Effectiveness and mechanisms of ammonium adsorption on biochars derived from biogas residues

The solid by-product of anaerobic digestion derived biochar has been used as an adsorbent for the liquid by-product and is a viable and environmental friendly way to control waste. In this study, two biochars (BC-PM, BC-ST) from pig manure biogas residue (BR-PM) and straw biogas residue (BR-ST) were assessed for ammonium adsorption. The ammonium adsorption on both BC-PM and BC-ST followed the Elovich kinetic model and fit well with the Langmuir isotherm, whereas BC-PM was better than BC-ST at absorbing ammonium. In addition, the adsorption mechanism was elucidated by analysing the physicochemical properties of the biochars. The Brunauer–Emmett–Teller (BET) surface area, pore structure and micromorphology, which are mainly related to the carbon in the biochars, had no direct correlation with the ammonium adsorption capacity. However, the ash in the biochars played an important role in ammonium adsorption. The metal elements in biochar ash significantly decreased after adsorption, especially potassium, which nearly disappeared as a result of ion exchange with ammonium. However, the quartz and carbonate mineral in biochar ash also functioned as ammonium adsorption sites according to Fourier transform infrared spectroscopy (FT-IR) and X-ray diffraction (XRD) analysis. This study revealed the irreplaceable role of ash in biochars for ammonium adsorption, which will guide the preparation of biochars from different types of feedstocks for ammonium adsorption.


Introduction
Anaerobic digestion for waste management and renewable energy production contributes considerably to environmental protection and the attention on energy recovery has grown in recent years. 1,2[12][13] Ammonium is one of the most abundant contaminants in the liquid by-product (biogas slurry) aer anaerobic digestion and is also a common nutrient in wastewater from industrial, agricultural and household activities.It contributes to the eutrophication of rivers and lakes, 14,15 and even 0.2 mg L À1 can be toxic to aquatic life. 16Adsorption is a feasible process for ammonium removal and has been well documented by many researchers.][19][20] As low-grade biomass, biogas residue derived biochars are inexpensive.More importantly, biochars possess particular advantageous properties.For example, some studies note the cost advantage of biochars derived from fermentation residue, [21][22][23] and revealed that these biochars had better adsorption properties than undigested feedstocks used directly. 24Although there is little information on the ammonium absorption properties of biochars derived from biogas residues, the studies of biochars produced from other biomass also provide a wealth of inspiration.Biochars prepared from plant biomass, 25 corn cob, 13 giant reed, 26 exhibit good ammonium adsorption performance.8][29] However, the highest ammonium adsorption capacity among biochars dose not correlate with the biggest BET surface area, which indicates that biochar properties, such as BET surface area and pore structure have no direct correlation with the ammonium adsorption capacity. 13,25,26Therefore, there must be factors inuencing biochar ammonium adsorption capacity other than the above properties.In fact, biochar, consists of two main composites, carbon and ash.The former determines the BET surface area and pore structure, whereas the effect of ash has not been studied systematically.Although there are several studies reported on biochars, there are few in-depth studies on the ammonium absorption of biochars derived from biogas residues.Furthermore, although there are some preliminary experiment examining the adsorption mechanism of ion exchange in ash, 30,31 the specic contribution from SiO 2 and minerals in biochar ash are usually neglected, which might be a crucial factor in biochar ammonium adsorption.
Therefore, this study aimed to assess the value of low-cost biochars derived from biogas residues, as efficient alternative absorbents for ammonium removal.Furthermore, the structural properties (composition, porous structure and surface chemistry) and ammonium adsorption of biochars prepared from two typical biogas residues (pig manure biogas residue and straw biogas residue) were compared.More importantly, the effect of ash composition on ammonium adsorption was investigated, providing insight into different the ammonium adsorption mechanism of biochars.This study met the requirements of low-cost and effective ammonium removal, and describes a value-added and environmentally friendly method to treat biogas residue.

Source of raw materials
Biogas residues, the raw material for biochars, were by-products of anaerobic digestion from pig manure (PM) and straw (ST) in our own laboratory and were named BR-PM and BR-ST, respectively.All of the chemicals in this study were purchased from Sinopharm Chemical Reagent Co., Ltd. and used as received without further purication.

Preparation of biochar
The biochar precursors BR-PM and BR-ST were chemically activated before being pyrolyzed.One gram of precursors were impregnated in 1 g L À1 KOH solution for 12 h, and dried at 105 C.Then, the samples were pyrolyzed at 550 C for 2 h with a ramp rate of 5 C min À1 in a tube furnace (SK2-1-12, Tianye, China) under a N 2 atmosphere (100 mL min À1 ), which was similar to a previous study. 32Then, the biochars were washed with excess deionized water and stirred with a glass rod for 5 minutes aer pyrolyzation, which was repeated three times.Finally, they were dried at 80 C in an oven for 12 h.Accordingly, the two types of biochars were named BC-PM and BC-ST.

