Adsorption and desorption of phosphate on biochars

Journal of Environmental Chemical Engineering 4 (2016) 37–46

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering journal homepage: www.elsevier.com/locate/jece

Adsorption and desorption of phosphate on biochars P.A. Trazzi, J.J. Leahy, M.H.B. Hayes, W. Kwapinski* Carbolea Research Group, Department of Chemical and Environmental Science, University of Limerick, Limerick, Ireland

A R T I C L E I N F O

A B S T R A C T

Article history: Received 5 August 2015 Received in revised form 8 October 2015 Accepted 2 November 2015 Available online 5 November 2015

Biochar (BC) is regarded as a potential carbon sequesterer, a soil fertility enhancer, and a preventer of nutrient leaching. Phosphorus amended biochar could enhance soil fertility. This work investigated the adsorption and desorption of phosphate from aqueous solution on two different carbonized materials. Sugar cane bagasse (SC) and Miscanthus
giganteus(M) samples were carbonized at various temperatures (between 300  C and 700  C) for residence times of 20 or 60 min. The largest surface area and the highest P adsorption at 20  C and pH 7 was obtained for M BC prepared at 700  C and at the longer residence time, compared to the SC BC made under the same conditions (approximately 15.5 mg g1 and 12.8 mg g1 for 400 mg dm3 phosphate in solution, respectively). Adsorption of P on BCs was endothermic and increased with process temperature. The amount of desorbed P was proportional to its adsorption capacity. Two isotherm models (Freundlich and Langmuir) fitted the experimental results of phosphate adsorption onto the BC, and the Langmuir adsorption model described it better. Thermodynamic parameters are compared in the text with phosphate adsorption on other BCs reported in the literature. Our data suggest that adding phosphate to BC could provide a better way to apply P to soil in order to obetain better agronomic performances. ã 2015 Elsevier Ltd. All rights reserved.

Keywords: Miscanthus Sugar cane bagasse Carbonization Pyrolysis Energy

1. Introduction Phosphorus deficiency is a major factor restricting crop yields. This deficiency especially applies for tropical weathered soils, where the bioavailability of phosphorus (P) has a major impact on of crop production [1]. Under natural conditions, the weathering of rocks and the release of elements essential for crop growth is a slow process, and much of the P applied as fertilizer becomes fixed in forms such as calcium iron, and aluminium phosphates that are relatively insoluble and unavailable for plant growth needs. In these situations non-amended soils are capable of supporting only slow-growing vegetation and crops adapted to low soil phosphate availability. Biochar (BC), produced by pyrolysis under limited air supply and temperatures greater than 300  C, is applied to soil to facilitate carbon storage, the filtration of percolating soil water, and the enhancement of crop productivity. Variabilities in the properties and performances of BC are attributable to various process parameters related to its formation, such as: temperature, the residence time and the heating rate during pyrolysis [2], the feedstock [3], the particle size [4], and the method of pyrolysis [5]. The degree of alteration of the original structures of the biomass,

* Corresponding author. E-mail address: [email protected] (W. Kwapinski). http://dx.doi.org/10.1016/j.jece.2015.11.005 2213-3437/ ã 2015 Elsevier Ltd. All rights reserved.

through microstructural rearrangement, attrition during processing, and the formation of cracks all depend upon the processing conditions to which they are exposed [6]. Piterina and Hayes [7] have shown that arbuscular mycorrhizal fungi in associations with BC can dissolve tricalcium phosphate and would make the locked P available for the plant having a symbiotic relationship with the fungus. Recognition of the properties and mechanisms of BC adsorption and desorption of phosphate applications to soil is very important for its potential uses in soil fertilizer applications. Xu et al. [8] concluded that BC application to soil altered P availability by changing the P sorption and desorption processes. These effects were dependent on soil acidity, and have important implications for improving soil productivity. The phosphorus concentration in carbonized materials vary depending on crops and growing conditions; e.g. the BC content of rice husk is 4.7 mg P g1 [9], that of Miscanthus
giganteus (M) is 2.5 mg P g1 [10], and of sugar cane bagasse (SC) is 6.1 mg P g1 [11]. Residue P in BC form is readily available for plants [12]. The conditions under which BC is produced can influence its effects on P availability. As the temperature of pyrolysis increases, the specific surface area increases, and the volatile matter and surface functional groups decrease [13–15]. All those changes can influence P bioavailability. There is evidence to indicate that BC can decrease P-fixation in soil resulting in greater bioavailability of added inorganic phosphate [5,16,17]. Cui et al. [16] observed that

