Introduction

Arsenic (As) is a toxic element that affects human health (Jomova et al. 2011), and hence, the study of arsenic in environmental media (i.e., groundwater, soil, and air) and the associated human health risks has been an active area of research for many decades (Lewis et al. 2012). Arsenic is widely distributed in the environment. Its abundance ranks 20th in the Earth’s crust, with a mean concentration of 1.6 μg/g (Holmes and Miller 2004), 14th in seawater, and 12th in the human body (Mandal and Suzuki 2002; Jomova et al. 2011).

Plenty of studies have studied human exposures to arsenic in drinking water and its adverse health effects due to the relatively high concentrations of naturally occurring arsenic in drinking water in some parts of the world (Chappell et al. 2003). Similarly, the exposure of arsenic arising from soils nearby mining and smelting activities has also been paid attention (Lewis et al. 2012). On the other hand, information regarding exposures associated with arsenic in ambient air is scarcer. Nonetheless, high concentrations of arsenic and other potentially toxic elements of environmental interest may be present in atmospheric aerosol (Sanchez de la Campa et al. 2011). Anthropogenic sources of arsenic are often related with combustion of fossil fuels such as coal, with copper smelting, and with incineration of municipal waste (Pacyna et al. 2007; Fernandez-Camacho et al. 2010). Arsenic is also released to the atmosphere from other smelting metals, use of some pesticides, and some glass raw materials (Sanchez de la Campa et al. 2011). Arsenic aerosol has also been associated with terrestrial crustal, volcanic, and groundwater origins (Sanchez de la Campa et al. 2011). Nevertheless, anthropogenic emissions of arsenic to the atmosphere are about three times higher than natural ones (World Health Organization (WHO) 2000).

In the area of study, which is the main Spanish and second most important European industrial ceramic cluster (Delgado-Saborit and Esteve-Cano 2007), the main source of arsenic aerosol is related to the ceramic industry. Arsenic is present as an impurity in some boracic minerals, namely, colemanite, hydroboracite, and ulexite, which are widely used for reducing the melting point of the ceramic frits (Esteve and Ramos 1999). Vickery et al. (1998) calculated that volatile losses of arsenic trioxide during the ceramic process were in the range of 0.15–0.21 mg/g of frit for colemanite and 0.03–0.07 mg/g of frit for hydroboracite. These very high values highlight the importance of the use of boracic minerals in the ceramic industry as an important source of airborne arsenic in the study area and the possible impact in the exposure to arsenic for the local population, all the same for the occupationally exposed workers (Chen et al. 2007).

Arsenic is considered as one of the most toxic elements for human health (Chappell et al. 2003). Continued exposure to a high concentration of arsenic may produce acute toxic effects on humans, which can be quickly diagnosed. However, chronic exposure to low doses of arsenic may give rise to cancer (Hayes 1997; Roy and Saha 2002; IARC 2009). Arsenic is recognized as a carcinogenic element producing skin, lung, and urinary bladder cancers (Englyst et al. 2001; Cantor and Lubin 2007). In 2000, the WHO Air Quality Guidelines for Europe calculated 1.5 × 10−3 cancer risks for a lifetime exposure to a concentration of 1 μg/m3 (WHO 2000). Several epidemiological studies and the Ariadna Data Base show that the region of Castellón, where the study area is located, holds the fifth and ninth positions in incidence of urinary bladder cancer and lung cancer, respectively, in Spain (Centro Nacional de Epidemiología (CNE) 2011). Both cancers are related with arsenic exposure (Englyst et al. 2001; Cantor and Lubin 2007). The map of cancer occurrence in Spain points out the study region as one with the greater risk of cancer of all Spain (Benach et al. 2003). On the other hand, besides promoting various types of cancer, arsenic produces other adverse effects on human health, such as hypertension, cardiorespiratory disease (Navas-Acien et al. 2005), immunological system disease (Duker et al. 2005), diabetes (Diaz-Villasenor et al. 2007), neurological disorders (Vahidnia et al. 2007), and dermal effects (Cohen et al. 2006).

