effects of rice straw–derived dissolved organic matter on pyrene sorption by soil

Environmental Toxicology and Chemistry, Vol. 29, No. 9, pp. 1967–1975, 2010# 2010 SETAC

Printed in the USADOI: 10.1002/etc.253

EFFECTS OF RICE STRAW–DERIVED DISSOLVED ORGANIC MATTER ON PYRENE
SORPTION BY SOIL
JIANGMIN ZHOU,*y HUALIN CHEN,y and WEILIN HUANGzySchool of Life and Environmental Sciences, Wenzhou University, Wenzhou 325035, Zhejiang, China

zDepartment of Environmental Sciences, Rutgers, The State University of New Jersey, New Brunswick, New Jersey 08901-8551, USA

(Submitted 23 November 2009; Returned for Revision 5 March 2010; Accepted 13 April 2010)

* T(rosech

Pub(wileyo

Abstract—The objectives of the present study were to elucidate the chemical and structural properties of dissolved organic matter(DOM) derived from aerobic decay of rice straw and to quantify the effect of the DOM on the sorption of pyrene on soil. The DOMsamples were obtained from microcosms incubated at 0, 21, 63, and 180 d. The bulk DOM samples were fractionated to four fractions:hydrophilic matter (HIM), acid-insoluble matter (AIM), hydrophobic acid (HOA), and hydrophobic neutral (HON) fractions. The bulkDOM and the four DOM fractions were characterized for their elemental compositions and functionalities. The results showed that HIMhad the highest H/C atomic ratios, whereas HOA and AIM had the lowest H/C atomic ratios. These DOM samples were used as thebackground DOMs in the initial aqueous solutions for measuring sorption of pyrene on a paddy soil. The results indicated that, among thefour DOM fractions, HOA, HON, and AIM significantly lowered the pyrene sorption coefficients, but HIM had little or no effect on thepyrene sorption by the soil. It appears that less polar AIM and HON had stronger binding affinities for pyrene in water, reducing thesorption coefficient for the soil, whereas more polar and less aromatic HIM had much weaker binding affinity for pyrene in water,causing little or no effect on the pyrene sorption by the soil. The present study showed that rice straw–derived DOM may enhancedesorption and transport of organic pollutants in soil–water systems. Environ. Toxicol. Chem. 2010;29:1967–1975. # 2010 SETAC

Keywords—Dissolved organic matter Heterogeneity Elemental composition Pyrene sorption

INTRODUCTION

Returning rice straw to the paddy field after harvest is apopular strategy for farming management. A major advantageof this practice is that it increases the soil’s organic carboncontent, hence improving the fertility of the soil and enhancingfixation of both organic and inorganic constituents in soils[1–3]. However, in addition to yielding higher organic carboncontent in soils, microbially facilitated decomposition of thedisposed rice straw also releases elevated levels of dissolvedorganic matter (DOM) in soil water, causing adverse environ-mental impacts in surface aquatic systems. Research on agri-cultural soils has shown that the addition of manure increasedthe concentration of water-extractable organic carbon [4,5].Dissolved organic matter is widely present in aquatic environ-ments and is a key constituent affecting the fate and transport ofhydrophobic organic contaminants (HOCs) in aquatic and soilenvironments [6–8]. It has been shown that DOM increases theapparent water solubility of various HOCs and promotes theirmobilization from soils and sediments [9,10]. Several priorstudies have shown facilitated transport of anthracene, phenan-threne, and other organic pollutants through different types ofsoils in the presence of DOM [11–13]. The magnitude of theenhanced aqueous concentrations and speed of transport of theorganic pollutants depend on the type and concentration ofDOM as well as on the characteristics of the contaminants[14,15].

The objectives of the present study were to characterize theheterogeneous properties of DOM derived from rice straw and

o whom correspondence may be [email protected]).

lished online 19 May 2010 in Wiley Online Librarynlinelibrary.com).

1967

to quantify the effect of the DOM on the equilibrium sorption oforganic pollutants in soil–water systems. According to ourrecent study [16], rice straw–derived DOM is highly heteroge-neous in chemical composition. It comprises a mixture ofpolysaccharides, polyphenols, proteins, lipids, and heteroge-neous molecules. Accurate determination of DOM structures isvery difficult because of this complexity of the chemical andstructural compositions. As a result, operational fractionation ofDOM into different fractions is an effective approach for betterunderstanding different physicochemical properties of wideranges of DOM [17]. Thus far, resin adsorbents have oftenbeen used in prior studies to fractionate DOM to hydrophobicand hydrophilic fractions; the latter have been further fractio-nated to acid, neutral, and base subfractions [18]. Although it isvery time consuming, this preparative DOM fractionationappears to be instrumental in evaluating environmental reac-tivities of DOM. Polubesova et al. [19] fractionated DOMderived from sewage sludge into hydrophobic acid (HOA)and neutral (HON) and found that pyrene binding was higherfor HON because of the combined effects of greater hydro-phobicity, aromaticity, and the large molecular size of thisfraction relative to HOA. Greater pyrene binding to HONcompared with HOA has been reported previously for thebinding of pyrene and other polycyclic aromatic hydrocarbonswith DOM fractions isolated from treated wastewater [20].

