Thermodynamic stability of mercury(II) complexes formed
with environmentally relevant low-molecular-mass thiols
studied by competing ligand exchange and density
functional theory
Introduction
Inorganic divalent mercury (HgII) has a very high affinity for reduced sulfur functional groups. Reports from laboratory experiments suggest that HgII complexes with specific low-molecular-mass (LMM) thiol (RSH) ligands control rates of HgII transformation reactions. Because of methodological limitations for precise determination of the highly stable HgII complexes with LMM thiol ligands, constants reported in the literature remain inconsistent. This uncertainty impedes accurate modelling of the chemical speciation of HgII and the possibility to elucidate the role of HgII complexes with LMM thiols for Hg transformation reactions. Here, we report values of thermodynamic stability constants for 15 monodentate, two-coordinated HgII complexes, Hg(SR)2, formed with biogeochemically relevant LMM thiol ligands. The constants were determined by a two-step ligand-exchange procedure where the specific Hg(SR)2 complexes were quantified by liquid chromatography-inductively coupled plasma mass spectrometry. Thermodynamic stability constants (log β2) determined for the Hg(SR)2 complexes ranged from 34.6, N-cysteinylglycine, to 42.1, 3-mercaptopropionic acid, for the general reaction Hg2+ + 2RS- ⇌ Hg(SR)2. Density functional theory (DFT) calculations showed that electron-donating carboxyl and carbonyl groups have a stabilising effect on the HgII-LMM thiol complexes, whereas electron-withdrawing protonated primary amino groups have a destabilising effect. Experimental results and DFT calculations demonstrated that the presence of such functional groups in the vicinity of the RSH group caused significant differences in the stability of Hg(SR)2 complexes. These differences are expected to be important for the chemical speciation of HgII and its transformation reactions in environments where a multitude of LMM thiol compounds are present. Environmental context. The chemical speciation of mercury (Hg) largely controls its biogeochemical cycling and exposure to biota. Here, we investigate the thermodynamic stabilities of complexes formed between inorganic divalent Hg (HgII) and 15 biogeochemically relevant low-molecular-mass (LMM) thiol ligands. This information is critical for accurate modelling of the chemical speciation of HgII and to clarify the role of HgII-LMM thiol complexes in the cycling of Hg in the environment.
Mercury (Hg) pollution is of great concern globally, in particular with respect to the formation and bioaccumulation of the neurotoxic methylmercury (MeHg) molecule. Inorganic divalent mercury (HgII) has a high affinity for thiolate groups (RS-), which largely controls its chemical speciation and reactivity in environmental systems.[1] The formation of MeHg is mediated by phylogenetically diverse microorganisms carrying the hgcA and hgcB gene clusters.[2,3] Recent laboratory studies have demonstrated that the addition of specific low-molecular-mass (LMM) thiol (RSH) compounds to bacteria culture systems can greatly enhance cellular uptake and subsequent methylation of HgII.[4,5] Current explanation models for these observations are based on the formation of specific Hg(SR)2 complexes with a high bacterial uptake rate and decreased partitioning of HgII to outer-cell-membrane functional groups suppressing cellular uptake. It has also been suggested that the formation of HgII-LMM thiol complexes with other coordinations than 1 : 2 Hg : RSH, i.e. HgSR+ , Hg(SR)3 and Hg(SR)4, reduces the bacterial uptake rate of HgII. However, because of uncertainties in thermodynamic constants reported, the composition of HgII-LMM thiol complexes in typical HgII methylation assays remains uncertain. For the same reason, the role of HgII-LMM thiol complexes behind MeHg formation in natural waters, soils and sediment remains elusive. The concentration of such complexes in natural environments is far lower than current detection limits of analytical methods for their direct measurements. The concentrations of HgII-LMM thiol complexes therefore need to be established by chemical speciation modelling. Such modelling, however, requires accurately determined concentrations of LMM thiol compounds, and accurate thermodynamic stability constants for the corresponding HgII-LMM thiol complexes. Only quite recently, methods capable of detecting different types of LMM thiols in wetlands and marine ecosystems, in which MeHg formation is an issue, have been reported.[6-8] Further, as additional support for the constants reported, a theory needs to be established to explain how the thermodynamic stability of HgII-LMM thiol complexes varies with chemical structure of the thiol ligands. With such a theory, the stability of HgII complexes with as yet unidentified LMM thiols in natural environments could be predicted to further help explain the role of LMM thiols in chemical speciation and transformation processes of HgII.
Recent spectroscopic work, in particular extended X-ray absorption fine-structure (EXAFS) spectroscopy, has established structures for HgII-LMM thiol complexes. Combined with 1H and 13C nuclear magnetic resonance (NMR) spectroscopy, EXAFS results show that HgII forms monodentate bonds with thiol groups belonging to different molecules.[9-12] In a series of works, it has been established that Hg(SR)2, Hg(SR)3 and Hg(SR)4 complexes may all form depending on pH, the HgII/RSH molar ratio and the thiol concentration. For glutathione (GSH), which is a comparatively large molecule, Hg(GSH)2 is the predominant form at acidic and neutral pH values. At a GSH to HgII molar ratio of 22, the percentage composition of the complexes Hg(GSH)2, Hg(GSH)3 and Hg(GSH)4 is 95 : 2 : 3 at pH 7.4.[10] The thiol-containing ligands N-acetylcysteine (NACCys), penicillamine (Pen) and cysteine (Cys) have a slightly higher tendency to form Hg(SR)3 and Hg(SR)4 complexes, but Hg(SR)2 is still expected to be the predominant form in most soils and waters having a pH below 7.[11-13]
Because of the very strong bonding between HgII and RS- groups, traditional methods, such as potentiometry, have failed to determine accurate stability constants for HgII-LMM thiol complexes. The unreasonably wide range (>20 orders of magnitude) of overall stability constants (β) reported for HgII-LMM thiol complexes before 1980 was pointed out by Casas and Jones.[14] They concluded that the log β2 constant for the formation of Hg(SR)2, the predominant complex in presence of excess ligand, should be in the range between 40 and 45 for LMM thiols like Cys, Pen and mercaptoacetic acid (MAC).[14,15] These results were based on experiments with a direct determination of the very low concentration of free Hg2+ ions by means of changes in the electrode potential at the surface of a mercury electrode. With a similar approach, Van Der Linden and Beers reported a log β2 of 39.4 for Hg(Cys)2.[16] A similar magnitude of the stability constant for HgII-LMM thiol complexes has been reported using a methodology in which the thermodynamics of the HgII speciation are shifted towards a separable, identifiable and measurable HgII complex with the addition of a competing ligand such as iodide (I-) or bromide (Br-) ions or the lipophilic thiol dithizone.[9,17-19] Even with this progress in methodological development, it remains a challenge to determine the exceptionally high stability constant for HgII-LMM thiol complexes, as exemplified by log β2 constants reported for the Hg(Cys)2 complex in recent literature ranging between 38.2[9] and 43.5.[18] Thus, there is still a substantial uncertainty remaining before a consensus on HgII-LMM thiol stability constants can be reached.
In the present work, we determined stability constants for HgII complexes formed with 15 different LMM thiols that have been reported in terrestrial and aquatic ecosystems.[6,8,20] We used a novel methodology based on competing ligandexchange experiments that takes advantage of a selective direct measurement of specific Hg(SR)2 complexes using liquid chromatography-inductively coupled plasma mass spectrometry (LC-ICPMS).Our resultswere evaluated in light of current knowledge provided by EXAFS spectroscopy measurements of several HgII-LMM thiol complexes.[11-13] Density functional theory (DFT) calculations were used to identify intramolecular interactions that could explain observed differences in stabilities among HgII-LMM thiol complexes. The study focusses on the two-coordinated symmetric type complexes Hg(SR)2, but also considers HgSR+ complexes and hetero complexes R'SHgSR''.
Experimental
Chemicals and reagents
All thiol compounds were purchased from Sigma-Aldrich. Their structures and the abbreviations used throughout this paper are given in Fig. S1, mercaptoethanol (ETH), monothiolglycerol (Glyc), mercaptoacetic acid (MAC), 3-mercaptopropionic acid (3-MPA), 2-mercaptopropionic acid (2-MPA), mercaptosuccinic acid (SUC), N-acetyl-cysteine (NACCys), N-acetyl-penicilamine (NACPen), cysteine (Cys), homocysteine (HCys), N-cysteinylglycine (CysGly), γ-glutamylcysteine (GluCys), glutathione (GSH), penicillamine (Pen), cysteamine (Cyst). In addition to a thiolate group, the different LMM thiol ligands contained hydroxyl, carboxyl, carbonyl, primary and/or secondary amino functional groups. Mercury nitrate monohydrate $99.99 %, (Hg(NO3)2 ⋅ H2O), potassium bromide ≥$99.5% (KBr), sodium chloride ≥$99.5% (NaCl), formic acid (FA), EDTA, sodium perchlorate ≥$98% (NaClO4) and 1-propanol were purchased from Sigma-Aldrich, analytical-grade potassium iodide (KI) from Fisher Scientific and Suprapur nitric acid from Merck. Ultrapure water (>18 MΩ cm) was obtained through a Milli-Q Advantage A10 Ultrapure water purification system (Merck Millipore). All stock solutions and reagents were prepared in a glove-box with a nitrogen (N2) atmosphere (<100 ppm O2). Deoxygenated Milli-Q water was prepared by purging with N2 overnight at a flow rate of 300 mL min-1 in the glove box. Stock solutions of HgII (6.5 mM) were prepared from Hg(NO3)2 ⋅ H2O in 0.12MHNO3. The concentration of HgII stock solution was verified Using reverse isotope dilution analysis with ICPMS and using combustion atomic absorption spectrometry (AMA 254 Leco Corporation). Stock solutions of KI (13.5 mM), KBr (100 mM), KCl (100 mM), EDTA (12 mM), NaClO4 (1.0 M) and LMM thiols (5 mM) were prepared in deoxygenated Milli-Q water inside the glove box. Mobile phases for LC included 1-propanol and Milli-Q water and pH was adjusted with FA to the same pH as the HgII-LMM thiol sample solutions.
