First investigations on removal of nitrazepam from water using biochar derived from macroalgae low-cost adsorbent: kinetics, isotherms and thermodynamics studies

Pharmaceutical compounds and their metabolites as emerging contaminants in the environment have recently been given more attention. They are considered as potentially serious threats to the environment and human health (Desbiolles et al. 2018). Although the concentrations of pharmaceutical compounds and their metabolites are low in the environment, their harmfulness and risk to humans and animals are very high (Boxall et al. 2012). Their presence in the environment may create drug resistance amongst pathogens and lead to many negative impacts such as convulsion, brain damage, cancer diseases, reproductive behavioral disorder, liver damage, cardiovascular diseases and lung defects as well as disturbances in gene expression and other deleterious eco-toxicological effects on aquatic organisms (Saravanan et al. 2014; Peltzer et al. 2017). Pharmaceutical compounds are discharged to the environment and water resources through different sources (Ilyas et al. 2020). Several studies indicate that a huge number (∼10,000) of pharmaceutical compounds are prescribed for human and veterinary treatments (Boxall et al. 2012; Caldwell et al. 2014) and about ∼ 10–90% of the evoked pharmaceutical compounds are excreted in their parent form, whereas others are excreted as metabolites or conjugated forms from the human body (Hirsch et al. 1999; Kummerer 2009). The pharmaceutical compounds excreted by patients reach the soil, drinking water, groundwater, lakes, rivers and oceans (Li et al. 2013; Schaider et al. 2014). Other major sources of pharmaceutical compounds in aquatic environments are the disposal of expired pharmaceutical products, hospital wastewater, manufacturing wastes, and veterinary contaminants (Vieno Tuhkanen & Kronberg 2007; Phillips et al. 2010; Vulliet & Cren-Olivé 2011; Tewari et al. 2013). Pharmaceutical compounds entering the soil infiltrate the groundwater, and it can reach seawater through agricultural runoff from fields that are treated with livestock slurries contaminated with pharmaceutical compounds (Kummerer 2010). Among the pharmaceutical compounds, psychiatric drugs such as benzodiazepines (e.g. Diazepam, Temazepam and Nitrazepam) are used to treat short-term sleeping problems (insomnia) and sometimes applied to treat epilepsy when other medications fail. They are widely prescribed and frequently detected in different environmental media such as drinking water (Lv et al. 2019), wastewaters (Thiebault et al. 2017a), surface water (Calisto & Esteves 2009) and sediment (Thiebault et al. 2017b) due to their inefficient treatment (Thiebault et al. 2017c). Hence, proper removal of these compounds from the aquatic environment is highly essential. The above facts have been considered by many researchers, and many techniques have been applied for removal of pharmaceutical compounds from wastewater such as adsorption (Naghipour et al. 2018; Bhowmik et al. 2020; Debnath et al. 2020; Zahra et al. 2020), ion exchange, precipitation, coagulation, chlorination, sedimentation, sand filtration, biological treatment, membrane separation, advanced oxidation and ozonation (Westerhoff et al. 2005; Snyder et al. 2007; Ikehata et al. 2008). Among these techniques, adsorption is one of the most widely used methods for the removal of pharmaceuticals from water. Adsorption technique is simple, effective, well established, adaptable to many treatment formats and relatively inexpensive. Other advantages include the fact the adsorption does not produce sludge and has a capability to remove most forms of organic material (Fakhri & Adami 2014; Khadir et al. 2020). The cost of adsorption depends mainly on the type of adsorbent used. This was the driving force to investigate different adsorbents from different sources such as activated carbon, powdered activated carbon, activated alumina, charcoal, brick powder, activated sludge, zeolites and biomass. Each of these materials has advantages and disadvantages. However, biomass is renewable as well as considerably cheaper and more abundant than other commercially available materials. In particular, naturally available macroalgae have many advantages. Their cell wall structure and components contain cellulose, alginate and other polysaccharides. These components have many chemical functional groups such as carboxylic acid, hydroxyls, and amines that provide high affinity and selectivity towards organic compounds such as Nitrazepam. These ecofriendly macroalgae and their derived biochar adsorbents showed promising results in removing inorganic and organic pollutants such as heavy metals (Nazal 2019) and phenolic compounds (Eugenia et al. 2006; Nazal et al. 2020). However, to the best of our knowledge, there is no information available on the adsorptive removal of Nitrazepam from water. Therefore, in this study, for the first time, the adsorptive performance of biochar derived from Sargassum macroalgae for Nitrazepam was investigated. The effects of pH, adsorbent mass, and adsorbate concentration on the adsorption efficiency were studied. The experimental results at equilibrium were fitted with the most frequently adsorption isotherms and kinetics models. The thermodynamics parameters of Nitrazepam adsorption were also calculated.

