Study of the interactions of dissolved organic matter with zinc ion and the impact of competitive metal ions (Ca²⁺ and Mg²⁺) by in situ absorbance

Abstract

The bioavailability and toxicity of zinc to aquatic life depend on dissolved organic matter (DOM), such as Suwannee River Fulvic Acid (SRFA), which plays an important role in the speciation of zinc. This study examined reactions of SRFA with zinc at different concentrations from pH 3.0 to 9.0, and competitive binding of calcium/magnesium and zinc to SRFA at pH 6.0, using in situ absorbance. Interactions of Zn²⁺ with SRFA chromophores were evidenced by the emergence of features in Zn-differential spectra. Among all Zn²⁺–SRFA systems, dominant peaks, located at 235, 275 and 385 nm, and the highest intensity at 235 nm indicated the replacement of protons by the bound Zn²⁺. The Zn²⁺ binding with SRFA could be quantified by calculating the changes of the slopes of Zn-differential log-transformed absorbance in the wavelength range of 350–400 nm (denoted as DS_350–400) and by comparing the experimental data with predictions using the Non-Ideal Competitive Adsorption (NICA–Donnan) model. DS_350–400 was correlated well with the bound Zn²⁺ concentrations predicted by NICA–Donnan model with or without Ca²⁺ or Mg²⁺. Ca²⁺ and Mg²⁺ only affect intensity of the Zn-differential and Zn-differential log-transformed absorbance, not shape. In situ absorbance can be used to gain further information about Meⁿ⁺–DOM interactions in the presence of various metals.

Introduction

Dissolved organic matter (DOM) affects the speciation of zinc in natural waters (Mueller et al. 2012) and plays an important role in protecting aquatic organisms against toxicity of excessive zinc (Heijerick and Janssen 2000, Schamphelaere et al. 2005). The composition of DOM relies on sources of organic matter, ionic strength, pH and major cation in water (Leenheer and Croué 2003). Current techniques for quantifying DOM–metals interactions are very diverse, such as ion exchange technique (IET) (Fortin et al. 2010), ion-selective electrodes (ISEs) (Valilue et al. 2016), cathodic stripping voltammetry (Wasag and Grabarczyk 2016) and Donnan membrane technique (Jones et al. 2016). Although these methods have been used to determine free metal ion concentrations in natural waters, the performance of many assays remains inadequate for the complexity of analytical matrices and very low ambient metal concentrations.

Absorbance spectroscopy has been used widely to test the reactivity between DOM and free metal ions. It was proved to be high sensitive to the variation of the composition of DOM and could track trace metal ion concentrations without any sample pre-concentration (Yan et al. 2013; Gao et al. 2015a, b). However, the absorbance spectroscopy of DOM–metals is featureless. An alternative in situ method, differential absorbance spectroscopy (DAS) (Dryer et al. 2008), was proved to be useful for revealing interpretable features associated with deprotonation of DOM molecules and the reaction with metal ions. In addition, Windermere Humic Aqueous Model (WHAM) and the Non-Ideal Competitive Adsorption (NICA)–Donnan model can predict concentrations of free metal ions to DOM binding in natural waters (Dudal and Gérard 2004).

In this study, we used in situ absorbance spectroscopy to track interactions between DOM and Zn²⁺ per se at a wide range of pHs, as well as Zn²⁺ in the presence of different concentrations of Ca²⁺/Mg²⁺ at pH 6.0. These background ions affect various heavy metals binding on SRFA. The spectroscopic data were processed and compared with the predicted data of NICA–Donnan model, which allowed the quantitation of metal binding on SRFA.

