Assessment of heavy metal pollution resulting from informal E-wastes recycling in the Greater Accra Region of Ghana


 This study investigated the levels and spatial distributions of four selected heavy metals in the soil and drainage components emanating from informal E-waste recycling activities at Ashaiman scrapyard, in the Greater Accra Region, Ghana. The metals are Cadmium (Cd), Chromium (Cr), Copper (Cu) and Lead (Pb). Five sampling sites were randomly selected, with top and sub-soil sampled from the two open burning areas (hereafter H and F). Three sites in the drainage that runs through the scrapyard were similarly selected for sample collection; a control upstream (soil-sediment-water; WSC) and two experimental units downstream (soil-sediment-water; WS1 and surface water only; WS2). Four control topsoil samples were taken at distances of 25, 50, 75 and 100 m away from the scrapyard. Composite sample of three sampling units per site, including pH analysis, with two replications per treatment, were investigated using standards methods. Spatial distribution of the metals in the scrapyard were analyzed using Inverse Distance Weighted (IDW) interpolation method. Coefficient of variation (CV) was used to investigate the source of pollution. The pollution levels were investigated using three criteria, namely Geoaccumulation Index (Igeo), Contamination Factor (CF) and Pollution Load Index (PLI). Correlation analysis was used evaluate the relationships between the metals. Mean CV of 88.5% suggests that the scrapyard pollution is anthropogenically-driven. Igeo of soil samples from the scrapyard revealed the following: (i) Cd and Pb (unpolluted to strongly polluted), (ii) Cu (unpolluted to moderately polluted), and (iii) Cr (unpolluted). CF revealed the following: (i) Cd (moderate to strong pollution), (ii) Cu (moderate pollution), (iii) Cr (low pollution), and (iv) Pb (high pollution), but the metals exhibited moderate PLIs. Spatial distribution maps revealed heavy metal pollution decline with distance away from the scrapyard, which was inversely related to pH levels. WSC showed lower heavy metal concentrations than WS 1, while the lowest levels were detected in WS 2. Generally, moderate to very strong correlations existed among the metals in the scrapyard. In conclusion, the scrapyard was the epicenter of E-waste pollution primarily driven by human activities.


Introduction
E-wastes cover Electrical and Electronic Equipment (EEE) and their parts that have been discarded by their owners as wastes without the intent of re-use or recycling. Due to their different lifespan pro les, different E-wastes have different environmental and health impacts as well as different economic value [1]. Owing to rapid changes in technological updates and upgrades of EEE, industrialization and modernization, increase in disposal income and the popularized increase in the use of EEE, there is an upsurge in the acquisition and utilization of electrical and electronic products. Consequently, the generation of E-wastes is one with the largest and fastest growth rates among wastes in the world. Asia contributed most to the generation of E-waste in 2019, generating close to 24.9 million metric tons (Mt), followed by Europe (12.0 Mt), Americas (13.1 Mt), Africa (2.9 Mt) and Oceania (0.7 Mt) [1,2].
As a result of free and illegal trading activities and the lack of implementation of environmental policies, African countries receive high quantities of potential E-waste materials from developed countries. Nigeria and Ghana are notably mentioned [3]. Around 600,000 of used EEE were imported into Nigeria in 2010. E-waste recycling activities in Ghana is dominated by the informal sector, largely due to poor implementation of environmental related laws, and also for socio-economic reasons. It is estimated that between 121,800 to 201,600 individuals are involved in the informal E-waste sector in Ghana. The formal recycling sector in Ghana receives only about 0.2% of E-waste for treatment [11]. Unsurprisingly, Ghana is noted for having one of the biggest informal E-waste recycling in Africa, at the Agbogbloshie scrapyard.
This research was to assess the levels and spatial distributions of selected four heavy metals (Cd, Cr, Pb, Cu) at an informal E-waste recycling site and evaluate their levels based on permissible ranges speci ed by the World Health Organization (WHO), Food and Agricultural Organization (FAO) of the United States and the Environmental Protection Agency (EPA) of Ghana. The second objective was to investigate the degree of pollution using geoaccumulation index (Igeo), contamination factor (CF) and pollution load index (PLI). The rest of the paper are organized into the following sections: (i) Sect. 2 describes data and methods, (ii) Sect. 3 is devoted to results and discussion, and (iii) Sect. 3 describes the conclusion, recommendation and future work.

