Chemical Compositions of PM 2 . 5 at Two Non-Urban Sites from the Polluted Region in Europe

The study presents the analysis of measurement results for the ambient mass concentrations of fine particulate matter (PM2.5), PM2.5-bound carbonaceous matter (OC, EC) and water-soluble ions (Cl, NO3, SO4, Na, NH4, K, Ca, Mg). The 24-h PM2.5 samples were collected in the heating and non-heating seasons at two regional background sites in Southern Poland in 2011–2013. The percentage of the secondary organic and inorganic matter in PM2.5 was calculated. Over the whole measurement period, the mean PM2.5 concentration was 31.56 μg m and 24.92 μg m in Racibórz and Złoty Potok, respectively. Regardless of the season, the total carbon percentage in PM2.5 was comparable at both sites and amounted ~40%. There were no visible seasonal variations in the secondary organic carbon (SOC) share in PM2.5. The mean percentage of the primary organic carbon (POC) in PM2.5 was higher than the SOC percentage at both locations. The mean contribution of the water-soluble ions in the PM2.5 mass was lower than the TC percentage, with values 20.35% (Złoty Potok) and 33.56% (Racibórz). The total share of the secondary ions (SO4, NO3 and NH4) in PM2.5 was comparable in both measurement periods. It was shown that PM2.5 at regional background sites in Southern Poland is significantly different than at similar stations across Europe. It is reflected by higher concentrations of PM2.5 and its main components and lower percentage of the secondary ions in the PM2.5 mass. The carbonaceous matter percentage in PM2.5 is higher than in other parts of Europe.


INTRODUCTION
The recent interest has been focused on the chemical composition of particulate matter (PM), especially carbonaceous fractions and water-soluble ions, as they have predominant contribution to the PM 2.5 mass (according to Directive 2008/50/EC: "fine PM; shall mean particulate matter which passes through a size-selective inlet as defined in the reference method for the sampling and measurement of PM 2.5 , EN14907, with a 50% efficiency cut-off at 2.5 µm aerodynamic diameter") (Rogula-Kozłowska et al., 2014;Zhang et al., 2014;Zhao et al., 2015).
Elemental carbon (EC), inorganic/carbonated carbon (IC/CC) and organic carbon (OC) constitute the total content of carbonaceous matter, i.e., total carbon (TC), in PM (Seinfeld and Pandis, 2006).IC can be found in the crustal matter, present mainly in the coarse particles, so IC content in PM 2.5 can be neglected.Although EC is believed to be solely of primary origin, OC can be either primary or secondary (Jones and Harrison, 2005;Saylor et al., 2006;Plaza et al., 2011).Primary organic carbon (POC) is formed during incomplete combustion of organic materials and emitted mainly as very fine particles (Jones and Harrison, 2005).Mechanical processes, such as abrasion of tire rubber, as well as biological sources, i.e., emission of plant spores and pollen, vegetation debris, give rise to POC related to coarse PM (PM 2.5-10 , ambient particles with aerodynamic diameters from 2.5 to 10 µm) (Castro et al., 1999).Secondary organic carbon (SOC) comes from condensation of semivolatile organic vapours onto the particle surfaces and via atmospheric photo-oxidation reactions of precursor gases, such as terpene (Yang et al., 2011).The SOC content in the air is related to the emission of its precursors, whose amounts are expected to increase due to human activities (IPCC, 2007).The climatic conditions also have impact on the ambient SOC concentrations, as SOC production increases during periods with high photochemical activity (clear sky without clouds or fog present, high O 3 levels) (Saylor et al., 2006;Rogula-Kozłowska and Klejnowski 2013).
Organic aerosols make an important part of the PM mass.
