Introduction

Over the last decades, surface and groundwater pollution has dramatically increased as a result of the release of several pollutants in the environment, such as organic contaminants and metal trace elements (MTE). In particular, the presence of MTE in water is due to the massive use of agro-chemicals in agriculture, mining activities as well as the discharge of wastewater from metallurgic industries in the environment (Blazquez et al. 2005). MTE can also come from natural inputs, such as soil erosion, volcanic eruptions and alteration of bedrock. In addition to water resources contamination, the presence of these pollutants has adverse effects on human health (Fijalkowska et al. 2021, Tomczyk et al. 2020). For instance, ingestion of high quantities of copper in drinking water may cause gastrointestinal bleeding, hypotension, convulsions and significant DNA damages (Yu et al. 2001). Therefore, it is urgent to remove these MTE from water through the use of different technologies, such as precipitation, electrolysis or (bio)sorption (Chakraborty et al. 2020, Cheraghi et al. 2015, Haghighat and Ameri 2016, Tyagi et al. 2020). Among these techniques, (bio)sorption is the least expensive and the most effective technology for removing contaminants from water; the other techniques are likely to generate toxic byproducts and are more difficult to apply on a large scale. Indeed, a wide range of adsorbents, such as activated carbons, zeolites, clays and lignocellulosic materials, have been used for the retention of inorganic pollutants (Miguel et al. 2006, Rajec et al. 2015). Recently, there has been a growing interest for the development of new bio-based, low-cost and sustainable adsorbents (Viglašová et al. 2018) whose production is based on renewable biomass. This is the case of biochar, a carbon-rich solid material obtained through pyrolysis i.e. the thermochemical conversion in an oxygen-poor environment (Lehmann and Joseph 2009). Indeed, biochar has gained attention since the discovery of the Terra Preta soils (approximately 7000 years old anthropogenic soils) in the Amazon (Parikh et al. 2014). Researchers found out that these soils, initially characterized by a low fertility, had been enriched through historical applications of charcoal, providing them a higher cation exchange capacity and nutrient retention (Sohi et al, 2010). Besides, the contribution of biochar in climate change mitigation through carbon sequestration in soils is the subject of great interest (Agegnehu et al. 2016). For these reasons, much research is focused on the investigation of biochar potential for the remediation of organic and inorganic contaminants in water. Biochars are known as very effective adsorbents (Inyang et al. 2016), and numerous studies considered the valorization of biowastes through biochar production and the use of charred materials for the removal of MTE from water (Yu et al. 2018). A large variety of feedstocks were used for this purpose, such as agricultural wastes (manure, wheat straw, rice husk, waste wood, pistachio, peanut and hazelnut shells), energy crops (corn, wood pellets and oil seed rape), sewage sludge and several other wastes (Sun et al. 2018). Shan et al. (2020) have shown that a biochar obtained from the pyrolysis of peanut shells at 400 °C was very efficient for metal removal from aqueous solution. Indeed, the maximum adsorption capacities of this biochar for Pb2+, Cd2+, Ni2+, Cu2+ and Zn2+ were between 82 and 239 mg.g−1 (Shan et al. 2020). In a similar study, Park et al. (2016) used a biochar produced from sesame straw at 700 °C to eliminate MTE from water. The retained amounts were between 55 and 102 mg.g−1 for Pb2+, Cd2+, and Cu2+. Biochars prepared from anaerobically digested sludge (Jin et al. 2016), date seed (Mahdi et al. 2018), rice husk (Masih et al. 2018), orange pomace (Santos et al. 2015), animal manure (Wang et al. 2020), were also used for copper sorption and proved high removal efficiencies. The cationic exchange capacity, the specific surface area, porosity and the pH value of biochars were underlined as the main parameters governing copper adsorption. In the different studies, the physico-chemical and structural properties of biochars were shown to be strongly dependent on feedstock nature and pyrolysis conditions, notably temperature, residence time and type of reactors/kilns.

