Chemical Composition of Particulate Matters in Makkah – Focusing on Cations , Anions and Heavy Metals

The size of particles mainly determines where the particle is deposited in the respiratory system, whereas the composition determines the type of body response. This study analyses Total Suspended Particles (TSP), Particulate Matter with aerodynamic diameter up to 10 micron (PM10) and 2.5 micron (PM2.5) and their chemical compositions. TSP, PM10 and PM2.5 samples were collected from August 2012 to September 2013 in Makkah, Saudi Arabia. Composition of PM was analysed, focusing on heavy metals, cations and anions. The heavy metals considered in this study were Lead (Pb), Nickel (Ni), Cadmium (Cd), Chromium (Cr), Vanadium (V), Arsenic (As), Mercury (Hg) and Aluminium (Al). Cations and anions considered were Chloride (Cl), Bromide (Br), Fluoride (F), Ammonium (NH4), Phosphate (PO4), Sulphate (SO4), Nitrate (NO3) and Nitrite (NO2). Average TSP, PM10 and PM2.5 (μg m) levels were 366.38, 233.38 and 143.49, respectively. The ratios of PM2.5/PM10, PM2.5/TSP and PM10/TSP were 0.61, 0.39, and 0.64, respectively. NO3 and SO4 were the most abundant anions in all three types of PM. NO3 contents (%) were 28.53, 39.79, and 30.77 in TSP, PM10 and PM2.5, whereas SO4 contents (%) were 22.52, 17.02 and 21.37, respectively. In TSP and PM2.5, As showed the highest content, which was about 42% and 43%, whereas in PM10, Al showed the highest content (47%).


INTRODUCTION
Atmospheric Particulate Matter (PM) is considered an important air quality issue due to the complex nature of its composition, growing levels in urban areas, and various health and environmental impacts.PM has both long and short-term effects on human health, such as cardiovascular, lung and skin diseases, which sometimes lead to premature death (Krewski et al., 2000).Particles smaller in size are considered more harmful in terms of their adverse impacts on human health (Habeebullah et al., 2014a).Fine particles go deeper into the respiratory system and, therefore, cause more health problems than coarse particles (Habeebullah, 2013).Rapid industrialisation and urbanisation in the past couple of decades have resulted in a world-wide increase of airborne particulate matters.Therefore, PM monitoring, modelling, investigating their chemical composition, and studying their health and environmental impacts have become a routine part of the air quality management programme.
Air pollution caused by PM is one of the major air quality issues due to the arid nature and presence of vast sandy deserts in Makkah, Saudi Arabia.Recently, several investigations were made in Makkah to model PM levels (Munir et al., 2013), to study their health impacts (Habeebullah, 2013), and to analyse their spatial and temporal variability (Habeebullah et al., 2014a, b).These authors have reported that PM levels in Makkah exceed air quality standards, however, the high levels of PM are probably not solely due to anthropogenic sources (e.g., road traffic) but due to the arid nature of the region.Recently, Habeebullah et al. (2015a) compared air quality in Makkah, Saudi Arabia and Leeds, UK and reported that in Makkah PM 10 did not correlate well with other anthropogenic air pollutants, such as Nitric Oxide (NO), and Carbon Monoxide (CO), whereas in Leeds PM 10 had much stronger association with other pollutants.This probably indicates that in Makkah PM 10 is not predominantly emitted by combustion sources like road traffic, and is rather emitted by other sources, such as windblown dust, re-suspension, construction and demolition activities, and sea spray.Munir et al. (2013) using a generalised additive model (GAM) analysed PM 10 in Makkah.They used wind speed, wind direction, temperature, relative humidity and traffic related air pollutant concentrations, such as CO, NO, nitrogen dioxide (NO 2 ), sulphur dioxide (SO 2 ) and lag_PM 10 (previous day PM 10 ).Temperature and wind speed showed strong positive, whereas traffic related air pollutants showed a weak association with PM 10 concentrations, suggesting road traffic is probably not the major source of particulate matters.Several authors (e.g., Alharbi, 2009;Othman et al., 2010;Khodeir et al., 2012) have reported that PM in Saudi Arabia, particularly in Makkah are emitted by a wide range of sources, which include road traffic, construction and demolition work, resuspension of dust particles, and windblown sand and dust particles.Sand and dust storms are common in Makkah and in the surrounding regions, which add to the load of atmospheric particulate matters (Alharbi, 2009).
Makkah is considered the holiest city in the Muslim world, therefore millions of Muslims visit the city every year.Makkah is a densely populated city, which accommodates about 1.7 million people (Central Department of Statistics and Information, 2010), however due to visitors from all over the world, especially during the month of Ramadhan and the Hajj season this number is increased by several folds, which put extra burden on the available resources, including road traffic.As a result, in addition to other environmental conditions, healthy and clean air becomes an issue, which require continuous monitoring.Despite the fact that several investigations have been made in the past (e.g., Habeebullah, 2013;Munir et al., 2013;Habeebullah et al., 2014a, b;Munir et al., 2014;Sayegh et al., 2014;Habeebullah et al., 2015b;Munir et al., 2015) to analyse different aspects of air quality in Makkah, the chemical composition of PM requires further investigation to analyse various species in PM, which may provide an insight into the origin and health impacts of PM.
The toxicity of particles is dependent on both size and chemical composition of the PM (Walters and Ayres, 2001).The size of particles mainly determines where the particles are deposited in the respiratory system, whereas the composition determines the type of response by the body, for instance iron (Fe) will have different response than arsenic (As) (Walters and Ayres, 2001).In this paper both particle size and composition are analysed in a view to better understand the levels of different size particles, such as Total Suspended Particles (TSP), Particulate Matters with aerodynamic diameter up to 10 micron (PM 10 ) and 2.5 micron (PM 2.5 ) and their chemical composition, focusing on cations & anions and heavy metals.