LC-ICPMS
The LC-ICPMS instrument consisted of an LC system with two micro pumps (PerkinElmer series 200), a column oven (PerkinElmer series 200), a vacuum degasser (PerkinElmer series 200) and an auto-sampler (PerkinElmer series 200). The temperature of the LC column and sampler tray was thermostatted at 25 °C. The ICPMS (ELAN DRCe, PerkinElmer SCIEX) included a PFA ES-2040-54 nebuliser and a cyclonic spray chamber (thermostatted to +4 °C) from Elemental Scientific Inc. The nebuliser and auxiliary gas flow rates were set to 0.6 and 1.2 L min-1 respectively. An ICP radio frequency power of 1350Wand an ion lens voltage of 10 V were used. The eluting Hg(SR)2 complexes were detected by monitoring the 202Hg+ isotope signal intensity.
The LC electrospray ionisation mass spectrometry (ESIMS) (Thermo Scientific LCQ Fleet Ion Trap) instrument consisted of quaternary LC pumps, an autosampler and a vacuum degasser. The operation parameters of the ESIMS were 300 °C for the capillary temperature, 4.3 kV for the electrospray voltage, 31 V for the capillary voltage, 90 V for the tube lens voltage and 20 and 5 arbitrary units for the sheath and auxiliary gas flow rate respectively. Both negative and positive ionisation modes were used with a mass scan range from m/z 200 to 1000.
A Phenomenex Kinetic Biphenyl LC column (150 mm x 3mm x 5 µm), with a 4 x 3.0-mm guard column, was used with mobile phases includingMilli-Q water and 1-propanol. 1-Propanol was used as organic modifier in the mobile phase owing to its low volatility (minimising excessive solvent loading of the ICP) and enhanced aerosol formation efficiency of the nebuliser, causing increased sensitivity of the ICPMS measurements at a low percentage (2-10 %) of 1-propanol in the eluent.[21] The pH of the mobile phase was adjusted with FA. A flow rate of 0.4 mL min-1 and injection volume of 10 mL were used for the LC. Isocratic elution was used with adjusted mobile phase composition to obtain complete separation of each pair of analysed Hg(SR)2 complexes. Three different proportions of 1-propanol were used depending on whether MAC (3.5 or 8.5% 1-propanol) or 2-MPA (11% 1-propanol) was used as competing ligand. A post-column flow rate of 0.1 mL min-1 of an aqueous solution containing 10 ng mL-1 thallium (Tl) was applied to monitor and correct for signal drift of the ICPMS instrument over time. The chromatographic peak areas of the complexes were calculated with OriginPro 9.1.0 from OriginLab Corporation. Stability constants of the complexes were corrected to certain ion strengths by specific interaction theory (SIT) using the free software Ionic Strength Corrections for Stability Constants from IUPAC, version 1, 2004.[22] Stability constants were calculated with the software WinSGW from MaJo.[23]
Determination of pKa of the RSH group
The pKa values of the RSH group for the 15 investigated LMM thiols were calculated with the atomic charges model presented by Ugur et al.[24] The method builds on a linear relationship between computed molecular charge distribution and empirically determined pKa values for 25 thiol compounds with different functional group composition. The dataset showed the best linear relationship with natural population analysis (NPA)[25] atomic charges on optimised geometries of the anionic form using the Minnesota functional M06-2X[26,27] with the 6-311G basis set and the conductor-like polarisable continuum model (CPCM)[28] with default atomic radii.
Step 1: Determination of the stability constant of Hg(MAC)2 and Hg(2-MPA)2 using I- as a competing ligand
A series of KI solutions with concentrations of 4, 8, 20, 50, 100, 300, 1000 µM were prepared with pH of 2.9 and 3.6 (adjusted with nitric acid) in 15-mL Falcon tubes, in a constant ionic medium of 0.1 M or 1.0 M NaClO4. Lower ionic strength conditions were established by the KI and HNO3 concentrations alone, without NaClO4 addition. Under these conditions, the ionic strength varied between 0.0003 and 0.003 M, which for the experimental purposes of the present study can be considered close to an ion strength, I, of zero (I=0). Samples were prepared at 25 ± 1 °C with a thermostat in a N2-filled glove box. To prepare solutions, different volumes of a KI stock solution (13.5 mM) were mixed with 9.6 µL of a LMM thiol solution (5 mM) and shaken for 10 s. Then, 120 mL of 100 µMHgII nitrate solution, and NaClO4 ionic medium, were added to give final HgII and LMM thiol concentrations of 4 and 16 µM-respectively in a final sample volume of 3 µL. The samples were rotated (end-over-end) inside the glove box for 2 h before measurement to assure equilibrium was achieved. The time required to reach equilibrium for a system with a mixture of two different LMM thiols was shown to be less than 30 min and the complexes formed were shown to be stable up to 5 days or more by replicate injection of the same samples (Figs S2 and S3).
The absolute concentrations of Hg(MAC)2 and Hg(2-MPA)2 complexes were determined by LC-ICPMS using calibration curves and the equilibrium pH of samples was measured after 2 h. Reactions 1-5 are represented by either step-wise (K) or overall (β) thermodynamic constants. Stability constants for the formation of Hg(MAC)2 and Hg(2-MPA)2 complexes (Table 1) were calculated for Reactions 1a and 1b using WinSGWsoftware, with Ka values for the thiol group (Reaction 2) of MAC and 2-MPA determined from DFT calculations and well-established thermodynamic constants for the formation of
complexes (Reactions 3a-d).[24,29] We also determined the concentration of mixed HgI(MAC, 2-MPA) complexes (Reaction S6, supplementary material) from the area of those peaks in relation to the peak areas of the Hg(MAC)2 and Hg(2-MPA)2 complexes. Details on the calculation of the stability constant for the Hg(MAC)2 complex are reported on in S1 (supplementary material).
(1a) Hg2+ + 2RSH ⇌ Hg(SR) KHg(SR)[sub 2]
(1b) Hg2+ + 2RS- ⇌ Hg(SR)2 β2
(2) RSH ⇌ RS- + H+ Ka
(3a) Hg2+ + I- HgI+ K1
(3b) Hg2+ + 2I- ⇌ HgI2 β2
(3c) Due to image rights restrictions, multiple line equation(s) cannot be graphically displayed.
(3d) Due to image rights restrictions, multiple line equation(s) cannot be graphically displayed.
(4) Hg2+ + RS- ⇌ HgSR+ KHgSR+
(5) Hg2+ + R'S- + R'S- ⇌ R'SHgSR'' KR'SHgSR''
Step 2: Determination of the stability constant for Hg(SR)2 complexes using 2-MPA and MAC as competing ligands
The stability constants for Hg(2-MPA)2 and Hg(MAC)2, as determined by competition with I- ions were first validated against each other. An equilibrated solution of 4 µM Hg(NO3)2, 8 µM MACand 8 µM 2-MPA, adjusted to pH 3.0, was examined by LC-ICPMS. First, 10 µL of 2.4 mM MAC and 10 µL of 2.4 mM 2-MPA and 2860 µL of pH 3.0 water solution (pH adjusted with HNO3 acid) were mixed in a 15-mL Falcon tube by strong hand-shaking for 10 s, and then 120 µL of 100 µM Hg (NO3)2 was added. The sample was rotated end-over-end in a glove box filled with N2 for 2 h. After 2 h of equilibration, concentrations of Hg(MAC)2 and Hg(2-MPA)2 were determined from LC-ICPMS chromatogram peak areas and stability constants were calculated by use of WinSGW. As shown in Fig. S2, the concentrations of Hg(SR)2 complexes did not change in the time-window of 10 min to 4 h. Once stability constants for Hg(MAC)2 and Hg(2-MPA)2 were established, constants for Hg(SR)2 complexes with the other 13 LMM thiols were determined in experiments where MAC or 2-MPA were added as competing ligands at concentrations equal to the LMM thiol studied. By using MAC or 2-MPA as competing ligands, we avoided the interference effects of I- (suppressing ionisation efficiency of Hg in the ICP) and
complexes (causing enhanced Hg spectral background signals) on the signal of Hg (SR)2 complexes. The signals of I- and
were well separated from Hg(MAC)2 and Hg(2-MPA)2 (as shown in Fig. S4c, d). Stability constants for Hg(SR)2 complexes were calculated for Reactions 1a and 1b. The complete calculation scheme is exemplified for Hg(Cys)2 in S2 (supplementary material) using MAC as the competing ligand.
Investigation of the possible formation of one-coordinated HgSR complexes and hetero-ligation R'SHgSR'' complexes
The possible formation of one-coordinated HgII complexes (Hg2+ + RS- ⇌ HgSR+ , Reaction 4) with the LMM thiols Cys, HCys, GSH, MAC, Glyc or NACCys was investigated. Samples were prepared at molar ratios of LMM thiol ligands to HgII between 1.0 and 10, at pH 3.0 in an ionic medium approximating 0 M(pH 3.0, HNO3). The samples were rotated inside a N2-filled glove box for 2 h to assure equilibrium was reached. The concentration of Hg(SR)2 complexes was determined by LC-ICPMS. In absence of an apparent peak for the HgSR+ complex (Reaction 4), stability constants were calculated from fitting measured (by LC-ICPMS) and modelled (in WinSGW) Hg(SR)2 concentrations at different HgII to LMM thiol molar ratios using a model including both two- and one-coordinated HgII-LMM thiol complexes. The existence of possible R'SHgSR'' hetero complexes (Reaction 5) was investigated with direct infusion of sample solutions in ESI-MS and with the use of LC-ESIMS and LC-ICPMS.