In 2018, 2,894,000 tons of olive oil were produced worldwide (Hogan 2018), and the annual production of wood in the world was estimated to over 3 million cubic metres in 2004 (Planetoscope 2020). Thus, huge amounts of olive pomace (OP) and wood sawdust (SD) are generated every year as the main byproducts of olive oil extraction process and wood industry. A reasonable fraction of these ligno-cellulosic residues can be used as raw materials for biochar preparation. Nevertheless, throughout the scientific literature in this field, biochars were produced in laboratory under controlled pyrolysis conditions. To our knowledge, no study explored the possibility of employing traditionally produced biochars for MTE removal from polluted water, whereas several traditional practices exist since ancient times and are still used today for charcoal production, such as Kon-Tiki kilns in Asia, Brazilian beehive kilns, earth pits and earth mound heaps in Africa (FAO, 2020).

In this context, the aims of the current research work are (i) the traditional production and characterization of biochars prepared from raw OP and pine SD and (ii) the study of their copper adsorption capacities in aqueous solution. The traditional biochars were prepared at Mnihla (Tunisia) in 2018, and copper experiments were carried out at ICMR, UMR CNRS 7312 (Reims, France) in 2019.

Materials and methods

Raw materials

OP and pine SD were collected from an oil mill and a carpentry shop both located in the area of Mnihla (North Tunisia). The collected materials were air-dried, ground and then sieved to 200 µm before undergoing an acido-basic treatment (with 0.1 mol.L−1 KOH and 0.1 mol.L−1 HNO3) in order to obtain an insoluble solid in a wide pH range. Finally, both biosolids were washed several times with distilled water and then dried at 40 °C for 48 h.

Biochar preparation

Biochar production in this study was inspired from a traditional method employed by Tunisian farmers to produce charcoal which is locally called “merdouma” (literally means “buried”). In a pottery plate (used to replace the soil stratum needed in the merdouma practice), 250 g of raw materials (OP or SD) was stacked in a small heap, about 25 cm in diameter at the base and 7 cm high. Then, the heap was covered with wood and branches before adding an impermeable layer of wet soil to prevent air entry. A small hole was left at the top of the pile to light the fire to biomass. Once flames appeared, the heap was completely closed to prevent air entry (Fig. 1). Temperature was monitored during the pyrolysis process using an infrared thermometer. The heap was left to cool gradually for half a day, and then opened and biochar was extracted. The obtained biochars (named OP-B and SD-B for OP and pine SD, respectively) were air-dried, ground and then sieved to 200 µm before undergoing an acido-basic treatment in the same way as described for the raw materials.

Fig. 1
figure 1

Photographs of traditional biochar preparation showing heap construction (a) firing point (b) and white smoke release from the top indicating that fire has taken hold (c)

Full size image

Characterization of raw materials and biochars

The main physico-chemical properties of raw biosolids and their derived biochars were determined. The pH value was measured for a 1/25 solid/solution ratio in ultra-pure water (ALPHA Q, 18 MΩ cm−1). The specific surface area and the total pore volume were measured using the BET method by N2 adsorption/desorption at 77 K (Micromeritics ASAP 2020A, United States). Elemental contents of carbon, hydrogen, nitrogen, sulphur (oxygen content was estimated by difference) were determined using an elemental analyzer (CHNS analyzer, Thermo Scientific). The different solids were characterized by Fourier Transform Infrared (FTIR) spectroscopy (Perkin Elmer FTIR Spectrometer) in the transmission mode by dispersing the solids in KBr pellets.

Copper adsorption experiments

The adsorption capacities of the four studied solids (SD, OP, SD-B, OP-B) towards Cu(II) were determined using the batch equilibrium technique at room temperature (293 K) in ultra-pure water (ALPHA Q, 18 MΩ cm−1), and experiments were duplicated. Controls containing Cu(II) without any solid were also prepared to check the absence of retention onto vessels during the whole duration of the experiments. The effects of adsorbent dose (solid charge), contact time, pH and copper initial concentration were investigated.