Given the health effects and its ubiquity in the atmosphere, airborne arsenic aerosol is, nowadays, a pollutant of major interest in the European Union (EU). Arsenic and its ambient air concentration in PM10 are regulated by the EU Directive 2004/107/EC of the European Parliament and of the Council of 15 December 2004 relating to arsenic, cadmium, mercury, nickel, and polycyclic aromatic hydrocarbons in ambient air. This directive proposed a target value for arsenic in PM10 of 6 ng/m3 as an annual mean to be met from December 2012. No limits are established for arsenic in PM2.5. On the other hand, the WHO guidelines on arsenic exposure state that a safe level for inhalation exposure cannot be recommended (WHO 2000) due to its carcinogenic potential.

The present work builds on previous studies by Querol and coworkers that characterized the quality of the air in the ceramic cluster area of Castellón (Querol et al. 2001) and compared it with other Spanish cities (Querol et al. 2004a, b, 2008).Those studies showed the occurrence of relatively high ambient air concentrations of arsenic in PM10 in the area of study (8–25 ng/m3). However, no data on arsenic content in PM2.5 was reported, which has a higher potential of penetration in the respiratory system. The cited studies of Querol et al. (2001, 2004a, b, 2008) were performed in the late 1990s and early part of the 2000s. Since then, several technical actions have been implemented by the ceramic industry to reduce the content of arsenic in the air. Considering several factors listed hereunder, there is the need of reassessing the levels and trends of arsenic in PM10 and characterizing the content of arsenic in PM2.5 in the ceramic area of interest. These factors are (a) the outlined health effects in the area of study, (b) the technological actions implemented, (c) the higher potential of penetration of PM2.5 in the respiratory system, and (d) the recent economic downturn on the ceramic industry undergoing a period of severe reduction in its manufacturing output.

The aim of this work focuses on characterizing the levels and trends from 2006 to 2010 of arsenic on PM10 and PM2.5 and to assess the compliance with the Directive 2004/107/EC in the ceramic industry study area.

Materials and methods

Area of study

Samples were collected in two sites located in L’Alcora (Eastern Spain), a city which is located within the ceramic industry cluster. The cluster is a geographical triangle of 300 km2 containing more than 200 ceramic industries producing 94 and 43 % of the Spanish and the European tiles, respectively (Querol-Balaguer et al. 2004; Delgado-Saborit and Esteve-Cano 2007). Further details of the area of study can be found in Delgado-Saborit and Esteve-Cano (2007).

Aerosol sampling and sample preparation

PM10 and PM2.5 samples were collected at two sampling sites at L’Alcora. The first site (site 1—industrial) was located near a municipal sports center within an industrial area (0°12′24″ W, 40°4′05″ N, 240 m a.s.l.). The second site (site 2—urban background) was located at a retirement home (0°12′44″ W, 40°4′25″ N, 260 m a.s.l.), considered as urban background. At the first station, PM10 sampling was carried out by means of a medium volume Partisol sampler model 2000-H (Rupprecht and Patashnick, USA) furnished with a PM10 inlet operating at 2.3 m3/h. At the second station, PM10 samples were collected using a high-volume sampler (model MCV-AV; MCV, Spain) operating at 30 m3/h furnished with a Digitel PM10 inlet. PM2.5 samples were collected using a medium volume sampler (Derenda Medium Volume Sampler Model MV 56.1 LV53.1, Germany) furnished with a PM2.5 inlet. All sampling equipment was made of noncontaminating materials such as anodized aluminum, stainless steel, polyethylene, and polypropylene.

