Nickel Binding A ﬃ nity with Size-Fractioned Sediment Dissolved and Particulate Organic Matter and Correlation with Optical Indicators

: In rivers, the distribution and reactivity of heavy metals (HMs) are affected by their binding affinitywithsedimentdissolvedorganicmatter(DOM)andparticulateorganicmatter(POM).The HM-OM binding a ﬃ nity a ﬀ ected by the interaction between DOM and POM is not well studied. This study investigated the Ni binding a ﬃ nity to size-fractioned overlaying water DOM and alkaline extracted sediment POM solution (AEOM). The DOM / AEOM ﬁltrates ( < 0.45 µ m) were sequentially separated into ﬁve nominal molecular weight (MW) solutions. The AEOM optical indicators had lower autochthonous, higher terrestrial sources, and lower aromaticity than the DOM. The Ni mass (72.3 ± 6.4%) was primarily distributed in the low molecular weight DOM ( < 1 kDa), whereas the Ni (93.5 ± 0.4%) and organic carbon (OC) mass (85.3 ± 1.0%) were predominantly distributed in the high molecular weight AEOM. The Ni and DOM binding a ﬃ nity, ([Ni] / [DOC]) DOM ratio ranging from 0.76 to 27.32 µ mol / g-C, was signiﬁcantly higher than the ([Ni] / [DOC]) AEOM ratios, which ranged from 0.64 to 2.64 µ mol / g-C. The ([Ni] / [DOC]) AEOM ratio correlated signiﬁcantly with the selected optical indicators (r = 0.87–0.92, p < 0.001), but the ([Ni] / [DOC]) DOM ratio correlated weakly with the optical indicators (r = 0.13–0.40, p > 0.05). In the present study, the Ni binding a ﬃ nity with size-fractioned DOM / AEOM agrees with the hypothesis of the DOM and POM exchange conceptual model in sediment. The POM underwent a hydrolysis / oxidation process; hence, AEOM had a high molecular weight and stable chemical composition and structure. The Ni mainly attached to the high molecular weight AEOM and the ([Ni] / [DOC]) AEOM ratios had a strong correlation with the AEOM optical indicators. In contrast, DOM had a high ([Ni] / [DOC]) DOM ratio in low molecular weight DOM. DOM had higher ratios than the HMW DOM. The ([Ni] / [DOC]) AEOM ratios had a signiﬁcant correlation with the optical indicators, but the ([Ni] / [DOC]) DOM ratios correlated weakly with the optical indicators. These results are in agreement with the exchange process between the DOM and POM in sediment. The present study suggests the exchange between the DOM and POM might play an essential role in understanding Ni and organic matter binding a ﬃ nity in sediment.


Introduction
The sediment organic matter (SOM) includes particulate (POM) and dissolved (DOM) organic matter [1][2][3][4][5]. The distribution of heavy metals in DOM/POM affects heavy metals toxicity to microorganisms and the safety of ecology that has received extensive attention worldwide [6][7][8][9]. The binding affinity between heavy metals and organic matter is affected by the quality and quantity of organic matter, and the type and total concentration of metals [2,6].

Sampling Site and Samples Collection
The sampling site (22 • 37 50.2 N 120 • 32 12.1 E) was downstream from a food industrial park in Pingtung County, Taiwan. The wastewaters from the individual manufacture were collected in a central wastewater treatment plant. The water quality in the effluent of the treatment plant wastewater agreed with the wastewater standard of the Taiwan Environmental Protection Administration (BOD < 30 mg/L, COD < 100 mg/L, and SS < 30 mg/L). The sampling location was about 1000 m downstream of the effluent outlet of the wastewater treatment plant. The samples were collected in triplicate in October 2016. The DOM and POM samples were collected from the surface sediment with a grab method. Five liter samples (including liquid and solid phases) were collected in each capture group. The samples were taken back to laboratory within 4 h. In the laboratory, sediment samples were centrifuged at 4500 rpm for 30 min. The liquid sample was filtered (<0.45µm) to collect the DOM sample and was stored at 4 • C for further separation. The solid phase was air dried for two weeks and then passed through a 2 mm sieve to collect the POM. The POM was stored in a 4 • C for further extraction.