Study area
The Ashaiman scrapyard is located at the entry into the township from the Tema metropolis, about 0.12 km from the Accra-Tema Motorway. Covering a land size of about 0.07 km 2 , it is located on latitude 05 o 41' 4.99" N and longitude 00 o 01' 37.28" W. The region is generally at with savannah grasses and shrubs being the dominant vegetation. The topsoil is mostly sandy-clay, with the subsoil being predominantly clay [12].
The scrapyard (Fig. 1), adapted from Okine [13], houses large metal containers which store E-waste materials until they are ready to be worked on. Dismantling  Samples were taken in the early morning of 16 th July 2019. At each burning site, ve top soil samples were randomly taken at ve different sites (marked as 1A, 2A, 3A, 4A and 5A) within a soil depth of 0-10 cm and ve subsoil samples from the sites but different level (marked as 1B, 2B, 3B, 4B, and 5B) within a depth pro le of 10-20 cm. Thus, from each burning site 10 soil samples were taken, and a total of 20 samples were obtained from the two burning areas. Four other top soil samples were randomly taken at distances of 25, 50, 75 and 100 m from the scrapyard (marked as HV 20, HV 50, HV 75 and HV 100, respectively) to determine the horizontal distance from the scrapyard at which levels of heavy metals will be detected and also serve as control as well as background conditions for assessing the pollution levels.
At all the sampling sites, a composite sample made up of three sampling units was collected. Soil samples were collected using newly purchased stainless-steel garden shovel and a standard measuring rule to determine the vertical depth of the soil pro le. Samples were collected into plastic bowls with tightly tting lids pre-cleaned with nitric acid. They were then sent to the Ghana Standards Authority for further treatment and analysis.
The coordinates at sampling points were recorded using GPS software.

Water samples
A composite of three sampling units of soil-sediment-water was collected upstream about 140 m from the scrapyard and as control (WS C). A similar procedure was replicated for two experimental units, namely soil-sediment-water samples (WS 1) at 50 m and purely surface wastewater (WS 2) at about 70 m both downstream of the scrapyard. These sampling stations were randomly selected from a range of stations that were accessible. Samples from WSC, WS 1 and WS 2 were used to test the hypothesis that high levels of heavy metals are expected to be sediment than the surface wastewater. Coordinates were measured using GPS software. pH of these samples was measured on site using a handheld HANNA pH meter calibrated with buffer solutions of pH 4, 7 and 10. The samples were collected and sent to the same place just as the soil samples.

Soil sample preparation and pH determination
Soil samples were air dried at around 105 o C to eliminate wetness and obtain constant weights representative of the soil only. They were then passed through a 2 mm non-metallic mesh to separate and remove rocks exceeding 6.35 mm. The soil particles passing the mesh were thoroughly homogenized by manual milling with a mortar and pestle. These preparations were necessary for good dissolution during chemical treatments and increasing the accuracy of the sample analysis [14]. To three grams of each of the dried and sieved soil samples in a 25 mL beaker (which had been pre-cleaned and thoroughly washed with distilled water), 15 mL of aqua regia were added and the resulting sample solution digested in a fume chamber for about 30 minutes to remove any foreign material that might interfere with the analytical results. Following cooling, there was addition of distilled water to the digested soil sample. This was then ltered into a 100 mL volumetric ask using the Johnson test paper lter paper with a diameter of 125 mm. Distilled water was then added to the solution to the 100 mL mark. Spatial distributions of topsoil and subsoils were determined using Inverse Distance Weighted (IDW) interpolation method [15].
The soil pH analysis was conducted following the procedure described by Al-Busaidi et al.
[16] by dissolving two grams of each sample in distilled water in a 1:1 ratio and stirred to a uniform suspended mixture using a clean glass rod. The samples were then allowed to settle for about 10 minutes and then continually stirred for about 15 minutes using a magnetic stirrer on a magnetic sitter plate. The samples were allowed to settle and their pH determined using the same calibrated instrument just as in the wastewater sample case.

Water sample preparation and pH determination
Two replicates each from WS 1 and WS 2 samples were digested by drawing 100 mL of each replicate into a beaker, to which 25 mL mixture of 3:2 conc. HNO 3

Indices for determination of soil pollution
Three pollution indices were employed to evaluate the pollution levels of the four heavy metals in the scrapyard and its environs.

