Seasonal evaluation of total organic carbon removal from river samples using scrap metal-based coagulant and local salt modified-biomaterials for point-of-use water treatment

ABSTRACT This study evaluated the potentials of activated carbon prepared from coconut shell (CS) (Cocos nucifera), counter wood (CW) (Afzelia africana) and iron (III) sulphate coagulant (ISC) prepared from scrap iron, in removing total organic carbons (TOCs) from contaminated Ebonyi river (EBRW) and Ezeiyiaku river samples (ERWA) at point – of – use (POU). A 250 ml of the river samples at ambient pH were mechanically (hands) batch – agitated with optimised doses of 0.5, 1.5, and 0.025 g of granulated activated coconut shell carbon (GACSC), granulated activated counter wood carbon (GACWC), and ISC at the two seasons. The results showed a 90 and 100% reduction in the TOCs of the treated ERWA and EBRW samples during the rainy season, while and at dry season, 88% reduction in TOCs was observed in both treated river samples.


Introduction
Water is highly essential for the sustenance of life; a basic criterion for a healthy living. However, the availability of safe drinking water has increasingly been elusive and a burden due to pollution further caused by rapid growth in human population, science and technology [1][2][3]. Pollution of natural water with trace organic chemicals has been documented [4][5][6][7][8]. These pollutants, most often are mutagenic or endocrine disruptors [9]. Some are non-biodegradable and recalcitrant such as the humic substances. Their presence in source water becomes potentially harmful during water disinfection process with chlorine due to formation of carcinogenic substances such as trihalomethanes and haloacetic acids, while some may persist after treatment and finally accumulate in the organisms [10,11]. Furthermore, the presence of dissolved organic carbon (DOC) in natural water promotes the growth and multiplication of pathogenic bacteria, and consequently constitutes a means of possible exposure to cancer and liver damage, anaemia, gastrointestinal disorder, urinary tract infections and skin irritation in humans [12]. Other illnesses likely to occur from water contaminated with dissolved organic substances include hormonal imbalance, biological dysfunction and defects of major internal organs [13].
The Ebonyi River, a tributary of the Cross River system, flows through Ezillo, Abakaliki (Izzi), Ikwo and Onisha-Oshiri. It serves for domestic purposes for the rural communities and for municipal water treatment and reticulation. The Ezeiyiaku River traverses all the communities of Akaeze (e.g., Umobor, Ihenta, Iyioji, Akaezeukwu and Mgbede), and it is a source for drinking water mainly during the dry season. Bathing, laundry and swimming activities take place in the rivers, especially the Ezeiyiaku River, during dry season and early rainy season.

Sample collection
Amber bottles were thoroughly washed with warm liquid soap, and separately rinsed with warm de-ionised water and dilute HCl. The glass bottles were heated in an oven at 300°C for 4 h. The teflon-lined caps were also washed with warm liquid soap, rinsed with deionised water and allowed to air-dry. The bottles were removed from the oven, capped and moved in an ice box for sample collection. At the point of sampling, the river was agitated with the sample bottle, uncapped and dipped about 30 cm below the surface of the water in inverted position. The containers were filled and capped under water and against the direction of the flow [29]. The pH of the water samples was measured in-situ, while the other physicochemical parameters of the river samples including turbidity, electrical conductivity and total dissolved solids were measured at the laboratory prior to analysis within 48 h using Labtech digital turbidity metre (probe), and Electrical conductivity metre WKW, Germany.

Carbonisation and activation of coconut shell and counter wood
The activation of the CS and CW (Cocos nucifera and Afzelia africana) were carried out by adapting the methods of [30]. CS and CW were broken and sieved with Sieve Number 12 (ASTM E 11-87) to obtain 1-3 mm particle sizes, and then dried in the oven at 110°C for 4 h. Different ratios of impregnation (1:1, 2:1, 3:1 and 3:2) of the biomaterials were then carried out by boiling separately with 10, 20, 30, 40, 50, 60, 70 and 80% concentrated solutions of indigenous sodium chloride salt, until the mixtures became pasty. The pasty materials were dried at 110 °C for 2 h in the oven, allowed to cool and repeatedly washed with de-ionised water to remove traces of activating agent. They were again dried in the oven at 110°C for another 2 h, and were separately carbonised in a muffle furnace at 450°C for 3 h. The activated biomaterials, after cooling, were separately rinsed with de-ionised water to a neutral pH solution and then dried in oven at 110°C for 3 h.

