by78 in studies conducted in China. In the case of CDJ, two components were identified. PC1 accounted for 93.9% of the variance, comprising CRY, BaA, BbF, BaP, BkF, IP, and DahA. This component likely reflects diesel combustion due to the heavy truck traffic in the region, as well as wood burning since many households use wood for heating during winter. PC2, explaining 5.5% of the accumulated variance, included BgP, IP, BbF, BaP, BaA, and CRY, pointing to gasoline engine emissions as a probable source.

Figure 8
figure 8

PCA of PAHs in PM\(_{10}\) across MMA, CHI, and CDJ. This figure illustrates the loadings (a, c, e) and cumulative explained variance (b, d, f) for each principal component, highlighting the major sources and patterns of PAHs pollution in these urban areas.

Full size image

Health risks

The assessment of BaPeq and potential cancer risk from the inhalation of PAHs in PM\(_{10}\) samples are shown in Table 5. The highest mean BaPeq was recorded in CHI (22.26 ng/m\(^3\), range: 14.31-162.69 ng/m\(^3\)), followed by MMA (20.07 ng/m\(^3\), range: 14.99-68.17 ng/m\(^3\)) and CDJ (15.02 ng/m\(^3\), range: 13.76-96.65 ng/m\(^3\)). These values align with high pollution levels reported in densely populated urban areas, such as Beijing, China80. However, they remain below the maximum thresholds observed in some studies conducted under different conditions. The elevated BaPeq in CHI suggests significant contributions from highly carcinogenic compounds, notably BaP (59.80%), DahA (14.96%), and BbF (11.39%), consistent with findings from other metropolitan areas80,81. In MMA, BaP (62.57%), BbF (10.57%), and BaA (7.57%) were predominant, while in CDJ, BaP (57.32%), followed by DahA (13.58%) and BbF (11.44%), were the main contributors. These results underscore the ubiquitous nature of BaP across urban environments as a primary driver of carcinogenic risk.

Table 5 BaPeq levels and ECR from inhalation of PAHs and NPAHs in PM\(_{10}\) samples.
Full size table

The elevated BaPeq values observed across the three cities highlight the urgent need for effective air quality management policies. The higher concentrations of CHI may result from a combination of industrial emissions, vehicular traffic, and meteorological conditions that contribute to pollutant accumulation. The persistent presence of high BaP levels emphasizes the importance of targeted interventions to reduce emissions from major sources, including vehicular exhaust and industrial activities. While CDJ exhibited lower mean BaPeq levels than CHI and MMA, this does not imply a negligible risk. Even at moderate concentrations, prolonged exposure can lead to cumulative health effects, aligning with studies on the chronic impacts of sustained pollutant exposure, particularly in vulnerable populations19.

The BaPeq results for NPAHs also present a notable concern. Table 5 shows that MMA recorded a BaPeq of \(2.392 \times 10^{-1}\) ng/m\(^3\) (range: \(1.976 \times 10^{-4}\) ng/m\(^3\) - \(7.033 \times 10^{-1}\) ng/m\(^3\)), indicating a moderate but non-negligible carcinogenic potential from NPAHs. CDJ showed a higher BaPeq at 0.6436 ng/m\(^3\), which may indicate increased emissions and conditions that favor the formation or persistence of these compounds. The significant contributions from 1N-PYR and 3N-FLA show the complexity of atmospheric chemistry, where both direct emissions and secondary formation processes influence exposure risks. These compounds, often associated with diesel emissions and industrial activities, underscore the need for expanded monitoring programs to include a broader range of pollutants to better characterize health risks.

The comparison of ECR values for a 70-year lifetime exposure highlights significant public health concerns. In MMA, the ECR was \(2.183 \times 10^{-4}\) using the WHO method and \(2.760 \times 10^{-6}\) based on CalEPA, indicating a moderate cancer risk. CHI showed slightly higher ECRs at \(2.420 \times 10^{-4}\) and \(3.060 \times 10^{-6}\), respectively. In contrast, the values for Ciudad Juárez were lower but still noteworthy, with ECRs of \(1.743 \times 10^{-4}\) and \(2.203 \times 10^{-6}\). Notably, the higher ECR from NPAHs in CDJ (\(5.600 \times 10^{-5}\)) underscores their significant contribution to the carcinogenic risk, despite their lower concentrations compared to PAHs.

While these ECR values offer a theoretical framework for understanding risk, they are crucial for guiding environmental and public health policies. The findings emphasize the need for strengthened regulatory measures to control emissions and reduce exposure, especially in urban areas with significant industrial and vehicular pollution. It aligns with global trends that advocate for comprehensive air quality management strategies. The substantially higher BaPeq and ECR values reported in this study compared to previous findings for MMA (e.g., BaPeq = 0.51 ng/m\(^3\), ECR\(_{WHO}\) = \(4.47 \times 10^{-5}\)) indicate that pollutant levels have risen, necessitating immediate action. These findings align with trends observed in other Mexican and Latin American cities, as well as internationally, where urban growth and industrial activities pose significant challenges to air quality76,82,83.

