J Soils Sediments DOI 10.1007/s11368-015-1162-y

SOILS, SEC 1 • SOIL ORGANIC MATTER DYNAMICS AND NUTRIENT CYCLING • RESEARCH ARTICLE

Sorption characteristics and contribution of organic matter fractions for atrazine in soil Qianqian Wu 1 & Qi Yang 1 & Wenjun Zhou 1,2 & Lizhong Zhu 1,2

Received: 25 March 2015 / Accepted: 20 May 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstract Purpose Atrazine, as one of most widely applied herbicides in the world, caused significant concern due to long-term threat to the environment and risk for human health. Soil organic matter (SOM) is considered as one of the most important soil component controlling the transport and fate of contaminants in soil/water environments. Therefore, it is important to explore the sorption characteristic as well as distribution of atrazine among different SOM fractions in their respective soils. Materials and methods To further our understanding in this area, various SOM fractions (humic acid1 (HA1), humic acid2 (HA2), humin (HM)) were sequentially extracted using Na4P2O7 and NaOH from two kinds of agricultural soils, and batch sorption experiments for atrazine onto extracted SOM fractions were examined. Results and discussion Atrazine sorption isotherms were nearly linear and were well-fitted to the Freundlich equation with nonlinearity factors (n) ranging from 0.887 to 0.977. The sorption capacity of atrazine by the extracted SOM fractions followed the order HA1 > HA2 > HM, and the organic carbon normalized distribution coefficients (logKOC) were significantly related to (N+O)/C ratios of the extracted SOM fractions, demonstrating that the polarity of SOM fractions dominates atrazine sorption by polar interaction (e.g., H bonding). The relative contributions of different SOM fractions to the Responsible editor: Gabriele E. Schaumann * Wenjun Zhou [email protected] 1

Department of Environmental Science, Zhejiang University, Hangzhou, Zhejiang 310058, China

2

Zhejiang Provincial Key Laboratory of Organic Pollution Process and Control, Hangzhou, Zhejiang 310058, China

overall sorption of atrazine depend on their sorption capacities of atrazine and relative contents in the SOM, and the combined contribution of HA fractions in their respective soils was higher than that of HM although their lower content in SOM. Conclusions The results in this study indicated the important roles of SOM polarity and the HA fractions in the sorption of atrazine by the investigated soils. Keywords Atrazine . Contribution . Organic matter fraction . Polarity . Soil . Sorption

1 Introduction Atrazine is one of the most widely applied herbicides in the world (Short and Colborn 1999; Du Preeza et al. 2005; Postigo et al. 2010; Sun et al. 2010) and has frequently been detected as an organic contaminant in agricultural soils, groundwater, and surface water (Gerecke et al. 2002; Vonberg et al. 2014). It is considered a potential long-term threat to the environment because of its persistence (Jablonowski et al. 2011). The effects of atrazine on human health have been reported to include an increased risk of cancer (Lima et al. 2010), and atrazine can also act as an endocrine-disrupting compound (EDC) affecting the central nervous, endocrine, and immune systems (Sun et al. 2010). Therefore, atrazine has attracted extensive studies of its environmental behaviors, such as transport and fate (Cao et al. 2009; Wang and Keller 2009; Liu et al. 2010; Lima et al. 2010; Vonberg et al. 2014). Sorption onto soils is one of the most important processes that affect the transport and fate of contaminants in natural environments and determine their distribution, transport, bioavailability, and toxicity in soil/water environments. Moreover, soil organic matter (SOM) is considered as the most

J Soils Sediments

important soil component controlling the sorption and transport of hydrophobic organic compounds (HOCs) in soil environments. Numerous studies have been conducted to clarify the interaction between HOCs and SOM (Jeong et al. 2008; Chefetz and Xing 2009; Sun et al. 2013; Zhang et al. 2013), and it has been determined that the sorption of HOCs onto soil is significantly affected by the structural components of SOM (Kang and Xing 2005; Wen et al. 2007; Ran et al. 2007) such as humic acids (HAs) and humin (HM). Different SOM fractions have various impacts on the sorption of HOCs on soils (Führ et al. 1998; Jeong et al. 2008; Chefetz and Xing 2009; Zhang et al. 2013) depending on their chemical composition and structural properties such as polarity, aromaticity, and aliphaticity (Sun and Zhou 2008; Zhang and He 2010; Chien et al. 2010; Zhang et al. 2013). Therefore, it is essential to ascertain the sorption capacities and contributions of SOM fractions when investigating the sorption characteristic and evaluating the environmental risk of HOCs in soil environments. Because of the widespread occurrence of atrazine contamination in natural environments, it is crucial to explore the sorption behavior of atrazine on soil to assess its environmental risk and remediate its adverse effects. The sorption, leaching, and transport of atrazine on different soils have already been intensively studied, as well as the roles of different soil components in the sorption and desorption processes (Liu et al. 2010; Baez et al. 2013). Recently, atrazine sorption by humic substances, black carbon, and biochars were reported on separately (Cornelissen et al. 2005; Cao et al. 2009). However, few studies have been conducted on the sorption characteristics of individual SOM fractions (e.g., HAs and HM) in their respective soils and their quantitative contributions to the overall sorption of atrazine, which probably should be considered in understanding the sorption, transport, and fate of atrazine in soil environments and then assessing its environment risk. Therefore, it is necessary to further explore the sorption characteristic and distribution of atrazine among different SOM fractions in their respective soils, which may provide some basic data to understanding the sorption behavior of atrazine on the soils. In this study, the main SOM fractions of HAs and HM for two typical agricultural soils in China were sequentially isolated from their respective demineralized soils in the classical way, which have been used when we study the sorption characteristic of organic matters. Moreover, the sorption of atrazine herbicide by the extracted SOM fractions was examined. The objectives of this study were to further investigate the sorption characteristics of atrazine on SOM fractions and to quantify the relative contribution of SOM fractions to the overall sorption of atrazine. The results will be useful for

further understanding the sorption and transport behavior of atrazine in soils and assessing its environmental risk