For experiments as a function of solid charge, 12.5 mg to 100 mg of solid was suspended in 20 mL of ultra-pure water and stirred using an orbital shaker (IKA KS 4000i control) at 200 rpm for 24 h to ensure solid hydration. Then, copper was introduced from a stock solution (Cu(NO3)2 at 10–3 mol.L−1 in ultra-pure water) to reach a final concentration of 10–4 mol.L−1 (6.35 mg Cu.g−1) in a total volume of 25 mL (adjusted by ultra-pure water addition). The suspensions were stirred during 12 h to be sure to reach adsorption equilibrium.

Adsorption kinetic experiments were performed at an introduced Cu concentration of 10–4 mol.L−1 in a 1 g.L−1 suspension (selected after experiments as a function of solid charge) and stirred during a contact time varying in each batch from 15 min to 8 h.

As for experiments as a function of pH, they were carried out at an introduced Cu concentration of 10–4 mol.L−1 in a 1 g.L−1 suspension with varying pH value in each batch from 3 to 9 by dropwise addition of HNO3 (10–1 mol.L−1) or KOH (10–1 mol.L−1). The suspensions were stirred for 2 h (equilibrium time deduced from adsorption kinetic experiments).

Finally, experiments as a function of Cu concentration were conducted at 1 g.L−1 and pH = 5.3 with an introduced Cu concentration varying between 2. 10–5 and 10–3 mol.L−1 (1.27 to 63.5 mg Cu.L−1) in each batch (2 h stirring).

At the end of each adsorption experiment, the suspensions were filtered through a 0.22 µm cellulose acetate membrane, and the filtrates were acidified with a drop of concentrated nitric acid in order to avoid the metal precipitation in the filtrates before analysis.

The remaining Cu concentration in these filtrates was measured by ICP-OES using an Icap 6300 Duo (Thermo Scientific) in axial mode. Yttrium was used as internal standard and λ = 324.754 and 327.396 nm were selected for Cu. The limits of detection and quantification were equal to 5 and 17 nmol.L−1, respectively. The Cu adsorbed amount was calculated by difference between the introduced and remaining Cu concentrations.

Adsorption isotherm modelling

The Langmuir, Freundlich and Sips adsorption models were used to describe the adsorption isotherms of copper on SD, OP and their derived biochars using Microcal™ Origin® software according to Eqs. 1, 2 and 3, respectively.

$$q_{e} = q_{\max L} \frac{{K_{L } C_{e } }}{{1 + K_{L } C_{e } }}$$
(1)
$$q_{e} = K_{F } C_{e}^{n}$$
(2)
$$q_{e} = q_{\max S} \frac{{K_{S } C_{e}^{m} }}{{1 + K_{S } C_{e}^{m} }}$$
(3)

where Ce is the equilibrium copper concentration in solution (mg.L−1), qe is the adsorbed amount of copper per gram of adsorbent at equilibrium (mg.g−1), KL is the Langmuir constant (L.mg−1), qmaxL is the Langmuir maximum adsorption capacity (mg.g−1), KF (Ln mg1−n g−1) is the Freundlich constant, n is the nonlinearity constant, KS is the Sips constant (Lm.mg−m), qmaxS is the Sips maximum adsorption capacity (mg.g−1), and m is the Sips model exponent.

Results and discussion

Temperature monitoring during traditional pyrolysis

The evolution of temperature in sawdust and olive pomace heaps is reported in Fig. 2. For OP, immediately after ignition, temperature was equal to 224 °C, and within 5 min, it increased to reach a maximum of 303 °C. This phase corresponded to combustion due to air entrapped in the heap which produced heat. Once air was consumed (5 min) the temperature decreased gradually to reach a temperature of 102 °C after cooling. As for SD, the starting temperature was equal to 322 °C, but the cooling phase was much faster than for OP. Indeed, within only half an hour, temperature fell drastically to about 40 °C, to finish at 27 °C after two hours. This rapid cooling suggests that some fresh air was able to get into the sawdust heat during the process. Indeed, during traditional pyrolysis based on earth mound kilns, the earth covering the heap may sink, and any cracks or holes which form must be closed to prevent air leakage (FAO 2020). In our experiment, some small cracks may have formed and would be responsible for the rapid cooling of the sawdust heap.