The EEA report (2013) shows that, on average, organic substances make ~30% of the PM 2.5 concentration and ~20% of the PM 10 concentration measured at regional background stations in Europe.The same report also reveals that the secondary inorganic aerosol (SIA) constitutes ~35% of the PM 10 concentrations and ~50% of the PM 2.5 concentrations.Sulphate (SO 4 2-), nitrate (NO 3 -) and ammonium (NH 4 + ) are reported to be the major SIA components (Deshmukh et al., 2010;Błaszczak et al., 2016).
SIA is produced in the atmosphere through (photo-) chemical reactions of gaseous precursors (such as NO x , SO 2 , or NH 3 ) that may react with O 3 and other reactive molecules (including radicals) to form mainly ammonium nitrate (NH 4 NO 3 ), ammonium sulphate ((NH 4 ) 2 SO 4 ), and ammonium bisulphate (NH 4 HSO 4 ).The SIA formation strongly depends on the atmospheric conditions and availability of its precursor gases.First, ammonia neutralizes sulphuric acid to ammonium bisulphate (NH 4 )HSO 4 and ammonium sulphate (NH 4 ) 2 SO 4 .The remaining NH 3 may also react with nitric acid to ammonium nitrate (NH 4 NO 3 ) (Squizzato et al., 2012;Pay et al., 2012).Moreover, nitrate and sulphate can easily react with the sea salt and crustal aerosols in the atmosphere that is poor in ammonia.It results in the formation of calcium and sodium sulphate (respectively: CaSO 4 and Na 2 SO 4 ), as well as calcium and sodium nitrate (respectively: Ca(NO 3 ) 2 and NaNO 3 ) in the coarse particles (Seinfeld and Pandis, 2006;Squizzato et al., 2012).
To assess the percentage of the secondary matter in PM, it is particularly important to have information on the concentrations of its components at the regional background stations (Minguillón et al., 2012;Squizzato et al., 2012).The air quality degradation caused by PM in polluted areas is often characterized by high ambient concentrations of the regional background aerosols.There is not much data on the PM 2.5 concentrations and chemical composition available in Eastern Europe (EMEP, 2013).The data on the PM properties and chemical composition in Poland is even more incomplete (Rogula-Kozlowska et al., 2012, 2014).
Southern Poland, where both stations are located, is one of the most industrialized and polluted areas in Europe (EEA report, 2013;Rogula-Kozłowska et al., 2014).Therefore, in order to clarify the chemical characteristics of the regional background aerosols in this region of Poland, PM 2.5 collected from Racibórz and Zloty Potok was analysed for its main compounds, i.e., carbonaceous matter (EC, OC) and watersoluble ionic species.The seasonal and spatial variations were also investigated.Furthermore, the secondary organic and inorganic aerosol (SOA and SIA) concentrations, were also discussed.

Research Area
Fine PM was sampled at two regional background sites located in Southern Poland, i.e., Racibórz and Złoty Potok.Fig. 1 presents the geographical location of both sampling sites.
The station located in Złoty Potok constitutes the regional background point for the Silesia Province.Its nearest surroundings consist of meadows and arable fields.There are also a few chalets and a forester's houses heated with coal at a 150 m distance from the station.
In Racibórz, the research was conducted on the outskirts of the town.Its nearest surroundings are constituted by arable fields.The national road no.45 is located ~100 m west of the station.The closest loose residential land development is located ~100 m west of the station.
Southern Poland belongs to the regions of Poland with the highest degree of urbanization and pollution of all the components of the environment.The unique position of coal as an energy source, especially in power generation sector, is the reason why large amounts of gaseous and particulate pollutants are emitted to the atmosphere.Moreover, an important threat to air quality are also numerous industrial plants (e.g., cocking plants, iron works, waste incineration plants), as well as public and individual transport (Rogula-Kozłowska, 2014;Rogula-Kozłowska et al., 2014).The region is also charged with low-level emission sources, in particular fossil fuel combustion in households and local boiler-room (Rogula-Kozłowska et al., 2014).
The above considerations explain the specificity of both considered stations.Even though selected regional background sites are far away from any large energy production and industrial sources, both measurement points could be affected by the municipal emission and the pollutant inflow from the surrounding urban and industrial areas (Pastuszka et al., 2010).