2 Materials and methods 2.1 Chemicals Atrazine (2-chloro-4-(ethylamino)-6-isopropylamino-1,3,5triazine) was purchased from Acros Organics (USA) with a reported purity >97 %. All reagents used for the analytical determination of atrazine and organic matter fractions were of analytical or HPLC grade. 2.2 Soil samples and isolation of SOM fractions Paddy soil (PS) and black soil (BS), two typical agricultural soils in China, were collected from Zhejiang and Heilongjiang provinces, respectively. The total organic carbon (TOC) of the PS and BS samples was 1.79 and 2.97 %, respectively. A classical way was used to sequentially extract the various SOM fractions from the given soils, which has been reported by Gelinas et al. (2001) and Kang and Xing (2005). Briefly, each soil sample was first decarbonated with 1 M HCl, and then, the residue was separated and demineralized with 1 M HCl/10 % HF for 5 days (five times). After centrifugation (4000 rpm, 30 min), the solid residue was rinsed with deionized water and collected as the demineralized fraction (DM). Second, the humic substance was progressively extracted from the DM fraction. A predetermined amount of DM solid was extracted with 0.1 M Na4P2O7 seven times, and the supernatant after centrifugation was combined for acidification (pH=1.5 with 6 M HCl) to obtain the first HA fraction (HA1). Then, the precipitated residue was sequentially extracted seven times with 0.1 M NaOH, and the supernatant after centrifugation was combined and acidified to obtain the second HA fraction (HA2). Finally, the residue precipitated after the extraction of HAs was collected as HM. The SOM fractions of DM, HA1, HA2, and HM were freeze-dried and gently ground to pass through a 100-mesh sieve. 2.3 Sorption experiments All sorption experiments for atrazine were conducted in triplicate using a batch equilibration method. A certain amount of soil or SOM fraction was weighed into 22-mL Corex centrifuge tubes with Teflon-lined screw caps, and 20 mL of aqueous solution containing atrazine was then added to each tube. These tubes were equilibrated on a reciprocating shaker for 7 days at 20±1 °C and then centrifuged at 4000 rpm for 20 min. An

J Soils Sediments

(Thermo Scientific Nicolet 6700) in the wavelength band between 4000 and 400 cm−1.

appropriate aliquot of the supernatant was sampled for HPLC analysis. The sorbed concentrations of atrazine were calculated by the mass difference between the initial and equilibrated concentrations in the aqueous solutions. The data for atrazine sorption onto soils and SOM fractions were fitted to the Freundlich isotherm model: logqe ¼ logK F þ nlogC e

3 Results and discussion

ð1Þ

3.1 Content and properties of SOM fractions

where qe is the solid-phase concentration (mg/kg) and Ce is the aqueous-phase concentration (mg/L). KF is the sorption capacity-related parameter ((mg/kg)/(mg/L)n ), and n is the nonlinear coefficient. The organic carbon (OC) normalized Freundlich sorption coefficients (KFoc) were calculated by dividing the KF value by the OC content (f oc ) in the sample (K Foc = K F / f oc ), and the single-point OC-normalized distribution coefficients Koc (K oc = q e / (C e f oc )) were also calculated at different concentrations.

The relative yields of HA1, HA2, and HM were calculated on the basis of the TOC in the DM using following equation: f i ðwt%Þ ¼ ðTOCi Mi Þ=ðTOCDM MDM Þ

where fi is the relative yield of SOM fractions; Mi and MDM (g) are the weights of a given SOM fraction and DM fraction, respectively, and TOCi and TOCDM are the TOC contents of the SOM fraction and DM fraction, respectively. According to the results shown in Table 1, the relative yields of SOM fraction in the two soils both followed the order HM > HA1 > HA2. The HM fraction constituted 61.3 and 56.0 % of the TOC in the DM of PS and BS, respectively, and was much larger than the other SOM fractions, indicating that HM is the dominant fraction of SOM in the two soils. The relative yields of the HA2 fraction in the two soils were both less than 10 %, making HA2 a minor fraction in the SOM of the two soils. The sum of the OC contents of the HM, HA1, and HA2 fractions accounted for 96.14 and 87.4 % of the TOC in the DM of PS and BS, respectively, indicating that low OM loss occurred during the isolation of SOM fractions from the DM. The elemental composition and the atomic ratios of H/C, O/C, and (O+N)/C of the extracted SOM fractions are also given in Table 1. The O/C and H/C atomic ratios of the extracted SOM fraction in the two soils both followed the order HA1 > HA2 > HM and decreased with the sequential extractions. A higher H/C ratio may suggest a higher degree of

2.4 Analysis and characterization The concentrations of atrazine in aqueous solution were determined with an Agilent 1100 HPLC (USA) fitted with a UV detector and an Agilent Eclipse XDB-C18 column (4.6 mm× 15 cm, 5 μm) using methanol/water (v/v, 70:30) as the mobile phase at a flow rate of 1 mL/min. The UV wavelength for atrazine was set at 225 nm. The OC contents of the soil samples and SOM fractions were determined using a TOC analyzer (Shimadzu Model 5000A, Japan), and the elemental compositions (C, H, N, O) were analyzed using an elemental analyzer (Thermo Finnigan Flash EA 1112). The ash contents of the samples were measured by heating at 900 °C for 4 h. The FTIR spectra of the SOM fractions were obtained using a FTIR spectrometer

