Synthesis and performance evaluation of adsorbents derived from sewage sludge blended with waste coal for nitrate and methyl red removal

Low-cost adsorbents were synthesized using two types of sewage sludge: D, which was obtained during the dissolved air flotation stage, and S, which was a mixture of primary and secondary sludge from the digestion and dewatering stages. The sewage sludge was mixed with waste coal before being activated with potassium hydroxide (KOH) and oxidized with ammonium persulfate (APS). The nitrate and methyl red removal capacities of the synthesized adsorbents were evaluated and compared to those of industrial activated charcoal. The oxidation surface area of adsorbents derived from sludge S shrank by six fold after modification i.e., from 281.72 (unoxidized) to 46.573 m2/g for the oxidized adsorbent with a solution of 2M ammonium peroxydisulfate, while those derived from D only varied narrowly from 312.72 to 282.22 m2/g, but surface modification had no effect on inorganic composition in either case. The adsorption of nitrate and methyl red (MR) was performed in batch mode, and the removal processes followed the pseudo second order kinetic model and the Langmuir isotherm fairly well. The adsorption capacities of nitrate and MR were higher at pH = 2 and pH = 4, respectively.

A waste coal sample was collected (using the ply sampling method) from a coal beneficiation facility located 15 kilometers south-west of Witbank in the South African province of Mpumalanga. The coal sample was dried in the laboratory for 48 h and then blended and divided using a Rotary splitter (Eriez) before being placed in a tight bag. A portion of the coal sample was pulverized to a particle size of 0.150-0.250 mm for further www.nature.com/scientificreports/ physicochemical tests, while another portion was crushed to a particle size of − 1 mm (80%) for petrographic examination. The physicochemical analyses of the waste coal sample utilized in this study are summarized in the Table 2 below. Sewage sludge reject from wastewater treatment facilities can be designated as primary or secondary sewage sludge based on its origin. The first type is the underflow from settling vessels following inlet physical treatment (sedimentation, flocculation, filtration, etc.) and is predominantly inorganic; the secondary, also known as biological sludge, is generated from biological process (activated sludge). In this paper we employ two types of sewage sludge collected at the East Rand Water Care Company (ERWAT) wastewater treatment plant facility: D, which was acquired during the dissolved air flotation stage, and S, which was a combination of primary and secondary sludge from the digestion and dewatering stages respectively.
Methods. Raw sewage sludges were sun dried for 24 h before being oven-dried. The following nomenclature is used throughout this paper to refer to all synthesized adsorbents: the letter C denotes the combination of sludge D or S and discard coal C, i.e. DC or SC, followed by the reagent concentration and finally the pyrolysis temperature. DC-5-750, for example, refers to sewage sludge D combined with waste coal C-The reagent concentration is 5M KOH, and the pyrolysis temperature is 750 °C, as stated in Table 2. The capital letter M denotes that the adsorbent has been changed, while the numbers 1 or 2 indicate the oxidation conditions using 1M or 2M ammonium peroxydisulfate (APS), respectively. SBAC ((Sludge Based Activated Carbon) refers to all activated carbons synthesized using sewage sludge alone or in conjunction with waste coal in this study. Dried and crushed precursors (sewage sludge with/without waste coal in a proportion of 0% and 50%) were impregnated with KOH solution with concentrations of 3, 5 and 7 moles/l, and a ratio of activating reagent to the sludge of 1.5:1. After impregnation, the mixture was dried for 24 h at 105 °C. The dried impregnated precursors were crushed to 100% passing through a 150 µm sieve, then placed in a crucible and pyrolyzed in an inert atmosphere for 90 min in a muffle furnace connected to a nitrogen bottle with a flowrate of 50 ml/min. The furnace regulator was turned off when the dwell time was reached, and the pyrolyzed sewage sludge was cooled to room temperature and then washed. The washing stage is divided into two stages: acidic and water. The