Utilization of hydrochar derived from waste paper sludge through hydrothermal liquefaction for the remediation of phenol contaminated industrial wastewater

Phenol and phenol derivatives are well-known for their toxicity to all forms of life as well as being non-biodegradable and bio-accumulative (Zhou et al. 2017; Wong et al. 2018). Therefore, the maximum allowable phenolic content in potable water and industrial effluent must be lower than 1 μg/L and 1 mg/L respectively (Tabassi et al. 2017). The most common methods used to treat industrial wastewater are adsorption, chemical precipitation, membrane filtration solvent extraction, ion-exchange, flotation and electrocoagulation (Aziz et al. 2008; Ariffin et al. 2017). Although these methods have been proven to be efficient in removing organic pollutants, the associated high operating cost and low efficiency when low contaminant concentrations are present outweigh their operational feasibility (Aziz et al. 2008; Barakat 2011). The high cost of adsorption can, however, be lowered if less expensive adsorbents, compared to activated carbon, could be found. Recent literature has reported on the use of hydrochar as such an adsorbent (Tan et al. 2015).

Hydrochar is a solid coal-like product that is produced from the thermal decomposition of biomass waste by employing either pyrolysis or HTL (Gollakota et al. 2018), with HTL being favoured because of the transitional structure of the hydrochar and the many oxygenated functional groups and mineral compounds, which promote adsorption, present on the hydrochar surface (Saleh et al. 2016). Various studies have reported on the production of pyrolysis-derived hydrochar as an adsorbent from waste materials such as pine waste (Amerkhanova et al. 2017), paper mill sludge (Yoon et al. 2017), sugarcane bagasse (Wong et al. 2018) and papaya peels (Abbaszadeh et al. 2016). However, limited literature is available on the use of industrial waste such as paper sludge for the production of hydrochar adsorbents through HTL for the removal of organic and inorganic pollutants from industrial wastewater streams.

Paper sludge is a waste product from the pulp and paper industry and approximately 300–350 million tons per of paper sludge is produced annually (Ioelovich 2014), the disposal cost of which amounts to approximately 60% of the total plant operating cost (Hojamberdiev et al. 2008; Gurram et al. 2015). The application of this waste product for the production of affordable adsorbents will lower the environmental impact of landfilled paper sludge and reduce the operating costs for paper mills. A study on the use of paper sludge-based hydrochar produced via pyrolysis for removal of fluoxetine from synthetic and real solutions (Jaria et al. 2015) showed no correlation between the surface area of the hydrochar and the adsorption capacity. Furthermore, activating agents added to the sludge prior to pyrolysis did not significantly improve the adsorption properties of the produced hydrochar. The effectiveness of pyrolysis-derived paper sludge-based hydrochar to remove phosphate and methylene blue from synthetic solution were investigated by Hojamberdiev et al. (2008) and it was reported that maximum multifunctional adsorption capacity of 0.11 mmol/g was observed with a maximum hydrochar surface area of 70 m²/g. Furthermore, phosphate was removed as Ca-phosphate that precipitated on the surface of the hydrochar. Méndez et al. (2009) found that de-inking paper sludge leads to mesoporous material during pyrolysis while paper sludge from virgin pulp mills forms microporous materials. Both materials were used to adsorb Cu²⁺ from wastewater. It was found that the de-inking sludge-based hydrochar had a higher adsorption capacity, which was ascribed by the authors to a higher Ca and Mg exchange content, a higher concentration of oxygenated functional groups on the surface and the larger pore size facilitating easy diffusion transport compared to the mesopores of the virgin sludge-based hydrochar.

