Oxidation kinetics of Iron(II) complexes of trans-1,2-diaminocyclohexanetetraacetate (cdta) with dissolved oxygen: Reaction mechanism, parameters of activation and kinetic salt effects

https://doi.org/10.1016/j.ces.2005.12.015Get rights and content

Abstract

The present investigation takes concern about a spiny environmental problem afflicting the pulp mill industry exploiting the Kraft sulfate-pulp process where dilute total reduced sulfur contaminants are co-mixed with oxygen in large-volume gas effluents. A potential Redox process for removing the total reduced sulfurs consists in oxidizing them by means of iron(III) organometallic complexes while the co-mixed oxygen mediates the oxidative regeneration of iron(II) into iron(III) complexes. In this work, the oxidation kinetics of iron(II) trans-1,2,-diaminocyclohexanetetraacetate (cdta) complexes with molecular oxygen (O2) as the source oxidant was investigated for a wide pH range (1.75<pH<12) in a 3.2 dm3 single-phase stirred cell reactor within the [281–323 K] temperature range. Simultaneous measurements of iron(II)–cdta (50–400μmoldm-3) and O2 (0.5–6mgdm-3) were used to clarify the reaction mechanism which has been interpreted differently in previous works. The observed kinetic data in alkaline solutions could be accounted for in terms of three forward [Fe2+cdta4-+O2 (rate-limiting, k1,app), Fe2+cdta4-+O2·-(k2,app), 2Fe2+cdta4-+H2O2] and one reverse [Fe3+(OH-)ncdta4-+O2·- (k-1,app,n=0 or 1)] elementary steps. Assessment of the rate-limiting apparent rate constant led to the following results (k1,app=38.0±2.4dm3mol-1s-1 at T=297.2K and Ic=0.05molNaCldm-3, ΔH10=92.6±0.6kJmol-1, ΔS10=38.6Jmol-1K-1). Fe3+OH-cdta4-, being the dominating iron(III) product at pH>10, was found to be less reactive than Fe3+cdta4- with the superoxide intermediate (O2·-), thus reducing the effect of the reverse step at higher pH. A study on the effect of electrolytes on the reaction rate led to the conclusion that salts increase the rate constant k1,app. Finally, kinetic results in acidic conditions leading to the formation of other iron(II)–cdta complexes (i.e., Fe2+cdta4-H+) and another superoxide intermediates (HO2·) are reported and discussed.

Introduction

The oxidation of iron(II) polyaminocarboxylate complexes (FeIILn-) with molecular oxygen (O2) and its derivatives (i.e., superoxides (Hn+O2·-), hydrogen peroxide (H2O2), hydroxyl radical (OH·)) has been studied for some time now in areas like biochemistry (i.e., superoxide dismutation, H2O2 decomposition—Lati and Meyerstein, 1978, Cohen et al., 1981, Bull et al., 1983, Rush and Koppenol, 1988, Rahhal and Richter, 1988, Seibig and van Eldik, 1997, Seibig and van Eldik, 1999), environmental science (i.e., catalytic removal of SOX and NOX, gas desulphurization—Brown and Mazzarella, 1987, Sada et al., 1987, Zang and van Eldik, 1990, Wubs and Beenackers, 1993, Demmink and Beenackers, 1997, Sharma et al., 2004) and water science (Liang et al., 1993). Recently, gas desulphurization processes based on ferric chelate chemistry have received increasing attention in the area of hydrocarbon processing (i.e., LO-CAT process: U.S. Filter Engineered Systems, Hardison and Ramshaw, 1992; Sulfint process, Mackinger et al., 1982). Application of such methods for hydrogen sulfide removal in the pulp and paper industry would differ on the flue gas composition because of the large amounts of oxygen associated with the atmospheric emissions. In line with this latter category of atmospheric emissions, Iliuta and Larachi (2003) recently established a modelling framework characterizing the potential of a bifunctional redox process where an iron(III) chelate (FeIIILn- where L denotes an organic ligand of n-charge) is used to throttle hydrogen sulfide emissions (Eq. (1)) whilst dissolved oxygen regenerates simultaneously the iron(II) chelate product into the active iron(III) form (Eq. (2)). Emphasizing on the regeneration step of such process, this work will strictly focus on the kinetics and mechanism of iron(II) chelate oxidation with dissolved oxygen as the source oxidant.H2S(aq)+2FeIIILn-+2OH-S0+2FeIILn-+2H2OO2(aq)+4FeIILn-+2H2O4FeIIILn-+4OH-