Future research should focus on expanding the toxicological database for NPAHs and developing more precise methods for assessing their health impacts. Such efforts would enhance the reliability of BaPeq and ECR estimates, leading to better-informed public health initiatives and regulatory actions.

Implications for public health policy and environmental management

Amid rapid industrialization, Northern Mexico faces the challenge of balancing economic growth with public health and environmental sustainability. The cities of MMA, CHI, and CDJ illustrate how industrial activities, while driving economic development, significantly compromise air quality. Without targeted, evidence-based policy interventions, the health of millions will remain at serious risk, threatening the long-term viability of these regions. This situation calls for a critical reassessment of current policies and the adoption of comprehensive strategies that align economic advancement with public health and environmental priorities.

This research lays the groundwork for developing targeted public health and environmental policies. By identifying the sources and concentrations of air pollutants, policymakers can create more precise and effective strategies. However, qualitative recommendations alone are not sufficient for lasting mitigation. Integrating quantitative assessments and data-driven policies is essential to address key sources, such as biomass burning and vehicle emissions, and to implement stricter emissions standards and cleaner industrial technologies with measurable outcomes.

Significant industrial pollution from sectors such as cement production, chemical processing, and metalworking further complicates the situation. Addressing these challenges requires not only qualitative insights but also data-driven frameworks supported by real-time monitoring and predictive modeling.

Based on the findings of this study, the following recommendations should guide public health and environmental policy development:

  1. 1.

    Enforce stricter emissions standards: Set clear, science-based limits for key pollutants like PAHs and NPAHs, supported by health risk assessments, regular audits, and penalties for non-compliance. Regular updates informed by new research and technological advancements are essential for maintaining effectiveness.

  2. 2.

    Provide financial incentives for cleaner technologies: Offer subsidies and tax credits to encourage industries to adopt advanced emission control systems, linking incentives to verified emission reductions. Support research and development (R&D) to drive innovation in sustainable industrial practices.

  3. 3.

    Establish a comprehensive air quality monitoring network: Implement sensors for real-time data collection and use predictive models to guide proactive measures and public health advisories. Ensure data is publicly accessible to promote transparency and community engagement.

  4. 4.

    Integrate urban planning that prioritizes green spaces and active transportation: Expand green areas to help absorb pollutants and invest in infrastructure that supports walking and cycling, thereby reducing vehicle emissions and promoting healthier lifestyles.

  5. 5.

    Strengthen international partnerships: Collaborate with global organizations to share knowledge, access resources, and adopt proven policy approaches. Periodic joint assessments can enhance local strategies and ensure alignment with international best practices.

While these measures may not have an immediate impact, they lay a strong foundation for raising awareness and fostering gradual change. With continued scientific contributions and a sustained focus on these urban areas, significant progress can be made in developing and implementing effective public policies. Coordinated actions and collaboration among researchers, governments, and local communities are essential for achieving lasting improvements in environmental protection and public health.

Conclusions

This study analyzed the concentrations, sources, and health risks associated with PAHs and NPAHs in PM\(_{10}\) particles across urban areas in northern Mexico, revealing significant spatial variability. The highest levels were detected in MMA and CHI, with total NPAH concentrations of 470 pg/m\(^3\) for MMA, 247 pg/m\(^3\) for CHI, and 193 pg/m\(^3\) for CDJ, where 3N-FLA was the dominant compound across all sites.

HMW such as BbF, IP, and BghiP were prevalent in all three cities, indicating substantial emissions from high-temperature combustion sources. Notably, the highest concentrations of BaP, a known carcinogen, were recorded in MMA and CHI, emphasizing significant public health risks in these regions. PCA and DRs identified primary emission sources, including biomass burning, vehicular emissions, and industrial processes. The BaPeq calculations and potential cancer risk assessments revealed that CHI had the highest BaPeq value at 22,257 ng/m\(^3\), followed by MMA at 20,074 ng/m\(^3\) and CDJ at 15,022 ng/m\(^3\). These values indicate moderate cancer risks, with particularly elevated risks associated with MMA and CHI. Additionally, the results in CDJ for NPAHs indicated a significant carcinogenic risk, highlighting the need for targeted control measures.

These findings provide essential data for understanding air quality challenges in northern Mexico and lay the groundwork for future research and policy development. Immediate actions should prioritize enhancing air quality monitoring networks, enforcing stricter emissions regulations, and promoting cleaner technologies to mitigate health risks and improve air quality. The results support the development of policies tailored to the unique pollution profiles of each city and emphasize the importance of targeted strategies that address key emission sources, fostering sustainable urban development. The study calls for prompt and effective policy measures to mitigate the risks associated with PAHs and NPAHs, contributing to public health advancements and stronger environmental protection in Mexico.

Future work

This study focused on winter sampling (January to March), chosen for its higher pollutant concentrations to establish a baseline representing worst-case scenarios. Future research will extend to other seasons sampling to capture seasonal variability in PAHs and NPAHs concentrations across the studied urban centers. This expanded approach will provide deeper insight into how seasonal factors influence pollutant levels and their associated health risks. Additionally, future analyses will incorporate PM\(_{2.5}\) to evaluate better health risks linked to finer particulate matter. These efforts will support the development of more effective, year-round public health policies and air quality management strategies.