To date, most of the studies that produced paper sludge-based hydrochar employed pyrolysis rather than hydrothermal liquefaction (HTL). The latter is better suited to the processing of wet feedstock compared to pyrolysis and operates at milder process conditions. Also, the adsorption performance of paper sludge-based hydrochar produced from HTL for the removal of phenolic compounds from industrial wastewater has not previously been reported. Therefore, in this study, the removal of phenol from industrial wastewater using HTL-derived hydrochar from industrial paper sludge was investigated. Isotherms were drawn to determine the mechanism of adsorption for the individual components and pseudo-second-order kinetic models were used to determine design parameters for industrial application. Mass transfer analysis elucidated the rate-determining step of the adsorption process.

Paper sludge was collected from an industrial paper recycling plant in South Africa at a point after the wastewater treatment and dewatering activities. The mill utilizes both bleached hardwood virgin fibres and recycled fibres as feedstock. Industrial wastewater was collected from the same paper mill at a point after the primary water treatment step. Activated carbon, used as a baseline, was procured from Associated Chemical Enterprises (ACE), South Africa. Both the paper sludge and industrial wastewater were stored at 4 °C after collection until used.

An SS316 batch autoclave with a working volume of 945 mL was filled with approximately 200 g of paper sludge without the addition of water as the paper sludge already contained sufficient water for the HTL reactions to proceed. The autoclave was pressurized to an initial pressure of 0.50 MPa at room temperature and heated to 300 °C at a rate of 3.42 °C/min, where it was kept constant for 15 min before the reactor was allowed to cool naturally and the hydrochar was recovered from the reactor via filtration. The filtered hydrochar was washed with enough acetone to remove all residual bio-oil components. The clean hydrochar was then dried overnight at 105 °C to a moisture content below 10% and stored in airtight containers until used as adsorbents.

The paper sludge, hydrochar, and activated carbon were characterized by proximate, elemental and Fourier-transform infrared spectroscopy (FTIR) analysis. The surface morphology was studied by energy-dispersive X-ray (EDX) analysis. The surface area and mineralogy of the hydrochar and activated carbon samples were determined by Brunauer–Emmett–Teller (BET) and X-ray diffraction (XRD) and X-ray fluorescence (XRF) analysis, respectively. The point of zero charge analysis was done by standard methods. Compositional analysis of the paper sludge was done through fibre analysis according to the Van Soest method (Van Soest 1963). The average particle size of the produced hydrochar was determined by a Hydro 2000 MU Malvern Mastersizer.

The collected wastewater was characterized by measuring the pH, conductivity, chemical oxygen demand (COD), total phenol content, 5-day biological oxygen demand (BOD₅), total dissolved solids (TDS), total suspended solids (TSS), calcium content and total phenolic content. The pH was measured using a Metrohm 692 pH meter and conductivity was measured with an EDT Instruments FE 280 conductivity meter. COD was determined by a photo spectrometric method using a test kit (Macherey – Nagel 100–1500 ppm COD test kit) (analogous to the DIN 150 15705-H45, EPA 4104, APHA 5220 D and DIN 38409-H-41-1 standard methods). BOD₅ analysis was outsourced to an accredited wastewater testing laboratory that used the W044-43-W reference method for BOD analysis according to the DIN EN 25813-G21 standard BOD testing method. TDS and TSS were determined according to the SANS 6049:2010 and SANS 5213:2013 methods respectively. Calcium concentration was determined using an ICP-OES instrument. The total phenolic content was determined using a standard test kit (Merck 0.002–5 ppm phenol test kit) (method analogous to the EPA 420.1, APHA 5530 C and D, ISO 6439 and ASTM D1783-01 standard for determination of phenol in wastewater).

Adsorption experiments were performed in 250 mL glass Duran bottles, using 100 mL adsorbate solution with an initial pH of 8. All the sample bottles were closed once the experiment was started. The pH of the solutions was adjusted by adding either 0.1 M HCl or 0.1 M NaOH solutions, dropwise. Sample bottles were continuously shaken at a rotary speed of 150 rpm in a Labcon platform incubator shaker at 25 ± 2 °C. Samples taken during the synthetic and industrial wastewater experiments were filtered with 0.22 μm Grafiltech and 0.45 μm Pall GHP membrane syringe filters respectively.

