Assessment of metals and anions in tap, river, wastewater, and sludge: comparison of hotplate-and microwave-assisted digestion

,


INTRODUCTION
The contamination of water bodies by heavy metals can originate naturally from the weathering of minerals and rocks, or from anthropogenic sources, such as sewage discharge, urban and agricultural runoff (Jaishankar et al., 2014).Surface treatment processes using heavy metals, as well as industrial products that are discharged at the end of their life, are the major industrial sources of heavy metals that end up in wastewater.Major urban inputs to sewage water are effluents from households and businesses, and traffic-related emissions as these can be transported with stormwater into the sewerage system.These can all lead to an increased amount of heavy metals reaching wastewater treatment plants, whereafter they are discharged with the treated effluent where they may accumulate in aquatic life and enter the food chain.Also, metal contamination from river water and metal leaching from water distribution systems may lead to the presence of heavy metals in drinking water, which may result in severe human health effects, especially under significant exposure to high concentration levels (Atlas et al., 2017;Smirnova et al., 2021).Moreover, the pollution of the environment by these heavy metals is a long-term and irreversible process (Olujimi et al., 2012).
Anion (chlorides, sulphates, phosphates, nitrates, ammonia) contamination is another common environmental problem if present in high concentrations.For example, nitrates (commonly used in agricultural activities) can affect the transportation of oxygen in the blood whilst excessive phosphates can result in eutrophication and algal blooms.Chlorides are common since chlorination is a widely used water treatment process that may increase the chloride ions present in water bodies (Altundag et al., 2019).Despite chloride being an essential nutrient, high levels of consumption can lead to kidney disorders and increased blood pressure.In addition, chloride may lead to the formation of toxic disinfection by-products resulting in cancers of vital organs (Lehtonen et al., 2019).
In wastewater treatment processes, large amounts of sludge are generated and heavy metals in the wastewater influent may become concentrated in the sludge.When this sludge is used as manure on agricultural land or wastewater effluent is used for irrigation, it may transfer metals to crops.This can negatively affect the productivity of crops, and threaten animal and human health (Yamgata et al., 2010).Despite many European countries having managed to decrease environmental pollution through the implementation of legislation, improved treatment processes, and eco-friendly industrial activities, developing countries still struggle to control environmental pollution.Therefore, new and efficient methods of treatment and consistent monitoring of water resources including wastewater are required (Olujimi et al., 2012).
Environmental sample matrices can be complex, thus requiring sample preparation prior to instrumental analysis.As a result standard acid digestion method is often used as a sample preparation method.The method is then validated by evaluating the effectiveness of metal recovery to assess the method's ability to completely digest the sample matrix (Jaishankar et al., 2014).Microwaveassisted digestion and ultrasonic digestion have become increasingly adopted due to their higher metal recovery rate as a result of minimal sample contamination (Hu et al., 2014).Microwave-assisted digestion has several advantages over conventional hotplate digestion, including retaining volatile analytes, and rapid heating and cooling (Hu et al., 2014).Generally, an acid digestion reaction depends on several factors including the type of acid used and its concentration, the time of digestion, and the metal form present in the sample matrix (Das and Ting, 2017).
The aim of this study was therefore to compare microwave and hotplate digestion methods in terms of total dissolved and recoverable metal concentrations in different water matrices when assessing the heavy metal and anion contamination in tap, river, and wastewater.To the best of our knowledge, no reported work has been conducted to assess the concentration levels of the studied metals and anions in the selected study areas.

EXPERIMENTAL Sample storage and collection
Tap water samples were obtained from Richmond Crest, Mkhondeni, Woodlands, Boughton and Scottsville, suburbs in the Pietermaritzburg area.River water samples were collected along the Msunduzi River, at Bishopstowe, College Road, Camp's Drift, Woodhouse, and YMCA (Fig. 1) Wastewater samples were obtained from three wastewater treatment plants (WWTPs) in the city of Durban: Amanzimtoti, Umhlathuzana and Northern.In all WWTPs, water samples were taken in the influent, the effluent and liquid sludge.The Umhlathuzana WWTP receives influent from Marianridge and Shallcross which are then combined into one effluent after treatment, and discharged into the Umhlathuzana River.Amanzimtoti WWTP discharges into the Mbokodweni River; Northern WWTP discharges into the Mgeni River.River water samples were taken from the rivers where the WWTPs of interest discharge their effluent.About 2.5 L of water was collected in polyethylene bottles and immediately placed in a portable ice chest.Samples were transported to the laboratory and filtered using a 0.45 µm membrane filter consisting of biologically inert mixtures of cellulose acetate and cellulose nitrate (Merck, Darmstadt, Germany).