Synthesis of Iron (III) Sulphate coagulant
The iron (III) sulphate coagulant was prepared by adapting the method of [31]. About 50 kg of degreased and washed scrap iron metal was soaked with enough quantity of 40% concentrated sulphuric acid and left to react for 24 h in a capped amber bottle. The resulting ferrous sulphate crystals was removed, washed with enough de-ionised water and crushed to fine size. Oxidisation of the iron (II) sulphate was done by heating the fine crystals with H 2 O 2 solution at 80°C in 500 ml beaker using temperature, and time regulated magnetic hot-plate stirrer until a brownish yellow paste was formed. The iron (III) sulphate paste was then dried in the oven at 120°C for 2 h. Iron (III) sulphate lump formed was later ground into anhydrous powder.

Determination of specific surface area (S BET ) of granulated activated biomaterials
The Brunauer, Emmett and Teller (BET) method was adopted in the assessment of the total surface area of the granulated activated coconut shell carbon (GACSC) and granulated activated counter wood carbon (GACWC). Analysis was done using a Micromeritics ASAP 2020 Surface Area and Porosity Analyser: Norcross, GA 30,093-2901, U.S.A. Before the gas adsorption, 0.1 g of each sample was degassed at a constant temperature of 120°C in a vacuum condition for a period of 1.5 h. Nitrogen adsorption isotherms were then measured between a relative pressure (P/P 0 ) range of 0.001-1.000. BET surface areas of the GACSC and GACWC samples were calculated using the BET equation in equation 1 based on the assumption that the surface area occupied by each physisorbed nitrogen molecule was 0.146 nm 2 [32].
Where: N = Avogadro's number; s = cross sectional area of adsorbed nitrogen; v = molarvolume of adsorbed nitrogen; w = mass of the modified sample; V liq = volume of liquidN 2 in pore [33].
The total and specific surface areas, S T and S BET , of the samples were obtained by using the equations 2 and 3, respectively, while the pore size was calculated using equation 4.

Determination of physical surface structure/morphology of the granulated activated biomaterials
Scanning electron micrograph of the activated biomaterials was examined using a DSM 9872 Gemni Scanning Electron Microscope (SEM) [34,35]. The sample surface was sputter-coated with a thin film of platinum before it was viewed with SEM.

Determination of surface charges of the granulated activated biomaterials and prepared coagulant
The method used by [35,36] was adapted in this determination. Seventeen sets of 100 ml beaker containing 50 ml of 0.1 M potassium chloride (KCl) solution with initial pH values ranging from 2.00 to 10.00 were prepared with 0.1 M NaOH and HCl solutions, respectively. A 0.150 ± 0.002 g of GACSC, GACWC and ISC were separately soaked in the initial pH of KCl solutions and allowed to attain equilibrium in 48 h. After the 2 days, the final pH values of the separate content of the beakers were measured.

Sedimentation beaker test of prepared coagulant and determination of dose optimisation of granulated-activated biomaterials
The sedimentation beaker test was aimed at accessing the optimum dose of the coagulant, ISC, which when used to treat the water samples would give the least values of turbidity. Adapting the method of [37], a 0.005, 0.010, 0.015, 0.020, 0.025 and 0.030 g of the prepared ISC were separately added to 250 ml, 86.4 NTU of EBRW and 250 ml, 99.0 NTU of ERWA samples in 250 ml beakers during rainy season and dry seasons, respectively. Each of the mixtures was mechanically stirred for 60 min without adjusting the pH of the water samples and allowed to sediment for a period of 60 min. They were then filtered with a cloth (cotton) strainer, and after which the turbidity of the filtrates was measured using labtech digital turbidity metre (probe).

