3.1 Characterizations of biochar
3.1.1 XRD and FT-IR
The XRD results are shown in Fig. 1(a), BAP, MFBAP, MEBAP and LBAP have similar peak shapes. The diffraction pattern of four biochars shows diffraction peaks at 14.8°(110) and 24°(002), which can be distinguished as carbon with JCPDS Card No. 50–0926 and JCPDSCard No. 46–0944, indicating that they were made of amorphous carbon (Cai et al. 2019), the other peaks of BAP belonged to the crystal form of whewellite. In addition, by comparing the PDF card (JCPDS Card No.391346), the new characteristic peaks of MFBAP, MEBAP and LBAP at 30.2°(220), 35.5°(311), 43.2°(400), 57.3°(511) and 62.9°(440) were discovered to be the crystal form of γ-Fe2O3 (Xiao et al. 2019), which indicated that γ-Fe2O3 particles were embedded in the pristine biochar.
The FT-IR patterns of four biochars were shown in Fig. 1(b). It can be seen that the biochar, before and after modification, had the homologous peaks shape except for BAP which not possessed a peak at 582 cm− 1 that was assigned to stretching vibration of Fe-O (Cai et al. 2019). Besides, seven main functional groups are contained on it, such as -OH stretching vibration of carboxy and hydroxyl group, C = O bending vibration, aromatic C = C, N-C = O stretching vibration, -COOH flexural vibration or conjugated C-N stretching vibration, alcoholic CO- stretching vibration (Jiang et al. 2017, Li et al. 2016, Sun et al. 2011) for the 4 samples which characteristic absorption peaks are at 3384 cm− 1, 2858 cm− 1, 1628 cm− 1, 1452 cm− 1, 1368 cm− 1, 1058 cm− 1. Furthermore, it is convincing that adding FeCl3·6H2O is the reason why the surface of biochar can form iron oxide which is consistent with the XRD results.
3.1.2 SEM
Figure 2(a) shows that BAP displayed neat and porous structures with no Fe existing on its smooth surface. Unlike BAP, MFBAP, MEBAP and LBAP present a rough surface in which many irregular particles are uniformly contained in Fig. 2(b-c). These particles are mainly composed of γ-Fe2O3, which has been confirmed by previous XRD results. In addition, MFBAP and MEBAP show a circular structure and under the same scale, the diameter of their circular structures are much larger than BAP, which is also far from the structures of BAP. Hence, it can be proved that the introduction of HAD and Melamine can dramatically alter the surface and porous structure of pristine biochar so that the porous structure can transform and extend into a kind of circular structure on which the surface is covered with a myriad of γ-Fe2O3. The possible reason is that adding HDA and ME can increase the pH of the reaction solution, which makes the decomposition of lignin or cellulose in AP more thorough during the HTC process (Wang et al. 2018), and thus leaving a large number of circular structures. In addition, it can be observed notably from Fig. 2(d) that the skeleton structures of the LBAP were conspicuous. However, its gap between the skeleton structures is blocked, which this phenomenon was explained that L-Glutathione could hinder the formation of the circular skeleton structures and blocked the pore channel during the HTC process, which is why the SBET and pore volume (Table 2) decreased after being modified by coating L-Glutathione onto pristine biochar. The different structures of the four materials account for the dissimilarity of BAP, MFBAP, MEBAP, LBAP for adsorption efficiency of Cr(VI) in a way.
In addition to studying the typical SEM appearance of four kinds of biochar, the surfaces before and after adsorption were also researched. As is shown in Fig. 2(e-h), it is found that the surface of four kinds of biochar has changed dramatically before and after adsorption. The surface of BAP has become rough and uncomely, indicating that its surface had been seriously damaged after adsorption. Followed by MFBAP, MEBAP and LBAP, despite their surfaces having also been damaged to some extent, it can be seen from SEM images that their surfaces are much less damaged than BAP. Therefore, it can be said that the introduction of amines could enhance the external resistance of biochar.
The elemental composition of the surface of BAP and MFBAP after adsorption was analyzed by SEM-EDS in Fig.3. It can be seen that the surface of MFBAP is loaded with a ton of Cr(VI) ions, and its load scale and quantity are more tremendous than that of BAP. The Cr(VI) loaded on the surface of BAP accounts for 19.24%, which is lower than 21.44% of that MFBAP, demonstrating that the Cr(VI) loading performance of MFBAP has been strengthened. In addition, the N content of MFBAP also increased by 3.05% relative to BAP, which also shows that the introduction of HDA increases a large number of N-containing functional groups on the surface of MFBP. The Fe peak in the EDS spectrum of MFBAP further proves that Fe is successfully introduced into biochar by HTC.