Sampling and Chemical Analysis
24-h PM 2.5 samples were collected during the field campaigns carried out under the framework of two research projects.The campaigns covered two distinct seasons: the heating and non-heating ones of 2011-2012 (Racibórz) and 2013 (Złoty Potok) (Table 1).The year division was caused by differences in the air temperatures and the resulting energy consumption.In Poland, the increased energy demand in the heating season (October-March) results in the increase in the PM emissions from the fossil fuel combustion and biomass burning (Rogula-Kozłowska et al., 2014).The meteorological conditions during the study were typical for study area and were presented in Table 1.
The 24-h PM 2.5 samples were taken with low-volume PNS-15 samplers (Atmoservice) and were collected on quartz fibre filters.Before and after the exposure, the filters were conditioned in a weighing room (48 h; relative humidity: 45 ± 5%; air temperature: 20 ± 2°C).They were weighed on the Mettler Toledo microbalance (resolution: 2 µg).The mass concentration of PM 2.5 was determined with the gravimetric method in accordance with the standard of PN-EN 14907: 2006.A 1.5-cm 2 piece was cut out from each filter and was analysed for the OC and EC contents.The remaining part of the filter was analysed for the contents of Cl -, NO 3 -, SO 4 2-, Na + , NH 4 + , K + , Ca 2+ , and Mg 2+ .For analyses of the OC and EC contents in the 24-h PM 2.5 samples, thermal-optical carbon analyzer was used (Sunset Laboratories Inc.; "eusaar 2" protocol).The ion contents in water extracts were determined with a Metrohm ion chromatograph (Herisau Metrohm AG, Swiss).The details of measurement methods were specified elsewhere (e.g., Rogula-Kozłowska et al., 2014).

PM 2.5 Concentrations and Major Components
The results obtained in present study were compared with the data from selected regional background stations in Europe (Tables 2 and 3).The data on the PM concentrations and chemical composition in Europe are collected within the framework of the European Monitoring and Evaluation Programme (EMEP) (http://ebas.nilu.no).They are reported by the EU member countries to the European Commission and collected in the open-access database of the European Environment Agency (EEA), i.e., AirBase (http://www.eea.europa.eu/data-and-maps/data/airbase-the-european-air-qualitydatabase-7).The research on the chemical composition of PM 2.5 at background stations in Europe is usually carried out in a random way (selected days of the month) or it concerns folded weekly samples.In fact, the German Melpitz station is the only institution where the investigations of the concentrations of anions, cations and carbon are conducted in the 24-h PM samples over the whole year (Tørseth et al., 2012).
The mean 24-h PM 2.5 concentrations were higher and more diverse in Racibórz (4.04-217.49µg m -3 ) than in Złoty Potok (7.23-120.80 µg m -3 ).The mean PM 2.5 concentration in the whole measurement period equaled 24.92 and 31.56 µg m -3 , respectively in Złoty Potok and Racibórz.Therefore, at both stations located in southern Poland, the exposure reduction target (18 µg m -3 ), as well as exposure concentration obligation (20 µg m -3 ), had not been met (Directive 2008/50/EC, Annex XIV).Moreover, in the case of Racibórz, the limit value for yearly averaged PM 2.5 concentrations (25 µg m -3 ) had been exceeded; mean PM 2.5 concentration at Zloty Potok was very close to the limit value.Thus, the obtained results indicate a serious problem in terms of exposure of the residents to fine PM and signal the need to take actions aimed at improving the air quality of the considered areas.