METHODOLOGY
This study was conducted at the Hajj Research Institute (HRI), Umm Al-Qura University in the Holy City of Makkah.The Holy City of Makkah is situated at an average elevation of around 277 m above sea level (Habeebullah, 2013).Makkah is threatened by seasonal flash floods despite the low amount of annual precipitation (about 110 mm annual), mainly falling in the winter months.High temperature is observed in summer months when temperature reaches as high as 48°C.Makkah is a highly populated city and in addition it hosts several millions of pilgrims each year coming to perform Hajj and Umrah.This increases the demand for transportation, accommodation and other facilities (Habeebullah, 2013).
High Volume Sampler (HVS) was used to collected PM samples.The loaded filters were weighed before and after sampling to determine the weight of suspended particulate matter.HVS measures different sizes of dusts, depending on the weather proof hood, which fits on the top of the HVS.It allows air to flow downward to the sampling filter.A standard filter was used having dimensions 203 mm × 254 mm (Hassan, 2006).After collection the samples were analysed in the Laboratory and the concentrations of heavy metals, cations and anions were determined in TSP, PM 10 and PM 2.5 using standards laboratory equipments.Quality control was assured by duplicate samples, blanks and the method of standard additions.
The flow of air through the glass filter used in HVS paper was 1-1.5 m 3 min -1 .Filtration method was used to estimate the total concentration of suspended particulate.The filter paper was weighed dried in the laboratory and transported in an airtight container to the site.At the end of sampling period (24 hours), samples were transferred carefully to the laboratory, and weighed again.The difference in weight was considered as the weight of suspended particulates collected on the filters.Water Soluble Ions (WSIa) were extracted in deionised water using ultrasonic water bath (Ultra-sons-H, J.P., and Selecta, Spain) and then the sample was filtered.Filtrate was used to estimate the concentration of WSIs using ion-chromatography system (850 Professional IC-Metrohm) (JIS, 1992).
For metals detection filters were digested three times (10 minutes each time) with 10 mL of nitric acid (HNO 3 ) (1 M) using ultrasonic water-bath (Mohammed, 2012).The obtained filtrate were analysed by using ICP technique (model: 700 series_ICP OES_ spectrometers Agilent) to detect the concentrations of Al, As, Cd, Cr, Ni, Pb, V, and Hg.Each element was quantified under specific wavelength condition with the corresponding dilutions using deionised distilled water and standards which were simultaneously analysed with experimental samples.The instrument deviation was checked at the beginning and end after the detection of each trace element.
In this study TSP, PM 10 and PM 2.5 samples were collected from August 2012 to September 2013 in Makkah, Saudi Arabia from 4 monitoring stations (Fig. 1).Samples were collected for every 24 hour period.Composition of PM was analysed, focusing on heavy metals, cations and anions.The heavy metals considered in this study were Lead (Pb), Nickel (Ni), Cadmium (Cd), Chromium (Cr), Vanadium (V), Arsenic (As), Mercury (Hg) and Aluminium (Al), which are the most predominant heavy metals in the atmosphere.Cations and anions considered in this paper were Chloride (Cl -), Bromide (Br -), Fluoride (F -), Ammonium (NH 4 + ), Phosphate (PO 4 3-), Sulphate (SO 4 2- ), Nitrate (NO 3 -) and Nitrite (NO 2 -).Data were analysed using the statistical software R programming language (R Development Core Team, 2012), and associated package openair (Carslaw and Ropkins, 2012).In addition, graphical presentations and correlation analysis were used to investigate the similarities and dissimilarities between various species.