Materials and methods

Reagents and chemicals

All chemicals used are reagent grade, unless mentioned specifically. All solutions were prepared using Milli-Q water (18.2 MΩ cm⁻¹, Millipore Corp., MA, USA). Suwannee River Fulvic Acid (SRFA) (standard number 1S101F) was acquired from the International Humic Substances Society (IHSS). SRFA is deemed to be representative of a wide range of DOM types in the environment (Maurice 2015). The dissolved organic carbon (DOC) concentration of SRFA was 5.0 mg L⁻¹ as DOC. The concentrations of residual metal cations in the SRFA sample were negligible (Kuhn et al. 2015). Ionic strength of SRFA solutions (0.01 mol L⁻¹) was regulated by adding NaClO₄ background electrolyte. Stock solutions (Zn²⁺, Ca²⁺ and Mg²⁺) were prepared using ZnCl₂, CaCl₂·2H₂O and MgCl₂·6H₂O salts from Aladdin reagent Shanghai Co. Ltd. (Aldrich Chemical Company, Shanghai, China).

Titrations

We conducted Zn²⁺ titrations according to the procedures for different metal ions previously described (Yan and Korshin 2004; Yan et al. 2013, 2015; Gao et al. 2015a, b). A certain amount of Zn²⁺ stock solution was added into a series of 40 mL-jars SRFA at pH 5.0–9.0, and a series of 40 mL-jars SRFA with or without 0.00025 and 0.0025 mol L⁻¹ Ca²⁺ or 0.0004 and 0.004 mol L⁻¹ Mg²⁺ at ionic strength 0.01 mol L⁻¹, pH 6.0.

Total Zn²⁺ concentrations were varied from 0 to 10⁻⁶ mol L⁻¹, which was below its precipitation level determined using Visual MINTEQ. We analyzed SRFA–metal complexation using the NICA–Donnan model (Benedetti et al. 1996; Kinniburgh et al. 1999; Milne et al. 2003). Complexation constants from Visual MINTEQ database used in the calculations are marked in Table S1. We adjusted the pH of the solutions by adding small amounts of HClO₄ or NaOH. Also, we added 0.01 mol L⁻¹ MES buffer to the solutions with low buffering capacity (e.g., at pH 6.0). This buffer does not affect the ionic strength of SRFA solutions or absorbing light above ca. 230 nm (And and Rorabacher 1999). After adding metal stock solution, followed by mixing and standing for 30 min, absorbance spectra of aliquots from solutions with various metal concentrations were conducted on a Perkin-Elmer Lambda 2550 UV–Vis spectrophotometer with a 5-cm cell at wavelengths from 200 to 800 nm.

Absorbance data processing

Data processing of SRFA absorbance spectra was done as described in previous studies (Yan and Korshin 2004; Yan et al. 2013, 2016; Gao et al. 2015a, b). The differential linear spectra and differential log-transformed spectra were calculated using Eqs. (1), (2), respectively (Dryer et al. 2008; Gao et al. 2015a, b):

$${\text{DA}}_{\lambda } = \frac{1}{{{\text{DOC}} \cdot L}}\left[ {A_{\lambda ,i} - A_{{\lambda ,{\text{ref}}}} } \right]$$

(1)

$${\text{DLnA}}_{\lambda } = {\text{LnA}}_{\lambda ,i} - {\text{LnA}}_{{\lambda ,{\text{ref}}}}$$

(2)

In these formulas, A _λ,i is DOM absorbance measured at the wavelength λ for any selected condition (i), and A _λ,ref is DOM absorbance measured at the wavelength λ for an applicable reference (ref, e.g., zero total metal or zinc concentration, DOC is the concentration of organic carbon (mg/L), and length is the cell length (in cm). The slopes and differential slopes of log-transformed absorbance spectra of DOM were calculated as follows:

$$S_{350 - 400} = \left| {\frac{d\ln A(\lambda )}{d\lambda }} \right|_{350 - 400}$$

(3)

$${\text{DS}}_{350 - 400} = \, S_{350 - 400,i} - \, S_{{350 - 400,{\text{ref}}}} .$$

(4)

In the above equations, S _350–400 is the slope of the linear correlation that fit the log-transformed DOM absorbance spectra in the range between 350 and 400 nm. S _350–400,i and S _{350–400,ref} are the spectral slopes determined for any selected experimental condition and applicable reference (ref, e.g., zero total metal or zinc concentration), respectively. The prefix D denotes the differential between any selected experimental condition and the applicable reference against which that differential is calculated.

