Microchim. Acta 146, 229–237 (2004) DOI 10.1007/s00604-004-0179-5

Original Paper Electrospray Ionization Mass Spectrometric Analysis of Aqueous Polysulfide Solutions Jenny Gun1 , Alexander D. Modestov1, Alexey Kamyshny Jr.1 , Dan Ryzkov1, Vitaly Gitis1 , Anatoly Goifman1 , Ovadia Lev1;
, Veit Hultsch2 , Thomas Grischek2 , and Eckhard Worch2 1 2

Division of Environmental Sciences, Fredy and Nadine Hermann School of Applied Science, The Hebrew University of Jerusalem, Jerusalem 91904, Israel Institute of Water Chemistry, Dresden University of Technology, 01062 Dresden, Germany

Received May 5, 2003; accepted December 15, 2003; published online May 25, 2004 # Springer-Verlag 2004

Abstract. The speciation of polysulfides in aqueous solutions was investigated by electrospray – ion trap and electrospray – time of flight mass spectrometry. The pH dependence of the observed total dissolved polysulfides’ concentration followed the trend calculated based on reported thermodynamic constants. However, the observed species’ distributions were substantially different from those calculated based on thermodynamic coefficients derived by UV spectroscopy. Notably, large abundances of heptasulfide, octasulfide and nonasulfide species were observed throughout the pH range 6 to 11. The large molecular weight anions had not been reported before in aqueous solutions although indirect evidence had suggested their existence. Key words: Polysulfides; hydropolysulfides; ESI-MS; collisioninduced dissociation; CID.

Polysulfides (Sn2 ) are reactive sulfur species that play an important role in batteries and photoelectrochemical cells. The impact of polysulfides and their products on the global sulfur cycle and on water and wastewater treatment have gradually unraveled in the last years [1–5].
Author for correspondence. E-mail: [email protected]

Polychalcogenides and particularly polysulfides undergo rapid comproportionation and disproportionation reactions in aqueous systems, thus precluding the elucidation of their direct speciation by chromatographic and other separation techniques [6]. Our knowledge of the speciation of these compounds in aqueous solution is derived exclusively from deconvolution of the complex UV-VIS spectra of high pH solutions of unknown distributions of polysulfide mixtures. The analysis at near neutral and moderately basic pH is further complicated by the formation of hydropolysulfides and polysulfanes. The shortcoming of the spectral deconvolution method is underscored by the fact that the same UV absorption technique led different researchers to qualitatively different conclusions regarding the distribution of polysulfides in aqueous systems. In fact, even the type and number of polysulfide species that prevail in aqueous solutions is still disputed. Some [7–12] claim that pentasulfide is the largest polysulfide species in aqueous solutions, while others [13, 14] believe that hexasulfide is also present in aqueous solutions and it is even the most dominant species under certain conditions. Steudel [15] recently suggested, based on indirect chemical reasoning, that higher polysulfides, with n>6 should also prevail in aqueous systems. As far as we know there is no experimental

230

evidence supporting this hypothesis. Since the UV spectra of aqueous polysulfides show at most two peaks in the UV range, the decision about the number of dominant polysulfide species in aqueous solutions and their distribution should be based on an alternative speciation technique. Electrochemical techniques, [16, 17] chromatographies, electrophoresis [18] and FTICR-MS [19] failed to resolve the speciation quest. Electrospray mass spectrometry (ESI-MS) is a powerful, soft ionization tool that is increasingly used for speciation of inorganic as well as organic compounds in aqueous and nonaqueous solutions [20, 21]. In fact, the current investigation was largely stimulated by the successful speciation of polyselenides, Sen2 , in aqueous solutions. Dorhout and co-workers [22] showed the prevalence of di-, tri-, tetra- and penta-selenides in aqueous solutions and unraveled the distribution of polyselenides as a function of pH and the concentration of the counter-cation. In this study we compare the ESI-MS spectra of polysulfide anions in a series of solutions prepared from sodium tetrasulfide salt, Na2S4. The qualitative distribution of polysulfides as a function of the pH was compared to the calculated distribution based on Boulegue’s [13] disproportionation constants and Schwarzenbach and Fischer [8] proton dissociation constants. ESI-MSTOF (Mariner ESI-TOF, Applied Biosystems) and ESIMS-ion trap (LCQ, ThermoQuest, Finnigan) spectra were compared revealing very similar species distributions for the same polysulfide compositions. Experimental Reagents NH4OH and glacial acetic acid were purchased from Frutarom (Haifa, Israel). Alkali metals (sodium and potassium) were obtained from BDH (Poole, UK). 98% pure Na2S  9H2O was purchased from Sigma (St. Louis, MO, USA). Elemental sulfur was purchased from Merck (Darmstadt, Germany). Pure (<18.2 M

cm), free of organics water was used in all tests. Analytical grade reagents were used unless otherwise stated.

