The association between natural organic matter (NOM) and iron (Fe) minerals was widely found in soil and sediments and has been shown to impact the fate of Fe minerals and NOM. Ferrihydrite, a ubiquitous Fe mineral, serves as important sink for NOM and rapidly transforms to secondary Fe minerals in the presence of Fe(II). The associated NOM has been found to influence the Fe(II)-catalyzed ferrihydrite transformation pathway, but it remains unclear how various NOM affects this transformation and the implication. This study specifically investigates how different species of NOM affect Fe(II)-catalyzed ferrihydrite transformation under different C/Fe ratios. A series of Fe isotope tracer experiments were conducted to measure Fe atom exchange and electron transfer between aqueous Fe(II) and ferrihydrite in the presence of diverse NOM species. The fate of Ni during Fe(II)-catalyzed transformation of NOM-Fh coprecipitate was also investigated. Ferrihydrite was found less susceptible to Fe(II)-catalyzed transformation with increasing C/Fe ratio and fulvic acids and Suwannee River NOM (SRNOM) in the coprecipitates need lower C/Fe ratio than humic acids to completely inhibit formation of secondary Fe minerals. At C/Fe ratios where ferrihydrite transformed to secondary minerals, goethite was dominant in ferrihydrite coprecipitated with humic acids, whereas lepidocrocite was favored in ferrihydrite coprecipitated with fulvic acids and SRNOM. Adsorbed SRNOM may be more inhibitive than coprecipitated SRNOM on Fe(II)-catalyzed ferrihydrite transformation under similar C/Fe ratios. Despite no secondary mineral transformation at high C/Fe ratios, Mössbauer spectra indicated electron transfer still occurred between Fe(II) and ferrihydrite coprecipitated with fulvic acid and SRNOM. In addition, isotope tracer experiments revealed that a significant fraction of structural Fe(III) in the ferrihydrite mixed with the aqueous phase Fe(II) (~85%). After reaction with Fe(II), Mössbauer spectroscopy indicated some subtle changes in the crystallinity, particle size or particle interactions in the coprecipitate. The effect of coprecipitated SRNOM on Ni(II) distribution during Fe(II)-catalyzed ferrihydrite transformation was investigated with adsorbed Ni(II) and coprecipitated Ni(II). Ni(II) adsorbed on ferrihydrite was more resistant to acid extraction after Fe(II)-catalyzed transformation and suggested that structural incorporation of Ni into secondary Fe minerals occurred. With coprecipitated SRNOM, ferrihydrite did not transform to secondary minerals in the presence of Fe(II) but extensive Fe atom exchange between aqueous Fe(II) and structural Fe(III) still occurred. Limited change in Ni stability was observed, suggesting there was only small portion of Ni redistributed in the presence of Fe(II). Pre-incorporated Ni(II) in Ni-SRNOM-Fh coprecipitate was partially released (6-8 %) in the presence of Fe(II), but the distribution of remaining Ni(II) in the solid did not change measurably. Our observation suggests that the presence of SRNOM limited the redistribution of Ni most likely because of limited transformation of ferrihydrite to secondary minerals.