Reagents, reference materials, and standards
Ultrapure water was employed in the preparation of standard solutions for the calibration of the ICP−OES.55% v/v nitric acid (HNO 3 ) (Merck, Darmstadt, Germany) was used for cleaning glassware and to digest the water samples since it liberates the trace metal elements as the soluble nitrate salt.Standard solutions were prepared by appropriate dilutions of a stock standard of 1 000 mg/L (Sigma Aldrich, South Africa).The standard reference material for trace elements in water (Ultraspec Multi-Element Aqueous CRM) was employed to evaluate the accuracy of the method employed for quantification of heavy metals in water samples.

Instrumentation
The sample digestion was performed using Multiwave 5000 (Anton Paar, Johannesburg).The Varian 720-ES ICP−OES (inductively coupled plasma−optical emission spectroscopy) was used for the determination of metals in water samples.The instrument operated at a frequency of 40 MHz and RF power of 1.00 kW, and consisted of a pneumatic concentric nebulizer with a flow rate of 0.75 L/min and a pump rate of 15 r/min.The inert carrier gas used was argon (Ar), with a plasma flow of 1.50 L/min.Three replicates were read, with a replicate read time of 1 s.These conditions remained constant throughout the analysis.Table 1 summarizes the conditions and optimal wavelengths used for each metal element.

Sample preparation
The digestion of the samples was done using a hotplate or microwave in order to determine the total recoverable and total dissolved metals in all the water samples.Spiked recovery tests were conducted for the optimization studies where the recoveries were calculated for all digestion methods.

Acid digestion by heating
The United States Environmental Protection Agency (EPA 3005A) digestion method was used to determine heavy metals in water.For total recoverable metals, a 100 mL water sample was transferred into a glass beaker.5 mL of 55% v/v nitric acid was added and the beaker heated on a hotplate to allow the contents to evaporate and reduce to around 20 mL.The sample was cooled for 5 min, and another portion of 5 mL nitric acid was added and further heated for 15 min.The sample was then cooled and transferred into a 100 mL volumetric flask and filled up to the mark with ultrapure water.For the determination of total dissolved metals, a 50 mL water sample was transferred into a 100 mL volumetric flask followed by the addition of 10 mL of 55% v/v nitric acid, and made up to the mark with ultrapure water.No heating was required for the determination of total dissolved metals as evaporation alters the amount of the sample.The samples were then analysed using ICP−OES.

Microwave-assisted acid digestion
The American Society for Testing and Materials method (ASTM-D4309-18) was followed for sample digestion to determine the total recoverable metals in water.To a 50 mL water sample, 5 mL of nitric acid was added and gently swirled.The sample−acid mixture was then digested by heating it to 170 ± 5°C for 10 min and maintaining this temperature for 10 min.For the determination of total dissolved metals, the EPA 3015A method was followed.The procedure was similar to that used for the total recoverable metals; however, the microwave programme involved heating to 170 ± 5°C for 20 min and maintaining this temperature for 10 min.After the digestion process was complete, the vessels were removed from the microwave reaction and transferred into ICP tubes (in a fume hood) for analysis.Both digestion methods were validated based on spiked recovery tests to assess the accuracy of the digestion methods.