Experimental efforts in the literature regarding the FeIILn-+O2 reaction kinetics and procedures are largely summarized in Wubs and Beenackers (1993) work. Sada et al. (1987) proposed a reaction scheme based on experiments conducted in a bubble column using iron(II) nitrilotriacetate (nta) and iron(II) ethylenediaminetetraacetate (edta) complexes. Kinetic measurements were acquired based on O2 gas/liquid mass transfer. They established the reaction rate to be first-order with respect to O2 while ranging between 0 and 1 for FeIILn-. However, the proposed rate expression becomes irrelevant for general situations since the O2 concentration was regarded to be negligible in contrast to the FeIILn- concentration. In addition, this assumption was not well established and could have led to misinterpretation considering that other studies on the subject report a FeIILn- reaction order between 1 and 2. Brown and Mazzarella (1987) developed a reaction scheme involving 4 forward and 1 reverse steps (Eqs. (3)–(6)).This was obtained from experiments conducted in O2-saturated solutions of iron(II) aminopolycarboxylate complexes, in which a rotating pyrolitic graphite electrode was immersed to monitor the FeIILn- oxidation rate.O2+FeIILn-FeIIILn-+O2·-O2·-+FeIILn-+2H+FeIIILn-+H2O2H2O2+FeIILn-FeIIILn-+OH-+OHOH·+FeIILn-FeIIILn-+OH-

The final steps (Eqs. (5)–(6)) were considered very fast due to stronger oxidants and so the rate law turned out to be first-order with respect to O2 while evolving from first to second order with respect to FeIILn-. The latter partial order would indeed depend on the FeIIILn- concentration and the relative strength of the reverse step (O2·-+FeIIILn-) with the subsequent forward step (O2·-+FeIILn-, Eq. (4)). Later on, Zang and van Eldik (1990) proposed a slightly different reaction mechanism hypothesizing that FeIILn- and O2 binds together to form a non-oxidized intermediate (FeIILn-O2) which is consumed via three parallel steps (Eqs. (7)–(9)). Superoxides (O2·- or HO2·) and H2O2 would then oxidize FeIILn-molecules following reaction steps similar to those depicted by Eqs. (4)–(6).FeIILn-(O2)FeIIILn-+O2·-FeIILn-(O2)+H+FeIIILn-+HO2·FeIILn-(O2)+FeIILn-(FeIIILn-)2O22-2H+2FeIIILn-+H2O2

Although molecular oxygen is a weak oxidant, it seems unusual that an intermediate like FeIILn-O2 would accumulate to such level that Reaction (9) could take place significantly. Unless large amounts of FeIILn- are used, FeIILn-O2 should be considered an activated complex immediately forming FeIIILn-O2·- or FeIIILn-+O2·-. Still, Wubs and Beenackers (1993) came with the same reaction scheme based on experiments conducted in a flat-interface stirred cell reactor containing iron(II)–edta or iron(II) (hydroxyethyl)ethylenediaminetetraacetate (hedta) solutions. Demmink and Beenackers (1997) studied the reaction with iron(II)–nta solutions. Kinetic measurements were obtained by a manometric method monitoring the O2 head space pressure in the stirred cell. They noted a slight but consistent increase in O2 consumption from expected stoichiometry. Accordingly, they modelled an iron chelate degradation step into their overall scheme stipulating that a fraction of the hydrogen peroxide produced during the reaction would attack functional groups (i.e., ethylene moiety of edta, carboxylate groups) splitting the iron from the organic matrix. This phenomenon was previously ascribed to the formation of hydroxyl free radicals (Chen et al., 1995).