Geoaccumulation Index (Igeo)
Igeo determines the contamination of heavy metals by assessing their concentrations in sampled soils relative to background concentrations during pre-industrial periods [17,18]. Igeo is computed using the mathematical formula given as follows: where

Contamination Factor (CF)
CF evaluates quantities of an element in a sample normalized over that of pre-industrial baseline value for the element. Mathematically, CF [20] is expressed as (2) where Ce and Ci are respectively the concentration levels of the heavy metal in the sample of interest and the background value of the heavy metal of interest. Based on values obtained, soil or sediments can be classi ed as follows: no or low contamination where CF < 1; moderate contamination where 1 < CF < 3; considerable contamination where 3 < CF < 6; very high contamination where CF > 6 [21].

Pollution Load Index (PLI)
PLI [20] examines the mutual contribution of a group of metals to the pollution of a site. Mathematically, where CF represents the contamination factor of each heavy metal element in a sampled soil and n is the number of heavy metals under consideration. The PLI gives an indication of whether the site under consideration is: lightly polluted, where PLI ≤ 1; moderately polluted, where 1 < PLI ≤ 3; highly polluted, where PLI > 3 [21,22].

Statistical analysis
Means of the heavy meatal concentrations were computed and compared to reference values. The associated standard deviations were computed to determine the distributions of the metals. Differences in means were analyzed using t-test at 95% con dence level. These parameters were computed using Microsoft Excel software 2016 version. The relationships among the metals in the topsoil and subsoil at sites H and F were quanti ed using Pearson correlations. Coe cient of variations (CVs) of the heavy metal concentrations were computed to determine the source of pollution, using by SPSS version 21.0.  (Table 3), pH ranged from 6.07 to 7.78, with a mean of 6.94. Heavy metal adsorption and retention by soil increases generally occur within a pH range of 4-7 [23,24], thus the pH range could partly account for the elevated levels of heavy metals in the samples. pH values recorded were within the WHO benchmark of 6.5-8.5, except for three samples (5.88 at Site F and 6.07, 6.38 at site H), where pH values were below the 6.5 minimum threshold. Relatively high pH value recorded in sample 5B (8.03) could be due to the presence of alkaline batteries, steel mill, and ashes from the incineration processes at the E-waste site.

Heavy metal concentration
Several factors, such as electron activity, soil texture, soil pH, ionic strength and level of organic matter, affect the metallic forms in soil matrix. Cd was found to be the least in concentration among the heavy metals in the two soil pro les at site F, from non-detection levels to a maximum of 1.57 ppm and an average concentration of 0.48 ppm ( Table 2). With the exception of one sub-soil sample with a concentration of 1.57 ppm, all Cd concentrations at Site F were below the Ghana EPA permissible limit of 1 ppm and the WHO/FAO standard of 3 ppm (Tables 1-2). Cd mobility is dependent on several factors such pH and presence of organic matter that has strong a nity for the metal. The mean pH at this site was alkaline (7.13), which limits its availability, thus accounting for its low concentrations at this site [25,[26][27][28]. At site H, Cd concentrations were relatively higher, with a minimum of 0.29 ppm and a maximum of 13.56 ppm and an average concentration of 4.14 ppm, which exceeds both Ghana EPA and the WHO/FAO standards (Tables 1 and 3). At this site, the mean pH was 6.94. The slightly acidic conditions may have contributed to the high levels of the metal. E-waste materials with Cd at the scrapyard include printed circuit boards, batteries, accumulators, cathode ray tubes and ultraviolet lights. Site H had higher concentrations of Cd than Site F because Cd-containing E-waste materials were located more at the former site than the latter.  (Tables 1 and 3). Comparatively, Cr concentrations at site F were higher than those at site H, which could be due to the fact that the metal containers that house E-waste materials were closer to site F. They are typically composed of steel and chromium, so any wear and tear on the metal adds on Cr concentrations to the soil, pH also being a factor.
At site F (Table 2), the minimum and maximum concentrations recorded for Cu were 29.97 and 253.42 ppm, respectively, with an average concentration of 114.85 ppm. The concentrations exceeded permissible levels of Ghana EPA (20 ppm) and the WHO/FAO standards of 100 ppm (Table 1). Also, the minimum and maximum concentrations of Cu at site H ranged from 5.24 to108.76 ppm with an average concentration of 48.37 ppm, which were above the national and international limits (Tables 1 and 3). Cu nds application in most electrical and electronic appliances, such as printed circuit boards, cathode ray tubes, bare/insulated wires and in refrigeration units. Majority of the samples had high levels of Cu in the topsoil than the subsoil. These may be attributable to the strong binding between Cu and organic matter and minerals in the soil. Consequently, its mobility is supressed, and hence cannot be leached into the subsoil [29,30]. Also, the prevailing alkaline conditions in the topsoil played a critical role (Table 3).
Ranging from 13.58 to 276.78 ppm, with an average concentration of 77.07 ppm, at site F, Pb level was found to have exceeded the WHO/FAO and Ghana EPA standard of 50 and 20 ppm, respectively (Tables 1-2). Elevated levels were detected for Pb, ranging from 17.81 to 1000.85 ppm, with an average concentration of 341.43 ppm at site H, with so obvious exceedance (Tables 1 and 3). The lowest and the highest levels were all detected on the topsoil. E-waste materials with Pb include cathode ray tubes, uorescent bulbs, batteries and fuses. The elevated Pb levels from site F to H is re ective of low organic matter in the presence of slightly alkaline to near-neutral mean soil pH. It is contemplated that the elevated levels of Pb is a consequence of metal accumulation arising for all the operational years of the scrapyard. The extent of pollution at site F and H can be respectively expressed as follows: Cu > Cr > Pb > Cd and Pb > Cu > Cr > Cd.
Heavy metals concentrations in this study were similar to other E-waste research [4,31,32]. Generally, Cu and Pb were in high concentrations in most of the research studies, suggesting an extensive use of the two metals in electrical appliances, whereas Cd concentrations seem to be on the lower side in most Ewaste soils in other research works (e.g., Table 4).

