Chloramphenicol and methylene blue adsorption by modestly treated paper sewage sludge-based activated carbon

Refractory pollutants like pharmaceuticals and dyes have become excessively prevalent in most Malaysian water bodies because of the growing textile and pharmaceutical industries. Hence, this work employed activated carbon prepared from freely available paper mill sewage sludge for removing chloramphenicol (CAP) and methylene blue (MB). Modest treatment of low-temperature carbonisation assisted with a short activation time of microwave radiation had been used. Analysis of variance of central composite design resulted in the optimum conditions of 440-W radiation power and 3-min activation time for optimum removal of 70% CAP and 51% MB. The surface area of the paper mill sewage sludge activated carbon (PMSSAC) improved greatly from 1.14 to 412 m2/g, with the highest adsorption capacity of 13 mg/g. The scanning electron microscope images demonstrated the efficiency of microwave radiation treatment, where more cavities and pores were observed on activated carbon for improved adsorbate penetration. The Freundlich isotherm and the pseudo-second-order model appeared to best fit the kinetic data. Furthermore, the high affinity of adsorbate towards the PMSSAC surface could be the plausible mechanism, as indicated by the high amount of adsorption within the initial stage of adsorption. Thus, it is envisaged that our PMSSAC could be effectively employed in actual wastewater systems, as evidenced by excellent CAP and MB removal.


Introduction
Adsorption as a mean for wastewater treatment has received vast attentions owing to its low-cost preparation and operation with high performance in removing various types of contaminants.Among other chemical and physical wastewater treatment methods, such as membrane filtration, coagulation, and sedimentation, adsorption which employs physicochemical effects to adsorb pollutants appears to be highly practical to endure stricter environmental regulations for final wastewater effluent.Thus, it is crucial to continuously explore the adsorption process and its operating elements in effectively treating countless hazardous contaminants present in the wastewater system.
Activated carbon (AC) as a main tool in adsorption has always been a central element in many reported studies.This material has been undeniably approved as an excellent adsorbent for removing various harmful contaminants including methylene blue (MB) (Somsesta et al. 2020;Han et al. 2020;Lu et al. 2022), polyaromatic hydrocarbons and phenolic compounds (Sullivan et al. 2019;Lϋtke et al. 2019), pharmaceuticals (Delgado et al. 2019;Varga et al. 2019), and other organic wastes (Piai et al. 2019;Anjum et al. 2019;Jun et al. 2021).Activated carbon, which is also known as activated charcoal, is produced from the carbonisation of a waste biomass precursor, after which the resulting char is gasified in either steam or carbon dioxide (CO 2 ).Carbonisation and gasification are essential steps in obtaining AC with high surface area and adjustable porosity for optimum removal of those recalcitrant contaminants.
Several studies demonstrated that AC prepared at lower carbonisation temperatures (200-500 °C) appeared effective in adsorption (Teimouri et al. 2019;Taher et al. 2023;Chen et al. 2013).However, optimal contaminant adsorption could only be accomplished by combining this low-temperature requirement with additional activation techniques, such as acidic activation, or with prolonged heating periods and higher heating rates (Zhao et al. 2023;Vinayagam et al. 2023;Ge et al. 2023).Besides, it has been proven that carbonisation assisted by moderate microwave heating could successfully increase the surface area and pore diameter of AC (Baytar et al. 2018).Microwave heating could efficiently modify the properties of AC in removing oxygenated functionalities from carbon surface for better adsorption as compared to using a conventional muffle furnace, which consumes high energy but produces less yield of AC (Yang et al. 2017).Activated carbon prepared from phosphoric acid activation assisted with microwave heating demonstrated significant adsorption capacity for dyes and organic matter (Dolas 2023;Abderrahim et al. 2023).In other words, microwave treatment could improve the conditions for AC due to its rapid heating and brief impregnation, which help to reduce the high-temperature requirement during carbonisation.
As a result of various surface-structure modifications, AC has been better for removing a wide range of refractory compounds, including pharmaceutical contaminants.The contaminants often easily bypass the wastewater treatment process and contaminate water reservoirs due to their polarity and/or chemical persistence (Varga et al. 2019).Antibiotics are among the frequently detected pharmaceutical compounds even in surface water, which include ampicillin, chloramphenicol, quinolone, metronidazole, etc. Praveena et al. (2019) detected pharmaceutical residues (e.g.caffeine and diclofenac) with concentrations ranging from 0.14 to 0.38 ng/L in Malaysian water bodies alone.Thus, a great concern to remove these harmful antibiotics from the wastewater has led to numerous studies being conducted, especially by using AC adsorption (Teixeira et al. 2019;Danalıoğlu et al. 2017;Nazari et al. 2016;Xiang et al. 2019).This is because the adsorption of pharmaceutical products by carbon-based adsorbent has shown better performance as it does not generate toxins or pharmacologically active products while operating (Katsigiannis et al. 2015).
Another critical pollutant with a similar adverse effect on the environment is the colouring dye.The ominous presence of dyes in the wastewater system is highly contributed by the increasing trend of textile industrial activities in this country.Despite their good income to the Malaysian economy, the environment would innocuously bear the consequences.According to Mohamed Sunar et al. (2019), the volume of effluent from the textile industry is in the range of 2000 m 3 against domestic wastewater, which is only 0.22 m 3 per capita per day.This is even more disastrous when the dyes which are generally found in textile wastewater are difficult to decolourise due to their complex structure, stability to light, heat and oxidising agents; thus, these dyes are becoming highly insusceptible to normal biodegradation.Similar to pharmaceutical pollutants, adsorption by AC looks like a highly favoured treatment method owing to its strong intact and interfacial adhesion for optimum removal of these recalcitrant dye contaminants.
Apart from wastewater pollution problems that need extra attention, the abundance of solid biomass and sewage sludge produced by various agricultural sectors, including palm oil, paper, etc., has seriously impacted our ability in sludge management.Other than to be applied on land as compost, sludge waste that is continuously produced needs to be properly treated and/or alternatively transformed into beneficial products, such as AC for wastewater treatment.In this study, AC was prepared from paper mill sewage sludge by microwave heating to adsorb chloramphenicol (CAP) and MB.This microwave-assisted technique requires low-temperature carbonisation without the use of an acid agent in order to provide a modest treatment for AC.Detailed characterisation of paper mill sewage sludge-based AC with complete optimisation of operating conditions, kinetic, and isotherm studies was performed.

