4, SULFIDE CHEMISTRY IN AQUEOUS SOLUTIONS 4.1. Solubility of H2S Gas Hydrogen sulfide gas, HySp, is quite soluble in water (Fig 5). In the laboratory, bubbling H)S, at 0.1 MPa into water results in a sulfide concentration of around 0.1 M and a pH of about 4. The reasons for this are related to a series of reactions which occur when H2S, is dissolved in water. An aqueous H2S species, H2Sqq, is initially formed which dissociates to form the bisulfide, HS" , and the sulfide, $°~. In geochemistry, we usually drop the subscript aq when referring to aqueous H2S unless we wish to distinguish this from H2S, The series of equilibria is then HySz = H2Suq @) H.S = HS" + H* ® HS =S* +H (5) ‘The net result of dissolving HyS in water is then the production of protons through reactions (4) and (5). The amount of gas dissolved will depend on the HS saluity mol % % 40 20 30 4 50 80 7 80 90 100 Temperature °C. FIGURES Solubility of HS, in water up to 100 “C and at a partial pressure of 0.1 MPa HS. “The solid line Is a best-fit polynoenial similar 1 Eqn (6) and thls is compared with data poinus fiom six experimenial reports, (See Carvoll and Masher. 1989 for detail. partial pressure or fugacity of HS, fH)S, the temperature, the nature of the solution and the three equilibrium constants, Ko, for Eqn (3), Ky for Eqn (4) and K> for Eq (5). Douabul and Riley (1979) measured HS solubility of over the temperature range of 2-30 °C and the salinity range of 0-40 and showed that increasing temperature and salinity decreased the solubility of HS. The solubility of HS in water and seawater can be represented by the equilibrium constant, Ky:sanoya fu par tcuv oun sony eye kaa ag + | 0.60583 + 0.379753(1/100) = 0.602140(7/100)"] C) where is the salinity and Tis in K. Equation (6) is valid' when the fugacity of HS, HS, is 1 aum or 0.1 MPa. In the low-temperature and moderate salinity range characteristic of marine systems, the equation by Douabul and Riley is quite robust. Carroll and Mather (1989) extended the data up to 90°C and Suleimenoy and Krupp (1994) provide a table of Henry's law constants up to 320 °C and an ionie strength of 3M. Il is difficult experimentally 10 obtain higher temperature and higher salinity data—such as might be applicable to deep ocean hydrothermal vent systems. The Hg$ solutions at temperatures >200 °C are extremely corzosive and experimental data collected at higher temperatures have shown large variations. In general, however, the solubility of H2S follows the normal pattern Chapter | 2 Sulfur Chemistry in Aqueous Solutions with rising temperature, revealing a minimum around 160 °C after which there is a continuous increase as the critical point of the solvent is approached (Fig. 6). However, the odd thing about H2S solubility is its change in salt solutions, so that in solutions with ionic strengths >1 M, the solubility of H2S decreases with increasing temperature. ‘The decrease in the solubility of H2S with the addition of salt is called the salting out and can be represented by the Setchenow equa In(Co/C) = In(y,) = KS o where Cp and C are the solubilities in water and ionic medium, 7y is the activity cocfficient of the gas. k is the salting coefficient and 5 is the salinity. The values for H2S are lower than other gases at low temperatures, but similar to other weak acids since the activity coefficient of HS in seawater is near unity, the interactions between the major sea salts and HyS are quite small, At higher temperatures, however, the values become much lower, probably due to H2S- water interactions, and the activity coefficient for H25 must deviate signifi- cantly from unitThe chemistry of (IN in aqueous environmental systems is well constrained (Morse eal, 1989). pRraue is lose (eg. Suleimenov and Seward, 1997) ‘which means that HS dominates the system a acid pH values and HS” i the dhominant species i alkaline solutions, 24 aol, iro £ = 35 = z m2 40 tao 48 1 L 0 59 100 150 200 250 300 360 mperature °C FIGURE 6 Solubility of HAS in terms of log Ky for Ean (5) in HLO at tmperatires between 1 and 360°C and att onic engi, J, betcen O and 31M from the enpesimental ta hy Swleimenov and Krupp (1994), ‘The value of Ky is related to observed concentrations and activity coeff- cients by Ky = rw (HY /tus: (HS")/tas(HS) = Kirw rus [ts @) where Kj = (H*)(HS~)/(H)S) is the stoichiometric constant and 7; is the activity coefficient for