J. Gun et al. Foster City, CA, USA) equipped with electrospray ionization (ESI) interfaces were used for data acquisition. The following experimental parameters were used for the LCQ experiments: the ESI was operated either in positive or in negative ion mode. In the negative ion mode we used 4.0 kV spray voltage and 20 V cone voltage. In the positive ion mode the spray voltage was 3.6 kV and the capillary voltage was set to 25 V. The sample injection flow rate was 5 mL min1 , the auxiliary liquid (MeOH) flow rate was 50 mL min1 . The source temperature was the subject of a separate optimization study that revealed optimal performance at 100  C. Mass spectra were acquired by scanning the mass analyzer from m=z 100 to 2,000 with 5 total microscans. Maximum injection time into the ion trap was 50 ms. The analyzer was operated at a background pressure of 2
105 Torr. In all experiments, helium was introduced at an estimated pressure of 1 m Torr to improve the ion trapping efficiency and as a collision gas during collision-activated decomposition events. During MSn experiments the compounds were isolated in the ion trap with the isotope pattern isolation width, and the collision energy was gradually increased to obtain the collision-dissociation profile of each hydropolysulfide. The following experimental parameters were used for the Mariner ESI-TOF experiments: the ESI was operated in negative ion mode. Other parameters optimized and used were: 4.0 kV spray voltage, 120 V nozzle voltage, 600 V quadrupole RF voltage, 2100 V detector votage, 100  C nozzle temperature, 100  C quadrupole temperature, the sample injection flow rate was 10 mL min1 . Curtain and nebulizer gas was pure nitrogen (pure gas generator model 75-72, Whatman, Kent, UK), the curtain gas flow rate was 1 L min1 and the nebulizer gas flow rate was 0.2 L min1 . The calibration of peak resolution and mass accuracy was performed prior to ESI-MS measurements. The calibration of peak resolution was performed in positive ion mode using a peptide standard solution (angiotensine, neurotensine and bradykinin). In all experiments the resolution factor was at least 5000. Exact mass calibration in negative ion mode was based on a standard solution of benzoic acid, 4-nitrophenol, and dinitro-o-cresol. Mass spectra were acquired by scanning the mass analyzer from m=z 50 to 1000 using a scan speed of 2 scans min1 . Using 600 V quadrupole RF voltage, the sensitivity of the TOF analyzer was reduced for ions with m=z below 100. The effective mass range was from 90 to 600 mass units. The analyzer was operated at a background pressure of 2.5
106 Torr. All samples were injected using a sample loop (1 mL PEEKtubing). Carrier solvent was 10 mL min1 methanol (gradient grade, Merck, Darmstadt, Germany). ESI-MS studies of K2S solutions: in order to exclude the possibility of substantial redox reactions under the specified conditions, we investigated the ESI-MS spectra of 15 mM K2S (pH 9.75) solutions under the ESI conditions of the polysulfide studies for LCQ. We did not observe any artificial polysulfide formation.

Polysulfide Synthesis

X-ray powder diffraction of inorganic polysulfides was measured with a Phillips automated powder diffractometer (Cu K
 radiation) at 1.2  C min1 scan rate. Elemental analysis was made in the Microanalysis Laboratory of the Hebrew University.

Na2S4 was synthesized by the method described previously [23, 24]. The polysulfide salts were characterized by X-ray diffraction and found to have the reported diffraction pattern [25]. The exact composition was found by elemental analysis, which showed that the synthesized compounds have alkali metal to sulfur ratios which are very close to the theoretical one, namely Na2S3.86.

