Effects of excessive impregnation, magnesium content, and pyrolysis temperature on MgO-coated watermelon rind biochar and its lead removal capacity

61 Abstract MgO-coated watermelon rind biochar (MWRB) is a potentially highly-effective waste-derived material in environmental applications. This research aims to provide valuable insights into the optimization of the production of MWRB for superior environmental performance. It was found that the Mg content of the MWRB could be easily controlled by adjusting the Mg/feedstock mass ratio during excessive impregnation. The BET surface area was found to first increase and then decrease as the Mg content of the MWRB (produced at 600 °C) increased from 1.52% to 10.1%, with an optimal surface area of 293 m 2 /g observed at 2.51%. Similarly, an optimum pyrolysis temperature of 600°C was observed in the range of 400-800 °C for a maximum surface area of the MWRB at a fixed Mg/feedstock ratio of 0.48% (resulting in MWRBs with Mg contents of 1.89-2.51%). The Pb removal capacity of the MWRB (produced at 600°C) increased with increasing Mg content, with a greatest Pb removal capacity of 558 mg/g found for the MWRB with the highest Mg content (10.1%), an improvement of 208% over the 181 mg/g Pb removal capacity of unmodified WRB produced at 600°C. The Pb removal capacity of the MWRB (produced with 1.89-2.51% Mg) was also discovered to increase from 81.7 mg/g (at 400 °C) to 742 mg/g (at 700°C), before dropping to 368 mg/g at 800 °C. These findings suggest that the MWRB can be more efficiently utilized in soil and water remediation by optimizing its synthesis conditions.

Heavy metals represent a widely distributed environmental concern (e.g., in soil and water) due to their presence in industrial emissions and consumer products (O'Connor et al., 2018a;Wang et al., 2019c;Zhang et al., 2018a). For example, lead (Pb) is a problematic soil and water contaminant worldwide, posing a significant hazard to human health, especially affecting the development of brain and nervous system children (Shen et al., 2018b;Wang et al., 2019b;Zhang et al., 2019). It was estimated in 2013 that 0.6% of the world's disease and approximately 853000 deaths were caused by Pb pollution (Shi et al., 2019). The development of engineered reactive materials is garnering much attention, as an ever improving approach to remediation (Kang et al., 2019;O'Connor et al., 2018b;Zhang et al., 2017b), in particular, biochar is seeing a growing interest in its use for heavy metal removal owing to its greener credentials in comparison to other energy-intensive technologies (e.g., electrokinetic remediation and soil washing) (Song et al., 2019).
Biochar, a carbon rich material, can be produced from various biomass feedstocks via pyrolysis in an oxygen limited environment. Because of its high surface area, alkalinity, abundant oxygen-containing functional groups, and high cation exchange capacity (CEC) (Ahmad et al., 2014;Inyang et al., 2016;Mohanty et al., 2018), biochar can be used as an effective sorbent for many heavy metals. Furthermore, biochar can be used to mitigate climate change by gathering carbon from atmosphere, or to enhance soil fertility by improving nutrient retention and moisture holding capacity (Tang et al., 2013;Yang et al., 2018). Therefore, biochar is increasingly regarded as a sustainable solution for water and soil remediation (Hou et al., 2018;O'Connor et al., 2018c;Song et al., 2019).
Recently, MgO modified biochar, or MgO-biochar composite, has been found to be promising in environmental applications (Ling et al., 2017;Shen et al., 2019b;Zhang et al., 2012). For example, MgO coating was observed to significantly enhance the adsorption capacity of Cr 6+ for sugarcane harvest residue biochar from 20.79 mg/g to 54.64 mg/g (Xiao et al., 2018). Likewise, MgO-coated corncob biochar was found to effectively immobilize Pb to reach Toxicity Characteristic Leaching Procedure (TCLP) regulatory levels when applied to a soil washing residue, while the original corncob biochar was ineffective (Shen et al., 2019b).
In order to coat MgO on biochar, MgCl 2 is typically applied to the feedstock as a pre-treatment before pyrolysis (Jellali et al., 2016;Jung and Ahn, 2016;Li et al., 2016;Ling et al., 2017;Shen et al., 2019b). Two methods have been used for the pre-treatment: "ultrasonic impregnation" and "excessive impregnation" (Wang et al., 2015c;Zhu and Zhang, 2008;Zhu et al., 2015). With the ultrasonic impregnation method the remaining MgCl 2 may produce large aggregates during thermal decomposition under pyrolysis, which may result in the heterogeneity of magnesium (Mg)-containing minerals on biochar and the blockage of biochar pores after synthesis (Shen et al., 2019b). In the excessive impregnation method the excess MgCl 2 solution is removed by filtration after thorough mixing, therefore this approach has been favored. Excessive impregnation may improve in the homogeneity of MgCl 2 mixing and the uniform distribution of Mg-containing minerals in biochar after pyrolysis.