When compared to the values observed at regional background stations in Europe, the PM 2.5 concentrations in both considered stations were high.At the background stations in Europe, relatively high PM concentrations are normally observed in the regions exposed to the strong influence of the natural aerosol sources, e.g., desert dust (Rodríguez et al., 2002;Pederzoli et al., 2010;EMEP, 2013).Both considered locations were not affected by such sources.Taking into account the location of Racibórz and Złoty Potok in Southern Poland, the seasonal variations in the PM 2.5 concentrations revealed that the concentrations were determined mainly by the variations in anthropogenic emission, especially municipal (low-level) emission (Rogula-Kozłowska et al., 2012, 2014).Moreover, the impact of meteorological conditions is crucial, especially in winterwhen the low height of the missing layer and frequent temperature inversions prevent from pollutants propagation in the atmosphere (Juda-Rezler et al., 2011;Rogula-Kozłowska et al., 2014;Reizer and Juda-Rezler, 2016).
The mean concentrations of the main PM 2.5 components in Złoty Potok and Racibórz were higher (similarly to the PM 2.5 concentrations) than the values observed at other sites in Europe (Tables 2 and 3).This finding particularly concerned carbon compounds, which were the dominant PM 2.5 component at both measurement stations in Southern Poland.What is more, it was also a dominant PM 2.5 component in the Central and North-East Poland (Zielonka and Puszcza Borecka, respectively).The percentage of the three dominant ions (SO 4 2-, NO 3 -and NH 4 + , a group known as secondary ions (SI) (Galindo et al., 2013;Moroni et al., 2015) in the PM mass in Poland was either lower (see values for Waldhof, Neuglobsow, Cabauw-Zijdeweg, Risoe, or Auchencorth Moss) or comparable (see values for Üto, Iskrba, Ispra, or Montseny).
It seems that the PM air pollution pattern for background locations in Southern Poland is different in comparison to the locations in other parts of Europe.Its specificity results from the different chemical characteristics of PM 2.5 in this region and concerns low SI percentage and high TC percentage in the fine PM mass.It corresponds with the conclusions included in the previous works of the authors, which showed that in cities of Southern Poland the main reasons for these differences are biomass burning and fossil fuel combustion in residential sector (Pastuszka et al., 2010;Juda-Rezler et al., 2011;Klejnowski et al., 2012;Rogula-Kozłowska et al., 2012, 2014).One of the reasons for the presence of very high PM 2.5 concentrations and the differences in the PM 2.5 chemical composition are also specific conditions that affect both measuring stations.Złoty Potok and Racibórz, differently from the regional background stations located in other parts of Europe (Table 2 and 3), throughout the year, and mostly in the heating season, remain under the influence of air masses originating from the polluted regions of Upper Silesia and South-Eastern parts of Europe (Juda-Rezler et al., 2011;Rogula-Kozłowska et al., 2014;Błaszczak et al., 2015;Reizer and Juda-Rezler, 2016).
The 24-h concentrations of carbonaceous compounds in Złoty Potok and Racibórz demonstrated a strong seasonal variability, with higher values in the heating season and more reduced levels in the non-heating season (Table 4).The differences can be attributed to the changes in the emission profile (domestic and public heating combustion devices were generally not in operation from April to September) and unfavourable meteorological conditions (low height of the mixing layer, frequent temperature inversions) that prevented the pollutant dispersion and removal (Yang et al., 2011;Rogula-Kozłowska et al., 2012).Importantly, the extremely high 24-h concentrations were observed in January and February and were associated with the periods when very high PM 2.5 mass concentrations occurred.In such cold winter months the activity of local emission sources of PM increases, especially in the case of biomass burning and fossil fuel combustion in domestic furnaces and in small local heating plants.
Regardless of the season, the TC percentage in PM 2.5 was comparable at both stations.The mean values were 39.59% for Złoty Potok and 41.52% for Racibórz (Table 4).The TC percentage values in PM 2.5 were mainly determined by the variations in the OC share in the PM 2.5 mass (mean values: between ~28% in the non-heating season in Złoty Potok and ~40% in the heating season in Racibórz).The OC percentage values in the PM 2.5 mass were higher at    both locations in the heating season than in the non-heating one.Nonetheless, a relatively high OC percentage in the PM 2.5 mass was also observed in the non-heating season.