Density functional theory modelling of Hg(SR)2 complexes
In DFT calculations, the Hg(SR)2 complex was taken as the initial geometry. The two ligands form a linear configuration with protonation states of the molecules' functional groups at pH 3.0. For each complex, geometry optimisations were performed in the gas phase using the B3LYP[30,31] level of theory and a mixed basis set comprising the Def-2TZVPP[32] basis set for Hg and 6-31++G(d,p) for all other atoms. The free energy was determined by the frequency calculation at the same level of theory, followed by a single-point energy calculation at the B3LYP level of theory with Grimme's dispersion and Becke-Johnson damping[33] and the basis set consisting of Def2-TZVPPD[32] for Hg and 6-311++G(d,p) for all other atoms respectively. The single-point energy at the larger basis set was introduced to correct the energy determined with the smaller basis set used in the geometry optimisations and frequency calculations. The B3LYP functional and the basis sets used in the present work were shown previously to produce results comparable with the CCSD(T) level of theory.[34] The same calculations were performed for the Hg2+ ion and each ligand to determine the free energy of complex formation (in the gas phase). The Gaussian 09[35] program suite was used in all DFT calculations.
Results and discussion
Spectroscopic methods (i.e. 1H NMR, Hg LIII-edge EXAFS) have established that HgII forms a two-coordinated linear structure with two separate LMM thiol ligand molecules, i.e. Hg (SR)2, in which each thiol group forms a monodentate complex. It has also been shown that three- and four-coordinated HgII-LMM thiol complexes may form at neutral to alkaline pH, but not at acidic pH. Because the primary aim of the current study was to characterise two-coordinated Hg(SR)2 complexes, the experimental pH was kept at acidic conditions of 2.9 or 3.6 to prevent formation of complexes with higher coordination.[ 11-13,36] The two-coordinated structure Hg(SR)2 was verified by LC-ICPMS and ESIMS as the only detectable molecular stoichiometry of HgII-LMM thiol complexes formed in this study. ESI mass spectra showing the molecular mass and isotope pattern of Hg(SR)2 complexes are given in Fig. S5.
Calculation of pKa values for LMM thiols
Depending on the pH value, the pKa value of the RSH group is of great importance for the stability constant of Hg(SR)2 complexes formed via Reaction 1a. The pKa value must therefore be known for each thiol compound, and should ideally be determined with a consistent methodology for all compounds to be compared. In the literature, only 10 of the 15 LMM thiols included in the present study have reported pKa values for the RSH group. Those values vary owing to different ionic strengths and experimental approaches used. Therefore, in order to make our determined stability constants comparable for the different Hg(SR)2 complexes, we calculated the pKa value for the RSH group of all LMM thiol compounds using the atomic charges model developed by Ugur et al.[24] These values, reported in Table 2, are in good agreement with the pKa values reported for most of the LMM thiols available in the literature when corrected for ionic strength.[9,13,15,17-19,37,38] The pKa value of the RSH group of LMM thiols studied here ranged between 7.3 and 10.8, with CysGly having the lowest and 3-MPA the highest value. Amino groups, particularly the protonated primary amino moiety (
), are electron-withdrawing and thus lower the pKa value of the corresponding RSH group. LMM thiols with one primary amino group such as Cys, HCys, Cyst, CysGly, GSH, GluCys and Pen have pKa values in the range between 7.3 and 10.3, with an average of 9.1. LMM thiols without an amino group, such as MAC, 2-MPA, 3-MPA, ETH, Glyc and SUC, have higher pKa values, in the range between 9.4 and 10.8, with an average of 10.2. For LMM thiols comprising the same functional groups, compounds with the amino group located closer to the RSH group have smaller pKa values: for example, Cys: pKa 8.6, versus HCys: pKa 9.9.
Determination of stability constants for Hg(MAC)2 and Hg(2-MPA)2 with I- as the competing reference ligand
In the first step of our analyses, we determined stability constants of Hg(MAC)2 and Hg(2-MPA)2 in competition with I- In the second step, MAC and 2-MPA were used as competing ligands for the determination of HgII complexation with the other 13 LMM thiols. The rationale for using I- as competing ligand in the first step is that, in contrast to most Hg(SR)2 complexes, the stability constants for
complexes are well established.[29] Other potential competing ligands such as Br-, Cl- and EDTA were also investigated. Our results showed that the concentration of Hg(MAC)2 was not significantly lowered even when adding up to 10 mM of those ligands (at [HgII] = 2 µM and [MAC] = 8 µM) (Fig. S6). This clearly demonstrates that these three ligands are all too weak to compete with MAC and 2-MPA to the extent that complexes are detectable by LC-ICPMS. Given that the stability constants for Hg(MAC)2 and Hg(2-MPA)2 were established by competition with I- , it is advantageous to determine the stability constants for the remaining 13 Hg(SR)2 complexes in a second step using MAC or 2-MPA as competing ligands. The main reason for selecting MAC and 2-MPA in the first step was that they were well separated in the LC column from both I- ions (causing signal suppression) and
complexes (Fig. S4). The Hg (MAC)%5F and Hg(2-MPA)2 were also well separated from all the other Hg(SR)2 complexes. The retention times on the LC column of all investigated Hg(SR)2 complexes are shown in Fig. S7. Further, the absence of amino groups in MAC and 2-MPA may result in better-resolved pKa values of the RSH group.
In Figs 1a, b and 2a, b, the determined concentrations of Hg(MAC)2 and Hg(2-MPA)2 are illustrated as a function of I- concentration. Log KHg(SR)2 and log β2 (Reactions 1a and 1b respectively) for the formation of Hg(MAC)2 and Hg(2-MPA)2 were determined according to the calculation scheme S1 (supplementary material). The constants were calculated for each addition of I- based on measured concentrations of Hg(MAC)2 and Hg(2-MPA)2 from LC-ICPMS and established constants
complexes (Table S1). Average values for log β2 for Hg(MAC)2 and Hg(2-MPA)2 were determined from all experiments and the deviation of data points from the linear equation of the plots illustrated in Figs 1c, d and 2c, d was used to calculate uncertainties in this constant, as reported in Table 1. The average values of log β2 of Hg(MAC)2 and Hg(2-MPA)2 were determined to be 40.9 ± 0.2 and 41.5 ± 0.1 (I = 0 M) respectively, which both are in fair agreement with previous reports by Basinger et al.[15] and Cardiano et al.[19] The formation of a mixed iodide-thiol ligation complex (HgISR) was observed forMACand 2-MPA when the molar ratio of I- to HgII was higher than 25 but not exceeding 250 (the molar ratio of LMM thiol to HgII was kept constant at 4), as shown in Fig. S4. Mixed halide-thiol complexes, in particular HgClSR, have been suggested to form by Hilton et al. as determined by NMR experiments.[39] In the present study, HgISR complexes were identified based on observed LC-ICPMS signals of both Hg and iodine, shown in Fig. S4. The average log K of HgI(MAC) and HgI(2-MPA) following the reaction Hg2+ + RS- + I- ⇌ HgISR was determined to be 32.2 ± 0.1 and 32.3 ± 0.1 respectively.
Determination of the stability constants for 13 LMM Hg(SR)2 complexes using MAC and 2-MPA as competing reference ligands
The stability constants for Hg(MAC)2 and Hg(2-MPA)2 determined with I- as competing ligand were cross-validated against each other using MAC as competing ligand for Hg(2-MPA)2. The log β2 (Reaction 1b) for Hg(2-MPA)2 determined with this approach was 41.6, essentially identical to the log β2 of 41.5 determined using I- as the competing ligand. This validation enabled us to use either MAC or 2-MPA as competing ligands for the determination of log β2 for the remaining 13 LMM thiols.
In Table 2, stability constants for the formation of Hg(SR)2 complexes between HgII and all 15 LMM thiols are reported. At low pH, the RSH group is fully protonated and complex formation with Hg2+ may be described by Reaction 1a: Hg2+ + 2RSH ⇌ Hg(SR)2 + 2H+ . The log KHg(SR)2 of this reaction varies substantially between 19.6 and 21.0, with Hg (Cyst)2 having the smallest and Hg(2-MPA)2 the largest value. When the reaction is written with the deprotonated thiolate group (Reaction 1b), log β2 varies from 34.6 for Hg(CysGly)2 to 42.1 for Hg(3-MPA)2. The relation between the two constants is log β2 = log KHg(SR)2 + 2pKa, where pKa relates to Reaction 3. The pKa value thus has a strong influence on the value of log β2. This quite large variability in thermodynamic constants shows that there are significant differences in the stability of different Hg-LMM thiol complexes depending on the chemical structure of the thiol ligand.