Determination of anions
The Aquakem 250 discrete selective photometric analyser was employed for the determination of anions in river and wastewater samples.Sulphate (SO 4 2-) ions were precipitated by barium chloride in a strongly acidic medium.The resulting turbidity was measured photometrically at 405 nm.Chlorides (Cl -) reacted with mercury (II) thiocyanate to form a soluble non-ionic compound.The thiocyanate ions released reacted with iron (III) nitrate to form a red/brown iron (III) thiocyanate complex.The resulting intensity of the stable colour produced was measured spectrophotometrically at a wavelength of 480 nm (Aquakem Labmedics, 2006;ALS, 2016).Nitrates (NO 3 -) were reduced to nitrites with hydrazine sulphate under alkaline conditions.The total nitrite ions were then reacted with sulphanilamide and N-1-naphthylenediamine hydrochloride under acidic conditions to form a pink azo-dye and the absorbance was measured at 540 nm.For the determination of phosphate ions (PO 4 3-), orthophosphate ions reacted with ammonium molybdate and antimony potassium tartrate (catalyst) under acidic conditions to form a 12-molybdophosphoric acid complex (Aquakem Labmedics, 2006;ALS, 2016).The complex was then reduced with ascorbic acid to form a blue heteropoly compound.The absorbance of this compound was measured spectrophotometrically at a wavelength of 880 nm.Ammonia (NH 3 -) reacted with hypochlorite ions generated by the alkaline hydrolysis of sodium dichloroisocyanurate to form monochloramine.This was then reacted with salicylate ions in the presence of sodium nitroprusside at pH 12.6 to form a blue compound.The absorbance of this compound was measured spectrophotometrically at 660 nm.

Effect of digestion acid type on the recovery of metals
The choice of acid or acid-mixture is crucial since it controls the effectiveness of the digestion process.Concentrated hydrochloric and nitric acids were used as pure or in ratios of 3:1 and 1:3 to digest the water samples spiked with a concentration of 0.50 mg/L.The percentage recoveries for the hotplate digestion ranged from 78-117%, 74-111%, 57-102% and 88-116% for 3:1 (HNO 3 :HCl), 1:3 (HNO 3 :HCl), HCl and HNO 3, respectively (Fig. 2).The microwave-assisted digestion yielded recoveries ranging from 62-95%, 67-111%, 66-113% and 83-103% for 3:1 (HNO 3 : HCl), 1:3 (HNO 3 : HCl), HCl and HNO 3 respectively (Fig. 3).The concentrated HNO 3 provided recoveries above 80% for all metals in both methods and hence was chosen as the most suitable.This could be due to the good ability of HNO 3 to extract a wide variety of metal salts, while HCl is suitable for metals in the form of carbonates, phosphates, borates, sulfides, and some oxides.Also, metals in the waste matrix tend to form soluble metal salts when subjected to oxidative acid digestion reactions (Das and Ting, 2017).

Effect of sample volume on the recovery of heavy metals in hotplate digestion
The effect of sample volume was investigated using 25, 50, and 100 mL of tap water samples.The results showed an increase in all metal recoveries with an increase in sample volume, with recoveries ranging from 83-99% for 100 mL volume (Fig. 4).This could be due to the fact that increasing the sample volume also increases the amount of metals available for digestion and ultimately the concentration recovered in the digestion solvent.Also, the digestion process for large sample volumes takes longer, and this may improve the concentration of metals recovered due to increased contact time.The statistical analysis also confirmed that the mean recovery result for 100 mL is significantly different at the 5% level from that for 25-and 50-mL sample volume (Table A1).
The 100 mL sample was then taken as the optimum volume for further analysis.

Recovery of metals from different water matrices
The tap, river, and wastewater matrices were spiked with a 0.50 mg/L metal mixture, digested, analysed, and the percentage recoveries calculated.There was no trend in metal recoveries from all water samples which indicated that the recoveries are independent of the sample matrix.The recoveries were found to be within an acceptable range of 72-119% (depicted in Fig. 5a-d).
Total recoverable metals were higher than the total dissolved metals for both digestion methods.This is expected since the total recoverable determination takes into consideration both the suspended and dissolved metal concentrations.However, it was also observed that for some metals the total dissolved recoveries were higher, and this could be due to the sample reduction step in the total recoverable determination where the analyte is lost via evaporation during the heating process (Sastre et al., 2002).A t-test showed that the mean recoveries were not significantly different at the 5% level (Table A2).