Assessments of concentration of organic substance and total organic carbons removal from water samples at point-of-use treatment
A 250 ml of water sample was batched separately and agitated mechanically with five different composite masses of GACSC, GACWC and ISC (0.500, 0.500, and 0.005; 0.500, 1.000, and 0.010; 1.000, 0.500, and 0.015; 1.500, 0.500, and 0.020; 0.500, 1.500, and 0.025 g) in five sets of 500 ml beakers during the rainy and dry seasons to determine the optimal doses of the composite materials for the amount of organic substance removed in mg/L from the water samples at POU. Each of the mixtures was mechanically stirred for 60 min without adjusting the pH of the water sample, and then allowed to sediment for 60 min. It was then filtered with a cloth strainer, and the amount of organic substance removed in mg/L from the untreated and treated water samples determined by Walkley-Black method as described by [38,39].
About 10.0 ml of the respective water samples was reacted with 10 ml of 1.0 M K 2 Cr 2 O 7 solution and 20 ml of concentrated stock solution of H 2 SO 4 in a 250 ml beaker. The mixtures were gently swirled severally and allowed to stand in a fume hood for 30 min. A 100 ml of de-ionised water was then added to dilute the mixtures followed by an addition of 10 ml phosphoric acid and 1 ml of diphenylamine indicator. Phosphoric acid was added to eliminate interferences of ferric (Fe 3+ ) that may be present in the untreated river samples or from the coagulant in the treated river samples [39,40].
The mixtures and a blank (without ferrous ammonium sulphate) were then titrated with a 0.2 M ferrous ammonium sulphate [Fe(NH 4 ) 2 (SO 4 ) 2 6H 2 O] solution until the colour had transitted from deep violet to blue and finally bright green. The percent carbon of the water samples was evaluated with equation 5.
The calculated % C is converted to mg/L by multiplying the % C by 10,000 Where: 0.003 is (equivalent weight of carbon/10 3 ), Vws is volume of water sample, Vs and Vbare volume of the ferrous ammonium sulphate used for the sample and blank,respectively.
During the rainy season, 5 ml of the treated samples (with the least % C) and the untreated samples were then analysed for TOC using GC-MS (gas chromatography 7890A GC system, 5675 C Inert MSD from Agilent 19,091-433HP-5 Ms USA) with triple axis detector equipped with an auto injector (10 µl syringe), and helium gas as the carrier gas.
However, during the dry season, the treated samples and the untreated samples were initially extracted into 10 ml n-hexane before injecting into GC-MS. The percentage removal of the TOC from the water samples was then calculated using equation 6. % TOC Removal ¼ TOCs identified in untreated sample À TOCs recalcitrant after treatment x 100 OCs identified in untreated sample (6)

Figures 2 and 3 show the nitrogen adsorption-desorption isotherm of the activated biomaterials at 77 K.
Where V ads denotes adsorption volume.

Total surface area of granulated-activated biomaterials
The pore/total surface areas and pore size distributions of GACSC and GACWC obtained using the standard BET equation at a relative pressure in the range of 0.0-1.0 were 52.11 m 2 /g and 27.37 nm, and 46.26 m 2 /g and 21.69 nm, respectively.

Physical surface structures of the granulated-activated biomaterials
The results for the physical surface structure of the activated biomaterials investigated using the scanning electron micrographs were presented in Figures 3 and 4. The micrographs of GACSC and GACWC ( Figure 3 a and b; and 4a and b) showed the development of several pores with honeycomb shape, which is predominant all over their surfaces, and transport pores. The high pore distribution is an indication of the effectiveness of the locally processed NaCl agent to create pores. However, GACSC showed a clearer developed pore structure, and higher pore volumes (i.e. rich in porous cavity); this indicates more suitability for adsorption. SEM micrograph of GACWC indicates the presence of some flaky structure and rudimentary pores but was not rich in porous cavity. A network of crack and crevice was observed in the matrix structure of GACWC probably due to the thermal stress on the feedstock caused by increase in temperature during pyrolysis. It can be seen that pores in open and external surfaces are irregular and heterogeneous.