Trace metal speciation in aquatic environments is inherently complex due to the large number of possible interactions with dissolved and particulate components. Adsorption onto iron oxyhydroxide and bacterial surfaces, as well as the formation of metal-ligand complexes can play important roles in controlling the fate and transport of trace metals in natural environments. The objective of this study is to describe and understand metal speciation and distribution in a complex biogeochemical system by incrementally increasing the complexity from simple binary systems to a dynamic quaternary system containing a trace metal, iron oxide and bacteria that are active and metabolizing an organic ligand. Copper, cadmium, and phthalic acid (H2Lp) adsorption onto ferrihydrite in binary systems was well reproduced using the diffuse layer model (DLM). The adsorption of H2Lp adsorption was analogous to that of inorganic diprotic acids in terms of the relationship between the adsorption constants and acidity constants. In ternary systems H2Lp caused Cu2+ or Cd2+ adsorption to be either enhanced (due to surface ternary complex formation) or inhibited (due to solution complex formation) depending on the conditions. The DLM could only describe the effect of H2Lp on metal ion sorption by including ternary complexes of the form ≡FeOHMLp (0), where ≡FeOH is a surface site and M is Cu or Cd. The relationship between binary metal adsorption constants and the ternary complex adsorption constants from this and previous studies suggest several properties of ternary complexes. First, ternary complex structures on both ferrihydrite and goethite are either the same or similar. Second, those cations having large adsorption constants also have large equilibrium constants for ternary complex formation. Third, ligands forming stronger solution complexes with cations will also form stronger surface ternary complexes but because of the strong solution complexes these ligands will not necessarily enhance cation adsorption. The bacterial strain Comamonas spp. was isolated from the activated sludge of a wastewater treatment plant. Comamonas spp. could effectively degrade H2Lp in the presence of Cd2+ and ferrihydrite and was therefore chosen to study the effect of H2Lp degradation on Cd2+ speciation. Proton, cadmium and H2Lp adsorption onto Comamonas spp. were measured. The Comamonas spp. titration curve is flatter than that of ferrihydrite, indicating a higher degree of site heterogeneity at the bacterial surface. Adsorption edges of Cd2+ adsorption onto Comamonas spp. occurred over about 4~5 pH units compared to those of ferrihydrite which occurred over ≈ 2 pH units on a dry weight basis. Comamonas spp. can accumulate a larger amount of Cd2+ than ferrihydrite especially under lower pH conditions. Proton and Cd2+ adsorption onto Comamonas spp. cells over a wide sorbent/sorbate and pH range was reasonably well described by a four site non-electrostatic model. The acid-base and Cd2+ adsorption behaviour of Comamonas spp. in this work were within the range of studies of bacteria adsorption. Phthalic acid adsorption onto inactive Comamonas spp. was negligible over a pH range of 3 to 8 and became significant only at pH < 3 where H2Lp was fully protonated. This is consistent with the proposed mechanism for ligand adsorption onto bacterial surfaces which involved a balance between hydrophobic interaction and electrostatic repulsion. The presence of H2Lp decreased Cd2+ adsorption onto Comamonas spp. due to competition for Cd2+ between the bacterial cell surface and the formation of solution complexes of Cd2+. This was accurately modelled with the Cd-Lp solution species indicating that no significant surface ternary interaction occurred between Cd2+, phthalic acid and Comamonas spp.. Cadmium adsorption onto ferrihydrite-Comamonas spp. mixtures was slightly less than the simple additive predicted adsorption of ferrihydrite plus Comamonas spp.. This suggests there is a weak interaction between ferrihydrite and Comamonas spp. and this interaction could be modelled by including a generic reaction between the ferrihydrite and Comamonas spp. surface sites. Cadmium distribution in a system of inactive Comamonas spp.-ferrihydrite in the absence and presence of H2Lp could be predicted by combining the ferrihydrite and bacteria models with the inclusion of the ferrihydrite-bacteria interaction. The effects of H2Lp degradation on Cd2+ distribution were investigated in dynamic systems with live bacteria. Results showed that Cd2+ adsorption in these dynamic systems was reasonably estimated with the model parameters developed in the proceeding experiments though uncertainty exists in the dynamic process with regards to H2Lp biodegradation products and changes in the bacteria population. This thesis was therefore able to provide a better understanding of metal speciation in complex and heterogeneous realistic environments by experimentally examining and modelling metal speciation and distribution in various systems with increasing complexity. This helps to bridge the gap of quantitative description of metal speciation from simple laboratory experiment systems to real world systems, both natural and engineered.