3.2 BET
Table 2
The pore structure parameters of BAP, MFBAP, MEBAP and LBAP.
Sample
|
SBET
(m2/g)
|
Average pore size
(nm)
|
Average mesoporous volume
(nm)
|
Pore volume
(cm3/g)
|
BAP
|
5.6308
|
17.5925
|
19.8232
|
0.0248
|
MFBAP
|
11.8252
|
30.6540
|
35.1388
|
0.0906
|
MEBAP
|
7.0464
|
17.2099
|
13.7804
|
0.0303
|
LBAP
|
3.4583
|
15.7163
|
27.4638
|
0.0136
|
Their N2 adsorption/desorption isotherms are displayed in Fig. 4(a). It can be seen from Table 2 that the four biochars all show a pseudo-type IV curve. The SBET of the biochar were 5.6308 m2/g (BAP), 11.8252 m2/g (MFBAP),7.0464 m2/g (MEBAP), 3.4583 m2/g (LBAP), respectively, which means HAD and ME modification can improve SBET. The outcome may owe to the process of manufacturing modified biochar, which was produced by mixing biomass with FeCl3 and amines, and the pore structures of pristine biochar were changed during the HTC process. Among them, MFBAP is the highest in terms of SBET, average pore size, average micropore volume as well as pore volume. The SBET of MFBAP is almost twice that BAP and the pore volume is almost four times. More importantly, it can be seen that the SBET and pore volume of LBAP is the lowest, which was again proved that L-Glutathione can block the pores of biochar, reducing its SBET and pore volume. These results all were corresponding with the previous results of SEM images partially. In this study, the SBET of the four biochar is not large, but previous studies (Yi et al. 2019a) have shown that the adsorption capacity of magnetic biochar to degrade Cr(VI) does not depend on its SBET. In other words, the SBET of magnetic biochar is not the key factor to determine the adsorption capacity, which means that the removal mechanism of Cr(VI) onto MFBAP relies mostly on functional groups.
3.3 Adsorption kinetics and isotherms
Table 3
Kinetic parameters of Cr(VI) adsorption by BAP, MFBAP, MEBAP, LBAP.
Sample
|
qe
(mg/g)
|
Pseudo-first-order
|
Pseudo-second-order
|
Elovichs
|
R2
|
Qe
(mg/g)
|
\({k}_{1}\times {10}^{3}\)
|
R2
|
Qe
(mg/g)
|
\({k}_{2}\times {10}^{3}\)
|
R2
|
β𝑬
|
α𝑬
|
BAP
|
42.4740
|
0.9563
|
40.5171
|
4.97
|
0.9763
|
46.2992
|
0.1405
|
0.9559
|
0.1405
|
1.5309
|
MFBAP
|
80.5804
|
0.9315
|
71.7952
|
9.31
|
0.9749
|
80.3924
|
0.1669
|
0.9889
|
0.0768
|
4.6688
|
MEBAP
|
62.2600
|
0.9669
|
61.6264
|
2.91
|
0.9765
|
74.8279
|
0.0415
|
0.9118
|
0.0986
|
1.3514
|
LBAP
|
55.6647
|
0.9179
|
50.0845
|
6.98
|
0.9634
|
55.6755
|
0.1886
|
0.9851
|
0.1116
|
2.6622
|
Table 4
Isotherm parameters of Cr(VI) adsorption by BAP, MFBAP, MEBAP, LBAP.