Whatman GF/A 150 mm (for MCV sampler) and Whatman QMA 47 mm (R&P and Derenda samplers) quartz glass microfiber filters were used in media collection. During a 5-year period from 1 March 2006 to 31 December 2010, a total of 3,272 24-h samples were collected, in which 1,241 were PM10 samples collected at site 1 (industrial), while 1,310 were PM10 and 721 were PM2.5 samples collected at site 2 (urban background). At the end of every 24-h sampling period, the aerosol samples were sealed in plastic bags and were brought back to the laboratory for chemical analysis. In the laboratory, samples were conditioned (20 ± 1 °C; 50 ± 5 % humidity) for 24 h in a desiccator before weighing. For chemical analysis, one sample corresponding to a day per week was selected randomly for every sampler; therefore, a total of three samples per day were selected. A total of 275 PM10 and 99 PM2.5 samples were analyzed.

Elemental analysis

Elemental analysis of PM10 and PM2.5 was performed. After gravimetric determination of PM levels, filters were digested in an acid media (HNO3, 65 %, J.T. Baker and H2O2, 33 %, PA-ACS-ISO). Extractions were carried out with 10 ml of acid and 10 ml of MQ ultrapure water, using Teflon jars and a Merck microwave oven (Esteve and Peris 2000). Once samples were digested and filtered with a 0.45-μm polypropylene membrane syringe filter and diluted to 25 ml using MQ ultrapure water, arsenic levels were analyzed with an inductively coupled plasma–mass spectrometer (ICP-MS) at the Technical–Scientific Services of the Universitat Jaume I. The ICP-MS (Agilent 7500 CX) contains a collision cell that removes any polyatomic interference, especially those derived from Ar and Cl. The limit of detection for arsenic was 0.073 ng/ml, which, when converted to atmospheric concentration, represents 0.05 ng/m3 for a sampling volume of 36 m3.

A QA/QC protocol was enforced to account for possible arsenic traces in reagents and quartz glass filters analyzing blank filters and to control the performance of the analytical technique by analyzing a certified reference material and a standard control check. Extraction efficiency was assessed by analyzing a sample of the certified reference material SRM 1648 (urban particulate matter standard) (NIST, Gaithersburg, MD, USA), using a half filter loaded with about 5 mg of SRM 1648 and digested with the other samples in every batch. Recovery rates ranged between 85 and 91 %. In each batch of samples, a blank filter, a sample of the standard reference material, and a 10-ppb standard alongside the samples (N = 12) were analyzed. The analyst was blinded in the knowledge of which vials were samples, blanks, reference materials, or standard check.

Results and discussion

PM10 and PM2.5 concentrations

Figure 1 shows the annual averages of PM10 (industrial and urban sites) and PM2.5 (urban site) and the number of exceedances of the current regulation per year. The values of PM10 measured in the urban background site during the period 2006–2010 are similar, albeit decreasing, to those reported previously by Querol et al. (2007) for the same area for the period 2002–2005. The values measured at the industrial station are higher than those measured at the urban background site (Fig. 1), which reflects the influence of the ceramic industry emissions on the site. The PM10 levels measured at the background and industrial sites are consistent with values reported for other areas in Spain, with levels ranging from 30 to 40 μg/m3 in urban background sites and levels ranging from 46 to 50 μg/m3 in industrial areas (Querol et al. 2008).

Fig. 1
figure 1

Evolution of annual mean concentrations PM10–PM2.5 from years 2006 to 2010

Full size image

Levels of PM2.5 measured at the area of interest (Fig. 1) are lower than those measured at other urban background areas in Spain, the latter ranging from 20 to 30 μg/m3, while they are similar to those measured at rural background areas (in micrograms per cubic meter) in Spain (Querol et al. 2008) and Puerto Rico (Figueroa et al. 2006), but lower than PM2.5 levels in Italy (Marcazzan et al. 2001), China (Ho et al. 2002), or Taiwan (Fang et al. 1999).