Most previous studies reporting stability constants for HgII-LMM thiol complexes rely on methods that do not directly quantify the Hg(SR)2 complexes, typically electrochemical or radiochemical detection of HgII in the presence of a competing ligand.[15,17] Chemical shifts measured by 1H NMR spectroscopy for LMM thiols with and without addition of HgII have also been used to calculate stability constants for Hg(SR)2 complexes.[18] Compared with previous studies reporting log b2 constants of the order of 38 to 44 for Hg(SR)2 complexes, the constants determined in our study are in fair agreement for HgII complexes with LMM thiols lacking amino groups: Hg(2-MPA) 2, Hg(SUC)2, Hg(3-MPA)2, Hg(MAC)2 (Table 2). For complexes with molecules containing amino groups: Hg(Pen)2, Hg(GSH)2, Hg(Cys)2 (Table 2), our determined constants are significantly lower than in previous studies.[9,13,15,17-19]
Our experimental approach, based on a direct quantification of specific Hg(SR)2 complexes, sets a new standard for the determination of thermodynamic constants for HgII-thiol complexes. The constants determined for the two complexes Hg (MAC)2 and Hg(2-MPA)2 are central in our methodology because they are used as references for the other 13 Hg(SR)2 complexes. The constants for Hg(MAC)2 and Hg(2-MPA)2 were determined with a fairly large number of data points covering extensive ranges of pH and ionic strength (Figs 1 and 2). Constants for the other 13 Hg(SR)2 complexes covered quite a diversity of LMM thiols, with a multitude of different environmentally relevant functional groups (Fig. S1).
Formation of one-coordinated HgSR+ complexes
According to previous studies,[18,19] model fitting to experimental data suggests the presence of a complex having a 1 : 1 stoichiometry between the Hg2+ ion and LMM thiol ligand (Reaction 4). Because HgII prefers two-coordination, it is expected that such HgSR+ complexes in addition to the thiol group involve coordination with one carboxyl or carbonyl oxygen group, or amino group, from the same LMM thiol molecule.[14,15,18,19] In order to test the existence of such complexes, we conducted experiments at RSH to HgII molar ratios of 1.0. Candidate peaks indicative of HgSR+ complex formation were detected by LC-ICPMS, but with increased peak broadening as compared with the Hg(SR)2 peak (Fig. S8). The reason could be that the fairly low stability of the one-coordinated complexes led to some degradation during separation on the LC column (possibly by interactions with FA in the mobile phase). Because of the quite large uncertainty in quantifying the area of these small and broad chromatographic peaks, log K for the HgSR+ complexes was instead determined by fitting the measured concentration of Hg(SR)2 to a model including both Hg (SR)2 and HgSR+ complexes. During fitting, log KHgSR+ for the reaction Hg2+ + RS- ⇌ HgSR+ (Reaction 4) was optimised, while keeping the value of log β2 for the Hg(SR)2 complex fixed. Model fitting was done on data for the formation of Hg(SR)2 and HgSR+ in experiments with Cys, HCys, GSH, MAC, Glyc, and NACCys ligands present. Fig. 3 shows the average Hg(SR)2 concentration for these six Hg(SR)2 complexes measured by LC-ICPMS and calculated by model fits with fixed log β2 for Hg(SR)2, pKa (RSH) (Table 2) and pKa (RCOOH). The log K for the formation of HgSR+ was successively varied from a value of 29 to 32, in steps of 0.5. The best fit was obtained with a log KHgSR+ value in the range between 30.5 and 31. Log KHgSR+ for the formation of HgSR+ was thus ~8.5 orders of magnitude smaller than the average log β2 of 39.2 for the formation of Hg(SR)2. This difference in binding strength between HgII coordinated by two or one RS- group is in fair agreement with previous results.[14,15,18,19] Owing to a lower stability of the HgSR+ complex as compared with Hg(SR)2, the concentration of HgSR+ can be considered negligible when the molar ratio of RSH to HgII exceeds 2. This implies that the Hg (SR)2 complex was by far the predominant form of HgII complexes under the experimental conditions used for the determination of stability constants for Reactions 1a and 1b.
Formation of R'SHgSR'' hetero complexes
The thermodynamic stability of R'SHgSR'' hetero complexes (where two different types of LMMRSH groups are involved in complex formation) relative to Hg(SR)2 complexes is important because the complexity of HgII speciation models may increase considerably if R'SHgSR'' hetero complexes need to be included in the model. The LC-ICPMS and LC-ESIMS chromatograms did not include information indicative of any R'SHgSR'' hetero complexes, such as CysHgMAC (Figs S9 and S10). The molar ratio of individual LMM thiols to HgII was fixed at 2.0 in these experiments (8 µM Cys and 8 µM MAC to 4 mM HgII). The absence of detectable R'SHgSR'' complexes could be due to these complexes being less stable than Hg(SR)2 complexes or the lifetime of hetero complexes being shorter. Pei et al.[40] observed a shift in the retention time of Hg(GSH)2 with increased concentration of Cys in the mobile phase of an LC system and proposed that the CysHgGSH hetero complex was formed. We observed the signal of R'SHgSR'' complexes with direct infusion of sample solutions in ESIMS (without an LC column, shown in Fig. S11). This observation may, however, be explained by the artefactual formation of R'SHgSR'' in the ion source. In the electrospray ion source, complex re-formation reactions are common and products with fast kinetics are preferentially formed. The formation of HgII-LMM thiol complexes has been demonstrated by 1H-NMR to be thermodynamically stable but kinetically very labile.[40-43] In line with the NMR spectroscopy observations, previous studies have shown that HgII reached equilibrium with LMM thiols within 30 s.[40,42,43] Liquid chromatography with ICPMS and ESIMS detection indicated that 100 ± 2% of the HgII concentration of 4 µM was represented by the Hg(SR)2 complex in our experiments. If it is assumed that a maximum 2% of HgII was in the form of the R'SHgSR'' hetero complex, the stability constant for this complex would be at least 1.5 log units smaller than the average stability constant for the corresponding Hg(SR)2 complexes.
Molecular modelling of Hg(SR)2 complex structure and stability
To gain insights into how chemical interactions and structures may control differences in the stability constant for HgII-LMM thiol complexes, molecular modelling based on DFT was performed for the Hg(SR)2 complexes. There was a good qualitative agreement between the relative stability of HgII-LMM thiol complexes determined experimentally and predicted by in silico modelling of the complexes in the gas phase, with the Hg(GSH)2 complex being the only exception. The experimentally determined log KHg(SR)[sub 2] values were continuously distributed in the range 19.6 to 21.0, whereas the modelled DG values separated the complexes into two groups (Fig. 4). The calculated Hg-S bond distances and S-Hg-S bond angles varied between 2.36-2.38 Å and 171-1808 respectively for all the complexes investigated (Table S2). The Hg-S bond distance for HgII-thiol complexes has previously been estimated at 2.34 Å ,[44] which is also the typical bond distance determined by EXAFS spectroscopy for both LMM thiols[10-13] and thiols associated with natural organic matter.[45] Interestingly, the measured and modelled differences in the stability of HgII-LMM thiol complexes were not random but systematically dependent on functional groups neighbouring the RSH groups. The DFT calculations indicated two predominant intrinsic intracomplex interactions that can explain these differences.
The presence of a primary amino group resulted in weaker HgII-LMM thiol complexes compared with molecules lacking such a functional group (blue versus red symbols respectively in Fig. 4). The difference can be understood by the strong electron-withdrawing effect of the
group (protonated at pH < ~10), which lowers the stability of the Hg-S bond by making the -S- group less negative. By contrast, electron-donating groups, such as the carbonyl and carboxylic groups, can contribute to the stabilisation of the HgII-LMM thiol complexes through additional electrostatic attraction to HgII besides the covalent bonding with linearly coordinated thiol groups (Fig. 5 and Table S2).[46] The opposing effects between the electron-withdrawing primary amino group and electron-donating oxygens on the stability of Hg(SR)2 complexes can be illustrated by comparing the Hg(Cys)2, Hg(3-MPA)2 and Hg(2-MPA)2 complexes (Fig. 5). The only structural differences between the Hg(Cys)2 and Hg(3-MPA)2 complexes is the presence of a primary amino group in Hg(Cys)2, contributing to the log KHg(SR)2 for Reaction 1a being half a log unit lower for Hg(Cys)2 than for Hg(3-MPA)2. The DFT modelling further suggests that an attraction between Hg and carboxylic oxygen in the Hg(2-MPA)2 complex, but less so in the Hg(3-MPA)2 complex, results in half a log unit higher log KHg(SR)2 for Hg (2-MPA)2 compared with Hg(3-MPA)2. Thus, the degree of destabilisation or stabilisation differs for the 15 complexes depending on the location of the additional functional groups relative to the -S position in the molecule. DFT modelling further predicts a lower ΔG value for Hg(GSH)2 compared with the other Hg(SR)2 complexes containing a primary amino group. Because GSH involves many rotatable bonds, free energies estimated based on a single optimised geometry may not be appropriate and multiple geometry sampled by advanced sampling techniques, such as molecular dynamics and Monte Carlo simulations, would enhance the accuracy of the determined free energies.
Environmental implications
There are several studies investigating the detailed Hg-S structure in Hg(SR)n complexes for variable n,[11,44,47-49] often with the perspective of designing optimised HgII chelating compounds (at physiological pH). There are few studies investigating the more subtle differences in the stability of Hg(SR)2 complexes induced by weaker electrostatic interactions with O and N functional groups[12,13] in addition to the covalent S-Hg-S coordination. Our experimental and modelling results show that despite the very strong Hg-S bond, there are differences in the stability of Hg(SR)2 complexes (KHg(SR)2 spans 1.5 orders of magnitude, Table 2) that can be explained by the presence of electron-withdrawing and electron-donating functional groups in the vicinity of the RSH group. These systematic differences may have substantial effects on the chemical speciation of HgII in environmental and biological systems where several LMM thiol ligands are present in similar concentrations, which in turn affects rates of environmentally important Hg transformation reactions, including the formation of the very toxic MeHg molecule. The present work thus significantly advances our fundamental understanding of interactions between HgII and thiol ligands at the molecular level. Our view is that the experimental approach reported here, combined with the DFT calculations, further presents a robust and comprehensive set of thermodynamic constants for HgII-thiol complexes. This new information is critical for an accurate modelling of the chemical speciation of HgII in natural environments, which in turn is necessary when elucidating the role of HgII-LMM thiol complexes behind Hg transformation reactions in natural waters, soils and sediment.