species. Miller (1983) used the data of Almgren e al. (1976) in NaCl solutions and activity coefficients calculated using Pitzer (1979) equations to estimate a value for pk, =698-+0.02 at 25°C. His general conclusion was that the activity coefficient for HS~ is similar to that of CI- ‘Suleimenoy and Seward (1997) reported the results of an experime! study of pKis up © 35°C, Their results are shown in Fig. 7 together With the results from previous experimental investigations, The experimental results are quite consistent up to 100°C but there is some vatiation at higher temperatures between their data and earlier measurements of Ellis and Giggenbach (1971). ‘Millero (1986) found that the equation pK} = pKir + AS'? 4 Bs co eseribed the relationship between pKy,9,5, salinity and temperature. In this ‘equation, the temperature dependent value for pKi.x,s, PK is deseribed by pir = 32.55 4 1519.44/T — 15.672 log? +0.027227 (10)10 ° ae & oe a ° 100 200 300 400 Temperature °C FIGURE7 pK ye ver wrapersuace fr al experoenial data at saute wer vapor presse (1976): A= -0.2391 435.685/T ay B = 0.0109 0.3776/T ay As can be seen fiom Fig. 7, Egn (10) is quite robust up t© 100°C. The equation is consistent with experimentally derived pKyi,s values up 0 seawater ionic strengths (Fig. 8) Ko is Tess well known but is o wgenbach, 1971 Schoonen and Barnes, 1988; Theberge et al., 1997). Figure 9 shows the reported measurements for this constant sinee 1946, As Morse ot al. (1987) noted, there was a measurement of 14.92 in 1906 but this is generally regarded as lucky. The uncertainty in che measurement of pK2,n,s is caused by problems ‘with polysultide contamination in the experimental measurement of pK 24,5 at high pH values. The reporied inerease in the value with time is related to improved experimental methods. In Fig. 9, Fnote that the regression of the values (with an R° of 0.88) is quite robust and itis likely that future workers will report even higher values. ‘Although it has been reported that the high value of pK>1,5 means that 's?~ has no significant concentration in aqueous solutions, this is not true. AS shown in Fig. 10, the activity of the $*~ ion in equilibrium with a total SCID) activity of 10-* varies across the range of pH values in most natural aqueous systems between approximately 10"! at pH = 10 and 10-7? at pH = thermodynamic terms, these are significant activities compared with the soluta 10 15% 2580 40 satnty FIGURE 8 p11 vers salinity for wo lfferent temperatires showing the fit of the Millero (1980) equation versus experimental data from Almaren ctl. (1976) and Goldaaber and Kaplan 1975). 4, Sulfide Chemistry in Aqueous Solutions GD 247 @ y = 0.1688x - 316,62 e- R? = 0.9835 1940 1960 1980 1970 1980 1999 2000 Year FIGURE 9 Measured values of piK,5 venus date (Data from Elis and Golding, 1959: Elis and Milestono, 1967; Giggenbach, 1071: Konopicand Leber, 1940; Kubli, 1046; Licht and Manasson, 1987; Maronny, 1959: Meyer est, 1983; Mahatma ind Sundaran, 1961; Schoonen and Birwes, 1988; Theberge ot al., 1997; Widmer and Schwarzenbach, 1964),temperature aqueous sulfide chemistry is not due to its Tow concentration but because of the uncertainty in the value of pK2.1,5. This is why the solubility ate the wulfides is usually written in terms of the HS~ species. | dest ity written in this way as Kyp2 Thus for FeS, the solubility is written as FeS +H! = Fe?! +HS «aay rather than in terms of the solubility, Kap, FeS = Fe? +S? «aay ‘The problem is that some compilations of stability constants still include older pK j1,5 values around 12 or 14, or include sulfide solubility constants which are based on these older values. These still slip readily into the literature FIGURE 10. Aqueous sulfide species ina sotution feation ofarouad 1 MG SCH)y— 10"). Note that yen with pRGCHSS) 2 18 cally meaningful cuvity AU the pH values of most natural waters. (From lokand end Lather 2007) out originally by Schoonen and Barnes (1988) 4.3. Molecular Structures of Sulfide Species in Aqueous Solutions HpS has a similar molecular suuucture to H>O (Fig. 11). It isa bent molecule an S-H-S bond angle of 92° which compares with 107.2° for water. ince sulfur is not as electronegative as oxygen, HS is nol as polar a molecule as HO. The HS bonuling electrons spend much of their time over the molecule as a whole rather than being centered on one atom, like