The decreasing pattern of PM10 and PM2.5 measured during the study period reflect the effect of the combination of measures that ceramic manufacture had already set in place in the previous years (Celades et al. 2012). Nonetheless, this decreasing trend might also be attributed to the fact that the economic crisis considerably hit the ceramic sector with a strong reduction in manufacturing, which has halved the production to levels similar to those experienced in 1995 (Celades et al. 2012). This is consistent with a recent study which documented that levels of PM10 followed a similar trend of economic indicators (Arruti et al. 2011).

To assess the degree of compliance in the area of study for the period 2006–2010, the PM10 and PM2.5 concentrations measured were compared with the guidelines given in the European Directive 2004/107/CE of the European Parliament on ambient air quality, the Directive 2008/50/EC of the European Parliament and of the Council of 21 May 2008 on cleaner air for Europe, and the Spanish R.D. 102/2011.

Annual PM10 levels in the industrial site were higher than the limit value proposed by the European Directive 1999/30/EC (i.e., annual average below 20 μg/m3 and a maximum of seven exceedances of the level of 50 μg/m3/year). However, the PM10 levels measured in 2009 and 2010 at the urban background site would meet the new European Directive 2008/50/EC and Spanish regulation R.D. 102/2011, which sets a limit value for PM10 of 40 μg/m3, with a maximum of 35 exceedances per year of the value 50 μg/m3. The same regulation established a limit annual value for PM2.5 of 25 μg/m3, which has been met in the area of study. It is likely that values of PM10 and PM2.5 will continue to be within the values established in the current regulation under the following two circumstances. The first one is associated with current low ceramic production, which reduces direct emissions of PM to the atmosphere. The second one is due to recent changes in storage conditions of raw materials that limit fugitive emissions. Therefore, in the event that there is a boost in ceramic production again, further emission reduction measures should be considered, as discussed in detail in the “Evolution in levels of arsenic in the period 2006–2010” section.

Arsenic content in PM10 and PM2.5

Levels of arsenic in the atmosphere are varied. In remote and rural areas, the mean arsenic airborne concentrations range from 0.02 to 4 ng/m3 (WHO 2000), while in urban areas, arsenic concentrations range from 3 to about 200 ng/m3 (Querol et al. 2007; Halek et al. 2010; von Schneidemesser et al. 2010). Typical PM10 arsenic levels in EU countries are currently between 0.2 and 1.5 ng/m3 in rural areas, between 0.5 and 3 ng/m3 in urban areas, and lower than 50 ng/m3 in industrial areas (Putaud et al. 2004). Querol et al. (2004b) reported arsenic PM10 levels for rural, urban, and some industrial sites in Spain in the range of 0.3, 0.3–1.8, and 5 ng/m3, respectively. Table 1 presents the arithmetic mean, maximum, and minimum concentrations of arsenic found in 275 PM10 samples and 99 PM2.5 samples randomly selected to represent the entire sampling period 2005–2010 at both sampling sites.

Table 1 Minimum, mean, and maximum levels of arsenic (in nanograms per cubic meter) in PM
Full size table

The mean arsenic content of the 181 PM10 samples analyzed representative of the period 2005–2007 in both sampling sites (3.33 ng/m3 at the industrial site and 2.25 ng/m3 at the urban background site) is above the usual range of levels (0.3–1.8 ng/m3) found in urban background areas in other Spanish cities (Querol et al. 2004b), with the exception of Huelva industrial area, that show higher values. On the other hand, arsenic levels representative of the period 2008–2010 are below the Spanish average arsenic content (0.99 ng/m3 at the industrial site and 0.93 ng/m3 at the urban background site), coinciding with the economic crisis.