Paper mill sewage sludge activated carbon preparation
Paper mill sewage sludge (PMSS) was used as a precursor for the preparation of AC.The sludge was obtained from Muda Paper Mill Sdn.Bhd., Seberang Perai Selatan, Penang, Malaysia.It was dried in an oven at 100 °C for 24 h.Then, the sample was crushed mechanically into smaller sizes and sieved to obtain a particle size of 1-2 mm to facilitate the pre-treatment process, which would increase the rate of drying of the sample.Carbonisation of PMSS was performed by heating the sample (500 g) in the furnace at 250 °C.The produced char was then placed in a microwave oven under CO 2 at a specific radiation power and activation period.The sample was subsequently stored for further characterisation and adsorption study after being cooled to room temperature under nitrogen flow.

Adsorbate preparation
Two adsorbates were selected in this study, namely, CAP and MB for the adsorption experiment.The adsorbate sample was collected using a disposable syringe every 1 h over 24 h for batch adsorption study.A double-beam ultraviolet-visible spectrophotometer was used to measure adsorbate concentrations.Different wavelengths of 665 and 278 nm were used for CAP and MB, respectively.The linear relationships between the adsorbate concentrations adsorbed onto the AC were plotted against time.

Characterisation of paper mill sewage sludge activated carbon
The prepared paper mill sewage sludge activated carbon (PMSSAC) was characterised using a Micromeritics ASAP 2020 volumetric adsorption analyser for its pore volume and average pore diameter.The surface area of PMSSAC was determined using Brunauer-Emmett-Teller (BET) method.The total pore volume was determined based on the relative pressure of the liquid volume of nitrogen at 0.98.The surface morphology of the samples was characterised using a scanning electron microscope (SEM) (Quanta 450 FEG, Netherlands).The proximate analysis was carried out using a simultaneous thermal analyser (Perkin Elmer STA 6000, USA).The elemental composition of the samples was determined using an elemental analyser (Model Perkin Elmer Series II 2400, USA).The sample was sealed inside a tin capsule and placed inside the combustion chamber to undergo pyrolysis at 975 °C.Lastly, the functional groups and bonds of the sample's compounds were determined by a Fourier transform infrared spectrometer (IR Prestige 21 Shimadzu, Japan).