Table 1 Physicochemical properties of the analyzed soils and their various organic matter fractions

Sample

PS PS-DM PS-HA1 PS-HA2 PS-HM BS BS-DM BS-HA1 BS-HA2 BS-HM a

ð2Þ

Yielda C (%)

Elemental composition C (%)

H (%)

O (%)

N (%)

H/C

O/C

(O+N)/C

1.85 30.8 39.4 35.7 28.3 3.04 35.5 40.7 36.6 30.8

– 3.71 5.52 4.14 2.44 – 4.54 5.86 4.52 2.74

– 15.97 23.06 16.65 11.03 – 18.48 27.35 15.73 13.59

– 1.03 1.14 1.71 1.13 – 2.02 0.89 1.65 0.67

– 1.44 1.68 1.39 1.03 – 1.52 1.73 1.48 1.07

– 0.38 0.44 0.35 0.29 – 0.39 0.50 0.32 0.33

– 0.42 0.46 0.39 0.33 – 0.44 0.52 0.36 0.35

Ash (%)

48.8 30.9 41.8 57.1 39.5 25.2 41.5 52.2

Yield is calculated by dividing the organic carbon of each fraction by the demineralized material

– 100 28.4 6.44 61.3 – 100 23.9 7.50 56.0

J Soils Sediments

aliphaticity (Wen et al. 2007). All of the H/C atomic ratios were greater than 1.0, suggesting that the samples contain a great amount of original organic residues such as polymeric CH2, fatty acids, lignin, and cellulose (Zhang and He 2010). The polarity index [(O+N)/C] was found to be highest for the HA1 fraction and lowest for the HM fraction. Thus, the HA1 fraction exhibited the highest aliphaticity and polarity among the extracted SOM fractions. The ash content of the SOM fractions ranged from 25.2 to 57.1 % and increased with the sequential extractions, which suggests that considerable mineral residues were still present in the extracted SOM fractions even after the soil was treated with the acid. From the FTIR analysis of SOM fractions shown in Fig. 1, the peaks were assigned as follows: O–H stretching vibration of carboxylic C (3450 cm−1); C–H stretching vibration of aliphatic C (2928 cm−1, 2850 cm−1); C=O stretching vibration of carboxylic C (1740 cm−1); C=C stretching vibration of aromatic C (1640 cm−1); C–H bending in methyl groups (1380 cm −1 ); C–O stretching (1230, 1160, 1100, and 1058 cm−1); C–H stretching vibration of aromatic C (885 and 750 cm−1), and Si–O–Si stretching (466 cm−1) (Alice et al. 2000; Kang and Xing 2005). The absorbance at 1230, 1160, 750, and 466 cm−1 was stronger for HM than for the HA fractions, whereas that at 1058 cm−1 was weaker. The high

absorbance at 466 cm−1 for HM suggests that more SiO2 was enriched, and the high ash content of HM was primarily the result of the accumulation of minerals during the extraction process. 3.2 Sorption kinetics of atrazine by SOM fractions Kinetics experiments were conducted to establish the time required to reach sorption equilibrium, which is one of the most important characteristics in representing the sorption capability of sorbents. Figure 2 illustrates the time schedules for atrazine sorption onto the extracted SOM fractions of the two soils. The amounts of atrazine sorbed onto the extracted SOM fractions (DM, HA1, HA2, and HM) of PS and BS increased sharply in the first 10 h, then attained 75 % of the corresponding equilibrium sorption amounts within 48 h, followed by much slower progress toward an apparent equilibrium during the next 120 h, after which the maximum sorbed amount was observed. To further understand the kinetics of atrazine sorption onto SOM fractions, experimental data were fitted to three 500 400

750

1100 1058 1380 1640 1160

2925

qe (mg/kg)

466

3450

885

1230 1030

2850 1740

PS-HM

300 200 PS-DM PS-HA1 PS-HA2 PS-HM pseudo-second order

100

PS-HA2 PS-HA1

0 0

400

800

1200

1600

2000

2400

2800 -1

Wavenumbers (cm

3200

3600

50

100

150

200

250

t (h)

4000

700

)

600 466

750

1058

2360

1380

500

2928

qe (mg/kg)

1160 1230

3450

2850

1640

BS-HM

400 300 200

BS-HA2

BS-DM BS-HA1 BS-HA2 BS-HM pseudo-second order

100 BS-HA1

0 400

800

1200

1600

2000

2400

2800 -1

Wavenumbers (cm

)

Fig. 1 FTIR spectra of the SOM fractions from soils

3200

3600

4000

0

50

100

150

200

t (h) Fig. 2 Sorption kinetics of atrazine onto SOM fractions

250

J Soils Sediments Table 2 Sample

PS PS-DM PS-HA1 PS-HA2 PS-HM BS BS-DM BS-HA1 BS-HA2 BS-HM

Kinetic parameters for the sorption of soils and their SOM fractions Pseudo-first-order modela

Intraparticle diffusionb

Pseudo-second-order model

qmax (mg/kg)

k1(h-1)

R2

qmax(mg/kg)

k2(mg/kg/h)

R2

kint (mg/kg/h1/2)

c (kg/mg)