For comparison, concentrations of arsenic in PM10 obtained from the regional air quality network representative of the period 2005–2007 measured at two rural background monitoring stations are 0.42 ng/m3 in San Jorge and 0.76 ng/m3 in Cirat. Arsenic levels representative of the period 2008–2010, coinciding with the economic crisis, are 0.34 ng/m3 at San Jorge and 0.38 ng/m3 at Cirat. These villages are located 90 and 40 km from L’Alcora, respectively. A similar behavior was observed in the rural background sites of Morella and Zorita (Santacatalina et al. 2011) The PM10 concentrations at these locations, which are not directly influenced by the emissions of the ceramic smelter industries, are lower than those found at L’Alcora in the present study and those found at other Spanish urban monitoring stations (Querol et al. 2004b). However, the reduction in arsenic levels measured during the period of the economic downturn shows that the ceramic industry not only impacts the local concentrations but also impacts on the background concentrations in the larger area. The impact of local emissions in areas far from the ceramic cluster might be consequence of the regional transport inland of local emissions associated with the regional meteorology, as described in detail by Querol et al. (2007) and Millán et al. (1997).

Previous studies of arsenic content in PM10 in the area of study showed higher levels than current ones with 25 ng/m3 in 1999 (Querol et al. 2001), 16 ng/m3 in 2002 (Pallares et al. 2007), 9.9 ng/m3 in 2003, 6 ng/m3 in 2004, and 2.5 ng/m3 in 2005 (Minguillon et al. 2009). The lower concentrations of airborne arsenic in PM10 measured since 2004 are associated with the introduction of several technical actions, such as the substitution of colemanite used as additive by the local frit industry (Esteve and Ramos 1999), which was the main source of arsenic as an impurity (Arslan et al. 1999). Actual measured levels are within the same range of concentration compared to those measured in another similar industrial area in Greece (1.9 ng/m3 in PM2.5 and 1.1 ng/m3 in PM10 in 2006) (Tsopelas et al. 2008). On the other hand, arsenic content of PM10 and PM2.5 are lower than those measured in industrial zones in Murano (Italy), with values as high as 60 ng/m3 in PM10 in 2003 (Rampazzo et al. 2008). The levels measured in the area of study are also lower than those reported in another urban–industrial area in Huelva (Spain). During the period 2004–2005, the area of Huelva presented average arsenic levels in PM10 and PM2.5 ranging from 4.7 to 11 ng/m3 and from 3.0 to 9.2 ng/m3, respectively, and maximum arsenic content in PM10 and PM2.5 ranging from 22 to 62 ng/m3 and from 19 to 60 ng/m3, respectively (Fernandez-Camacho et al. 2010).

The European Directive 2004/107/EC establishes a target value for arsenic of 6 ng/m3 to be met by 2013. If the concentrations of airborne arsenic remain similar to those measured during the study period (2006–2010), it is very likely that the arsenic target value will be met in 2013.

Ratios of As in PM: As-PM2.5, As-PM10, and As-PM(10–2.5)

The ratios of PM2.5 to PM10 and to coarse fraction (i.e., PM10–PM2.5) at the urban site for the years 2007–2010 are presented in Table 2. The arsenic PM2.5/PM10 ratio ranges between 0.65 and 0.87 for the urban site (Table 2), while the PM2.5/(PM10–PM2.5) ratio varies between 1.9 and 6.8 for the urban site, which also indicates the accumulation of arsenic in the fine size particles. This values show that most of the arsenic content of the aerosol is found in the finer fraction PM2.5. This represents a risk for human health, since the smaller the diameter of the particle, the greater the capacity to enter the organism through the respiratory system and penetrate deep into the alveolar region of the lung (Sanchez-Rodas et al. 2012). Hetland et al. (2000) suggested that metallic particles could play a role in the induction of inflammation and cytotoxicity in the human epithelial cells.