Supplementary material
Details on stability constant calculation schemes and analytical method optimisation results are available from the Journal's website.
Acknowledgements
This work was financially supported by the Kempe Foundations (Grant no. SMK-2745, SMK-2840), the JC Kempe Memorial Scholarship Foundation, the Swedish Research Council (Grant no. 2015-04114), the Swedish University of Agricultural Sciences and Umeå University. Ilke Ugur and Gerald Monard, Université de Lorraine, are gratefully acknowledged for pKa determinations of the LMM thiols. Tomas Hedlund is gratefully acknowledged for assistance with the WinSGW software. All DFT calculations were conducted using the resources provided by the Swedish National Infrastructure for Computing (SNIC) at the High-Performance Computing Center North (HPC2N) and National Supercomputing Center (NSC).
Table 1. Overall thermodynamic constants (i.e. log α2, Reaction 1b) for the formation of Hg(MAC)2 and Hg(2-MPA)2 complexes and corresponding constants for mixed complexes including one I- and one low-molecular-mass (LMM) thiol ligand at different ionic strengths (I) and pH
MAC, mercaptoacetic acid; 2-MPA, 2-mercaptopropionic acid; RSH, thiol compound
| Complexes | Log β[sub 2] (±s.d.) |
| pH 2.9 | pH 3.6 |
| I = 0 (M)A | I = 0.1 (M)B | I = 1.0 (M)B | I = 0 (M)A | I = 0.1 (M)B | I = 1.0 (M)B |
| Hg(MAC)[sub 2] | 40.9 (±0.2) | 40.5 (±0.2) | 40.1 (±0.1) | 40.1 (±0.1) | 40.3 (±0.2) | 40.2 (±0.2) |
| Hg(2-MPA)[sub 2] | 41.5 (±0.1) | 40.9 (±0.1) | 40.3 (±0.1) | 40.3 (±0.1) | 40.7 (±0.1) | 40.4 (±0.2) |
| HgIMAC | 32.2 (±0.1) | 32.0 (±0.1) | 31.7 (±0.2) | 31.7 (±0.2) | 31.8 (±0.2) | 31.7 (±0.1) |
| HgI(2-MPA) | 32.3 (±0.1) | 32.1 (±0.2) | 31.9 (±0.10) | 31.9 (±0.10) | 31.9 (±0.2) | 31.8 (±0.1) |
A The pKa values of the RSH groups of MAC and 2-MPA at I = 0 were determined to be 10.2 and 10.3 respectively based on the model developed by Ugur et al.[24]
B The pKa values of the RSH groups were reported by Cardiano et al.[19] to be 10.0 at I = 0.1, and 9.8 at I = 1.0 M for MAC, and 10.1 at both I = 0.1 M and 1.0 M for 2-MPA.
Table 2. Step-wise (log K) and overall (log β2) thermodynamic constants for the formation of Hg(RS)2 complexes with low-molecular-mass (LMM) thiols, as described by Reactions 1a and 1b, and pKa values for RSH groups in Reaction 2
Complexes are sorted according to decreasing value of log KHg(SR)2 for Reaction 1a at an ion strength (I) of 0 M and pH 3.0. Literature values are reported with and without correction for ion strength effects and pKa values. ETH, mercaptoethanol; Glyc, monothiolglycerol; MAC, mercaptoacetic acid; 3-MPA, 3-mercaptopropionic acid; 2-MPA, 2-mercaptopropionic acid; SUC, mercaptosuccinic acid; NACCys, N-acetyl-cysteine; NACPen, N-acetyl-penicilamine; Cys, cysteine; HCys, homocysteine; CysGly, N-cysteinylglycine; GluCys, γ-glutamylcysteine; GSH, glutathione; Pen, penicillamine; Cyst, cysteamine
| Complexes | Thermodynamic constant (±s.d.) |
| Present study, I = 0 (M) | LiteratureA log β[sub 2] (Reaction 1b) | LiteratureB log β[sub 2] (Reaction 1b) |
| log K (Reaction 1a) | log β[sub 2] (Reaction 1b) | pK[sub a] (Reaction 2) |
| Hg(2-MPA)[sub 2] | 21.0 | 41.5 (±0.1) | 10.3C | 42.68[19] | 43.2 |
| Hg(NACPen)[sub 2] | 20.9 | 40.1 (±0.2) | 9.6C | | |
| Hg(NACCys)[sub 2] | 20.6 | 40.2 (±0.2) | 9.8D | 41.81[9] | 42.0 |
| Hg(SUC)[sub 2] | 20.6 | 41.7 (±0.2) | 10.6C | 42.92[19] | 43.3 |
| Hg(3-MPA)[sub 2] | 20.6 | 42.1 (±0.2) | 10.8C | 39.54[19] | 40.9 |
| Hg(MAC)[sub 2] | 20.6 | 40.9 (±0.2) | 10.2D | 40.5[15] | 40.8 |
| Hg(Glyc)[sub 2] | 20.6 | 39.4 (±0.2) | 9.4C | | |
| Hg(GluCys)[sub 2] | 20.5 | 40.3 (±0.3) | 9.9C | | |
| Hg(ETH)[sub 2] | 20.5 | 40.3 (±0.2) | 9.9C | | |
| Hg(Pen)[sub 2] | 20.3 | 36.9 (±0.2) | 8.3C | 38.2,[9] 43.51[18] | 39.0, 44.0 |
| Hg(GSH)[sub 2] | 20.2 | 38.8 (±0.2) | 9.3D | 40. 36[18] | 41.8 |
| Hg(Cys)[sub 2] | 20.1 | 37.5 (±0.2) | 8.6C | 40.0,[17] 43.41[18] | 40.3, 44.2 |
| Hg(CysGly)[sub 2] | 19.9 | 34.6 (±0.3) | 7.3C | | |
| Hg(HCys)[sub 2] | 19.7 | 39.4 (±0.2) | 9.9C | | |
| Hg(Cyst)[sub 2] | 19.6 | 40.3 (±0.2) | 10.3D | | |
A Literature values without correction for differences in ion strength.
B Literature values corrected to I = 0 and based on pKa values reported by Ugur et al.[24] to be comparable with constants determined in the present study.
C Computed from natural population analysis (NPA) atomic charges on optimised geometrics of the anionic form using M062X/6-311G and conductor-like polarisable continuum model (CPCM) model developed by Ugur et al.[24] (Reaction 1a: Hg2+ + 2RSH ⇌ Hg(SR)2 + 2H+; Reaction 1b: Hg2+ + 2RS- ⇌ Hg(SR)2; Reaction 2: RSH ⇌ RS- + H+).
D Reference from Ugur et al.[24]
Fig. 1. (a, b) Determined concentration of Hg(MAC)2 as a function of the concentration of competing ligand I- between 0 and 1000 µM. Black dotted lines are the modelled concentration of the Hg(MAC)2 complex from WinSGW using the optimised stability constant of log β2=40.9 for Hg(MAC)2. (c, d) Correlation between measured and calculated Hg(MAC)2 concentrations from WinSGW using the optimised stability constant of log β2 = 40.9 for the formation of Hg(MAC)2 following Reaction 1b. Experiments were conducted at T = 25 °C with two different pHs of 2.9 and 3.6 and three different ionic strengths (NaClO4) of 0, 0.1 and 1 M. (MAC, mercaptoacetic acid.).
Fig. 2. (a, b) Determined concentration of Hg(2-MPA)2 as a function of the concentration of competing ligand I- between 0 and 1000 µM. Black dotted lines are the modelled concentration of Hg(2-MPA)2 complex using the optimised stability constant of log β2 = 40.9 for Hg(2-MPA)2. (c, d) Correlation between measured and calculated Hg(2-MPA)2 concentrations using the optimised stability constant of log β2 = 40.9 for the formation of Hg(2-MPA)2 following Reaction 1b. Experiments were conducted at T = 25 °C with two different pH of 2.9 and 3.6 and three different ionic strengths (NaClO4) of 0, 0.1 and 1M. (2-MPA, 2-mercaptopropionic acid.).
Fig. 3. Average concentration of the Hg(SR)2 complex as a function of molar ratio of low-molecular-mass (LMM) thiol to HgII between 1.5 and 10. Experimental data (average±s.d., as represented by the solid blue line) were collected from separate experiments with 4 µMHgII and varying concentrations of six different LMM thiols (Cys, HCys, GSH, MAC, Glyc and NACCys). Dashed lines represent the modelled average concentrations of the Hg(SR)2 complex with log KHgSR+ (Reaction 4) for the HgSR+ complex varying between 30.0 and 31.5. (Cys, cysteine; HCys, homocysteine; GSH, glutathione; MAC, mercaptoacetic acid; Glyc, monothiolglycerol; NACCys, N-acetyl-cysteine.).
Fig. 4. Comparison of the experimentally determined stability constant log KHg(SR)2 and the modelled (gas-phase) Gibbs free energy (in kcal mol-1; 1 kcal mol-1 = 4.186 kJ mol-1) for the formation of the 15 Hg(SR)2 complexes. Thiol ligands with a primary amino group are indicated by blue circles and their chemical structures are displayed at the top of the figure. Thiol ligands lacking a primary amino group are indicated by red circles and their chemical structures are displayed at the bottom of the figure. The thiol structures are arranged from the top left to the bottom right according to increased experimentally determined stability constant for the corresponding Hg(SR)2 complex.