Fig. 2
figure 2

Temperature evolution during traditional pyrolysis of OP and SD

Full size image

Characterization of the raw materials and their biochar

Characterization of adsorbents

The composition and main physico-chemical properties of the raw biomasses and their derived biochars are given in Table 1.

Table 1 Composition and physico-chemical properties of the solids
Full size table

The results show that SD and OP presented pH values of 5.8 and 5.3, respectively, comparable to those obtained by other authors, who found a pH of 5.7 for sawdust (Ahmad et al. 2009) and a value of 5.4 for olive pomace (Martinez-Garcia et al. 2006). This acidic nature can be attributed to the abundance of carboxylic and phenolic moieties at the solid surface. After biomass pyrolysis, their pH values became alkaline (9.3 for SD-B and 8.1 for OP-B). This pH increase after pyrolysis was reported by numerous studies (Nzediegwu et al. 2020) and can be explained by the fact that high temperatures eliminate the acid functional groups present on the surface of biochars (-COOH and -OH) and favour the formation of inorganic carbonates, e.g. MgCO3 and CaCO3 (Nanda et al. 2016).

The two raw materials presented similar surface area values (≈ 1.5 m2.g−1). After pyrolysis, the surface area of the sawdust-derived biochar increased to 4.1 m2.g−1, value very close to the one reported by Pariyar et al. (2020) for biochar prepared from sawdust at 350 °C (3.9 m2.g−1). The increase in surface area is generally attributed to the progressive degradation of hemicellulose, cellulose and lignin (Soni and Karmee 2020). In particular, hemicellulose degradation occurs at temperature ranging between 200 and 350 °C (Dorez et al. 2014) and is responsible for a major part of volatile compounds release. Thus, the lower specific surface area of OP-B may be explained by a lower amount of hemicellulose of OP compared to SD (Elalami et al. 2020) leading to a limited thermal degradation.

The elemental analysis showed that the studied raw materials presented relatively similar carbon, nitrogen, hydrogen, and oxygen contents (Table 1). However, after traditional pyrolysis, SD and OP exhibited a different evolution of their elemental composition. Indeed, the C content, which was approximately 50% for both raw materials increased to 65.5% for OP-B, while it dropped down to 19.3% for SD-B. While the increase in C content during pyrolysis is widely accepted among researchers (Chowdhury et al. 2016), the observed strong decrease in the SD-B carbon content was unexpected. Hadroug et al. (2019) also noticed a decrease in the C content in the case of manure-derived biochars, but this loss did not exceed 5–6% and corresponded to a higher process temperature (500–600 °C). These authors attributed this decrease to the presence of labile carbon fraction in the raw materials. In our case for SD-B, the C loss was much higher (around 30%) with a lower temperature (300 °C). It is thus unlikely that this high loss was only due to the removal of labile carbon. Thus, our hypothesis is that, as suggested by temperature profile of SD heap, it is probable that some oxygen entered the artisanal kiln and led to the partial combustion of biomass, hence the decrease in C content of the pyrolyzed SD. According to Lee et al. (2013), the uncontrolled process of biomass burning in open pits (ancient practice), one of the most common processes for biochar production, may produce significant amounts of CO2 and CO, which can exceed 20% wt. of the C content.

Atomic ratios H/C and O/C or (O + N)/C have been used to investigate the aromaticity (Wang et al. 2015) and polarity (Usman et al., 2015) of biochars. The lower the H/C ratio is, the higher the polycondensation degree of organic compounds. The H/C ratio decreased for both SD-B and OP-B indicating the formation of higher proportions of aromatic C during pyrolysis (Yue et al. 2017), in accordance with previous studies (Stylianou et al. 2020). O/C and (O + N)/C ratios slightly decreased for OP-B suggesting a less polar and hydrophilic character, while they were higher for SD-B indicating an increase in biochar polarity.