The present study was designed to assess how chemicallyand structurally heterogeneous DOM derived from rice strawimpacts the distribution of organic pollutants between soil andwater. Bulk DOM samples were obtained at four differentincubation times using a microcosm technique described byChen et al. [16], and we fractionated each DOM into fivedifferent fractions. These DOM fractions along with the bulkDOMs were characterized for their elemental and spectroscopicproperties. They were used as background dissolved organic

1968 Environ. Toxicol. Chem. 29, 2010 J. Zhou et al.

carbon in batch sorption experiments. The measured equili-brium sorption properties were found to relate closely to theproperties of the DOMs.

MATERIALS AND METHODS

Rice straw and soil sample

A rice straw sample was taken in the autumn of 2006 afterharvest from a paddy field in the suburban area (N2880.1280,E120847.6460) of Wenzhou, Zhejiang Province, China. The ricestraw is the hybrid rice species most commonly grown in China.The sample was air dried in the laboratory, ground, passedthrough a 2-mm sieve, and stored in a glass jar at 48C. The strawconsisted of 27, 33, and 9% weight hemicelluloses, cellulose,and lignin, respectively, and had 15 and 16% weight of extract-able organic matter and ash, respectively. A topsoil sample wastaken at depths of 0 to 20 cm from the same paddy field. Thesample had 21.4 g kg�1 organic carbon, 120 mg kg�1 dissolvedorganic carbon, 13.4 mEq 100 g�1 cation exchange capacity,and pH 5.92. The soil was homogenized in a glass container andwas divided into two subsamples in the field. A subsample wasair dried in the laboratory, ground, passed through a 0.25-mmsieve, and stored in a glass jar at 48C for characterization andsorption experiments, and the other was stored at 48C for theinoculation preparation.

Inoculation

The inoculum used for degradation of rice straw was pre-pared following a procedure described previously [21,22]. Inbrief, the wet topsoil sample was incubated for 4 d at 258C toreactivate soil microorganisms without addition of water ornutrients. The soil was then suspended in a flask at a soil-to-water (w/w) ratio of 1:45. After being shaken for 2 h at roomtemperature, the flasks were left to settle. The supernatant wasfiltered through 5mm, and the filtrate was retained as theinoculum. An aliquot of the filtrate was filtered again through0.45mm and analyzed for DOC content with a liquid TOC (totalorganic carbon) II analyzer (Elementar). According to the

Fig. 1. Flowchart of the fractionatio

DOC data (�1.0 mg L�1), the inoculum was diluted with MiliQ(Millipore) water to yield DOC less than or equal to0.1 mg L�1).

Preparation of DOM solutions

A total of four different DOM solutions were prepared at258C under aerobic and constant-moisture conditions and overfour different time periods of straw decay. Glass bottles(250 ml) without caps were used as the batch reactors, eachof which contained 3 g rice straw, 30 g precleaned quartz sand,5 ml of the inoculum, and 10 ml MilliQ water. During incuba-tion, the contents of each bottle were mixed intermittently byhand once every other day. All bottles were weighed every otherday, and the weight loss resulting from water evaporation wascompensated for by addition of MilliQ water. The reactors runat day 0 had not been mixed, representing the initial extractableDOM associated with the rice straw. At each time interval of 0,21, 63, and 180 d, five replicate reactors were taken forextraction of DOM. Each reactor was filled immediately with60 ml MilliQ water, capped, and shaken on a shaker at 200 rpmfor 2 h. The contents of the five reactors were combined andtransferred to centrifuge bottles (250 ml) and centrifuged at13,000 g for 30 min. The supernatant was withdrawn from eachbottle and filtered through a 0.45-mm membrane filter to obtainthe bulk DOM solution. The filtrate was split to two subsamples;one was used for fractionation and characterization of DOM andthe other for DOC measurements and subsequent sorptionexperiments.

Fractionation of DOM

A modified procedure of Leenheer [17] was used to frac-tionate the bulk DOM into five different DOM fractions. Asschematically shown in Figure 1, the first step of the fractio-nation procedure was to acidify the bulk DOM to pH 2.0 withHC1 (0.1 mol L�1), and the precipitate, which is defined asthe acid-insoluble matter (AIM), was obtained by centrifuga-tion. The solid AIM was then dissolved with 50 ml NaOH(0.1 mol L�1). The acidic supernatant was neutralized with

n of dissolved organic matter.

Pyrene sorption by soil Environ. Toxicol. Chem. 29, 2010 1969

NaOH (0.1 mol L�1) to near neutral and passed through a DAX-8 resin column (60 ml, 2 cm diameter� 20 cm height) atapproximately 30 bed volumes/h to separate the operationallydefined hydrophobic bases (HOB) by adsorption. The columnwas then rinsed with 2.5 bed volumes of Milli-Q water, and theeffluents collected from these procedures were combined forfurther fractionation of the three DOM fractions in the follow-ing steps. The HOB adsorbed on the DAX-8 column was theneluted using 0.25 bed volumes HCl (0.1 mol L�1). In the secondstep, the combined effluents obtained as described above wereacidified to pH 2.0 with HCl (0.1 mol L�1). The acidic super-natant was pumped through a second DAX-8 (Supelco, USA)column at 30 bed volumes h�1, followed by 1 bed volume ofHCl (0.1 mol L�1). The effluents were combined and neutral-ized with NaOH (0.1 mol L�1) to obtain the hydrophilic matter(HIM). The organic matter sorbed on the resin was extractedsequentially to obtain two different DOM fractions. First, theresin was extracted with 0.25 bed volumes NaOH (0.1 mol L�1),followed by 1.5 bed volumes distilled water. The effluents werecombined to obtain the hydrophobic acids (HOA). Second, theresidual organic matter on the resin was Soxhlet extracted with100 ml high-performance liquid chromatograph (HPLC)-grademethanol for 24 h to recover the hydrophobic neutral (HON)fraction. The methanol solution was evaporated at 408C on arotary evaporator to dryness. The solid HON was then dissolvedwith 100 ml MiliQ water.

