Surfactant modified waste ash for the removal of chloro and nitro group substituted benzene from wastewater

Water pollution is among the major consequences of rapid industrialization in many parts of the world. To ensure sustainable fresh water supplies, effective wastewater treatment technologies are needed. Treatment technologies such as advanced oxidation processes, biochemical, coagulation, ion-exchange, electrolysis, adsorption and reverse osmosis have been deployed for wastewater treatment and conservation (Gupta et al. 2012; Ayanda 2014). The application of wastewater treatment techniques is typically gauged based on installation, operational and maintenance cost at commercial level. The cost of water treatment using different processes ranges from 10 to 450 $/m³ of purified water, except the adsorption process where the cost usually varies from 5 to 200 $/m³ of treated water. Adsorption is considered as an attractive technology because of its effectiveness in treating wastewater containing pollutants at low concentration and the availability of different inexpensive materials from industrial and agriculture waste that could be potentially used as adsorbents (Mo et al. 2018; Hossain et al. 2020). Among them, waste fly ash is a promising material that can be utilized as an adsorbent. Wang et al. worked on microwave irradiated coal fly ash and used it as Fenton like catalyst for the treatment of polyacrylamide contaminated water. They reported 75% reduction in total organic carbon (TOC) under optimized experimental conditions (Wang et al. 2019).

He et al. synthesized A-type zeolite from fly ash sampled from thermal power plant for the removal of nickel cations from wastewater. The adsorbent achieved 94% removal efficiency and corresponding uptake capacity of 47 mg/g (He et al. 2020). A low-cost filter media from fly ash was prepared and characterized by Sanas et al. The synthesized filter media was used for the reduction of total solids, biochemical oxygen demand (BOD) and chemical oxygen demand (COD) present in aqueous effluent. With a 10 cm thick layer of ash, they reported 70% reduction in total dissolved solids and 65 and 66% decrease in BOD and COD load, respectively (Sanas et al. 2016). Laghari et al. examined the treatment of burnt coal, in fly ash form, for dyes adsorption from industrial effluent. The results showed that by using optimum dosage (4 g/L) of synthesized adsorbent, there was a significant reduction of chemical oxygen demand (54%), color (76%), turbidity (72%) and total suspended solids (98%) from dye-containing effluent (Laghari et al. 2015). Literature reveals that waste coal fly ash has great prospects in applications for environmental remediation, particularly in wastewater treatment. Considering the potential of fly ash for wastewater treatment, surface modification of coal fly ash without converting it to secondary material could evolve as a straightforward and simple route to synthesize an adsorbent for treatment of wastewater.

Several surface active agents have been studied and reported by many researchers under different experimental conditions (Mao et al. 2015; Liu et al. 2016). Surface modification of substrate with large organic molecules such as surfactants, which cannot be intercalated into pores and stick to the external surface only, is a promising way to enhance the adsorption capacity of materials. Surfactants are divided into cationic, anionic, nonionic, and amphoteric based on the type of attached hydrophilic functional group. The surfactants could act as promising surface-active agents because of charged head groups in their structure. Although many studies have been reported on the utilization of fly ash as an adsorbent, the removal of organic compounds like benzene derivatives using sodium dodecyl sulfate (SDS) has been rarely reported. The aim of this research is to develop an SDS modified fly ash adsorbent for the removal of organic pollutants from wastewater. To investigate the properties of the developed adsorbent, various characterization techniques, including scanning electron microscopy (SEM), surface porosity analyzer, thermogravimetric analysis (TGA) and Fourier transform infrared (FTIR) were used. Chlorobenzene (CB) and nitrobenzene (NB) pollutants were taken as model pollutants to investigate the efficacy of modified fly ash through batch adsorption experiments. Both benzene derivatives act as solvent and are used as raw material to manufacture other chemicals. The Environmental Protection Agency (EPA) has classified these chemicals as main concerned pollutants because of their higher toxicity and potential to build up in the environment. Even at small concentrations, these can cause neurasthenia, dizziness, and increase the risk of cancer after exposure (Nematollahzadeh et al. 2021; Ren et al. 2022). Batch adsorption runs were conducted at different experimental conditions to eradicate the NB and CB from polluted water. Details on the adsorption process, mechanism and kinetics have also been studied. This work contributes to the ongoing efforts towards developing the effective adsorbents for the abatement of water pollution.