Analytical methods
X-ray diffraction (XRD) analysis was performed on an Ultima IV X-ray Diffractometer (Rigaku, Japan) operated at 35 kV and 10 mA, with a scanning rate of 20 min À1 and a 2q angle ranging from 5 to 70 using CuKa radiation (l ¼ 0.15418 nm).The carbon, nitrogen, hydrogen and sulphur content of the samples were analysed with an elemental analyser (Vario EL III, Elementar, Germany).Ash content was determined by heating the samples at 600 C for 4.0 h with a ramp rate of 10 C min À1 in a muffle furnace.Fourier transform infrared spectroscopy (FT-IR) spectra were obtained on a Nicolet Avatar 660 (Thermo Electron, USA) using the KBr pellet method.The cation exchange capacity (CEC) of the biochar was measured using the Schollenberger and Simon method. 33The biochar was saturated with 1 mol L À1 neutralized ammonium acetate and subsequently washed with 95 v/v% alcohol.It was then transferred into a 150 mL kjeldahl ask to distill and titrate.The Brunauer-Emmett-Teller (BET) surface area and average pore size of the biochars were measured using an Automated Physisorption and Chemisorption Analyser (Tristar 3000, USA) with N 2 adsorption/ desorption at À196 C. All of the samples were degassed at 300 C for 2 h before measurement.Biochar morphology and surface composition were investigated using a scanning electron microscope (SEM) with Energy Dispersion X-ray Spectrometry (EDX) capability (ZEISS SIGMA, Germany).

Adsorption of ammonium
The adsorption performance of the biochars toward ammonium was evaluated using two types of wastewater.One NH  1.
Kinetics experiments in the two types of wastewater were performed to examine how the adsorption behaviour of ammonium on biochars relate to contact time.Adsorbent (0.03 g, 1.0 g L À1 ) was mixed with 30 mL wastewater solution without adjusting the pH in a thermostatic shaker at 30 C/180 rpm, and the concentration of residual ammonium at different time points was measured.The supernatant was ltered through a 0.45 mm membrane, and the residual ammonium concentration (C t ) was measured using the Nesslerization method with a Thermo scientic Evolution 220 UV-Visible spectrophotometer at 420 nm. 34he ammonium adsorption capacity of the biochar was calculated according to the mass balance equations in which q t is the adsorption capacity at time t (min), C 0 is the initial concentration of ammonium (100 mg L À1 ), C t is the concentration of ammonium at time t of adsorption (mg L À1 ), m is the mass of adsorbent (g), and V is the volume of solution (L).Adsorption isotherms were obtained during the batch mode adsorption experiments in a thermostatic shaker at 30 C/180 rpm.C e was measured aer 0.03 g (1.0 g L À1 ) biochar was mixed with 30 mL wastewater for 4.0 h (reaching complete equilibrium according to our preliminary study).In particular, the C 0 of the two types of wastewater varied from zero to the initial ammonium concentration by dilution.Then, the equilibrium adsorption capacity (q e , mg g À1 ) was still calculated according to eqn (1).All of the adsorption assays were performed in triplicate.