All five DOM fractions (HOB, HOA, HON, HIM, and AIM)and the bulk DOM collected as aqueous solutions were desaltedwith an electrodialysis system. The obtained fraction and bulkDOM solutions were analyzed for TOC. The results showed thatthe four DOM fractions (HOA, HON, HIM, and AIM) hadsufficiently high TOC contents for further characterization andsorption study. However, the HOB fraction had insufficientmass for further studies. Each of the four fractions and bulkaqueous DOM solutions was divided into two subsamples. Onesubsample was used for examining the role of different DOMfractions in the sorption of pyrene on a soil. The other sub-sample was freeze dried, and the solid DOM samples were usedfor the analyses of carbon, hydrogen, and nitrogen contents withan EA-1112 elemental analyzer (Thermo-Fisher) and ash con-tent with a combustion method (5508C for 4 h). The oxygencontent in each DOM fraction sample was calculated by massdifference. Fourier-transform infrared (FTIR) spectra of theDOM samples were obtained using KBr pellets (1–2 mg sampleand 200 mg KBr) on an FTIR spectrophotometer (Equinex 55;Bruker Optics).

Sorption experiments

To quantify the effect of different DOM fractions on thesorption of organic pollutants on soil, a series of sorptionisotherms was measured using a batch reactor system andwith different initial aqueous solutions. The same soil sampledescribed above for preparation of the inoculum was used hereas the sorbent, and pyrene, a four-ring polynuclear aromatichydrocarbon, purchased from Sigma-Aldrich Chemical Com-pany, was used as the sorbate. The initial aqueous solutionscontained CaCl2 (0.01 mol L�1) as the background electrolyte,200 mg L�1 of sodium azide (NaN3) to inhibit bacterial growth,either no DOM or 200 mg L�1 (TOC) of the bulk DOM or theDOM fractions isolated above. Each of these backgroundaqueous solutions was spiked with a pyrene methanol solution(200 mg L�1) to make a series of the final aqueous solutions atpyrene concentrations ranging from 10 to 120mg L�1.

The batch reactors used were 35-ml glass vials, each ofwhich contained 0.2 g soil and 30 ml initial aqueous pyrenesolutions concentrations. The vials were sealed and shaken for24 h in the dark at constant temperature (208C). A preliminarykinetic experiment over a time period of 72 h indicated that 24 hwas sufficient for achieving apparent sorption equilibrium.After 24 h of mixing, the vials were centrifuged at a constanttemperature of 208C at 800 g for 20 min. After centrifugation,the vials were set upright in the dark for 12 h, and an aliquot ofthe supernatant was taken from each vial for analysis of theaqueous phase pyrene concentrations (Ce in mg L�1) via anHPLC method described below. The pyrene concentrations onthe soil phase (qe in mg kg�1) were calculated with a massbalance equation for the two-phase batch system. For all thesorption experiments, triplicate reactors were used for eachisotherm data point. Batch reactors with no soil were set up andrun similarly for assessing the loss of pyrene during mixing,centrifuging, and sampling procedures. The results indicated nomeasureable solute loss, so no correction was made for thecalculated qe data.

Both initial (Co) and final (Ce) aqueous phase pyrene con-centrations were quantified on an Agilent 1100 serials HPLCsystem equipped with a diode array and multiple wavelengthdetectors (DAD). A hypersil reversed-phase ODS-C-18 column(Agilent; 4.6� 150 mm) was used for separation. The mobilephase used was a mixture of water (5%) and methanol (95%) ata flow rate of 1.00 ml min�1. Pyrene was detected at an ultra-violet wavelength of 330 nm and a reference wavelength of360 nm. External pyrene standards in methanol matrix rangingfrom 10 to 120mg L�1 were used for quantification.

RESULTS AND DISCUSSION

Characterization of DOM

The relative contents of the five DOM fractions isolated fromeach bulk DOM obtained at different time intervals are pre-sented in Figure 2. As the figure shows, the DOM is composedmainly of AIM, HIM, and HOA. From 0 to 180 d, the content ofAIM increased from 7.9 to 63.9% on the total DOC base,whereas the content of HIM decreased from 76.3 to 9.2%.The content of HOA increased from 10.3% at 0 d to 30.4% at 63d and then decreased to 20.9% at 180 d. The relative content ofHON varied within a small range (4.6–9.8%), whereas HOBhad the lowest content (0.4–1.4%). The change trend of eachfraction with decay time was different from the composition ofthe DOM previously extracted from composted municipal solidwaste [23], which were fractionated into six fractions (hydro-phobic acid, hydrophobic base, hydrophobic neutral, hydro-philic acid, hydrophilic base, and hydrophilic neutral), and thetotal hydrophilic components increased from 30% at 47 d to40% at 161 d. The difference may be due largely to the differentsource material of DOM.