The hypnotic drug Nitrazepam was purchased from Sigma-Aldrich, USA. All the chemicals (i.e. sodium hydroxide, sodium chloride and nitric acid) were analytical reagent (AR) grade and purchased from Merck, Germany. The Acetonitrile (99.9%) HPLC grade solvent, used as mobile phase, was purchased from Fisher Scientific, Germany. Deionized water was used for the experiment.

Sargassum macroalgae (SM), collected from the coastal water of the Arabian Gulf close to the Saudi Arabia, was thoroughly washed, and then dried in the oven. A biochar derived from Sarrgassum macroalgae (BSM) was prepared through thermal pyrolysis of the dried SM under an inert nitrogen atmosphere without any chemical treatment. The detailed preparation procedure and characterization of adsorbents have been reported elsewhere (Nazal et al. 2020). Briefly, the dried SM was thermally treated under nitrogen atmosphere in a furnace at 500 °C for 2 hours. Washed material was dried at 110 °C for 48 hours. The resulting biochar was preserved in a tightly closed container to be used later for the adsorption experiment. The point of zero charge pH, at which the net charge of the adsorbent is neutral, was measured by mixing 50 mg of adsorbent in 20 mL of 0.5 M NaCl solutions having different pH values (2–10) for 24 hours at room temperature. Initial and final pH values were recorded using an OAKTON PC2700 pH meter.

A standard stock solution of 1000 mg/L of Nitrazepam (NZP) was prepared by dissolving the required amount of NZP in methanol. Subsequent test solutions were prepared from the stock solution by appropriate dilution with deionized water. Figure 1 shows the chemical structure of NZP, and Table 1 presents its main physical and chemical properties.

Concentrations of NZP before and after the adsorption experiments were analyzed in triplicate and measured using High Performance Liquid Chromatography coupled with a Diode Array Detector (HPLC-DAD) at 273 nm wavelength. The chromatographic conditions are summarized in Table 2.

To study the effect of the weight of added adsorbents (SM and BSM), different masses (5–250 mg) of adsorbents were added to 25 mL of NZP (5 mg/L) solutions. The resultant solutions were agitated at 24 °C and 140 rpm for 5 days.

The effect of pH on the adsorptive removal of NZP using Sargassum macroalgae adsorbents before and after pyrolysis was studied in pH ranging from 2 to 10. The pH was adjusted by 1.0 M NaOH and 2% HNO₃. 25 mg of adsorbent were mixed with 25 mL of NZP (5 mg/L) and kept in a water bath shaker (BS-31, Lab Companion, Korea) for 5 days, at 24 °C, at a shaking speed of 140 rpm.

The effect of NZP concentration, in the range of 1 mg/L to 30 mg/L, on the removal efficiency of SM and BSM adsorbents was studied at a constant temperature of 24 °C. Fixed quantities of adsorbents of 1 g/L and 0.2 g/L of SM and BSM, respectively, were used. The resulting solutions were agitated for 5 days. The adsorption isotherms of NZP onto SM and BSM were investigated by fitting the equilibrium adsorption data with the most frequently used isotherm models (i.e. Freundlich, Langmuir and Temkin). Table 3 shows the linear equations of the used isotherm models.

where, q_e (mg/g) is the adsorption capacity, C_e (mg/L) is the NZP concentration at equilibrium. K_f (mg/g)(L/mg)^1/n and n (dimensionless) are the Freundlich model parameters. Q_max (mg/g) is the maximum monolayer adsorption capacity in the Langmuir model and b is the Langmuir constant. In the Temkin model, A_T (L/g) is the equilibrium binding constant, b_T is the Temkin constant, R (J/K mol) is the ideal gas constant and T (K) is the temperature in Kelvin.