Identification of functional groups of raw materials and their biochar by FTIR spectroscopy

SD and OP raw materials spectra (Fig. 3) evidenced an intense broad band centred on 3425 and 3456 cm−1, respectively, attributed to hydroxyl groups corresponding to phenol and alcohol moieties (Ahmed and Khaled, 2019). Their pyrolysis led to the decrease in these bands due to dehydration during the pyrolysis process (Hadroug et al., 2019). Peaks at 2930 and 2925 cm−1 for SD and OP, respectively, corresponding to the νC-H stretching band, indicated the presence of hemicellulose and cellulose in raw materials. These peaks disappeared in their biochars, which suggests the degradation of cellulosic compounds during pyrolysis. Broad medium peaks at 1644 and 1634 cm−1 for SD and OP, respectively, corresponded to C = O stretching and non-conjugated or aromatic ring C = C stretching vibrations. The intensity of these bands dramatically decreased in their corresponding biochar. The broad medium band around 1385 cm−1 in the raw material spectra was attributed to C–H and O–H (phenol) bending vibrations. These bands were strongly modified in the corresponding biochars (decreasing and/or shifted), which may be an indicator of the degradation and depolymerization of cellulose, hemicelluloses and lignin (Cantrell et al. 2012), since these compounds are rich in phenolic groups. Bands between 1300 and 1000 cm−1 corresponded mainly to C–O (alcohol, ether, ester … moieties) and C–N (aromatic and aliphatic amines) stretching vibrations.

Fig. 3
figure 3

FTIR spectra of SD, OP and their traditional biochars

Full size image

Copper retention

The adsorption of Cu(II) was studied on raw materials and their corresponding biochars in order to test their performance in the removal of this metallic cation from aqueous solution. In this way, the influence of various parameters was considered.

Effect of solid charge

Four solid/solution ratios from 0.5 to 4 g.L−1 were tested for the retention of Cu(II) onto SD, OP, SD-B and OP-B (Fig. 4).

Fig. 4
figure 4

Effect of solid charge on Cu(II) adsorption onto SD (dark grey), SD-B (light grey) (a) and OP (dark grey), OP-B (light grey) (b) (histograms in mg.g−1 and symbols + lines in %), [Cu2+]0 = 10–4 mol.L−1, contact time = 6 h, T = 20 °C, natural solid pH

Full size image

As expected, the percentage of adsorbed Cu(II) increased with the solid charge, whereas the adsorbed amount in relation to the total solid weight (mg.g−1) decreased. For both solids, the traditional pyrolysis leads to an enhancement of adsorption, in terms of both percentage and amount. A solid charge of 1 g.L−1 was selected for further experiments, in order to obtain a good compromise between significant adsorbed amount and percentage.

Effect of contact time

Adsorption kinetic curves obtained at 1 g.L−1 for the four solids are reported in Fig. 5.

Fig. 5
figure 5

Adsorption kinetics of Cu(II) onto SB, OP, SD-B, and OP-B ([Cu2+]0 = 10–4 mol.L−1, solid charge = 1 g.L−1, T = 20 °C, natural solid pH)

Full size image

For all solids, the kinetics were very fast with an adsorption equilibrium time lower than 15 min for SD, SD-B, and OP, and inferior to 1 h for OP-B, corresponding to the beginning of the plateau. For further experiments, the batches were stirred during 2 h in order to have similar conditions for the four solids. As for the percentage of adsorbed Cu(II) at equilibrium, it was greater onto OP (51%) than onto SD (17%). After pyrolysis, these percentages were increased: 1.6 and 5.9 times for OP and SD, respectively. The high enhancement obtained for SD (100% retention onto SD-B) was in accordance with the increased specific surface area (from 1.5 to 4.1 m2.g−1) and pH value (from 5.8 to 9.3) of the solid after pyrolysis.

Effect of pH

The effect of pH on Cu(II) adsorption is presented in Fig. 6 for the four solids.