Results and discussion

Effect of Zn²⁺ binding on SRFA chromophores

As illustrated in Fig. 1a for SRFA at pH 9.0 and ionic strength 0.01 mol L⁻¹, the zero-order absorbance spectra of SRFA have no conspicuous features and the intensity of them decreases near exponentially with the observation wavelength. By increasing zinc concentration, subtle but consistent changes in zero-order absorbance spectra were detected, which was consistent with previous studies of metal–SRFA interactions (Yan and Korshin 2004; Yan et al. 2013; Gao et al. 2015a, b). To gain more information of the interactions of Zn²⁺–SRFA, we calculated DOC-normalized differential spectra (DAS) using Eq. 1, presented in Fig. 1b. Three distinct bands with peaks at ca. 235, 275 and 385 nm were identified, and the strongest was at ca. 235 nm. The shape change of absorbance spectra was stable when zinc concentration increased. As previously described (Dryer et al. 2008), the band at ca. 280 nm was likely associated with the deprotonation of carboxylic groups in SRFA, while the bands at ca. 244 nm and in the 300–390 nm region reflected the deprotonation of phenolic group in SRFA.

Fig. 1

Absorbance spectra of SRFA recorded at various concentrations of zinc at pH 9.0 and ionic strength 0.01 mol L⁻¹. a Zero-order spectra; b DOC-normalized differential spectra; c log-transformed spectra; and d differential log-transformed spectra at a constant DOC concentration 5.0 mg L⁻¹ with cell length 5 cm and ref. without zinc

Figure 1c demonstrates that the logarithms of absorbance of Zn²⁺–SRFA decreased linearly with the wavelength. However, the log-transformed spectra were made of several regions (e.g., <250, 250–400 and >400 nm) with slightly different slopes.

Figure 1d shows that the slopes of the log-transformed spectra have consistent changes at various zinc concentrations, and the slope of log-transformed absorbance of SRFA in wavelengths 350–400 nm would be the most sensitive range to the change of Zn²⁺ concentrations. Thence, the absolute slope values of log-transformed spectra in this wavelength range and their changes (denoted as S _350–400 and DS_350–400, respectively) were used to quantify the binding of Zn²⁺–SRFA.

The slope changes of log-transformed spectra in 350–400 nm (DS_350–400) calculated for Zn²⁺–SRFA at pH values ranging from 5.0 to 9.0 and an ionic strength at 0.01 mol L⁻¹ are shown in Fig. 2.

Fig. 2

Slope changes (DS_350–400) of the differential spectral as a function of Zn²⁺ concentrations at variations of pH at a constant ionic strength 0.01 mol L⁻¹

The datasets of DS_350–400 values in comparison with the concentrations of Zn²⁺–SRFA complexes estimated using the NICA–Donnan model are shown in Fig. 3. These calculations used the complexation constants from Visual MINTEQ database (Supporting Information Table S1). For all examined pHs, DS_350–400 values were linearly correlated with the modeled concentrations of SRFA-bound Zn²⁺.

Fig. 3

Correlation between DS_350–400 and the amount of SRFA-bound Zn²⁺ predicted by using the NICA-Donnan model for variations of pH at ionic strength 0.01 mol L⁻¹

The competitive binding of Zn²⁺ and Ca²⁺ by SRFA was examined at pH 6.0. Background Ca²⁺ concentrations were set at 0.0, 0.00025 and 0.0025 mol L⁻¹, respectively. Differential spectra for this system were calculated and compared with the reference conditions with the same pH, ionic strength and the selected background Ca²⁺ concentration in the absence of Zn²⁺. The differential spectra of the Ca²⁺–Zn²⁺–SRFA system for the three Ca²⁺ background levels and various total Zn²⁺ concentrations are shown in Fig. 4.