Effect of spiking concentration on metal recovery
The effect of sample spiking concentration on total recoverable and total dissolved metals for the certified reference material was investigated at 0.10, 0.50, and 1.00 mg/L spike levels.There was no trend observed in the percentage recoveries for the different spiking concentrations therefore it can be reasoned that the recoveries are independent of the sample spiking concentration (Figs 6 and 7).A t-test confirmed that the results were not significantly different at the 5% level (Tables A3, A4, A5).

Physicochemical properties of water samples
Dissolved oxygen (DO), total dissolved solids (TDS), temperature, pH, conductivity, and salinity were measured before the determination of metal concentrations (Table 2).The temperature of tap and river water ranged from 17.4-22.6°Cand 17.1-23.2°C,respectively, while for wastewater it ranged from 13.2-24.0°C.
Studies have shown that an increase in temperature can result in higher maximum sorption of metals by minerals; however, the average dissolved metal concentrations showed no dependence on temperature at 4-25°C (Huang et al., 2017).A study conducted by Li et al. (2013) showed that Pb concentrations increased with increasing temperature (15-35°C).However, no significant concentration variation was observed.It was also observed that Cd was only detected at temperatures of 30 and 35°C.This is because the oxidiz-able fraction of the metal is transformed easily in chemical reactions that occur when the temperature is increased (Li et al., 2013).
The pH of the collected tap water ranged from 6.2-6.9, within the WHO recommended range for drinking water (6.5-8.5 ).
An acidic pH can result in the presence of metals (e.g., Fe, Mn, Cu, Pb, and Zn) in drinking water due to leaching from plumbing systems (Rahmanian et al., 2015).The pH in river and wastewater samples was from 7.2-9.0and 7.1-7.6,respectively, which is slightly basic, and this could be due to the presence of carbonates, bicarbonates, and hydroxides originating from limestone found in the riverbed (Reeve, 2002).Metal ions can also be converted into poorly soluble forms which tend to adsorb on suspended materials present in slightly basic water if there is a high amount of dissolved oxygen.
The DO levels in tap water ranged from 2.41-3.66mg/L, in river water from 0.64-2.90mg/L and in wastewater from 0.33-2.61mg/L.The presence of organic and or inorganic material in water depletes oxygen.For example, Fe 2+ can deplete oxygen via oxidation to form Fe 3+ (Reeve, 2002) and oxidation processes can be a possible explanation for the differences between the DO levels in tap and river water samples.Oxidation processes are used in treatment of drinking water, thus increasing DO levels.Therefore, tap water is expected to have higher DO levels compared to river water since microorganisms will significantly decrease DO (Li et al., 2013).Salinity in wastewater ranged from 0.29-0.64psu and was higher than that of river water (0.10-0.15 psu) and tap water (0.19-0.44 psu).Conductivity of tap, river and wastewater samples ranged from 187-758 µS, 210-888 µS, and 608-1 312 µS, respectively.The maximum allowable limit of conductivity in water as per the NDWQS guidelines (Rahmanian et al., 2015) is 1 000 µS.
Total dissolved solids (TDS) in tap, river, and wastewater samples were from 106-243 mg/L, 105-163 mg/L and 304-658 mg/L, respectively, which were all below the acceptable limit of 1 000 mg/L in drinking water (WHO, 2004)