Fig. 6
figure 6

Effect of pH on Cu(II) retention onto SD, OP, SD-B, and OP-B ([Cu2+]0 = 10–4 mol.L−1, solid charge = 1 g.L−1, T = 20 °C, contact time = 2 h), superimposed with the experimental precipitation curve of Cu(II), and with the speciation diagram of Cu (black = Cu2+, green = Cu(OH)+, orange = Cu2(OH)22+, blue = Cu(OH)2(s))

Full size image

The adsorption of Cu(II) by raw materials and their biochar classically increased with increasing pH. In the studied conditions, the maximum retention was equal to 75, 95, 100 and 95% for SD, OP, SD-B, and OP-B, respectively. For comparison, the experimental precipitation curve of Cu(II) is superimposed in Fig. 6. This curve was obtained in the same conditions than adsorption data, but without adding any solid in the batches. This experimental precipitation curve and adsorption data are also superimposed with the speciation diagram of Cu considering the corresponding constants of Cu(OH)+ (log β =  − 7.9), Cu(OH)20 (log β =  − 16.8), Cu(OH)3− (log β =  − 27.5), Cu(OH)42− (log β =  − 40.4), Cu2(OH)3+ (log β =  − 5.8), Cu2(OH)22+ (log β =  − 11.0), Cu3(OH)42+ (log β =  − 22.5), and Cu(OH)2(s) (log Ksp = 9.04) from the constant database IUPAC, and using the HySS program. The experimental precipitation curve was very close to the theoretical curve (blue curve in Fig. 6) and evidenced the precipitation of copper as Cu(OH)2(S) at pH 6.7. Before Cu(II) precipitation, Cu was mainly present as Cu2+ cations (black line in Fig. 6). More than 80% of Cu(II) were retained below the copper precipitation pH value (except for SD at 70%) in accordance with an adsorption process. At higher pH values precipitation cannot be ruled out, and Cu(II) was retained by adsorption and/or precipitation.

In the case of SD, one can notice a decrease in Cu(II) retention above pH 7, certainly due to a partial dissolution of the solid in spite of the preliminary basic treatment. This partial dissolution led to a Cu(II) complexation in solution with released organic compounds instead of copper precipitation, since the percentage of retained copper decreased in basic conditions. This behaviour prevents the use of SD for the removal of Cu(II) in alkaline solutions. The obtained results were in accordance with previous studies reporting that the highest adsorbed amounts onto biochars were generally found between pH 5 and 6 (Bouhamed et al. 2016, Demirbas et al. 2009, Kalavathy et al. 2010).

Cu(II) adsorption capacities

In order to compare the solid removal performances, the influence of the introduced copper concentration on the adsorbed amounts was studied at pH 5.3 in order to obtain adsorption isotherms (Fig. 7). This value far from the Cu(II) precipitation pH value (6.7) was selected (i) to ensure high adsorption capacities according to Fig. 6 and (ii) to prevent the precipitation given that the precipitation pH value is decreased when the initial Cu concentration is increased. The Langmuir and Freundlich classical models were used to fit these isotherms. The adsorption curves relative to OP and OP-B were well fitted with the Langmuir model, whereas the Freundlich model was not suited to describe the data. On the contrary, the adsorbed amounts onto SD and SD-B were well described by the Freundlich model, but not by the Langmuir one (Table 2). In order to compare the performances of the four adsorbents, the adsorption isotherms were thus fitted using the Sips model which is a combined form of Langmuir and Freundlich equations that can simulate both behaviours. It palliates the main drawbacks of these two models: the Langmuir model hypothesizes a homogeneous adsorbent surface with identical adsorption sites, and according to the Freundlich model, infinite adsorbed amounts could be retained on the sorbent surface which is unlikely at solid/liquid interfaces. The Sips model presents the advantage to predict the maximum adsorption capacity of the adsorbent by considering the general case of heterogeneous surfaces (Sips 1948, Sayen et al. 2018). The adsorption data of OP, OP-B, SD and SD-B were very well fitted by this model with determination coefficients R2 comprised between 0.975 and 0.996 (Fig. 7, Table 2).