AEOM Extraction and Separation
We followed Hur et al. [32] where bulk AEOM was extracted with an alkaline solution. Briefly, 5 g air dry sediment was added to 0.1 N NaOH 100 mL solution. The sediment suspension was shaken for 24 h and centrifuged at 4500 rpm for 30 min. Four-liter bulk AEOM and DOM solutions (<0.45 µm) were separated into the five molecular weight AEOM and DOM fractions. A cross-flow ultrafiltration system sequence equipped with 10, 3, 1, and 0.3 kDa nominal molecular weight cutoff ceramic membrane cartridges (Filtanium, France) was used to separate the five molecular weight AEOM/DOM solutions: MW-A (10 kDa-0.45 µm), MW-B (3-10 kDa), MW-C (1-3 kDa), MW-D (0.3-1.0 kDa), and MW-E (<0.3 kDa). The feed flow rate was 1.7-2.0 L/min. In the concentration process, the volume concentration factor (C f ) was kept at 10, the ratio of the sum volume of the retentate and permeate to retentate volume. In each separation process, the retentate flow was sent back to the feed flow container and the permeate flow was collected in another container. The mass balances (R%) of the DOC and Ni were calculated by Equation (1). The mass percentages (M i ) of organic carbon and Ni in each fractioned DOM/POM were calculated following Equation (2).
where R(%) is percentage of mass balances for DOC, and Ni. M i (%) is mass percentages for individual size-fractioned DOC and Ni. C i and V i were the concentrations and volumes for each size-fraction in the separation processes for DOC and Ni. The volume ratios for the study were 1.0, 0.1, 0.09, 0.081, 0.0729, and 0.6561 for bulk and the five size-fractioned solutions: MW-A, MW-B, MW-C, MW-D, and MW-E, respectively. In the separated process the membranes were carefully cleaned following the suggestion by the manufacture.

UV/Vis and Fluorescent Measurement
When the UV-Vis and fluorescent spectra were measured, the bulk and fractioned AEOM and DOM solutions were diluted to 5.0 and 1.0 mg-C/L, respectively, with ultrapure water. The absorbance was measured with an ultraviolet/visible spectrophotometer (Hitachi, U-2900) and fluorescence spectra were recorded on a fluorescence spectrometer (Hitachi F-7000). The absorbance at 700-800 nm was set as the background value. The absorbance of the sample was subtracted from the average of the absorbance at 700-800 nm per Helms et al. [33]. The UV-Vis spectrophotometric scanning wavelength was 800-200 nm. The excitation/emission matrixes (EEMs) were generated by recording emission spectra from 250 to 550 nm at 2.0 nm steps for an excitation wavelength between 200 and 450 nm at 5 nm increments. The scanning rate was 2400 nm/min. The value of the blank sample was subtracted from the sample fluorescent data.

Optical Indicator
Three widely used optical indicators were selected to qualify and quantify the chemical properties of bulk and fractioned DOM/AEOM. SUVA 254 (L/mg-C/m) was the absorption coefficient (cm −1 ) at 254 nm divided by DOC concentration (mg-C/L) times 100 following Weishaar et al. [34]; Matilainen et al. [27]. The fluorescence index (FI) was calculated as the intensity ratio of emission at 450 nm over 500 nm with an excitation at 370 nm per Hansen et al. [24]; Birdwell and Engel [28]; Derrien et al. [25]. The Biological index (BIX) was calculated as the intensity ratio at emission wavelength 380 nm over 430 nm with an excitation of 310 nm as was done by Huguet et al. [29]; Birdwell and Engel [28]. Nickle binding affinity to DOM/AEOM was calculated as the Ni concentration divided by DOC concentration ([Ni]/[DOC], µmol/g-C) per Baken et al. [12]; Kikuchi et al. [17]; Hsieh et al. [22].

