Trace Metals in PM₁₀ and Associated Health Risk in Two Urban Sites Located in Campeche

Abstract

This study reports the trace metal concentrations in PM₁₀ collected in two urban sites of Campeche, Mexico (the INAH site, located in the downtown, and the TNL site, located in the Tecnologico Nacional de Mexico campus Lerma). Trace metals (Cd, Co, Cu, Fe, Mn, and Zn) were determined by flame atomic absorption spectroscopy, with Fe and Mn being the most dominant species. Cd and Co showed high values of enrichment factors, indicating that they were highly enriched by anthropogenic sources. A health risk assessment was carried out according to the IRIS-EPA methodology considering both carcinogenic and non-carcinogenic effects and different age and gender population groups. The TNL site showed cancer risk coefficients that exceed the maximum limits established by the EPA and the WHO (1 × 10⁻⁶ and 1 × 10⁻⁵, respectively), being higher for Cd and Co. The non-carcinogenic risk exceeded the limits recommended by the EPA, being higher in the INAH site for cobalt (HQ = 28.92). There is a latent risk that the exposed population may develop cardiovascular and respiratory diseases due to inhalation of the metals measured in PM₁₀. It is necessary to carry out more continuous monitoring campaigns of trace metals linked to PM₁₀ in this area to enable a better understanding of the effects of these contaminants on the health of the exposed population. The results of the present work constitute the first efforts to evaluate the trace metals concentrations in PM₁₀ in urban areas of the city of Campeche, Campeche and can be used to develop programs focused on improving air quality.

1. Introduction

Urban growth and industrial development have resulted in the deterioration of air quality, with subsequent effects on people’s health and damage to ecosystems and materials, as well as global warming [1].

Air quality in urban areas depends on the presence and distribution of pollutants, meteorological and geographic conditions, solar radiation, and deposition and dispersion processes [2]. In recent years, fine particles (PM₁₀ and PM_2.5) have been of particular interest due to their adverse effects on human health, have been associated with an increase in respiratory problems in highly polluted urban areas and are associated with greater morbidity and mortality rates [3,4,5]. Acute and chronic exposure to PM₁₀ can exacerbate asthma and other respiratory symptoms, as well as cause premature mortality from cardiopulmonary diseases and lung cancer [4]. In addition to the mass composition of PM₁₀, the health effects of this pollutant also depend on its chemical composition, since these particles contain toxic elements such as heavy metals (Cd, As, Fe, Zn, Cr, Cu, Al, Ni, Co, Pb, etc.), some of which have been identified to be potentially carcinogenic [6]. Many studies are devoted to the analysis of pollutant levels in urban areas, also involving advanced modelling techniques. Levels of NOx emitted by road traffic have been simulated to evaluate the effects of urban canopy in favoring critical situations of exposure to traffic induced air pollutants [7]. Simulations have been used to assess the cancer risk induced by urban pollutants in two Italian cities, to detect the areas where the population is more directly exposed to potential risks for health [8,9]. Despite the relevance of this issue, in southeastern Mexico, there is little information on levels of suspended particles (PM₁₀) and their heavy metal content, as well as the risk that these pollutants represent for the health of the exposed population. There is only one air quality monitoring station in Yucatan peninsula located in Merida city, and the information is not continuously recorded (https://sinaica.inecc.gob.mx/) (accessed on 10 June 2022) [10]. To the southwest of the Yucatan Peninsula is located the city Campeche better known as “San Francisco de Campeche”, which is affected by an environmental phenomenon on a regional scale: the traditional burning of land for the preparation of crops and numerous forest fires that occur in a characteristic way during dry periods. In the last 10 years, the city has increased its population by 17.3%, which has led to a greater demand for services and an increase in the vehicle fleet [11]. On the other hand, in the surroundings of San Francisco de Campeche, there is a town named Lerma, where an electric generation plant is located and that supplies electricity to the region. This power plant constitutes an important source of PM₁₀ atmospheric particles. There is not enough information on the content of trace metals in PM₁₀ particles in urban areas of Campeche, Campeche in Mexico. Most studies report only the gravimetric concentrations of the particles and do not carry out the chemical characterization of their content. In this research work, the health risk represented by these atmospheric particles was determined based on their heavy metal content in the exposed population in two sites in the city of San Francisco de Campeche, Campeche, considering the risk of cancer in the lifetime and the risk of developing cardiovascular and respiratory diseases (non-carcinogenic effects). The objectives of this work were the following: (1) to determine the exposure to trace metals contained in PM₁₀ considering different age groups and gender at different values of frequency and exposure duration, (2) to determine the coefficients of cancer risk in the lifetime by inhalation of carcinogenic metals in PM₁₀ (cadmium and cobalt) applying the methodology based on the EPA Risk Guidelines, (3) to determine the risk coefficients of non-cancer (potential to manifest diseases of the heart and respiratory system) by inhalation of toxic metals in PM₁₀ (cadmium, manganese and cobalt) applying the methodology based on the EPA Risk Guidelines, and (4) to evaluate the potential risk to health represented by the air pollutants studied in two studied sites.

2. Materials and Methods

2.1. Study Area and Sampling Method

Samples were collected during the rainy season (June–July, 2022), taking a sample every 3 days with a duration of 24 h in two sites, the first site located in the downtown of Campeche (the National Institute of Anthropology and History: INAH) (19.843044 N, 90.536749 W), and the second one located in the Tecnologico Nacional Campus Lerma facilities (TNL) (19.79348 N, 90.61634 W). A total of 50 samples were collected during this study (25 samples in each sampling site). The location of the study sites is shown in Figure 1. PM₁₀ particles were collected on quartz filters of 47 mm in diameter (Whatman©), using Minivol samplers (Airmetrics^®) brand, operating at a controlled flow of 5 L min⁻¹ in accordance with the Official Mexican Standard NOM-035-SEMARNAT-1993 [12]. Filters were previously conditioned at a temperature not higher than 30 °C and relative humidity of 50% for 24 h, and weighed on an analytical balance with a precision of 1 mg (Model 130S-F, Sartorius). The weights of the filters were recorded before and after sampling. The gravimetric concentration was obtained by difference, considering the flow and sampling time and the filter area according to the US EPA method IO2.1 [13]. Quality control was carried out by performing simultaneous measurements of three control filters, which were treated in the same way as the samples, at controlled temperature and humidity inside a desiccator.