However, the amount of MgCl 2 adsorbed on the feedstock remains unclear, bringing difficulty in controlling the final Mg content in biochar.
The enhancement of MgO coating for biochar mainly involves enhanced surface area and pore structure fashioned by MgCl 2 activation, and increased buffering capacity owing to the MgO content (Li et al., 2016;Shen et al., 2018a). These enhanced 6 properties result in biochar with better performance in environmental applications. It was observed that MgO coating (Mg content of 11.96%) greatly increased the surface area of corncob biochar from 0.07 to 26.56 m 2 /g (Shen et al., 2019b). In comparison, the surface areas of MgO-coated magnetic sugarcane harvest residue biochars with Mg contents from 2.12% to 20.47% were examined, and it was observed that the surface area of the engineered biochar decreased with increasing Mg content from 118 to 27.2 m 2 /g. These suggest that the surface area of MgO-coated biochar is closely related to its Mg content, as a result of the trade-off between fashioned pore structure by MgCl 2 activation and pore blockage by coated MgO.
The maximum pyrolysis temperature also plays an important factor determining the properties of MgO modified biochar. The temperature will determine the degree to which the MgCl 2 solution will thermally decompose, and the amount of MgO produced.
The heating temperature also influences the formation of intermediate products (Huang et al., 2011), and the purity, crystal structure, and reactivity of MgO (Huang et al., 2011;Wang et al., 2019a). In addition, the properties of biochar itself are also significantly affected by pyrolysis temperature (Shen et al., 2019a;Suliman et al., 2016;Zhao et al., 2018).
In view of these considerations, the effects of excessive impregnation, Mg content, and pyrolysis temperature on the characteristics of MgO-coated watermelon rind biochar (MWRB) were investigated. Watermelon rind was used as the feedstock based on pre-trial investigations, watermelon rind biochar (WRB) revealing significantly higher adsorption capacities for heavy metals among the different fruit wastes tested (watermelon rind, pomelo peel, banana peel, mandarin orange peel, and sugarcane peel). As a common and abundant agricultural by-product, watermelon rind is rich in ash and carbohydrates, giving WRB the potential to have excellent performance in soil and water remediation (Lam et al., 2016). However, the surface area of WRB was found to be relatively small. The objective of this work is to find out the optimal synthesis parameters to produce MWRB with the highest surface area The watermelon rind was obtained from Zhaolanyuan Canteen, Tsinghua University, Beijing, China. The watermelon rind was oven dried at 80°C for 48 h, and then crushed and sieved through #20 mesh (≤ 0.83 mm). The feedstock was pretreated by MgCl 2 using excessive impregnation (Wang et al., 2015c), involving 10 g of watermelon rind added to 200 mL of MgCl 2 solution, and agitated at 240 rpm for 24 h by magnetic stirrer. After stirring, the treated feedstock and excess MgCl 2 solution were separated by vacuum filtration. The filtration residue was oven dried at 80°C for 24 h, crushed and sieved through #40 mesh (≤ 0.83 mm).
The dried mixture was pyrolyzed in a muffle furnace with the heating rate of 10°C/min from room temperature to 600°C (or designated highest temperatures for the temperature-dependence study), and maintained for 1 h (Shen et al., 2019b;Wang et al., 2015b). The produced biochar was washed with deionized water three times, and oven dried at 80°C for 24 h. The dried biochar was sieved (≤ 0.38 mm) and stored in sealed sample bags.

Biochar characterization
After obtaining the biochars under different conditions, they were thoroughly characterized. The details of the characterization are presented in the supporting information (SI). Briefly, the biochar pH was measured in aqueous solution with a pH meter; the BET surface area was measured by N 2 adsorption/desorption; the Mg content was determined by acid digestion and inductively coupled plasma optical emission spectrometry (ICP-OES); the molecular structure was tested by Fourier transform infrared spectroscopy (FTIR); and, the mineral composition was analyzed by X-ray diffraction (XRD).

Lead removal test
Removal tests were carried out in order to understand the ability of the MWRBs to remove Pb. Pb was selected as a representative heavy metal contaminant, because unmodified WRB revealed a higher adsorption capacity for Pb among different heavy metals and there is an urgent need to remediate lead in soil and water as mentioned.
For each removal test, 0.1 g of biochar was added to 20 mL of 50 mM Pb(NO 3 ) 2 solution. The mixture was shaken at 250 rpm for 24h to reach adsorption equilibrium.