The raw coal fly ash (CFA) was collected from ICI soda ash plant, Khewra Salt Range, Pakistan. The modification of CFA was done by using reagent grade sodium dodecyl sulfate (SDS, Duksan Pure Chemicals, Korea). Nitrobenzene (97%, NB) was purchased from Riedel-de Haën (Germany), hydrochloric acid was acquired from Sigma—Aldrich (China), sodium hydroxide was purchased from Merck (Germany), and chlorobenzene (99.5%, CB) was supplied by RCI Labscan Limited (Thailand).

Raw CFA was sieved using W.S. Tyler Ro-Tap Shaker (EW-59986-62) and undersize of 80 mesh Tyler standard screen was collected, washed with distilled water and oven-dried at 105 °C. The dried CFA was then modified using anionic surface-active agent, SDS. Required amount of SDS surfactant was dissolved in the specified amount of distilled water to prepare the surfactant solution of 55 mmol/L. For the activation of the fly ash, CFA powder was mixed in SDS solution in a ratio of 2:125 (W/V, g/mL) and the resultant mixture was stirred using a magnetic stirrer (Wiggen Hauser, MSC digital, USA) at 1,000 rpm for 4 h. The particular ratio of fly ash to surfactant solution was selected to keep the weight ratio of fly ash to surface active agent as 1:1 to avoid aggregation of surfactant molecules on the surface of CFA. After 4 h of mixing, dispersed particles were allowed to settle down and solid residue was collected after filtration. The solid residue was then dried in a vacuum oven (Chincan, DZF 6051, China) at 60 °C for 2 h. The modified CFA was then stored in closed glass bottles for subsequent use. The prepared material was abbreviated as SDS-modified coal fly ash (SCFA).

The efficacy of the SCFA as an adsorbent was investigated through batch adsorption experiments for the removal of chlorobenzene (CB) and nitrobenzene (NB) from polluted water. The benzene derivatives stock solutions having concentration of 1,000 ppm were prepared by adding appropriate amounts of NB and CB in distilled water and later, the test solutions of desired strength were prepared by diluting the stock solution. In a typical run, 100 mL of each NB and CB having 50 ppm concentration was taken in separate conical flasks and SCFA was added to each flask at room temperature to observe the removal efficiencies of SCFA. Fifty ppm concentration was kept constant for all the experiments except isothermal runs at different initial concentrations to avoid dilutions before analysis. The flasks were placed on an orbital shaker (Model SB08, USA) for 1 h to attain the adsorption equilibrium. The solution was then filtered, and the filtrate was analyzed using UV-vis spectrophotometer (SP-3000 Optima Inc., Tokyo, Japan) and the intensity of light absorption corresponding to wavelength λ_max = 209 nm and λ_max = 268 nm for CB and NB, respectively, was recorded. The absorbance value of filtrate was compared with standard calibration curve of each pollutant to calculate the pollutant concentration in the filtrate. The removal efficiency of SCFA was calculated using Equation (1). The effect of adsorbent dosage was investigated by varying the adsorbent concentration from 0 to 0.5 mg/mL (0, 0.05, 0.1, 0.15, 0.2, 0.25, and 0.50). The adsorption kinetics were determined by analyzing the adsorption of both organic solutes at different time intervals (1, 3, 5, 10, 20, and 30 min). The influence of pH was evaluated at room temperature by varying pH within the range of 3–10. The pH adjustment was achieved using 0.1 M HCl and 0.1 M NaOH solutions. Adsorption tests were also performed at 293, 303 and 323 K by varying the concentration of CB and NB from 10 to 350 mg/L for the purpose of isotherm and thermodynamic analysis. Each test run was repeated three times and average values are reported in the results.