The TNL site is of particular interest in this study because it is located 500 m from the town of Lerma, where approximately 9000 people live (INEGI, 2020), which are exposed to the negative effects generated by the Power Generation Plant (Thermoelectric plant) located on that site. On the other hand, the INAH site is located in the downtown of the city, on 59th street, where many commercial, educational and touristic activities are carried out. In addition, the study zone is commonly affected by agricultural activities carried out in the surrounding municipalities, causing numerous forest fires that release particulate matter and gases derived from combustion into the atmosphere (PROAIRE, 2019).

2.2. Chemical Analysis

Samples were properly digested and then analyzed by atomic absorption (AA) spectrophotometry based on US EPA-standardized methodologies (methods IO.3.1 and IO.3.2) [14,15]. Prior to analysis, the calibration curves for each metal were prepared. The filters were placed in 150 mL glass beakers, adding 10 mL of aqua regia (25% HNO₃ + 75% HCl) and 1.065% HClO₄, leaving them in contact for 18 h. The content of each glass was heated at 60 °C for approximately 70 min, until almost dry. Subsequently, 20 mL of hot deionized water were added, and then were filtrated using acrodisks, once the contents of each glass were cooled. These solutions were placed in flasks and were gauged to 50 mL using deionized water, after which the samples were stored in polypropylene containers for further analysis [16]. The standard solutions of Cd, Co, Cu, Fe, Mn and Zn were prepared by successive dilution from their respective 1000 ppm HYCEL brand standard solution [17]. For each metal, from each stock solution, 5 dilutions were prepared at different concentrations according to the requirements of each one of the metals, which were gauged to 100 ml and used for the calibration curve of each metal in a concentration range from 0.0035 mg/L to 2 mg/L. A Thermo Scientific™ model iCE 3000 spectrophotometer (AAS Series) was used to analyze the samples. A deuterium lamp was used as a background corrector for all measurements [16,17].

The basic instrumentation for AA spectrophotometer consists of a monochromatic radiation source (specific for each element to be analyzed), an atomizer to produce the excited atoms of the substance to be analyzed; a monochromator to select the wavelength of the characteristic radiation of each element to be analyzed; a detector sensitive to the emitted radiation and an output signal processor. The beam emitted by the source passes through the atomization system that contains the sample in the state of atomic gas, and it reaches the monochromator that eliminates the radiation that is not of interest for this study, thus passing to the absorbed radiation detector, which is then processed and amplified, resulting in an output reading. The type of flame most used for the determination of metals such as Cd, Co, Cu, Fe, Mg, Mn and Zn is obtained using acetylene gas as fuel, which has a temperature between 2100 and 2400 °C, being optimal for carrying to the atoms to their ground state [18].

The wavelengths used to make the determinations were 279 nm, 248 nm, 240 nm, 325 nm, 214 nm and 229 nm for Mn, Fe, Co, Cu, Zn and Cd, respectively. The limits of detection (LD) of trace metals in PM₁₀ were in the range of 0.003–0.056 µg L⁻¹ (

Table 1. Recovery percentage, detection limits (LOD) and uncertainties of the measured trace metals in PM₁₀.

Trace Metals	LD (µg L−1)	Recovery % + SD	urep	usol	ux	UC	UE
Mn	0.020	97.20	0.00224	0.00779	0.00883	0.01173	0.02648
Fe	0.050	98.56	0.00075	0.00820	0.01975	0.02314	0.04325
Co	0.056	97.61	0.01120	0.01298	0.01975	0.02718	0.05135
Cu	0.033	97.40	0.00265	0.00327	0.00454	0.00561	0.01462
Zn	0.010	103.70	0.00126	0.00800	0.00427	0.00842	0.01927
Cd	0.013	99.30	0.00243	0.00235	0.00461	0.00678	0.01396

u_rep: uncertainty associated with repeatability; u_sol: uncertainty associated with solutions preparation; u_x: uncertainty associated with stocks solution of calibration standards; U_C: combined uncertainty; U_E: extended uncertainty.

). The percent recovery of trace metals by the spike method (n = 3) ranged from 97.2% to 103.7% and extended uncertainties ranged from 0.01396 to 0.05135 (Table 1). To verify the reproducibility and low background metal concentrations of the reagents and filters, 5% of the total number of samples was taken as blank and analyzed for the presence of trace metals.

2.3. The Health Risk Assessment

Quantitative cancer risk assessment (CR) was performed for men, women, and children in the age groups 0–2 years and 2–16 years. In this study, only Cd and Co were considered for the evaluation of the cancer risk scenario (Methodology I), since these metals are considered to be carcinogenic according to the EPA [19,20]. For the estimation of non-carcinogenic risk, only cadmium, cobalt and manganese were considered since they are the only metals measured in this study that are evaluated under the EPA-IRIS Program [19].

To determine cancer risk coefficients, exposure was calculated based on the lifetime average daily dose (LAD), considering population groups (men, women and children) and ages (0–2 years and 2 to 16 years). Two daily exposure periods were considered. In the case of the population who live and work or study in the place of exposure, a period of 24 h/day was considered.

For people living at the exposure site (place of residence) and work or study outside that region, a period of 14 h/day was considered. An exposure period of 350 days/year was considered and the values for the inhalation rate (InR) and body weight (BoW) were selected based on the different age groups. (

Table 2. Parameters used in the calculation of exposure according to the considered age groups [19,20].