38

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46

the presence of BC decreased P adsorption on the Fe-oxides and thereby enhance P bioavailability. Parvage et al. [17] concluded that BC from wheat straw can act as a source of soluble P, and low and high additions of BC can have different effects on soil solution P concentration. Morales et al. [5] found that BCs, depending on how produced, can have very different P sorption and desorption properties: a BC produced by fast pyrolysis decreased the P-fixing capacity of a degraded tropical soil, whereas that formed in slow pyrolysis had the opposite effect. Xu et al. [8] found that the BC application altered the availability of P by changing the capacities of the soils to sorb and desorb P, and these effects were influenced by the soil acidity which has important implications for soil productivity. Some studies presented cation modified BCs, such as Fe [16,18], Mg [19] or La [20] which influence sorption as well as desorption properties. These cation-modified BCs enhance the sorption of P but might negatively influence the desorption of P and its plant availability. Research on the sorption and desorption properties of BCs is still in the early stages and number of observations need to be confirmed. Most studies have worked with the capacities of BCs or of BC-complexes to adsorb. An understanding of the properties of BCs and of their influences on the mechanisms of the sorption and desorption of P is very important for a better management of soil P applications. The objective of this study was to investigate the influences on the adsorption and desorption of phosphate of the temperature and of the residence time on the carbonization process of two feedstock bio-materials. 2. Material and methods BCs were produced by the pyrolysis of sugar cane bagasse (SC) and of Miscanthus
giganteus (M). The materials were air-dried and shredded into pieces ca. 2–3 cm in length. These pieces were pyrolyzed at 300, 500 and 700  C for residence times of 20 and 60 min, and are referred to as 300/20; 300/60; 500/20; 500/60; 700/20; and 700/60 BCs, respectively. BCs were produced by slow pyrolysis in a fixed-bed reactor. The pyrolysis apparatus consisted of a temperature controller cabinet, a quartz tube reactor, and an electric furnace heater. The BCs obtained were ground, washed with deionized water, dried in a furnace at 105  C for 24 h, and then stored in plastic containers. The surface area was measured by N2 adsorption using the BET surface area analyser. Ultimate analyses of BC samples were carried out with a CHNS/O analyzer (Elementar Vario LE Cube). Moisture contents of the BC samples were determined according to the standard: ICS 75.160.10, DD CEN/TS 14774-3:2004. Ash contents and volatile matter contents were determined according to ICS 75.160.10, DD CEN/TS 14775:2004 and ICS 75.160.10, DD CEN/TS 15148:2005, respectively. A Fourier transform infrared spectrometer (FTIR) (Cary 630 FTIR spectrometer, Agilent Technologies Inc.) with the resolution of 4 cm1 and 64 scans per sample was used to collect spectra in the range of 600–4000 cm1. Samples of 300/60, 500/60 and 700/60 SC BCs were selected before and after adsorption (BCs residual solids from 100 mg P dm3 at pH 7). A preliminary study showed that SC and MC BCs were similar in their peaks and absorbance, and no differences were found between residence times of 20 and 60 min for the same pyrolysis temperature. Determination of zeta potential was carried out for all BCs. To each sample (0.2 g) 100 ml of de-ionized water at pH 6.5 was added. The change in the zeta potential from P adsorption was tested by adding also 0.2 g of 500/60 M BC, with or without the addition of P solution (100 mg P dm3) to each flask, when the pH of suspensions had been adjusted to within 3.0 to 8.0 with NaOH or HCl. Then the analysed solutions were shaken at 250 rpm for

30 min using a mechanical shaker. The suspensions were dispersed ultrasonically for 1 h at 30  1  C in a bath-type sonicator at a frequency of 40 kHz and a power of 300 W. After that the solution was filtered using a filter paper (Whatman 42 filter paper). The zeta potencial of each supernatant solution obtained was determined using a Malvern Zetasizer Nano (Malvern Instruments). To study the effects of pH, 0.2 g of 500/20 SC and of M BC was mixed with 100 cm3 of solution containing 100 mg KH2PO4. The pH of the solutions were adjusted to values of 3–10, before adding the biochar, using a PHM 84 pH meter (Radiometer, Denmark) with glass REF 451 and calomel pHG 201-8 electrodes. A separate set of experiments was set up for each pH measurement. The suspensions were agitated on a shaker at 250 rpm and 25  C for 24 h. Each sample was filtered through a 0.5-mm syringe filter and the pH of the filtrate was measured. For the different phosphate adsorption experiments, to BC (0.2 g in a conical flask) was added 100 cm3 of phosphate solution each containing: 25, 50, 100, 200 and 400 mg P dm3 at pH 7 (the pH values were adjusted by adding HCl or NaOH (0.1 mol dm3)). An initial study indicated that equilibrium for the adsorption of inorganic P was attained in a matter of minutes. Nevertheless the suspensions were agitated on a shaker at 200 rpm and 25  C for 24 h. The suspensions were filtered through a 0.5-mm paper filter. The phosphate contents of the filtrates were meansured, and the solids were collected for measurements of phosphate desorption. The experiments were carried out in triplicates and the BCs from each sample were combined for the desorption experiments. For desorption, to BCs residual solids (0.2 g) from each phosphate concentration applied were added to 100 cm3 of 0.01 mol dm3 of citric acid in a conical flask, and the suspensions were agitated for 24 h on a shaker at 200 rpm and 25  C. Samples were filtered as described for the adsorption experiments and the phosphate concentrations in solution were determined. The equilibrium adsorption capacity was calculated using Eq. (1) qe ¼

VðC e  C 0 Þ m

ð1Þ

where: qe is the adsorption capacity at equilibrium (mg g1), V the volume of solution (dm3), C0 and Ce are the initial and equilibrium concentrations of phosphate (mg dm3), and m is the weight of adsorbent (g). Phosphate concentrations were determined using a UV–vis4000, Varian Spectrophotometer, using the stannous chloride method [21]. Each experiment was carried out three times, and mean values are presented. Differences between sorption for BCs in each treatment were tested for significance using a factorial analysis of variance and Duncan’s multiple range tests. Differences are reported as significant at p < 0.05. The adsorption data of the phosphate on the BCs were analyzed using the Langmuir and Freundlich isotherm models. The Langmuir model, described Eq. (2) [22] is. qe ¼

ðkL
C e Þ ð1 þ qm
C e Þ

ð2Þ

where: Ce is the equilibrium concentration of phosphate (mg dm3), the constant qm (mg g1) and KL are the characteristics of the Langmuir equation (dm3 mg1) and can be determined from the linearized form (plots of Ce/qe vs. Ce). The Freundlich model is expressed by Eq. (3) [23]: qe ¼ K F
C e n

ð3Þ 1

where: KF is the Freundlich adsorption capacity (mg g ), n is the Freundlich constant. The above equation can be linearized to calculate the parameters KF and n (plots of log qe vs. log Ce).