Experimental design for PMSSAC preparation
The variables for the preparation of the PMSSAC experiment were proposed by the central composite design (CCD) of response surface methodology.The two independent variables used were X 1 (radiation power, W) and X 2 (activation time, min).Three variable levels of 264, 440, and 616 W and 2, 3, and 4 min were used for both X 1 and X 2 , respectively.A total of 13 experimental runs were proposed with varied values of X 1 and X 2 (four factorial points, four axial points, and five centre points).Two responses (i.e.CAP removal (%) and MB removal (%)) were used to develop an empirical model, as shown in Eq. 1, where Y is the predicted response; b 0 and b i are constant and linear coefficients, respectively; b ij is a quadratic coefficient; and x i and x j are the coded values of variables.

Batch equilibrium, isotherm, kinetic, and mechanism studies
The batch equilibrium experiments were carried out for the adsorption of CAP and MB onto PMSSAC.First, 0.20 g PMSSAC was put into the test tubes filled with 200 mL of CAP and MB solutions with different initial concentrations ranging from 25 to 300 mg/L.The test tubes were then sealed and placed in an isothermal water bath shaker at 30 °C with an agitation speed of 150 rpm.For the equilibrium study, the amount of adsorption at equilibrium, q e, was calculated using Eq. 2, where C o is the liquid-phase adsorbate concentration at the initial stage (mg/L), C e is the liquid-phase concentration of adsorbate at the equilibrium stage (mg/L), V is the volume of adsorbate solution (L), and W is the mass of adsorbent used (g).The percentage of adsorbate removal, %, was calculated using Eq. 3, where C t is the liquid-phase adsorbate concentration at time t.
The experimental data were fitted using three isotherm models, namely Langmuir, Freundlich, and Temkin.The coefficient of correlation coefficient value, R 2 , which is the closest to unity, can be used to identify the isotherm that fits data the best.The three isotherms' equations of Langmuir, Freundlich, and Temkin are shown in Eqs. 4, 5, and 6, respectively.The favourability of the nature of adsorption can be described by a separation factor, R L , as shown in Eq. 7.
where q m is the maximum adsorption capacity (mg/g), K L is the Langmuir adsorption constant (L/mg), n F is the Freundlich isotherm constant related to adsorption intensity, K F is the Freundlich isotherm constant related to capacity, R is the universal gas constant (8.314J/mol.K), T is the absolute temperature (K), A T is the Temkin equilibrium binding constant (L/mg), B T = RT b is the Temkin constant related to the heat of adsorption (L/mg), b is the Langmuir constant, and C i is the initial concentration.
For the kinetic study, the amount of adsorption at time t, q t (mg/g) was determined using Eq. 8, where C t is the liquid-phase concentration of adsorbate at time t (mg/L).Pseudo-first-order and pseudo-second-order models (Eqs.9 and 10, respectively) were used in this study, where k 1 is the rate constant of first-order sorption (h −1 ) and k 2 is the rate constant of pseudo-second-order sorption (g/mg•min).

Analysis of variance for CAP and MB removal
The data obtained for the 13 experimental runs were analysed by analysis of variance (ANOVA).The model F-value of 21.47 and 62.49 for CAP and MB dye, respectively, implied that both models were significant (Prob > F data supplied in Supplementary File).In this case, radiation power and activation time in terms of X 1 and X 1 2 for CAP and X 1 , X 2 , X 1 •X 2 , and X 1 2 for MB were significant model terms.Based on the model equations, it could be deduced that most of the significant terms present a positive impact on the removal of both adsorbates.Thus, PMSSAC could offer high performance, especially when X 1 , X 2 , and X 1 •X 2 are varied from low to high levels (Ab Ghani et al. 2017).
Figure 1a and b shows the three-dimensional (3D) response surface of the preparation parameters for CAP and MB removal by PMSSAC, respectively.The removal for both adsorbates started to decrease when the radiation power increased above 440 W. Longer activation time of 4 min with a smaller range would be preferred for MB removal as compared to CAP removal, which is deemed more suitable at a larger range of activation time from 2.5 to 4.0 min.Prolonged activation time might be explained by the volatilisation of surface carbon atoms, which became predominant, increasing adsorbent weight loss and causing the formation of new pores and pore widening (Ab Ghani et al. 2017).Hence, this phenomenon offers greater availability of adsorbent surface area for higher adsorbate removal.