Electrospray Mass Spectrometry

Preparation of Polysulfide Solutions

LCQ (Thermo-Quest, Finnigan, San Jose, CA, USA) quadrupole ion trap mass spectrometer and Mariner ESI-TOF (Applied Biosystems,

Polysulfide solutions were prepared by dissolving a desirable mass of polysulfides in the specified oxygen-free, aqueous 50 mM ammo-

Instrumentation

Electrospray Ionization Mass Spectrometric Analysis of Aqueous Polysulfide Solutions nium acetate solutions followed by pH adjustment with NH4OH or acetic acid. The solutions were thermostated for at least 12 hours to reach equilibrium and sulfur sedimentation.

Results and Discussions ESI Mass Spectra of Polysulfide Solutions ESI-MS studies of polysulfides were carried out under the optimized conditions of the electrospray as dictated by the tuning programs of the two instruments. The specific parameters for every instrument are given in the experimental section. Preliminary studies revealed that protonated polysulfanes were never observed in the positive ion mode, which could be anticipated in view of the high acidic character of polysulfanes. Therefore, the rest of the study was carried out in negative ion mode. We also carried out a preliminary study proving that bisulfide and hydrogen sulfide (introduced as K2S in aque-

231

ous solution and buffered to the different pH levels) did not form polysulfides in the ESI chamber. This part of the research was important in view of cumulative evidence showing that both oxidation and reduction processes can take place in the ESI chambers even in negative ion mode of operation [26]. The absence of polysulfides in K2S studies suggests that oxidation processes in the ESI are not significant in polysulfide speciation. The change of capillary voltage (for LCQ) did not significantly influence the spectra of ESMS in the range of 50 to þ80 V. For the TOF instrument, an influence of nozzle voltage on mass peak intensity was demonstrated in the range of 30 to 120 V. The distribution of polysulfide ion species was not influenced by nozzle voltage. In the following experiments the TOF was operated with a nozzle voltage of 30 V to minimize the dissociation of molecules in ESI. Figure 1 presents two typical ESI-MS spectra of 8 mM Na2S4 by ESI-TOF (A) and ESI-ion trap (B)

Fig. 1. Typical electrospray mass spectra of 8 mM Na2S4 in 50 mM NH4Ac pH 9.75. The solution was diluted by a 9 times larger methanol stream in the transfer line to the ESI inlet. (A) ESI-TOF MS, (B) ESI-MS ion trap. MS peak assignments are marked

232

J. Gun et al.

mass spectrometers. The studies were conducted using 50 mM NH4Ac buffer corrected to pH 9.8. Identification of the peaks of the ESI-TOF spectra was done by previous exact mass calibration of the device and further comparison of the experimental m=z values with those calculated using Mariner Data Explorer software version 4.0 (Applied Biosystems, Foster City, CA, USA). Identification of the peaks of the ESI-ion trap spectra was done by largest peak in the mass=charge region, and identification was confirmed by comparison with the simulated isotopic pattern [27]. Similar mass spectra were obtained when injecting from the solution with phosphate buffer instead of ammonium acetate as a supporting electrolyte. Surprisingly, ammonium ions did not affect the observed relative distribution of polysulfides. The peak of m=z ¼ 113 was assigned as HS2O3 . The peak appeared only at the high pH studies and gradually increased with time, which is explained by the OH induced disproportionation of polysulfide solutions [28]: Saq þ OH ! 1=4 S2 O3¼ þ 1=2 HS þ 1=4 H2 O or S5¼ þ 3 OH ! S2 O3¼ þ 3 HS We have assigned the other peaks as follows: 215 – NaS6 ; 247 – NaS7 ; 279 – NaS8 ; 311 – NaS9 . These assignments are supported by the increase of the corresponding peaks when the Naþ concentration was increased. As expected, sulfate was not detectable in the ESIMS studies. Doubly charged polysulfide species were never observed in our ESI-MS studies. This is understandable since it is now generally accepted that acid – base speciation cannot be distinguished by ESI-MS due to proton exchange in the ESI chamber. Likewise, thiosulfate species were observed as HS2O3 . We never observed hydrogen disulfide species, probably due to the more polar nature of these compounds. Enke et al. [29] have recently shown that ESI-MS exhibits much lower sensitivity for polar compounds due to their tendency to concentrate in the core of the spray droplets, making them less susceptible to ion evaporation. Extended H€uckel calculations [30] indeed showed that the charge is more evenly distributed on high molecular weight polysulfides. Thus, it is reasonable to expect that electrospray mass spectrometry will show preference for the high molecular weight polysulfides over the smaller ones.