former was performed using a 3 M HCl solution to remove excess activating reagent and ash-laden soluble ash, whilst the latter was performed using distilled water until the filtrate achieved a neutral pH value of between 6 and 7. SBAC (Sludge Based Activated Carbon) was washed and dried for 24 h at 105 °C, then crushed and sieved to a mesh size of 125 μm before being kept in an airtight container in the desiccator for further testing. The dried sludge was crushed to 100% passing through 125 µm sieves, and pyrolyzed at different temperatures (DC-5-750, DC-7-900, SC-3-600, and SC-5-900). The products from the pyrolysis were mixed with discard coal in a proportion of 1:1 and activated with KOH under N 2 atmosphere at a flow rate of 5 l/min. To achieve surface modification via oxidation, 1 g of pyrolyzed sewage sludge was incubated in 25 ml of 1M or 2M APS dissolved in 1M H 2 SO 4 at 60 °C for 240 min before being washed with distilled water until the pH neutral value was reached. Batch adsorption experiments were carried out with an adsorbent dose of 0.5% (5 mg of adsorbent in 10 ml of solution) and the performance of adsorbents was compared to commercial activated carbon. The best oxidized and/or unoxidized SBACs were chosen based on preliminary evaluation test results conducted at 303 K in an incubator shaker. The initial concentration of nitrate was 50 mg/l, with contact times of 180 and 360 min, respectively, while the initial concentration of MR was 75 ppm, with contact times of 90 and 180 min, and initial pH of 2 and 4 for nitrate and methyl red, respectively. The initial pH value of 2 in the case of nitrate was chosen because adsorbent surfaces are prone to being positively charged as pH decreases, and nitrate ions cannot compete with hydroxyl anions 25 Chemical analysis. SEM and EDS analysis were done using a ZEISS SIGMA FESEM 03-39 apparatus on coated samples on a carbon tape support using Gold-Palladium techniques. TGA was performed with the Pioneer (SDT-Q500) equipment. Proximate analysis of sewage was performed using a Perkin-Elmer STA 6000 simultaneous thermal analyser. FT-IR analyses were done using the Perkin-Elmer FT-IR Spectrometer Spectrum 2. Ultimate analysis was conducted with the thermoscientific flash 2000. XRD study was done using a Bruker 2D phaser instrument. The NH 4 OAC method 27 was used to determine cations exchange capacity. The Boehm titration method was used to determine the acidity of SBAC 28 . The pH point of zero charge was determined using the procedure described by Leng et al. 29 . Surface functionalities and graphitization or carbon disorder structure of the adsorbent were assessed using FT-IR analysis and Raman spectroscopy, respectively. For the pH point of zero charge, 50 mg of adsorbent were placed in vials with 40ml of 0.01 M NaCl solution and shaken for 48 h, the solution pH ranged from 2 to 12 with an incremental of 2. The concentrations of nitrate and MR were calculated using an IC dionex-120 and a U-vis Shimadzu 1800, the latter at a wavelength of 480 nm, as per Ding et al. 30 .

Results and discussion
Preparation and characteristics of precursors. The precursors were sun dried, and the moisture content of the sample was determined by mass losses after oven drying at 105 °C for 24 h as shown in Table 3. The elemental analysis (Table 4) reveals that there is only a minor difference between the two samples, indicating that digestion has no effect on the organic content of sewage sludge, despite the volatile solid content being slightly different. However, although the surface area of sludge from D was approximately 1.5 m 2 /g, the surface area of sludge S could not be determined, possibly due to a lack of porosity in the precursor materials.
In Figure 1, the TGA curve of dried sludge shows that between 30 and 190 °C, there is a slight mass loss, and the mass of both sludge was roughly 5%, possibly due to adsorbed water. The most mass loss occurred between 200 and 600 °C, with mass loss of D and S increasing from about 5% at 200 °C, to about two-thirds (64%) and half (46%) at 600 °C respectively. The mass loss from 600 to 1000 °C is less pronounced for both sludge with 17 and 13% for D and S respectively. As a result, 600 °C was chosen as the minimum temperature for pyrolysis. S appears to be slightly more stable than D based on the TGA curves.