A Fourier transform infrared (FTIR- JASCO-4100) spectrophotometer was used to detect various functional groups present on the surface of the modified sample. In this regard, the transmittance spectrum was recorded within the 4,000–500 cm⁻¹ wavenumber range. Thermal behavior of CFA and SCFA was analyzed using thermogravimetric analyzer (Q600 SDT, TA instruments, USA). The accurate weight sample was heated at a rate of 10 °C/min from atmospheric to 800 °C under continuous flow rate of N₂ set at 20 mL/min. Field emission scanning electron microscope (FESEM, MIRA3 TESCAN) was used to capture the images of CFA and SCFA for surface morphology analysis. Surface area and pore size distribution were determined by analyzing the CFA and SCFA through Micromeritics ASAP 2020 surface area analyzer. During analysis, the sample was degassed under vacuum at high temperature and then N₂ adsorption-desorption curve was recorded to determine the surface area and pore size distribution.

The minor peaks observed in the band range of 2,800–3,000 cm⁻¹ in SCFA are assigned to the C-H symmetric and asymmetric stretching vibrations corresponding to –CH₂ and –CH₃ alkyl groups. The new sharp peaks appearing in SCFA at 1,384 and 1,429 cm⁻¹ may correspond to the C-H bending vibrations of alkane and strong S = O stretching of sulfate in SDS. The doublet bands at 798 and 788 cm⁻¹ and the peak at 563 cm⁻¹ can be attributed to the rocking motion associated with CH₂ groups and C-C present in the hydrocarbon part of the surfactant. The appearance of these bands in the FTIR scan of SCFA clearly indicates the presence of surfactant on the fly ash surface. The thermogravimetric (TG) and derivative thermogravimetric (DTG) profiles of CFA and SCFA are presented in Figure 1(b). The degradation of both samples can be viewed in terms of three distinct regions. The first region (up to 100 °C) covers the loss of moisture in CFA and SCFA and, as a result, approximately 2% and 0.5% weight loss was observed in this region, respectively. The second region (100–450 °C) is showing desorption of water of hydration along with devolatilization of both CFA and SCFA, thereby, 2% and 1.5% weight loss can be seen, respectively. A kink in the DTG curve of SCFA in the second degradation region is attributed to thermal decomposition of the carbon chain of surfactant in the sample, thus confirming its deposition on CFA. The third stage spans between 450 and 700 °C and approximately 10 and 7% weight loss are experienced by CFA and SCFA, respectively. This weight loss is attributed to the decomposition of residual coal present in the ash (Aslam 2022). The third region is represented by a large peak in the DTG curve and a shift of peak temperature towards lower value for SCFA (∼574 °C) as compared to CFA (610 °C) can also be observed. The resistance to thermal degradation by CFA is probably due to the strong bonding and electrostatic interactions of zeolitic material with alumina and silica present in the ash (Li & Ishiguro 2016; Nguyen et al. 2018). The tailing region in TGA curve represents the residual weight of the samples and it is more in SCFA than CFA which probably resulted from the charring of the organic chain present in SCFA (Tanzifi et al. 2017; Siddiqui et al. 2018; Şenol et al. 2020).

After modification with anionic surfactant, the reduction in the surface roughness of fly ash is observed, this is probably caused by the organic layer developed after the modification with SDS. The disappearance of broken, needled and spherical particles can be noticed in Figure 2(b). The SCFA surface looks desolated, having numerous cracks and fractures, which could be beneficial for the adsorption of pollutants.

The porosity features (i.e., Barrett-Joyner-Halenda (BJH) pore size distribution and Brunauer-Emmett-Teller (BET) surface area) of CFA and SCFA are presented in Figure S1. The BET surface area of the CFA slightly changed from 9.07 m²/g to 9.46 m²/g after surfactant treatment. In the pore size distribution graph, SCFA curve is broadened and lies above the CFA curve, which indicates that surfactant treatment produces more mesopores. As a result, average pore width (inset table) in SCFA is 88.7 Å as compared to 172.9 Å average pore width of CFA surface pores.

Interaction of anionic surfactant with CFA surface both electrostatically (through –SO³⁻ groups) and hydro-phobically (through dodecyl chain) promote the rupture of low-energy intermolecular bonds and reform the CFA surface which is also supported through SEM images. This modification in the surface texture and porosity enhances the adsorption capability of CFA as noticed in the subsequent adsorption experiments.