Statistical Analysis
In this study, linear correlation, and the difference tests used the S-Plus software (V 6.2) at significance levels at p < 0.05. Two group difference tests between AEOM and DOM (such as concentration, mass fraction, [Ni]/[DOC] ratio, and optical indicators) were used in the t-test method. The three group difference tests used the ANOVA test method such as fraction difference of mass fractions, [Ni]/[DOC] ratios, and optical indicators for DOM/AEOM. Fluorescence indicators were calculated at R script as developed by Lapworth and Kinniburgh [35].

DOC and Ni Concentrations and Mass Fractions in DOM and AEOM
DOC concentration represents the abundances of dissolved organic matter [3,36]. The mass fractions of Ni and DOC in size-fractioned DOM and AEOM were important to understand the Ni and DOC distribution between DOM and POM. Table 1 lists DOC and Ni concentrations of bulk and fractioned DOM and AEOM. In two other studies, concentrations of DOC in bulk DOM were similar to river water, which ranged from 1.3 to 20 mg/L with an average of 6 mg/L for DOC [12] but lower than sediment pore water, which ranged from 15-28 mg/L [37].  In our research, bulk alkaline extracted organic carbon (AEOC) averaged 14.78 g-C/kg and total organic carbon was 73.1 g-C/kg based on sediment mass. The AEOC/TOC ratio was 20.2%. The AEOC content in present study was higher than water extracted sediment organic carbon, which averaged 226 and 828 mg-C/kg [38] and alkaline extracted organic carbon from sediment, which averaged 5.94 ± 0.56-6.40 ± 0.96 g-C/kg [39]. The AEOC/TOC ratios in this study were higher than the WEOC/TOC ratios extracted from soil reported by Hsieh et al. [22], and Xu et al. [38] and comparable to the AEOC/TOC ratio, which averaged 23.7% from sediment [39]. The high AEOC concentration in this study suggested the sediment contained highly extractable organic matter, which could be discharged from food wastewater.
The concentration of Ni in DOM was higher than river water and sediment pore water in other studies, which ranged from 1.05 to 7.52 µg/L [14,23,37,40]. In present study, the average Ni concentration of bulk AEOM was 1.70 mg/kg and the total Ni concentration averaged 33.0 mg/kg, based on the sediment mass. The AEOM-Ni accounted for 5.2% of the total Ni. The AEOM-Ni concentration in the present study was higher than the water and NaOH-extracted Ni concentrations 0.21-1.21 mg/kg from soil [19,22]. In addition, the ratio of AEOM-Ni/total Ni was higher than the ratio 1.07%, as reported by Hsieh et al. [22]. The ratio of extracted organic carbon was higher than the Ni ratio, which suggested the organic matter was more readily extracted by alkaline solution than the Ni.
In Table 1, the DOC and Ni concentrations of size-fractioned DOM/AEOM were the measured concentrations. The mass fractions, in each fractioned DOM/AEOM, were calculated by Equation (2). Figure 1a,b shows DOC and Ni mass percentages of size-fractioned DOM/AEOM. In most of the separation studies, molecular weight 1 kDa was a common size to distinguish high and low molecular weight DOM/AEOM [21,22,38,39,41]. In the DOM solution, the Ni mass percentage (27.7 ± 6.4%) of high molecular weight DOM (HMW, >1 kDa) was significantly lower than the percentage (72.3 ± 6.4%) of low molecular weight DOM (LMW, < 1 kDa, p = 0.001). The HMW DOM organic carbon (OC) mass percentage (45.5 ± 6.7%) was insignificantly different to the LMW DOM OC percentage (54.5 ± 6.7%) (p = 0.18). In the AEOM solutions, the mass distribution of OC and Ni had similar patterns. The HMW AEOM mass percentages (85.3 ± 1.0% for OC and 93.5 ± 0.4% for Ni) were significantly higher than the LMW AEOM (p < 0.001). The results suggested that Ni preferred binding with LMW DOM, but Ni preferred binding to HMW AEOM. Moreover, AEOM had a high fraction of organic carbon.
The mass fractions of fractioned OC and heavy metal in DOM were varied, which depended on the DOM sources and type of heavy metals. In addition, when metal and OC are extracted from soil and sediment, the fractions may depend on the matrix, type of metal, extraction solvent, extraction method and solid/liquid ratio, and the separation method and conditions [41,42]. For example, in one study, Ilina et al. [43] size-fractioned soil solution, lake water, and river water, where the HMW Ni percentages were 83%, 55%, and 43%, respectively. In other studies, the HMW OC percentage was 44.1% for soil AEOM solution [22] and 81% for sediment AEOM solution [39]. In a municipal wastewater treatment plant study, Hargreaves et al. [21] reported the HMW fractions were 67-75% for Ni and 58% for OC.
Hargreaves et al. [21]  ) AEOM (>1 kDa). The ratio difference followed the different molecular weights in the present study, which may be caused by the exchange of POM and DOM in sediment.  [24,25,27,34]. SUVA 254 values of AEOM was significantly lower than values of DOM (Figure 3a). SUVA 254 values of HMW were significantly higher than the LMW DOM/AEOM, respectively. However, all SUVA 254 values of DOM and AEOM were <3.0, which suggested that the chemical properties of DOM/AEOM attributed to the hydrophilic compounds and poor aromaticity [27].