Both samples D and S show heterogeneous morphology and composition as revealed by microscopic examination by SEM analysis in Fig. 2a-d, and EDS analysis in Figs. 3a, b and 4a, b respectively. The sponge-like morphology shown in Fig. 2b may be attributed to an organic water-repellent compound that adhered to the bubbles during the dissolved air flotation process. Both Figs. 3a, b and 4a, b reveal that the main elements in the sewage www.nature.com/scientificreports/ sludge samples were carbon and oxygen, with silica and aluminium at trace concentrations. In comparison to D sludge, energy dispersive X-ray (EDX) analysis of sample S indicated a high inorganic content, which was most likely created by domestic wastewater particles recovered during decantation or solid-liquid separation. This can be explained by the fact that excess air is released during the dissolved air flotation (DAF) process in the form of microbubbles that adhere to the dispersed phase, causing the particles to float, resulting in the formation of aerophobic compounds that are predominantly inorganic, whereas digestion reduces the amount of organic compounds, most likely due to the production of biogas.  www.nature.com/scientificreports/ Figure 5 shows the FT-IR spectra of the raw sewage sludge with wavenumbers ranging from 450 to 4000 cm −1 . Reading the peaks spectra wavelengths from left to right, the peak observed at 3688-3619 cm −1 is attributed to OH-kaolinite and gibbsite lattice stretching 31,32 , 2988-2901 cm −1 to -C-H group vibration 7,33 , 1631 cm −1 and 1538 cm −1 to sulphur and nitrogen functional groups, respectively 33 . The shape of the shoulder peaks at 1050-1090 cm −1 was attributed to Si-C or Si-O-Si bands (Liang et al. 15 ), C-O-C vibration 33 , and finally, under 1000 cm −1 , the peaks at 749 cm −1 , 535 cm −1 , and 467 cm −1 were attributed to silica or calcium carbonate stretching 32,34 . Discard coal has a lower transmittance from 1498 cm −1 peaks than FT-IR spectra sludge 1007 cm −1 ,   www.nature.com/scientificreports/ 1030 cm −1 , implying that waste coal contains more mineral elements whereas sludge S has lower transmittance than D, implying that the former contains more functional classes.
Characteristics of synthesized activated carbons. Figure 6.a shows the microscopic examination of the sorbent SC-3-600 after activation. Pyrolysis and washing stages may have facilitated cavity formation of highly formed cavities due to the remarkable depletion of inorganic and organic components 35 . Furthermore, some particles ( Fig. 6a, b) lack or have insignificant cavities, which may be attributable to a lack of volatile and decomposed matter escaping to facilitate porosity 9 . The EDS qualitative analysis ( Fig. 6c) of the local particle ( Fig. 6b) revealed inorganic characteristics as well as the presence of K and Cl from activation and washing, respectively. The FT-IR spectrum of the various SBAC (as per Table 4) exhibited almost identical shape and peaks with different intensities regardless of the parameter involved in the synthesis process, as shown in Figs. 7 and 8. The spectra have six main peaks, which are 3676 cm −1 , 2901-2998 cm −1 , 1394 cm −1 , 1225 cm −1 , and 892 cm −1 in all SBAC samples.
In comparison to feedstock spectra, the disappearance of the wavelength peak at 470 cm −−1 and 500 cm −1 in SBAC could be due to inorganic matter solubilisation during the acid washing phase [36][37][38] , while 798 cm −1 could be due to dehydrogenation reactions 24 , 1631 cm −1 and 1538 cm −1 could be due to thermal degradation of protein for nitrogen related compound or sulphur 33 .
