Environmental Pollution 204 (2015) 306e312

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Effect of humic acid (HA) on sulfonamide sorption by biochars Fei Lian a, *, Binbin Sun a, b, Xi Chen b, Lingyan Zhu b, Zhongqi Liu a, Baoshan Xing c a

Agro-Environmental Protection Institute, Ministry of Agriculture, Tianjin 300191, China College of Environmental Science and Engineering, Nankai University, Tianjin 300071, China c Stockbridge School of Agriculture, University of Massachusetts, Amherst, MA 01003, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 March 2015 Received in revised form 11 May 2015 Accepted 12 May 2015 Available online

Effect of quantity and fractionation of loaded humic acid (HA) on biochar sorption for sulfonamides was investigated. The HA was applied in two different modes, i.e. pre-coating and co-introduction with sorbate. In pre-coating mode, the polar fractions of HA tended to interact with low-temperature biochars via H-bonding, while the hydrophobic fractions were likely to be adsorbed by high-temperature biochars through hydrophobic and p-p interactions, leading to different composition and structure of the HA adlayers. The influences of HA fractionation on biochar sorption for sulfonamides varied significantly, depending on the nature of interaction between HA fraction and sorbate. Meanwhile, co-introduction of HA with sulfonamides revealed that the effect of HA on sulfonamide sorption was also dependent on HA concentration. These findings suggest that the amount and fractionation of adsorbed HA are tailored by the surface properties of underlying biochars, which differently affect the sorption for organic contaminants. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Biochar Humic acid Fractionation Sulfonamide Sorption

1. Introduction Humic acid (HA), as an important dissolved organic matter (DOM), is ubiquitous in the environments, which shows complicated chemical compositions, reactivity, and varying molecular weights (Hyung and Kim, 2008; Karanfil et al., 1999; Pignatello, 2012). The special physicochemical properties including hydrophobicity, aromaticity and functionality make HA have strong affinity to environmental solids, e.g., minerals and black carbon, resulting in the formation of HA adlayers (Gu et al., 1994; Jonker et al., 2004; Kalbitz et al., 2005; Kasozi et al., 2010; Pignatello et al., 2006) and altering the surface properties. HA sorption is crucial for various biogeochemical processes in both terrestrial and aquatic environments. For example, sorption to minerals could increase the stabilization of HA by protecting it from biodegradation and thus preserve soil organic carbon (Kalbitz et al., 2005). On the other hand, the HA adlayer may change the physicochemical properties and reactivity of the coated surface and modify its sorption behavior to various contaminants (Jonker et al., 2004; Kasozi et al., 2010; Pignatello et al., 2006). The interactions

* Corresponding author. Institute of Agro-Environmental Protection, Fukang road 31, Nankai District, Tianjin 300191, China. E-mail address: [email protected] (F. Lian). http://dx.doi.org/10.1016/j.envpol.2015.05.030 0269-7491/© 2015 Elsevier Ltd. All rights reserved.

between HA and surfaces are regulated by diverse mechanisms such as hydrophobic adsorption (Kasozi et al., 2010) and electrostatic interactions (Armanious et al., 2014) which are greatly dependent on the surficial properties of the sorbents. Biochar (BC), intentionally produced by pyrolysis of carbon-rich biomass, has gained increasing research attention as a soil amendment for agricultural and environmental purposes (Jha et al., 2010). Owing to its high surface activity, BC contributes greatly to the overall sorption of organic compounds in the amended soil and thus alters the mobility and bioavailability of the chemicals. The presence of DOM including HA has been identified to influence the sorption of organic chemicals on BC in the past decades (Guo et al., 2007; Kwon and Pignatello, 2005). However, the roles of HA in the sorption process varied significantly, to a great extent depending on the properties of BC, sorbate, and adsorbed HA. To date, a systematic investigation on impact of interaction between HA and BC on chemical composition and structure of loaded HA and BC sorption for ionizable compounds remained missing in the literature. BC is commonly referred as a “combustion continuum”, which shows distinct physicochemical characteristics according to the charring temperatures (Keiluweit et al., 2010). The hightemperature (>600  C) BC develops high content of graphitic carbon (Zheng et al., 2013) while contains less polar domains. In contrast, the low-temperature ( CeO (19.3%) > COOH (3.92%). Noticeably, HA coating displayed different impacts on the compositions of surface oxyl groups for BCs300 relative to BCs600. Generally, the contents of CeO and COOH on BCs300 decreased and that of C]O increased after HA loading. However, the opposite trend was observed for BCs600 (Fig. S1). This suggested that the loading level and/or sorption mechanism of HA on BCs300 was different from that on BCs600. BCs300 had more Ocontaining groups than BCs600 and were likely to interact with polar components of HA via H-bonding, thus depleting plenty of hydroxyl and carboxyl groups. In comparison, BCs600 tended to interact with HA mainly through hydrophobic and p-p EDA interactions due to their more hydrophobic and graphitic surfaces (Kasozi et al., 2010; Pignatello et al., 2006). Xie et al. (2014) recently