Fig. 2. The percent distribution of polysulfides as a function of the heated capillary temperature of the LCQ-MS (normalized to the base peak of HS6 for every temperature). Legends: ^: HS4 ; &: HS5 @: HS6 , : HS7 ; &: HS8

Temperature Dependence We carried out the ESI-MS studies of 1 mM Na2S4, pH 9.75, at different heated capillary temperatures (100– 150  C) for the Mariner MS-TOF, and at 80–200  C for the LCQ – ion trap – MS). Figure 2 presents the percent distribution of polysulfides as a function of the heated capillary temperature of the LCQ-MS. In Fig. 2 the peak abundances of the different polysulfides is expressed relative to the base peak of HS6 for every temperature. High-n polysulfide (n ¼ 7–9) was observed over the whole temperature range, yet above 140  C there was a rapid increase in the concentration of the high polysulfide species. Since the relative distribution of the polysulfides was almost constant at 80–150  C and since their distribution was practically constant for T<140  C, we conducted the rest of the research using 100  C capillary. Prior research has also indicated that the electronic spectra of distribution of polysulfides remained virtually constant for T<150  C [31]. Flow Injection Analysis Figure 3 demonstrates a flow injection analysis of ESI MS-TOF of sodium tetrasulfide solutions in ammonium acetate buffer, pH 10.6. Specific ion – time traces corresponding to polysulfides with n from 4 to 8 are shown for a series of injections of different sodium tetrasulfide concentrations. The calibration curves, showing peak areas versus concentration, were linear (R2 ranged between 0.97 and 0.99) for all the species, and the intercepts ran through the initials for all species. The linear range was from 0 to 30 mM, and the minimum detection limit determined at 5 times the noise level of HS4 was 0.10 mM of (injected) sodium tetrasulfide solutions.

Electrospray Ionization Mass Spectrometric Analysis of Aqueous Polysulfide Solutions

233

Fig. 3. (A) Time trace of selected polysulfide anions recorded during flow injection analysis with ESI-TOF MS (loop 1 mL, flow of carrier liquid (MeOH) 10 mL min1 ). The labels C1, C2, C3 and C4 mark the instant of injection of 1.0, 3.0, 8.9 and 26.8 mM Na2S4 solutions containing 50 mM NH4Ac buffer, pH 10.6. (B) Calibration curves depicting the abundance of each of the polysulfide peaks as a function of the concentration of sodium tetrasulfide used for the preparation of the solution ^: HS4 (line formulae, y ¼ 1.26x; correlation coefficient, R2 ¼ 0.978); &: HS5 , (y ¼ 2.25x; R2 ¼ 0.982) @: HS6 , (y ¼ 5.64x; R2 ¼ 0.993); : HS7 , (y ¼ 1.41x; R2 ¼ 0.973); &: HS8 , (y ¼ 1.39x; R2 ¼ 0.986)

The linear abundance – concentration curves exhibited by all the dominant polysulfide species confirmed that polysulfides are not formed by bimolecular interactions in the ESI chamber, since bimolecular interactions should exhibit pronounced nonlinearity. pH Dependence and Distribution of Polysulfides The pH dependence of the total concentration of dissolved polysulfides and their speciation was studied using a series of sodium tetrasulfide solutions that were prepared from the same ammonium acetate – buffered stock solution (8 mM Na2S4 and 50 mM NH4Ac). The pH of the different solutions was adjusted to the set levels by addition of about 25% ammonium hydroxide solution. Figure 4A depicts the relative abundances of the polysulfide species at the specified pH. The Figure

demonstrates that the solubility limit is met at ca. pH 9.3 for 8 mM sodium tetrasulfide. Below pH 9.3 the dissolved concentration of all polysulfide species decreases monotonically reaching near-background abundances at pH 7. The distribution of the different polysulfide species is even more pronounced in Fig. 5A which depicts the molar % distribution of each species, normalized by the sum of the abundances of all the different polysulfide species, assuming – as a very rough approximation – identical sensitivity for all polysulfides. S62 is by far the most dominant species throughout the pH range, although S52 , S72 and S82 are also abundant. Surprisingly, the distribution of the polysulfides was not very dependent on the pH. Speciation of Polysulfides For comparison we have calculated the speciation of polysulfides for the same initial concentration of