Determination of metals in wastewater
There was no trend observed between total recoverable and total dissolved metal concentrations in wastewater (Table 3).Co and Li concentrations were found to be higher in influent samples, while they were either below quantification or detection limit in the corresponding effluent samples, indicating their partial removal by the WWTPs.Mn was present in all sludge samples (192-785 µg/L) and higher concentrations were observed for total recoverable than total dissolved Mn.This could be due to the fact that liquid sludge contains some solid particles which can increase the adsorption of Mn and, since total recoverable determination includes dissolved and suspended metals, high concentrations are expected (Addo-Bediako et al., 2018).Ni was only quantified at the Amanzimtoti WWTP influent for total recoverable and total dissolved (118.9 and 235 µg/L).It was below the quantification limit in the effluent and other samples -an indication of efficient removal of Ni at the Amanzimtoti WWTP.Pb, Sr, and Zn concentrations were quantified in all waste and corresponding river water samples (where WWTPs discharge), and were found to be below the maximum permissible limits, with the exception of Pb.The highest Pb concentration was found at the Northern River WWTP for both total recoverable and dissolved determinations.Amanzimtoti WWTP had higher Zn concentrations in the effluent for total dissolved determination.This might be due to particulate Zn in the influent being transferred to the aqueous phase in aeration tanks used in the WWTPs.Also, a higher amount of metals in the activated sludge may be transferred to the aqueous phase in aeration tanks which might increase the amount of Zn present in the effluent (Yamagata et al., 2010).Co was detected in Amanzimtoti influent and sludge at 37.4 µg/L and 28.6 µg/L, respectively, as well as in Northern Works influent and sludge at 37.6 µg/L and 40.7 µg/L, respectively.Li was detected in Amanzimtoti influent and sludge at 13.9 µg/L and 20.3 µg/L, respectively, and in Northern Works sludge (17.6 µg/L) and Umhlathuzana influent (7.3 µg/L).Lastly, Cd, Cr, Cu, Co, Li, Ni, and Tl were below detection or quantification limits.

Determination of metals from tap and river water
The average concentrations observed for total recoverable metals were much higher than for total dissolved metals in tap and river water (Table 4 and 5).This was expected since the total recoverable metal concentrations consider the soluble and insoluble metals (unfiltered samples) whilst the total dissolved metal concentration only considers the soluble metals as particulates (insoluble) are removed by filtration (USEPA, 1994).However, some metals had higher total dissolved compared to total recoverable concentrations, especially when using microwave digestion.This could be due to the microwave method being a closed-system digestion, which, apart from a considerable reduction in digestion time, also results in minimal sample contamination and loss of volatile metals such as As, Hg, and Cr (Sastre et al., 2002).In tap water, Li concentrations were approximately the same (4.9-5.5 µg/L) for all samples for both digestion methods, whereas for river water, Li was only quantified in Woodhouse River water (34.5 µg/L) using the microwave digestion method.There is no maximum permissible limit set for Li in drinking water; however, the obtained values were below the oral reference dosage which is 700 µg/L (USEPA, 2003); hence the analysed tap water can be assumed to be safe for consumption.Li has also been detected in tap water at 20-160 µg/L and 0.7-59.0µg/L from Texas and Japan, respectively (Ohgami et al., 2009), and the maximum concentrations were higher than those obtained in this work.The highest concentration obtained for Sr was 90.8 µg/L (Richmond Crest) which is lower than that reported in drinking water from China (1 690.0 µg/L) by Zhang et al. (2018).
The highest concentration of Zn (142.3 µg/L) was observed in the Scottsville tap water sample.The presence of Zn (147.6-307.1 µg/L) in drinking water from Jordan has also been reported (Massadeh et al., 2020), at higher concentrations than those reported here, but still below the permissible limits.Pb was the only metal found in all tap water samples (59.2-155.1 µg/L) that was present above the permissible limit of 50.0 µg/L.This indicates that tap water from all the sampling areas is not safe for human consumption, as Pb is one of the most toxic metals, which can lead to permanent damage to the nervous system, brain, and kidneys in humans and animals.Pb has various industrial applications which can result in lead contamination of water supplies through indirect pathways.
It is commonly found in batteries, water distribution piping, and paints, and occurs as an organic compound, alkyl lead, in gasoline.
The presence of Pb in tap water can be from the dissolution of household plumbing systems where the pipes, solder, fittings, or service connections to homes contain Pb (Mebrahtu et al. 2011;Mehdizadehtapeh et al., 2017;Fajri et al., 2023).The dissolution of Pb is generally increased in soft water, i.e., water containing low levels of calcium carbonate (SAWQG, 1996).These results agree with those reported by Massadeh et al. (2020), where concentrations of Pb above permissible limits were observed in drinking water from Jordan (7.7-60.6 µg/L), suggesting that Pb contamination of drinking water is a prevalent global problem.Cu and Co were not found in tap water samples, while Cr, Tl, Mn, Ni, and Cd were either below detection or quantification limits.Even though the concentration levels for all other heavy metals in this study are within the permissible limits, the high Pb concentration needs to be continuously monitored to ensure that tap water is safe for human consumption, with appropriate treatment processes of coagulation with alum, ferric salts or lime thereafter followed by settlement and filtration (SAQGW, 1996).The increased concentration may be due to the sorption of Zn by hydrous metal oxides, clay minerals and organic material commonly found in river systems.However, Zn can be toxic to organisms when present in higher concentrations (Mebrahtu et al., 2011).The presence of Zn could also be due to pesticide and fertilizer contamination through agricultural runoff (Oguzie et al., 2010).Sr and Pb concentrations were detected in all samples and were below the permissible limits, with the exception of Pb at YMCA and College Road (Table 5).The presence of Sr in water could be due to the weathering of natural rocks as well as the direct discharge of wastewater into rivers.The possible sources of high Pb concentrations in river water could be exhaust emissions from motor vehicles that can make their way into river systems.This can have adverse effects on the surrounding environment such as inhibiting the growth of plants and affecting the central nervous system of humans upon consumption (Mebrahtu et al., 2011).The maximum concentration of Pb (51.8 µg/L) obtained in this work is lower than that reported by Olujimi et al. (2018) for river water from Gauteng (86.73 µg/L); however, both studies recorded Pb levels above the permissible limits.The concentrations of Cd, Cr, Mn, Ni, Ga, and Tl were found to be below the detection or quantification limits in river water samples.
In general, it was observed that microwave-assisted digestion was more sensitive as higher concentrations of metals were detected and quantified in all samples.Some metal concentrations quantified using the microwave-assisted method were either below the detection or quantification limits for the hotplate method.
For both digestion methods, statistical analysis showed that the concentrations obtained were not significantly different (Table A6).In addition to the t-test, a one-way ANOVA was conducted on the two digestion methods which revealed that there is no significant difference between the methods as F critical > F value with p ≥ 0.05 (Table A7).It can be concluded that the hotplate method can be recommended for daily routine analysis as it is a cheaper technique compared to microwave-assisted digestion.The reason for the differing concentrations of heavy metals in water samples is that despite the assumption that applications are the same in nearly all countries, consumption patterns for chemicals may be different.For instance, some applications that may have been phased out in some countries may be widely used in other countries resulting in the presence or absence of certain heavy metals in water systems (Olujimi et al., 2012).
There were some inconsistencies concerning total dissolved metals being higher than total recoverable metals.These non-correlations may be subjected to sample matrix interferences, by the loss of sample volume during the digestion process, sample contamination during analysis, and high volatility of metals in the presence of high temperatures (Sastre et al., 2002;Lomonte et al., 2008).