Figure 4(a) reveals that the uptake capacity of SCFA for CB increases with increasing pH while it decreases with increasing pH for NB. At pH < pH_PZC, SCFA will have net positive charge and attract the electronegative chlorine (–Cl^−δ). As a result, uptake slowly increases from 195 mg/g to 220 mg/g. Furthermore, at pH > pH_PZC, the adsorbent surface is negatively charged but the adsorption of CB continues to rise from 220 mg/g to 240 mg/g because of electropositive carbon ring which is attracted towards the negative surface of the adsorbent (Aramendía et al. 2003). On the other hand, the behavior of NB is opposite that of CB. It is observed that the increase in solution pH led to a gradual decrease in the NB uptake because of the reduction in the surface positivity of SCFA. This indicates that the dominating nucleophilic oxygen atom (^−δO = N⁺–O⁻) has high affinity towards the adsorbent surface due to strong electrostatic force of attraction. But this attraction diminishes as the surface gets negative (i.e., pH > PZC) and it causes strong electrostatic repulsion with the adsorbate, which is induced by the nitro group (^−δO = N⁺–O⁻) in nitrobenzene. Thus, the adsorption occurs because of the inductive effect of the benzene ring. Overall, the adsorption of CB is higher than NB, because chlorobenzene possess a high magnitude of the inductive effect, become more reactive (less non-polar, low dipole value) to the electrophilic attack, as compared to nitrobenzene (Heaney 1970).

The effect of adsorption time on the uptake of CB and NB is presented in Figure S2 and it is observed that the removal of both benzene derivatives increases with time. The initial rapid increase reflects the availability of abundant vacant sites. The adsorption of CB is equilibrated at ∼ 225 mg/g while NB settled at 90 mg/g. Further contact of SCFA with pollutants showed no substantial improvement in the uptake. The experimental data was further analyzed by applying the kinetic models to explore the mechanism of adsorption. The mathematical expressions of both pseudo first and second order kinetic models are presented in Table 1.

Table 1

Model equations and regression results

Kinetic model .	Regression results .	CB .	NB .
Pseudo first order qe & qt are equilibrium and instantaneous uptake (mg/g) and k1 (min−1), first order rate constant	qe,cal	224.9	99.55
	k1 (min−1)	2.036	0.961
	R2	0.927	0.966
	RSS	32.36	0.534
Pseudo second order k2 (g mg−1 min−1), second order rate constant	qe,cal (mg g−1)	233.6	100.8
	k2 (min−1 g mg−1)	0.022	0.046
	R2	0.99	0.978
	RSS	1.529	0.336

Kinetic model .	Regression results .	CB .	NB .
Pseudo first order qe & qt are equilibrium and instantaneous uptake (mg/g) and k1 (min−1), first order rate constant	qe,cal	224.9	99.55
	k1 (min−1)	2.036	0.961
	R2	0.927	0.966
	RSS	32.36	0.534
Pseudo second order k2 (g mg−1 min−1), second order rate constant	qe,cal (mg g−1)	233.6	100.8
	k2 (min−1 g mg−1)	0.022	0.046
	R2	0.99	0.978
	RSS	1.529	0.336

RSS, residual sum of squares.

View Large

Adsorption equilibrium was investigated using two and three parameter models as summarized in Table S1 (Chen et al. 2022; Nirmala et al. 2022). The Freundlich isotherm model is suitable for evaluating the heterogeneity of the surface and to investigate multilayer adsorption. The Dubinin Radushkevich (D-R) isotherm model explains the adsorption on heterogeneous surfaces on the basis of gaussian energy distribution. Its parameters are temperature dependent and through them, one may be able to differentiate the physical and chemical adsorption mechanism by evaluating the value of mean free energy per unit molecule of adsorbate. The Hill isotherm model assumes cooperative adsorption phenomenon where the adsorbate influences the other binding sites of the same homogeneous adsorbent. Khan and Koble Corrigan (KC) isotherm models are generalized and represent both Langmuir and Freundlich isotherms under limiting conditions of solution concentration. At low concentrations, these models reduce to Langmuir isotherm and at higher concentrations, they approach the Freundlich isotherm model (Al-Ghouti & Da'ana 2020).