Sample
|
qe
(mg/g)
|
Langmuir
|
Freundlich
|
R2
|
Qe
(mg/g)
|
kL
|
R2
|
n
|
kF
|
BAP
|
42.4740
|
0.9921
|
46.1405
|
0.6031
|
0.9522
|
8.6398
|
25.0175
|
MFBAP
|
80.5804
|
0.8979
|
90.6798
|
0.3817
|
0.8719
|
6.4624
|
39.4584
|
MEBAP
|
62.2600
|
0.9476
|
66.5468
|
0.7089
|
0.8917
|
8.2141
|
34.6154
|
LBAP
|
55.6647
|
0.9231
|
59.1987
|
0.8952
|
0.8806
|
9.3205
|
33.2846
|
The study of adsorption kinetics and isotherms are of great significance to determine the adsorption efficiency in practical application (Tao et al. 2018). The pseudo-firs-order, pseudo-second-order models and elovich were used to study the adsorption kinetics of Cr(VI) on biochar, and the adsorption isotherms of Cr (VI) on biochar were simulated by Langmuir and Freundlich model. The three kinetics models (Wei et al. 2020) and two isotherm models (Zhang et al. 2020) are described by equations (1)-(5):
Pseudo-first-order: \({\text{q}}_{\text{t}}={\text{q}}_{\text{e}}\left(1-{\text{e}}^{-{\text{k}}_{1}\text{t}}\right)\) (1)
Pseudo-second-order: \({\text{q}}_{\text{t}}=\frac{{\text{k}}_{2}{{\text{q}}_{\text{e}}}^{2}\text{t}}{1+{\text{k}}_{2}{\text{q}}_{\text{e}}\text{t}}\) (2)
Elovich: \({\text{q}}_{\text{t}}=\frac{1}{{{\beta }}_{\text{E}}}\text{ln}{{\alpha }}_{\text{E}}{{\beta }}_{\text{E}}+\frac{1}{{{\beta }}_{\text{E}}}\text{ln}\text{t}\) (3)
Langmuir: \({\text{q}}_{\text{e}}=\frac{{\text{q}}_{\text{m}}{\text{k}}_{\text{L}}{\text{c}}_{\text{e}}}{1+{\text{k}}_{\text{L}}{\text{c}}_{\text{e}}}\) (4)
Freundlich: \({\text{q}}_{\text{e}}={\text{k}}_{\text{F}}{{\text{c}}_{\text{e}}}^{\frac{1}{\text{n}}}\) (6)
Where \({\text{q}}_{\text{e}}\) and \({\text{q}}_{\text{t}}\) (mg/g) are the adsorption amounts of Cr(VI) by biochars at equilibrium and time t, respectively. \({\text{k}}_{1}\) and \({\text{k}}_{2}\) are the corresponding adsorption rate constants of the pseudo-first-order (h− 1) and pseudo-second-order (g·mg− 1·h− 1) models, respectively. \({{\alpha }}_{\text{E}}\) and \({{\beta }}_{\text{E}}\) are the initial adsorption rate (mg·g− 1·h− 1) and desorption rate (g·mg− 1) constants, respectively. \({\text{q}}_{\text{m}}\) (mg/g) is the maximum monolayer adsorption capacity, \({\text{k}}_{\text{L}}\) and \({\text{k}}_{\text{F}}\) is the Langmuir and Freundlich isotherm constant, respectively. n is the measure of adsorption intensity.
The experimental results of kinetics are shown in Fig. 4(c). The adsorption of Cr(VI) by four kinds of biochar is rapid, which reached 59.45% (MFBAP), 45.73% (BAP), 35.26% (MEBAP) and 51.29% (LBAP) within 120 min. When adsorption reached 800 min, the adsorption capacity of the four biochars reached equilibrium. The reason why the adsorption equilibrium took a long time is that the active surface sites of biochar were occupied by Cr(VI), which made the binding sites on the material surface decrease rapidly (Zhang et al. 2015a). As illustrated in Table 3, it is the MFBAP that possessed the maximum adsorption capacity of 80.58 mg/g, which is higher than that of the BAP (42.47 mg/g) and that of the MEBAP and LBAP (62.26 mg/g and 55.66 mg/g). The adsorption data were further fitted by the pseudo-first-order model and the pseudo-second-order model. The fitting degree of kinetic was judged by the correlation coefficients (R2). The pseudo-second-order model is better fitted to BAP (R2 = 0.9763) and MEBAP (R2 = 0.9765), indicating that the adsorption of Cr(VI) by BAP and MEBAP is mainly a chemical adsorption process (Liu et al. 2017). Meanwhile, the Elovich model was found to better fit MFBAP (R2 = 0.9889) and LBAP (R2 = 0.9851), demonstrating that the MFBAP and LBAP have a variety of adsorption mechanisms for Cr(VI).
