Copper removal from aqueous solution using biochar: Effect of chemical activation

2.1 Solution preparation and characterization

Copper(II) was the adsorbate studied in all experiments. In this case, two different types of aqueous solutions containing this metal were used. First, synthetic solution was prepared by dissolving CuNO₃·3H₂O (p.a.), (Lach-Ner, CZE) into a previously prepared background electrolyte (0.01 M NaNO3). Second, soil solution was obtained from Cu-contaminated soil by 0.01 M CaCl₂ leaching procedure overnight. This soil originated from the vicinity of Le Palais copper refinery (Limoges, France) and all characteristics of this soil solution are listed in Table 1. Additionally, such prepared synthetic and soil solutions were analysed for: (i) pH value; (ii) all metal concentrations using ICP-OES (Agilent Technologies 700 series, USA), (iii) dissolved organic carbon (DOC) using TOC-L CPH Analyser (Shimadzu, Japan), (iv) available nutrients using ion chromatograph ICS 90 (Dionex, USA) and flame atomic absorption spectrometer F-AAS (Varian 280FS, Australia), respectively.

2.2 Biochar preparation and characterization

Local brewers draff obtained from the brewery at the Czech University of Life Sciences Prague was used due to its low cost and large amounts produced (Šillerová et al., 2013), was air dried overnight and consequently pyrolysed at 650 °C under 100 mL min⁻¹ nitrogen flow rate. Such high temperature was chosen for the biochar preparation in order to get the brewers draff fully carbonized with low volatile amounts and high fixed carbon values (Agirre et al., 2013). Additionally, the yield of final char was measured at the temperature rate of 5 °C min⁻¹ using thermogravimeter Stanton Redcroft TG 750. The resulting biochar (BC) was then cooled overnight under the same nitrogen flow rate (as before). Such prepared pyrolysed product was ground, sieved to less than 0.25 mm, washed by ultra-clean water MilliQ Integral (Merck Millipore Corp., USA) and dried at 105 °C. Biochar activation was implemented according to Regmi et al. (2012), when 4 g of BC was mixed with 1000 mL of 2 M KOH solution and stirred for 1 h. This solution was then filtered using 0.45 μm nylon filters (Whatman Ltd, USA). In the next step, filtered biochar was mixed with ultraclean water and the pH value of both biochars was adjusted using 0.1 M NaOH/HCl. Next, the biochar (BCact) was again filtered and dried overnight at 105 °C.

Surface area, total pore volume and pore distribution were determined using the BET surface and pore analyser SA 3100 (Beckman Coulter, USA). The BET surface area was measured by nitrogen adsorption isotherms at 77 K, total pore volume was determined from adsorbed volume of nitrogen at pressure p = 1 atm and pore distribution was then calculated from the desorption isotherm. Elemental composition was measured in triplicate by CHNS-O Analyser Flash 1112 (Thermo, USA). Moreover, surface functional groups of brewers draff, BC and BC_act were analysed using FTIR Nicolet 6700 analyser connected with a microscope Continuum (Thermo-Nicolet, USA). Point of zero charge (pH_zpc) was determined for both biochars according to Fiol and Villaescusa (2009) using potentiometric mass titration (PMT) and the immersion technique (IT). Additionally, SEM analyses of both biochar samples were also implemented using a scanning electron microscope JEOL JSM-7401F Fesem (USA).

2.3 Batch experiments

First, kinetic experiments were implemented in order to establish equilibrium for the subsequently prepared sorption/desorption experiment. More precisely, synthetic Cu-solution at 0.1/0.25 mM was mixed with biochar samples at the ratio 1:500 (w/v). Suspensions prepared in this way were shaken (at 250 rpm), centrifuged, filtered (0.45 μm nylon filter) and the obtained supernatants were analysed at regular time steps (1, 2, 4, 8, 12, 24 and 48 h) in order to determine the Cu concentration. Next, 20 mL of synthetic solution (with Cu concentration steps of 0.02–2 mM) was added into 0.04 g of both prepared biochar samples (BC × BC_act) respectively, to establish their sorption maxima. Following the previously implemented kinetic experiments, the suspensions of the solution with the biochar samples were shaken for 24 h in order to reach equilibrium. Additionally, desorption experiment was also carried out. Background electrolyte (0.01 M NaNO₃) was used to obtain information about the desorbed volume of copper from the previously Cu-saturated biochar samples.