In order to study the effect of contact time on the removal efficiency, fixed weights of 25 mg of SM and BSM adsorbents were added to 25 mL of NZP (5 mg/L) solutions in capped vials. The mixtures were agitated for different time intervals (i.e. 0.5, 1, 2, 4, 6, 8, 24, 72,120 and 144 hours). The solutions were then filtered and analyzed to measure the NZP concentration. The adsorption data were then fitted with different adsorption kinetic models (presented in Table 4).

where q_t (mg/g) is the adsorption capacity at a certain time t (min), in pseudo-first order and pseudo-second order kinetics, and k₁ (min⁻¹) and k₂ (g/mg/min) are the rate constants respectively. In the Elovish model, a (mg/g/min) is the initial adsorption rate and b (g/mg) is the desorption constant. k_id (mg/g/min^0.5) and C (mg/g) in the Weber equation are the intra-particle diffusion rate constant and the constant related to the thickness of the boundary layer, respectively.

The detailed characterization was performed, discussed and reported in our previous publication (Nazal et al. 2020). Briefly, the surface morphology of SM before and after pyrolysis is shown in Figure 2. It has been found that the surface of SM (shown in Figure 2(a)) is covered with biological species (Diatoms), which may result in covering the active adsorption sites and explains the low adsorption capacity of macroalgae, before pyrolysis, for the NZP compound. After pyrolysis, the surface of BSM (shown in Figure 2(b)) became more clear with some openings and holes, which may explain the improvement in the adsorption performance of BSM as a result of increasing the contact area between NZP and the active adsorption sites on the surface.

The proximate analysis results showed that in the SM the percentage of fixed carbon (46.45%) increased after pyrolysis to be 74.57%, which indicates the increase of hydrophobic carbon contents of BSM. On the other hand, the FTIR showed the surface of both SM and BSM has different functional groups (e.g. hydroxyl (OH), carboxylic acid (COOH) and carbonyl (C = O)) that may contribute to the interaction between the used adsorbents and Nitrazepam, the emerging pollutant, through the hydrogen bonding. Moreover, the surface area (3.8 m² g⁻¹) and pore volume (0.006 mL g⁻¹) of BSM were larger than the surface area (1.34 m² g⁻¹) and pore volume (0.005 mL g⁻¹) of SM before pyrolysis. These findings may contribute to elaborating the difference between the adsorption capacity of MA and BSM adsorbents as well as explaining the adsorption mechanism onto their surfaces.

Figure 3 shows the effect of the loaded SM and BSM adsorbents' dose on the removal efficiency of NZP. It has been found that the maximum removal efficiencies of 60 and 98% are obtained using 4 g/L and 1 g/L of SM and BSM adsorbents respectively. After the treatment of SM adsorbent, the removal efficiency was significantly improved. This might be attributed to the increase of the BSM adsorbent's surface functional groups, which in turn enhanced the acid-base and electrostatic interactions between the pharmaceutical compound NZP and the BSM adsorbent. Figure 3 also shows that the removal efficiency increases by increasing the adsorbent weight. This is due to the increase of adsorption sites.

The effect of pH on the removal efficiency of NZP onto SM and BSM adsorbents is shown in Figure 4. It has been found that the pH of the solution slightly influences the removal efficiency. This may be attributed to the slight change in the adsorbate-adsorbent electrostatic interaction. Considering the pKa values of NZP (3.2 and 10.8) and the pH_pzc of SM (5.06) and BSM (9.87), it is expected that, in the range of pH between 5 and 10, the anionic form of NZP dominates and the surface charges of SM and BSM are negative and positive, respectively. Therefore, at pH 5, the removal efficiency of NZP using SM slightly decreased, while it slightly increased using the BSM adsorbent. Hence, in all the adsorption experiments, the NZP solutions having a pH of 5.8 were used without any adjustment.