Fig. 5. Illustration of intracomplex interactions affecting the stability of HgII-LMM (low-molecular-mass) thiol complexes. The presence of an electron-withdrawing primary amino group destabilises the Hg-S bond as illustrated by comparing the Hg(Cys)2 and Hg(3-MPA)2 complexes. The only structural differences between the complexes are the presence of a primary amino group in Hg(Cys)2. Coordination to Hg2+ of electrondonating groups in addition to the linear S-Hg-S configuration enhances the stability of the HgII-LMM thiol complex as illustrated by comparing the Hg(3-MPA)2 and Hg(2-MPA)2 complexes. Density functional theory (DFT) modelling suggested that additional coordination to HgII by carboxylic oxygen is present in the Hg(2-MPA)2 complex but not in the Hg(3-MPA)2 complex. (Cys, cysteine; 3-MPA, 3-mercaptopropionic acid; 2-MPA, 2-mercaptopropionic acid.).
[
References
[1]
U. Skyllberg, Competition among thiols and inorganic sulfides and polysulfides for Hg and MeHg in wetland soils and sediments under suboxic conditions: illumination of controversies and implications for MeHg net production. J. Geophys. Res. Biogeosci. 2008, 113, G00C03.
[2]
J. M. Parks, A. Johs, M. Podar, R. Bridou, R. A. Hurt, S. D. Smith, S. J. Tomanicek, Y. Qian, S. D. Brown, C. C. Brandt, A. V. Palumbo, J. C. Smith, J. D. Wall, D. A. Elias, L. Y. Liang, The genetic basis for bacterial mercury methylation. Science 2013, 339, 1332. doi:10.1126/ SCIENCE.1230667
[3]
C. C. Gilmour, M. Podar, A. L. Bullock, A. M. Graham, S. D. Brown, A. C. Somenahally, A. Johs, R. A. Hurt, K. L. Bailey, D. A. Elias, Mercury methylation by novel microorganisms from new environments. Environ. Sci. Technol. 2013, 47, 11810. doi:10.1021/ ES403075T
[4]
J. K. Schaefer, F. M. M. Morel, High methylation rates of mercury bound to cysteine by Geobacter sulfurreducens. Nat. Geosci. 2009, 2, 123. doi:10.1038/NGEO412
[5]
J. K. Schaefer, S. S. Rocks, W. Zheng, L. Y. Liang, B. H. Gu, F. M. M. Morel, Active transport, substrate specificity, and methylation of Hg (II) in anaerobic bacteria. Proc. Natl. Acad. Sci. USA 2011, 108, 8714. doi:10.1073/PNAS.1105781108
[6]
V. Liem-Nguyen, S. Bouchet, E. Björn, Determination of sub-nanomolar levels of low molecular mass thiols in natural waters by liquid chromatography-tandem mass spectrometry after derivatization with p-(hydroxymercuri)benzoate and online preconcentration. Anal. Chem. 2015, 87, 1089. doi:10.1021/AC503679Y
[7]
C. L. Dryden, A. S. Gordon, J. R. Donat, Seasonal survey of copper-complexing ligands and thiol compounds in a heavily utilized, urban estuary: Elizabeth River, Virginia. Mar. Chem. 2007, 103, 276. doi:10.1016/J.MARCHEM.2006.09.003
[8]
J. Zhang, F. Wang, J. D. House, B. Page, Thiols in wetland interstitial waters and their role in mercury and methylmercury speciation. Limnol. Oceanogr. 2004, 49, 2276. doi:10.4319/LO.2004.49. 6.2276
[9]
H. Kõszegi-Szalai, T. Paal, Equilibrium studies of mercury(II) complexes with penicillamine. Talanta 1999, 48, 393. doi:10.1016/S0039-9140(98)00258-6
[10]
V. Mah, F. Jalilehvand, Glutathione complex formation with mercury (II) in aqueous solution at physiological pH. Chem. Res. Toxicol. 2010, 23, 1815. doi:10.1021/TX100260E
[11]
F. Jalilehvand, B. O. Leung, M. Izadifard, E. Damian, Mercury(II) cysteine complexes in alkaline aqueous solution. Inorg. Chem. 2006, 45, 66. doi:10.1021/IC0508932
[12]
B. O. Leung, F. Jalilehvand, V. Mah, Mercury(II) penicillamine complex formation in alkaline aqueous solution. Dalton Trans. 2007, 4666. doi:10.1039/B711436B
[13]
F. Jalilehvand, K. Parmar, S. Zielke, Mercury(II) complex formation with N-acetylcysteine. Metallomics 2013, 5, 1368. doi:10.1039/ C3MT00173C
[14]
J. Casas, M. M. Jones, Mercury(II) complexes with sulfhydryl containing chelating agents: stability constant inconsistencies and their resolution. J. Inorg. Nucl. Chem. 1980, 42, 99. doi:10.1016/0022-1902(80)80052-2
[15]
M. A. Basinger, J. Casas, M. M. Jones, A. D. Weaver, N. H. Weinstein, Structural requirements for Hg(II) antidotes. J. Inorg. Nucl. Chem. 1981, 43, 1419. doi:10.1016/0022-1902(81)80058-9
[16]
W. Van Der Linden, C. Beers, Determination of the composition and the stability constants of complexes of mercury(II) with amino acids. Anal. Chim. Acta 1974, 68, 143. doi:10.1016/S0003-2670(01) 85155-2
[17]
J. Starý, K. Kratzer, Radiometric determination of stability constants of mercury species complexes with L-cysteine. J. Radioanal. Nucl. Chem. 1988, 126, 69. doi:10.1007/BF02164804
[18]
P. Cardiano, G. Falcone, C. Foti, S. Sammartano, Sequestration of Hg2+ by some biologically important thiols. J. Chem. Eng. Data 2011, 56, 4741. doi:10.1021/JE200735R
[19]
P. Cardiano, D. Cucinotta, C. Foti, O. Giuffre, S. Sammartano, Potentiometric, calorimetric, and 1H NMR investigation on Hg2+-mercaptocarboxylate interaction in aqueous solution. J. Chem. Eng. Data 2011, 56, 1995. doi:10.1021/JE101007N [20] H. Hu, S. E. Mylon, G. Benoit, Distribution of the thiols glutathione and 3-mercaptopropionic acid in Connecticut lakes. Limnol. Oceanogr. 2006, 51, 2763. doi:10.4319/LO.2006.51.6.2763
[21]
Y. Nygren, E. Bjorn, Mobile phase selection for the combined use of liquid chromatography-inductively coupled plasma mass spectrometry and electrospray ionisation mass spectrometry. J. Chromatogr. A 2010, 1217, 4980. doi:10.1016/J.CHROMA.2010.05.062
[22]
IUPAC, Ionic Strength Corrections for Stability Constants using Specific Interaction Theory (SIT), Version 1 2004 (International Union of Pure and Applied Chemistry: Zürich, Switzerland).
[23]
M. Karlsson, J. Lindgren, WinSGW 2006 (Majo: Umeå, Sweden). Available at http://www.winsgw.se/WinSGW%5Feng.htm [verified 4 June 2017].
[24]
I. Ugur, A. Marion, S. Parant, J. H. Jensen, G. Monard, Rationalization of the pKa values of alcohols and thiols using atomic charge descriptors and its application to the prediction of amino acid pKas. J. Chem. Inf. Model. 2014, 54, 2200. doi:10.1021/CI500079W
[25]
A. E. Reed, R. B. Weinstock, F. Weinhold, Natural population analysis. J. Chem. Phys. 1985, 83, 735. doi:10.1063/1.449486
[26]
Y. Zhao, D. G. Truhlar, The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor. Chem. Acc. 2008, 120, 215. doi:10.1007/S00214-007-0310-X
[27]
Y. Zhao, D. G. Truhlar, Density functionals with broad applicability in chemistry. Acc. Chem. Res. 2008, 41, 157. doi:10.1021/AR700111A
[28]
V. Barone, M. Cossi, Quantum calculation of molecular energies and energy gradients in solution by a conductor solvent model. J. Phys. Chem. A 1998, 102, 1995. doi:10.1021/JP9716997
[29]
A. Martell, R. Smith, R. Motekaitis, National Institute of Standard and Technology (NIST), Critically Selected Stability Constants of Metal Complexes. 2004 (PC-based Database: Gaithersburg, MD).
[30]
C. Lee, W. Yang, R. G. Parr, Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B 1988, 37, 785. doi:10.1103/PHYSREVB.37.785
[31]
A. D. Becke, Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98, 5648. doi:10.1063/1.464913
[32]
F. Weigend, R. Ahlrichs, Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297. doi:10.1039/B508541A
[33]
S. Grimme, S. Ehrlich, L. Goerigk, Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 2011, 32, 1456. 10.1002/jcc.21759.