The effect of hydrochar dosage on adsorption performance was investigated at hydrochar dosages between 2 g/L and 12 g/L. Once the optimum adsorbent dosage was determined, isotherms and kinetic experiments were performed at initial phenol concentration of between 10 and 150 ppm. The effect of the presence of calcium as a contaminant during phenol adsorption was done at an initial phenol concentration of 10 ppm and a calcium (as Ca(NO₃)₂ 4H₂O) concentration of 600 ppm. The phenol concentration was quantified by an Agilent Technologies 1260 Infinity II high-performance liquid chromatography (HPLC) instrument fitted with an InfinityLab Poroshell 120 EC-C18 column (4.6 × 100 mm and 4 μm with column ID). The HPLC had two mobile phases namely: A – 1% formic acid in water and 1% formic acid in acetonitrile, mixed initially: 92% A, 8% B, and at 10 minutes: 70% A, 30% B. The column temperature was set as 40 °C. An injection volume of 10 μL was used with a Diode Array Detector with a wavelength of 275 nm and a bandwidth of 4 nm. Quantification of phenol components was done using a standard set of calibration curves. The detection and quantification limits for phenol were 2.67 μg/L and 8.9 μg/L respectively.

The equilibrium data were interpreted by fitting Langmuir and Freundlich isotherms and Henry's law isotherms to the data.

The compositional analysis of the paper sludge showed that it consisted of cellulose (34.88 wt. %), hemicellulose (12.03 wt.%), lignin (24.53 wt.%), cutin (14.05 wt.%), protein (2.25 wt.%) and fat by ether extract (1.01 wt.%). The large lignin content of the sludge (22.54%, dry basis) was attributed to the presence of corrugated boxes in the feed to the paper recycling plant, which contain lignin to improve the stiffness of the boxes (Antonides 2000).

A hydrochar yield of 723 ± 84 g/kg (dry basis) was obtained at a reaction temperature of 300 °C, which was comparable to the hydrochar from paper sludge yield of 640 g/kg reported by Jaria et al. (2015) at 800 °C under pyrolysis conditions. Some physical characteristics of the paper sludge, hydrochar, and activated carbon are given on a dry basis (db) in Table 1.

The BET analysis showed a relatively low microporous (1.54 m²/g) and mesoporous surface area (24.78 m²/g) for the hydrochar compared to un-activated pyrolysis-based hydrochar surface areas reported in the literature (Hojamberdiev et al. 2008; Méndez et al. 2009; Jaria et al. 2015; Yoon et al. 2017). The activated carbon had a large micro-and mesoporous surface area of 550.45 m²/g and 277.72 m²/g. Proximate analysis showed the presence of ash species in the hydrochar (41 ± 4 wt. %, db) originating from the paper sludge that was also rich in mineral-based species (22 ± 0.7 wt. %, db). EDX analysis revealed that the mineral-based particles were mostly calcium, silicon and aluminium. XRD analysis indicated that the main mineral constituent of the hydrochar was calcite (CaO), identified by the sharp narrow peaks at 2θ = 27°, 34.2°, 42°, 46.1°, 50.6°, 55.2°, 55.9°, 57°, 67.9° and 76.8°. The presence of a small amount of kaolinite (Al₂Si₂O₅(OH)₄) was indicated by narrow peaks produced at 2θ = 14.2° and 28.9°. XRF analysis validated the distribution of mineral constituents in the hydrochar to be calcite (47.68%), quartz (27.23%) and aluminium oxide (18.50%). Calcium carbonate (CaCO₃) was observed in the FTIR spectra of the paper sludge as a sharp peak at a wavelength of 876 cm⁻¹ and the sharp peaks at wavelengths of 1,032 cm⁻¹, 3,619 cm⁻¹ and 3,694 cm⁻¹ were assigned to kaolinite. During HTL, CaCO₃ decomposes to CaO, which explains the origin of CaO in the hydrochar. FTIR spectra of the hydrochar also showed the presence of many oxygenated functional groups that can act as adsorption sites.