Table 2 Annual average ratios at the urban site
Full size table

Table 2 shows an enrichment of the content of arsenic on the fine fraction during years 2009 and 2010 similar to the enrichment of fine particulate matter in PM10. It can also be observed that the ratios from 2007 to 2008, prior to the economic downturn, are significantly different (p < 0.05) from the ratios from 2009 to 2010. The concentration of arsenic in the fine size fraction of the PM is a consequence of the combustion processes in frit production (Vickery et al. 1998). Therefore, the evolution of the fine fraction will be mainly linked to the evolution of frit production. The ceramic tile production in 2007–2008 was 550 M/m2/year, while during 2009 and 2010, this figure decreased to approximately 350 M/m2/year (Celades et al. 2012). Nonetheless, the impact of the decline of manufacturing is expected to similarly affect the content of PM10 and PM2.5, as observed in Table 1 and Fig. 1. The enrichment of the fine fraction and the arsenic content of the fine fraction are related with a decrease of the emission of PM10. The main source of PM10 in the ceramic industry is linked with fugitive emissions from storage and handling of raw materials (Minguillon et al. 2009). From 2001 to 2007, abatement technologies were implemented. Therefore, while practically all companies stored the raw materials outdoors in 2001, a reassessment of the situation showed that over 50 % of companies stored their raw materials in closed compartments in 2007. In recent years, the handling and storage of raw materials in closed compartments continued to increase and many more companies have installed corrective measures. As a result of the implementation of storage of raw material indoors, the concentrations of fugitive emissions of PM10 have decreased significantly.

Arsenic seasonal pattern

In order to identify possible seasonal variations of arsenic in PM10 and PM2.5, an ANOVA and a Kruskal–Wallis test have been performed to the concentrations measured in spring, summer, autumn, and winter at the two sampling sites. The statistical analysis shows that there is no difference (p > 0.10) between concentrations measured at different seasons for none of the sites, indicating that a seasonal pattern does not exist in the area of study. The lack of a seasonal pattern in the area of study differs from the common trend of higher concentrations measured in winter in comparison with summer, as widely reported in the literature (e.g., Gao et al. 2002).

In addition to the statistical analysis, the seasonal levels of arsenic in PM10 and PM2.5 for the period (2006–2010)—plotted in Fig. 2—neither show a clear seasonal pattern. This may indicate that the arsenic airborne concentrations are mainly related with the industrial production of ceramic products. This suggestion is supported by the fact that maximum arsenic values occur at random order across all the months during the period of study (Table 1). Other causes that could lead to high concentrations in the levels of arsenic in PM10 and PM2.5 might be associated with lack of dispersion process associated with meteorology (Millán et al. 1997; Querol et al. 2007) and stagnant atmospheric conditions, which would not facilitate the dispersion of locally emitted arsenic. Nonetheless, stagnant conditions are generally more frequent in winter than in summer, and hence, this alone would not explain fully the spikes of concentrations observed during the warmer months (Fig. 2 and Table 1). Saharan intrusions would have an effect in spikes of particulate matter concentrations (see Table S1, supporting information), but it is not expected to substantially increase the concentrations of arsenic. Therefore, the lack of a seasonal pattern, the negligible influence of stagnant condition in warmer months, and the randomness across the year on the observed maxima values of arsenic in particulate matter suggest a prevalence of the industrial origin of airborne arsenic levels in the area of study associated with the batch operating production of frits and enamels.

Fig. 2
figure 2

Evolution of monthly average levels of arsenic in PM for the period 2006–2010 at two sampling sites, namely, site 1—industrial (PM10, solid line) and site 2—urban background (PM10, dotted line; PM2.5, dashed line)

Full size image

Evolution in levels of arsenic in the period 2006–2010

Figure 2 shows the monthly evolution of arsenic content in PM10 and PM2.5 during the period 2006–2010 at an industrial and urban background site within the ceramic industrial cluster. Figure 2 shows a zigzag pattern, which may be due to the batch mode operation of smelters that operate with different quantities of raw material containing variable contents of arsenic. This zigzag pattern has been previously related with a clear industrial origin (Sanchez de la Campa et al. 2008).

Figure 2 also shows a gradual decrease in the levels of arsenic across the period of study, albeit the major reductions can be observed prior to 2008. This sharper decrease can be attributed to the introduction of cleaner raw materials in the frit fusion kilns, which used colemanites with lower content of arsenic (Arslan et al. 1999; Karagolge et al. 2002).