Parameters	Men	Women	Children (0–2 Years)	Children (2–16 Years)
Inhalation rate (InR, m−3 day−1)	16.4	12.6	4.9	10.8
Body weight (BoW, kg)	76	63	10.3	32.5

) [19,20].

2.3.1. Methodology I: Cancer Risk Estimation for Cobalt and Cadmium

The LAD (Equation (1)) is the inhaled amount of a specific chemical per kg of body weight that is likely to cause health effects once absorbed into the human body over a prolonged period of time [21]. Equations (1) and (2) are used to estimate the LAD (lifetime average daily dose) and are as follows:

$$\mathrm{L}\mathrm{A}\mathrm{D}=\mathrm{E}\mathrm{x}\times \mathrm{C}$$

(1)

where LAD is determined from the measured concentration (C) of metals in PM₁₀ in the study site (mg m⁻³) and the exposure (Ex) in mg kg⁻¹ day⁻¹ from Equation (2):

$$\mathrm{E}\mathrm{x}=\frac{\mathrm{I}\mathrm{n}\mathrm{R}\times \mathrm{T}\times \mathrm{F}\times \mathrm{D}}{\mathrm{B}\mathrm{o}\mathrm{W}\times \mathrm{A}\mathrm{v}\mathrm{T}\times 365 \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s}/\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}}$$

(2)

where InR in m³ h⁻¹ is the air inhalation rate; T is the exposure period (24 h day⁻¹); F is the exposure frequency (350 day years⁻¹); D (years) is the duration of exposure (for D, used the most conservative assumption is used, and it means that the duration of exposure is equivalent to the period of a person’s life expectancy); BoW (Kg) is body weight; AvTc (years) is the mean time to carcinogens; and AvTn (years) is the average time for non-carcinogens [21]. CRc represents the increased probability of cancer occurrence above the general average due to inhalation of carcinogenic substances (cancer risk). The risk range in humans considers CRc values from 10⁻⁴ (lifetime risk of developing cancer is 1 in 10,000) to 10⁻⁶ (lifetime risk of developing cancer is 1 in 1,000,000). CRc values less than 10⁻⁶ indicate that the cancer risk is within an acceptable level, while a cumulative cancer risk greater than 10⁻⁴ is not tolerable. The acceptable limit value is 10⁻⁶. For the carcinogenic metals tested in this study (Co and Cd), the cancer risk (CRc) was determined as reported in Equation (3) [21,22].

$$\mathrm{C}\mathrm{R}\mathrm{c}=\mathrm{L}\mathrm{A}\mathrm{D}\times \mathrm{C}\mathrm{S}\mathrm{F}$$

(3)

where CRc is the probability of cancer occurring in the exposed population considering a lifetime of 70 years, and is determined from the LAD (lifetime average daily dose) (mg kg⁻¹day⁻¹) and the CSF (cancer slope factor) (mg kg⁻¹day⁻¹) [21,22]. Carcinogenic risks were defined as the incremental probability that a person will experience cancer over the lifetime, due to exposure to a specific potential carcinogen (i.e., an incremental increase in cancer over the lifetime). The CSF has been calculated according to Equation (4) [22]:

$$\mathrm{C}\mathrm{S}\mathrm{F}=\frac{\mathrm{I}\mathrm{U}\mathrm{R}\times \mathrm{B}\mathrm{o}\mathrm{W}}{(\mathrm{I}\mathrm{n}\mathrm{R}\times \mathrm{T})}\times 1000$$

(4)

The inhalation unit risk (IUR) is the upper limit of the excess lifetime risk of cancer, which is estimated from continuous exposure to a carcinogenic chemical at a concentration of 1 mg/m³ in air. IUR values are reported in the US EPA database [20,22,23]. The IUR and RfC (reference concentration) values for the considered metals are shown in

Table 3. Inhalation unit risk (IUR) and reference concentration (RfC) data from the USEPA database [20,22,23,24].

Metal	CAS	IUR (μg m−3)−1	RfC (mg m−3)
Cd	7440-43-9	1.8 × 10−3	1 × 10−5
Co	7440-48-4	9 × 10−3	6 × 10−6
Mn	7439-96-5	-	5 × 10−5
Ni	7440-02-0	2.6 × 10−4	5 × 10−5

[20,22,23,24].

2.3.2. Methodology 2: Estimation of the Non-Carcinogenic Risk for Cadmium and Cobalt and Manganese

The quantitative non-carcinogenic risk assessment (THQ: total hazard quotient) was performed for men, women, and the two groups of children living in the Campeche area. We calculate the risk quotient (THQ) (dimensionless) from Equation (5).

$$\mathrm{T}\mathrm{H}\mathrm{Q}=\frac{\mathrm{A}\mathrm{D}\mathrm{I}}{\mathrm{R}\mathrm{f}\mathrm{D}\mathrm{i}}$$

(5)

where ADI is the average daily intake (mg kg⁻¹day⁻¹), that is the estimated dose to which the recipient is exposed by a route of exposure; RfDi is the reference concentration (mg kg⁻¹day⁻¹), which is the dose, via a given route (inhalation in this case) that is believed to have no effect; the cumulative THQ should be seen as the sum of the THQ calculated as in Equation (5) for each pollutant. The THQ considers that there is an exposure level (RfDi) below which the population (including vulnerable groups) is unlikely to experience adverse health effects. If the exposure level (ADI) exceeds unity, there may be concern for possible non-carcinogenic effects; the highest THQ values (greater than unity) constitute the highest levels of concern [25], indicating that inhalation of these metals may be associated with the development of cardiovascular and respiratory diseases. RfDi (mg kg⁻¹day⁻¹) represents r, the inhalation dose at which no effects occur and is defined as:

$$\mathrm{R}\mathrm{f}\mathrm{D}\mathrm{i}=\frac{\mathrm{R}\mathrm{f}\mathrm{C}\times \mathrm{I}\mathrm{n}\mathrm{R}\times \mathrm{T}}{\mathrm{B}\mathrm{o}\mathrm{W}}$$

(6)

ADI (mg kg⁻¹day⁻¹) is the estimated dose received by the recipient due to exposure to air and is calculated according to the following Equation (7):

$$\mathrm{A}\mathrm{D}\mathrm{I}=\mathrm{E}\mathrm{x}\times \mathrm{C}$$

(7)

where C is the concentration in (mg m⁻³) and Ex is the exposure (mg kg⁻¹ day⁻¹) [21,22].