Characterization of biogas residues and biochars
The elemental analysis and ash content of BR-PM, BR-ST, BC-PM and BC-ST, in addition to the BET surface areas, average pore size and biochar yields for both biochars are listed in Table 2.The two types of biogas residue contained similar elemental (CHNS) content.The ash content of BR-PM and BR-ST was 31.21AE 1.53% and 17.75 AE 0.89%, respectively.However, the derived biochars exhibited a quite different carbon content, ash content and yield.The carbon content of BC-PM was 15.75 AE 0.14%, whereas that of BC-ST reached 37.59 AE 0.16%.The ash content of BC-PM (62.93 AE 2.66%) was higher than that of BC-ST (49.67 AE 2.05%).More importantly, another distinct difference was the BET surface area, which was higher for BC-ST (260.21 m 2 g À1 ) than BC-PM (167.88 m 2 g À1 ).The average pore radius on BC-PM and BC-ST were calculated as 4.08 and 2.57 nm, respectively, using nitrogen adsorption isotherms and classied as mesoporous according to the International Union of Pure and Applied Chemistry (IUPAC). 35e XRD patterns of BC-PM, BC-ST and their ashes are shown in Fig. 1, which was used to probe the short range ordered biochar structures and the presence of miscellaneous crystalline minerals. 36here was a sharp peak at 20.99 and 26.74 in the two biochars conrming the presence of quartz in both of them. 37,38his nding indicated that both of the biochars were amorphous materials, lacking a distinct crystalline structure, and no other signicant peaks were observed.However, a few new peaks appeared on Ash-BC-PM, the sample of BC-PM aer calcination at 600 C in air for 2 h.Specically, peaks at 26 and 33 were assigned to (CaMg) 3 (PO 4 ) 2 /Whitlockite; peaks at 28 and 41 were due to sylvite (KCl), and a peak at 31 conrmed the presence of CaMg(CO 3 ) 2 /dolomite. 37,39These results suggested that BC-PM actually contained appreciable amounts of metals and carbonate that was absent from the XRD pattern because of the carbon coverage.The minerals might be present within the biochar pores rather than on the biochar surface, therefore, washing the biochars with water is ineffective, and these minerals are evident via XRD only aer removing carbon (via calcination).The mineral XRD peaks emerged once the carbon was burnt off.Cao et al. 40 also similarly reported that biochars derived from dairy manure commonly contained high levels of Ca, Mg and P. Furthermore, Herrera et al. 41 reported that these minerals were from superuous dietary intakes.As for the Ash-BC-ST XRD pattern, the BC-ST sample aer calcination, only one weak peak at 28 emerged to conrm trace amounts of KCl.Based on the above analysis, KCl was present in BC-ST as well.Therefore, it can be speculated that BC-ST  contained fewer minerals than BC-PM, which might affect their ammonium adsorption behaviour.

Adsorption kinetics study
The adsorption kinetics on biochars was investigated to determine the time required to reach the adsorption equilibrium and further examine the mechanisms of ammonium adsorption.Fig. 2 shows time dependent ammonium adsorption on BC-PM and BC-ST in articial wastewater and biogas slurry.In articial wastewater (Fig. 2a), the ammonium adsorption equilibrium for BC-PM and BC-ST was reached aer approximately 4.0 h with an adsorption capacity of 13.66 mg g À1 and 11.36 mg g À1 , respectively, and approximately 80% and 90% of the equilibrium adsorption capacity was reached within the rst 1.0 h, respectively.In biogas slurry (Fig. 2b), equilibrium was also established in 4.0 h for BC-PM and BC-ST with an adsorption capacity of 26.82 mg g À1 and 19.16 mg g À1 , respectively.The results indicated that ammonium adsorption on the two biochars was a rapid process.
In addition, the kinetics data were also analysed using pseudo-rst-order, pseudo-second-order and Elovich models, and the calculated model parameters are shown in Table 3. Elovich models were suitable for describing the ammonium adsorption process in both articial wastewater and biogas slurry.

Adsorption isotherms study
The adsorption isotherms for ammonium on BC-PM and BC-ST in articial wastewater (concentration range from 0-100 mg L À1 ) and biogas slurry (concentration range from 0-855 mg L À1 ) t to the Langmuir and Freundlich models at 30 C are presented in Fig. 3.The Langmuir model t the data better than the Freundlich model for BC-PM, whereas the Freundlich model described the data better for BC-ST.The simulation parameters for the two models are presented in Table 4.In particular, the Langmuir maximum capacity for BC-PM in articial wastewater and biogas slurry were 37.26 mg g À1 and 48.89 mg g À1 , respectively, which were both higher than those of BC-ST for the adsorption of ammonium from articial wastewater and biogas slurry, respectively.
In general, BC-PM exhibited better ammonium adsorption than BC-ST in articial wastewater or biogas slurry.

Adsorption performance comparison
To illustrate the potential of the BC-PM and BC-ST prepared in this study as promising biochars for ammonium adsorption with improved ammonium adsorption behaviours, the maximum adsorption capacities of various agro-based biochars in the literatures prepared in similar pyrolyzation conditions were also investigated and compared, as listed in Table 5.The maximum ammonium adsorption capacities of BC-PM and BC-ST reached 37.26/48.89mg g À1 and 21.17/28.86mg g À1 in articial/real wastewater, respectively, which were much higher  than biochars prepared from rice husk or hardwood in similar adsorption conditions.It also indicated that biogas residue could be an alternative precursor of the biochars for ammonium adsorption.