reported that p-p interaction was a more important driven force relative to surface O-functionalities in HA sorption by hightemperature BCs. Thus, the relatively hydrophobic components of HA were preferentially adsorbed on carbon surface of BCs600, and leaving polar moieties (e.g., hydroxyl and carboxyl groups) on the external surface of HA adlayer. Owing to C]O group of HA could probably facilitate and/or involve in the formation of EDA interactions between HA and BCs600 (MacKay and Vasudevan, 2012), its content on the HA adlayer was relatively decreased. The FTIR results (Fig. 1) also shed light on the nature of sorptive interactions between HA and BCs. A number of strong peaks (e.g., 1580 cm1 for C]O (Zheng et al., 2013) and 1031 cm1 for CeO (Qian et al., 2013)) were identified in the spectrum of HA, indicating it was rich in oxyl groups. The FTIR spectra of two representative BCs (i.e., CR300 and CR600) before and after HA sorption were compared to examine the variations of surficial groups (Fig. 1). After coating with HA, the spectra of CR300 changed significantly and became more similar to that of free HA, indicating strong modification of HA to BCs prepared at lower temperature. Relative to original CR300,

(a)

O-H 3421

aromatic C=C, C=O 1580 -CH2-CH2C=O 1700 1380 2925 C-O 1436 1265 1031

1095

HA-CR300 HA CR300 4000

3415

3500

3000

O-H 3428

(b)

2000

1500

1000

C=C, C=O 1560

HA-CR600

CH2-

C-O 1072

500

C-H 876-789

1385

CR600

4000

3500

3000

2500

2000

1500

1000

500

-1

Wavenumber (cm ) Fig. 1. FTIR spectra of HA, corn straw biochars (CR300 and CR600) and their respective HA-loaded samples.

F. Lian et al. / Environmental Pollution 204 (2015) 306e312

the vibrations of carboxyl (1700 cm1) and C]O groups (1580 cm1) after adsorbing HA could be attributed to H-bonding mechanism. Further, the disappearance of CeO band (1031e1095 cm1) for HA after binding to CR300 also supported the interaction between HA and polar components of CR300. In the case of CR600, the increased peak intensity of OeH (3428 cm1) and CeO (1072 cm1) groups after loaded with HA was consistent with the observation by XPS that HA coating enhanced the polarity of BCs600. The shifts of aromatic C]C and C]O bands were indicative of hydrophobic and p-p interactions between HA and CR600. UV/vis spectra provided additional information for structural change of HA adlayer upon sorption to BC. Negative correlation was established between the ratio of absorbance at l ¼ 465 and 665 nm (E4/E6) and HA aromaticity in previous studies (Davis et al., 1999; Wang and Xing, 2005). Table S2 listed the E4/E6 values of the solution HA before and after sorption on CR300 and CR600. Relative to the free HA (E4/E6 ¼ 1.46), the E4/E6 of HA-BC complexes was reduced, indicating that there was relatively more aromatic structures for the complexes than free HA. With increasing the ratio of HA to BC, the E4/E6 value increased from 1.08 to 1.17 for HA-coated BCs300, while slightly decreased from 1.07 to 1.03 for HA-coated BCs600. This suggested preferential sorption of more aliphatic fractions to BCs300, while more aromatic fractions to BCs600. The combined results of XPS, FTIR and UV/vis reveal that the HA sorption to BCs resulted in fractionation of the humic materials, which was highly dependent on the surficial properties of BCs. Then, the structural and fractional changes of adsorbed HA could greatly affect SMX and SNA sorption on BCs (see more discussion below).