234

J. Gun et al.

disproportionation (3) reactions. H2 Sn . HSn  þ Hþ ;

Ka;1;n

ð1Þ

HSn  . Sn 2 þ Hþ ;

Ka;2;n

ð2Þ

Kðnþ1Þ=n

ð3Þ

H2 Sn . H2 Sn1 þ Saq ;

Fig. 4. Observed and calculated polysulfide speciation as a function of pH. (Composition: 8 mM Na2S4 solutions, prepared in NH4Ac buffers of different pH) (A) Experimental abundances obtained by flow injection analysis with ESI-TOF MS. (normalized by the sum of the abundances of all the different polysulfide species at pH 10.75). Legends: ^: HS4 ; &: HS5 @: HS6 , : HS7 ; &: HS8 . The full-circle symbols represent the sum of the concentrations of the HS2 –HS6 species. (B) Calculated dissolved concentrations of polysulfides normalized by the total molar concentration of polysulfides at the specified pH. Legends: : S22 overlaps with ^: S32 , ^: S42 , &: S52 , @: S62 . The full-circle symbols represent the sum of the concentrations of the S22 –S62 species

We used the set of constants derived by Boulegue [13] taking into account the presence of hexasulfide species. Schwarzenbach’s proton dissociation constants were used [8]. Schwarzenbach assumed prevalence of Sn2 with n ¼ 2–5 only. The dissociation constant of H2S6 was taken from Meyer and co-workers [30]. The relevant constants are depicted in Table 1. This set of 18 equations was solved subject to electroneutrality and sulfur preservation conditions – Eq. (4). The electroneutrality condition implies that CT;Sn ¼ð½Sn 2  þ ½HSn   þ ½H2 Sn Þ for n ¼ 1 . . . 6

ð4Þ

CT,Sn is also the molar concentration of the dissolved tetrasulfide salt precursor. A simple way to understand Eq. (4) is by describing Sn¼ as comprised of one divalent and (n  1) zerovalent atoms [Sn  1(0)Sð2Þ ]¼ . Then Eq. (4) is just a conservation of the sum of the divalent species. The concentration of the polysulfide salt that is introduced into the reaction vial is different from the sum of the dissolved polysulfide concentrations. Sulfur mass balance implies that the sum of the zero valent sulfur atoms is constant, taking into account that the divalent sulfur is preserved by Eq. (4). ðn  1Þ  CT;Sn ¼ ðn  1Þð½Sn 2  þ ½HSn   þ ½H2 Sn Þ þ ½Saq  ð5Þ When elemental sulfur concentration (marked as S(aq) although S8 (aq) is the dominant sulfur form) exceeds sulfur solubility, then condition (5) is violated, and it should be replaced by a constant activity condition.

Fig. 5. Observed and calculated speciation of polysulfides (composition: 8 mM Na2S4 solutions, prepared in NH4Ac buffers of different pH). A) Observed relative abundance of the different polysulfides obtained by flow injection analysis with ESI-TOF MS. (normalized by the sum of the abundances of all the different polysulfide species at the specified pH). Legends: ^: HS4 , &: HS5 , @: HS6 , : HS7 , &: HS8 . B) Calculated dissolved concentrations of polysulfides in the same solution. Legends: : S22 , ^: S32 , ^: S42 , &: S52 , @: S62

tetrasulfide (8 mM). In order to calculate the actual distribution of dissolved polysulfides, we used the following set of diprotic acid dissociation (1 & 2) and

fSs g . fSaq g;

KS

ð6Þ

In Eq. (6), {Ss} ¼ 1, and the activity of the dissolved species is taken as their concentrations. Figures 4B and 5B depict the calculated speciation of the polysulfides in absolute concentration units and as a molar fraction of the total polysulfides, respectively. A comparison of the calculated polysulfide distribution and the observed MS abundances reveals a similarity in the general solubility trend. Additionally, the observed abundances of the disulfide and trisulfide