Table 1 lists the elemental compositions for the four bulkDOM samples and the four fractions (HOA, HON, HIM, andAIM) fractionated from each of the four bulk DOM samplesobtained at incubation times of 0, 21, 63, and 180 d. No data areshown in the table for the HOB at all four decay times and theAIM at 0 d because of their insufficient mass obtained duringfractionation. According to Table 1, among the four bulk DOMsamples, the one at 0 d of decay had the lowest carbon contentand the greatest oxygen and hydrogen contents, whereas theother three bulk DOM samples had similar contents on each ofthe three elements. The hydrogen to carbon (H/C) atomic ratiosfor the bulk DOM samples decreased as a function of straw

Fig. 2. Relative contents of fractions in bulk dissolved organic matter(DOM). The abbreviations represent acid-insoluble matter (AIM),hydrophobic bases (HOB), hydrophilic matter (HIM), hydrophobic acids(HOA), hydrophobic neutrals (HON). Bars represent standard errors.


decay time, suggesting that the bulk DOM derived at laterstages might have more aromatic structures.

Among the four DOM fractions fractionated from the bulkDOM, the HIM fraction, the dominant DOM in the bulk DOMof earlier stages (0 and 21 d), had the lowest relative contents ofC and N and highest relative contents of H and O, yielding thehighest H/C and O/C atomic ratios, suggesting its distinctaliphatic nature. Both AIM, the dominant DOM in the bulkDOM of the later stages (63 and 180 d), and HOA had low H/Catomic ratios, indicating dominant aromatic structural compo-nents. The HON had the higher carbon content than HIM, andits high H/C atomic ratios (1.70–1.88) suggest coexistence of

Table 1. Elemental compositions of d

Straw decay time (d)

Elemental composition (weight %)

N C H

DOM0 3.0 (0.08) 47.4 (1.9) 7.0 (0.09)21 3.8 (0.01) 53.3 (0.1) 5.9 (0.01)63 3.2 (0.05) 51.9 (0.4) 5.8 (0.01)180 2.6 (0.25) 53.0 (0.2) 5.4 (0.01)

HIM0 0.7 (0.10) 40.6 (0.66) 7.0 (0.16)21 1.8 (0.10) 44.2 (0.21) 8.4 (0.06)63 2.3 (0.05) 40.3 (0.31) 8.1 (0.08)180 2.4 (0.02) 48.6 (0.27) 7.6 (0.01)

HOA0 2.9 (0.06) 47.5 (0.13) 6.9 (0.01)21 2.5 (0.16) 49.8 (0.08) 6.0 (0.03)63 2.8 (0.05) 54.3 (0.06) 5.7 (0.02)180 2.8 (0.02) 53.2 (0.15) 5.5 (0.01)

HON0 3.0 (0.35) 51.9 (0.33) 8.8 (0.07)21 4.0 (0.11) 50.1 (0.23) 6.9 (0.03)63 2.8 (0.02) 53.0 (0.28) 7.5 (0.04)180 2.8 (0.15) 56.2 (0.25) 8.8 (0.03)

AIM21 3.6 (0.00) 54.2 (0.10) 6.1 (0.02)63 3.0 (0.04) 55.4 (0.13) 5.6 (0.01)180 2.3 (0.03) 56.4 (0.52) 5.2 (0.02)

a Standard deviation (�s); HIM¼ hydrophilic matter; HOA¼ hydrophobic acids

aliphatic with aromatic components. The polarity, defined as theatomic ratio of (NþO)/C, was highest with HIM (0.68–0.97)and lowest with HON (0.47–0.65) and AIM (0.52–0.56), whichindicates that the HON and AIM are less polar than HIM.

As straw decay proceeds, different DOM fractions evolvedifferently as a function of time. Table 1 shows that both HIMand HON fractions obtained at different time intervals retainedsimilar elemental compositions, suggesting that they mighthave similar structural components at different stages of strawdecay. Conversely, both HOA and AIM exhibited decreases inrelative hydrogen contents and hence decreasing H/C atomicratios as a function of decay time, suggesting that these DOMfractions become enriched in aromatic structures at the laterstages. At the decay time of 180 d, both the bulk DOM and AIMhad similar elemental compositions, because AIM becamedominant in the bulk DOM.

It should be noted that the DOM isolated in the present studyhad elemental compositions slightly different from those of theDOMs extracted from both soils and surface water. Accordingto Dilling and Kaiser [24], the DOMs extracted from three forestsoils have carbon and oxygen contents of 42.0 to 45.8% and48.5 to 52.2%, respectively, and O/C and H/C atomic ratios of0.79 to 0.93 and 0.98 to 1.17, respectively. These data arecomparable to those for the bulk DOM of the last stage (180 d)of straw decay but differ from those of the bulk DOMs of earlydecay (Table 1). Ma et al. [25] used XAD-8 resin and wetchemical procedures to fractionate a bulk DOM sampleobtained from surface water into three DOM fractions, includ-ing fulvic acid (FA), humic acid (HA), and hydrophilic acid(HyI). They found that HyI had the highest H/C atomic ratios(1.34–1.78), and FA and HA had H/C atomic ratios of 1.04 to1.12 and 0.98 to 1.07, respectively. Grossl and Inskeep [26]reported H/C and O/C atomic ratios of 1.12 and 0.48, respec-tively, for an aqueous FA isolated from land-disposed wheatstraw. According to Table 1, the HOA and AIM fractions

issolved organic matter (DOM)a

Atomic ratio

O H/C O/C (NþO)/C

42.6 (2.0) 1.77 (0.09) 0.67 (0.08) 0.73 (0.08)37.0 (0.1) 1.33 (0.00) 0.52 (0.00) 0.58 (0.00)39.1 (0.5) 1.34 (0.01) 0.56 (0.01) 0.62 (0.01)39.0 (0.5) 1.21 (0.00) 0.55 (0.00) 0.59 (0.01)