The finding is agreed with similar studies (Plaza et al., 2011;Pio et al., 2011;Zhang et al., 2014) and indicates that biological matter and secondary organic aerosol (SOA) could have been important OC sources in Racibórz and Złoty Potok.There was also a seasonal variation for the EC/PM 2.5 ratio.However, it was less significant than the one discovered for OC/PM 2.5 .
The 24-h OC/EC ratios in PM 2.5 revealed a broad range of values from 3.61 to 14.03 and from 3.06 to 9.25, respectively at Raciborz and Zloty Potok.They were relatively high, which showed clear prevalence of the organic carbonaceous species over EC.The finding also indicates the SOA formation (Plaza et al., 2011;Satsangi et al., 2012).At both stations, there were no clear seasonal changes in the OC/EC ratios (Table 4).However, the maximum OC/EC ratios were found in the non-heating season.
The OC/EC ratio can be used to gain some insight into the emission and transformation characteristics of the carbonaceous aerosol (Yang et al., 2011).The contributions of the primary and secondary organic carbon (POC and SOC, respectively) to OC have been difficult to quantify.The lack of a direct chemical analysis method to identify these components has led the researchers to employ several indirect methods (Strader et al., 1999).In this study, the SOC and POC levels were calculated according to the methodology proposed by Castro et al. (1999): where (OC/EC) pri is the ratio of primary OC to EC, assumed to be relatively constant for a specific location, season and local meteorology (Castro et al., 1999).To estimate (OC/EC) pri , the least-squares regression was performed for 10% of the samples with the lowest 24-h OC/EC ratio (Strader et al., 1999) (Fig. 2).The regression curve slope represents the ratio of the primary OC to EC (OC/EC) pri .
The 24-h SOC concentrations varied within broad limits (Fig. 3).Their values were 0.00-52.45µg m -3 and 0.00-17.06µg m -3 for Racibórz and Złoty Potok, respectively.Over the whole measurement period, the mean SOC concentrations were 4.29 µg m -3 (Racibórz) and 3.24 µg m -3 (Złoty Potok), which accounted for 33.54% and 36.06% of the mean PM 2.5 -bound OC mass at both locations, respectively.The mean ambient concentration of SOC was slightly higher in Racibórz, because of the location of the station close to the urban area and this could contribute in higher concentrations of PM 2.5 and its chemical constituents.The station in Złoty Potok lies at a distance of about 20 kilometers to the south-east of Częstochowa and about 25 kilometers to the north from Zawiercie.The direct surroundings of the station are the meadows and fields of crops.Therefore the concentrations of PM 2.5 and its chemical components were quite lower.
The mean SOC percentage in the PM 2.5 mass was comparable for both locations ~12%).No clear seasonal dependence was observed (Table 4).Moreover the SOC content in the PM 2.5 mass from both measurement stations was nearly twice as low as the POC content (Table 4; Fig. 3).The mean POC percentage in PM 2.5 was 27.62% and 21.00% (Racibórz) and 25.40% and 17.66% (Złoty Potok) in the heating and non-heating seasons, respectively.The higher POC percentage in the PM 2.5 mass in the heating season suggests that biomass burning was an important OC source, because it often generates fairly substantial amounts of organic matter of primary origin (Pio et al., 2007;Plaza et al., 2011).Local heating systems, particularly in small towns and villages, are often based on the combustion of low quality coal, wood, dung and domestic waste, which leads to the increase of the amount of carbonaceous particles of primary origin (Braniš et al., 2007).On the other hand, abundant vegetation is present in rural areas.It releases significant amounts of primary organic matter, such as biological matter (e.g., fungal spores, vegetation detritus, and plant waxes), especially in the warm season (Yang et al., 2011).