coated BCs are presented in Fig. 2. The sorption data were fitted to the Freundlich model, Q ¼ KfCne , by nonlinear regression, where Q (mg/kg) and Ce (mg/L) are the solid-phase and liquid-phase concentration of the solute, respectively, Kf (mg1nLn/kg) is the Freundlich sorption coefficient, and n (unitless) is the Freundlich linearity index. Freundlich model fitted well to the sorption isotherms especially for SMX (R2 > 0.97) in the experimental Ce ranges, and the fitting parameters are listed in Table 2. For original BCs, the isotherms of SMX on low-temperature samples were less nonlinear as suggested by the higher linearity index (n ¼ 0.74e0.79). As the pyrolytic temperature increased from 300 to 600  C, the n value (0.40e0.58) decreased, implying that more rigid and porous structures were developed, which was evidenced by the increased surface C content and N2-SBET (Table 1). In contrast, sorption of SNA on the BCs was generally more nonlinear than that of SMX with lower n values (0.42e0.65), suggesting a more heterogeneous distribution of the sorption sites for SNA. Sorption capacities of BCs were compared by distribution coefficient (Kd ¼ Q/Ce, L/Kg) according to the fitting results of Freundlich model (Table 2). Within the examined concentration ranges, SNA exhibited higher logKd values (1.60e3.63 L/Kg) than SMX (0.83e2.90 L/kg) on the BCs. Furthermore, it was noted that SMX and SNA exhibited different sorption patterns on original BCs (Fig. 2a, b). BCs300 had higher sorption capacity for SMX than BCs600. On the contrary, BCs600 showed higher sorption for SNA with the exception of CT600. It is well documented in the literature that sulfonamide antibiotics can interact with carbonaceous materials via various mechanisms including H-bonding with polar functional groups (Zheng et al., 2013), hydrophobic interaction (Lian et al., 2011), and p-p EDA interaction (Ji et al., 2009). Compared to BCs600, BCs300 had more O-containing groups (Table 1) and non-condensed structures. Thus, the sorption of SMX

3.2. SMX and SNA sorption by original BCs Sorption isotherms of SMX and SNA on the original and HA-

R300

R600

309

CR300

CR600

CT300

CT600

2000

(a) BC+SMX

(b) BC+SNA

4000

1600 3000 1200 2000

800

1000

Q (mg/kg)

400

0

0 0

10

20

30

40

50

60

70

0

10

20

30

40

50

60

1600 3000

(c) HA-coated BC+SMX

(d) HA-coated BC + SNA

2500

1200

2000 800

1500 1000

400 500 0

0 0

10

20

30

40

50

60

0

5

10

15

20

Ce (mg/L) Fig. 2. Sorption isotherms for SMX and SNA on original biochars (a, b) and HA-coated biochars (c, d).

25

30

310

F. Lian et al. / Environmental Pollution 204 (2015) 306e312

Table 2 Freundlich model parameters for SMX and SNA by original and HA-coated biochars. Biochar

SMX

SNA

Kf (mg1nLn/kg)

R300 HA-R300 R600 HA-R600 CR300 HA-CR300 CR600 HA-CR600 CT300 HA-CT300 CT600 HA-CT600

91.1 128 200 383 71.4 89.2 147 69.5 60.9 85.9 17.1 24.5

± ± ± ± ± ± ± ± ± ± ± ±

5.70 8.46 8.43 26.3 3.06 6.38 3.84 10.5 2.50 13.1 1.39 0.95

n

0.76 0.91 0.40 0.46 0.79 0.93 0.40 0.49 0.74 0.79 0.58 0.77

± ± ± ± ± ± ± ± ± ± ± ±

0.02 0.02 0.01 0.03 0.01 0.02 0.01 0.04 0.01 0.05 0.02 0.01

R2

logKd (L/kg)

Kf (mg1nLn/kg)