The aqueous solution was separated from the solids by centrifugation at 4000 rpm for 4 min and filtration (0.22 μm), and analyzed for Pb concentration by ICP-OES.
The XRD patterns of WRB and MWRBs are shown in Figure 3. The unmodified WRB revealed the presence of KCl and aphthitalite ((K,Na) 3 Na(SO 4 ) 2 ). Potassium (K) and sodium (Na) are abundant elements in watermelon rind (Egbuonu, 2015). Therefore, the formation of minerals containing K and Na after pyrolysis could be expected (Li et al., 2019;Shen et al., 2017a;Shen et al., 2017b). After modification, characteristic peaks of MgO (Jin and Al-Tabbaa, 2014a) occurred on the XRD patterns, with their strength significantly increasing with increased Mg content, suggesting successful coating of MgO, and the MgO content being directly related to the Mg content. As the MWRB Mg content increased to 10.1%, two new peaks representing Mg(OH) 2 occurred on the XRD pattern. This may be due to the greater energy demand for thermal decomposition of hydrated MgCl 2 , and, therefore, a certain amount of Mg(OH) 2 was formed as an intermediate product (Huang et al., 2011). This assumption is supported by the previous finding that Mg(OH) 2 formed during the thermal decomposition of hydrated MgCl 2 and corncob mixture at a Mg content of 17.9% (Shen et al., 2019b). Due to the introduction of chlorine (Cl) with Mg, the strength of the peaks representing KCl also increased with increasing Mg content.
However, at Mg contents higher than 5.76%, their strength decreased because the formed MgO will have diluted the KCl concentrations, and it may also have coated over the KCl surfaces. These may also explain the decrease in the peak intensity representing (K,Na) 3 Na(SO 4 ) 2 . The FTIR spectra of WRB and MWRBs with different Mg contents are shown in Figure   4. Two peaks were observed for WRB. The peaks at 1570 cm -1 represent aromatic C=C, which were typically observed for plant-based biochar (Wang et al., 2017;Zhang et al., 2017a). The peaks at 1060 cm -1 are attributed to aliphatic C-O-C originated from the raw watermelon rind (Jawad et al., 2018), which are also typical FTIR peaks for biochar (Shen et al., 2018d). After modification, the aromatic C=C peak (1060 cm -1 ) significantly strengthened as the Mg content of biochar increased to 2.51%. Similarly, the peaks representing substituted aromatic C-H (1040 cm -1 ) (Keiluweit et al., 2010) was greatly enhanced after modification. Three new peaks representing aromatic C-H (870, 810, and 750 cm -1 ) occurred on the FTIR spectra of MWRBs. These suggest that the aromaticity of biochar was significantly enhanced after modification. As mentioned, the presence of MgCl 2 activates biochar during heating, facilitating the volatilization of organic materials. The increased volatilization 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 13 resulted in the increased aromatic structure of biochar. It was observed that 2.51% Mg was the optimal content for a maximum surface area of MWRB (Figure 2). The FTIR results further confirm this, because all the peaks (1060, 1040, 870, 810, and 750 cm -1 ) related to aromatic C on FTIR spectra reached their highest strength for the MWRB with 2.51% Mg content. These peaks decreased for MWRB with further Mg content, suggesting that the activation efficiency decreased. Another plausible reason may be the MgO on the biochar surface affecting the detection of aromatic C by the FTIR test.
It is of note that the unmodified WRB had a BET surface area of only 1 m 2 /g, therefore, it can be inferred that the MgCl 2 greatly contributed to the volatilization of the organics in watermelon rind, aiding the development of biochar porosity. The Mg content of the MWRBs produced at different temperatures were in the range of 1.89-2.51%, as the original Mg/feedstock ratio for the MWRBs produced at different temperatures was fixed at 0.48%. This suggests that the biochar yield at the temperature range of 400-800°C did not vary significantly, which is in line with previous findings that the yield of rice straw biochars produce in the range of 500-700°C (Shen et al., 2019a) .   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63 (Huang et al., 2011). For the MWRBs produced at 500-800°C, the thermal decomposition of hydrated MgCl 2 was complete, therefore, the produced MgO content was similar.