Speciation governs transport and toxicity of trace metals and is important to monitor in natural waters. Geochemical models that predict speciation are valuable tools for monitoring. They can be used for risk assessments and future scenarios such as termination of liming. However, there are often large uncertainties concerning the speciation of iron and aluminium in the models, due to the complicated chemistry of these metals. Both are important in governing the speciation of other metals, due to (i) their capacity to form minerals onto which metals can adsorb and (ii) their ability to compete for binding sites to natural organic matter (NOM). Aluminium is also potentially toxic and is therefore closely monitored in acidified freshwaters. In this study different phases of iron in Swedish lakes were characterised. This required a good method for preconcentrating the iron colloids. A new method was developed in this thesis that uses an anion-exchange column to isolate the iron colloids prior to characterisation with extended X-ray absorption fine structure (EXAFS) spectroscopy. Iron was present as ferrihydrite in particles but was also strongly monomerically complexed to NOM in two Swedish lakes. Based on the results an internally consistent process-based geochemical equilibrium model was presented for Swedish freshwaters. The model was validated for pH (n = 9 400) and inorganic monomeric aluminium (Ali) (n = 3 400). The model could simulate pH and Ali simultaneously, and be used for scenario modelling. In this thesis, modelling scenarios for decreases and complete termination of liming are presented for the 3 000 limed Swedish lakes. The results suggest that liming can be terminated in 30 % of the Swedish lakes and decreased in many other lakes. / <p>QC 20120919</p>

Original estimates of phosphorus (P) concentrations in the Precambrian oceans before 1.9 Ga gave a budget of ~10-25% of modern day levels. This budget was challenged by accounting for high silica (Si) concentrations that were believed to have outcompeted P for binding sites on precipitating iron oxide-hydroxide particles during the chemical oxidation and burial of iron (Fe). Such iron oxide-hydroxide particles are considered as proxies of ancient iron-rich sedimentary rocks, such as banded iron formations, which are often used to infer the dissolved chemistry of trace elements in the ancient oceans. This study raises the question of wether arsenic (As) had an effect of the binding of P to precipitating iron minerals, during the co-precipitation of Iron oxide- hydroxide in elevated Fe and Si concentrations characteristic of the early oceans. This hypothesis is based on the chemical similarities seen between P and As. Results show a more pH dependent competition between P and AsIII, whereby P outcompetes AsIII at a pH &lt;7. The effect decreases as the pH rises until pH ~8 at which the effect cancels out and AsIII becomes somewhat predominant over P. AsV on the other hand, an analogue to P, is outcompeted by P throughout pH 5-10. Distribution coefficients (Kd) of P on iron oxide-hydroxide particles were not affected by the concentration of Si in solution. Average Kd and standard error between concentrations of Si, across the sample pH of 5-10 revealed an average Kd of 0.072 (±0.01) μM-1. This is strikingly similar to another experimental Kd at 0.075 (±0.003) μM-1, when the effects of Si are excluded. The average Kd in this study is also consistent with the average Kd of 0.06 μM-1 from a range of As-rich hydrothermal systems reported in a previous study, supporting the original idea of Precambrian P levels being low. The average Kd between concentrations of Fe revealed a Kd of 0.12 (±0.03) μM-1 although this was not statistically significant from the average Kd between groups of Si. In addition to low levels of P, the Precambrian oceans likely also contained high levels of As, due to the high hydrothermal activity. This scavenging of P from oceanic waters would have become increasingly important as surface oceans became more oxygenated and the presence of AsV would have been greater. Because the availability of Si does not show any great effect on the uptake of P by precipitating iron oxide-hydroxides, Si concentration is likely not a proxy for oceanic P concentrations. It is proposed that low dissolved P levels are consistent with early oceans that w!ere a lot more hydrothermally influenced than the oceans of today. / Prekambriska fosfor (P) nivåer var ursprungligen estimerade till ca 10-25% utav koncentrationen funnen i dagens havsvatten. Denna budget blev motsagd i och med att kisel (Si) sades kunna ersätta bundet fosfor på järn oxid-hydroxid partiklar som precipiterade genom kemisk oxidation och sedimentering av järn (Fe). Dessa järn oxid-hydroxid partiklar anses användbara som proxy för formationen av uråldriga järn-rika sedimentära bergarter såsom