Effect of activating agent on the surface charges of the composite materials
Figure 5, Figure 6 shows the values obtained for the points of zero charge (pH pzc ) of GACSC, GACWC and the developed iron (III) sulphate coagulant (ISC). The pH pzc of GACSC, GACWC and ICS were observed to occur at pH 7.0, 7.0 and 2.5, respectively. Table 1 presents the results obtained for the mean values of the characteristics of the untreated and treated river samples during rainy and dry seasons.

Effect of dose optimisation of prepared iron (III) sulphate coagulant
The results obtained for dose optimisation of ISC by sedimentation tests during the rainy and dry seasons are presented in Table 2. Table 3, Table 4 present the results for the optimisation of the composite materials for the mg/L C removed during the rainy and dry seasons at POU. Figure 7, Figure 8, Figure 9, Figure 10, Figure 11, Figure12, Figure 13, Figure 14 show the corresponding gas chromatograms -mass spectrometry showing the total ions concentrations (mg/L) of the identified probable organic substances in the untreated and treated water samples during the rainy and dry seasons. The results obtained showed a total of 25, and 31 peaks during rainy season in untreated ERWA, and EBRW samples, respectively. However, a total of 19 different organic compounds were identified in ERWA sample while 26 different organic compounds were identified in EBRW sample during rainy season. The results obtained for the untreated ERWA, and EBRW, samples at dry season showed 99, and 89 peaks, respectively. However, a total of 75 and 49 different probable organic compounds were identified in the samples, respectively. These were achieved through a library search to identify the base peak from the mass spectra, retention time and similarity with the molecular masses of the reference spectra. The identified chromatographic peaks of organic carbons for the untreated ERWA sample during the rainy and dry seasons in Figure 7, 9, 11 and 13 were located at t R = 3.733-26.493 and 8.008-21.376 min, and scan # 36-2760 and 473-2608, respectively. The identified chromatographic peaks of organic compounds for the untreated EBRW sample during both seasons were located at t R = 5.570-19.839 and 5.631-39.153 min, and scan # 324-1962 and 94-5447, respectively. About 19 and 11 peaks were observed from the GC-MS chromatograms after the POU treatment of the ERWA and EBRW samples during rainy season. Out of these observed peaks, a total of 18 and 10 different probable organic compounds were identified from the peaks. At dry season, about 14 and 72 peaks were observed, respectively, from the samples after the POU treatment, while a total of 22 and 57 different probable organic compounds were identified, respectively, from ERWA and EBRW samples. The 22 different organic compounds identified in ERWA sample was comprised of 9 recalcitrant probable organic compounds and 13 organic compounds probably formed after the POU treatment. Similarly, the 57 different organic compounds identified in EBRW sample was comprised of 6 recalcitrant probable organic compounds and 51 organic compounds formed after the POU treatment.

Total removal of TOCs for water after treatment process
About 15 and 9 different probable organic compounds, respectively, were observed to have been introduced from the n-hexane extract into untreated ERWA and EBRW samples, while about 11 and 13 different organic compounds were introduced after the POU treatment into the respective water samples during the dry season ( Figure 15). The % TOCs removal from ERWA and EBRW samples during rainy and dry seasons was 90 and 100%, and 88 and 88%, respectively. Tables 5 to 8 show the names of the identified organic substances from the GC-MS chromatograms.