Ionic Compounds. SI Contribution
Taking into consideration the entire measurement period, the mean ionic concentrations in PM 2.5 were in the following order (Table 2): The common mass share of above mentioned ions equalled 20.35% (Złoty Potok) and 33.56% (Racibórz) of PM 2.5 .Secondary ions (SI; SO 4 2-, NO 3 -and NH 4 + ) dominated in the PM 2.5 ionic composition from both measurement stations.Higher SI concentrations were observed at both stations in the heating season (Table 3), which was probably caused by the enhanced intensity of the local emission sources (emission of gaseous precursors of SI) and atmospheric conditions that were unfavourable to the spreading and favourable to the formation of secondary inorganic compounds (Rogula Kozłowska et al., 2014;Błaszczak et al., 2016).The SI concentrations revealed also spatial variability, with higher levels recorded in Racibórz, because of the localization of the station on the outskirts of the town, where mean SI concentrations were 14.167 µg m -3 (26.76% of PM 2.5 ) (heating season) and 4.270 µg m -3 (30.32% of PM 2.5 ).In Złoty Potok, the mean SI concentrations were 5.845 µg m -3 (18.50% of the PM 2.5 mass) in the heating season and 2.795 µg m -3 (17.09% of the PM 2.5 mass) in the non-heating one.
The mean seasonal percentage of nss-SO 4 2-(non-sea salt sulphates), NO 3 -and NH 4 + in the SI sum and total PM 2.5 mass is shown in Table 5.Interestingly, the SO 4 2-concentration was almost identical with the nss-SO 4 2-concentration (Fig. 4).As both stations were located inland, the sea salt influence on PM 2.5 could be ignored (Deshmukh et al., 2010).The occurrence of Na + and Cl -in fine PM at different locations in Southern Poland should be solely linked to the emission from the coal combustion and biomass burning (Rogula-Kozłowska and Klejnowski, 2013;Rogula-Kozłowska et al., 2012, 2013).The situation in Racibórz and Złoty Potok was probably the same as higher Na + and Cl -concentrations were observed in the heating season along with high contents of carbonaceous compounds (particularly OC) in the PM 2.5 mass.
At both stations, the SI percentage in the PM 2.5 was visibly higher than the SOC share (Tables 4 and 5).The considered stations did not only differ in the total SI content in the PM 2.5 but also in the share of each secondary ion in the SI and PM 2.5 masses.The mean nss-SO 4 2-percentage in SI was higher in Złoty Potok (48.59% and 64.60% in the heating and non-heating seasons, respectively).In Racibórz, the mean seasonal nss-SO 4 2-percentage in SI was 34.89% (heating period) and 56.19% (non-heating period).The values for NO 3 -in SI were higher (44.16% and 32.97%) in Racibórz than in Złoty Potok (34.97% and 24.93%) (Table 5).
The content of NH 4 + in SI was twice as low as the contents of SO 4 2-and NO 3 -and was comparable at both stations.The SI percentage in the PM 2.5 did not demonstrate significant seasonal differences (Table 5).However, contributions of individual inorganic ions to PM 2.5 and SI revealed different seasonal variations, which agreed well with the literature data (Rogula-Kozłowska et al., 2014;Zhang et al., 2014).At both stations, the nss-SO 4 2-percentage in the SI and PM 2.5 masses was higher in the non-heating season, which was related to the increased intensity of the photochemical transformations (Mirante et al., 2014).In the heating season, the observed higher values for NO 3 -/SI, NH 4 + /SI, NO 3 -/PM 2.5 and NH 4 + /PM 2.5 were due to low temperatures and stable meteorological conditions that were favourable to the reactions of the nitric acid transformations (mainly in the gaseous phase) into nitrates (Deshmukh et al., 2010;Mirante et al., 2014).