51.8 (0.93) 2.07 (0.27) 0.95 (0.18) 0.97 (0.18)45.7 (0.37) 2.27 (0.04) 0.77 (0.01) 0.81 (0.01)49.3 (0.44) 2.40 (0.08) 0.92 (0.04) 0.97 (0.02)41.5 (0.30) 1.88 (0.02) 0.64 (0.01) 0.68 (0.01)

42.7 (0.20) 1.74 (0.01) 0.67 (0.00) 0.73 (0.00)41.8 (0.27) 1.43 (0.01) 0.63 (0.01) 0.67 (0.00)37.1 (0.13) 1.25 (0.00) 0.51 (0.00) 0.56 (0.00)38.4 (0.18) 1.25 (0.01) 0.54 (0.00) 0.59 (0.00)

36.4 (0.74) 2.04 (0.00) 0.53 (0.01) 0.57 (0.01)38.9 (0.38) 1.66 (0.00) 0.58 (0.01) 0.65 (0.01)36.6 (0.34) 1.70 (0.02) 0.52 (0.01) 0.56 (0.01)32.2 (0.44) 1.88 (0.02) 0.43 (0.01) 0.47 (0.01)

36.1 (0.12) 1.35 (0.01) 0.50 (0.00) 0.56 (0.00)36.0 (0.17) 1.22 (0.00) 0.49 (0.00) 0.53 (0.00)36.2 (0.57) 1.11 (0.02) 0.48 (0.01) 0.52 (0.01)

; HON¼ hydrophobic neutrals; AIM¼ acid-insoluble matter.


obtained in this study had atomic ratios of H/C and O/C similarto the FA.

Figure 3 presents the FTIR spectra for the bulk DOMsamples of the four decay times and the fractionated DOMfractions. Each spectrum had a very broad absorption peakin high frequencies with a maximum at 3422 cm�1 resultingfrom adsorbed water and �OH stretching of organic functionalgroups. Additional bands can be identified and assigned to thevibration of carboxyl groups (1,723 cm�1). The aliphatic C–Hstretching and bending bands can be identified at high (2,925and 2,930 cm�1) and low (1,383 and 1,391 cm�1) frequencies,respectively. The peaks at 1,078 cm�1 (or 1,088 cm�1) and1,050 cm�1 usually result from �O� and C–O stretching vib-rations, respectively [27]. According to Figure 3, among thefour bulk DOM samples, DOM at 0 d is characterized by strong

Fig. 3. Fourier-transform infrared spectra of the five fractions of bulk dissolved orghydrophobic acids (HOA), and hydrophobic neutrals (HON) extracted from straw

absorbance at 1,078 cm�1, 1,383 cm�1, and 1,723 cm�1, indi-cating that carbohydrate and carboxylic acids might be dom-inant organic compounds. At 21 d, the bulk DOM had strongpeaks at 1,049 cm�1, 1,078 cm�1, and 1,383 cm�1 resultingfrom sugars and aliphatic components. After 63 d of decay,the peak at 1,049 cm�1 was diminished and the peak at1,078 cm�1 shifted to 1102 cm�1, suggesting increased contentof lignin-derived constituents. At 180 d, the DOM had peaksat 1,648 cm�1 and 1,512 cm�1 resulting from stretching vibrantof C––C in aromatic rings and an even stronger peak at1,424 cm�1 resulting from aliphatic C-H deformation. Thisindicates greater contents of aromatic carbon constituents inthe bulk DOM at the later stages of rice straw decay.

Figure 3 shows noticeable differences in FTIR spectraamong the four different DOM fractions. First, HIM was

anic matter (DOM), acid-insoluble matter (AIM), hydrophilic matter (HIM),at the decay of 0 d (a), 21 d (b), 63 d (c), and 180 d (d).


characterized by a high intensity of vibration bands assignedto carboxyl vibration bands (1,723 cm�1) and the strongabsorption peaks at 1,050 cm�1 and 1,113 cm�1 resultingfrom C–O stretching vibration, indicating carbohydratesand carboxylic acid structures. The rich carboxylic acid indi-cates high contents of polar functional group in HIM.The intensity of the main peaks for the samples obtained atdifferent stages was similar, suggesting that the HIM fractionswere composed of similar components at different decaytimes. The HOA was characterized by strong absorbance at1,711 cm�1 (carboxyl vibration), 1,647 cm�1 (stretchingvibrant of C––C in aromatic rings), and 1,040 cm�1 (C–Ostretching of polysaccharides and polyols), indicating higharomaticity and polarity. This result is similar to the pro-perties of HOA reported previously [19], suggesting thatHOA may have components similar to those of the fulvicacid isolated from groundwater [28]. The increased intensityof peak at 1,647 cm�1 with decay time indicates increasedaromaticity of HOA at the later stages. Both HON and AIMhad similar FTIR spectra characterized by well-resolvedpeaks at 2,926 cm�1 (aliphatic C–H stretching), 1,646 cm�1

(stretching vibrant of C––C in aromatic rings), 1,462 cm�1 (C–H deformations and aromatic ring vibrations), and 1,050 cm�1

(C–O stretching vibrations). The dominant peaks were at2,926 and 1,462 cm�1, indicating low polarity and high aro-maticity. The intensity of the peak at 1,648 cm�1 increasingwith decay time indicated the accumulated aromatic compo-nents in HON and AIM at later stages. The only differencebetween HON and AIM is that AIM had a higher intensity ofpeaks at 1,646 and 1,462 cm�1, suggesting that the aromaticityof AIM is greater than that of HON.