The point of zero charge analysis showed that the point of zero charge of the hydrochar was close to 8. Since the pH of the industrial wastewater used in this study was also close to 8, all experiments were performed at a fixed pH of 8. The effect of the hydrochar dosage on phenol adsorption is shown in Figure 1. The increase in effective surface area with an increase in adsorbent dosage from 2 g/L to 12 g/L led to an expected increase in removal efficiency as more active sites create more opportunities for the adsorption of phenol molecules.

The overall adsorption capacity, however, decreased initially from 3.16 mg/g to 1.51 mg/g as a higher dosage resulted in an increase in the amount of unsaturated active sites while the number of phenol molecules in the solution remained the same (Abbaszadeh et al. 2016). At hydrochar dosages of above 6 g/L, the hydrochar particles started forming agglomerates due to overcrowding, which resulted in some active adsorption sights being masked and thus no increase in adsorption capacity was observed (Gholizadeh et al. 2013; Xu et al. 2013). A dosage of 12 g/L was chosen to use in further studies to limit the effect of the experimental error on the results.

Adsorption isotherms obtained from equilibrium adsorption data is necessary for the design of adsorption systems (Figure 2).

The equilibrium isotherm in Figure 2 conforms to a class L (normal Langmuir) isotherm (Bonilla-Petriciolet et al. 2013), which is typically found for adsorption in aqueous media. Langmuir type isotherms imply that adsorption capacity is directly proportional to the phenol concentration in the aqueous medium until the adsorption sites become limited and molecules start competing for available sites. Furthermore, Langmuir type isotherms indicate that molecules are adsorbed flat on the surface and binding between the molecules and active sites are through weak Van der Waals forces, i.e. physical adsorption (Bonilla-Petriciolet et al. 2013). Further classification of the isotherm as a subclass 4 isotherm indicates the formation of a fresh adsorption surface as seen by the slight change in the slope of the isotherm at an equilibrium phenol concentration of approximately 74 mg/L. The latter can be explained by molecule-molecule interactions on the surface, resulting in fresh adsorption sites. The corresponding parameters calculated for each isotherm model at a 95% confidence level are given in Table 2.

From the accuracy of fit, it can be seen that the data were best described by the Freundlich isotherm. The constants, K_F and n in the Freundlich isotherm is related to the macroscopic behaviour of the adsorption system and the microscopic behaviour of the molecules being adsorbed (Yang 1998). A heterogeneity factor (n) larger than 1 indicates adsorption onto a heterogeneous surface through physical adsorption. The Freundlich isotherm has also been associated with multilayer adsorption and molecule-molecule interactions, which confirms the isotherm classification as a subclass 4 Langmuir type isotherm. Adsorption onto more than one type of active site was confirmed with FTIR analysis (Figure 3).

Some changes in the spectra of the fresh and used hydrochar were observed. New peaks were observed in the spectra of used hydrochar at 2,360 cm⁻¹, the single peak at a wavelength of approximately 460 cm⁻¹ split into two small weak peaks and the peak observed at approximately 400 cm⁻¹ increased to a large, sharp peak in the spectra of the used hydrochar. The weak peak in the spectra of the fresh hydrochar at 1,600 cm⁻¹ became a shoulder in the spectra of the used hydrochar and the peaks at 1,450 cm⁻¹ and 1,050 cm⁻¹ were slightly larger. The peak at 2,360 cm⁻¹ (H bond in –COOH (Abdel-Ghani et al. 2015) together with the peak at 1,050 cm⁻¹ (primary alcohol C-O stretch) was assigned to the physical adsorption bonds of phenol to active carbon and ether groups on the hydrochar surface. The presence of additional phenyl groups attached to the hydrochar surface after batch adsorption is confirmed by the slight increase in absorption at 1,450 cm⁻¹ (C = C-C aromatic ring stretch) and the change in the weak shoulder peak at 1,600 cm⁻¹. The FTIR analysis thus confirms adsorption onto different surface groups and thus the good fit of the Freundlich isotherm to the equilibrium adsorption data.