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46

3. Results and discussion

Several methods are used [7] for the determination of the thermodynamic parameters for adsorption systems (liquid–solid). The common point in all of these methods is the calculation of the standard free energy change of the adsorption. The thermodynamic parameters, such as Gibb’s free energy (DG0), enthalpy (DH0) and entropy (DS0), were estimated using Eqs. (4) and (6):

DG0 = –R
T ln (Kc)

3.1. Material properties Hemicelluloses and cellulose will be converted mainly to gaseous products during the pyrolysis process, whereas lignin is transformed at higher temperatures, mainly to char [27]. Increasing carbonization temperature trended to raise the fixed carbon, the ash content, and the surface area (BET), and to decrease the volatile matter in the BCs (Table 1). Increasing the residence time tended to raise the content of fixed carbon and to decrease the content of volatile matter. Increasing the temperature and/or residence time during the carbonization process decreased the atomic ratios of H/C and O/C. These observations and the results presented in Table 1 are typical for the carbonization of lignocellulosic biomass. The degree of carbonization of BC can be promoted by increasing the pyrolysis temperature and the residence time, indicating the BCs obtained might be beneficial for carbon sequestration. Similar results were found for different carbonization temperatures in references [20,28]. The adsorption capacity of materials can be identified from their physical characteristics, including porosity, surface area, and pore size, and especially in case of BC, to ion-exchange capacity. The surface area of BC increases with increasing temperature until it reaches a stage at which deformation occurs, resulting in subsequent decreases in surface area [6]. The surface area depends largely upon the C mass removed during processing, creating pores in the material [29]. This is attributed to a sintering effect, followed by shrinkage of the BC, and realignment of a structure resulting in decreased pore sizes [30]. Lua et al. [2] evaluated the importance of pyrolysis temperature, holding time, nitrogen flow rate, and heating rate on the properties of the BC produced. They concluded that temperature has the most significant effect, followed by the pyrolysis heating rate. The increase in pyrolysis temperature leads to the increase of the surface area of BC [13,31], and this facilitates higher sorption of chemicals [32]. Char made from wheat straw at 500–700  C was well carbonized and its surface area was relatively high (>300 m2 g1), whereas chars formed at 300–400  C were partially carbonized and had a lower surface area (<200 m2 g1) [32].

(4) 1

1

is the gas constant, and T is where: R = 8.314 J
K mol temperature in K. Kc is the equilibrium constant, and can be calculated from Eq. (5). Kc ¼

qe Ce

ð5Þ

The equilibrium constant was determined for 500/20 SC and M BCs, for three temperatures, 20, 35 and 50  C. The equilibrium constant can be calculated based e.g. on the Langmuir equation [24] or the Freudlich equation [25], or Eq. (5) [26]. Because the absolute values of DG0 depend on the method applied, it is important that the same method be used when comparing values for different materials. The DH0 and DS0 values were determined from the slope and intersection, respectively, from Eq. (6). lnðK c Þ ¼

DH0 DS 0 þ ðR
TÞ R

ð6Þ

The energy used for temperature rise (Qd) from ambient (Ta) to the water boiling temperature can be obtain from Eq. (7). Qd = m
cp,wet
(100 – Ta)

(7)

where: m is the mass of wet biomass (kg), and cp,wet is the specific heat capacity of a wet material (J kg1 K1). The energy for water evaporation (Qw) was determined from Eq. (8) Qw = mw
lw

(8)

where: mw is the mass of the water in the biomass, and lw is the water vaporization heat J kg1. Carbonization energy (Qc) can be calculated from Eq. (9) Qc = (m – mw)
cp,dry
(Tc – 100)

39

(9) 3.2. FTIR



where: Tc is final carbonization temperature ( C), and cp,dry is the specific heat capacity of the dry material (J kg1 K1).

FTIR spectral analysis is important for identifying characteristic functional groups which are responsible for adsorbing ions [33].