R2

5.70 208.80 318.33 373.23 186.62 18.03 207.85 512.03 399.17 299.50

0.77 0.07 0.26 0.14 0.07 1.30 0.13 0.08 0.17 0.12

0.957 0.924 0.795 0.968 0.934 0.887 0.856 0.803 0.898 0.881

6.12 256.41 434.78 400.00 222.22 20.01 250.00 625 454.5 277.78

0.189 0.000302 0.000252 0.000239 0.000386 0.0731 0.000407 0.000203 0.000499 0.000362

0.999 0.997 0.998 0.996 0.997 0.999 0.996 0.997 0.999 0.996

0.61 15.62 26.18 24.35 13.01 1.86 14.62 34.69 25.28 16.17

2.24 49.44 96.57 77.12 49.44 8.35 54.58 162.6 138.92 60.60

0.699 0.883 0.853 0.880 0.839 0.649 0.884 0.867 0.781 0.895

ln(qmax −qt)=lnqmax −k1tqt (mg/kg): sorbed quantity per unit mass of sorbent at time t; t(h): solid-solution contact time; qmax (mg/kg): maximum sorbed amount per unit mass of sorbent; k1(h−1 ): the rate constant of pseudo-first-order

a

qt =kintt1/2 +ckint (mg/kg/h1/2 ): the rate constant of intraparticle diffusion kinetics; c: constant related to the thickness of the boundary layer

t=qt ¼ 1= qmax k 2 Þ þ t=qmax 2

ð3Þ

where qt and qmax are the sorbed amount of sorbate at time t and the maximum sorbed amount at equilibrium, respectively, and k2 is the rate constant of the pseudo-second-order kinetic equation. The slopes and intercepts of the plots of t/qt versus t were used to calculate the rate constants qmax and k2, and the results are given in Table 2. The experimental data points correspond well with the simulated curves plotted with the calculated parameters of the pseudo-second-order model in Table 2. The pseudo-second-order model is based on the assumption that the rate-limiting step may be chemical interaction during the sorption process (Chen et al. 2014). As a result, the sorption of atrazine onto SOM fractions from aqueous solution may be primarily controlled by chemical sorption rather than physical sorption.

atrazine sorption by all of the sorbents fit the Freundlich model well, as indicated by the correlation coefficients (>0.99). PS-DM PS-HA1 PS-HA2 PS-HM

800

600

qe (mg/kg)

conventional kinetic models: a pseudo-first-order model, pseudo-second-order model, and intraparticle diffusion model. The fitted model parameters for the three kinetic models are listed in Table 2. The high correlation coefficients (R2 >0.995) indicate that the pseudo-second-order model (expressed via Eq. (3)) is more closely represented the sorption of atrazine onto SOM fractions.

400

200

0

0

2

4

6

8

10

12

14

16

18

12

14

16

18

Ce (mg/L) 1200 BS-DM BS-HA1 BS-HA2 BS-HM

1000

qe (mg/kg)

b

800 600

3.3 Sorption isotherms for atrazine on SOM fractions

400

The equilibrium sorption isotherms are important in determining the sorption capacity of atrazine by soils and their corresponding SOM fractions. The sorption isotherms and Freundlich model fitting parameters for atrazine sorption by the extracted SOM fractions are presented in Fig. 3 and Table 3, respectively. With the fitted parameters, the isotherms of

200 0

0

2

4

6

8

10

Ce (mg/L) Fig. 3 Sorption isotherms of atrazine on SOM fractions

J Soils Sediments Table 3 Fitting parameters for the Freundlich model and the single point KOC for samples

Sample

PS PS-DM PS-HA1 PS-HA2 PS-HM BS BS-DM BS-HA1 BS-HA2 BS-HM

Log Koc at Ce (mL/g)

Freundlich sorption Log KF ((mg/kg)/ (mg/L)n)

n

R2

Log KFOCa ((mg/kg-OC)/ (mg/L)n)

=0.005SWb

=0.05SW

=0.5SW

0.095±0.013c 1.573±0.035 1.871±0.028 1.688±0.007 1.454±0.021 0.516±0.017 1.644±0.003 1.949±0.007 1.743±0.019 1.568±0.016

0.828±0.022 0.878±0.011 0.907±0.016 0.964±0.019 0.887±0.028 0.856±0.029 0.848±0.008 0.941±0.023 0.977±0.021 0.895±0.013

0.997 0.998 0.998 0.995 0.998 0.998 0.999 0.999 0.999 0.974

1.827 2.084 2.276 2.135 1.995 2.033 2.094 2.340 2.179 2.079

1.962 2.179 2.255 2.163 2.091 2.146 2.213 2.386 2.197 2.161

1.790 2.057 2.228 2.127 1.978 2.002 2.061 2.327 2.174 2.056

1.618 1.936 2.163 2.091 1.865 1.858 1.909 2.268 2.150 1.952

a

KFOC =KF /foc and the focis the carbon content

b

SW: aqueous solubility (mg/L)