2.4 Column experiments

The retention column experiment used two identical glass columns (Bio-rad Laboratories, USA) of 13 cm length and 2 cm in internal diameter which were filled with 3 g of BC and BC_act, respectively. Such prepared columns were gravitationally leached from the top for 24 h under a constant flow rate (5 mL min⁻¹), which was maintained by a peristaltic pump EP-1 Econo Pump (Bio-rad Laboratories, USA). Two types of aqueous solutions were used for the retention test mentioned previously: (i) soil leachate; and (ii) synthetic Cu-solution, the concentration of which was adjusted to equal the Cu concentration in the soil leachate (approx. 0.02 mM). The pH value of the synthetic solution has also been adjusted to equal that of the soil leachate.

Both leachates were then automatically sampled over a 24 h period of time (at gradually increasing time steps) during the whole retention test, using the fraction collector FC 204 (Gilson, USA). Finally, obtained samples were analysed for Cu and DOC concentration in the leachate. Furthermore, other metals and all (in)organic anions/cations in the soil solution were also analysed.

2.5 Copper speciation model

The Minteq.v3.dat thermodynamic database (derived from the code MINTEQ+, version 3.0, released by the U.S. EPA in 1999) was used for all calculations. The initial soil solution (see Table 1; obtained by 0.01 M CaCl₂) with all characteristics (DOC; all (in)organic anions and cations as well as metal concentrations) was used for the calculation. Additionally, Cu speciation was also calculated at the final time step of the column retention test for both the BC and BC_act variants, respectively.

3.1 Thermogravimetric analysis of brewers draff

The thermogravimetric mass loss (TG) and its derivative curve (DTG) are shown in Fig. 1. Initial mass loss (approx. 10% reduction) ranging from 25 to 150 °C was mainly due to moisture evaporation. Next, devolatilization and biochar thermal decomposition were responsible for other mass losses (approximately 80% reduction) from 300 °C up (Gil et al., 2012; Yi et al., 2013). In more detail, there were two obvious peak values for the DTG curve of −0.42 and 0.28% °C⁻¹ at 295 and 580 °C, respectively. In summary, total mass loss was approximately 90%, which demonstrated a very low yield of biochar during pyrolysis at 650 °C of brewers draff. Due to this fact a different char preparation would be more suitable, such as hydrothermal carbonization (HTC), hence the term hydrochar (Regmi et al., 2012; Titirici et al., 2007; Xue et al., 2012).

Close

Figure 1

Thermogravimetric mass loss (TG) and derivative thermogravimetric mass loss (DTG) curves of used brewers draff.

Export to PPT

3.2 Chemical activation by 2M KOH

According to Regmi et al. (2012) cold alkali solution (2 M KOH) was chosen for the chemical activation of prepared biochar to enhance its porosity and clean up blocked pores mainly by tar particles. Additionally, surface area of activated biochar was expected to increase which might have enhanced copper removal from both aqueous solutions. All characteristics of activated biochar compared with biochar are shown in Table 2; Figs. 2–4 and the results are discussed in the next subsection in more detail.

Close

Figure 2

SEM images of biochar (a) and chemically activated biochar (b). 1 μm.

Export to PPT

Close

Figure 3

Point of zero charge (pH_zpc) using potentiometric mass titration (PMT) for biochar (BC) (a), activated biochar (BC_act) (b) and immersion technique (IT) for biochar (BC) (c) and activated biochar (BC_act) (d).

Export to PPT

Close

Figure 4

FTIR spectrum of (i) brewers draff before pyrolysis; (ii) biochar (BC), and (iii) activated biochar (BC_act).

Export to PPT

3.3 Characterization of prepared biochar samples

3.4 Batch experiments

First, the effect of contact time on Cu sorption onto both biochar samples was investigated. Fig. 5 provides information about the reached plateau of Cu sorption (removal) from the synthetic solution on both biochar samples. The 24-h period was chosen for practical reasons as the equilibrium in this case. Furthermore, as stated by the paper of Chen et al. (2011) the equilibrium should be expected as early as 2 h (see also Fig. 5). Sorption efficiency (E) of both tested biochar samples at different time steps was calculated according to Eq. [1].