Evolution of the kinetic profile of an adsorption system is an important part of an adsorption process as it offers insight into the rate of adsorption, the residence time required to reach equilibrium conditions and the effectiveness of the adsorbent used (Bonilla-Petriciolet et al. 2013). The kinetic profile obtained for this study is shown in Figure 4.

The adsorption of phenol onto the hydrochar produced required between 30 and 360 min to reach equilibrium conditions. The intermediate equilibrium plateaus observed in the data for all initial phenol concentrations are attributed to adsorption of phenol molecules onto different types of active sites (as confirmed by the FTIR analysis). As one type of site is saturated, a plateau is reached in the adsorption capacity and another type of adsorption site is then filled, until all available sites have been saturated and true equilibrium adsorption is reached.

In order to determine the rate of adsorption of phenol onto the hydrochar produced, the pseudo-first-order and pseudo-second-order rate models were fitted to the experimental data. However, only the results of the fitting of the pseudo second-order model are given here, because the pseudo-first-order model did not fit the equilibrium data (R² ≤ 0). The kinetic parameters obtained from fitting a pseudo-second-order kinetic model at a 95% confidence level are given in Table 3.

The equilibrium adsorption capacities predicted by the model correlated well with the experimental values obtained. It can be seen from Table 3 that both the specific rate of adsorption and equilibrium adsorption capacity (q_e) increases with an increase in initial phenol concentration. The initial rate of adsorption (h₀) increases as the initial phenol concentration increases and the driving force for mass transfer is increased.

Adsorption isotherm and adsorption kinetics analysis has shown that more than one type of active site is involved in the adsorption process and there is a change in the mechanism of adsorption, as seen from the changes in h₀ and q_e with a change in initial phenol concentration.

A number of mass transfer steps are involved in the adsorption of a solute from an aqueous media onto a solid surface. It is assumed from the general theories of momentum and mass transfer that each adsorbent particle is surrounded by a transport film across which the concentration of the adsorbate changes from the bulk fluid to the surface of the solid. The first step in the overall adsorption process is thus the bulk diffusion of adsorbate molecules to the outer layer of the transport film surrounding the solid adsorbent particle. Molecules thus need to be transported across the particle film before attaching to the outer surface of the adsorbent, from where it can be transported through pore diffusion into the internal pore structure of the solid particle for adsorption onto available adsorption sites. The rate of transport in each step is a function of the adsorbate concentration and properties as well as the characteristics of the adsorbent. Mass transfer, momentum transfer and thermodynamics all play a role in determining the rate of transport during each step, but the slowest of the transport steps determines the overall rate of adsorption for a specific adsorbate. External transport is usually rate limiting when adsorbate molecules are relatively small and present at low concentrations, while internal transport is usually rate limiting for larger molecules present in higher concentrations, but the rate limiting step needs to be determined for each adsorption system. Mathematical theory to describe the different phenomena influencing adsorption is by nature complex and difficult to solve. However, it has been noted that for intra-particle transport, most solutions show a square root relationship of adsorption time with the fraction of adsorption sites occupied (Kouyoumdjiev 1992). The presence of more than one rate-determining step can thus be determined by examining the relationship between q_t and t^0.5 (Jaman et al. 2009). This relationship represents initial intra-particle diffusion and if the relationship shows more than one straight line, there is more than one mechanism controlling the adsorption process. From a plot of q_t and t^0.5 (Figure 5), there is one straight line relationship for an initial phenol concentration (C₀) of 10 to 30 mg/L. At C₀ of 60 mg/L, there was an initial linear relationship up to 120 minutes, followed by a plateau for the rest of the contact time.