[34]
M. M. Montero-Campillo, A. M. Lamsabhi, O. Mó, M. Yáñez, Alkyl mercury compounds: an assessment of DFT methods. Theor. Chem. Acc. 2013, 132, 1328. doi:10.1007/S00214-012-1328-2
[35]
M. J. Frisch, G. W. Trucks, H. B. Schlegel, et al., Gaussian 09, Revision D. 01 2013 (Gaussian. Inc.: Wallingford, CT).
[36]
T. Warner, F. Jalilehvand, Formation of Hg(II) tetrathiolate complexes with cysteine at neutral pH. Can. J. Chem. 2016, 94, 373. doi:10.1139/CJC-2015-0375
[37]
W. Stricks, I. Kolthoff, Reactions between mercuric mercury and cysteine and glutathione. Apparent dissociation constants, heats and entropies of formation of various forms of mercuric mercaptocysteine and -glutathione. J. Am. Chem. Soc. 1953, 75, 5673. doi:10.1021/JA01118A060
[38]
P. D. Oram, X. Fang, Q. Fernando, P. Letkeman, D. Letkeman, The formation constants of mercury(II)-glutathione complexes. Chem. Res. Toxicol. 1996, 9, 709. doi:10.1021/TX9501896
[39]
B. Hilton, M. Man, E. Hsi, R. Bryant, NMR studies of mercurial-halogen equilibria. J. Inorg. Nucl. Chem. 1975, 37, 1073. doi:10.1016/0022-1902(75)80707-X
[40]
K. L. Pei, M. Sooriyaarachchi, D. A. Sherrell, G. N. George, J. Gailer, Probing the coordination behavior of Hg2+, CH3Hg+, and Cd2+ towards mixtures of two biological thiols by HPLC-ICP-AES. J. Inorg. Biochem. 2011, 105, 375. doi:10.1016/J.JINORGBIO.2010.11.019
[41]
N. Ballatori, T. W. Clarkson, Biliary secretion of glutathione and of glutathione-metal complexes. Toxicol. Sci. 1985, 5, 816. doi:10.1093/ TOXSCI/5.5.816
[42]
D. L. Rabenstein, A. A. Isab, A proton nuclear magnetic resonance study of the interaction of mercury with intact human erythrocytes. Biochim. Biophys. Acta, Mol. Cell Res. 1982, 721, 374.
[43]
D. L. Rabenstein, A. A. Isab, R. S. Reid, A proton nuclear magnetic resonance study of the binding of methylmercury in human erythrocytes. Biochim. Biophys. Acta, Mol. Cell Res. 1982, 720, 53.
[44]
J. Fu, R. E. Hoffmeyer, M. J. Pushie, S. P. Singh, I. J. Pickering, G. N. George, Towards a custom chelator for mercury: evaluation of coordination environments by molecular modeling. J. Biol. Inorg. Chem. 2011, 16, 15. doi:10.1007/S00775-010-0695-1
[45]
U. Skyllberg, P. R. Bloom, J. Qian, C. M. Lin, W. F. Bleam, Complexation of mercury(II) in soil organic matter: EXAFS evidence for linear two-coordination with reduced sulfur groups. Environ. Sci. Technol. 2006, 40, 4174. doi:10.1021/ES0600577
[46]
J. Watts, E. Howell, J. K. Merle, Theoretical studies of complexes between Hg(II) ions and L-cysteinate amino acids. Int. J. Quantum Chem. 2014, 114, 333. doi:10.1002/QUA.24565
[47]
G. N. George, R. C. Prince, J. Gailer, G. A. Buttigieg, M. B. Denton, H. H. Harris, I. J. Pickering,Mercury binding to the chelation therapy agents DMSA and DMPS and the rational design of custom chelators for mercury. Chem. Res. Toxicol. 2004, 17, 999. doi:10.1021/ TX049904E
[48]
V. Mah, F. Jalilehvand, Mercury (II) complex formation with glutathione in alkaline aqueous solution. J. Biol. Inorg. Chem. 2008, 13, 541. doi:10.1007/S00775-008-0342-2
[49]
A. Manceau, C. Lemouchi, M. Rovezzi, M. Lanson, P. Glatzel, K. L. Nagy, I. Gautier-Luneau, Y. Joly, M. Enescu, Structure, bonding, and stability of mercury complexes with thiolate and thioether ligands from high-resolution XANES spectroscopy and first-principles calculations. Inorg. Chem. 2015, 54, 24. doi:10.1021/ACS.INORG CHEM.5B01932
]
Supplementary material
S1. Calculation of stability constants for Hg(MAC)2 and HgIMAC using I- as competing ligand. Components, reactions and a calculation scheme for determination of the stability constant for the Hg(MAC)2 and HgIMAC complexes with I- as the competing ligand, using WinSGW software. The pKa of the RSH and COOH groups for MAC are 10.2 and 3.5, respectively. A similar set of corresponding equations were set-up when calculating constants for Hg(2-MPA)2
| [I[sup -]][sub total] = [I[sup -]][sub free] + [HgI[sup +]] + 2[HgI[sub 2]] + 3[HgI[sub 3-]] + 4[HgI[sub 4-]] + [HgIMAC] | (Sa) |
| [MAC][sub total] = [MAC][sub free] + 2[Hg(MAC)[sub 2]] + [HgIMAC] + [HgMAC] | (Sb) |
| [Hg[sup 2+]][sub total] = [Hg[sup 2+]][sub free] + [HgI[sup +]] + [HgI[sub 2]] + … + … + [HgIMAC] + [Hg(MAC)[sub 2]] +[HgMAC] + [Hg(OH)[sub n]] | (Sc) |
| Hg[sup 2+] + I[sup -] ⇌ HgI[sup +] | Log K[sub 1] = 13.42 | (S3a) |
| Hg[sup 2+] + 2I[sup -] ⇌ HgI[sub 2] | Log K[sub 2] = 24.60 | (S3b) |
| Hg[sup 2+] + 3I[sup -] ⇌ … | Log K[sub 3] =28.32 | (S3c) |
| Hg[sup II] + 4I[sup -] ⇌ … | Log K[sub 4] = 30.07 | (S3d) |
| Hg[sup 2+] + MAC ⇌ HgMAC | Log K = 32.5 | (S4a') |
| Hg[sup 2+] + 2MAC ⇌ Hg(MAC)[sub 2] | Log β[sub 2] = ? | (S1b') |
| Hg[sup 2+] + I[sup -] + MAC ⇌ HgIMAC | Log K = ? | (S6) |
Example of calculation
[I-]total = 20 µMa
[MAC]total = 16 µMa
[Hg2+]total = 4 µMa
[Hg(MAC)2] = 3.2 µMb
[HgIMAC] = (Peak Area (HgIMAC) / Peak Area (Hg(MAC)2)) x [Hg(MAC)2] = 0.45 µMb
[
] = [HgII]total - [Hg(MAC)2] - [HgIMAC]= 0.29 µMb
a Concentration of ligand used for complexation
b Concentration of complexes determined by LC-ICPMS.
In the experiment, the concentration of I- was varied, and in the calculation scheme above it is exemplified for 20 µM I-. Constants were recalculated to an ionic strength of 0 M and experimental pH was 2.9. Based on the equations (Sa-Sc); reaction S3(a-d), (S4a'), (S1b'), (S6), and total concentration of HgII and ligands, and measured concentration of Hg(MAC)2 and HgIMAC complexes, the log K of Hg(MAC)2 and HgIMAC was determined to be 40.9 and 32.2, respectively.
S2. Calculation of the stability constant for the Hg(Cys) 2 complex using MAC as competing ligand. Calculation scheme for determination of the stability constant for Hg(Cys)2 with MAC as the competing ligand using WinSGW software. The pKa of the RSH group for MAC and Cys are 10.2 and 8.6, respectively and pKa of the COOH group for MAC and Cys are 3.4 and 2.0, respectively. The same corresponding calculation scheme was used for the other Hg(SR)2 complexes using MAC or 2-MPA as the competing ligand
| [MAC][sub total] = [MAC][sub free] + 2[Hg(MAC)[sub 2]] + [HgMAC] | (Sd) |
| [Cys][sub total] = [Cys][sub free] + 2[Hg(Cys)[sub 2]] + [HgCys] | (Se) |
| [Hg[sup 2+]][sub total] = [Hg[sup 2+]][sub free] + [HgMAC] + [Hg(MAC)[sub 2]] + [HgCys] + [Hg(Cys)[sub 2]] +[Hg(OH)[sub n]] | (Sf) |
| Hg[sup 2+] + MAC ⇌ HgMAC | Log K = 32.5 | (S4') |
| Hg[sup 2+] + Cys ⇌ HgCys | Log K = 29.0 | (S4") |
| Hg[sup 2+] + 2MAC ⇌ Hg(MAC)[sub 2] | Log β[sub 2] = 40.9 | (S1b') |
| Hg[sup 2+] + 2Cys ⇌ Hg(Cys)[sub 2] | Log β[sub 2] = ? | (S1b") |
Example of calculation
[MAC]total = 18 µMa
[Cys]total = 8 µMa
[Hg2+]total = 4 µMa
[Hg(MAC)2] = 2.27 µMb
[Hg(Cys)2] = 1.72 µMa
a Concentration of ligand used for complexation
b Concentration of complexes determined by HPLC-ICPMS.
Constants were recalculated to an ionic strength of 0 M and experimental pH was 2.9. Based on the equations (Sd-Sf); reactions (S4'), (S4"), (S1b'), (S1b'') and total concentration of the HgII and ligands and measured complexes concentration, the log β2 of Hg(Cys)2 was determined to 37.5.