The peaks associated with C-H group stretching not only shifted slightly from2859 to 2922 cm −1 in feedstock to a higher value (2901-2988 cm −1 ) in SBAC, which can be related to the presence of saturated group 11 but also transmittance increased after activation for all samples, which is in contrast to some literature 7,33,39 in which it was argued that the disappearance of the peaks was due to the the decomposition of fatty organic matter and dehydration 7 . Organometallic formation may be a possible explanation for the increased transmittance of SBAC pyrolyzed at lower temperatures followed by depletion as temperature rises; for example, the abundance of functional groups (hydroxyl, carboxyl) on the biochar surface synthesized from sewage sludge pyrolyzed at 300 °C reduced extractable cations due to the formation of organometallic compounds 40,41 . The functional groups present in SBAC can be summarized as O-H, C=C, C=O, aliphatic C-H, Si-C, Si-O-Si, phosphate, and carbonate, based on the observations. Com-AC has broad peaks (2 = 25.3° and 2 = 44.6°) linked to its amorphous phases, while the XRD pattern (Fig. 9a, b) demonstrated mineral phases transformation from broad peak in the precursors to sharp in the Figure 6. SEM microscopic analysis of SBAC SC-3-600, (a) multiple particle (b) focus on particle without cavity after activation and (c) qualitative analysis of the targeted particle. www.nature.com/scientificreports/ manufactured absorbent, which clarified transition from amorphous to crystalline phase due to pyrolysis. XRD verified the existence of minerals such as wustite, quarts, illite, and feldspars in SBAC. It is worth noting that the presence of alkaline earth elements in the minerals (feldspars and illite) led to magnesium, calcium, and iron being classified as exchangeable cations.   (Fig. 10), which can be linked to symmetric COO-and nanoaromatic C=O stretching entailed in formation of carboxylic functional groups or C=C bond of the aromatic skeleton ring of adsorbent 13 . On further notice the reduction in -CH (2901-988 cm −1 ) and -OH (3640 cm −1 ) stretching substances which may be ascribed to their solubility/affinity with H 2 SO 4 . On the other side, Com-AC did not have functional group. Raman spectra pattern is shown in Fig. 11, The peak located at 1587cm −1 (G band) and 1357 cm −1 (D band) are associated to sp 2 -bonded carbon and disorder of carbon structure respectively and their intensity I G and I D reveal the adsorbent degree of graphitization and carbon disorder structure from which I D /I G ratio enunciates prevalence of carbon disorder structure over graphitization and vice-versa 13 .
As shown in Table 5, graphitic structure is more predominant than carbon disorder structure and augmented with oxidation, partial graphitization of activated carbon was reported with acidic oxidation treatment with HNO 3 42 . The surface of oxidized adsorbent (Fig. 12b) exhibited less irregularities and soften surfaces than the unoxidized adsorbent (Fig. 12a) and Com-AC (Fig. 12c) probably as results of corrosive H 2 SO 4 -adsorbent surface interaction 18 and also disintegration of pore structure situated at the carbon edge 17 . XRD spectra of adsorbent are represented in Fig. 13 after oxidation treatment the unoxidized sorbent exhibited a broader peak (2θ = 25.3°) associated to amorphous or graphite structure as endorsed by Raman spectra analysis, on the other hand augmentation in peak (2θ = 30°) ascribed to peak might be corollaries of other mineral (feldspar and illite) solubilisation/decline.
The oxidation with APS did not impact significantly carbon composition as per ultimate analysis results since the biggest difference was 6.771% (from DC-5-750: 23,612% to DC-5-750M2:30,383%) while with other sorbents (SC-3-600,SC-5-900,DC-7-900) variation was less than 2%, alike tendency was reported by Ang et al. 22 Furthermore, textural properties in Table 6 receded severely after modification, in case of adsorbent derived from sludge S, for instance prior oxidation SC-3-600 had surface area of 281.72 m 2 /g that shrunk to 68.22 m 2 /g and 46.673 m 2 /g when treated with a solution of 1M and 2M APS respectively, probably due to thinness of walls www.nature.com/scientificreports/ pores, which are prone to collapse 43 and/or micropore occlusion 44 . In the case of DC-5-750, surface area varied slightly (7.28-30.57 m 2 /g: variation) and micropore areas changed from 124.14 to about 157.36 m 2 /g and 176.11 m 2 /g after modification with 1M APS and 2M APS respectively. After modifications DC-7-900 surface area (247.57 m 2 /g) increased probably as results of micropore and mesopore collapse emerging from activated carbon surface etching by reagents, it was 285.59 m 2 /g for DC-7-900M1 and 295,31 m 2 /g for DC-7-900M2.

Adsorption kinetics.
To better understand kinetics order and intra particle diffusion, the effect of time was investigated using the following relationships:  where q e (mg/g), C 0 (mg/l), C e (mg/l), m (mg) and V (ml) represent the adsorption capacity, initial concentration, equilibrium concentration, adsorbent mass and volume of liquid in contact with adsorbent respectively. The removal (%) is calculated via Eq (2). The first and second order kinetics models, as well as interparticle diffusion, are commonly used to understand the adsorption mechanism of pollutants with activated carbon 45 .