The isotherm model parameters obtained through nonlinear regression of experimental data are presented in Table 2. The Freundlich isotherm reasonably fits the data and produces a fairly high R² (>0.97) for both CB and NB pollutants. The Freundlich constant, K_F, varies from 2.81 × 10⁻⁴ to 3.57 × 10⁻⁵ with increase in temperature for CB, indicating that it is an exothermic adsorption. For NB, the K_F extends from 7.08 × 10⁻⁹ to 5.89 × 10⁻³ when the temperature changes from 293 to 323 K, suggesting the endothermic nature of adsorption. According to D-R isotherm, the value of mean free energy (E_d) is less than 8 kJ/mol for all temperatures. This reflects that the adsorption follows the physisorption mechanism where hydrophobic/organic interactions play a role in the removal of CB and NB from the polluted water (Siddiqui et al. 2018; Şenol et al. 2020).

For the three-parameter isotherm models, Hill isotherm shows a good correlation with the experimental data (i.e. R² > 0.94) for both NB and CB. The n_H value of Hill isotherm model for both CB and NB adsorption is greater than one, which revealed the positive cooperativity in binding (Tanzifi et al. 2017). However, the q_s obtained from Hill isotherm model does not follow the exothermic nature of CB, i.e. q_s decreases with increase in temperature, but remains constant for NB adsorption. Consequently, the Hill isotherm is not suitable for describing the mechanism of adsorption of CB and NB. The regression of isotherm data using the Koble Corrigan model gives the lowest R² for CB among all tested isotherm equations. But the R² is fairly high for the experimental isotherm data of NB. Besides this, n_KC is less than 1 for NB adsorption data as shown in Table 2. This indicates that the KC model is not suitable, despite a good correlation coefficient (Ayawei et al. 2017). The Khan isotherms produce high regression coefficient and the uptake calculated through this model agrees with the exothermic nature of CB and endothermic behavior of NB as exhibited by the q_s. Therefore, this model can be considered suitable for the adsorption of CB and NB onto SCFA.

Thermodynamic analysis sheds light on the degree of spontaneity and mechanism of adsorption. The thermodynamic variables, ΔG°, ΔH°, ΔS°, were calculated from adsorption equilibrium data. Pertinent equations relating to these quantities are presented in Table S2. A graph of ln K vs 1/T produces ΔS° and ΔH° from the intercept and slope of the plot, respectively. The Van't Hoff plot for both CB and NB are shown in Figure S3. The thermodynamic quantities calculated on the basis of Figure S3 are summarized in Table 3. It is clearly seen that ΔG° is negative for CB in a given temperature range, which indicates that the process is globally spontaneous and feasible. The value of ΔG° becomes less negative with rise in temperature at a constant concentration but increases at constant temperature with increase in initial concentration of pollutant as shown in Table 3. Therefore, it can be concluded that adsorption of CB onto SCFA is advantageous at lower temperature and higher concentration. The negative value of ΔH° at all initial concentrations indicates that the process is exothermic and follows the physisorption mechanism. This is because ΔH° is much less than 41.8 kJmol⁻¹, which is the threshold number differentiating physisorption and chemisorption mechanisms (Radhakrishnan et al. 2018; Tahira et al. 2019). Moreover, the negative ΔS° for CB removal reveals that the adsorption process is enthalpy driven with no significant structural changes during the adsorption. It shows less disorderliness at the interface during the adsorption process. This is probably because of multilayer adsorption which tends to organize the system and reduces surface randomness (Antunes et al. 2012; Maszkowska et al. 2014). For NB, ΔG° is positive at all temperatures in the studied concentration range except Co = 300 mg/L where it becomes negative. The ΔG° values in Table 3 reveal that the degree of non-spontaneity of NB adsorption decreases as the temperature increases. A similar behavior was observed for the initial concentration of NB. These results may suggest that the adsorption of NB onto SCFA is favorable at high temperature as well as at higher initial concentration. Positive ΔH° suggests endothermic nature and it can also be considered as physisorption, where the Vander Waal's forces and electrostatic interactions play a role in the adsorption of NB. Positive ΔS° shows increased randomness at solid/liquid interface during adsorption of NB onto SCFA (Saha & Chowdhury 2011; Yu et al. 2020).