Figure 5(a) shows the effect of NZP concentration on the removal efficiencies of SM and BSM adsorbents. It has been found that using SM (1 g/L) is not efficient for removing the NZP compound under the experimental conditions. In the case of using BSM (0.2 g/L), the maximum removal efficiency for NZP was 75%, which is 20 times higher than the maximum removal efficiency of an SM adsorbent. However, their removal efficiencies decreased as the initial concentration of NZP increased. This is attributed to the saturation of the adsorbents’ surface and the increase of repulsion between the adsorbed molecules on the surface and free molecules in solution. Because of the low adsorption capacity of SM, the adsorption experimental data of SM could not be fitted with the isotherm models. However, different isotherm models were tested for NZP adsorption onto BSM adsorbents. The plots of the linear forms of the tested adsorption isotherm models are presented in Figure 5(b)–5(d).

The models’ parameters for NZP adsorption onto BSM are summarized in Table 5. These parameters were calculated from the slope and the intercept of the linear form of the corresponding adsorption isotherm model. It has been found that the Freundlich isotherm is the best fit to the experimental adsorption results with a squared correlation coefficient (R²) of 0.9980 and summation of relative error (SRE) of 0.3448. This reveals that the BSM adsorbent's surface is heterogeneous and the physical adsorption mechanism is predominant. The n and K_f values higher than 1 indicate that the adsorption of NZP onto BSM is favorable. The higher magnitude of these values corresponds to greater heterogeneity and higher adsorption capacity respectively (Li Quinlivan & Knappe 2002). The value of R_L (0.34), between zero and one, shows the favorability of NZP adsorption. The maximum monolayer adsorption capacity of BSM for NZP is 143.12 mg/g. The interactions between NZP and the BSM surface involve three possible mechanisms. These are (i) electrostatic interaction, (ii) π-π and n-π interactions between the π and n electrons on NZP and π-electrons on the surface of the BSM adsorbent and (iii) hydrogen-bonding interaction.

Contact time is a vital parameter affecting the removal efficiency and giving an insight into adsorption kinetics. As shown in Figure 6, the SM and BSM removal efficiencies for NZP increased rapidly in the first 24 hours, which is attributed to the maximum availability of unoccupied adsorption sites on the surface of the adsorbents. Then, they reached equilibrium within 48 and 120 hours using SM and BSM adsorbents, respectively. It is also apparent that the removal efficiency of BSM is higher than SM. This is due to the dominance of ionic forms of NZP molecules as well as the negative and positive charge of the SM and BSM surfaces, respectively, at the pH (5.8) of the solutions.

Figure 7 shows the plots of the linearized kinetics models (i.e. PFO, PSO and Elovich). The adsorption kinetics’ parameters are calculated from the slopes and the intercepts of the corresponding linear plots and summarized in Table 6.

The obtained squared correlation coefficients (R²) are close to 1 and the q_e experimental (q_{e exp.}) and q_e predicted (q_{e pred.}) values are very close to each other in PSO for both adsorbents. Although the R² values obtained in the PFO are high, the q_{e exp.} and q_{e pred} values are not close to each other. In addition, the summation of relative error (SER) between and was calculated for the tested kinetics models. It has been found that the lowest SER is for PSO. This clearly indicates that the adsorption of NZP on SM and BSM follows the PSO kinetics. This also indicates the involvement of chemisorption in the rate-determining step in the adsorption mechanism (Ho & McKay 1999).