Adsorption kinetics can be expressed in terms of the hypothesis that adsorbate removals obey a first-order kinetics:  www.nature.com/scientificreports/ where q t and q e are the amount of pollutants adsorbed per mass of adsorbent (mg/g) at the targeted time and equilibrium, respectively, and k 1 is the constant rate (min −1 ). After integration with conditions that q e = 0 if t = 0, Eq (3) can be written: Or alternatively The adsorption capacity at equilibrium (q e ) and the first-order sorption rate constant (K 1 ) can be evaluated from the slope and the intercept respectively from plot of ln (1 − q t /q e ) vs t.
The second pseudo order kinetics is defined by equation: where K 2 is the second order rate constant, integration of Eq (6) with initial conditions when t = 0 and q e =0, lead to: K 2 and q e can deducted from the slope and the intercept of the plot t/q t vs t, where q t is the adsorption capacity at a specific time. The intraparticle diffusion model is a convenient means to depict diffusion mechanism and examine whether intraparticle diffusion is the rate-limiting step in the adsorption process. The intraparticle model diffusion is represented by: K int and C are determined from the slope and intercept of q t vs t 1/2 .where K int is the intraparticle diffusion rate constant (mg g −1 min −1/2 ) and C is the boundary layer effect intercept; the larger C, the greater the contribution of surface sorption to the rate-controlling step.
The "preferred" oxidized and unoxidized SBAC were chosen based on removal evaluation tests performed at a constant temperature of 303 K for 180 and 360 min with an initial concentration of 50 mg/l of nitrate, 90 min and 180 min for MR with 75 ppm as an initial concentration and an initial pH of 2 for nitrate and 4 for methyl red. The initial pH of 2 in the case of nitrate was chosen based on the hypothesis that adsorbent surfaces become positively charged as pH decreases, preventing competition with hydroxyl anions and pH 4 in the case of methyl red to avoid competition with H + cations while attempting to keep the surface deprotonated (below the pH pzc for the oxidized adsorbent) as MR pKa = 5.1. Initially, unoxidized SBAC outperformed oxidized SBAC as shown in Fig. 14, probably due to the former's proclivity for having a highly deprotonated surface at lower pH. The adsorption capacity decreased as time progressed, with the exception of Com-AC, where it changed slightly from 16.32 to 13.2 mg/g for contact times of 180 and 360 min, respectively. Furthermore, SBAC (DC-5-750M1 t q e (8) q t = K int t 1/2 + C Table 6. Structural properties and ultimate analysis results of SBAC after and before oxidation.

Modification conditions
S BET (m 2 /g) S meso (m 2 /g) S micro (m 2 /g) V tot (cm 3 /g) V micro (cm 3 /g) www.nature.com/scientificreports/ and SC-5-900M2) released a significant amount of nitrate ion in solution after 360 min of agitation, while this amount was lower when the process was carried out with other adsorbents. This situation may be related to the release of compound contained in the adsorbent ash in the liquid phase and/or the adsorption equilibrium phenomenon, in which optimum equilibrium contact was reached. DC-5-750 and SC-3-600, as well as the oxidized sorbents DC-5-750M2 and SC-3-600M2, were chosen for further nitrate adsorption experiments. The adsorbent had mesopores, making it ideal for removing medium-sized substances from liquid process 1 . Despite having a larger surface area, SC-5-900 (422.09 m 2 /g) and SC-5-900M2 (313.05 m 2 /g) had a lower adsorption capacity than other SBAC. This may be attributed to lower carbon content (Table 1) or the disappearance of acidic functional groups as temperatures rose 36 , and this result confirmed the significance of functional group presence. Based on the test results in Fig. 6, DC-5-750 M1 and DC-7-900 M1, with adsorption capacities of 127.634 mg/g and 124.376 mg/g, respectively, were chosen for additional experiments in relation to Com-AC (121.10 mg/g) performances, as well as unoxidized SBAC (DC-5-750 and DC-7-900). Figure 7a depicts the nitrate removal pattern over time. Prior to 30 min of contact time, all adsorbents had negligible adsorption power, most likely due to film diffusion resistance (external diffusion) 46 , but it increased significantly up to 120 min for Com-AC, DC-5-750, and DC-5-750M1 at 15.132 mg/g, 17.46 mg/g, and 10.72 mg/g, respectively.