More recently, Seibig and van Eldik, 1997, Seibig and van Eldik, 1999 revisited the reaction with help of stopped-flow units connected to a diode array spectrophotometer. Their study has put a lot of importance on iron(II) chelate complex chemistry making clear that different FeIILn- species will result in different reactivity with O2. For example, it was alleged that iron(II)–edta having one protonated acetate arm (Ln-edta4-H+) self-arranges into a pentagonal bipyramidal structure, which is believed to bind O2 much faster than mono-capped trigonal-prismatic species such as nonprotonated (Ln-edta4-) and diprotonated (Ln-edta4-H2+) iron(II)–edta. Accordingly, equilibrium reactions illustrating the pH-dependence of different FeIILn- complexes were added to the reaction scheme without significantly modifying the elementary steps from the ones proposed by Zang and van Eldik (1990).

Overall assessment of previous studies leads to the following conclusions with respect to pH. The FeIILn-+O2 reaction rate is normally pH-independent in the alkaline-to-neutral region (DeBerry, 1997). A reaction enhancement is observed at lower pH and a maximum is reached in the vicinity of pH=3 depending on the ligand in use (Sada et al., 1987, Zang and van Eldik, 1990, Seibig and van Eldik, 1997, Seibig and van Eldik, 1999). This is likely ascribable to the formation of more reactive complexes as mentioned by Seibig and van Eldik, 1997, Seibig and van Eldik, 1999. The oxidation rate then falls drastically at pH lower than 2 where FeIILn-dechelation is triggered. The exact pH where the iron(II) aminopolycarboxylate complex will start to split into its building species (i.e., Fe2+ and Ln-) is characterized by the formation constant (Smith et al., 1985a, Smith et al., 1985b; Smith and Martell, 1987) along with the ligand deprotonation constants. The lack of reactivity at very low pH supports the fact that hydrated iron(II) ions weakly react with O2.

Up till now, all known experimental setups (i.e., bubble column, stirred cell) were used to investigate the FeIILn- oxidation rate using FeIIILn- spectrophotometrical, FeIILn-/FeIIILn- electrochemical or O2 pressure methods. In no case was the dissolved oxygen concentration monitored during the reaction apart from being known at the reaction start or maintained at saturation using O2 vigorous sparging. In this work, an experimental setup was developed to quantify FeIILn-, FeIIILn- and O2simultaneously in a completely filled stirred cell reactor using trans-1,2-diaminocyclohexanetetraacetatic acid (cdta) as the ligand. Concentration measurements from a UV–Vis spectrophotometer (iron(II)–cdta, iron(III)–cdta) and a polarographic electrode (O2) were used to clarify the reaction mechanism and determine thermodynamic properties (parameters of activation, kinetic salt effects) with help of elements discussed in previous studies. The experimental settings involve broad intervals of temperatures (281<T<323K), ionic strengths (0<Ic<0.8moldm-3) and pH (1.75<pH<12). From an engineering perspective, this work focuses on the importance of knowing the effect of every possible aqueous setting on the oxidation rate in order to prepare accurate design procedures for scrubbing processes exploiting the iron chelate chemistry in the mitigation of the odors associated with the emissions of total reduced sulphurs in the pulp and paper atmospheric air emissions.

Section snippets

Experimental section

Kinetic experiments were conducted in a hybrid aluminum/Plexiglas stirred cell of 12.7 cm i.d. and 25.4 cm height (Fig. 1) as described in a previous work (Piché and Larachi, 2005) allowing some modifications: (1) two extra six-blade turbine stirrers (6.35 cm i.d.) for a total of four were added in the upper section to maximize liquid mixing in the 3200±10cm3 stirred cell reactor; (2) a Clark-type polarographic electrode (Omega DOB-930) was placed in the cell with its tip positioned at about 7 cm