c

Standard deviation

The sorption isotherms of atrazine on the original bulk soils of PS and BS are relatively nonlinear with n values of 0.828 and 0.856 (Table 3), respectively, and the nonlinearity of atrazine sorption on the bulk soils is close to that on the corresponding DM fractions (n=0.878 or 0.848), implying that the nonlinear sorption of atrazine on the two soils are closely related to the sorption of the organic phase rather than the inorganic minerals (Ran et al. 2007; Zhang et al. 2013). The n values for atrazine sorption on the extracted SOM fraction ranged from 0.887 to 0.977 and followed the order of HA2 > HA1 > HM, along which the sorption isotherms of atrazine exhibited increasing linearity. In addition, the sorption of atrazine onto the HA1 and HA2 fractions displayed nearly linear isotherms with n values higher than 0.90, showing characteristics of solute partitioning (Chiou and Kile 1998; Chen et al. 2008; Sun et al. 2010), and the sorption of atrazine was more linear on the HA2 fraction than on any of the other SOM fractions. It has been suggested that the nonlinearity factor n can be taken as an index of site energy distribution, and it has been related to the heterogeneous glass, hard, or condensed SOM domain and to the maturation degree of SOM (Kang and Xing 2005; Ran et al. 2007; Sun and Zhou 2008). The lower the n value is, the more heterogeneous the sorption site energy distribution or the higher the degree of SOM maturation is (Sun and Zhou 2008). Therefore, the later-extracted HA2 fraction with high n values may have a more homogeneous structure or chemical composition than the earlier-extracted HA1 with small n values. The HM fractions in this study exhibited greater nonlinearity for atrazine sorption compared with the HAs fractions, which is consistent with previous results for the

sorption of HOCs on SOM (Kang and Xing 2005), in which a trend of increasing nonlinearity after studying phenanthrene sorption to sequentially extracted soil HAs and HM was observed. The high nonlinearity of HM for atrazine can be explained by the greater maturation of the HM fraction, which consisted of diagenetically mature kerogen and black carbon (BC), compared to the HA fractions (Nam and Kim 2002). 3.4 Sorption capability of SOM fractions for atrazine The logKFOC values for atrazine in the two bulk soils of PS and BS were 1.827 and 2.033, respectively. In the extracted SOM fractions of PS and BS, the logKFOC for atrazine ranged from 1.995 to 2.176 and from 2.079 to 2.340, respectively (Table 3), and followed the order of HA1 > HA2 > DM > HM. A high KFOC value generally indicates a high sorption capacity for atrazine. The HA1 fractions in the two soils both showed the highest KFOC value and thus sorption capacity for atrazine. However, a precise comparison cannot be made based on the KFOC values because the units differ as a result of nonlinearity (Kang and Xing 2005; Wen et al. 2007). Therefore, the concentration-dependent OC-normalized sorption coefficient K OC at three selected concentrations (C e = 0.005SW, 0.05SW, and 0.5SW, where SW is the water solubility of atrazine) was employed to compare the sorption capacity of SOM fractions for atrazine in this study. The logKOC value measured for atrazine decreases as a function of C e due to the isotherm’s nonlinearity (Table 3). Regardless of Ce levels, the atrazine logKOC values of the extracted SOM fractions, including HA1, HA2, and HM, were higher than that of their respective

J Soils Sediments

Moreover, the sorption capacity (logKOC) of the extracted SOM fractions for atrazine increased with the increasing H/C ratios in the sequence HA1 > HA2 > HM, further demonstrating that the SOM fractions with high H/C ratios or rich aliphaticity have a stronger sorption capacity for atrazine. This observation corresponds well with the results reported for phenanthrene sorption by several SOM fractions (Kang and Xing 2005; Ran et al. 2007; Wen et al. 2007). Piccolo et al. (1998) also reported that the aliphatic carbon content of HAs was one of the key factors regulating their sorption to atrazine. In addition to aliphatic carbons, the effect of SOM polarity on the sorption capacity should also be considered, which can significantly affect the sorption capacity of HOCs (Kang and Xing 2005). In our current study, the polar atomic ratios ((N+ O)/C) for the SOM fractions extracted from the two soils decreased with the sequential extraction and followed the sequence HA1 > HA2 > HM. A similar trend was also obtained for the sorption capacity (logKOC) of atrazine for the extracted SOM fractions. Therefore, the sorption capacity (logKOC) of atrazine is also significantly related to the polarity of the SOM fractions (Fig. 4), indicating that the polarity of the SOM 2.4

2

2.3

Ce=0.5SW

2

y=0.49x+1.40 R =0.920

2.2 2.1 2.0 1.9 1.8

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

H/C 2.4

b

2

Ce=0.005SW y=1.78x+1.42 R =0.938 2

Ce=0.05SW y=1.83x+1.37 R =0.900

3.5 Effects of fractionated SOM properties on atrazine sorption

2.3

Ce=0.5SW

2

y=1.93x+1.29 R =0.788

2.2

Log KOC

According to the results discussed above, different SOM fractions extracted from soils displayed various sorption characteristics for atrazine. To further investigate the effect of SOM properties on atrazine sorption and the corresponding sorption mechanisms, the sorption parameters of atrazine were related to the chemical composition and structural properties of SOM fractions. It was observed that the sorption coefficients logKOC at each of the three atrazine concentrations are significantly related to the H/C ratios or aliphatic carbon of the extracted SOM fractions (Fig. 3), revealing the importance of aliphatic groups for atrazine sorption.