2.4. Statistical Analysis

Descriptive statistical analysis were carried out for the PM₁₀ and PM₁₀-bound trace metal concentrations data. Non-parametric tests (the Friedman test) were applied at a significance level of alpha = 0.05 in order to determine if there were significant differences between the measured variables for both sampling sites.

A multivariate analysis (principal component analysis: PCA) was conducted to determine if there were associations among the measured variables. PCA was performed to reduce the number of variables within a large set to a small set of uncorrelated variables called principal components. Chemical components with common emission sources present a strong correlation with each other. PCA examines the number of factors and quantifies the loading of each factor, considering all measured variables. These factors are used to explain the variability of changes in the data analyzed when there are a large number of samples. Higher factor loading values indicate greater representativeness. Thus, the two principal components contribute to the greatest variability in the data.

Principal components are a normalized linear combination of the original variables of a dataset [26]. The principal components are eigenvectors that are taken from the correlation matrix (where the elements of the diagonal are equal to 1), not from covariances (since with standardized variables both matrices coincide). Generally, it will be possible to obtain as many different principal components as there are variables available. The choice is made so that the first principal component is the one that collects the greatest variance; the second must collect the maximum variability not collected by the first, etc., choosing a number that collects a sufficient percentage of total variance. The eigenvalues are calculated, which are the values with which the eigenvector is multiplied and which give rise to the original vector. Eigenvalues measure the amount of variability retained by each principal component (being greater for the first principal component than for the rest), so they can be used to determine the number of principal components to retain. An eigenvalue > 1 indicates that the principal component explains the greatest variability in the standardized data [27]. The number of significant factors takes into account the cumulative variance (>50%) and the eigenvalue (must be greater than 1). It is common to represent the two main components in a biplot. All data analyses were performed using XLSTAT 2017 software.

2.5. Enrichment Factors Analysis

The contribution of anthropogenic sources to metal levels measured in PM₁₀ was determined using the enrichment factors (EF), which are a useful tool to evaluate the degree of enrichment of a single element in comparison to its relative abundance in the crust [28]. Because iron is the metal that is present in the greatest proportion in the crust, it is usually used as a tracer of natural origin. The average compositions of the upper continental crust (UCC) for the metals studied were taken from the study by Wedepohl [29]. EF was calculated using the following equation:

$$EF=\frac{{\left(\frac{X}{Ref}\right)}_{PM10}}{{\left(\frac{X}{Ref}\right)}_{UCC}}$$

(8)

where ${\left(\frac{X}{Ref}\right)}_{PM10}$ is the quotient between the concentration of the considered element present in PM₁₀ and the concentration of the reference element (Fe was used as the reference metal in this study). ${\left(\frac{X}{Ref}\right)}_{UCC}$ is the quotient between the concentration of the considered element in the UCC and the concentration of the reference element in UCC. The average upper continental crustal (UCC) compositions (ppm) for the studied metals are de following; 0.102 for Cd, 11.6 for Co, 14.3 for Cu, 30,890 for Fe, 527 for Mn and 52 for Zn [29]. Metals with EF values close to 1.0 have a probable natural origin (the crust can be considered the main contributor), while metals with higher values are strongly influenced by anthropogenic sources [21]. Di Vaio et al. [21] and Mianka et al. [30] reported a classification that relates the enrichment factor with the probable origin of a considered element, according to

Table 4. Classification of enrichment factors according to the origin of a considered element [21,28].

EF Value	Probable Origin of a Considered Element
<10	Suggest that the element crustal origin
10–100	The element concentration was influenced by an anthropogenic source (suggest a moderate enrichment)
100–1000 (or higher)	It can be considered that the element was highly enriched by anthropogenic sources

[21,29].

3. Results

3.1. PM₁₀ and PM₁₀-Bound Trace Metal Concentrations

Table 5. Descriptive statistics for PM₁₀ concentrations in both sampling sites.

	Mean (µg/m3)	Maximum (µg/m3)	Minimum (µg/m3)	Standard Deviation (µg/m3)
INAH site	41.043	106.016	10.250	26.589
TNL site	28.591	76.659	2.47	22.014

shows the descriptive statistics for PM₁₀ concentrations in both sampling sites. The mean value of PM₁₀ concentration was higher in the INAH site than the TNL site. It can be attributed to the fact that this site is located within the downtown of Campeche city and many commercial activities are carried out in this area. Additionally, government buildings, banks and schools are located in this area, causing frequent intense vehicular traffic, which may have contributed to the recorded PM₁₀ levels. Figure 2a,b show the PM₁₀ concentration values and their comparison with the maximum permissible limit established in the current Mexican standard. As can be observed, both sampling sites presented 3 exceedances of the maximum permissible limit established by the Mexican Air Quality Regulation (NOM-025-SSA1-2021) [31] for the protection of the health of the population.

It can be observed in Figure 3 that trace metal concentrations showed the following relative abundance for the INAH site: Fe > Zn > Co > Mn > Cu > Cd. Fe was the most abundant trace metal in PM₁₀ in the TNL site, followed by Mn and Zn. Cd showed the lowest values of concentrations in both study sites.

According to the Friedman test, the measured trace metals showed significant differences among each other (p < 0.05), indicating that they probably had different sources. Cd has been classified as carcinogenic in humans within the group 1 [23]. We found mean values for this metal of 0.033 µg m⁻³ and 0.024 µg m⁻³ for the INAH and TNL sampling sites, respectively. These values were 5- to 6-fold higher than the limit values established by European Union regulations (5 ng m⁻³) [32]. The Mn mean values were 0.256 µg m⁻³ and 0.432 µg m⁻³ for the INAH and TNL sampling sites, respectively. These values exceeded almost 2-fold the limit value established by the World Health Organization (150 ng m⁻³) [31].