The isosteric heat of adsorption was also evaluated from equilibrium data using Clausius-Clapeyron equation as given in Table S2. Isosteric heat gives an indication about the surface heterogeneity of sorbent, which is important for describing the adsorption mechanism. The isosteres corresponding to fixed uptakes of both pollutants are presented in Figure S4. It is observed that isosteres for both CB and NB showed linear behavior. The isosteres with negative slope for CB (Figure S4a) substantiates the exothermic nature of CB adsorption. However, the slopes are positive (Figure S4b) for NB adsorption over SCFA. The corresponding isosteric heats are evaluated from slope values and results summarized in Table 4. The values of isosteric heats of adsorption are found to be lower than 80 kJ/mol, which support that CB and NB are physically sorbed onto the SCFA surface (Naseem et al. 2022). This is also in agreement with earlier findings of thermodynamic analysis. The different values of ΔHx for different uptakes indicate the heterogeneous surface of SCFA. A change in magnitude of ΔHx at different surface coverage reflects the lateral interaction that may exist among adsorbed moieties. At lower loadings, there are probably stronger interactions between pollutant and the active sites of the adsorbent, resulting in higher isosteric heat. With increase in loading, interaction between CB/NB and adsorbent surface decreases due to either pore filling and/or formation of monolayer, leading to a decrease in isosteric heat of adsorption (Murthy et al. 2019; Abdulsalam et al. 2020). To assess the reusability of prepared SCFA, used adsorbent was separated from the polluted solution and soaked in 50 mL distilled water. The mixture was continuously stirred at 40 °C for 1 h then passed through filter paper and solid residue was collected and dried at 60 °C in the vacuum oven. The removal efficiency of SCFA was examined for five consecutive runs and the results are presented in Figure S5. The graph shows that there is small change in the removal efficiency of both CB and NB. Drop in removal capacity is probably because of deterioration of surfactant layer on the SCFA because of repeated use. Table S3 compares the adsorption capacities of SCFA and various reported adsorbents for CB and NB removal from aqueous phase. The SCFA exhibited superior adsorption capacity compared to many adsorbents.

The surface properties of fly ash were modified using anionic surface-active agent, i.e. SDS under different experimental conditions. The chemical and physical properties of the surfactant treated fly ash sample were compared to that of the untreated fly ash. The modified ash sample was tested for the adsorption of organic pollutants (chlorobenzene and nitrobenzene). FTIR spectra had indicated that SDS is well deposited onto the fly ash surface. SEM micrograph revealed that loading of CFA with SDS resulted in the formation of numerous cracks and fractures on the surface of the SCFA, resulting in higher pore volume and surface area as confirmed by the BET data. The thermal degradation profiles had shown a lower weight loss in case of SCFA (approximately 10%) than CFA (approximately 15%), which indicate that SCFA is more thermally stable than the raw fly ash. The optimum dose of raw CFA and SCFA was found to be 20 mg for the maximum adsorption of NB and CB at solution pH. The SCFA showed maximum removal at 90 and 40% with CB and NB, respectively. The adsorption of NB and CB followed a pseudo second order kinetic model with the formation of a surface complex, indicating rate determining step in the adsorption process. The adsorption of chlorobenzene is exothermic over both adsorbents while nitrobenzene exhibited an endothermic behavior. The isotherm analysis depicted the multilayer adsorption of both pollutants onto the surface of the surfactant-modified fly ash at different temperatures. Overall, the modification of CFA is an attractive way to develop an effective adsorbent for wastewater treatment.

The authors acknowledge the support of Higher Education Commission, Pakistan through National Research program for Universities for funding this research through project 8039/Punjab/NRPU/R&D/HEC/2017. In addition, Department of Chemical Engineering, UET, Lahore is also acknowledged.

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

The authors declare there is no conflict.