Fig. 4

Differential spectra of SRFA generated at various total concentrations of Zn²⁺, with background Ca²⁺ concentration at: a 0 mol L⁻¹; b 0.00025 mol L⁻¹; and c 0.0025 mol L⁻¹, respectively, at a constant pH 6.0, ionic strength 0.01 mol L⁻¹ and DOC 5.0 mg L⁻¹

No obvious change in the shape of the Zn²⁺-differential spectra was observed at various Ca²⁺ levels and the intensity of Zn²⁺-differential spectra decreased in the presence of Ca²⁺, which were similar to previous study (Gao et al. 2015a, b). This indicates that competing Ca²⁺ reduced the amount of SRFA sites available for Zn²⁺ binding. When comparing the intensity data of Zn²⁺-differential spectra for different background concentrations of Ca²⁺, the conspicuous site of the Zn²⁺–SRFA absorbance with 0.00025 mol L⁻¹ Ca²⁺ was different from others. A hypothesis was made that low Ca²⁺ concentration would be more influential. Further researches were needed here.

Similar results of SRFA spectra in the increasing concentration of Zn²⁺ and Ca²⁺ are shown in Fig. S2, which proved that the slope of log-transformed absorbance of Zn²⁺–Ca²⁺–SRFA in the wavelengths range of 350–400 nm is most sensitive. The DS_350–400 was calculated at varying total zinc concentrations in three background levels of Ca²⁺ to illuminate the occurrence and contributions of these alternative mechanisms, as shown in Fig. 5.

Fig. 5

Correlation between changes of the spectral slope in the range of 350–400 nm and total Zn²⁺ concentrations, at various background Ca²⁺ concentrations at a constant pH 6.0 and ionic strength 0.01 mol L⁻¹

This figure shows that DS_350–400 increases gradually with total Zn²⁺ concentrations, but this effect is suppressed by background Ca²⁺.

The relationships between the changes of the spectral slope in the range of 350–400 nm and the Zn²⁺ concentrations bound to SRFA, simulated by the NICA–Donnan model, at three different Ca²⁺ concentrations, are shown in Fig. 6. These calculations used the complexation constants from Visual MINTEQ database (Supporting Information Table S1).

Fig. 6

Correlation between DS_350–400 and the amount of SRFA-bound Zn²⁺ predicted by using the NICA-Donnan model for various background Ca²⁺ concentrations at a constant pH 6.0 and ionic strength 0.01 mol L⁻¹

Correlation between DS_350–400 and the amount of SRFA-bound Zn²⁺ is predicted by the NICA–Donnan model for three different background Ca²⁺ levels at pH 6.0 and ionic strength 0.01 mol L⁻¹.

It demonstrates that for all examined background Ca²⁺ levels, DS_350–400 values measured in a wide range of Zn²⁺ concentrations are linearly correlated with the modeled concentrations of SRFA-bound Zn²⁺ irrespective of the background Ca²⁺ concentrations. Additionally, this figure also demonstrates that the DS_350–400 values did not form a simple dataset when compared with the sum of molar concentrations of Zn²⁺–SRFA and Ca²⁺–SRFA complexes ([Zn²⁺–SRFA] + [Ca²–SRFA]) in solutions, rather than α[Zn²⁺–SRFA] +β [Ca²⁺–SRFA] (α, β represent the account of the respective differential spectra).

Effect of Mg²⁺ on Zn²–SRFA interaction

The competitive binding of Zn²⁺ and Mg²⁺ by SRFA was examined at the same condition as in Ca²⁺–Zn²⁺–SRFA systems except background Mg²⁺ levels at 0.0004 and 0.004 mol L⁻¹. Differential spectra for this system were calculated versus the reference conditions at the same pH, same ionic strength, no Zn²⁺, and either selected background Mg²⁺ concentration. The differential spectra of Mg²⁺–Zn²⁺–SRFA are shown in Fig. 7.