The second sharpest decline occurred in 2008 when the economic crisis first struck Spain and ceramic production decreased significantly. A steady decrease is observed in the years 2009 and 2010 in which the economic crisis in Spain hit very strongly the construction sector, which was the main customer of ceramic materials, affecting considerably the production of ceramic industries settled in the study area. The concentrations of arsenic in PM levels measured during the economic crisis years are significantly lower (p < 0.001) than concentrations measured before the economic crisis.

Figure 3 shows the evolution between 2006 and 2010 of annual average levels of arsenic (in nanograms per cubic meter) in other locations belonging to the ceramic cluster of Castellón, which spans an area of 300 km2, with a population of 400,000 inhabitants. Data were obtained from the regional air quality monitoring network (i.e., Conselleria de Medio Ambiente de la Generalitat Valenciana) in the sites located in L’Alcora, Onda, Vila-real, Burriana, and Castellón. Onda and Vila-real are industrial cities like L’Alcora, Burriana is a coastal town and mainly agricultural, and Castellón is a major urban and coastal city (165,000 inhabitants).

Fig. 3
figure 3

Evolution during the years 2006–2010 of annual average concentrations of arsenic (in nanograms per cubic meter) measured in five sites within the ceramic cluster area

Full size image

Figure 3 follows a similar trend as that shown in Fig. 2, with arsenic levels declining in all cities within the ceramic cluster area. This suggests that the decline in arsenic emissions is generalized in all the ceramic manufacturing areas and not restricted to the local area of L’Alcora itself. The reduction of arsenic levels in the atmosphere is also benefiting the population located in nearby nonindustrial cities such as Burriana and Castellon or rural areas such as Zorita and San Jorge, as discussed previously. The fact that Burriana shows the highest levels of arsenic in aerosol might be attributed to the use of lead arsenates as pesticides (He et al. 2005) in orange tree orchards, which is the main economic resource of that city. It might also be attributed to the fact that Burriana is located downwind of the ceramic emissions (Querol et al. 2007).

If we compare the reduction of arsenic levels with the decrease in sales of ceramic materials such as tiles and flooring (Fig. 4), we can observe a strong correlation between both variables that clearly indicates the dependence on arsenic levels in the production industry in the area under study. Similar to the downward trend observed for the As content in PM, Fig. 1 shows a decrease in PM10 and PM2.5 levels that mirrors the reduction in ceramic sales.

Fig. 4
figure 4

Relationship between total ceramic sales in the ceramic cluster (in millions of euros considering CPI, source ASCER Spanish Association of Manufacturers of Tiles) and the arsenic content in PM for the period 2006–2010

Full size image

The combination of (a) the reduction of arsenic levels both locally and at rural background areas associated with the period of the economic downturn, (b) the zigzag pattern in arsenic content levels, (c) the relationship between particulate matter and arsenic content in air with ceramic sales, and (d) the use of boron compounds (such as colemanite, hydroboracite, and ulexite) in the production of frits (a raw material in tile manufacturing) provides evidence that the ceramic industry is the main source of airborne arsenic in the area of study.

Conclusions

The local ceramic manufacturing activity, especially that related to the smelters using boron compounds as raw materials, appears to be responsible for the arsenic airborne levels found in PM10 and PM2.5 in the study area (L’Alcora) and surrounding cities within the ceramic cluster area.

PM10, PM2.5, and its arsenic content have shown a sharp decrease in recent years associated with the economic downturn, which severely hit the production of ceramic materials in the area under study. The production decrease due to the economic crisis combined with several technological improvements in recent years (e.g., substitution of boron which contains As impurities as raw material) has reduced the concentrations of PM10, PM2.5, and As in air to an extent that currently met the existing European regulations. However, due to the association shown in this study between levels of PM and As in ambient air with ceramic production, when there is boost in ceramic production again, it is advisable that technological improvements are considered to ensure that the particulate matter and arsenic concentrations continue to meet the regulations.