Metal removal efficiency of WWTPs
The removal efficiency (%) for heavy metals in WWTPs was calculated using Eq.1: where: C influent and C effluent are the concentrations obtained in the raw influent and final effluent, respectively.12.2 ±0.09 bdl bdl -below detection limit, bql -below quantification limit, BS -Bishopstowe, CD -Camps Drift, CR -College Road, WH -Woodhouse, RC -Richmond crest, SV -Scottsville, MN -Mkhondeni, WL -Woodlands, BT -Boughton Cd and Li were completely removed from the Amanzimtoti WWTP.Also, Mn was completely removed at Umhlathuzana, while −106% and 3.9% were removed at Amanzimtoti and Northern Works, respectively.Pb removal was 25%, −125%, and 14.7%; Sr removal was 37.6%, 55.4%, and −20.3%;Zn removal was 07.9%, −9.1% and −30.1% at Amanzimtoti, Umhlathuzana and Northern, respectively.Zn showed negative removal in all WWTPs which indicated its high persistence within the wastewater treatment plants.A study conducted in Japan by Yamagata et al. (2010) revealed that Zn on adsorbed particulates could be easily removed; however, there was difficulty in removal of dissolved Zn in the influent during the activated sludge process.Amanzimtoti achieved better removal efficiency for most of the metals compared to Umhlathuzana and Northern Works.These results indicate that the WWTPs contribute to heavy metal pollution of the rivers they discharge their effluents into.