Fig. 7

Comparison of differential spectra of SRFA generated at different total concentrations of Mg²⁺: a 0.0004 mol L⁻¹; b 0.004 mol L⁻¹, respectively, at a constant pH 6.0, ionic strength 0.01 mol L⁻¹ and DOC 5.0 mg L⁻¹

In comparison with the data in Fig. 4a, the intensity of Zn²⁺-differential spectra in the presence of Mg²⁺ was also suppressed; however, the effect of Mg²⁺ on the spectral intensity of Zn²⁺–SRFA was less than Ca²⁺. This indicates that Mg²⁺ would also compete with binding sites available for Zn²⁺, the nature of these sites remained the same irrespective of the background Mg²⁺ concentrations, but differential spectra of Zn²⁺–SRFA have few obvious features in the presence of Mg²⁺. The DS_350–400 of Zn²⁺–Mg²⁺–SRFA at various Zn²⁺ concentrations and different Mg²⁺ concentration is displayed in Fig. 8.

Fig. 8

Correlation between changes of the spectral slope in the range of 350–400 nm and total Zn²⁺ concentrations at different background magnesium levels at a constant pH 6.0 and ionic strength 0.01 mol L⁻¹

This figure evidences that DS_350–400 increases gradually with total Zn²⁺ concentrations in the presence of various Mg²⁺, and the effect was suppressed similarly by 0.0004 mol L⁻¹ Mg²⁺ or 0.004 mol L⁻¹ Mg².

The relationships between DS_350–400 and concentrations of Zn²⁺ bound onto SRFA from the NICA–Donnan model in the presence of Mg²⁺ are shown in Fig. 9.

Fig. 9

Correlation between DS_350–400 and the amount of SRFA-bound Zn²⁺ predicted by the NICA–Donnan model for different background magnesium levels at a constant pH 6.0 and ionic strength 0.01 mol L⁻¹

It manifests that DS_350–400 values in a wide range of Zn²⁺ concentrations are linearly correlated with the predicted Zn²⁺ concentrations, bound onto SRFA, irrespective of the background Mg²⁺ concentrations. Additionally, the DS_350–400 values are formed by datasets compared with the sum of molar concentrations of α[Zn²⁺–SRFA] + β [Mg²⁺–SRFA] (α, β represent the account of the respective differential spectra).

Conclusions

The differential spectra of Zn²⁺–SRFA had consistent changes and performed metal-specific features, which reflected the nature and amount of interactions between Zn²⁺ and SRFA. The features in differential spectra were ascribed to carboxylic and phenolic functional groups in SRFA and caused by the replacement of the SRFA-bound protons by Zn²⁺.

The differences in the effects of Zn²⁺ and Ca²⁺/Mg²⁺ on SRFA chromophores allowed quantitating the competition between Zn²⁺ and Ca²⁺/Mg²⁺ binding on SRFA sites in Zn²⁺–Ca²⁺/Mg²⁺–SRFA system, which could be explained well by examining the intensity and shapes of differential spectra of SRFA. The features of differential spectra of Zn²⁺–Mg²⁺–SRFA system are not obvious. However, significant differences of the response were observed in Zn²⁺ differential spectra in the presence of 0.00025 mol L⁻¹ Ca²⁺, which require further investigation.

The binding of Zn²⁺ onto SRFA is not significantly affected by Mg²⁺ (e.g., 0.0005 or 0.005 mol L⁻¹) at pH 6.0. The competition of Zn²⁺/Ca²⁺/Mg²⁺ binding by SRFA can be quantified by calculating the DS_350–400 and predicted by the NICA–Donnan model. This study demonstrated the applicability of the NICA–Donnan model in elucidating interactions of the SRFA with Zn²⁺ at a wide range of pHs, and different concentrations of Ca²⁺ and Mg²⁺ by in situ absorbance. NICA–Donnan model can be used to gain further information about Meⁿ⁺–DOM interactions in the presence of various metals at different pHs and ionic strengths