235

Electrospray Ionization Mass Spectrometric Analysis of Aqueous Polysulfide Solutions Table 1. Thermodynamic constants Notation Sulfur solubility Proton dissociation

pKS

Hydrogen sulfide Hydrogendisulfide Hydrogentrisulfide Hydrogentetrasulfide Hydrogenpentasulfide Hydrogenhexasulfide

pKa,1,1; pKa,1,2; pKa,1,3; pKa,1,4; pKa,1,5; pKa,1,6;

Value (M) 6.81

pKa,2,1 pKa,2,2 pKa,2,3 pKa,2,4 pKa,2,5 pKa,2,6

7; 17.2 5; 9.7 4.2; 7.5 3.8; 6.3 3.5; 5.7 3.2; 5.2

Calculated from reference no. 32 8, 9, 10 8 8 8 8 30

Disproportionation Hydrogendisulfide=Hydrogensulfide Hydrogentrisulfide=Hydrogendisulfide Hydrogentetrasulfide=Hydrogentrisulfide Hydrogenpentasulfide=Hydrogentetrasulfide Hydrogenhexasulfide=Hydrogenpentasulfide

pK2=1 pK3=2 pK4=3 pK5=4 pK6=5

species are very small, as predicted by the numerical calculation. However, the predicted distributions of polysulfides (Fig. 5B) are different from the observed signals (Fig. 5A): 1) The calculated distribution plays down the significance of the large polysulfide species. This is only partly explained by the higher hydropho-

1.84 4.00 8.20 6.03 5.84

13 13 13 13 13

bicity of the large-n polysulfide, since the differences in hydrophobicity of compounds containing 6, 7, and 8 catenated sulfur atoms are not very large. 2) The calculated increase of the high Sn2 compounds and decrease of the low-n ions at high pH is not observed in our mass spectrometry tests.

Fig. 6. CID studies of polysulfides. (A) Dissociation profiles of the polysulfide ions as a function of the collision energy in the ESI-MS ion trap. (B) Comparison of the observed abundances of the hydropolysulfide peaks and their observed E1=2 values in CID tests. The dashed line corresponds to CID tests; the solid line corresponds to observed abundances

236

It should be noted that while the actual disproportionation constants of Boulegue [13] can – and probably should – be disputed, the general tendencies of the concentration – pH curves are of thermodynamic nature and should not be influenced by the numerical values of the disproportionation constants used in the calculation – at least for the high pH, undersaturated conditions. Collision-Induced Dissociation Studies Figure 6 depicts collision-induced dissociation (CID) studies of the hydropolysulfide ions in the ion trap mass spectrometer. Figure 6(A) shows the relative abundance of the starting ion as a function of the applied collision energy during the collision-activated decomposition events. The x-axis presents the percentage of the maximum tickling voltage. The figure reflects the relative gas phase stability of the different anions. All dissociation profiles reveal S-shaped curves. The half wave collision energy (E1=2) corresponds to the collision energy at which the relative abundance in the fraction of the total ion current of the starting ion is 0.5 [33, 34]. The observed stability ranking is HS6 >HS7 >HS8 >HS5 >HS4 . Figure 6B depicts the E1=2 of the initial hydropolysulfides (Fig. 6A) and the relative abundance of the different polysulfide species in the same frame. The similarity between the two curves suggests that the observed distribution of polysulfides is affected more by the gas phase stability of the hydrosulfides than by the actual distribution of polysulfides in the solution. This led us to believe that the observed distribution of the polysulfides was largely affected by their gas phase stability and their ion evaporation susceptibility. Conclusions Both ESI-TOF MS and ESI-MS ion trap can be used for quantification of polysulfides in aqueous solution. The two instruments gave a very similar distribution of polysulfide species. The minimum detection limit of S42 in both cases was about 10.0 mg mL1 as S. The article underscores the higher resolution power of the TOF mass spectrometer and the MS2 capability of the ion trap MS. Our studies provides first-time evidence for the prevalence of heptasulfide and the existence of even higher polysulfides (n ¼ 8, 9) in aqueous solutions, although accurate quantification must rely on reference

J. Gun et al.

materials for these compounds. Reference materials are unavailable and we doubt that they will become available in the near future due to the rapid disproportionation of polysulfides in aqueous solutions. Unfortunately – unlike polyselenides [22] – ESI-MS could not quantitatively resolve the polysulfide speciation quest, since the gas phase ion stability and the ion evaporation susceptibility biased the observed signals, favoring larger polysulfides over smaller ones. Acknowledgments. We thank the water technology program of the BMBF, Germany, and the MOS, Israel, and Mekorot Ltd, Israel, for supporting this research.