Table 1 Ultimate and proximate analyses of SC and of M BCs samples formed under different pyrolysis conditions. Biochar

Ultimate analysis wg%

BET

Proximate analysis wg%

N

C

H

S

O

Atomic ratios C/N

H/C

O/C

m2 g1

Volatile matter

Ash

Fixed carbon

Sugar cane 300/20 300/60 500/20 500/60 700/20

0.3 0.3 0.6 0.6 0.5

51.7 54.0 82.0 83.0 86.5

6.0 6.3 3.7 3.4 1.9

0.0 0.0 0.0 0.0 0.0

42.1 39.5 13.7 13.0 11.1

198 193 142 143 159

0.12 0.12 0.05 0.04 0.02

0.81 0.73 0.17 0.16 0.13

4.93 9.20 10.8 60.5 131

46.9 44.7 21.2 20.1 17.2

3.44 2.53 9.54 9.41 12.5

49.7 52.8 69.3 70.6 70.3

Miscanthus 300/20 300/60 500/20 500/60 700/20 700/60

0.4 0.4 0.5 0.6 0.5 0.5

63.1 63.6 83.0 86.3 89.2 90.4

6.0 5.6 3.6 3.2 1.9 1.7

0.0 0.0 0.0 0.0 0.0 0.0

30.5 30.3 12.9 9.92 8.45 7.46

174 143 163 147 179 192

0.10 0.09 0.04 0.04 0.02 0.02

0.48 0.48 0.16 0.11 0.09 0.08

6.17 6.39 21.8 81.0 228 244

45.0 42.7 23.5 22.5 18.9 13.9

6.54 7.12 10.0 11.3 13.3 14.9

48.5 50.3 66.4 66.3 67.7 71.2

40

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46

Fig. 1. The FTIR spectra of: (a) 300/60, 500/60 and 700/60 SC BCs, and (b) 300/60, 500/60 and 700/60 SC BCs before and after phosphate adsorption.

Fig. 1a shows a typical FTIR spectrum of 300/60, 500/60 and 700/ 60 SC BCs. It was observed that the total functional groups decreased when the pyrolysis temperature is increased. The peak assignments in the spectra represented methyl C H stretching compounds (2916 cm1), aromatic carbonyl/carboxyl C¼O (1699 cm1), aromatic C¼C and C¼O (1595 cm1), aliphatic COC and alcohol OH (1030 cm1), and aromatic C H 1 (815 cm ) [34,35]. These bands presented different changes when the pyrolysis temperature was increased, the observation is consistent with other studies [34,36]. Peaks between the

1000–600 cm1 interval are attributed to the aromatic C-H wagging vibration [33]. The polar groups (OH and C O) exhibited the lower magnitude of peaks upon heating at high temperature, suggesting a decrease in the polar functional groups with an increase in the pyrolysis temperature [34]. To compare the adsorption by the BCs, the second order derivatives of the FTIR spectra were obtained in order to give a better display of the peaks and their intensities (Fig. 1b). It was observed that, despite of showing the same peaks, the peak intensities were lower in the BCs after P adsorption, that was

Fig. 2. Effect of the adsorption of phosphate by SC and M BCs on the pH change in solution at various initial pH values.

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46

41

Fig. 3. Zeta potential of 500/60 M BC before and after phosphate adsorption.

especially evident for SC 700/20 BC, and the highest adsorption capacity was showed by this BC. Similar results were found by Zeng et al. [31] and by Halajnia et al. [37] as absence of peaks after adsorption and FTIR spectra did not show such a strong band for phosphate.

BCs, and the impact of specific adsorption increased with increased suspension pH. Tong et al. [44] and Xu et al. [41] found similar results working on Cu and methyl violet adsorption by crop straws BCs, respectively. 3.5. Adsorption of phosphate

3.3. Effect of initial pH The species of P and its stability and dominance depends on pH. As the pH rises, H2PO4 species begin to form at ca pH 4.0, and at pH 7 the concentration of H2PO4 and HPO42 are approximately equal [38,39]. It was also found that initially the pH value difference (pHe– pH0), increases between 3 and 4, and then decreases from pH 4 to pH 10, for SC as well as for M BC (Fig. 2). It indicates that ion exchange mechanisms are involved in the adsorption [24]. With an increase in pH, OH  competes strongly with phosphate for active sites, which affects the adsorption capacity. This implies adsorption is favored by low pH values and the capacity would be higher at low pH [40]. 3.4. Zeta potential

The data in Table 2 confirm that adsorption of P on BC was influenced by the phosphate concentration in solution, by the carbonization temperature and by the residence time. The surface area and the adsorption capacity increased with the carbonization temperature and the longer residence time (Fig. 4). There was no statistical difference for the adsorption from solution of each concentration by SC BC made under the conditions of 500/60, 700/20 and 700/60. Adsorption of phosphate in the 25, 50, 100 and 200 mg dm3 concentration range was statistically similar for M BCs. However, adsorption from 400 mg dm3 solution

Table 2 Mean values for the adsorption capacities (qe, mg g1) for phosphate at different concentrations on BCs from SC and M. Biochar

There was no significant difference between zeta potential values for SC and M BCs (F(11,22) = 1.02; p > 0.05). The zeta potential at pH 6.5 of the M BCs samples ranged from 26.6 to 36.3 mV, and of SC BCs samples from 28.4 to 34.9 mV, indicating that the BC particles carried negative charges on their surfaces. Although, the zeta potential of the 500/60 M BCs became more negative with increased pH, suggesting that the amount of negative charge increased with increased pH (Fig. 3), and adsorption should be lower as observed at Fig. 2. Similar and consistent results were found by Xu et al. [41], Inyang et al. [42], Yuan et al. [43]. The presence of P shifted the zeta potential pH curve of 500/60 M BC particles to positive values directions. The difference in zeta potential between the systems (500/60 M BC and 500/60 BC after sorption of P) raised with increased pH of the suspension. These results suggested that phosphate can be specifically adsorbed by