Ammonium adsorption mechanism
The above studies showed that BC-PM and BC-ST showed excellent ammonium adsorption performance.As mentioned in the section on BET analysis, BC-PM had a lower specic surface area.However, BC-PM had a higher adsorption capacity than BC-ST, suggesting that the BET surface area is not the main factor in ammonium adsorption.Biochars contain two main components, carbon and ash, and the BET surface area and pore structure were mainly determined by carbon content, leading us to examine the composition and content of the ash in both biochars.3.5.1 Metal ion exchange in biochars ash.Metal ion exchange is usually seen as an important aspect of ammonium adsorption. 25,44In this study, the CEC values of BC-PM and BC-ST were measured as 2.18 AE 0.3 mmol g À1 and 0.893 AE 0.3 mmol g À1 , respectively (Table 2).Obviously, BC-PM possessed a higher cation exchange capacity than BC-ST indicating that BC-PM might absorb more ammonium through ion exchange than BC-ST.This nding might explain some of the improved BC-PM ammonium adsorption capacity.On the other hand, XRD analysis veried that more metals were present in the BC-PM ash than the BC-ST ash.Thus, the CEC of the two biochars may come from the metals in them.To estimate the contribution of metals and observe the possible transformation of the biochar microscopic structure before and aer ammonium adsorption, SEM images and EDX of BC-PM, BC-ST and the corresponding samples aer adsorption in articial wastewater were obtained (Fig. 4a-d).Signicant differences in biochar pore structures from BR-PM and BR-ST were clearly observed.Specically, the surface of BC-PM was rough, whereas BC-ST possessed an apparently porous structure, which corresponded with the BET analysis.There was also no signicant change in the pore structures of the biochars aer adsorption, indicating they possessed good mechanical properties.However, the element composition varied greatly aer adsorption, which was reected by the SEM-EDX results.The weight percent of metal elements (Mg, K, Ca) in BC-PM and BC-ST decreased strongly aer adsorption, and K nearly disappeared, which might be caused by ion exchange with ammonium, an ammonium adsorption mechanism suited to both BC-PM and BC-ST.
3.5.2SiO 2 in biochar ash.The XRD and SEM-EDX results conrmed the effect of metals in biochar ash on ammonium adsorption.However, ammonium adsorption might also be related to the functional groups on the biochar surface.Ammonium adsorption performance of the biochar depends on porosity as well as the active groups on the surface. 45Thus, FT-IR spectra were collected for qualitative characterization of biochar surface functional groups.The FT-IR results for ammonium adsorption in articial wastewater aer 20 min and 240 min for BC-PM are shown in Fig. 5a, and those for BC-ST are depicted in Fig. 5b.
The broad bands at 3420 cm À1 (Fig. 5a) and 3445 cm À1 (Fig. 5b) were attributed to the O-H stretching vibration of BC-PM and BC-ST, respectively.The bands centred on 1630 (Fig. 5a) and 1618 cm À1 (Fig. 5b) were attributed to O-H bending vibrations. 46The strong bands for BC-PM and BC-ST at 1025 cm À1 and 1058 cm À1 were assigned to an asymmetric Si-O-Si stretching mode.Peaks at 572 cm À1 and 805 cm À1 , neither of which split in two, were assigned to a symmetric Si-O-Si stretching mode. 45,47The above results indicated that SiO 2 in both biochars was present in an amorphous state, which was also conrmed by the XRD patterns.However, the peaks for asymmetrically stretching Si-O-Si in BC-PM shied to 1040 cm À1 aer 20 min and 240 min adsorption.In addition, the peaks for asymmetrically stretching Si-O-Si were also detected to show a blue-shi aer adsorption by BC-ST.Specically, as shown in section B (Fig. 5b), 13 cm À1 was shied in the rst 20 min, and another 9 cm À1 in the later 220 min.These shis can be attributed to electrostatic interactions between SiO 2 and ammonium, and the SiO 2 in the biochars may have acted as ammonium adsorption sites, playing a signicant role in the process.Si-O-Si, the main functional groups on the two biochars, was from SiO 2 in biochar ash.Thus, another component in biochar ash was veried to contribute to ammonium adsorption.3.5.3Carbonate minerals in the ash of biochars.Compared with the FT-IR spectrum of BC-ST, that of BC-PM showed a distinctive region weakened aer adsorption (section A in Fig. 5a).The peaks at 1484 and 1426 cm À1 were assigned to CO 3 2À of the mineral components (carbonates) in biochars originating from pyrolyzation, 40,48 and they also belonged to BC-PM ash according to the XRD analysis.In addition, the EDX results shown in Fig. 4 suggest that the BC-PM metal content was much higher than BC-ST.The carbonates might serve as additional ammonium adsorption sites, contributing to BC-PM's high adsorption capacity.
To verify their exact effect on ammonium adsorption, BC-PM was treated with water for 8 h because carbonates dissolve in water aer a long treatment time.As expected, water treatment weakened the FT-IR peaks for carbonates, which is shown in section C of Fig. 6a.Furthermore, the ash content of water treated with BC-PM barely decreased, and the EDX spectra analyses of BC-PM treated by water (Fig. 4e) also showed that its metal content was barely decreased.Accordingly, BC-PM's ammonium adsorption capacity only decreased to 11.87 mg g À1 from 13.66 mg g À1 of the control as shown in Fig. 6b.Thus, the reduced ammonium adsorption capacity (1.79 mg g À1 ) on water treated BC-PM might be attributed to the slight dissolution of carbonate and the accompanying decreasing additional ammonium adsorption sites on mineral components.
For HCl-treated BC-PM, the ash content fell sharply to 35.27% from 62.93%.The SEM imaging and EDX spectral analyses (Fig. 4f) showed that the structure became more porous, and the metal content (Mg, K, Ca) was nearly zero, owing to the dissolution of mineral components by HCl.The results of FT-IR (Fig. 6a) showed that the carbonates peaks disappeared and SiO 2 peaks weakened aer HCl treatment.Accordingly, the ammonium adsorption capacity decreased from 13.66 mg g À1 to 4.72 mg g À1 .Thus, the reduced ammonium adsorption capacity (8.94 mg g À1 ) on HCl-treated BC-PM was not caused by the same phenomenon as water treatment, but rather by a combined action of missing metal ion exchange, carbonates and reduced SiO 2 .In addition, the porous structure of biochar enhanced by HCl treatment seemed to have little effect promoting ammonium adsorption capacity.In general, the effect of carbonate mineral components on additional  adsorption sites in biochar for ammonium adsorption was less than metal ion exchange and SiO 2 effect.