Table 6. Comparison of the results for trace metals in PM₁₀ of the present study with other authors.

Study Sites/Air Pollutants	Cd	Co	Cu	Fe	Mn	Zn
	(µg m−3)	(µg m−3)	(µg m−3)	(µg m−3)	(µg m−3)	(µg m−3)
INAH site	0.033	0.181	0.148	1.273	0.256	0.374
TNL site	0.024	0.116	0.122	2.214	0.432	0.218
Leon, Mexico (Ceron et al. 2019) [39]	0.099	0.120	0.241	2.337	0.136	1.433
Puebla, Mexico (Morales et al. 2013) [34]	0.005	0.0018	0.0419	1.2025	0.0335	-
Riohacha, Colombia (Rojano et al. 2023) [36]	0.001	0.001	0.065	1.419	0.014	0.001
Acerra, Italy (Di Vaio et al. 2018) [21]	0.0038	0.0025	0.0192	1.285	0.0508	0.0457
Constantine, Algeria (Ali-Khodja et al. 2017) [38]	-	-	0.630	4.110	-	1.47
Bolpur, India (Ghosh et al. 2017) [35]	0.0026	0.00002	-	-	0.028	0.204
Shangai, China (Wang et al. 2013) [37]	0.002	-	0.034	1.238	0.069	0.382
Pune, India (Yadav and Satsangi, 2012) [33]	0.240	0.530	0.339	2.09	0.14	0.43

shows the comparison of the found results in this study for PM₁₀-bound trace metal concentrations with those reported in other studies around the world. The concentrations of Cd and Co in this study are lower compared to those recorded in the city of Pune, India [33]. The average values of Co are higher in INAH and Lerma, than those reported in previous studies of cities such as Puebla [34], Bolpur [35], Acerra [21] and Riohacha [36]. However, the average Co values corresponding to the Lerma site are similar to those reported for Leon, Guanajuato.Comparison of Mn values with those of other regions of the world suggests higher values at the INAH site during the rainy season compared to Bolpur, India [35] and Riohacha, Colombia [36]. The average Fe values are in the same order of magnitude as many other cities in the world such as Riohacha [36], Shanghai [37] and Acerra in Italy [21] (except Constantine in Algeria [38], which reported values of 4.110 µg m⁻³ in 2017). Cu levels at both study sites indicate higher values than previous studies conducted in cities such as Puebla [34], Riohacha [36], Acerra [21] and Shanghai [37]. The Zn values in this study are low compared to those reported in Leon, Guanajuato [39] and Constantine, Algeria [38]. However, they are higher than those reported in Acerra, Italy [21] and Riohacha, Colombia [36].

3.2. Enrichment Factors

The enrichment factors (EF) of the measured trace metals in PM₁₀ are showed in Figure 4. It can be observed in Figure 4a that Cd showed the highest EF values. Co and Cu showed EF values between 100 and 1000 and higher, indicating that these metals were highly enriched by anthropogenic sources. Mn in the INAH sampling site was moderately enriched by anthropogenic sources, whereas this metal in the TNL sampling site was probably crustal. Mn and Zn in TNL showed EF values lower than 10, indicating that these metals were probably crustal (Figure 4b). In the case of Co, its enrichments can be attributed to industrial activities located in the study area. Cd enrichment can be associated with high temperature and coal combustion, waste incineration and steel and plastic production.

3.3. Principal Component Analysis

Principal component analysis (PCA) was used to identify correlation patterns between the measured metals, with the aim of reducing the original dataset to a smaller set of principal factors. The main components were determined using the XLSTAT statistical software 2017.4, taking into account only those components with an eigenvalue greater than 1, being able to infer the probable sources of metals in PM₁₀. PCA results are showed in Figure 5 and Figure 6 for the INAH and TNL sampling sites, respectively. As can be observed, two factors (F1 and F2) were necessary to explain 58.41% of the total variability of the dataset in the INAH sampling site (Figure 5). Factor 1 accounted for 37.511% of the total variance with an eigen value of 2.251. This factor was highly correlated with Fe and Mn, which are tracers of road dust sources [40,41]. Factor 2 contributed with 20.902% of the total variance and an eigen value of 1.254. This factor showed a high correlation with Co and Cd. These elements are associated with industrial emissions and application of coatings and paints [42]. Zn, Mn and Fe showed a significant correlation, indicating that these metals could have common sources (crustal and road dust sources). Co and Cu showed a significant correlation indicating that these elements could be from anthropogenic sources.

Two principal factors (F1 and F2) were required in order to explain 55.75% of the total variability of the dataset in the TNL site (Figure 6). Factor 1 contributed with 34.346% to the total variance in the dataset with an eigen value of 2.061. Factor 2 accounted for 21.403% of the total variance with an eigen value of 1.284. As can be observed in Figure 6, Mn, Fe, Cu and Zn showed a significant correlation indicating that these elements could be from common sources (crustal, mobile sources and re-suspension of dust) [43,44]. Zn and Cu have been associated with the mechanical wear of vehicles and tire wear [45]. Co and Cd showed a different behavior, indicating that they had their origin in different anthropogenic sources. Cd may be associated with coal combustion processes and waste incineration [24]. Co is associated with application processes of coatings and paints [42].

3.4. The Health Risk Assessment

In this work, the risk of developing cancer in a lifetime and non-carcinogenic effects (respiratory and cardiovascular diseases) were estimated in different age groups in the population exposed to heavy metals associated with PM₁₀ in two urban sites of Campeche city.

Table 7. Cancer risk coefficients for exposure to cadmium and cobalt in both study sites.