Anion concentration
Phosphates were detected in all water samples (Table 6).The presence of some phosphates in water is natural as it is an essential nutrient; however, agricultural (over-fertilization) and industrial resources tend to increase the phosphates in natural surface water which may result in eutrophication and excessive algal blooms (Altundag et al., 2019).The concentration of phosphates was found to be lower in the effluent compared to the corresponding influent which could be due to the treatment processes applied in the plant and also adsorption on the sludge (El-Nahhal et al., 2014).The highest phosphate concentration was found in sludge samples from Amanzimtoti WWTP (175 mg/L), possibly because of the solubility and pH effect, as low pH values permit the adsorption of phosphate on sludge.Sludge samples had a pH of 7.36 ± 0.03; hence, phosphates were detected as this pH influenced them to strongly bind to the sludge.Also, phosphoric acid is a weak acid with three dissociations; hence the phosphate levels observed in the sludge are those found in a neutral pH.However, more phosphates are expected to be found at a more acidic pH (El-Nahhal et al., 2014).The highest concentration of phosphates in river water was observed in Amanzimtoti River (13.0 mg/L), which was than that reported for the Sakarya River in Turkey (2.72 mg/L) (Altundag et al., 2019).
Chloride concentrations were found to be below the maximum acceptable values, except in the Amanzimtoti and Northern River water samples.Scottsville tap water had the highest chloride level of all tap water samples (8.12 mg/L), but this was still below the maximum allowable limit.Chlorides are expected to be present in tap water since chlorination is used in water treatment.In river water, the highest concentration above the maximum limit was found in Amanzimtoti (3 064.54 mg/L), which was higher than that observed at the Sakarya River in Turkey (78.52 mg/L) (Altundag et al., 2019).The chloride concentration in the river water could be due to natural sources such as weathering of rocks and concentrations can increase because of evaporation.
The highest nitrate concentration in river water was observed at Bishopstowe (6.42 mg/L), but this value is lower than that observed in the Sakarya River (920 mg/L) by Altundag et al. (2019).Nitrates can reach surface waters from agricultural activity (fertilizers), oxidation of nitrogenous wastes, human and animal excreta.Scottsville tap water recorded the highest nitrate concentration for all tap water samples.Nitrates in tap water may be due to nitrite being formed chemically in distribution pipes by Nitrosomonas bacteria during the stagnation of nitrate-containing and oxygen-poor drinking water in galvanized pipes.Another reason could be when chlorination (not a well-controlled process) is used as a disinfectant thus increasing the nitrate concentration in the tap water (Koch, 1984).The reduction of nitrate to nitrite gives rise to its toxicity in humans and high concentrations can cause the oxidation of haemoglobin (Hb) to methaemoglobin (metHb), which is unable to transport oxygen to tissues (Koch, 1984).In wastewater, Amanzimtoti WWTP sludge had the highest concentration of nitrates (39.57mg/L) while the influent had the lowest concentration (0.32 mg/L).The Northern Works and Umhlathuzana also produced high concentrations in the effluent compared to the influent, though still below acceptable levels.The high concentration of nitrates in effluent compared to influent could be due to the re-concentration of cations and anions which results in the conversion of ammonium hydroxide to nitrates in the presence of oxygen and nitrifying bacteria within the WWTP (El-Nahhal et al., 2014).
Ammonia levels were found to be the highest in Richmond Crest tap water (23.51 mg/L), while the highest concentration of ammonia in river water was observed at Woodhouse (3.91 mg/L).Ammonia can have toxic effects on humans when consumed in large amounts resulting in compromised capacity to detoxify.High levels of ammonia can also lead to toxic build-up in tissues and blood in aquatic organisms.The main source of ammonia is human faeces containing a high protein fraction due to high consumption of protein (El-Nahhal et al., 2014).All WWTPs successfully removed the ammonia from the influent water resulting in low concentrations in the effluent and river water samples.The high concentrations of ammonia in influent could be due to fertilizers and other agricultural products containing ammonia (Altundag et al., 2019).
The highest concentrations of sulfates above permissible limits were observed in the river samples for both Amanzimtoti (437.80 mg/L) and Northern Works (292.47 mg/L).Sulphate is the most common anion after bicarbonate and chloride (Altundag et al., 2019).Sulphates occur naturally in minerals such as barite, epsomite, and gypsum, and this can contribute to the sulphate content in drinking water.Other sources include fertilized agricultural lands and sewage treatment plants (Altundag et al., 2019).The high concentration of sulphates in the water samples may also be due to the high solubility of sulphates in river water which is not pH dependent (El-Nahhal et al., 2014;Jing et al., 2013).Although high sulphate concentrations can result in dehydration, it has been reported that humans can adapt to high sulphate levels with time (WHO, 2004).
The common sources of the analysed anions were assessed by performing a principal component analysis (PCA) on the results obtained (Table 7).In Component 1, phosphates, chlorides, and nitrates are closely associated with each other; this implies that they may originate from the same source, which could be agricultural fertilizers.Similarly, ammonia and sulphate are closely associated with each other.This is evident from the component plot in the rotated space (Fig. 8).Both ammonia and sulphates are key ingredients in soil fertilizer, e.g.ammonium sulphate, which is an inorganic salt, is commonly present in fertilizers and also has various commercial uses (Altundag et al., 2019).