Lead removal
The effect of Mg content on the Pb removal capacity of MWRBs was shown in Figure   2. The modification significantly enhanced the Pb removal capacity of MWRB, and the efficiency increased with increasing Mg content. The highest Pb removal capacity (558 mg/g) was found at the highest Mg content (10.1% Mg), an increase of 208% over the 181 mg/g of the unmodified WRB. Two factors are attributed to the increased Pb removal capacity of MWRB. One is the increased surface area of biochar; it is well established that higher surface area of biochar can aid the adsorption/removal of heavy metals due to enhanced contact between metals and biochar (Shen et al., 2015;Shen et al., 2017b). The other is the effect of the MgO coating. Although the pH value of MWRB was not significantly changed compared to WRB (Figure S1), the buffering capacity was enhanced considerably (Shen et al., 2018c;Wang et al., 2018b). Therefore, greater amounts of Pb can be precipitated after modification. In addition to precipitation, MgO and its hydration products Mg(OH) 2 can also aid the removal of Pb through physical sorption, cation exchange, and complexation (Jin and Al-Tabbaa, 2014b;Jin et al., 2016). The increased Pb removal capacity for MWRBs with the highest Mg contents suggests that the increased buffering capacity, rather than surface area, was the main contributor to Pb removal. If surface area had been the main factor, the Pb removal capacity would have decreased as Mg content of the MWRB was increased from 2.51% to 10.1% with decreased surface areas. Therefore, MgO was the main contributor to increased Pb removal. It is of note that the greatest increase in Pb removal (as a percentage of Mg added) was observed in the Mg range 1.52-2.51% ( Figure 2). This may be due to the strong increase of biochar surface area, and the increased formation of MgO in biochar at this Mg content range.
The effect of pyrolysis temperature on the Pb removal capacity of WRB and MWRB is shown in Figure 8. The Pb removal capacity of the unmodified WRB increased with increasing pyrolysis temperature owing to the formation of alkaline minerals (e.g. K 2 O) and greater aromaticity, aiding precipitation and cation-π adsorption of Pb (Dodson, 2011;Shen et al., 2019b).
The Pb removal capacity of MWRB increased from 81.7 mg/g (at 400°C) to 742 mg/g (at 700°C), and then dropped to 368 mg/g (at 800°C). The Pb removal capacity of the MWRB was generally much higher than WRB due to increased surface area and MgO coating. However, for the MWRBs produced at 400°C and 800 °C, the Pb removal capacity was lower than the respective WRB counterparts. As discussed above, MgO was not observed for MWRB produced at 400°C; on the contrary, other Mg-minerals formed may have blocked the pores of the biochar, thus inhibiting the adsorption/removal of Pb. At 800 °C, although considerable amounts of MgO formed, the reactivity of the formed MgO was lower than that formed at lower temperatures (Garcia et al., 2004;Jin and Al-Tabbaa, 2014a), which may have reduced its ability to remove Pb. In addition, the large amount of (low-reactive) MgO and other Mg-

Environmental implications
In order to optimize performance in environmental applications, such as soil remediation, water treatment, and catalysis, biochar selection is often based on BET surface area. In this study, a maximum surface area of 293 m 2 /g was achieved for MWRB by increasing the Mg content to 2.51%, and using an optimal pyrolysis temperature of 600°C.
The unmodified WRB produced at 600°C had a Pb removal capacity of 181 mg/g, which is relatively high in comparison to biochars produced from other feedstocks (Gao et al., 2019). Furthermore, the modification greatly enhanced the MWRB Pb removal capacity to 558 mg/g (at 600°C), at an Mg content of 10.1%. Although further increases of Mg content may result in a higher Pb removal capacity, the modification cost will also increase. In industrial manufacturing, higher Mg content and synthesis temperature can lead to higher costs. Therefore, a balance between the cost and performance for MWRB production should be found out, depending on the remediation scenario. It should also be noted that MgCl 2 is abundant in natural resources (sea water) and industrial wastes (brine water) (Ye et al., 2018 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65   23 controlled by adjusting the initial Mg/feedstock mass ratio, following a logarithmic relationship. The modification significantly enhanced the surface area of the MWRB, with the BET surface area increasing first and then decreasing as the Mg content of biochar increased from 0.63% to 10.1%, with an optimal value 293 m 2 /g found for MWBR with an Mg content of 2.51%, whereas the surface area of the unmodified WRB was only 1 m 2 /g. The increased Mg content of the MWRBs aided the formation of MgO, and FTIR peaks representing aromatic C was also increased, being at their maximum in the MWRB with an Mg content of 2.51%. For a fixed Mg/feedstock ratio of 0.48% (resulting in MWRBs with Mg contents of 1.89-2.51%), the BET surface area of MWRB also increased as the pyrolysis temperature was increased from 400°C to 600°C, and then decreased with increasing temperature from 600 to 800°C, with an optimal value of 293 m 2 /g found at 600°C.
Increased pyrolysis temperature also aided the formation of MgO, but the reactivity of the MgO formed may be reduced at the highest pyrolysis temperatures. The peaks representing aromatic C of MWRBs peaked at 600°C.