The film mass transfer coefficient can be determined from a plot of C/C₀ vs time, where C is the concentration of phenol in the adsorption media at time t and C₀ is the initial phenol concentration. The effective intra-particle diffusion coefficient can be determined through a plot of –ln(1-F²) versus time, where F is the ratio between adsorption capacity at time t (q_t) and the equilibrium adsorption capacity (q_e) (Girish & Murty 2016). The calculated k_f and D_e values at a 95% confidence level are given in Table 4. Values for 10 and 60 ppm initial concentration are not reported due to low fit of accuracy (R² < 0.2) during determination of the calculated parameters.

The industrial wastewater used in this study had a high calcium content that could influence the adsorption of phenol onto the hydrochar (Tabassi et al. 2017). The adsorption performance of hydrochar for adsorption of phenol with or without the presence of calcium is compared in Figure 6.

The presence of calcium nitrate (Ca(NO₃)₂·4H₂O) in the adsorption media did not have a significant effect on the removal efficiency of phenol. It is expected that the presence of dissolved salts will influence the electrostatic forces between the phenol molecules and the adsorbent surface. In this study, it was shown that the adsorption process is controlled by film mass transfer and the concentration of phenol molecules on the external surface of the hydrochar is thus expected to be low at any given time (Sobiesiak 2017). Therefore, FTIR analysis (Figure 7) was performed on the used hydrochar to determine the impact of the calcium salt on the interaction of phenol with the hydrochar.

The spectra for the used hydrochar in the presence of Ca(NO₃)₂·4H₂O show the typically expected peaks at 3,640 cm⁻¹ (-OH group in water), 2,350 cm⁻¹ together with 1,650 cm⁻¹ (H₂O) and the broad double peak at 1,360–1,480 cm⁻¹ (NO₃⁻). The peak at 2,360 cm⁻¹ that disappeared was previously shown to be one of the –COOH active sites on the hydrochar surface where phenol adsorbed. The shoulder associated with phenol adsorption onto the hydrochar surface at 1,600 cm⁻¹ disappeared and the peak indicating phenol adsorption at 1,450 cm⁻¹ is masked by the nitrate peak. The presence of nitrate ions thus shielded some of the preferred phenol adsorption sites, but not the extent where it negatively impacted the total phenol adsorption capacity. The slightly higher removal efficiency of phenols in the presence of the calcium salt observed in Figure 6 could be due to a slight decrease in the ionic strength of the solution due to adsorption of nitrate ions onto the hydrochar surface.

Characteristic properties of collected wastewater were determined to be: COD of 2,720 mg/L, BOD₅ 695 mg/L, BOD₅/COD ratio of 0.26, total phenol concentration of 16.7 mg/L, total calcium concentration of 579 mg/L, TDS of 3,794 mg/L and conductivity of 3.1 mS/cm. The wastewater had a slight alkaline pH (7.8) and a relatively high total suspended solids (TSS) concentration (320 mg/L) as expected for recycling paper mill wastewater (Pokhrel & Viraraghavan 2004; Rahman & Kabir 2010). The calculated BOD₅/COD ratio was lower than 0.4, indicating that the wastewater is not readily biodegradable and would require a tertiary treatment step to reach the effluent limits. The latter is due to the relatively high phenolic content of the wastewater. The paper mill that produced the wastewater used in this study uses recycled fibres with virgin fibres to ensure that paper products produced have increased mechanical properties. Therefore, low phenol concentrations in the wastewater are expected as most of the lignin was not broken down but contained within the desired paper products produced in the process. The high calcium concentration in the wastewater is in agreement with the calcium observed in the paper sludge.

The complex composition of the wastewater and the low concentration of some compounds made a qualitative analysis of individual components difficult and, therefore, the efficiency of the hydrochar to lower the COD of the wastewater stream was used as a quantitative measure to follow the removal efficiency. The COD removal efficiency of the hydrochar was compared to that of activated carbon (the industry standard) (Figure 8).