Optimisation of operating parameters
Based on the 3-D plots, 440-W radiation power and 3-min activation time were chosen as the optimum conditions for the validation test.As shown in Table 1, the percentage differences between the predicted and experimental results were 24.36 and 46.74% for CAP and MB removal, respectively.The optimum conditions for CAP adsorption seemed more fit and sufficient to predict the model compared to MB adsorption with a quite high percentage difference.

Surface area and pore characterisation
Table 2 provides a summary of the surface area and pore characterisation for the PMSS and PMSSAC samples, including the BET surface area, mesopore surface area, total pore volume, and average pore diameter.PMSSAC had a significant surface area of 412 m 2 /g for optimum adsorption of CAP and MB.The tremendous increase for all the characteristics tested in PMSSAC was the evidence of the impact of the activation process, which include the physical activation (CO 2 ) under microwave irradiation.The CO 2 gasification promoted the formation of mesopores and enhanced the surface area of AC (Wu and Tseng 2006).The pore structure of AC could also be altered by the increase in activation temperature, which would modify the shape of the pore size distribution (PSD) (Sethia and Sayari 2016;Dolas et al. 2011).The PSD in particular amplified and shifted towards larger pores as the activation temperature increased, which in turn simultaneously increased the surface area and average pore diameter of the AC.Interestingly, the AC prepared from black wattle bark waste achieved a similar surface area (414 m 2 /g), but required a higher carbonisation temperature of 700 °C and strict ZnCl 2 chemical activation (Lütke et al. 2019).A recent study also found that increasing the temperature and microwave irradiation power (400-800 W) resulted in increased pore diameter and oxygenic functional groups (Dongyang et al. 2023), which eventually enhanced adsorbate adsorption.Therefore, our significantly lower carbonisation temperature (250 °C) with simple microwave activation appeared to be a more sustainable method for producing good AC.

Proximate analysis
Through proximate analysis, it was found that PMSS had a suitable amount of fixed carbon content, making it a good precursor for AC production.From Table 3, the moisture and volatile content percentage decreased significantly from 27.2% to 10.5% and 45.8% to 12.1% for PMSS and PMSSAC, respectively.At high temperatures, the organic substances could not maintain their stability as the heat forced the molecules to break their bond during activation.The organic substances were released as liquid and gas, forming high-purity carbon materials for the development of highly porous AC (Ahmad and Alrozi 2011).With high carbon content of 74% and significantly low ash content, PMSSAC is a highly suitable AC for CAP and MB adsorption, as explained in the following subsections.

Surface morphology
The morphologies of PMSS and PMSSAC were characterised using SEM micrographs, as shown in Fig. 2. It was apparent that PMSS composed of a compacted surface with insufficient pore opening.However, after activation, PMSSAC showed a series of cavities with a quite uniform distribution of pore structures and openings.The presence of cavities on the PMSSAC structure was the evidence for the higher BET surface area and pore volume, which provided better penetration of adsorbate into the AC (Soltani et al. 2015).Furthermore, the activation process by microwave radiation could promote the development of the porous structure of PMSSAC.

Batch adsorption study
The CAP and MB adsorption behaviour onto PMSSAC can be analysed by performing batch adsorption studies based on their adsorption equilibrium, isotherm, kinetics, and mechanism.