Initial P concentration, mg dm3 25

50

100

200

400

Sugar cane 300/20 0.63 300/60 2.84 500/20 3.81 500/60 5.32 700/20 5.66

dB cB cb B ab C ab C

1.27 5.28 5.99 6.98 8.11

b AB a AB aA a BC a BC

1.63 5.28 6.68 9.34 10.0

c AB b AB bA a AB a AB

2.28 6.08 7.32 10.3 11.0

cA bA bA aA aA

2.46 6.52 7.51 11.4 11.5

cA bA bA aA aA

Miscanthus 300/20 0.69 300/60 3.36 500/20 3.99 500/60 5.89 700/20 6.28 700/60 7.66

dD cC cC bE ab D aD

1.40 5.53 6.60 7.72 8.95 9.47

e CD dB cB bD aC aC

1.83 5.92 7.64 10.5 11.4 12.1

e BC d AB c AB bC ab B aB

2.58 7.08 8.55 12.0 12.8 13.2

c AB b AB b AB aB aA aB

3.01 7.75 9.00 13.6 13.7 15.5

dA cA cA bA bA aA

Means followed by the same letter, capital letters in rows and lowercase in columns, do not differ by Duncan’s test at 5% of probability.

42

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46

Fig. 4. Equilibrium isotherm plots at 20  C for phosphate sorption on: (a) SC BCs, and (b) M BCs. Solid and dashed lines represent the Langmuir and Freundlich isotherm data model, respectively.

concentration by the 700/60 BC preparation was statistically higher than for the others BCs. For each BC, the maximum adsorption was achieved when phosphate concentration in the solution was highest. However, for the 300/20, 300/60 and 500/20 SC BC samples the maximum concentration was not significantly different from that for the 50 mg dm3. No statistical differences were observed between the adsorptions of the 200 and 400 mg dm3 phosphate solution concentrations by M BCs 300/20, 300/60, 500/20 and 700/20.

The phosphate adsorption on the BCs samples were analyzed using the Langmuir and Freundlich isotherm models. The Langmuir isotherm assumes a uniform sorption at all of the binding sites, whereas the Freundlich isotherm suggests a heterogeneous surface with a nonuniform distribution of heat of adsorption over the surface [45]. The Langmuir model gave the higher R2 values, showing that the adsorption on SC and M BCs could be better described by that model (Table 3). Similar results were found by Gao et al. [39] and by Kilpimaa et al. [46] for other chars.

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46 Table 3 Langmuir and Freundlich isotherm parameters for the adsorption of phosphate on SC and M BCs. System

Sugar cane bagasse BC 300/20

300/60

500/20

1.676 4.114 0.784

500/60

700/20

700/60

Freundlich

KF n R2

0.177 2.143 0.907

2.499 4.968 0.801

2.971 4.205 0.958

3.615 4.755 0.898

4.754 5.865 0.973

Langmuir

qm KL R2

2.953 6.887 7.819 0.014 0.043 0.069 0.994 0.998 0.999 Miscanthus
gigantus BC

12.07 0.042 0.999

11.95 0.065 0.999

13.21 0.057 0.996

Freundlich

KF n R2

0.167 1.974 0.907

1.867 3.981 0.873

2.387 4.163 0.832

3.149 3.903 0.974

3.845 4.378 0.936

4.960 5.157 0.987

qm KL R2

3.754 0.011 0.995

8.244 0.037 0.998

9.425 0.051 0.999

14.47 0.036 0.998

14.31 0.054 0.999

16.10 0.046 0.995

Langmuir

Based on the results presented in Table 2 we can say that there is a relationship between the amount of phosphate adsorbed and the conditions under which a BC was made. The higher the carbonisation temperature and the longer the residence time the higher is the adsorption capacity, as the process was endothermic. However, the difference in the amounts of phosphate adsorbed is smaller for higher temperatures. This rises a question of optimal energy usage and its relation to the BC sorption capacity. The energy required for the pyrolysis process is a sum of the energy (QS) required to raise the temperature of the wet material to the boiling water temperature (Qd) plus the energy for water evaporation (Qw), and the carbonisation energy (Qc). The first two energies depend on the moisture content and are independent from the carbonisation temperature and are the same for each material. While the energy needed for carbonisation rises with the final temperature. We assume no energy loses by conduction or radiation to the surroundings during these processes, this simplification does not change the final conclusions. In terms of energy expended, the BC formed at 500  C for 60 min performed best in relation to phosphate sorption, as described in Fig. 5. The amount of phosphate adsorbed on SC as well as M BCs divided by