a

2

Ce=0.005SWy=0.42x+1.56 R =0. 931 Ce=0.05SW y=0.44x+1.51 R =0.941

Log KOC

bulk soils and followed the order of HA1 > HA2 > HM, decreasing over the course of the extraction process. The HA1 fractions have the highest logKOC values and thus the highest sorption capacity for atrazine among the fractions, whereas the HM fractions have the lowest sorption capacity for atrazine (low logKFOC and logKOC). The KOC values (Ce =0.05SW) of the HA1 and HA2 fractions for atrazine were 1.78 and 1.41 times that of the HM fraction for PS and 1.87 and 1.31 times the HM fraction for BS, respectively, demonstrating the higher sorption capacity the HA fractions for atrazine compared to the HM fraction. Thus, the HA fractions may be important for the atrazine sorption of SOM on an OC basis. However, this observation does not correspond with previous results for the sorption of apolar organic pollutants onto SOM such as phenanthrene (Kang and Xing 2005; Wen et al. 2007; Sun and Zhou 2008), in which the HM fraction had a higher logKOC value and sorption capacity for phenanthrene than the HA fractions. One possible explanation for this discrepancy is that the sorption of atrazine, which is polar, by SOM fractions may follow a different mechanism from that of apolar HOCs due to structural and polarity differences. And also, some studies (Welhouse and Bleam 1993; Wang et al. 2011) have reported that H-bouding between atrazine and the acid functionalities of HAs is a key mechanism governing atrazine sorption onto SOM, while the contribution of the π-π interaction between phenanthrene and aromatic components in humin substances (HSs) is much stronger and important than others, such as van der Waals interactions and H bonding. The rich amount of O-containing functionalities of HAs and high containing of aromatic components in humins would cause the sorption of atrazine and phenanthrene followed opposite order. This description may be useful to explain the different tendency of phenanthrene and atrazine sorption. Moreover, the condensed domains of SOM would reduce the contact between atrazine and the O-containing moieties, resulting in a decreased sorption capacity for atrazine.

2.1 2.0 1.9 1.8 0.30

0.35

0.40

0.45

0.50

0.55

(N+O)/C Fig. 4 Correlation analysis of sorption parameters with SOM properties

J Soils Sediments

As the fact that different fractions play different roles in the overall sorption, the investigation of the sorption onto different organic fractions is efficient to quantify the relative contribution of each fraction to the overall sorption potential of atrazine by soils. Thus, atrazine risk assessment should consider the contribution of different fractions. To calculate the contribution of each SOM fraction to the overall sorption of atrazine, a similar approach described in the previous study was chosen (Ran et al. 2007) which has been also accepted by other authors in their article (Sun et al. 2010; Ran et al. 2007). Briefly, according to the extraction method utilized in this study, the SOM fraction of the DM was composed of three components: HA1, HA2, and HM. The relative contribution of each SOM fraction to the total sorption of DM was quantitatively evaluated using the following equations: qe ¼ Y HA1 K F;HA1 C e n;HA1 þ Y HA2 K F;HA2 C e n;HA2 þ Y HM K F;HM C e n;HM


f i ¼ Y i K F;i C e n;i =qe  100%

ð4Þ

ð5Þ

50

Contribution of SOM (%)

3.6 Contribution of SOM fractions to overall sorption

where qe is the sum of the atrazine sorbed onto each SOM fraction (mg/kg); YHA1, YHA2, YHM, and Yi are the relative yields of the corresponding fractions based on the OC of DM; KF,i is the Freundlich sorption coefficient of atrazine for SOM fraction i; and fi is the percent contribution of SOM fraction i to the total sorption. Therefore, the relative contribution of each SOM fraction depends on not only its sorption capacity but also its relative yield. The quantitative contributions of the HA1, HA2, and HM fractions to the total sorption of atrazine onto DM are shown in Fig. 5. It is clear that the fractions vary greatly in their contributions to atrazine sorption. Because of its extremely low yield (<10 %), the HA2 fraction of each soil type only accounted for a small percentage of atrazine sorption, with maximum sorption contributions of 10 and 15 % to the total sorption of atrazine by PS and BS, respectively. Therefore, the HA2 fraction made minor contributions to atrazine sorption by the two soils. The HA1 fraction exhibited the

45

40 10 PS-HA1 PS-HA2 PS-HM

5 0 0.0

0.2

0.4

0.6

0.8

Ce/SW 50

Contribution of SOM (%)

fractions has a significant effect on the sorption capacity of atrazine. It has also been reported that higher-polarity biopolymers have a higher sorption affinity for atrazine (Shechter and Chefetz 2008), and atrazine sorption by humic substances exhibited a positive correlation with the substances’ polar carbon contents (Wang et al. 2011). However, this result differs from the previous reports on the sorption of apolar HOCs such as phenanthrene and lindance by fractionated SOM, which described a negative relationship between SOM polarity and logKOC (Kang and Xing 2005; Wang et al. 2011). It is documented that the lone-pair electrons on the nitrogen atoms of atrazineare able to create polarity to form H bonding with the O-containing polar groups in SOM, and the amino nitrogen atoms of atrazine can also act as H-bonding acceptors with the adsorbed water molecule on the polar moieties of SOM (Wang et al. 2011). The positive correlation between the logKOC of atrazine and the polar index ((N+O)/C) of the extracted SOM fractions further demonstrates that the specific interaction of H bonding between atrazine and the Ocontaining moieties of SOM may be involved in the sorption of atrazine by SOM, and plays a key role in the sorption capacity of atrazine (Welhouse and Bleam 1993; Sun et al. 2010; Wang et al. 2011). Comparing Fig. 4a with Fig. 4b, the slopes of the correlation curves of atrazine logKOC with (N+O)/C ratios are higher than that with H/C ratios, indicating that SOM polarity had more significant effect on atrazine sorption than its aliphaticity, and polarity is the key factor regulating atrazine sorption in SOM.