oxygen. Qualitative moleculir orbital theory provides insights on how electron orbitals interact to conurol the outcome of reactions. For reactivity, the most imporant orbitals in molecules are the two frontier orbitals: the highest occupied. molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). The LUMO receives electrons donated by the HOMO (Fig. 12). ‘The frontier orbitals for the bent molecule H38 are welll known (Gimare, 1979), Figure 12 shows the molecular orbital energy level diagram for HoS. which results from the linear combination of the Ly orbitals of the wo hydrogen. ‘atoms with the 3s and 3p orbitals of the sulfur atom, It also compares the ener: level diagrams of HS~ with HS, ‘The relative energies of these orbitals suLUMO. In electron-transfer processes, the HOMO of the reductant overlaps the LUMO of the oxidant in order to initiate outer sphere electron transfer. The orbitals must have similar symmetries with respect to the bond axis so that they can overlap (Pearson, 1976). The reaction is symmetry-allowed if (1) the molecular orbitals are positioned for good overlap, (2) the energy of the LUMO is lower than, or less than 6 eV above, that of the HOMO and (3) the bonds thus ereated or broken are consistent with the expected end products of the reaction. The LUMO for HS~ is +8.015 cV (Rickard and Luther, 1997). This i a computed value and there are no experimental data available, so the accuracy of the result is unknown, However, the high positive energy indicates HS” cannot be an electron acceptor. The calculated HOMO for HS~ is 2.37 eV, whlch cavipaitarsvell with itie eenerinbarsl value de —24) eV (ele MMe 1984; Radzig and Smirnov, 1985). The HOMO for HS~ is less stable than that FIGURE 11 Molecular model of 1:5 compared & With HO. Note the smaller H-S-H angle andthe longer S-H bonds. (Keprinted with permission from RICKARD, D. & LUTHER (II), G. W. 2007. Chemistry of bron Sulphides. Chemical Reviews. Fig. 6. Copyright (2007) American Chemical Society) Hs HO HS" HS so1sev on" PER iia ane tb tt ota: 3ay 2). i ze SL soarev oe n(yy"5 by tt ow cs (3); Lay FIGURE 12 Molecular orbital cnersy level diagrams for HS~ and H,S. (ltickarid and Luther: 2007),is consistent with the known reactivity of HIS. For example. it forms the basis of the mechanisms of the formation of FeS in aqueous solution where the more nucleophilic HS~ attacks the Fe(ID) center of the [Fe(:0)6F* molecules whereas H2S initially forms an outer sphere complex (see Section 3.4 of Chapter 5). The calculated LUMO orbital for H2S (Fig. 13) is +0.509 eV (Rickard and Luther, 1997) whereas the experimental value is —1.1 eV based on electron affinity data (Radzig and Smirnov, 1985). These data show that HyS can be 3 electron acceptor or oxidizing agent. This is a counterintuitive result for many geochemists because sulfide-containing natural environments are casually referred 10 as “reducing.” However, this is 2 vague term chemically since everything is reducing with respect to something. Thus, the geochemical term “reducing environment” is often a portmanteau term with lie chemical mecning. In the case of HS, the computed and experimentally measured LUMO orbital (between — 1.1 and 0.5 eV) is of a similar magnitude to that for oxygen, the eponymous oxidizing agent, which is —0.47 eV. However, the calculated HOMO orbital for H2S is —9.646 eV which compares with the FIGURE 13. Three-dimensional representation of the HLS LUMO. The postive port of the orbital wave function is green and the neza- tive par ofthe orbital wave function is pink For interpretation of the references to color in this figure legend, the reader is refered to the online version of this book. (Re ted with peamission from D. & LUTHER (Ii, G. W. 2007. Cheanisiry of Irom Sulphides: Cheri cal Reviews. Fig 4. Copyright (2007) American Chemical Society). experimental value of —10.47 eV from ionization energy data (Drzaic et al, 1984; Redzig and Smirnov, 1985), Thus, Ha is not as good an electron donor as oxygen because the H38 HOMO js so stable. Even so, HyS can act as an electron donor or oxidizing agent to metals. This is because metal cations have LUMO oxbitals of similar energy or more stable energies compared to the HOMO of HS orhave an empty orbital due to water exchange. In the reaction of FeS with HoS to form pyrite, for example, HS is the oxidizing agent (Rickard, 1997; Rickard and Luther, 1997).