43

the sum of energy used for the BCs production was optimum for BC produced at 500 C for longer residence time. The adsorption of phosphate by different BCs is listed in Table 4. Our work shows that the amount of phosphate adsorbed depends on the carbonization process parameters, as well as on the initial phosphate concentration in solution. These parameters should be taken into consideration when comparing the sorption capacities for various BCs. The data presented in Table 4 indicate that none of the BCs listed had a significantly better sorption capacity. The largest difference presented in Table 4 is between two activated carbon samples made in relatively similar ways. Our work indicates that the method of preparation of BC samples has an important bearing on the sorption of phosphate. 3.6. Adsorption thermodynamics Table 5 presents the thermodynamic parameters for different temperatures. The positive values of Gibbs energy suggested that the adsorption of phosphate onto each BC was non-spontaneous. Additionally, the positive values of enthalpy changes indicated that the adsorptions were endothermic. The positive values of the entropy changes indicate increased randomness at the solid/ solution interface during the adsorption of phosphate onto these BCs [47]. Results similar to ours were obtained by Foo and Hammed [49] for activated carbon from palm oil fibres. A majority of researchers have also found phosphate adsorption on BCs to be endothermic [19,47,49,50] and random [19,47,49,50] However, in some cases [19,47,50] the Gibbs energy did not always fall with increasing temperature, indicating that phosphate was adsorbed efficiently at a high solution temperatures. The adsorption of phosphate increases with temperature, as indicated by Fig. 6. The data indicate that higher temperatures are favorable for phosphate adsorption onto these BCs. At higher temperature, the reactivity of the surface sites and the rate of intraparticle diffusion of sorptive ions into the pores of adsorbent increased [37,51]. The majority of researches confirm that phosphate sorption increases with temperature [19,37,39,47,49,50,52,53]. Opposite results were obtained for the sorption of phosphate on dolomite mineral [25,54] and on modified wheat residue [55], suggesting a tendency for the phosphate ions to escape from the solid phase to the bulk phase as the temperature of the solution increased.

Fig. 5. The amount of phosphate adsorbed divided by the energy required for the production of SC and M BCs under various conditions.

44

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46

Table 4 Phosphate adsorption capacities of different BCs formed under different conditions. Conditions

Material production

Pine sawdust BC Embauba Lacre Inga BCs Corn BC

Activated sugarcane bagasse BC

P Adsorption capacity, Refs. qe mg g1

126 219

[47]

388 495 382 490 – – 314

10 14 14 20 10 35 40 50 70 11 20 14

501 19.4

0.3 1.2

[18]

285

21

[46]

>350

0.06 to 1.11

[48]

P initial concentration mg dm3

550  C 750  C 500  C

for 15 min

300  C 600  C

for 3 h

100

500  C

for 30 min

61.3

30 40

Mg modified corn BC Oak sawdust BC La modified oak sawdust BC Ferrihydrite modified rice straw BC Orange peel BC Fe3+/Fe2+ modified orange peel BC Activated carbon residue BC

Surface area (N2) m2 g1

600  C

100 (pH 7) 2

700  C with ramp 5  C/min

downdraft gasifier 1000  C. kiln at 400  C for 2 days and activation at 900  C for 100 min

50 (pH 6) 1–30

[5]

[19]

[20] [16]

3.7. Desorption of P Table 5 Thermodynamic parameters at different temperatures for SC and M BCs at phosphate concentration of 100 mg dm3. Temperature ( C)

DG0

DS0

DH0

R2

Sugar cane

20 35 50

6.242 5.845 5.386

0.010

9.197

0.990

Miscanthus

20 35 50

5.861 5.289 4.982

0.013

9.509

0.959

Biochar

(kJ mol1)

(kJ mol1 K1)

(kJ mol1)

As for adsorption, the process for desorption of P was influenced by the P concentration in solution, by the carbonization temperature, and by the residence time. Results presented in Table 6 show that SC and M BC samples that adsorbed greater amounts of P, released more to the equilibrium solution during the desorption process. The phosphate desorbability can be defined as the ratio of the desorbed phosphate to the total adsorbed by the adsorbents, and it can be used to indicate the degree of phosphate desorption from the adsorbate [16]. In terms of the phosphate desorbed, there are no statistical differences for each P concentration between the percentage desorbed from the SC and M BCs (Fig. 7). For the SC BCs with the

Fig. 6. Effect of temperature on the sorption of phosphate from 50, 100 and 200 mg dm3 initial concentration of phosphate by 500/20 SC and M BCs.

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46 Table 6 Mean values of phosphate (%) desorbed from different concentrations of phosphate adsorbed on SC and M BCs. Biochar

Phosphate concentration, mg dm3 25

50

100

200

400

Sugar cane 300/20 19.6 300/60 20.6 500/20 22.1 500/60 20.9 22.1 700/20 700/60 22.4

a a a a a a

E D E E E E

25.4 26.4 26.2 27.5 26.3 29.2

a a a a a a

D C D D D D

29.0 28.9 30.7 32.6 31.1 32.5

a a a a a a

C C C C C C

36.3 34.2 36.6 38.0 38.1 38.3

a a a a a a

B B B B B B

44.4 44.8 43.5 45.3 46.3 45.6

a a a a a a

A A A A A A

Miscanthus 19.1 300/20 300/60 21.0 500/20 20.5 500/60 19.7 700/20 23.7 700/60 21.0

a a a a a a

D C C E E D

24.7 27.7 25.5 29.3 25.2 30.7

a a a a a a

D C C D D C

28.2 30.2 29.7 32.2 32.5 31.6

a a a a a a

C C B C C BC

36.9 32.7 37.8 38.3 40.5 36.3

a a a a a a

B B A B B B

44.5 48.1 40.7 44.5 48.3 47.2

a a a a a a

A A A A A A

Means followed by the same letter, capital letters in rows and lowercase in columns, do not differ by Duncan’s test at 5% of probability.