Approximately 44 ± 2% of the COD was removed by the activated carbon after 240 minutes and 31 ± 1% by the hydrochar after 360 minutes. After 24 hours, removal efficiency for the activated carbon increased slightly to 47 ± 3% and that of hydrochar decreased slightly to 29 ± 1%. The slight changes in concentration after 24 hours fall within or very close to the standard deviation in the COD measurement and were thus presumed to be the result of the normal variation in the data. The difference in removal efficiency between activated carbon and hydrochar is significant and probably the result of the much higher surface area of the activated carbon compared to that of hydrochar. The average COD removal efficiency obtained for hydrochar (∼30%) and activated carbon (∼41%) in this study is lower than what is reported in literature (Huggins et al. 2016), but higher than that reported for phenol removal using paper sludge-based biochar (Calace et al. 2002). Table 6 presents the phenol content of the wastewater before and after adsorption.

From Table 6, it can be seen that all the phenol derivatives were removed to a concentration below that of the detection limit of the HPLC, but only 9% of the other organic components present in the waste water were removed. These results show that hydrochar can be considered as a suitable and inexpensive adsorbent for continuous practical wastewater treatment of wastewater containing phenolic components as contaminants. Lastly, the COD removal efficiencies obtained for the hydrochar by adsorption were higher than other wastewater treatment methods such as ozonation (approximately 20%), but not aerobic treatment technologies such as aerated stabilisation basins (67%) and bio-filters (52%) (Amat et al. 2005; Ashrafi et al. 2015). An analysis of total phenolic components in the wastewater confirmed that the COD removal observed was mostly due to the adsorption of phenolic components onto the hydrochar. The concentration of residual phenolic components in the wastewater was well below the allowable acceptable limit of 1 mg/L for phenolic compounds present in industrial effluents (World Health Organization 1994). The phenol removal efficiency obtained by the hydrochar in real industrial wastewater was slightly higher than in a synthetic wastewater environment containing only phenol. According to Calace et al. (2002), the adsorption of substituted phenolic compounds with a more hydrophobic nature (due to having low pKa values) will adsorb more easily than phenol with a higher solubility and therefore higher pKa value. Furthermore, an increase in the ionic strength due to the presence of inorganic salt species such as divalent calcium, magnesium, sodium and potassium has been found to promote the adsorption of organic contaminants that are more hydrophobic than phenol such as phenolic compounds substituted with additional chlorine and nitrogen-based groups.

In this study, it was evident that hydrochar derived from paper sludge can be used as an adsorbent to remove hazardous organic pollutants such as phenol and/or phenol derivatives from both synthetic and real industrial wastewater streams. However, low phenol removal efficiencies were obtained in synthetic phenol solutions due to the low surface area and therefore, the limited amount of active sites available on the surface of the produced hydrochar. The low surface area was largely contributed to the blockage of pores by mineral-based compounds such as calcite and kaolinite. The presence of calcium in a binary solution with phenol did not significantly influence the adsorption of phenol. The hydrochar achieved a COD removal efficiency of only 31 ± 5.1%, but all phenolic components were removed from the wastewater. Therefore, it can be concluded that hydrochar derived from paper sludge through HTL has the potential to replace expensive adsorbents such as activated carbon to remove harmful organic pollutants such as phenolic components from real industrial wastewater streams. The direct usage of wet biomass waste products such as paper sludge for HTL purposes also offers cost-effective waste management solutions that reduce the environmental impact of the pulp and paper industry.

The authors declare that there is no conflict of interest.

The authors gratefully acknowledge the financial support of the Paper Manufacturers' Association of South Africa (PAMSA), Mpact and the National Research Foundation of South Africa (NRF grant UID 91635) for this research. This work is based on the research supported by the National Research Foundation. Any opinion, finding and conclusion or recommendation expressed in this material is that of the author(s) and the NRF does not accept any liability in this regard.

All relevant data are included in the paper or its Supplementary Information.