Equilibrium study
In this study, the initial concentration of adsorbates (50-300 mg/L) was applied to PMSSAC over 24 h.The results of the percentage removal and adsorption uptake versus time are plotted in Figs. 3 and 4, respectively.Generally, for both CAP and MB adsorption, the removal and uptake rate increased steeply at the initial adsorption time, reaching an optimum level at 4 h.Then, as the adsorption time was prolonged, the adsorption rate for all profiles remained stable for the next 24 h.A huge number of vacant/free functional groups on the PMSSAC surface might cause high initial CAP and MB removal rates.Then, steady occupation and saturation of AC active sites by adsorbate molecules resulted in a slight decrease until the equilibrium was reached (Vakili et al. 2020).The percentage removal of MB was relatively low at only 51% as compared to CAP with 70%, which could be explained by MB's lower anionic tendency and affinity towards PMSSAC active surface.Furthermore, the percentage removal of both CAP and MB decreased with increasing initial concentrations.The reduced adsorbate concentration improved CAP and MB removal by increasing the driving force for mass transfer between the liquid and solid phases.An opposite trend was observed for adsorption uptake plots, where the adsorption capacity, q t, increased with increasing initial concentrations for both CAP and MB, as depicted in Fig. 4.This trend agreed with Somsesta et al. (2020), where dye adsorption was enhanced with the increase of dye concentration by cellulose composite AC.The increase of adsorbate molecules improved the possibility of contact between the adsorbate molecules and AC, accelerating the adsorption onto the film surface (Deveci and Kar 2013).

Isotherm and kinetic studies
The adsorption isotherm study was performed to describe the progress of the adsorption mechanism between adsorbate solution and adsorbent at equilibrium conditions.The data obtained for CAP and MB removal at optimised radiation power and activation time were analysed using three isotherms: Langmuir, Freundlich, and Temkin.The data of the three isotherms are presented in Supplementary File.
The R 2 values the adequate fit of the models to the experimental data for CAP and MB adsorption were near 1 for all three isotherms.This testified that CAP and MB adsorption using PMSSAC was highly favourable.The sequence of the models for CAP adsorption followed Freundlich > Langmuir > Temkin, while MB adsorption followed the sequence of Freundlich > Temkin > Langmuir.Generally, both Freundlich and Langmuir isotherms fit CAP adsorption better as the R 2 values approach 1.0.Only Langmuir exhibited less favourable effects for MB adsorption with the lowest R 2 value.The better consistency by the Freundlich model indicated that the adsorbate's adsorption sites were heterogeneous, and the relationship between the solute concentration on the PMSSAC and the concentration of the solute in the bulk liquid was empirically developed by multi-molecular layers.In terms of K F value, CAP appeared to be slightly lower than MB, which could signify that PMSSAC has little influence on equilibrium, which is corroborated by the uptake of adsorption (Branko et al. 2022).
The constant n F designates the adsorption intensity in the Freundlich model.In this study, its values fall below 1.0 (n F < 1.0).This indicated that CAP and MB adsorption was best described by the chemisorption process rather than the physical process (Misran et al. 2022).In addition, CAP adsorption was also better described by the Temkin model with higher B T and A T values, suggesting that adsorption might occur through an equal combination of physical and chemical sorption.
The significant characteristic of the Langmuir isotherm could also be used to predict the affinity between the adsorbate and AC using a dimensionless constant called a separation factor or equilibrium parameter, R L (Malik 2004).The favourability of the adsorption process can be classified as unfavourable (R L > 1), linear (R L = 1), or favourable (0 < R L < 1).Based on Fig. 5, the R L values fall between 0.1 and 0.99, confirming that the adsorption of CAP and MB was favourable at the proposed optimum conditions.The R L values decreased gradually as the initial concentration of adsorbates increased, reaching an almost steady R L towards higher concentrations (Ajenifuja et al. 2017).This may imply that the adsorption efficiency was mostly favourable at higher concentrations for both CAP and MB.Additionally, this trend is consistent with the adsorption capacity profile, where optimum adsorption capacity was achieved at higher concentrations of adsorbates.The higher R L near 1 for MB could be explained by its low favourability, and this was consistent with the Langmuir R 2 value of only 0.785, the lowest data fit among other models.
The adsorption kinetics are very important to provide information about the mechanism of adsorption.Pseudofirst-order and pseudo-second-order kinetic models were employed (kinetic data are provided in Supplementary File).Among all initial concentrations tested, the pseudo-secondorder model showed a better fit to the kinetic data for CAP and MB adsorption with higher R 2 values.Furthermore, the theoretical (q e,calc ) values were relatively closer to the experimental (q e,exp ) values by the pseudo-second-order model compared to that of the pseudo-first-order model.This suggested that the pseudo-second-order model offered better adsorption correlation for CAP and MB removal with respect to their calculated rate constant, k 2 .This result was in good agreement with Kuang et al. (2020), where the dynamic of MB adsorption using surfactant-modified AC better fitted the pseudo-second-order model.
The kinetic mechanism of CAP and MB adsorption can be further analysed by using the intraparticle diffusion model.The acquired data can be displayed in an intraparticle diffusion plot with multiple linear profiles at various initial concentrations, as shown in Fig. 6.It is good to observe the virtual analogous trend of the adsorption mechanism demonstrated by both adsorbates.The first region of the graph shows a sharp increase of q t up to 1.8 t 0.5 , especially for higher initial concentrations.The q t then achieved a steady condition where there was no further increase of q t with increasing t 0.5 .The electrostatic interaction between the adsorbent and adsorbate, which led to a high affinity of adsorbate adsorbed on the PMSSAC surface, may be responsible for the high q t at the initial t 0.5 .By considering the mathematical relationship given by Eq. 13, with k p being the rate constant in mg/g/h 0.5 , three phases in the adsorption mechanism were previously outlined (Delgado et al. 2019), which were mainly (a) adsorbate transfer to the adsorbent particle's outer film, (b) movement of adsorbate solutes towards the active sites, and (c) adsorption of the adsorbates on the active sites.It is hypothesised that the present study rather corresponded to the three phases at which in the whole process, the adsorbates moved at the aforementioned rates from the film towards the active sites before reaching equilibrium at the designated t 0.5 .