Fig. 7
figure 7

Equilibrium adsorption isotherms of Cu(II) onto SD, OP, SD-B, and OP-B (symbols), and fitting using Sips model (dotted line), solid charge = 1 g.L−1, T = 20 °C, contact time = 2 h and pH = 5.3

Full size image
Table 2 Model parameters for the equilibrium isotherms relative to Cu(II) adsorption onto OP, SD and their derived biochars using Langmuir, Freundlich, and Sips models, R2 = coefficient of determination
Full size table

The maximum adsorption capacities qmaxS predicted by the Sips model increased in the following order: OP-B (4.9 mg.g−1) < OP (13.1 mg.g−1) < SD (14.9 mg.g−1) < SD-B (50.8 mg.g−1) (Table 2). Thus, the pyrolysis of SD enhanced its removal performance (> 3 times), which was not the case for OP, since its pyrolysis led to a reduction in over half from its adsorption capacities. These behaviours could be explained by an increase in both specific surface area (≈ 3 times) and O/C ratio (≈ 4.5 times) after pyrolysis for SD, and conversely a diminution of the O/C ratio (2 times) and a similar specific surface area for OP and OP-B.

The maximum adsorption capacities of OP, OP-B, SD, and SD-B, were compared to those of previous studies investigating Cu(II) retention in the most similar experimental conditions and reporting the maximum adsorption capacity of the adsorbent obtained by modelling (i.e. in terms of pH value, solid charge, adsorption temperature, and pyrolysis temperature in the case of biochars) but with different biomass-based materials (Table 3). Among the four studied sorbents, SD-B exhibited the highest maximum adsorption capacity which is significantly higher than the sorbents reported in Table 3. This is all the more supported by the fact that the pH value of the other studies was slightly higher than in the present work, which is in favour of an increase in adsorbed amounts.

Table 3 Comparison of Cu(II) maximum adsorption capacities of various adsorbents from biomass
Full size table

Concerning the economical aspect, the cost of the traditional biochar can be estimated from the price at which it is sold at the local markets in Tunisia, around 600 USD/ton, while other commercial sorbents, such as activated carbons, are purchased at more than 1200 USD/ton. Thus, the low price of the traditional biochar, very competitive, may increase the interest for this biomaterial and promote its attractiveness as a sorbent for the removal of pollutants from water.

It can thus be concluded that our SD-B artisanal biochar is a competitive biowaste rendering it a good potential candidate as sustainable, low-cost and efficient adsorbent for copper removal from polluted water.

Conclusion

The aim of this work was to investigate the performance of two raw organic wastes (SD and OP) and their derived biochars (SD-B and OP-B) in terms of copper removal from aqueous solution. Biochars were prepared following a soft and easy pyrolysis process traditionally used by Tunisian farmers (earth mound kiln). The physico-chemical characterization of the studied materials showed that the raw biomasses presented initially comparable properties, in terms of pH, elemental composition and specific surface area. After traditional pyrolysis, the charred materials exhibited alkaline pH, higher specific surface areas and more pronounced aromatic character than the source feedstocks. The study of Cu removal by the raw sawdust, olive pomace and their derived biochars showed that regardless of sorbent nature, the uptake of Cu from solution increased with pH and was maximum for alkaline pH values. At pH = 5.3, using a sorbent dose of 1 g/L and a contact time of 2 h, the maximum adsorption capacities deduced from Sips modelling followed the order OP-B (4.9 mg.g−1) < OP (13.1 mg.g−1) < SD (14.9 mg.g−1) < SD-B (50.8 mg.g−1). These behaviours may be explained regarding the specific surface area of the solids and their O/C ratio. The SD-B performance for Cu(II) removal makes it more competitive than biomass-based wastes previously studied in similar conditions.

However, it is worth noticing that the traditional pyrolysis of biomasses still remains an uncontrolled process which may lead to a variability in the quality of the produced biochars and constitutes a barrier to the application of the traditional pyrolysis at the industrial scale. Thus, as perspectives for this work, the production of biochar from sawdust and olive pomace will be investigated under laboratory controlled conditions in a future study.