45

40 10 BS-HA1 BS-HA2 BS-HM

5 0 0.0

0.2

0.4

0.6

0.8

Ce/SW Fig. 5 Contribution of each SOM fraction to total atrazine sorption

J Soils Sediments

800

qe (mg/kg)

600

400

200 PS-DM PS-Total

0 0.0

0.2

0.4

0.6

0.8

1.0

Ce/SW 1200 1000

qe (mg/kg)

strongest sorption capacity for atrazine among the three fractions, but the relative yields of HA1 from DM (28.4 and 23.9 %, respectively) were much lower than those of the HM fraction. The percent contribution of HA1 to atrazine sorption increased slightly from 44.5 to 45.5 % and from 43.6 to 44.0 % in PS and BS, respectively, as the aqueous concentrations of atrazine increased. Therefore, the HA1 fraction accounted for a large percentage of the atrazine sorption of the two soils and made relatively larger contributions to total sorption compared to the HA2 fraction. The percent contributions of the HM fraction to atrazine sorption were close to those of the HA1 fraction and decreased slowly from 48.0 to 44.1 % and from 46.6 to 42.0 % for PS and BS, respectively, with an increasing atrazine concentration. The high contribution of the HM fraction to atrazine sorption can be attributed to its high yields from the DM (61.3 and 56.0 % for PS and BS, respectively), which offset the fraction’s low sorption capacity for atrazine. The combined contribution of HA fractions (HA1 and HA 2) for atrazine sorption increased from about 52.0 to 55.9 % and from about 53.4 to 57.9 % in PS and BS, respectively. Therefore, HAs accounted for higher contribution to atrazine sorption than HM. This result also differs from the sorption of apolar HOCs onto SOM, in which the condensed HM fraction plays the key role. The different results can be attributed to the different effect of SOM polarity on the sorption of polar atrazine and apolar HOCs (Nam and Kim 2002). The total sorbed atrazine on the three extracted fractions (HA1, HA2, and HM) was compared with that on the DM (Fig. 6). The total sorption amount of atrazine on the three fractions was greater than that on the DM, which became more obvious as the aqueous concentration of atrazine increased. Thus, the SOM fractions exhibit much higher sorption capacities for atrazine when they are isolated. Previous research has found that inorganic mineral components may form complicated aggregate structures with SOM, affecting the sorption capacity of SOM in soils (Ran et al. 2007). With the removal of the inorganic components, some formerly occupied sorption sites on the SOM were freed, and the accessibility of the sorption sites on the isolated SOM fractions was also increased, both of which enhanced the sorption capacity of the SOM fractions. In addition, the spatial arrangement and physical makeup of the SOM would be altered by the extraction process, which may cause some hidden sorption sites to become exposed, significantly increasing the available sorption sites for atrazine (Kovaios et al. 2011). It may not fully reflect the nature condition in this way, but it can be used as a reference

800 600 400 200 0 0.0

BS-DM BS-Total

0.2

0.4

0.6

0.8

1.0

Ce/SW Fig. 6 The comparison between the total sorption of atrazine by SOM fractions and that by the DM

to assessing environmental risk of atrazine to a certain extent.

4 Conclusions In the present study, the sorption of atrazine in different SOM fractions isolated from two typical soils was investigated and the relative contribution of each extracted fraction to the overall sorption of atrazine in DM was quantified and evaluated. The sorption capacities (logKOC) of atrazine by the extracted SOM fractions were positively correlated with the atomic ratios H/C and (N+O)/C, and the polarity of SOM fractions showed more significant effect on atrazine sorption by polar interactions (e.g., H bonding). The relative contributions of different SOM fractions to the overall sorption of atrazine depend on their sorption capacities of atrazine and relative contents in the SOM, and the combined contribution of HA fractions in their respective soils was higher than that of HM although their lower content in SOM. The results of this study demonstrate that SOM polarity and the HA fractions play an

J Soils Sediments

important role in atrazine sorption, which differs from the sorption of nonpolar organic compounds like phenanthrene onto SOM, and then can enhance our understanding of the sorption and transport behavior of atrazine in soils. Acknowledgments This study was financially supported by the Zhejiang Provincial Natural Science Foundation of China (LZ12B07001), the National Basic Research Program of China (2014CB441103), the National Natural Science Foundation of China (21137003), and the Key Project in Science and Technology of Zhejiang Provincial, China (2013C03024).