Risk Coefficient	Cancer Risk Coefficient (CR)
Trace Metal	Cd (CR)		Co (CR)
Exposed Population Age Group and Gender	Cd INAH	Cd TNL	Co INAH	Co TNL
Men	Mean:	Mean:	Mean:	Mean:
	4.48 × 10−8	3.36 × 10−8	4.48 × 10−8	3.36 × 10−8
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	6.59 ×10−9	3.34 × 10−9	5.28 × 10−8	8.90 × 10−8
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	8.61 × 10−8	5.51 ×10−8	2.66 × 10−6	1.43 × 10−6
Women	Mean:	Mean:	Mean:	Mean:
	4.48 × 10−8	3.36 × 10−8	4.48 × 10−8	3.36 × 10−8
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	6.11 × 10−9	3.34 × 10−9	4.90 × 10−8	8.90 × 10−8
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	7.98 × 10−8	5.51 × 10−8	2.47 × 10−6	1.43 × 10−6
Children (age from 0 to 2 years)	Mean:	Mean:	Mean:	Mean:
	9.03 × 10−7	2.09 × 10−6	9.03 × 10−7	2.09 × 10−6
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	1.49 × 10−8	8.90 × 10−8	1.17 × 10−7	8.90 × 10−8
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	2.95 × 10−7	1.42 × 10−6	5.86 × 10−6	1.42 × 10−6
Children (age from 2 to 16 years)	Mean:	Mean:	Mean:	Mean:
	4.48 × 10−8	3.36 × 10−8	4.48 × 10−8	3.36 × 10−8
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	1.01 × 10−8	3.34 × 10−9	8.14 × 10−8	8.90 × 10−8
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	1.33 × 10−7	5.51 × 10−8	4.10 × 10−6	1.42 × 10−6

shows the results of the cancer risk coefficient for exposure to cadmium and cobalt in both study sites. It can be seen that the value of CR exceeds the limit recommended by the US EPA (1.0 × 10⁻⁶) only for the population group of children from 0 to 2 years of age for both metals (cadmium and cobalt) at the TNL site so this population group could develop cancer in their lifetime. It should be noted that this site is located in the vicinity of the Lerma thermoelectric power plant, which could be the main source of PM₁₀ and associated metals.

The risk of developing non-carcinogenic effects due to the inhalation of Cd, Co and Mn in the exposed population at both study sites is high, since in all cases the limit recommended by the US EPA of 1.0 is exceeded, being higher for Co and Mn at the INAH (center site) and TNL sites, respectively (

Table 8. Non-cancer risk coefficients for exposure to cadmium, cobalt and manganese in both study sites.

Risk Coefficient	Non-Cancer Risk Coefficient (HQ)
Trace Metal	Cd (HQ)		Co (HQ)		Mn (HQ)
Exposed Population Age Group and Gender	Cd INAH	Cd TNL	Co INAH	Co TNL	Mn INAH	Mn TNL
Men	Mean:	Mean:	Mean:	Mean:	Mean:	Mean:
	3.2	2.3	28.9	18.5	4.9	8.3
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	0.013	0.010	0.34	0.15	0.046	0.098
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	0.66	0.46	1.92	0.97	0.33	0.76
Women	Mean:	Mean:	Mean:	Mean:	Mean:	Mean:
	3.2	2.3	28.9	18.5	4.9	8.3
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	0.012	0.010	0.34	0.15	0.046	0.098
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	0.66	0.46	1.92	0.97	0.33	0.76
Children (age from 0 to 2 years)	Mean:	Mean:	Mean:	Mean:	Mean:	Mean:
	3.2	2.3	28.9	18.5	4.9	8.3
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	0.012	0.095	0.21	0.17	0.094	0.043
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	0.64	0.55	1.53	0.71	0.68	0.54
Children (age from 2 to 16 years)	Mean:	Mean:	Mean:	Mean:	Mean:	Mean:
	3.2	2.3	28.9	18.5	4.9	8.3
	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:	5% Lower Confident Limit:
	0.012	0.095	0.21	0.17	0.094	0.043
	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:	95% Upper Confident Limit:
	0.64	0.55	1.53	0.71	0.68	0.54

).

Table 7 shows the results of the risk coefficients for cancer (CR) for both sampling sites for different exposed population groups. The risk of developing cancer in both sampling sites was higher in the population group of children ages from 0 to 2 years and the higher cancer risk coefficients were observed in this age group due to the inhalation of Cadmium in both study sites (this metal has been classified by the International Agency for Research on Cancer (IARC) as carcinogenic to humans within the Group 1) [46,47].

Table 8 shows the results of the non-cancer risk coefficients (HQ) for both sampling sites for different exposed population groups. As can be seen, the risk of developing cardiovascular and respiratory diseases due to the inhalation of the measured metals is high, being higher for cobalt in both study sites. Exposure to cobalt can cause an asthma-like allergy, difficulty breathing, wheezing, cough, and/or chest tightness. This metal can affect the heart, thyroid, liver and kidneys, and if exposure is frequent it can cause scarring of the lungs (fibrosis) [48].

According to Sexton and collaborators [49], cancer risk can be classified into three categories: definitive cancer risk: CR value > 1.0 × 10⁻⁴; probable cancer risk: if the CR value is between 1.0 × 10⁻⁵ and 1.0 × 10⁻⁴; possible cancer risk: if the CR value is between 1.0 × 10⁻⁵ and 1.0 × 10⁻⁶. Therefore, the child population in the age range from 0 to 2 years residing in and near the study sites in the municipality of San Francisco de Campeche is at a possible risk of suffering from cancer due to exposure to cadmium and cobalt present in PM₁₀. This can be explained, since the younger population turns out to be more vulnerable to the toxic compounds that they inhale. The non-cancer risk coefficients (HQ) were greater than 1.0 in both sampling sites, exceeding the permissible limit established by the EPA, constituting a potential risk of developing diseases of the respiratory and cardiovascular systems, with Co being the metal that represents the greatest danger by acute exposure.