Discussion
Assessment of the total organic carbon in drinking water is very important to forestall formation of carcinogenic or mutagenic disinfection by-products (DBPs) which may result during treatment processes [41]. The results obtained for the dose optimisation of composite materials indicated that increase in the dose of the adsorbent decreased the concentration in mg/L of organic substance in water and attained optimum reduction at the treatment combination of 0.5, 1.5, and 0.025 g/250 ml of GACSC, GACWC and ISC, respectively, during the rainy and dry seasons. Beyond this optimum dose, the concentration in mg/L of organic substance in the water samples remained constant. This observation agrees with [42] that beyond this maximum dose, the amount of adsorbed  and non-adsorbed particles remained constant even with further net increase in the surface area of the activated biomaterials. From the results of the GC-MS analysis, it was observed that there were more organic substances identified in the untreated ERWA, and EBRW samples at the dry season than the rainy season. The increased in TOC budget for the river water samples during the dry season was attributed to pollution resulting from increased anthropogenic imputes in the rivers due to recession of the water levels and decomposition of leaf litter from trees along the banks of the rivers [43,44].
It could also be apportioned to some organic substances introduced into the water samples from the n-hexane used for the extraction of organic substances based on the n-hexane chromatogram ( Figure 15) which showed about 24 probable organic compounds present as impurities in the n-hexane. Out of these 24 possible organic  compounds, 15 were introduced into the untreated ERWA and 9 into the EBRW during the extraction process.
During the rainy season, two organic compounds were recalcitrant after treating ERWA samples at POU. About 10 residual TOCs identified in the treated EBRW sample were entirely new compounds. At dry season, a total of 9 and 6 organic compounds were recalcitrant after treating ERWA and EBRW while a total of 13 and 51 were new organic compounds identified in ERWA and EBRW samples after treatment. About 11 and 13 organic carbons were introduced into the treated ERWA and EBRW samples during extraction process. The new organic compounds identified in the treated water samples may have been formed from the reactions catalysed by the GAC surface and/or brought  about by the presence of bacteria in treated water samples probably due to increase in temperature [45,46]. On the other hand, the recalcitrant organic carbons identified in the treated water samples may have been as a result of displacement from the adsorption sites by other organic carbons or as a result of desorption process [46].
However, there were about 90, 88, and 100, 88% reductions in the TOCs in ERWA and EBRW samples during rainy and dry seasons, respectively. The reason for the reduction in the TOC content of the water samples may have been due to the ambient alkaline pH of the water samples which resulted in the formation of polynuclear cation of the iron (III) coagulant that neutralised the negatively charged colloid in the water samples [47]. Increasing pH of aqueous solution increases TOC removal by coagulation process due to the fact that more anionic sites are introduced into the carboxyl groups of the organic carbons by deprotonation [48]. This  results in the metal ion species interacting electrostatically with the anionic organic carbon to form flocs.
On the other hand, the removal of dissolved organic carbons from aqueous solution by adsorption mechanism depends on the characteristics of the activated  biomaterial, the adsorbate, and solution pH [49]. The author revealed that higher molecular weight organic carbons are adsorbed onto granulated activated biomaterials by accessing the mesopores while low weight molecules access the micropores. Consequently, it could be inferred that since the activation of the biomaterials in this research resulted in the formation of Type IV isotherm in which adsorptions were due to monolayer-multilayer adsorption process (Figures 2 and 3), the activated biomaterials produced is efficiently adsorbing through the mesopores and macropores as well as the micropores, and on the external surface by electrostatic interaction. The considerable reduction in the TOC of the water samples could also be attributed to the fact that the surface of the GACSC and GACWC were slightly basic or neutral ( Figure 5). As reported by [49], heat treatment (carbonisation) of activated carbon results in the progressive decrease of the acidic surface produced during activation due to the removal of carboxyl and phenolic groups, and this reduction increases adsorption of organic molecules during water treatment.

Conclusion
The results obtained showed that 19 and 26 probable organic substances were identified from GC-MS chromatograms of ERWA and EBRW samples during rainy season while 75 and 49 probable organic compounds were detected from the chromatograms of ERWA and EBRW samples. After treatment at POU, a total of 18 and 10 probable organic compounds were detected from ERWA and EBRW samples during rainy season while 22 and 57 probable organic compounds were identified during dry season from the respective river samples. Out of the 22 and 57 probable organic substances detected, 9 and 6 probable recalcitrant organic compounds, and 13 and 51 probable new organic compounds were identified from ERWA and EBRW samples. Consequently, 90 and 100% TOCs removal was recorded for ERWA and EBRW samples during the rainy season while 88 and 88% TOCs removal was observed for ERWA and EBRW at dry season after treating the river samples at POU with optimal doses of 0.5, 1.5, and 0.025 g of GACSC/GACWC/ISC composite materials.

Recommendation
Based on the findings of the present study, it is therefore suggested that the developed composite material be used to replace the commercial 'Water Guard' in treating water at POU.