0.1 mg/L

10 mg/L

0.997 0.997 0.993 0.980 0.999 0.997 0.997 0.971 0.999 0.979 0.990 0.999

2.2 2.2 2.90 3.12 2.06 2.02 2.77 2.35 2.04 2.14 1.65 1.62

1.72 2.02 1.70 2.04 1.64 1.88 1.57 1.33 1.52 1.72 0.83 1.16

and SNA by BCs300 was mainly regulated by H-bonding with surficial polar components, which was supported by the positive correlation between LogKd values and surface polar carbon contents (sum of CeO, C]O, and COOH contents, Table 1) especially at high solute concentration (Ce ¼ 10 mg/L) (Fig. S2). Comparatively, BCs600 contained more carbonized phase while less O-containing groups. Both of SMX and SNA molecules are strong p-acceptors due to the N-heteroaromatic ring and sulfonamide group. Thus, the two test sulfonamides were expected to interact with the graphitic carbon surface (p-donor) via p-p interaction. Moreover, BCs600 were highly porous, pore-filling mechanism may also contribute to the sulfonamide sorption (Ji et al., 2011; Zheng et al., 2013). Consistently, the logKd values increased with increasing N2-SBET of BCs600 (Fig. S3). Sulfonamide sorption on carbonaceous sorbents is highly affected by their speciation (Zheng et al., 2013). SMX and SNA have various species in aqueous solution which is dependent on solution pH. In this study, more than 95% of SMX molecules (pKa, 2 ¼ 5.7) were negatively charged in the test pH (7.0 ± 0.2), resulting in remarkable increase of hydrophilicity and decrease of electron accepting ability compared to the neutral species. In comparison, SNA molecule (pKa, 2 ¼ 10.6) existed as neutral species at the test pH, and it could be more strongly adsorbed on the hydrophobic carbon surface of BCs600 relative to SMX. This well explained the facts that SMX displayed much less sorption capacity than SNA as well as it was less adsorbed by BCs600 than BCs300. Similar results were reported by Yao et al. (Yao et al., 2012) that BC obtained at 450  C showed better sorption ability for SMX than 600  C BC. In addition, compared to SMX, SNA has smaller molecular size due to the absence of a pyridine group (Table S1). Thus, pore-filling effect could be more pronounced for SNA and partly accounted for its higher Kd values relative to SMX. Electrostatic interactions could also affect the sorption of ionic compounds (Sassman and Lee, 2005), for which the surface charge is a key governing factor. The negative Zeta-potential values (Fig. S4) of CR300 and CR600, measured under the same condition in sorption experiments, indicated that the surfaces of original and HA-coated BCs were all negatively charged, consistent with previous studies (Xie et al., 2014; Zheng et al., 2013). The electrostatic repulsion between SMX and negatively charged BCs could greatly depress SMX sorption. However, SMX still displayed relatively strong sorption affinity to the BCs, indicating that other factors might facilitate its sorption. One of the possible mechanisms was the negative charge-assisted H-bond [()CAHB)], which is expected for compounds with a comparable pKa to surface O-containing groups (Li et al., 2013). Our previous study examined the pKa value (around 5.0) of a BC prepared at 250  C using an acid-base titration