Conclusions
Two biochars (BC-PM and BC-ST) were prepared from pig manure and straw biogas residues through activation and then pyrolyzation.They both exhibited good ammonium adsorption in articial wastewater and biogas slurry.The kinetics of ammonium adsorption on BC-PM and BC-ST both follow the Elovich model.Adsorption isotherm analysis indicated that BC-PM and BC-ST both t the Langmuir model well, although BC-ST possessed a lower adsorption capacity than BC-PM.The BET surface area and biochar pore structure had no direct correlation with ammonium adsorption capacities.However, the ash in the biochars played an important role in ammonium adsorption.The metal elements in the biochars ash decreased as a result of ion exchange with ammonium.SiO 2 and carbonate mineral components in biochar ash also function as ammonium adsorption sites.
solution was prepared from NH 4 Cl.The concentration of NH 4 + -N was 100 mg L À1 , and the initial pH of the solution was approximately 5.0 without adjustment.A real sample of NH 4 + -containing wastewater was collected from the supernatant of the biogas slurry produced from pig manure anaerobic digestion in our own laboratory.Its NH 4 + -N concentration was approximately 855 mg L À1 aer centrifugation at approximately 10 000g (RCF) for 5 min before use.The detailed quality parameters of the biogas slurry are listed in Table

Fig. 1
Fig. 1 XRD patterns of BC-PM, BC-ST and their ashes.

Fig. 2
Fig. 2 Kinetics of ammonium adsorption on BC-PM and BC-ST in artificial wastewater (a) and biogas slurry (b) with data fit to pseudofirst-order, pseudo-second-order and Elovich.

Fig. 3
Fig. 3 Isotherms of ammonium adsorption on BC-PM and BC-ST in artificial wastewater (a) and biogas slurry (b) with data fit to the Langmuir and Freundlich models.

Fig. 5
Fig. 5 FT-IR spectra of BC-PM (a), BC-ST (b) and the same samples after ammonium adsorption at 20 min and 240 min.Fig.6 Comparison of the FT-IR spectra of BC-PM after various treatments (a) and the ammonium adsorption capacity contribution of different components (b).

Fig. 6
Fig. 5 FT-IR spectra of BC-PM (a), BC-ST (b) and the same samples after ammonium adsorption at 20 min and 240 min.Fig.6 Comparison of the FT-IR spectra of BC-PM after various treatments (a) and the ammonium adsorption capacity contribution of different components (b).

Table 3
Kinetic constants of the various models of ammonium adsorption on BC-PM and BC-ST in artificial wastewater and biogas slurry

Table 4
Langmuir and Freundlich constants for ammonium adsorption on BC-PM and BC-ST in artificial wastewater and biogas slurry