Generally, over the entire measurement period, the SO 4 2contents in the air significantly exceeded the NH 4 + contents (Fig. 5).The mean nss-SO 4 2-/NH 4 + values in Racibórz were 0.88 ± 1.01 (heating season) and 2.22 ± 1.20 (non-heating season).In Złoty Potok, the values were 1.26 ± 0.56 and 2.67 ± 1.13, respectively (Table 5).It may be said that (NH 4 ) 2 SO 4 was the main component of SIA at both locations in the non-heating season.The remaining SO 4 2 -ions in the PM 2.5 samples could have come from soluble salts (such as CaSO 4 , Na 2 SO 4 ) or from H 2 SO 4 .
The occurrence of H 2 SO 4 and HNO 3 in PM 2.5 at both locations (particularly in the non-heating season) was confirmed, to some extent, with analysing the ratios of the total cations Σ cations [µeq m -3 ] (µeq m -3 -micromole of a given ion in 1 m 3 of ambient air, e.g., 2 µeq m -3 of Na + means 2 µmol of Na + in 1 m 3 of ambient air) to the total anions Σ anions [µeq m -3 ] in PM 2.5 .Regardless of the measurement season, Σ cations /Σ anions seasonal values were less than 1 at both locations (Table 5).This means that anions extracted from PM 2.5 could have come also from compounds with other elements than those determined in the cation forms in the water extracts.As the sum of the crustal and trace elements usually makes a small percentage of the PM 2.5 mass (Sillanpää et al., 2006;Zhao et al., 2015), it seems probable that some part of SO 4 2-and NO 3 -in PM 2.5 could have formed compounds with hydrogen (undetermined in the extracts) before the extraction.The highest daily values of Σ cations /Σ anions ratio was observed in Racibórz in the heating season of 2011 (Fig. 6).The high Σ cations /Σ anions value was accompanied by very high concentrations of NH 4 + , K + and Na + .This situation could have been caused by the occurrence/inflow of a specific alkaline aerosol.At that time, the correlation of Σ cations vs. Σ anions was linear.However, its character was different from the remaining measurement period (Fig. 7).In Złoty Potok, concentrations of all the ions increased in the similar way and proportionately to the PM 2.5 concentrations in the heating season.
If it is assumed that all NH 4 + ions form (NH 4 ) 2 SO 4 in the periods when the SO 4 2-/NH 4 + ratio is greater than 1, (NH 4 ) 2 SO 4 concentration in the air can be easily calculated with the stoichiometric calculations (Rogula-Kozłowska, 2016).It may also be assumed that the content of NH 4 + reacting with SO 4 2-is sufficient to completely neutralize SO 4 2-and the remaining NH 4 + ions make NH 4 NO 3 in the periods when the SO 4 2-/NH 4 + ratio is less than 1 whereas the Σ cations /Σ anions ratio is higher than the ratio in the remaining periods.Those assumptions helped to calculate the 24-h and mean concentrations of (NH 4 ) 2 SO 4 and NH 4 NO 3 (Table 5; Fig. 8).
In the heating season, the mean concentrations of (NH 4 ) 2 SO 4 and NH 4 NO 3 were similar at both locations (Racibórz: 8.83 and 7.77 µg m -3 , respectively; Złoty Potok: 3.31 and 2.64 µg m -3 , respectively) (Table 5).For both locations, the sum of the ambient (NH 4 ) 2 SO 4 and NH 4 NO 3 concentrations slightly exceeded the SI concentration, which resulted from the highly simplified assumptions for the assessment of the ambient (NH 4 ) 2 SO 4 and NH 4 NO 3 concentrations.In the non-heating season, only trace amounts of NH 4 NO 3 were found in the sum of the two discussed secondary components of PM 2.5 .In the warm seasons, the sum of the (NH 4 ) 2 SO 4 and NH 4 NO 3 concentrations did not cover even the half of the SI concentrations.

CONCLUSIONS
Observed PM 2.5 concentrations at both regional background sites in Southern Poland are much higher than values from other stations of this type in Europe.This is most strongly pronounced during the heating season when the PM 2.5 concentration in Racibórz was above twice greater than in the most polluted European regional background stations  located in Ispra (IT) and Melpitz (DE).These results show the necessity of separate treatment of heating and nonheating seasons during analyses of air pollution.Moreover, at both considered Polish stations, the air quality standards for PM 2.5 had not been met, which may lead to adverse effects on human health.