Fig. 4. Sorption isotherms of pyrene on soil in the absence of dissolved organic mainsoluble matter (AIM), hydrophilic matter (HIM), hydrophobic acids (HOA), and hyat 0 d (a), 21 d (b), 63 d (c), and 180 d (d).

Pyrene sorption on soil

The final sorption data (Ce and qe) measured for eachaqueous solution system were fit to a linear isotherm equationhaving the following form:

qe ¼ KD Ce;

where KD is the sorption distribution coefficient (L kg�1). Theorganic carbon content normalized distribution coefficient (KOC)was calculated from the following equation:

KOC ¼ KD=fOC;

where fOC is the mass fraction of the total organic carbon contentin the soil. The pyrene sorption isotherms are plotted in Figure 4,and the corresponding reduced sorption parameters (KD andKOC), along with their standard deviations and R2 values, aresummarized in Table 2. According to Table 2 and Figure 4, thesorption isotherms are reasonably linear, and the linear equationfits the sorption isotherms well. This likely is due to the fact thatthe range of aqueous pyrene concentrations within which thesorption isotherms were measured is narrow and close to thesolubility limit of pyrene (135mg L�1) [29]. Note that the KOC

values were calculated for the convenience of discussion. Suchcalculations were based on an assumption that the added DOMwas not adsorbed on the soil so that the fOC of soil in the secondequation remained constant for all the reactor systems tested inthe present study. This assumption likely is not valid, becausesoil-bound organic matter may dissolve in water, whereas theadded DOM may adsorb to soil particles.

The KD and KOC data presented in Table 2 and the sorptionisotherms shown in Figure 4 indicate that the sorption of pyreneon the soil was variously lowered in the presence of DOM and

tter (CK) and in the presence of bulk dissolved organic matter (DOM), acid-drophobic neutrals (HON). Numbers indicate the days of the rice straw decay

Table 2. The partition coefficients of pyrene on soil in the presence ofdissolved organic matter (DOM)a

DOM fraction KD (L kg�1) R2 Log KOC (L kg�1)

CK 1,538� 84 0.985 4.54� 0.030 d

DOM 816� 52 0.975 4.27� 0.04HIM 2,146� 66 0.995 4.69� 0.02HOA 1,380� 31 0.990 4.49� 0.01HON 174� 21 0.978 3.59� 0.08AIM 214� 10 0.997 3.68� 0.03

21 dDOM 185� 9 0.985 3.62� 0.03HIM 1,038� 46 0.990 4.37� 0.03HOA 563� 83 0.974 4.10� 0.03HON 112� 12 0.987 3.40� 0.07AIM 138� 7 0.989 3.49� 0.03



a HIM¼ hydrophilic matter; HOA¼ hydrophobic acids; HON¼ hydro-phobic neutrals; AIM¼ acid-insoluble matter; CK¼ sorption of pyreneon soil in the absence of dissolved organic matter.


DOM fractions and that the magnitude of reduction in sorptionwas highly dependent on the type of DOM fractions. Accordingto Table 2, the pyrene KD value is 1,538 L kg�1 in the absence ofthe isolated DOM, and the corresponding log KOC value is4.54� 0.03, which is close to the value (4.70) reported bySchlautman and Morgan [29]. In the presence of the bulkDOM, the KD values were lowered by approximately 47%(816 L kg�1) for 0 d DOM to 94% (94 L kg�1) for 180 dDOM. This suggests that the bulk DOM obtained in the laterstages of straw decay has a strong impact on the observedpyrene sorption on soil.

The data presented in Table 2 also show that different DOMfractions isolated from the same bulk DOM sample have verydifferent effects on the sorption behavior. In the four DOMfractions, the presence of AIM, HOA, and HON in the back-ground solution lowered the KD values. For example, in thepresence of these fractions isolated from the bulk DOM at 0 d,the KD values were lowered by approximately 86% (AIM), 10%(HOA), and 89% (HON), respectively. Similar results wereobserved for the same DOM fractions isolated from the bulkDOM of the later stages of straw decay, but the magnitudes ofthe KD reduction were greater. For the fractions isolated fromthe bulk DOM at 180 d, the KD values were lowered by 98, 75,and 96% in the presence of AIM, HOA, and HON, respectively.HIM has different effects on the pyrene sorption. In the presenceof HIM (0 d), the measured KD value was increased by 39%.However, the HIM fraction of the later stages lowered thepyrene sorption, but the magnitudes of the KD decreases weremuch smaller compared with the other three DOM fractions.

The differential roles of different DOM fractions in theobserved effect of bulk DOM on the pyrene KD value couldbe evaluated quantitatively according to the content of eachDOM fraction in the bulk DOM. Apparently, the overallreduction of the pyrene KD value in the presence of a bulk

DOM is the sum of the contributions from all four significantDOM fractions isolated from the bulk DOM. The contributionfrom each DOM fraction can be calculated from its reduction(or enhancement) of the KD value multiplying by the relativecontent of the DOM fraction that are presented in Figure 2.According to this calculation, AIM was the most importantfraction and contributed 16, 46, 50, and 74% at 0, 21, 63, and180 d of decay time, respectively, to the overall reduction of thepyrene KD value by the bulk DOM. Similarly, the calculatedcontributions of HIM to the overall reduction in KD values bythe bulk DOM were 72, 23, 6, and 1% at 0, 21, 63, and 180 d,respectively. Because of their low relative contents, the HOAand HON fractions contributed 3 to 32% and 6 to 12%,respectively, to the overall reduction in the KD by the bulkDOMs. Conversely, the presence of ther HIM fraction for 0 and180 d increased the KD values by 39 and 10%, respectively,indicating that HIM facilitates sorption of pyrene on the soil.