RESULTS AND DISCUSSION
Average TSP, PM 10 and PM 2.5 (µg m -3 ) levels were 366.38, 233.38 and 143.49, respectively.This means TSP > PM 10 > PM 2.5 , which was expected.The levels of TSP, PM 10 and PM 2.5 demonstrated considerable temporal variability both in weekly and annual cycles.The ratios of PM 2.5 /PM 10 , PM 2.5 /TSP and PM 10 /TSP were 0.61, 0.39, and 0.64, respectively.Previously, Khodeir et al. (2012) calculated PM 2.5 /PM 10 ratios in Jeddah, where the ratios were significantly lower (0.33).In addition, Khodeir et al. (2012) reported much lower PM 2.5 (28.4 µg m -3 ) and PM 10 (87.3 µg m -3 ) levels in Jeddah than in Makkah.Makkah and Jeddah have different emission sources and atmospheric conditions and therefore differences in PM levels are expected.Temporal variations of TSP, PM 10 and PM 2.5 are presented in Fig. 2. In Fig. 2  lower-panel.TSP, PM 10 and PM 2.5 showed different levels during different hours and had different weekly cycles.TSP, PM 10 and PM 2.5 demonstrated highest levels on Monday, Saturday and Friday, respectively.The weekly cycles of these species are not associated with the traffic flow, which is higher on weekdays and lower on weekend.Habeebullah (2013) reported lowest levels of Carbon Monoxide (CO), Nitrogen Oxides (NO x ) and Sulphur Dioxide (SO 2 ) in Makkah on Friday.In Makkah the weekend is on Friday and Saturday, which have less traffic activities than the weekdays.Therefore on Friday the levels of these pollutants are less due to weekend effect.Habeebullah (2013) further reported that PM 10 had different weekly cycle than the rest of the traffic related pollutants, which was due to the fact that PM concentration is more related to the arid nature of the region, wind speed and wind direction rather than traffic flow.Annual cycles (Fig. 2-lower) of TSP, PM 10 and PM 2.5 also showed different trends and peaks levels were observed in different months.Highest levels of TSP was observed in October, that of PM 10 in January and October, and that of PM 2.5 in March.There seems to be a weak association between the different species and are affected differently by local meteorology, otherwise they would have shown the same annual trend.
Table 1 shows the average concentrations of different inorganic ions in TSP, PM 10 , and PM 2.5 .Among cations and anions secondary nitrate (NO 3 -) and sulphate (SO 4 2-) were the most abundant anions in all three species (TSP, PM 10 , PM 2.5 ).Table 2 shows percentage of these ion in particulate matters.NO 3 -contents (%) were 28.53, 39.79, and 30.77 in TSP, PM 10 and PM 2.5 , whereas SO 4 2-contents (%) were 22.52, 17.02 and 21.37, respectively.Although the absolute content of the cations and anions is much greater in larger particle size than the finer ones, this is not the case for the percentage amount.For example, the absolute amount of Cl in TSP > PM 10 > PM 2.5 , however the percentage amount of Cl in PM 2.5 > PM 10 > TSP.The percentage NO 3 in PM 10 > PM 2.5 > TSP.This shows that some cations and anions are more abundant in one particle size, while other are more abundant in another particle size.
In Fig. 3 the correlation plots show the association between various cations and anions in TSP (upper-panel), PM 10 (middle) and PM 2.5 (lower-panel).The association between various cations and anions is different in TSP, PM 10 and PM 2.5 .As shown in the Fig. 3, there is both positive and negative correlation coefficients (r), however the negative correlation coefficient values are much smaller than the positive ones.In TSP the highest positive correlation was observed between PO 4 3-and Br -(r = 0.74), followed by F - and NO 2 -(r = 0.34).In PM 10 the strongest positive correlation was found between Br -and Cl -(r = 0.76), followed by NH 4 + and NO 3 -(r = 0.59).In PM 2.5 the strongest correlation was observed between NH 4 + and NO 3 -(r = 0.61), followed by NO 3 -and Cl -(r = 0.36).Strong positive association probably indicates the same or similar emission source.However, most of the cations and anions are secondary pollutants, which are formed in the atmosphere from the primary pollutants.The conversion of the primary pollutants to the secondary depends on the concentrations of the primary pollutants (precursors) and meteorological parameters, such as temperature and solar radiation.Therefore, the positive association probably shows that the two ions are affected in a similar way by a meteorological parameters.
The average concentrations (µg m -3 ) of various heavy metals in TSP, PM 10 and PM 2.5 are shown in Table 3. Highest amounts of Hg, Cd, and Cr were found in PM 2.5 , whereas the highest amount of As and Pb were found in TSP.In percentage (Table 4) Ni (36.95%) and As (42.25%) were the two most abundant heavy metals in TSP, which accounted for about 80% of the total heavy metals in TSP.In PM 10 Al (46.86%) and As (22.67%) were the two most abundant heavy metals, whereas in PM 2.5 As (43.42%) and Hg (31.99%) were the two most abundant heavy metals.In each TSP, PM 10 and PM 2.5 As is one of the two abundant heavy metals.Arsenic is considered to be one of the most toxic elements for human health.Prolonged exposure to a high level of As is likely to produce an acute toxic effect in humans.However, low doses of As do not result in an acute toxic effect but may cause cancer in humans after prolonged exposure (Fang et al., 2012).The data used here are for a relative shorter period of time, therefore no final conclusion regarding whether As can cause any serious health problem or not can be drawn based on this dataset in Makkah.Further monitoring of As is required in Makkah for a longer period of time at several monitoring stations, including urban, suburban, rural, roadside and industrial locations.Correlation plots of various heavy metals in TSP (lower), PM 10 (middle) and PM 2.5 (upper) are presented in Fig. 4. In PM 2.5 , Cd, Cr, Al and Pb showed very strong positive correlation with each other (r > 0.90).Likewise, V    showed strong positive correlation with Hg (r = 0.80).However, As showed strong negative correlation with Cd, Cr, Pb and Al (r~-0.60).In PM 10 , the correlation of Cd, Cr and Pb is very strong with each other, however the correlation of Al with Cd and Cr has turned negative (r~-0.50).Also, As showed very strong correlation with V and Cr (r > 0.90).
In TSP, the correlation between various heavy metals is weaker, except between Cr and Cd, where r > 0.90, followed by V vs. Cd (r = 0.65) and V vs. Cr (r = 0.64).Both As and Ni showed negative correlation with Cd and Cr (r~-50).
It is useful to compare the results of this study with other studies conducted in Saudi Arabia and the surrounding regions.Table 5 compares the concentrations of TSP, PM 10 and PM 2.5 , whereas Table 6 compare the concentrations of various ions and heavy metals in PM.