Statistical Studies
Tables 5a-g present the relationships between the metal levels at the following sampling sites: (i) Topsoil at site F, (ii) Subsoil at F, (iii) Topsoil at site H, (iv) Subsoil at site H, (v) Topsoil at F and H, (vi) Subsoil at F and H, and (vii) Topsoil and subsoil over the scrapyard. These were quanti ed using Pearson correlation coe cient metric r.
The following were the key ndings at the following sampling sites: (i) In the topsoil at site F, only Cd and Cu were signi cant, but exhibited out-of-phase relation (r= 0.667; p<0.01; Table 5a); (ii) In the subsoil at site F (Table 5b), Cu, Cd, and Pb all exhibited in-phase relationships (Cu vs. Cd r=0.926; p<0.01; Pb vs Cd r=0.956; p<0.05; Pb vs. Cu r=0.889; p<0.05); In the topsoil at site H (Table 5c), it is only Pb and Cu that were signi cant and exhibited an in-phase relationship (r=0.675; p < 0.01). In the subsoil at site H (Table 5d), Cr, Cd, and Cu showed signi cant inphase relationships (Cr vs. Cd r=0.786; p<0.01; Cu vs. Cd r= 0.978; p<0.05; and Cu vs. Cr, r= 0.653; p<0.01).
In the topsoil at F and H (Table 5e), Cd, Cr and Cu at site F, and Cr and Pb at site H, showed good correlations. The in-phase relationship ranged from r=0.693 to 0.779; p<0.05 and p<0.01). These were captured for Pb vs. Cr, Cr vs. Cd, and Cd vs Cu. In contrast, Cd vs. Cr (r=-770; p<0.01) showed an out-ofphase relationship. Across the scrapyard, in the study area (Table 5g), the topsoil and subsoil revealed inphase relationship for Cr vs. Cr (r=0.821; p<0.05), Pb vs Cd (r=0.734; p<0.01) and Pb vs. Pb (r=0.79; p<0.05).
From these results, it has been revealed that the in-phase relationships dominate the relationships and suggest synchronization of activities and chemical processes emanating from human E-waste recycling activities, whereas the out-of-phase relationships suggest otherwise. For instance, the in-phase relationships observed between pairs of heavy metals may be due to the dual complementary usage they have in certain EEE products. Cd and Pb nd close applications in cathode ray tubes where Cd is used as the uorescent powder coatings to produce color while Pb is employed to absorb the UV lights and X-rays produced. Cd is added to Cu to form alloys in Cd-Cu wire which are more resistant to softening at higher temperatures. Pb is also alloyed to Cu to act as a lubricant and also assist in chip break up, thereby increasing the machinability of the Cu metal. Since site H is used as a burning site and dumping grounds for burnt E-waste products, heavy metals may be carried from site F to site H. This could explain the positive correlation between heavy metals at different site.
A coe cient of variation (CV) analysis carried out to determine if the presence of the heavy metals was due to natural or anthropogenic source showed CV for Cd, Cr, Cu and Pb to be 137, 58, 61 and 104%, respectively at site F, and 108, 22, 125 and 93% for Cd, Cr, Cu and Pb, respectively at site H. According to Guo et al. [33], a CV less than 20% indicates natural sources while values greater than 50% imply anthropogenic sources. By inference the heavy metal pollutions were due to anthropogenic sources, speci cally E-waste activities.
An independent t-test analyses showed statistically signi cant differences in mean concentrations of the heavy metals, which are as follows:   3.6. Concentration differences at increasing distance from scrapyard Figure 8 depicts the metal concentrations as a function of distance away from the scrapyard. Soil samples taken at 25, 50, 75 and 100 m away from the scrapyard were mostly sandy. pH values were mildly acidic, decreasing with distance away from the scrapyard, but within the WHO 6.5-8.5 thresholds.