Table 9. Comparison of the lifetime risk coefficients of developing cancer (CR) and non-cancer risk coefficients (HQ) for this study with the reported values by other authors.

Risk Coefficients	CR		HQ
	Cd	Co	Cd	Co	Mn
This study	INAH: 4.68 × 10−8 TNL: 3.36 × 10−8	INAH: 4.68 × 10−8 TNL: 3.36 × 10−8	INAH:3.1 TNL:2.2	INAH: 28.9 TNL:18.4	INAH: 4.9 TNL: 8.3
Di Vaio et al. 2018 (Acerra, Italy) [21]	2.54 × 10−6	7.40 × 10−6	0.14	0.14	0.29
Romanazzi et al. 2014 (Torino, Italy) [46]	1 × 10−5	1 × 10−7	1.5	0.001	0.10
Pongpiachan et al. 2018 (Bangkok, Thailand) [50]	1 × 10−3	5.71 × 10−6	4.13 × 10−6	1.57 × 10−4	6.0 × 10−3
Cerón et al. 2019 (Leon, Guanajuato, Mexico) [39]	2.17 × 10−8	7.07 × 10−7	0.450	1.285	0.169
Debanada et al. 2020 (Seoul, Korea) [51]	-	-	-	-	0.138
Truong et al. 2022 (a southern Vietnam megacity) [52]	-	-	-	-	6

shows a comparison of the obtained results for cancer risk and non-cancer risk coefficients for this study with other studies carried out by other authors. It can be observed that the risk cancer coefficients (CR) for Cd are similar than those reported in Leon, Guanajuato, Mexico [39] but lower than those registered in Acerra [21] and Torino in Italy [46], and those reported in Bangkok, Thailand [50]. Cobalt showed similar values for the cancer risk coefficients (CR) than those reported for Torino, Italy [46] and Guanajuato, Mexico [39], but lower than those found in Acerra, Italy [21] and Bangkok, Thailand [50]. Non-cancer risk coefficients (HQ: hazard quotients) were higher in this study in both sampling sites than those reported in Leon, Guanajuato, Mexico [39], in Acerra [21] and Torino in Italy [46], in Bangkok, Thailand [50], in Seoul, Korea [51] and in a southern Vietnam megacity [52].

The differences in the reported values are due to the fact that the methodologies commonly used to determine these parameters consider different values for the inhalation rate and body weight, which are key parameters for estimating the exposure dose. On the other hand, exposure depends on other factors such as age, sex and activity patterns. Age constitutes a key parameter in risk studies because during the first years of life, individuals are more vulnerable to the effects of toxic compounds. Other authors have also found that age is a key factor in determining cancer and non-cancer risk ratios [19,53,54,55].

4. Discussion

Mean PM₁₀ concentration values were lower than the WHO daily guideline value of 45 µg/m³ [56].

Table 10. Comparison of the PM₁₀ concentration values for this study with the reported values by other authors.

Author	PM10 (µg/m3)	Location
This Study	INAH: 41.043	Downtown area in Campeche, Mexico
This Study	TNL: 28.591	Lerma, Campeche, Mexico
Jena and Singh (2017) [57]	216	Dhanbad, India
Cerón-Bretón et al. (2019) [39]	27.992	Leon, Guanajuato, Mexico
Li et al. (2021) [58]	97.7	Arequipa, Peru
Kunt et al. (2023) [59]	93.2	Kayseri Province, Turkey
Smichowski et al. (2005) [60]	44	Capana, Argentina
Nuchdang et al. (2023) [61]	65.90	Pathum Thani, Thailand
Rodriguez et al. (2004) [62]	33.3	Onda, Spain

shows a comparison among the mean PM₁₀ concentrations found in this work with the reported values in other studies. Both sampling sites showed lower mean values for PM₁₀ than those found in Dhanbad, India [57], Arequipa, Peru [58], Kayseri, Turkey [59], Capana, Argentina [60] and Pathum Thani, Thailand [61]. Mean PM₁₀ concentrations found in the INAH sampling site were higher than those reported for Onda, Spain [62] and Leon, Guanajuato, Mexico [39]. The TNL sampling site showed similar mean concentration values of PM₁₀ to those reported in Leon, Guanajuato, Mexico [39].

Fe was the most abundant trace metal in PM₁₀, as expected, since Fe is the most abundant trace metal in the crust, whereas Cd showed the lowest average concentration value in this study, although this metal exceeded the maximum limits established by the European Union and the World Health Organization (Section 3.1). Average concentrations for the measured trace metals and the enrichment factors (Section 3.2) were higher in the INAH sampling site than the TNL sampling site. The INAH site is in the center of the city, where shopping centers, banks and government offices are located. This causes intense vehicular traffic frequently in this area, which may have contributed to the trace metals concentrations in PM₁₀. On the other hand, the TNL site is located on the outskirts of the city and at a higher altitude, so the dispersion is better due to altitude.

The enrichment factors analysis showed that all Cd, Co and Cu were highly enriched by anthropogenic sources in both sites. Mn showed a different behavior in both sampling sites, originating from crust in the TNL site (being the main source the entrainment of Mn-containing soils), whereas this element was moderately enriched by anthropogenic sources in the INAH site. This can be explained by the fact that this metal is released to the environment from fossil fuel combustion and sewage sludge incineration (typical urban sources).

Figure 7 shows the correlation between the enrichment factors for each trace metal with wind direction for the INAH and TNL sampling sites.

It can be observed in Figure 7b–d that in the INAH site, Cu, Zn and Mn showed higher EF when wind came from ENE, and Co (Figure 7b) and Cd (Figure 7a) showed higher EF values when winds came from SE and E, respectively. Residential and commercial areas are located to the ENE of this sampling site. To the east and southeast, there are avenues with high vehicular traffic (Francisco I. Madero, Cuauhtemoc and Luis Donaldo Colosio and Murrieta Avenues), and all of these sources could influence metal concentrations in PM₁₀ in this site.