References [1] Gun J, Goifman A, Shkrob I, Kamyshny A, Ginzburg B, Hadas O, Dor I, Modestov A D, Lev O (2000) Envir Sci Technol 34: 4741–4746 [2] Ginzburg B, Chalifa I, Zohary T, Hadas O, Dor I, Lev O (1998) Water Res 32: 1789–1800 [3] Ginzburg B, Dor I, Chalifa I, Hadas O, Lev O (1999) Envir Sci Technol 33: 571–579 [4] Lippa K A, Roberts A L (2002) Envir Sci Technol 36: 2008–2018 [5] Chen K Y, Morris J C (1972) Envir Sci Technol 6: 529–537 [6] Steudel R (1996) Ind Eng Chem Res 35: 1417–1423 [7] Teder A (1971) Acta Chem Scand 25: 1722–1728 [8] Schwarzenbach G, Fischer F (1960) Helv Chim Acta 43: 1365–1390 [9] Giggenbach W (1972) Inorg Chem 11: 1201–1207 [10] Licht S (1986) J Phys Chem 90: 1096–1099 [11] Janz G J, Downey J R Jr, Roduner E, Wasilczyk G J, Coutts J W, Eluard E (1976) Inorg Chem 15: 1759–1763 [12] Maronny G (1959) Electrochim Acta 1: 58–69 [13] Boulegue J, Michard G (1978) J Franc° ais d’Hydrologie 9: 27–33 (in French) [14] Williamson M A, Rimstidt J D (1992) Geochim Cosmochim Acta 56: 3867–3880 [15] Steudel R (2000) Environmental technologies to treat sulfur pollution. In: Lens P N L, Hulshoff Pol L (eds) IWA, London, p 1–31 [16] Kariuki S, Morra M J, Umiker K J, Cheng I F (2001) Anal Chim Acta 442: 277–285 [17] Rozan T F, Theberge S M, Luther G (2000) Anal Chim Acta 415: 175–184 [18] O’Reilly J W, Dicinoski G W, Shaw M J, Haddad P R (2001) Anal Chim Acta 432: 165–192 [19] Clarke E T, Solouki T, Russell D H, Martell A E, McManus D (1994) Anal Chim Acta 299: 97–111 [20] Stewart II (1999) Spectrochim Acta B 54: 1649–1695 [21] Henderson W, Nicholson B K, McCaffrey L J (1998) Polyhedron 17: 4291–4313 [22] Raymond C C, Dick D L, Dorhout P K (1997) Inorg Chem 36: 2678–2681 [23] Rule A, Thomas J S (1914) J Chem Soc (London) 105: 177–181 [24] Pearson T G, Robinson P L (1930) J Chem Soc (London) 115: 1473–1478 [25] International Centre for Diffraction Data, USA, 1992, Powder diffraction file, Alphabetical indexes, Inorganic phases, Sets 1–42

Electrospray Ionization Mass Spectrometric Analysis of Aqueous Polysulfide Solutions [26] De la Mora J F, Van Berkel G J, Enke C G, Cole R B, MartinezSanchez M, Fenn J B (2000) J Mass Spectrom 35: 939–952 [27] ‘‘Sheffeld Chemputer’’, http:==www.shef.ac.uk=chemputer= sheffild [28] Licht S, Davis J (1997) J Phys Chem B 101: 2540–2545 [29] Cech N B, Krone J R, Enke C G (2001) Anal Chem 73: 208–213

237

[30] Meyer B, Peter L, Spitzer K (1977) Inorg Chem 10: 27–32 [31] Giggenbach W F (1973) Inorg Chem 13: 1724–1733 [32] De Turi V F, Ervin K M (1999) J Phys Chem A 103: 6911–6920 [33] Talaty E R, Perera B A, Gallrdo A L, Barr J M, Van Stipdonk J M (2001) J Phys Chem A 105: 8059–8068 [34] Boulegue J (1978) Phosphorous, Sulphur Related Elements 5: 127–128