Correlation: DOM properties and pyrene sorption coefficient

The varied roles of different DOM fractions in the sorptionof pyrene on the soil are apparently related to the overallstructural and elemental properties of the DOMs. Intuitivelyit seems that the magnitude of equilibrium sorption of pyrene bysoil depends on the interactions between the DOM fractions andpyrene in the aqueous phase. The more strongly a DOM fractioninteracts with pyrene, the less pyrene is sorbed on the soil. It isknown that the van der Waals forces dominate the binding ofHOCs such as pyrene on colloids and surfaces in aquaticsystems. The DOM fractions such as AIM, having less polarityand lower O/C or (NþO)/C atomic ratios, are expected toexhibit greater interactions with HOCs, hence lowering theapparent sorption capacity for the soil. Conversely, the HIMhas high O/C atomic ratios, suggesting its greater polarity andits comparably weaker interactions with pyrene. The presenceof HIM in the background solution has little effect on themeasured KOC values.

Figure 5a to c presents correlations between the measuredlog KOC values and the elemental compositions of the bulkDOMs and their fractions. The positive correlations betweenlog KOC value and the O/C and (OþN)/C atomic ratios of theDOM fractions (Fig. 5a and b) indicate that polar DOMs withgreater O/C and (OþN)/C atomic ratios have less effect on themeasured log KOC values for the soil because of weakerinteractions between DOM and pyrene. Similarly, the weakpositive correlations between log KOC value and the H/C atomicratios of the DOM fractions (Fig. 5c) indicate that the DOMfractions with smaller H/C atomic ratios have stronger effectson the log KOC values measured for the soil, suggesting that theDOM fractions with higher contents of aromatic carbon mayinteract with pyrene more strongly. It should be pointed out thatthe KOC values presented here were calculated based on theassumptions that the added background DOM would not adsorbon soil particles and that the organic matter bound to the originalsoil would not desorb to the solution phase. Similarly, the H/Catomic ratios for the added DOM fractions could be used only ifthere was no exchange of organic carbon between the soil andthe water phases.

The correlations between log KOC values and the O/C and H/C atomic ratios of soil/sediment organic matter (SOM) havebeen reported in several prior studies [30,31]. The negativerelationship between the log KOC values and the O/C atomicratios of SOM reported in these prior studies is consistent withthe fact that the major driving force for sorption is hydrophobicinteractions. The lower the O/C atomic ratio of the SOM, the

Fig. 5. Correlations between the logKOC for pyrene sorption on soil and the atomic ratios of O/C (a), (NþO)/C (b), and H/C (c) of bulk dissolved organic matter(DOM) and the DOM fractions.


more hydrophobic the SOM, and the greater the driving forcefor sorption [32]. This is also consistent with our observations ofa positive correlation between the log KOC values and the O/Catomic ratios of DOM, because DOM acts to enhance solubilityof HOCs in aqueous phase and lower binding affinity of HOCson soils.

Significance

The present study shows that the DOM derived from ricestraw is chemically and structurally heterogeneous and that thechemical and structural properties of DOM evolve as a functionof straw decay time. The rice straw–derived DOM can signifi-cantly enhance desorption and leaching of organic pollutantsfrom soil. Because of the evolving chemical heterogeneity ofDOM, the facilitated organic pollutant desorption is likely at themaximum in the beginning of the straw decay. It appears that insitu disposal of rice straw could cause enhanced transport oftoxic organic chemicals such as pesticides and herbicides fromsoil to surface water systems, posing potential risk to livingaquatic organisms and human beings.

Acknowledgement—We thank two anonymous reviewers for their valuablecomments and suggestions on the original manuscript. The present study wassupported by the National Natural Science Foundation of China (NSFC;project 40501065), the Zhejiang Natural Science Foundation (ZJNSF;project Y307068), and Multistate Project W-1082 from the U.S. Departmentof Agriculture Cooperative State Research, Education, and ExtensionService.

REFERENCES

1. Guo L, Bicki TJ, Felsot AS, Hinesly TD. 1993. Sorption and movementof alachlor in soil modified by carbon rich wastes. J Environ Qual22:186–194.

2. Johnson A, Worral F, White C, Walker A, Besieu TJ, Williams RJ. 1997.The potential of incorporated organic matter to reduce pesticideleaching. Toxicol Environ Chem 58:47–61.

3. Flores-Cespedes F, Fernandez-Perez M, Villafranca-Sanchez M,Gonzalez-Pradas E. 2006. Cosorption study of organic pollutants anddissolved organic matter in a soil. Environ Pollut 142:449–456.

4. Rochette P, Gregorich EG. 1998. Dynamic of soil microbial biomass C,soluble organic C and CO2 evolution after three years of manureapplication. Can J Soil Sci 78:283–290.

5. Gregorich EG, Rochette P, McGuire S, Liang BC, Lessard R. 1998.Soluble organic carbon and carbon dioxide fluxes in maize fieldsreceiving spring-applied manure. J Environ Qual 27:209–214.

6. Chiou CT, Malcolm RL, Brinton TI, Kile DE. 1986. Water solubilityenhancement of some organic pollutants and pesticides by dissolvedhumic and fulvic acids. Environ Sci Technol 20:502–508.