In Figure 7e,f, it can be observed for the TNL site that Co and Cd had higher EF values when winds came from SE, Mn and Zn (Figure 7g,h) showed higher EF when winds came from E and the highest EF values were found for Cu when winds came from ESE. The Pablo Garcia and Montilla periphery, one of the main access roads to the City of Campeche, is located southeast of this site, so vehicular sources could have contributed to the Co and Cd levels. Air masses coming from the thermoelectric power plant located ESE of this site could have caused Cu enrichment. Transport of dust particles from non-urbanized areas and unpaved roads located to the E of this site could have caused the levels of Mn and Zn.

The main emission sources at the study sites are related to the re-suspension of road dust, industrial emissions and the burning of bark and biomass. These emission sources are the result of multiple activities carried out at the sampling sites and include medium-scale industries such as electric power generation plants, as well as commercial establishments, farming areas, and residential areas. According to the criteria air pollutants inventory base year 2016 (https://gisviewer.semarnat.gob.mx/wmaplicacion/inem/) (accessed on 10 September 2022), the main PM₁₀ source emissions in Campeche city (the INAH site) are: mobile sources (buses, tractor-trailers and heavy vehicles: 552 tons/year), construction activities (329.99 tons/year) and paved and unpaved roads 867.47 tons/year). On the other hand, the main emissions sources in Lerma (the TNL site) include power generation (451.88 tons/year), agricultural sources (391.88 tons/year) and agricultural burning (242.56 tons/year). All these sources could contribute to the levels of PM₁₀ and PM₁₀-bound trace metals in this area.

From the PCA (Section 3.3), it was corroborated that Zn, Fe and Mn probably had common sources at the INAH site (combined sources that includes crustal and road dust). It was possible to infer that Co, Cd and Cu probably had their origin in anthropogenic sources (combustion of fossil fuels and waste) [63,64]. PM₁₀ collected samples in the TNL site showed good correlations among Mn, Fe, Cu and Zn, indicating that these elements could have combined common sources (crustal, car exhaust and re-suspension of dust), whereas we could infer that Co and Cd could be from anthropogenic sources [63,65]. Cobalt is associated with coatings and paint application processes. Cadmium is associated with combustion processes that use carbon as fuel. At the TNL site, there is an electric power generation plant that uses combustoleo and coal, and in the surroundings of this site there are agricultural areas where biomass burning and waste incineration are common practices. For the above, all of these sources could have contributed to the cadmium levels at this study site.

All CR values were lower than the acceptable level (1.0 × 10⁻⁶) except for cobalt (Co) and cadmium (Cd) in the population group of children from 0 to 2 years of age in the TNL site (Section 3.4). It is important to mention that this site is located in the vicinity of the Lerma thermoelectric power plant, so these metals could be from this source. All HQ values were greater than 1, indicating that there is a latent risk of developing respiratory and cardiac diseases in the exposed population, including adverse health effects such as allergic dermatitis, rhinitis and asthma [66]. The results of this study constitute a first approximation and should be considered as preliminary; it is necessary to carry out more studies that consider other factors such as the distribution of mass of particles and fractions of different sizes that can be applied to other models in order to determine the risk to health, which could provide additional information on the health risk potential of trace metals present in PM₁₀ in the study sites.

5. Conclusions

The trace metal concentrations were similar in both sampling sites, with Fe and Mn being the dominant metals present in PM₁₀. Cd and Mn mean concentrations exceeded the maximum permissible levels established by the European standard and the WHO. The enrichment factor analysis revealed that Cd and Co were highly enriched by anthropogenic sources in both sampling sites. Mn, Zn and Fe showed low values for enrichment factors, indicating that these trace metals probably had their origin in crustal.

The PCA revealed that Fe, Mn and Zn had a significant correlation, indicating that these metals could have their origin in common sources (crustal). Co and Cu showed significant correlation, indicating that these elements could be from common sources (anthropogenic). Cd did not show a correlation with the rest of the metals, indicating that this element could have a different origin (industrial emissions).

This study focuses on the estimation of health risks caused by exposure to trace metals in PM₁₀ in the population of two urban sites in San Francisco de Campeche, Mexico. Therefore, the found results will be useful to constitute a base study since there are not enough studies in this regard in this area, and most of them only report gravimetric concentrations of PM₁₀ and the content of these particles is not determined. Health effects on different population subgroups were estimated for trace metals in PM₁₀ through different quantitative methodologies reported in the literature, considering different parameters such as age, sex and the average daily exposure time of the population. The results can serve as a basis to improve the control measures of metal and particle emission sources in the study sites to reduce the risk of the population due to exposure to these air pollutants.

The results demonstrate the need for continuous monitoring of PM₁₀ concentrations and associated trace metals in the study area, especially metals of greater toxicity such as cadmium and cobalt. The health risk analysis revealed that there is a possible risk of developing cancer in a lifetime, being higher in the child population, which is more vulnerable to air pollutants. The cancer risk coefficients at the INAH site were higher, exceeding the maximum allowable limits established by the US EPA and the WHO. The non-carcinogenic risk exceeded the limits recommended by the US EPA, so it can be concluded that there is a possible risk of developing cardiovascular and respiratory diseases in the study sites due to inhalation of the measured metals present in PM₁₀.

It is recommended to monitor the trace metals present in PM₁₀ in other seasons of the year (dry season and northern season) to determine the seasonal variation in these pollutants, as well as to include other metals that are highly toxic, such as Pb, Al and As. It is recommended to include in future assessments other toxic compounds such as volatile organic compounds (VOCs), carbonyls, and dioxins present in the air to evaluate the actual risk that air pollutants represent for the population of this city. The results provided by this work are useful data to be considered for air quality evaluation and contribute to understanding the potential risks due to the exposure to toxic metals in PM₁₀ in an urban area located in the southeast of Mexico.