Pore size (nm) C (%) H (%)
As illustrated in Fig. 15, adsorption rises with time except for Com-AC, where the change was insignificant (from 123. 8 mg/g at 90 min to 121. 2 mg/g at 180 min), owing to the fact that equilibrium had already been established. Despite having the highest surface area, SC-5-900 and SC-5-900M2 had a lower q e than other sorbents, most likely due to a lower carbon content (Table 6) or a lack of acidic functional groups due to their depletion during temperature augmentation 14 . This finding emphasizes the critical nature of functional group presence.
As depicted in Figure 15, the synthesised sorbents with the highest adsorption potential were DC-5-750 M1 (127.6 mg/g) and DC-7-900 M1 (124.4 mg/g), and they were therefore chosen for future experiments, along with the unmodified sorbents (DC-5-750 and DC-7-900).The upward tendency may be associated to pore diffusion and surface reaction which are deemed to present less component resistance than external diffusion 46 , while stagnant trend after 120 min may be ascribed to occupation of available adsorption site by nitrate ions as the process progress 25,47,48 , which weaken the interaction between sorbate and adsorbent surface 48 .
Preliminary tests shown in Fig. 6 revealed that the adsorption capacity increased significantly with time with a little fluctuation in Com-AC from 123.8 mg/g at 90 min to 121.172 mg/g at 180 min, implying that equilibrium had already been reached.
Although having a higher surface area, the adsorption capacity of SC-5-900 (422.1 m 2 /g) and SC-5-900M2 (313.1 m 2 /g) was lower than other SBAC. This could be due to lower carbon content (Table 5) and/or depletion of acidic functional groups as they vanished with temperature rise (28). This observation further corroborated the importance of functional group presence. www.nature.com/scientificreports/ Based on results in Fig. 6, DC-5-750 M1 and DC-7-900 M1 with adsorption capacities of 127.6 mg/g and 124.4 mg/g, respectively, were chosen for additional investigations in comparison to Com-Ac (121.1 mg/g) performances. Figure 16a depicts the nitrate removal trend as a function of time. Initially, all adsorbent adsorption capacities were negligible prior to 30 min of contact time possibly due to film diffusion resistance (external diffusion), and thereafter rose significantly for Com-AC (15.1 mg/g), DC-5-750 (17.5 mg/g), and DC-5-750M1 (10.72 mg/g). Figure 16.b depicts the changes in dye adsorption capacity with contact time. Com-Ac adsorption capacities increased slightly from 109.63 mg/g at 30 min to 123.868 mg/g at 90 min and 121.107 mg/g at 180 min, while DC-5-750 and DC-5-750M1 adsorption capacities increased from 83.08 mg/g and 113 mg/g at 30 min to 113.126 mg/g and 123.8 mg/g at 90 min and 114.36 mg/g and 127mg/g at 180 min.
The rapid increase in adsorption capability at the start of the process could be attributed to the abundance of adsorption sites 26,49 .
Nitrate pseudo first order (PFO) and Pseudo second order (PSO) plots are recorded in Fig. 17a, b and the corresponding parameters are recorded in Table 6. From the results in Fig. 17a, b Com-AC fitted better the PFO (R 2 = 0.9693) than PSO (R 2 = 0.8569) kinetic model, while other sorbents fitted better PSO kinetic model albeit lower coefficient correlation (R 2 ) 0.7859 and 0.793 for SC-3-600 and SC-3-600 M1 respectively. Figure 18 depicts intraparticle diffusion model line plots for nitrate adsorption that did not pass through the axis origin, indicating that the nitrate removal mechanism was not solely controlled by the intraparticle diffusion model. Observations of similar nitrate removal were recorded by others 25,50 .