For further investigation of the adsorption mechanism and to know if NZP adsorption onto SM and BSM are described as a diffusion-controlled process, the equilibrium adsorption data were fitted to Weber and Morris’ intra-particle diffusion kinetics model. As shown in Figure 8, plotting q_t versus t^0.5 reveals two linear parts. This indicates that NZP adsorption onto both adsorbents is governed by two steps of diffusion. The first step is fast and related to the diffusion of NZP from the bulk solution to the available adsorption sites on the surface of the adsorbent, while the second step arises from the adsorption of these compounds to the adsorption sites, and this step occurs more slowly than the first one (Khadir et al. 2020). Table 7 summarizes the calculated intra-particle parameters. As shown, k_id values of the first step for adsorption of NZP onto SM and BSM adsorbents are higher than k_id values of the second step, while the values of the C constant for the first steps are lower than those in the second steps. This suggests that the thickness of the boundary layer of SM is smaller than that of the boundary layer of BSM. As a result, the adsorption of NZP is faster onto an SM adsorbent.

The adsorption's enthalpy and entropy were obtained from the slope and intercept of log K_d vs 1/T plot (shown in Figure 9). Table 8 summarizes the thermodynamics parameters for NZP adsorption onto SM and BSM adsorbents. It has been found that the free energy (ΔG) values of NZP adsorption onto SM are positive in the studied temperature range (24–45 °C). This indicates that the adsorption is non-spontaneous and thermodynamically unfavorable. However, these values are negative for NZP adsorption onto BSM adsorbents, which confirms that the adsorption is spontaneous and thermodynamically favorable. Decreasing ΔG values with increasing temperature, and positive values of ΔH for NZP adsorption onto SM and BSM adsorbents, indicate that the adsorption process is endothermic. In addition, the low values of ΔG (ranging between −10.73 and − 0.28 kJ/mol) and the obtained value of ΔH (147.51 kJ/mol) reveals that the adsorption mechanism is mixed and the physical adsorption is predominant (Saha & Chowdhury 2011; Zaghouane-Boudiaf & Boutahala 2011). The small positive entropy values in the adsorption system in this study indicate that randomness increases at the adsorbate/adsorbent interface.

Generally, the adsorption process of organic compounds onto carbon based material depends on three main factors, namely the nature of the adsorbent and adsorbate (e.g. porosity, size of molecules, functional groups and net charge) and adsorption conditions (e.g. pH and temperature of solution) (Gupta et al. 2013). Considering the previously reported characterization results of the SM and BSM adsorbents, NZP molecules may interact with the surface of SM through hydrogen-bonding and electrostatic interactions, while it may interact with the prepared biochar through hydrophobic, π-π and n-π, electrostatic and hydrogen-bonding interactions. Figure 10 illustrates the key adsorption mechanism of NZP onto the surface of the biochar derived from Sargassum macroalgae. The proposed mechanism is supported by the obtained adsorption isotherms and thermodynamics results, where the adsorption of Nitrazepam onto both adsorbents is dominated by a physisorption mechanism.

For the first time, the adsorptive removal of the pharmaceutical active compound Nitrazepam (NZP) from water was investigated, and the nature and kinetics of its adsorption onto the Sargassum macroalgae (SM) and the biochar derived from this macroalgae (BSM) were studied in this work. The removal capability of the macroalgae adsorbent (MA) for the NZP compound was improved after thermal pyrolysis, with a removal efficiency up to 98% and an adsorption capacity of 143 mg/g. The adsorption equilibrium results showed that the adsorption was obtained through a mixed mechanism dominated by physisorption and followed by pseudo-second-order kinetics. The thermodynamics results revealed that the adsorption of NZP onto SM was non-spontaneous and thermodynamically unfavorable, with a positive value of ΔG. However, the adsorption of NZP onto BSM was spontaneous and thermodynamically favorable with a negative value of ΔG. In addition, the adsorption process of NZP onto both adsorbents was endothermic with a positive ΔH. Furthermore, the low values of ΔG and the obtained values of ΔH disclose that the adsorption mechanism is mixed and physical adsorption is predominant. The positive value of ΔS reflected the affinity of the adsorbent toward NZP, indicating an increase in randomness at the liquid/solid interface. The findings in this study showed that the biochar derived from Sargassum macroalgae (BSM) could provide a promising adsorbent for removing pharmaceutical compounds from contaminated water.

The authors declare that there is no conflict of interest regarding the publication of this article.

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