CONCLUSION
In atmospheric particulate matters both particles size and their composition play an important role in: (a) for how long the particles stay in the atmosphere; (b) where the particles deposit in the respiratory tract; and (c) what is the body response.In this paper the various particles sizes i.e., TSP, PM 10 and PM 2.5 and their compositions in terms of cations, anions, and heavy metals are analysed.Sample of particulate matters are collected from several locations in Makkah.The Concentrations of TSP, PM 10 and PM 2.5 are determined and their weekly and annual cycles are analysed.
Average concentrations (µg m -3 ) of TSP, PM 10 and PM 2.5 were 366.38, 233.38 and 143.49, respectively.NO 3 -and SO 4 2-ions were the most dominant, whereas arsenic (As) was found to be the most prominent among the heavy metals analysed in this paper.Further monitoring of As is required in Makkah for a longer period of time at several monitoring stations, including urban, suburban, rural, roadside and industrial locations.Correlation analysis is performed between various ions as well heavy metals.From the results of this study it can be concluded that a considerable proportion of PM is not associated with combustions sources (e.g., road traffic) and rather seems to be consists of particles generated by constructions and demolition activities, windblown dustand-sand particles, and resuspensions.Detailed receptor modelling studies are required over a larger temporal and spatial range to conclusively determine the emission sources of PM, which is part of the ongoing research work carried out for the improvement of air quality in Makkah.

Fig. 1 .
Fig. 1.Map of the monitoring stations in Makkah.
Fig. 2. Weekly (upper-panel) and annual (lower-panel) cycles of TSP, PM 10 and PM 2.5 from August 2012 to September 2013.

Table 1 .
The average concentrations (µg m -3 ) of various ions in TSP, PM 10 , and PM 2.5 in Makkah.

Table 2 .
The percent Concentrations (%) of various ions in TSP, PM 10 , and PM 2.5 in Makkah.