Close

Figure 5

Influence of contact time of used biochar (biochar BC and activated biochar BC_act) with Cu synthetic solution at two different concentrations (0.10 and 0.25 mM).

Export to PPT

Next, following the kinetic results, sorption capacity of both biochar samples to remove Cu from synthetic solution was determined at equilibrium state (t = 24 h). The sorption efficiency of BC and BC_act was evaluated using Freundlich [4] and Langmuir [5] equations.

Close

Figure 6

Sorption isotherms (a) and pH changes (b) during Cu sorption from synthetic solution for biochar (BC) and activated biochar (BC_act).

Export to PPT

Furthermore, decreases in the solution pH at a higher Cu concentration during the sorption experiment can be attributed to the hydrolysis effect (Ponizovsky et al., 2007; Mouta et al., 2008). The pH of both experimental solutions changed significantly due to the pH-buffering effect of both biochar samples (pH = 7.00).

3.5 Column experiments

In the final experimental step, Cu removal from aqueous solution was also measured over a period of time using the column retention test. As we had expected, the column studies showed that the Cu removal was more effective in the synthetic solution (Fig. 7a) solely due to Cu sorption over the period of time. In comparison with the synthetic solution, the effect of Cu removal from the soil solution was significantly reduced. For the synthetic solution, both BC and BC_act were fully saturated by Cu (when the ratio c/c_i ≈ 1) from t = 2160 min (Fig. 7a), whereas in the soil solution both biochars were fully saturated from t = 960 min.

Close

Figure 7

Breakthrough curves of copper retention for synthetic (a) and soil (b) solution over the period of time.

Export to PPT

Breakthrough curves with an early increase in the c/c_i ratio were observed for the biochar which had not been previously treated by 2 M KOH, for both synthetic and soil solutions, respectively, which is in agreement with the previously mentioned fact about better sorption of BC_act. In more detail, the activated biochar breakthrough curve point (BTCP; the final time step for the maximum sorption efficiency of biochar, when the ratio c/c_i ≈ 0) was obvious at t = 780 min for the synthetic solution and at t = 80 min for the soil solution, respectively. On the other hand, BC started to saturate sooner hence BTCP was already at t = 50 and 10 min for the synthetic and soil solutions, respectively. Such dramatic differences of BTCP between BC and BC_act are most likely caused by free micro-pores (<6 nm) in the BC_act. More precisely, these pores, which are created predominantly during high temperature pyrolysis (Tsai et al., 2012), were formed only in the case of BC_act due to the leaching out of the tar particles from these places using cold alkali solution stirring (discussed in subsections 3.2 and 3.3.1). Due to this fact, we assume that Cu was trapped inside these free micro-pores thus replacing the previously removed tar particles. Additionally, concentration of the dissolved organic carbon (DOC) was also measured over the period of time for both synthetic and soil solutions, respectively. All results show that DOC was unchanged over the period of time for both variants as well as for both solutions (data not shown).

Additionally, copper speciation of the initial soil solution was compared with the Cu-species in the soil solution, which was leached through both biochar samples (at the final time step). Table 5 shows Cu²⁺ species decreasing in the soil solution which was in contact with both BC and BC_act, respectively. On the other hand, Cu-species fixed on the organic matter significantly increased in this final soil solution. These Cu species in the final soil solution were thus changed mainly due to a slight continuous pH increase. Furthermore, minimum changes in Cu species have been registered for the BC_act in comparison with the BC.

Finally, desorption of Cu at the end of the column retention test was also implemented in the synthetic solution. More precisely, the Cu-fully saturated biochar samples (BC and BC_act) were shaken overnight using 0.01M NaNO₃. Obtained results showed minimum desorption of Cu from both biochar samples (0.48 ± 0.03 and 0.51 ± 0.05 mg L⁻¹). We can assume from these results that Cu is strongly bounded to the biochar surface. This is in agreement with other authors, i.e., Uchimyia et al. (2010), (2011), Trakal et al. (2011), Regmi et al. (2012).