Conclusion
The sewage sludge from paper mills was effectively used as a precursor for the synthesis of AC using low-temperature carbonisation and a quick microwave activation period.The ANOVA of CCD showed that the optimum conditions of 440-W radiation power and 3-min activation time resulted in the highest percentage removal of 69.48% and 50.51% for CAP and MB, respectively.The BET surface area, mesopore surface area, total pore volume, and fixed carbon content for PMSSAC increased significantly to 412 m 2 /g, 235.0 m 2 /g, 0.241 cm 3 /g, and 72.6%, respectively.The highly porous structure of PMSSAC with cavities was observed through SEM, leading to better penetration of CAP and MB onto the AC.The Freundlich isotherm appeared to best describe both CAP and MB adsorption, showing a heterogeneous relationship between the multi-layer adsorbate solutes and the solid surface of AC.The separation factor, R L, revealed that the adsorption capacity was highly favoured at higher concentrations of CAP and MB.The kinetic data for both adsorbates fit well with the pseudo-second-order model.A high amount of adsorption q t was found for both adsorbates, suggesting a good adsorption mechanism by high affinity adsorbates towards the active sites of PMSSAC.Our PMSSAC system had proved similar and effective performance to adsorb CAP and MB.This is an important finding, where our adsorption system could be used for the removal of multiple recalcitrant contaminants and further improved for application in the real industrial wastewater system.

Fig. 1
Fig. 1 Three-dimensional response plots for a CAP and b MB removal by PMSSAC

Fig. 3 Fig. 4
Fig. 3 Percentage removal against adsorption time at various initial concentrations of a CAP and b MB by PMSSAC at 30 °C

Fig. 5
Fig. 5 Separation factor against adsorbate initial concentrations by PMSSAC at 30 °C Fig. 6 Intraparticle diffusion models for a CAP and b MB adsorption by PMSSAC at 30 °C The removal of CAP (Y 1 ) and MB (Y 2 ) was in the range of 17.12%-50.30%and 52.85%-68.43%,respectively.Both responses were developed by quadratic models, as suggested by the CCD.Equations 11 and 12 are the suggested model equations for CAP and MB removal, respectively.The R 2 values obtained for CAP and MB removal were 0.8951 and 0.9624, respectively.These indicate that 89.51% and 96.24% of the total variation in CAP and MB removal were attributed to the experimental variables studied.

Table 1
Model validation for MB and CAP removal

Table 2
Surface area and pore characteristics of PMSS and PMSSAC

Table 3
Proximate and elemental analyses of PMSS and PMSSAC