The adsorption isotherm parameters and the fitting results of the two models are shown in Table 4 and Fig. 4(d). At low concentration, the adsorption capacity of the four biochar increased extremely and as the concentration gradually raised, their adsorption capacity approached saturation. Besides, it could be negligible for the residual Cr(VI) concentration after adsorption by MFBAP as the initial concentration in the solution is relatively low. Langmuir and Freundlich are the most common models used to describe the adsorption characteristics between adsorbent and Cr(VI). The Langmuir correlation coefficients of the BAP, MFBAP, MEBAP and LBAP (R2 = 0.9921, 0.8979, 0.9476, 0.9231) are higher than those of Freundlich (R2 = 0.9522, 0.8719, 0.8917, 0,8806) and the values of \({q}_{m}\) calculated by the Langmuir isotherm are quite close to experimental values \({q}_{e}\), indicating that Langmuir has a better fitting effect on the four biochar and the adsorption process of Cr(VI) onto four biochars was monolayer adsorption with uniform binding sites, equal adsorption energies and few interactions between adsorbed species (Liabc et al. 2019). Hence, Cr(VI) is mainly adsorbed to the surface of biochar through various chemical reactions.
3.4 Effect of pH
pH can affect the speciation of Cr(VI) and the surface charge of the biosorption in aqueous solution, therefore it has a vital impact on Cr(VI) removal (Qiua et al. 2020). The effect of the initial pH of the solution on adsorption is shown in Fig. 4(b). It can be seen that it is more efficient for MFBAP to remove Cr(VI) than BAP and the adsorption capacity of BAP and MFBAP for Cr(VI) becomes lower gradually with the increase of pH value (2–8). When pH = 2, the adsorption capacity reaches the maximum value, which are 80.58 mg/g and 43.21mg/g respectively. Meanwhile, the minimum adsorption capacity was observed at pH = 8 (13.89 mg/g and 11.70 mg/g). Research (Liabc et al. 2019) found that Fe2O3 can be protonated as FeOH+ that could be the active centers of Cr(VI) adsorption under lower initial pH, which is beneficial to adsorption of biochars.
These results proved that high pH value is unfavorable to the removal of Cr(VI) by MFBAP and BAP.
3.5 Effect of the coexisting cations.
The actual wastewater exists various kinds of ions, therefore the effects of different anions (Cl−, NO3−, SO42−, HPO42−) and cations (Cu2+, Ca2+, Ni2+, Cd2+) on the adsorption capacity of MFBAP were studied in Fig.4(e-f). Several anions had little effect on the adsorption effect, however, cations exerted great influence on adsorption capacity. The disturbance capability of the coexisting cations on the adsorption capacities of MFBAP was as follows: Cd2+>Cu2+>Ca2+>Ni2+ due to that their affinityto compete for the active site of the MFBAP is related to the valence, hydration state and chemical structure of ions (Dong et al. 2018). For the above results, it can also be said that MFBAP has a partial adsorption abili on these cations in a way due to the competition mechanism, especially had a high affinity for Cd2+, which lays a certain foundation for the study of Cd2+ adsorption by HAD-modified magnetic biochar. At the same time, a previous study has shown that the reason why Ca2+ can affect the adsorption is that Ca2+ can occupy the active adsorption sites on the surface of biochar. The above results show that MFBAP has an anti-ions interference ability to some extent.
3.6 Regeneration performance of biochar
Regeneration performance of adsorbent materials is very important for evaluating the feasibility of practical and large-scale applications (Fomina &Gadd 2014). The recycling experiment ofMFBAP after five times of adsorption/desorption was studied. The results are shown in Fig. 5.It can be seen from the figure that the adsorption capacity of MFBAP after regeneration decreased a lot, and the final adsorption capacity is still 33.68mg/g after 5 cycles, which is better than many other adsorbents reported in the literature (Table 5), demonstrating that the adsorbent has preferable adsorption potential.
Table 5
Comparison of Cr(VI) adsorption capacities with other adsorbents.