CONCLUSION
This study focused on the development of two commonly used digestion methods: microwave-assisted and hotplate digestion.
The parameters investigated included digestion acid type, acid combinations, sample volume, and spiking concentrations.Concentrated nitric acid (55% v/v) and a sample volume of 100 mL proved to be optimal conditions for sample digestion.Both digestion methods proved to be reliable; however, hotplate digestion was recommended for daily analysis as it is a more accessible and cheaper method.Microwave-assisted digestion provided evident advantages as a sample digestion method with high metal recovery.Total and dissolved metals were quantified using the ICP−OES.All metals (except Pb) and anions (except for chlorides and sulphates in Amanzimtoti and Northern rivers) were below the maximum permissible limits.PCA analysis grouped phosphates, chloride and nitrates (anions) indicating that they are potentially originating from the same source while sulphates and ammonia were also grouped suggesting that they could be from the https://doi.org/10.17159/wsa/2024.v50.i3.4044 same source.Lead in particular was present in tap water at levels above the legislative limit for drinking water.The possible sources included water distribution piping in households, paints, and other organic compounds in gasoline through indirect pathways.In some instances, total dissolved metals were higher than total recoverable metals.It was observed that WWTPs also contribute towards the presence of metals in rivers.The efficiency of the wastewater treatment plants was calculated by percentage removal.This indicated that treatment processes require improvement or perhaps new processes to be implemented to remove these pollutants before discharge into river systems, in order to safeguard human health on consumption and to ensure clean freshwater resources.https://doi.org/10.

Figure 1 .
Figure 1.Map showing river water sampling points along the Msunduzi River

Figure 5 .
Figure 5.The effect of different water matrices on the total (a) and dissolved (b) metal recovery using hotplate digestion; and total (c) and dissolved (d) metal recovery using microwave-assisted digestion.

Figure 6 .
Figure 6.The effect of different spiking concentrations on metal recovery using hotplate digestion

Table 1 .
Operation conditions of the ICP−OES

Table 6 .
Concentrations of anions (mg/L) in water samples

Table 7 .
PCA for anions in different water matrices component plot in rotated space Rotated component matrix for anions in different water matrices

Table A2 .
Statistical analysis of effect of digestion method on metal recoveries: t-test assuming unequal variances for the three water matrices

Table A3 .
Statistical analysis of effect of digestion method on total recoverable metals in wastewater influent: t-test assuming unequal variances for the three spike concentrations

Table A4 .
Statistical analysis of effect of digestion method on total dissolved metals in wastewater influent: t-test assuming unequal variances for the three spike concentrations

Table A5 .
Statistical analysis of effect of digestion method on metal recoveries in tap water: t-test assuming unequal variances for the (TR: total recovered metals; TD; total dissolved metals)

Table A6 .
Statistical analysis of effect of digestion method on metal recoveries in river water: t-test assuming unequal variances for the (TR: total recovered metals; TD; total dissolved metals)

Table A7 .
Results of ANOVA for effect of digestion method (hotplate and microwave) on hotplate and microwave digestion for total recoverable metals in tap, river and wastewater samples