Nitrate PFO and PSO parameters are shown in Table 7. It was noticed that Com-AC fitted better PFO (R 2 = 0.9693) than PSO (R 2 = 0.8569) kinetic model, whereas other sorbents fit better PSO kinetic model, albeit with lower coefficient correlation (R 2 ) 0.7859 and 0.793 for SC-3-600 and SC-3-600 M1, respectively. The plot of MR PFO and PSO kinetic models are depicted in Fig. 19a, b. .The MR adsorption process could not be depicted using the PFO kinetics model due to insufficient linearity fitting induced by the inherent equation formula where q e (adsorption capacity) is concomitantly fitting data and determining value. The parameters for parameters value of MR PFO and PSO kinetic and intraparticle model are presented in Table 8.  www.nature.com/scientificreports/ pH dependence. The effect of pH was measured by varying the pH solution from 2 to 10, as shown in Fig. 20. The process was pH dependable, as evidenced by adsorption decrease with pH increasing, in the case of DC-5-750 (pH pzc = 6.6) from 20.56 mg/g at pH 2, to 6.24 mg/g at pH 6 and 4.2 mg/g at pH 10, Com-AC (pH pzc = 10.3) and DC-5-750M2 (pH pzc = 3.1). At pH 2, adsorption potential was 16.32 mg/g and 12.24 mg/g, respectively, at pH 6, 11.12 mg/g and 9.6 mg/g, respectively, and at pH 10.6 mg/g and 1.8 mg/g. This pattern was more likely caused by: (1) favorable conditions of nitrate removal accentuated by electrostatic attraction as adsorbent surface bears positive charge at lower pH. (2) The presence of competition between nitrate ions and hydroxyl ions in basic solution, as also stated in other works using carbon-based activated carbon content 47,48,51 .
In addition to the above-mentioned justifications, it may be further hypothesized that the underperformance of oxidized SBAC is due to the introduction of an acidic surface functional group; in the case of deprotonation, if pH > pHpzc, more binding sites for cationic sorbate are created on the surface than if it was an unoxidized adsorbent 44,52 .
Taking into account that MR is negative if pH>pKa and positive 26 , the introduction of acidic functional groups caused a shift in pH pzc , from neutral 6.6 (DC-5-750) to acidic after oxidation 3.1 (DC-5-750-M1) and Com-AC was basic 10.2, surface functional group deprotonated when pH pzc < pH and adsorbents surface becomes negatively charged 53 . In contrast to SBAC, Com-AC (pH pzc = 10.2) has a wider range where the surface's net  www.nature.com/scientificreports/ charge is positive. Adsorption of MR with changed SBAC, which had a lower pH pzc than unoxidized, was more pH dependable due to electrostatic attraction between the adsorbent negatively charge and positive MR below pH4 26 ; at pH4, the adsorption potential of DC5-750M1 and DC-7-900M1 was 127.634mg/g and 117.176 mg/g, respectively, from pH 6, pH 8, and pH 10. Adsorption capacity decreased at pH6 from 97.488mg/g for DC-5-750M1 and 109.278 for DC-7-900M1 to 81.854 mg/g for DC-5-750M1 and 71.028 mg/g for DC-7-900M1.  Figure 19. MR PFO (a) and PSO (b) kinetic model. www.nature.com/scientificreports/ The results show that for the oxidized adsorbent at pH4 the adsorption mechanism was related to electrostatic attraction and hydrophobic associated to/or hydrogen bond 54 . Similarly, the adsorption potential of unoxidized SBAC (DC-5-750 and DC-7-900) decreased from 123.868 mg/g and 111.76 mg/g at pH 2 to 81.854 mg/g and 90.32 mg/g at pH 10. The outperformance in basic solution could be attributed to adsorption site rivalry between hydroxyl ion and MR negatively charged ions 49 . Similarly, in an acidic solution with a pH of 2, the rivalry may have been between H + and positive MR 55 or/and repulsion force between protonated adsorbent surface and MR 26 .
However, pH solution variation affected slightly the Com-AC adsorption capacity of MR, from 101.894 mg/g at pH 2 to 113.04 mg/g at pH 10, possibly because the electrostatic attraction mechanism was not very pronounced in the process because below pH 5.1 dye was charged positively and the protonated adsorbent had positive net charge, similar results were recorded on adsorption of cationic dye (Methyl blue) 56 .