DOM and AEOM Optical Indicators
FI is an indicator that describes the relative contribution of terrestrial and microbial sources to the DOM/AEOM pools [24,28]. Figure 3b shows the FI values of DOM/AEOM. The FI values of DOM/AEOM increased following the decreased molecular weight of fractioned DOM/AEOM. DOM FI values of MW-A (>10 kDa) were significantly lower than the LMW fractioned DOM (p = 0.02). The FI value of HMW AEOM (1.48 ± 0.06, >1 kDa) was significantly lower than the LMW AEOM (1.80 ± 0.12, p < 0.001). The low value of FI in the HMW AEOM may be generated by the lignin phenolic compounds when labile POM was hydrolyzed/oxidized [2,5].
BIX is an indicator used to assess the relative contribution of autochthonous DOM in water and soil samples [28,29]. Figure 3c shows BIX values with size-fractioned AEOM/DOM. The BIX values of total and bulk AEOM were significantly lower than the BIX values of DOM (p < 0.001). The AEOM autochthonous sources were lower than the DOM sources. BIX values of HMW AEOM/DOM were significantly lower than LMW AEOM/DOM, respectively (p = 0.001). Bulk AEOM contained median autochthonous sources but bulk DOM contained strong autochthonous sources. The high BIX value (>1.0) corresponded to recently produced DOM of autochthonous origin [29]. In our study, the BIX values of fractioned DOM ranged from 1.04 to 1.10, which indicated the DOM contained a strong autochthonous source [25,29].
In present study, the selected indicators suggested that AEOM had higher terrestrial and allochthonous sources than DOM. However, the aromaticity of DOM was higher than AEOM. Other work has shown that generally, the high terrestrial and low autochthonous sources have high aromaticity of DOM [10,44]. The low SUVA 254 values in this study attributed to hydrophilic substances and poor aromatics. The hydrophilic substances may not fully develop the aromaticity in DOM and AEOM. In addition, the DOM had a complex composition where the three selected indicators had an insignificantly correlation with each other. Nevertheless, the POM underwent a hydrolysis/bioxidation process where the three selected indicators had a significant correlation with each other in AEOM ( Table 2). The correlation suggested the extracted AEOM had more stable and uniform chemical properties.  The ratios of ([Ni]/[DOC]) DOM were insignificantly correlated with the indicators, which implied a complex composition of DOM, with aromatic content (SUVA 254 ); sources of DOM (FI and BIX) cannot predict the Ni and DOM binding affinity. The indicators had a strong correlation with ([Ni]/[DOC]) AEOM ratios, which suggested the AEOM chemical properties have the potential to represent the binding affinity of Ni. One investigation showed that the exchange conceptual model assumed that the POM in the sediment underwent hydrolysis and oxidation cleavage started from low molecular weight POM [2,5]

Conclusions
In this study, we studied OC and Ni distribution and Ni binding affinity on size-fractioned DOM/AEOM. The Ni and OC mass distribution in the DOM and AEOM was significantly different.