Raw material
|
Adsorption capacity (mg/g)
|
reference
|
Eichhornia crassipes
|
120
|
(Zhang et al. 2015a)
|
Astragalus membranaceus
|
23.85
|
(Shang et al. 2016)
|
Enteromorpha prolifera
|
88.17
|
(Chen et al. 2018)
|
Phoenix tree leaves
|
27.2
|
(Shi et al. 2018)
|
Melia azedarach wood
|
25.27
|
(Zhang et al. 2018)
|
Pine sawdust
|
25.25
|
(Li et al. 2019)
|
Loofah
|
30.14
|
(Xiao et al. 2019)
|
Steel pickling waste liquor
|
43.12
|
(Yi et al. 2019a)
|
Water hyacinth
|
43.48
|
(Yu et al. 2018)
|
Bamboo
|
38
|
(Huang et al. 2018)
|
Sugarcane bagasse
|
43.12
|
(Yi et al. 2019b)
|
This study
|
81.31
|
-
|
3.7 Cr(VI) removal mechanisms
In order to understand the valence state of elements on the surface of biochar and whether the functional groups on the surface of biochar are involved in the reaction, and mostimportantly, the mechanisms of Cr(VI) adsorption, XPS was used to study the chemical changes of the surface of biochar before and after Cr(VI) adsorption and explain the mechanism of Cr(VI) removal. The results were displayed in Fig.6. A new double peak appears in the spectrum after adsorption, which was confirmed to be the photoelectron peak of Cr 2p, indicating that the phenomenon of Cr(VI) ions adsorption onto MFBAP took place. The peaks at 585.20 and 574.5 eV attributed to the Cr 2p1/2 and Cr 2p2/3 of Cr(III), while the peaks at 585.4 eV and 578.2 eV are ascribed to the Cr 2p1/2 and Cr 2p2/3 of Cr(VI), respectively, indicating that on the surface of MFBAP after adsorption, there are two forms of chromium: Cr(III) and Cr(VI), and therefore it is the biochar that converted Cr(VI) to Cr(III) by reduction.The other significant peaks represent C 1s (83.25%), O 1s (12.73%) and N 1s (4.02%) before adsorption, and C1s (80.06%), O 1s(17.11%), N 1s (2.84%) after adsorption.Comparing the XPS spectra before and after adsorption, it can be found that the N/C atomic ratio decreases after Cr(VI) was adsorbed by biochar, indicating that the amino groups participated in the reaction and were consumed in the adsorption process. In the XPS spectrum of N 1s before adsorption, the peak at 397.70, 396.5 and 398.4 eV were identified as N-C, -NbN- and N-H respectively.Apparently, the three peaks of N 1s became weaker after adsorption, which means the amino-functional groups of MFBAP were consumed after participating in the degradation of Cr(VI) adsorption. In other words, the amino-functional groups are the significant adsorption mechanism of Cr(VI).Simultaneously, Cr(VI) exists as the form of HCrO4− and Cr2O72− under acidic conditions (Saha &Orvig 2010, Zhang et al. 2017), under which the amino-functional groups -NH2 could be protonated to -NH3+ as well as -OH and -COOH, which also could be protonated to -OH2+ and -COOH2+. Therefore, part of the process for the mechanisms of Cr(VI) can be portrayed as follows (Mohan &Pittman 2006):

Except for these three peaks, Fe 2p can also be observed in XPS survey spectra, which demonstrated the preparation of magnetic biochar was successful. The peaks of 284.8 eV and 286.06 eV, 284.2 eV and 286.7 eV were assigned to the C-C, C-O, C = C and C-N bonds(Li et al. 2019, Zhang et al. 2015a), respectively. Compared with C1s before and after adsorption, it was found that the C-O peak became weak, while the C = C peak increased, suggesting that functional groups participated in the process of Cr(VI) adsorption. In fact, some active functional groups, such as C = C and C-OH, which can be called Lewis base, can provide electrons for Cr(VI) in the adsorption process to reduce into Cr(III) (Hai et al. 2014, Zhang et al. 2015b). Therefore, the reduction process of Cr(VI) should be as follows:
$${\text{H}\text{C}\text{r}\text{O}}_{4}^{-}+7{\text{H}}^{+}+3{\text{e}}^{-}\to {\text{C}\text{r}}^{3+}+4{\text{H}}_{2}\text{O}$$
\({\text{C}\text{r}}_{2}{\text{O}}_{7}^{2-}+14{\text{H}}^{+}+6{\text{e}}^{-}\to\) 2\({\text{C}\text{r}}^{3+}+7{\text{H}}_{2}\text{O}\)
As is shown in Scheme. 2, the main interaction mechanisms of the removal of Cr(VI) include four key steps: (1) Complexation, where functional groups on biochar can be protonated at low pH to form positively charged sites (-NH3+, -OH2+, -COOH2+), which Cr(VI) can be easily bound to the electropositive surface of biochar and form complexes with the functional groups. (2) Electrostatic interaction, where the surface of the modified biochar material has a positive charge (H+) when the pH is low, and Cr(VI) in the valence state of negative ions can be trapped on the surface of the adsorbent material. (3) Reduction, where Some active functional groups can act as Lewis bases to provide electrons for the reduction of Cr(VI). (4) Ion exchange, where Cr(VI) was adsorbed onto active sites of the MFBAP.