45

lowest and the highest initial P concentrations, the desorption was 19.6 to 22.4% (F(5,10) = 0.39; p > 0.05) and 43.5 to 46.3% (F(5,10) = 0.38; p > 0.05), respectively. Very similar results were observed for the M BC, with corresponding figures of 19.1 to 23.7% (F(5,10) = 0.92; p > 0.05), for the lowest initial phosphate concentration, and 40.7 to 48.3% (F(5,10) = 1.38; p > 0.05), for the highest. Results are consistent with those obtained by Cui et al. [16]; the rate of phosphate desorption from the BC complex is clearly dependent on the desorption conditions employed, the gradient of phosphate concentration, and the concentrations of various ions in solution. According to Morales et al. [5], the feedstock makes a significant difference in phosphate desorption properties and opens the possibility for designing BC preparations for specific soil management objectives. 4. Conclusions Increasing the carbonization temperature and residence time for biomass gave rise to incresed fixed carbon and surface area.

Fig. 7. Effect of the initial concentration on phosphate desorption from SC and M BCs.

46

P.A. Trazzi et al. / Journal of Environmental Chemical Engineering 4 (2016) 37–46

Miscanthus biochar has a higher surface area and higher phosphate adsorption capacity than sugar cane bagasse biochar. The greatest changing pH adsorption corresponded to the highest adsorption. Desorption was greatest in cases where most phosphate was adsorbed. Biochars formed at 500  C for 60 min performed best in relation to phosphate sorption in terms of energy expended. The experimental data fitted with the Langmuir isotherm marginally better than with the Freundlich isotherm, indicating a uniform phosphate sorption at all of the binding sites. Adsorption of phosphate on the biochars was an endothermic non-spontaneous process and the adsorption incresed with the temperature. Acknowledgements Paulo Trazzi thanks to Conselho Nacional de Desenvolvimento Científico (CNPq) for receiving his Post-Doctoral grant. References [1] L.H. Bohn, B.L. McNeal, G.A. O’Connor, Soil Chemistry, John Wiley and Sons Inc., New York, 2001, pp. 322. [2] A.C. Lua, T. Yang, J. Guo, J. Anal. Appl. Pyrolysis 72 (2004) 279. [3] K.A. Spokas, K.B. Cantrell, J.M. Novak, D.W. Archer, J.A. Ippolito, H.P. Collins, A.A. Boateng, I.M. Lima, M.C. Lamb, A.J. McAloon, R.D. Lentz, K.A. Nichols, J. Environ. Qual. 41 (2012) 973. [4] S.D. Joseph, M. Camps-Arbestain, Y. Lin, P. Munroe, C.H. Chia, J. Hook, L. van Zwieten, S. Kimber, A. Cowie, B.P. Singh, J. Lehmann, N. Foidl, R.J. Smernik, J.E. Amonette, Aust. J. Soil Res. 48 (2010) 501. [5] M.M. Morales, N. Comerford, I.A. Guerrini, N.P.S. Falcao, J.B. Reeves, Soil Use Manag. 29 (2013) 306. [6] A. Downie, A. Crosky, P. Munroe, Biochar for Environmental Management: Science and Technology, in: J. Lehmann, S. Joseph (Eds.), Earthscan, London, 2009, pp. 13. [7] A.V. Piterina, M.H.B. Hayes, Influence of the addition of herbaceous biochar on the metabolic profiles of the maize rhizosphere microbial community, 4th International Biochar Congress, Beijing, 2012, pp. 74. [8] G. Xu, J.N. Sun, H.B. Shao, S.X. Chang, Ecol. Eng. 62 (2014) 54. [9] T.T. Qian, X.S. Zhang, J.Y. Hu, H. Jiang, Chemosphere 93 (2013) 2069. [10] C.M. Monterumici, D. Rosso, E. Montoneri, M. Ginepro, A. Baglieri, E.H. Novotny, W. Kwapinski, M. Negre, Int. J. Mol. Sci. 16 (2015) 8826. [11] R.Z. Yin, R.H. Liu, Y.F. Mei, W.T. Fei, X.Q. Sun, Fuel 112 (2013) 96. [12] R. Chintala, T.E. Schumacher, L.M. McDonald, D.E. Clay, D.D. Malo, S.K. Papiernik, S.A. Clay, J.L. Julson, Clean-Soil Air Water 42 (2014) 626. [13] R.A. Brown, A.K. Kercher, T.H. Nguyen, D.C. Nagle, W.P. Ball, Org. Geochem. 37 (2006) 321. [14] J. Lehmann, Front. Ecol. Environ. 5 (2007) 381. [15] S. Joseph, C. Peacocke, J. Lehmann, P. Munroe, Biochar for Environmental Management: Science and Technology, in: J. Lehmann, S. Joseph (Eds.), Earthscan, London, 2009, pp. 107. [16] H.J. Cui, M.K. Wang, M.L. Fu, E. Ci, J. Soils Sediments 11 (2011) 1135. [17] M.M. Parvage, B. Ulen, J. Eriksson, J. Strock, H. Kirchmann, Biol. Fertil. Soils 49 (2013) 245. [18] B.L. Chen, Z.M. Chen, S.F. Lv, Bioresour. Technol. 102 (2011) 716.