7. Uhle ME, Chin YP, Aiken GR, McKnight DM. 1999. Binding ofpolychlorinated biphenyls to aquatic humic substances: The role ofsubstrate and sorbate properties on partitioning. Environ Sci Technol33:2715–2718.

8. Blasioli S, Braschi I, Pinna MV, Pusino A, Gessa CE. 2008. Effect ofundesalted dissolved organic matter from composts on persistence,adsorption, and mobility of cyhalofop herbicide in soils. J Agric FoodChem 56:4102–4111.

9. Johnson WP, John WW. 1999. PCE solubilization and mobilization bycommercial humic acid. J Contam Hydrol 35:343–362.

10. Cho HH, Park JW, Liu CC. 2002. Effect of molecular structures on thesolubility enhancement of hydrophobic organic compounds by environ-mental amphiphiles. Environ Toxicol Chem 21:999–1003.

11. Liu H, Gary AG. 1993. Modeling partitioning and transport interactionsbetween natural organic matter and polynuclear aromatic hydrocarbonsin groundwater. Environ Sci Technol 27:1553–1562.

12. Magee BR, Lion LW, Lemley AT. 1991. Transport of dissolved organicmacromolecules and their effect on the transport of phenanthrene inporous media. Environ Sci Technol 25:323–331.

13. Li K, Xing BS, Torello WA. 2005. Effect of organic fertilizers deriveddissolved organic matter on pesticide sorption and leaching. EnvironPollut 134:187–194.

14. Johnson WP, Amy GL. 1995. Facilitated transport and enhanceddesorption of polycyclic aromatic hydrocarbons by natural organicmatter in aquifer sediments. Environ Sci Technol 29:807–817.


15. Akkanen J, Tuikka A, Kukkonen JK. 2005. Comparative sorption anddesorption of benzo[a]pyrene and 3,4,30,40-tetrachlorobiphenyl innatural lake water containing dissolved organic matter. Environ SciTechnol 39:7529–7534.

16. Chen HL, Zhou JM, Huang W, Yu WP, Wan ZX. 2009. Biodegradabilityof dissolved organic matter derived from rice straw. Soil Sci 174:143–150.

17. Leenheer JA. 1981. Comprehensive approach to preparative isolationand fractionation of dissolved organic carbon from natural waters andwastewaters. Environ Sci Technol 15:578–587.

18. Leenheer JA, Noyes TI, Rostad CE, Davisson ML. 2004. Character-ization and origin of polar dissolved organic matter from the Great SaltLake. Biogeochemistry 69:125–141.

19. Polubesova T, Sherman-Nakache M, Chefetz B. 2007. Binding of pyreneto hydrophobic fractions of dissolved organic matter: Effect ofpolyvalent metal complexation. Environ Sci Technol 41:5389–5394.

20. Ilani t, Schula E, Chefetz B. 2005. Interactions of organic compoundswith wastewater dissolved organic matter: Role of hydrophobicfractions. J Environ Qual 34:552–562.

21. Kalbitz K, Schmerwitz J, Schwesig D, Matzner E. 2003. Biodegradationof soil-derived dissolved organic matter as related to its properties.Geoderma 113:273–291.

22. Don A, Kalbitz K. 2005. Amounts and degradability of dissolved organiccarbon from foliar litter at different decomposition stages. Soil BiolBiochem 37:2171–2179.

23. Chefetz B, Hadar Y, Chen Y. 1998. Dissolved organic carbon fractionsformed during composting of municipal solid waste: Properties andsignificance. Acta Hydrochim Hydrobiol 26:172–179.

24. Dilling J, Kaiser K. 2002. Estimation of the hydrophobic fraction ofdissolved organic matter in water samples using UV photometry. WaterRes 36:5037–5044.

25. Ma HZ, Allen HE, Yin YJ. 2001. Characterization of isolated fractions ofdissolved organic matter from natural waters and a wastewater effluent.Water Res 35:985–996.

26. Grossl PR, Inskeep WP. 1992. Kinetics of octacalcium phosphate crystalgrowth in the presence of organic acids. Geochim Cosmochim Acta56:1955–1961.

27. Poirier N, Sohi SP, Gaunt JL, Mahieu N, Randall EW, Powlson DS,Evershed RP. 2005. The chemical composition of measurable soilorganic matter pools. Organ Geochem 36:1174–1189.

28. Christensen JB, Jensen DL, Grøn C, Filip Z, Christensen TH. 1998.Characterization of the dissolved organic carbon in landfill leachate-polluted groundwater. Water Res 32:125–135.

29. Schlautman MA, Morgan JJ. 1993. Effects of aqueous chemistry on thebinding of polycyclic aromatic hydrocarbons by dissolved humicmaterials. Environ Sci Technol 27:961–969.

30. Kang S, Xing B. 2005. Phenanthrene sorption to sequentially extractedsoil humic acids and humans. Environ Sci Technol 39:134–140.

31. Grathwohl P. 1990. Influence of organic matter from soils and sedimentsfrom various origins on the sorption of some chlorinated aliphatichydrocarbons: Impolications on KOC correlations. Environ Sci Technol24:1687–1693.

32. Huang W, Weber WJ Jr. 1997. A distributed reactivity model for sorptionby soils and sediments. 10. Relationships between desorption, hysteresis,and the chemical characteristics of organic domains. Environ SciTechnol 31:2562–2569.

effects of rice straw–derived dissolved organic matter on pyrene sorption by soil

Documents