Table S1. Thermodynamic stability constants for
complexes at different ionic strengths
The stability constants were corrected to different ion strength using free software Ionic Strength Corrections for Stability Constants from IUPAC with Specific Interaction Theory (SIT) method[1]
| Reaction | Stability constant (log K[sub 1], log β[sub 2]-β[sub 4]) |
| I=0.5* | I=0 | I=0.1 | I=1 |
| Hg[sup 2+] + I[sup -] = HgI[sup +] | 12.87 | 13.42 | 13.02 | 12.90 |
| Hg[sup 2+] + 2I[sup -] = HgI[sub 2] | 23.82 | 24.60 | 24.00 | 23.92 |
| Hg[sup 2+] + 3I[sup -] = … | 27.6 | 28.32 | 27.74 | 27.74 |
| Hg[sup 2+] + 4I[sup -] = … | 29.83 | 30.07 | 29.73 | 30.15 |
*Reference Martell et al. (2004)[2]
Table S2. Computed complex formation free energies, bond lengths and angles of the Hg(SR)2 complexes determined at the B3LYP level of theory
All geometries were optimized at the gas phase with a mixed basis set containing Def2-TZVPP for Hg and 6-31++G(d,p) for all other atoms, respectively, and an initial geometry with all ligands fully extended
| Complexes | ΔG[sub 2]a (kcal/mol) | Bond (Å) | Angle (°) |
| Hg-Sb | Hg-Oc | S-Hg-S | Hg-S-Cb | O-Hg-Oc |
| Hg(2-MPA)[sub 2] | 53.1 | 2.381 | 3.042 | 179.7 | 101.4 | 158.4 |
| Hg(NACPen)[sub 2] | 56.4 | 2.364 | | 175.6 | 105.0 | |
| Hg(NACCys)[sub 2] | 51.1 | 2.383 | 3.087 | 175.1 | 99.3 | 65.4 |
| Hg(SUC)[sub 2] | 54.0 | 2.376 | 3.146 | 176.7 | 102.8 | 96.4 |
| Hg(3-MPA)[sub 2] | 58.0 | 2.361 | | 177.6 | 103.0 | |
| Hg(MAC)[sub 2] | 56.6 | 2.373 | 3.176 | 177.6 | 102.9 | 173.4 |
| Hg(Glyc)[sub 2] | 55.2 | 2.361 | | 177.7 | 103.0 | |
| Hg(GluCys)[sub 2] | 77.4 | 2.365 | | 177.0 | 103.5 | |
| Hg(ETH)[sub 2] | 56.9 | 2.360 | | 177.7 | 102.7 | |
| Hg(Pen)[sub 2] | 86.6 | 2.383 | 2.876 | 176.8 | 108.4 | |
| Hg(GSH)[sub 2] | 54.7 | 2.380 | 3.062 | 170.7 | 104.6 | 63.6 |
| Hg(Cys)[sub 2] | 92.1 | 2.373 | | 177.4 | 105.7 | |
| Hg(CysGly)[sub 2] | 91.7 | 2.375 | | 176.6 | 106.0 | |
| Hg(HCys)[sub 2] | 87.3 | 2.372 | | 176.7 | 106.1 | |
| Hg(Cyst)[sub 2] | 90.7 | 2.374 | | 175.6 | 106.1 | |
a The ΔG values are for the reaction 1a, i.e., Hg2+ + 2RSH = Hg(SR)2 + 2H+, in the gas phase at 298.15 K.
b The Hg-S distances and Hg-S-C angles are the averages over the two Hg-S distances and Hg-SC angles, respectively.
c In the geometry optimisations, several complexes form an additional coordination between Hg and the ligand's carbonyl or carboxyl oxygen (bidentate coordination). The Hg-O distances and OHg-O angles are the averages over the two Hg-O distances and O-Hg-O angles, respectively, except Hg(Pen)2, which only forms a single Hg-O interaction.
Fig. S1. The structure and abbreviation of investigated thiol ligands. The thiols are grouped according to the presence of functional groups in addition to the thiol group.
Fig. S2. The integrated peak area of Hg(SR)2 complexes to Thalium (204Tl+) signal ratios with different reaction time for the Hg(SR)2 complex synthesis. (a) mixture of 1 µM of HgII and 2 µM of each HCys and NACCys. (b) mixture of 1 µM of HgII and 2 µM of each Cys and NACPen. The experiments were conducted at pH 3.0, and an ion strength of 0 M. A post column flow of a 10 ng ml-1 Tl solution (flow rate of 100 µl min-1) was used to monitor and correct for signal drift of the LC-ICPMS system over time.
Fig. S3. Peak area of Hg(SR)2 complexes to Thalium (204Tl+) signal ratios at different storage time of Hg(SR)2 complex solutions. Samples contain 1 µM of HgII and 4 µM of individual LMM thiols i.e. Cys, HCys, GSH and Glyc. A post column flow of a 10 ng ml-1 Tl solution (flow rate of 100 µl min-1) was used to monitor and correct for signal drift of the LC-ICPMS system over time.
Fig. S4. LC-ICPMS chromatograms showing 202Hg+ and 127I+ signals (counts per second, cps) for (a) 4 µM of HgII and 50 µM of I-. A gradient elution was used with initially 3.5% of 1-propanol during 12 min, then increased to 25% in a step gradient and kept for 4 min and then back to the initial 3.5% concentration. (b) 4 µM of HgII and 50 µM of I-. A gradient elution was used with initially 8.5% 1-propanol during 14 min, then increased to 25% in a step gradient and kept for 4 min and then back to the initial 8.5% concentration. (c) 4 µM of HgII, 16 µM of MAC and 100 µM of I-. The elution gradient was the same as in (a). (d) 4 µM of HgII, 16 µM of 2MPA and 100 µM of I-. The elution gradient was the same as in (b). The complexes of Hg(MAC)2, HgIMAC, Hg(2MPA)2, and HgI(2-MPA) eluted at 600 s, 670 s, 780 s, and 1010 s, respestively. The HgI2 complex eluted at 790 s with the gradient of (a) and at 1070 s with the gradient of (b). The peaks 1, 2 correspond to the complexes of
(n=1, 3, 4) and the peak 3 to a system background signal.
Fig. S5. The mass spectra of 15 investigated Hg(SR)2 complexes achieved by direct infusion to ESIMS, showing molecular mass and Hg isotope pattern of the Hg(SR)2 complexes. All Hg(SR)2 complex analyses were conducted in the negative ionization mode with the exception of Hg(Cyst)2, which was analysed in positive mode. The concentration of HgII was 0.1 mM, the molar ratio of RSH to HgII was 4 and pH was 3.0.
Fig. S6. Measured concentration of the Hg(MAC)2 complex (2 µM of HgII and 8 µM of MAC) at different added concentrations of the competing ligands EDTA, Cl-, and Br-. The experiments were carried out at constant ionic strength of 0.5 M (NaClO4) and pH of 3.0.
Fig. S7. LC-ICPMS chromatograms of 202Hg+ signals Illustrating of retention time of the 15 investigated Hg(SR)2 complexes on the Kinetic Biphenyl LC column used in the LC-ICPMS measurements. The retention time increased as, 1. Hg(Cyst)2, 2. Hg(CysGly)2, 3. Hg(Cys)2, 4. Hg(HCys)2, 5. Hg(GSH)2, 6. Hg(GluCys)2, 7. Hg(Pen)2, 8. Hg(Glyc)2, 9. Hg(NACCys)2, 10. Hg(ETH)2, 11. Hg(MAC)2, 12. Hg(SUC)2, 13. Hg(3MPA)2. 14. Hg(NACPen)2, 15. Hg(2MPA)2.
Fig. S8. LC-ICPMS chromatogram of 202Hg+ signals for a mixture of 4 µM of HgII and 4 µM of MAC with isocratic elution using 3.5% 1-propanol indicating the presence of the one-coordinated HgMAC complex with a retention time of 700 s.
Fig. S9. LC-ESIMS chromatograms with selected ion monitoring (SIM) mode targeting m/z of Hg(MAC)2, Hg(Cys)2, and CysHgMAC and their corresponding Hg isotope pattern 378-385 m/z, 436-443 m/z and 407-414 m/z, respectively. An elution gradient was used with initially 8% of MeOH, 0.1%FA and 92% H2O, 0.1%FA from 3 to 10 min followed by a linear gradient to 90% of MeOH, 0.1%FA which was kept for 3 min. From 13 to 16 min the concentration of MeOH, 0.1%FA was reduced from 90% to 8% in a linear gradient and was kept for 9 min. The increased background after 6 min is caused by the increase of MeOH in the mobile phase.
Fig. S10. LC-ICPMS chromatograms showing 202Hg+ signals of (a) a mixture containing 4 µM of HgII and 8 µM of each Cys and MAC, (b) a mixture containing 4 µM of HgII and 16 µM of MAC, (c) a mixture containing 4 µM of HgII and 16 µM of Cys. The retention time of Hg(Cys)2 and Hg(MAC)2 is 85 s and 540 s, respectively with a mobile phase gradient of initially 3.5% of 1-propanol for 12 min then a step gradient increase to 25% of 1-propanol. The appearance of a small peak at 850 s is caused by the changed concentration of 1-propanol in the mobile phase.
Fig. S11. Full scan mass spectrum of a solution containing 0.1 mM of HgII and 0.2 mM of each Cys and MAC by direct infusion to ESI-MS with a flow rate of 50 µl min-1. Observed signals indicate the presence of Hg(MAC)2, Hg(Cys)2 and CysHgMAC complexes, matching their molecular mass and Hg isotope pattern.
[
References
IUPAC, Ionic Strength Corrections for Stability Constants using Specific Interaction Theory (SIT), version 1. 2004.
A. Martell, R. Smith, R. Motekaitis, National Institute of Standard and Technology, NIST. Critically selected stability constants of metal complexes 2004 (PC-based Database: Gaithersburg, MD).
]
Received 11 March 2017, accepted 27 May 2017, published online 23 June 2017
By Van Liem-Nguyen, Department of Chemistry, Umeå University, SE-901 87 Umeå, Sweden; Ulf Skyllberg, Department of Forest Ecology and Management, Swedish University of Agricultural Sciences, SE-901 83 Umeå , Sweden; Kwangho Nam, Department of Chemistry, Umeå University, SE-901 87 Umeå, Sweden and Erik Björn, Department of Chemistry, Umeå University, SE-901 87 Umeå, Sweden' Email: erik.bjorn@umu.se