The adsorption capacity increased as initial concentration increased because driving force of concentration gradient prevailed and had propensity to subjugate mass transfer resistance barrier between solid and liquid interface. Conversely, the proportion of nitrate extracted decreased due to adsorbent site saturation, so a fraction of sorbate remained in solution 14 . As the initial pollutant concentration increased, the adsorbent's dye adsorption capacity increased, but the proportion of dye removed decreased due to a stronger driving force to overcome mass transfer resistance in terms of adsorption capacity and less available adsorption 55,57 For example, the adsorption capacity of Com-AC, DC-5-750, DC-5-750M2, and SC-3-600 at the lowest initial concentration (10 mg/l) was 5.78 mg/g, 8.50 mg/g, 6.28 mg/g, and 8.51 mg/g, respectively; at 50 mg/l it was 16.32 mg/g, 20.56 mg/g, 12.24 mg/g, and 18.32 mg/g, respectively; and at the highest initial concentration (90 mg/l 13 mg/g, 23.98 mg/g, 14.65 mg/g, and 18.778 mg/g, while the proportion of nitrate extracted displayed the opposite pattern. At the lowest initial concentration (10 mg/l), it was 27.78%, 40.87%, 30.19%, and 40.91%, and then decreased at the initial concentration of 50 mg/l to 16.32%, 20.56%, 12.4%, and 18.32%. Figures 21a, b and 22 a, b display Langmuir and Freundlich isotherms fitted to elucidate the nitrate and MR removal processes, respectively. Tables 9 and 10 provide data on the fitting parameters. For all adsorbents, the Langmuir isotherm suit the process better with a higher R 2 than the Freundlich model. The presence of R L values between 0 and 1 suggested that nitrate and MR adsorption were favorable on all sorbent surfaces. It is worth noting, however, that as the R L value reached zero, MR adsorption became more irreversible with increasing concentration 12,35 .

Isotherms.
Nitrate Langmuir and Freundlich isotherms were used to investigate the nitrate removal process. The findings are presented in Table 11.
As predicted, the Langmuir isotherm described the process better than the Freundlich model, with a greater R 2 for all adsorbents involved on a monolayer surface. The value of R L between 0 and 1 indicated that both sorbate adsorption was favourable on all sorbent surfaces of the Langmuir isotherm model, it is regarded as unfavourable if R L > 1 36 , for Freundlich model adsorption intensity (1/n) value was less than 0.5 in all cases, indicating that sorbate was easily adsorbed, it is hardly adsorbed if 1/n > 2 42 .
It is worth noting that MR adsorption became more irreversible with increasing concentration since the RL value approached zero at higher concentration 25,26 . The published data are also compared to the present study's nitrate adsorption capacities in Table 12.

Ionic strength.
To assess the effect of ionic strength, MR was diluted in NaCl solution with a concentration of 0.01 M and 0,05M and the pH was adjusted at 4. As shown in Fig. 23, presumably Na + cations competition and  Environmental consideration: toxicity characteristic leaching procedure. The toxicity contaminant leaching procedure was carried out as described elsewhere 39,62 , and the element concentrations were determined using a Perkin-Elmer AA spectrometer. Table 13 shows the results of the TCLP test; in general, the concentration of leachable heavy metal in the pyrolyzed adsorbent was lower than the precursor due to the higher thermal stability of heavy metal acquired through pyrolysis 62 .  www.nature.com/scientificreports/ However, after pyrolysis, SC-5-900 released more heavy metals (Fe, Cr, Co, and Ni) than its precursors. This may be due to the disintegration of some stable inorganic minerals (primarily silicate and carbonate) from sludge during pyrolysis with temperature augmentation, which caused the liberation of the fixed metals from the lattice 17 .

Conclusion
Two kinds of sewage sludge were used in this paper to create low-cost adsorbents: D, which was collected during the dissolved air flotation stage, and S, a combination of primary and secondary sludge from the digestion and dewatering phases. After mixing the sewage sludge with waste coal, it was activated with KOH and oxidized with APS. We evaluated and compared the capacity of the synthesized adsorbents to remove nitrate and MR to that of commercial activated carbon. The oxidation with APS (1) had a greater negative effect on the textural properties of adsorbent derived from sludge S than on those derived from sludge D, (2) had a negligible effect on the organic composition as revealed by ultimate analysis, and (3) induced the introduction of acidic functional groups as revealed by FT-IR and Raman spectroscopy, respectively. The adsorption capacity of adsorbents increased with time and initial contamination concentrations. The nitrate and MR removal processes followed  www.nature.com/scientificreports/ the pseudo-second order kinetic model and the Langmuir isotherm rather well. Nitrate and MR adsorption capabilities were greater at pH = 2 and pH = 4, respectively.

Data availability
Some of the data will be made available upon the journal acceptance.