O2, iron(II)–cdta and iron(III)–cdta quantification

Dissolved oxygen saturation depends foremost of the surrounding pressure, temperature and ionic composition. The operating pressure was never recorded but was assumed to be near atmospheric. A thermocouple ring attached onto the polarographic O2 electrode allowed automatic temperature correction. In contrast, no conductivity correction was directly imposed from the meter. Instead, salting-out constants (kS; kNaCl=0.131dm3mol-1, kLiCl=0.094dm3mol-1, kNa2SO4=0.334dm3mol-1—Lang and Zander, 1986)

Kinetics in alkaline solutions

Forty-three kinetic trials were used as basis for developing a consistent reaction mechanism for operation in alkaline solutions (pH=8–12). Temperature (T) and ionic strength (Ic) were maintained at 297.2±0.5K and 0.05±0.002mol of NaCl per dm3, respectively. In these conditions, only one iron(II)–cdta complex (Fe2+cdta4-) forms (Seibig and van Eldik, 1999) while the Fe3+cdta4- and Fe3+OH-cdta4- products co-exist in equilibrium. Initial Fe2+cdta4-(CFe2+0) and O2(CO20) concentrations were kept in

Conclusion

Oxidation of iron(II) trans-1,2,-diaminocyclohexanetetraacetate (cdta) complexes with molecular oxygen as the source oxidant was studied in alkaline (mechanism and kinetics, activation parameters, kinetics salt effects) and acidic (kinetics) media. Fe2+cdta4- and O2·- are the only iron(II)–cdta and superoxide species formed at pH larger than 8. So it led to the elaboration of a reaction rate model (Eq. (22)) coupled with two apparent rate constant functions (Eqs. (34)–(35)) characterized by

Notation

CFeiron concentration, moll-1
CO2oxygen solubility, moll-1
CSsalt concentration, moll-1
Icionic strength, molm-3
k1,applimiting step apparent rate constant in Eq. (22), dm3mol-1s-1
kSsalting out constant, dm3mol-1
K1equilibrium constant of activated complex, Eq. (16)
ncomplex coordination
rrate constant, moll-1s-1
Sstoichiometric ratio
ttime, s
Tabsolute temperature, K
Greek letter
γactivity coefficient
Sub/superscripts
0initial
#1first activated complex

Acknowledgement

Financial support from the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Fonds Québécois de la Recherche sur la Nature et les Technologies is gratefully acknowledged.

References (31)

  • E.V. Anslyn et al.

    Modern Physical Organic Chemistry

    (2005)
  • C. Bull et al.

    The mechanism of Fe-EDTA catalyzed superoxide dismutation

    Journal of the American Chemical Society

    (1983)
  • D. Chen et al.

    Studies on the mechanism of chelate degradation in iron-based, liquid redox H2S removal processes

    Canadian Journal of Chemistry

    (1995)
  • D. DeBerry

    Chemical evolution of liquid redox processes

    Environmental Progress

    (1997)
  • J.F. Demmink et al.

    Oxidation of ferrous nitrilotriacetic acid with oxygen: a model for oxygen mass transfer parallel to reaction kinetics

    Industrial and Engineering Chemistry Research

    (1997)
  • Cited by (9)

    • Electron shuttling catalytic effect of mellitic acid in zero-valent iron induced oxidative degradation

      2017, Catalysis Today
      Citation Excerpt :

      However, by addition of organic acid ligands capable of iron complexation, the reduction potential of the Fe(II)/Fe(III) redox cycle is lowered, which allows efficient oxygen reduction to form hydrogen peroxide (H2O2) [2]. The higher reactivity of the Fe(II)-ligand complex towards H2O2 decomposition compared to the Fe(II)-aquo species [3–5] also enhances the Fenton reaction and increases the effective hydroxyl radical (HO) concentration for oxidation of aqueous organics [6–9]. The ligands are also capable of reducing Fe(III) (formed during the Fenton reaction) back to Fe(II) [5,7].

    • Study of ferrous iron oxidation in Morocco drinking water in an airlift reactor

      2008, Chemical Engineering and Processing: Process Intensification
    View all citing articles on Scopus
    View full text