The concentrations of PM 2.5 and PM 2.5 -bound major components revealed clear seasonality, with the maxima in the heating season caused by the changes in the emission profile and unfavourable meteorological conditions.Regardless of the season, the PM 2.5 chemical composition was generally dominated by the carbonaceous matter, especially organic carbon, which concentration was at least twice higher than values observed in other background stations in Europe.The assessment of the SOC and POC concentrations revealed a much higher contribution of POC to the PM 2.5 mass at both stations.The mean SOC percentage in the PM 2.5 mass was similar at both locations.
No clear seasonal dependence was observed.
At both locations, the ionic content of PM 2.5 was dominated by SI (SO 4 2-, NO 3 -NH 4 + ), which generally followed the mass abundance pattern of SO 4 2-> NO 3 -> NH 4 + .The total SI percentage in the PM 2.5 was similar in heating and nonheating seasons.On the other hand, the contributions of individual inorganic ions to PM 2.5 and SI revealed different seasonal variations.
The rough assessments of the seasonal and daily mean concentrations of (NH 4 ) 2 SO 4 and NH 4 NO 3 indicated general correlations concerning both regional background stations in Poland.In the cold heating seasons, both NH 4 NO 3 and (NH 4 ) 2 SO 4 were equal components of PM 2.5 .In the warm non-heating seasons, NH 4 NO 3 could not be practically found in PM 2.5 .The specificity of both considered stations could have resulted from the higher SO 2 level in the air than at other background locations in Europe, which was observed in summer.Even though the level was even higher in the heating season, the emission of ammonia, which neutralizes sulphuric acid in the air, was also higher.
The PM 2.5 found at regional background sites in this area was different from the PM 2.5 found at similar sites in various parts of Europe.The obtained results suggest that the main reasons for these differences were: utilization structure of fossil fuels in Poland and the periodically occurring combination of meteorological conditions particularly unfavourable to the pollutant spread.Even though the results of the present study should not be generalized, the findings show that higher aerosol pollution may not necessarily be restricted to larger cities.It is also common in small villages.Due to the rising prices of oil, electricity and natural gas, people have returned to more traditional and cheaper fuels as low-quality coal, dust coal or waste wood.Consequently, traditional heating in villages may strongly contribute to the local air pollution to a great extent and may pose serious health problems.
data was taken from the stations where the concentrations of all the components discussed in this publication were measured within the air quality monitoring framework.a http://ebas.nilu.no;b http://www.eea.europa.eu/data-and-maps/data/airbase-the-european-air-quality-database-7.

Fig. 3 .
Fig. 3. Time series of the POC and SOC contributions to PM 2.5 in Racibórz (a) and Złoty Potok (b).

Table 1 .
PM 2.5 sampling periods and selected meteorological parameters during the sampling campaigns.RH -relative humidity; P -atmospheric pressure.The meteorological data was taken from Regional Inspectorate of Environmental Protection (RIEP) in Katowice website concerning air quality monitoring in Silesian Province (http://monitoring.katowice.wios.gov.pl/).a at station in Zloty Potok.b at the station in Wodzisław Śląski (the nearest RIEP`s station to Racibórz).

Table 2 .
Mean yearly concentrations of PM 2.5 and its main components at regional background stations in Europe.

Table 3 .
Mean concentrations of PM 2.5 and its main components at regional background stations in Europe in the heating and non-heating seasons.

Table 4 .
Mean ambient concentrations (± standard deviations) of EC, OC, TC, SOC and POC and mass contributions of EC, OC, TC, SOC and POC to PM 2.5 , OC in EC, EC in TC and SOC and POC in OC, at two regional background sites -Racibórz and Złoty Potok.