461 67.9 1110 29.9 443 39.0 876 22.9 281 77.1 146 13.3

± ± ± ± ± ± ± ± ± ± ± ±

28.1 8.72 100 5.52 68.1 7.22 132 2.62 26.1 8.25 6.33 1.78

n

0.65 1.06 0.42 0.98 0.49 1.09 0.50 0.86 0.62 0.84 0.44 0.98

± ± ± ± ± ± ± ± ± ± ± ±

0.02 0.05 0.04 0.06 0.05 0.07 0.06 0.04 0.03 0.04 0.01 0.04

R2

logKd (L/kg) 0.1 mg/L

10 mg/L

0.994 0.991 0.953 0.982 0.931 0.985 0.921 0.991 0.986 0.989 0.994 0.991

3.01 1.77 3.63 1.50 3.17 1.50 3.44 1.50 2.83 2.05 2.72 1.14

2.31 1.89 2.47 1.46 2.14 1.68 2.44 1.22 2.07 1.73 1.60 1.10

method (Lian et al., 2014), which was similar to the second pKa (5.7) of SMX. Thus the [()CAHB)] would partly contribute to SMX sorption in the current study especially for BCs300 due to their more abundant O-containing functionalities. 3.3. SMX and SNA sorption by HA-coated BCs HA coating increased SMX sorption on BCs, however, the opposite trend was observed for SNA sorption (Fig. 2c, d), suggesting that HA adlayer has different effects for the two sulfonamides. HA coating introduced O-containing groups to the surface of BCs except for R600 (Table 1), which facilitated SMX sorption because H-bonding dominated the sorption process as mentioned above. On the contrary, hydrophobic interaction played a more important role in SNA sorption relative to SMX. The newly introduced polar functionalities made the HA-loaded BCs energetically less favorable for SNA sorption in comparison with original BCs. Thus, HA coating strongly reduced the sorption of SNA by BCs. Previous studies reported that HA coating could highly reduce the accessibility of HOC molecules to hydrophobic sites on carbon nanotubes and activated carbons (Wang et al., 2008; Zhang et al., 2011). Furthermore, it was noting that HA adlayer displayed different degrees on the sorption of sulfonamides (Table 2): (1) for SMX sorption, the enhancement on BCs600 was higher than that on BCs300 with HA coating. For example, the logKd value (Ce ¼ 10 mg/ L) by CT300 increased by 13.2% (from 1.52 to 1.72 L/kg), and that by CT600 increased by 39.8% (from 0.83 to 1.16 L/kg); (2) for SNA sorption, however, the decline on BCs600 owing to HA coating was greater than that on BCs300, where the logKd value (Ce ¼ 10 mg/L) by R300 decreased by 18.2% (from 2.31 to 1.89 L/kg), and that by R600 decreased by 50% (from 2.44 to 1.22 L/kg). The results suggested that HA coating exhibited a more striking impact on sulfonamide sorption for BCs600 than that of BCs300. As discussed above, the surface properties of the underlying BC to a great extent determined the amount and fractionation of adsorbed HA. The more increase of SMX sorption by BCs600 relative to BCs300 after HA-loading was mainly attributed to the more significant enhancement of surface polarity of BCs600 (Table 1). For SNA, however, hydrophobic and p-p interactions predominated its sorption on BCs600. Thus, the more significant decrease of SNA sorption on BCs600 can be explained by the fact that plenty of hydrophobic sites and micropores for SNA sorption were covered by HA coating. Additionally, the highly increased linearity indicator for SNA sorption (n, Table 2) suggested that hydrophobic partitioning into HA phase probably contribute to SNA sorption by the HA-coated BCs. Similarly, Yang et al. (Yang et al., 2009) observed

F. Lian et al. / Environmental Pollution 204 (2015) 306e312

that the adsorbed surfactants could form a partitioning phase for naphthalene on carbon nanotube surface. It was expected that the sorption of SNA on BCs600 was evolved from a surface-adsorption dominant to a partition dominant process owing to HA coating, leading to more significantly reduced sorption capacities relative to BCs300. These results suggest that the sorption is a complex combination of forces and bonds formation in the BC-HA-sulfonamide ternary system. 3.4. Co-introduction of HA and sulfonamide to BCs To further explore the role of HA in sulfonamide sorption on BCs, various concentrations (2.5e50 mg/L) of HA were co-introduced to BCs with SMX or SNA in the sorption systems. In general, SMX sorption on BCs declined when HA concentration increased from 2.5 to 30 mg/L, and then increased with further increasing HA concentration (Fig. 3). Moreover, a slightly elevated SMX sorption on BCs600 in the presence of HA at lower concentrations ( CR600 > R600 for BCs600, which was opposite to that of sulfonamide sorption on BCs (Fig. 2), indicating their different interactions with BCs. For HA sorption on BCs300, it was only conducted on CR300 (Fig. S5b) because the released DOM by other BCs300 severely interfered HA sorption. The more linearity for HA sorption on CR300 relative to BCs600 further revealed the homogeneity of binding sites on BCs300 surfaces and was probably attributed to their higher amount of volatile organic components and functional groups. On the other hand, the interaction between sulfonamides and HA molecule could be neglected when compared with their sorption to BCs. A preliminary experiment (Fig. S6) and previous studies (Hou et al., 2010; Pan et al., 2013) demonstrated that the negatively charged SMX slightly interacted with HA at neutral pH. These results suggested that the reduced SMX sorption on BCs at relatively low HA concentrations (30 mg/L) (Fig. 3). The slightly increased SMX sorption on BCs600 at very low HA concentrations (

Effect of humic acid (HA) on sulfonamide sorption by biochars.

Effect of quantity and fractionation of loaded humic acid (HA) on biochar sorption for sulfonamides was investigated. The HA was applied in two differ...
417KB Sizes 1 Downloads 13 Views