Contamination Factor and Pollution Load Index
This is expected, as increasing distance from the scrapyard means decreasing heavy metals concentrations, which are mostly alkaline. This nding is comparable to a study by [34] where the pH at a dumpsite decreased from 5.9 to 4.7 at a distance of 18 m from the dumpsite.
The results practically revealed no Cd in these soil samples. Samples taken within the 25 m distance were found to contain respective concentrations of 20.73, 24.94 ppm for Cr and Cu and were within safe levels set by WHO/FAO but slightly above permissible levels of Ghana EPA with respect to Cu. However, m away from it. However, since soil samples taken at 25 and 50 m were close to the Accra -Tema motorway, contamination from road dust is still possible since heavy metals can be found in tires and brake abrasion, combustion exhaust and pavements wear [35], which can be transported by rain, runoff, dry deposition, and atmospheric drifts. and Further research will be needed to investigate this proposition.
With a general decline in the concentrations of heavy metals from the 75 and 100 m distance, the high levels of heavy metals within the scrap yard can be attributed mainly to that of the E-waste activities.
Comparably, concentrations of Cr, Cu and Pb were several times higher within the scrapyard than outside of it.
The decreasing concentrations of heavy metals with increasing distances from the scrapyard agrees with other research studies which explored the effect of increasing distance from source on concentration levels of heavy metals [36, 37].

Heavy metal concentrations in sediment and water
The metal concentrations in the drain both upstream and downstream are shown in Table 9. Soilsediment-water samples (WSC) taken outside the scrapyard showed lower concentration of heavy metals than those obtained within the scrapyard particularly at WS 1, while the four heavy metals were absent within the water at WS 2. Water sediment outside the scrapyard (WSC) was found to contain Cd and Cr, with respective concentrations of 0.03 and 11.95 ppm, whereas the levels of Cu and Pb were 5.84 and 5.89 ppm, respectively. Water sediments within the scrapyard (WS1) were found to contain 0.49 ppm Cd, and a concentration of 217.98 ppm, for Cu. Cr had a concentration of 12.28 ppm while Pb had a concentration of 44.77 ppm. The levels of the toxic metals in the water sediments of WS1 increased signi cantly within the scrapyard as one moves downstream from WSC. With the wastewater drain lying in a lower plain to the two burning sites, and with movement of air current across the drain from the two burning sites, it can be fairly postulated that the E-waste activities are a possible origin for the heavy metals in the soil-sediment water samples, through the actions of wind drift and dry deposition. One other possibility is the presence of E-waste materials found near or inside the drain, causing heavy metals to dissolve into the wastewater as has been reported elsewhere [38].
The relatively concentrated amounts of heavy metals of soil-sediment-samples (WS1), compared to no detection in the surface wastewater samples at WS2 supports the research hypothesis, that heavy metals tend to be high in sediments and settleable particles than surface water [39]. WSC and WS1 samples had mild alkaline pH, indicative of the presence of the heavy metals. Levels of heavy metals in the drain represented by WS1 were all above the standard permissible levels of Ghana EPA and WHO/FAO. This is of a major concern since the drain serves as irrigation source for farm crops as well as drinking water for herds of cattle near the scrapyard.

Conclusions
The research revealed soil The study has further shown that Cd levels (at sites F and H) and Pb levels ( at site F) in the scrapyard suggest skewed distributions relative to Cr and CU (at both sites) and Pb (at site H), which are normally distributed. This outcome provides insight into modeling the behavior of these metals in future. Finally, future studies can also focus on investigation of heavy metals contamination in workers at the scrapyard and herds of cattle around the environment.            Spatial distributions of Cu in subsoil (L) and topsoil (R) of sites F and H Figure 5