[19] C. Fang, T. Zhang, P. Li, R.F. Jiang, Y.C. Wang, Int. J. Environ. Res. Public Health 11 (2014) 9217. [20] Z.H. Wang, H.Y. Guo, F. Shen, G. Yang, Y.Z. Zhang, Y.M. Zeng, L.L. Wang, H. Xiao, S.H. Deng, Chemosphere 119 (2015) 646. [21] American Public Health Association (APHA) Standard Methods, Method 4500P D, 21st Ed., 2005, Washington, D.C. [22] I. Langmuir, J. Am. Chem. Soc. 38 (1916) 2221. [23] H.F.M. Freundlich, Z. Phys. Chem. 57 (1906) 385. [24] D. Kolodynska, R. Wnetrzak, J.J. Leahy, M.H.B. Hayes, W. Kwapinski, Z. Hubicki, Chem. Eng. J. 197 (2012) 295. [25] W.X. Xiaoli Yuan, J. An, Jianguo Yin, Xuejiao Zhou, Wenqiang Yang, J. Chem. 853105 (2015) 1 ID. [26] P.N. Diagboya, B.I. Olu-Owolabi, D. Zhou, B.H. Han, Carbon 79 (2014) 174. [27] R. Brown, Biochar for Environmental Management: Science and Technology, in: J. Lehmann, S. Joseph (Eds.), Earthscan, London, 2009, pp. 127. [28] W.P. Song, M.X. Guo, J. Anal. Appl. Pyrolysis 9 (2012) 138. [29] A. Zabaniotou, G. Stavropoulos, V. Skoulou, Bioresour. Technol. 99 (2008) 320. [30] J. Guo, A.C. Lua, J. Anal. Appl. Pyrolysis 46 (1998) 113. [31] Z. Zeng, S.D. Zhang, T.Q. Li, F.L. Zhao, Z.L. He, H.P. Zhao, X.E. Yang, H.L. Wang, J. Zhao, M.T. Rafiq, J. Zhejiang Univ.-Sci. B 14 (2013) 1152. [32] J.C. Tang, W.Y. Zhu, R. Kookana, A. Katayama, J. Biosci. Bioeng. 116 (2013) 653. [33] Z.H. Zhou, J.H. Yuan, M.H. Hu, Environ. Progress Sustainable Energy 34 (2015) 655. [34] X.P. Gai, H.Y. Wang, J. Liu, L.M. Zhai, S. Liu, T.Z. Ren, H.B. Liu, PLoS One 9 (2014) . [35] K. Sun, K. Ro, M.X. Guo, J. Novak, H. Mashayekhi, B.S. Xing, Bioresour. Technol. 102 (2011) 5757. [36] B.L. Chen, D.D. Zhou, L.Z. Zhu, Environ. Sci. Technol. 42 (2008) 5137. [37] A. Halajnia, S. Oustan, N. Najafi, A.R. Khataee, A. Lakzian, Appl. Clay Sci. 80–81 (2013) 305. [38] K. Kaikake, T. Sekito, Y. Dote, Waste Manag. 29 (2009) 1084. [39] Y. Gao, N. Chen, W.W. Hu, C.P. Feng, B.G. Zhang, Q. Ning, B. Xu, J. Solution Chem. 42 (2013) 691. [40] Y.J. Xue, H.B. Hou, S.J. Zhu, J. Hazard. Mater. 162 (2009) 973. [41] R.K. Xu, S.C. Xiao, J.H. Yuan, A.Z. Zhao, Bioresour. Technol. 102 (2011) 10293. [42] M. Inyang, B. Gao, P. Pullammanappallil, W.C. Ding, A.R. Zimmerman, Bioresour. Technol. 101 (2010) 8868. [43] J.H. Yuan, R.K. Xu, H. Zhang, Bioresour. Technol. 102 (2011) 3488. [44] X.J. Tong, J.Y. Li, J.H. Yuan, R.K. Xu, Chem. Eng. J. 172 (2011) 828. [45] K.L. Wasewar, M. Atif, B. Prasad, I.M. Mishra, Clean-Soil Air Water 36 (2008) 320. [46] S. Kilpimaa, H. Runtti, T. Kangas, U. Lassi, T. Kuokkanen, Chem. Eng. Res. Des. 92 (2014) 1923. [47] F. Peng, P.W. He, Y. Luo, X. Lu, Y. Liang, J. Fu, Clean-Soil Air Water 40 (2012) 493. [48] M.N. Liang, H.H. Zeng, Y.N. Zhu, Z.L. Xu, H.L. Liu, Environmental Biotechnology and Materials Engineering, Pts 1–3, in: Y.G. Shi, J.L. Zuo (Eds.), Trans Tech Publications Ltd., Stafa-Zurich, 2011, pp. 1046. [49] K.Y. Foo, B.H. Hameed, Fuel Process. Technol. 99 (2012) 103. [50] C. Namasivayam, D. Sangeetha, J. Colloid Interface Sci. 280 (2004) 359. [51] R. Chitrakar, S. Tezuka, A. Sonoda, K. Sakane, K. Ooi, T. Hirotsu, J. Colloid Interface Sci. 290 (2005) 45. [52] W. Jutidamrongphan, K.Y. Park, S. Dockko, J.W. Choi, S.H. Lee, Environ. Chem. Lett. 10 (2012) 21. [53] L. Zeng, X.M. Li, J.D. Liu, Water Res. 38 (2004) 1318. [54] S. Karaca, A. Gurses, M. Ejder, M. Acikyildiz, J. Colloid Interface Sci. 277 (2004) 257. [55] X. Xu, B.Y. Gao, W.Y. Wang, Q.Y. Yue, Y. Wang, S.Q. Ni, Colloids Surfaces BBiointerfaces 70 (2009) 46.