References Alice B, Csanad S, Mandolna H (2000) Comparison of Plioceneorganicrich lacustrine sediment in Twin Craters. Org Geochem 31:453–461 Baez ME, Fuentes E, Espinoza J (2013) Characterization of the atrazine sorption process on andisol and ultisol volcanic ash-derived soils: kinetic parameters and the contribution of humic fractions. J Agric Food Chem 61:6150–6160 Cao XD, Ma LN, Gao B, Harris W (2009) Dairy-manure derived biochar effectively sorbs lead and atrazine. Environ Sci Technol 43:3285– 3291 Chefetz B, Xing BS (2009) Relative role of aliphatic and aromatic moieties as sorption domains for organic compounds: a review. Environ Sci Technol 43:1680–1688 Chen CP, Zhou WJ, Yang Q (2014) Sorption characteristics of nitrosodiphenylamine (NDPhA) and diphenylamine (DPhA) onto organo-bentonite from aqueous solution. Chem Eng J 240:487–493 Chen BL, Zhou D, Zhu LL (2008) Transitional adsorption and partition of nonpolar and polar aromatic contaminants by biochars of pine needles with different pyrolytic temperatures. Environ Sci Technol 42:5137–5143 Chien SWC, Chen CY, Chang JH (2010) Sorption of toluene by humic acids derived from lake sediment and mountain soil at different pH. J Hazard Mater 177:1068–1076 Chiou CT, Kile DE (1998) Deviations from sorption linearity on soils of polar and nonpolar organic compounds at low relative concentrations. Environ Sci Technol 32:338–343 Cornelissen G, Gustafsson O, Bucheli TD (2005) Extensive sorption of organic compounds to black carbon, coal, and kerogen in sediments and soils: Mechanisms and consequences for distribution, bioaccumulation, and biodegradation. Environ Sci Technol 39:6881–6895 Du Preeza LH, Jansen van Rensburga PJ, Joostea AM (2005) Seasonal exposures to triazine and other pesticides in surface waters in the western Highveld corn production region in South Africa. Environ Pollut 135:131–141 Führ F, Burauel P, Dust M, Mittelstaedt W, Putz T, Reinken G, Stork A (1998) Comprehensive tracer studies on the environmental behavior of pesticides: the lysimeter concept. Lysimeter Concept. ACS Symp Ser 699:1–20 Gelinas Y, Prentice KM, Baldock JA (2001) An improved thermal oxidation method for the quantification of soot/graphitic black carbon in sediments and soils. Environ Sci Technol 35:3519–3525 Gerecke AC, Scharer M, Singer HP (2002) Sources of pesticides in surface waters in Switzerland: pesticide load through waste water treatment plants-current situation and reduction potential. Chemosphere 48:307–315 Jablonowski ND, Schäfferand A, Burauel P (2011) Still present after all these years: persistence plus potential toxicity raise questions about the use of atrazine. Environ Sci Pollut Res 18:328–331

Jeong S, Wander MM, Kleineidam S, Grathwohl P, Ligouis B, Werth CJ (2008) The role of condensed carbonaceous materials on the sorption of hydrophobic organic contaminants in subsurface sediments. Environ Sci Technol 42:1458–1464 Kang SH, Xing BS (2005) Phenanthrene sorption to sequentially extracted soil humic acids and humins. Environ Sci Technol 39:134–140 Kovaios ID, Paraskeva CA, Koutsoukos PG (2011) Adsorption of atrazine from aqueous electrolyte solutions on humic acid and silica. J Colloid Interface Sci 356:277–285 Lima DLD, Schneider RJ, Scherer HW, Duarte AC, Santos EBH, Esteves VI (2010) Sorption-desorption behavior of atrazine on soils subjected to different organic long-term amendments. J Agric Food Chem 58:3101–3106 Liu YH, Xu ZZ, Wu XG (2010) Adsorption and desorption behavior of herbicide diuron on various Chinese cultivated soils. J Hazard Mater 178:462–468 Nam K, Kim JY (2002) Role of loosely bound humic substances and humin in the bioavailability of Phenanthrene aged in soil. Environ Pollut 118:427–433 Piccolo A, Conte P, Scheunert I, Paci M (1998) Atrazine interactions with soil humic substances of different molecular structure. J Environ Qual 27:1324–1333 Postigo C, López de Alda MJ, Barceló D (2010) Analysis and occurrence of selected medium to highly polar pesticides in groundwater of Catalonia (NE Spain): an approach based on on-line solid phase extraction−liquid chromatography−electrospray-tandem mass spectrometry detection. J Hydrol 383:83–92 Ran Y, Sun K, Yang Y, Xing BS, Zeng E (2007) Strong sorption of Phenanthrene by condensed organic matter in soils and sediments. Environ Sci Technol 41:3952–3958 Shechter M, Chefetz B (2008) Insights into the sorption properties of cutin and cutan biopolymers. Environ Sci Technol 42:1165–1171 Short P, Colborn T (1999) Pesticide use in the US and policy implications: a focus on herbicides. Toxicol Ind Health 15:240–275 Sun HW, Zhou ZL (2008) Impacts of charcoal characteristics on sorption of polycyclic aromatic hydrocarbons. Chemosphere 71:2113–2120 Sun K, Gao B, Zhang ZY, Zhang GX, Zhao Y, Xing BS (2010) Sorption of atrazine and phenanthrene by organic matter fractions in soil and sediment. Environ Pollut 158:3520–3526 Sun K, Ran Y, Yang Y, Xing BS (2013) Interaction mechanism of benzene and phenanthrene in condensed organic matter: importance of adsorption (nanopore-filling). Geoderma 204:68–74 Vonberg D, Vanderborght J, Cremer N, Pütz T, Herbst M, Vereecken H (2014) 20 years of long-term atrazine monitoring in a shallow aquifer in western Germany. Water Res 50:294–306 Wang P, Keller AA (2009) Sorption and desorption of atrazine and diuron onto water dispersible soil primary size fractions. Water Res 43: 1448–1456 Wang XL, Guo XY, Yang Y, Tao S, Xing BS (2011) Sorption mechanisms of phenanthrene, lindane, and atrazine with various humic acid fractions from a single soil sample. Environ Sci Technol 45: 2124–2130 Welhouse GJ, Bleam WF (1993) Atrazine hydrogen-bonding potentials. Environ Sci Technol 27:494–500 Wen B, Zhang JJ, Zhang SZ (2007) Phenanthrene sorption to soil humic acid and different humin fractions. Environ Sci Technol 41:3165– 3171 Zhang YL, Ran Y, Mao JD (2013) Role of extractable and residual organic matter fractions on sorption of phenanthrene in sediments. Chemosphere 90:1973–1979 Zhang JH, He MC (2010) Effect of structural variations on sorption and desorption of phenanthrene by sediment organic matter. J Hazard Mater 184:432–438