banded iron formation (BIF), vilka används idag för att bestämma mängden spårämnen i de uråldriga haven. Denna studie ställer frågan huruvida arsenik (As) påverkar mängden P som binder till precipiterande järn mineral under procession av co-precipitering av järn oxid-hydroxid i lösning med förhöjda koncentrationer av Fe och Si, karakteristiska för the uråldriga haven. Denna hypotes är baserad på de kemiska likheter som finns mellan P och As. Resultaten påvisar en pH beroende konkurrens mellan P och AsIII där P utkonkurrerar AsIII vid låg pH. Effekten av denna konkurrans minskar med ökande pH tills effekten blir omvänd omkring pH 8 och P blir istället till viss del utkonkurrerad av AsIII. AsV å andra sedan, en verklig kemisk analog till P, är kontinuerligt utkonkurrerad av P genom alla utförda pH, pH 5-10. Distribueringskoefficienter (Kd) för P på järn oxid-hydroxid partiklar visade ingen påverkan av mängden Si tillgängligt. Medelvärdet av Kd och standard error mellan data av alls pH, grupperat av Si, gav ett värde av 0.072 (±0.01) μM-1. Detta är påfallande nära ett experimentellt framtaget Kd värde av 0.075 (±0.03) μM-1 då effekten av Si är borttagen. Medelvärdet i denna studie är också sammanfallande med det Kd medelvärde man finner idag från olika hydrotemala system av 0.063 (±0.01) μM-1. Detta ger support till den originala idén att de prekambriska haven troligen hade låga halter P tillgängligt. Medelvärdet av Kd mellan koncentrationer av Fe gav ett värde av 0.12 (±0.03) μM-1, dock var detta värde ej statistiskt significant från det Kd utifrån koncentrationer av Si. Förutom de låga nivåer av P i de Prekambriska haven så var det troligen även höga halter av As på grund av utbredd hydrotermal aktivitet. Detta uppfångande av P i de tidiga haven var troligen en alltmer viktigare process då ytvatten blev syrerikare och den oxiderade formen av As, det vill säga AsV hade varit mer vanligt förekommande. Framför allt då den konkurrerande effekten av Si kan bortses när P såväl som As inte påverkas av dess närvaro till den grad man hade trott. Detta gör även att mängden Si troligen inte är en tillförlitlig proxy för att estimera P nivåer i de uråldriga haven. Därmed föreslås det att de prekambriska haven var k!arakteriserade av låga P nivåer, jämfört med idag.

Trace metal speciation in aquatic environments is inherently complex due to the large number of possible interactions with dissolved and particulate components. Adsorption onto iron oxyhydroxide and bacterial surfaces, as well as the formation of metal-ligand complexes can play important roles in controlling the fate and transport of trace metals in natural environments. The objective of this study is to describe and understand metal speciation and distribution in a complex biogeochemical system by incrementally increasing the complexity from simple binary systems to a dynamic quaternary system containing a trace metal, iron oxide and bacteria that are active and metabolizing an organic ligand. Copper, cadmium, and phthalic acid (H2Lp) adsorption onto ferrihydrite in binary systems was well reproduced using the diffuse layer model (DLM). The adsorption of H2Lp adsorption was analogous to that of inorganic diprotic acids in terms of the relationship between the adsorption constants and acidity constants. In ternary systems H2Lp caused Cu2+ or Cd2+ adsorption to be either enhanced (due to surface ternary complex formation) or inhibited (due to solution complex formation) depending on the conditions. The DLM could only describe the effect of H2Lp on metal ion sorption by including ternary complexes of the form ≡FeOHMLp (0), where ≡FeOH is a surface site and M is Cu or Cd. The relationship between binary metal adsorption constants and the ternary complex adsorption constants from this and previous studies suggest several properties of ternary complexes. First, ternary complex structures on both ferrihydrite and goethite are either the same or similar. Second, those cations having large adsorption constants also have large equilibrium constants for ternary complex formation. Third, ligands forming stronger solution complexes with cations will also form stronger surface ternary complexes but because of the strong solution complexes these ligands will not necessarily enhance cation adsorption. The bacterial strain Comamonas spp. was isolated from the activated sludge of a wastewater treatment plant. Comamonas spp. could effectively degrade H2Lp in the presence of Cd2+ and ferrihydrite and was therefore chosen to study the effect of H2Lp degradation on Cd2+ speciation. Proton, cadmium and H2Lp adsorption onto Comamonas spp. were measured. The Comamonas spp. titration curve is flatter than that of ferrihydrite, indicating a higher degree of site heterogeneity at the bacterial surface. Adsorption edges of Cd2+ adsorption onto Comamonas spp. occurred over about 4~5 pH units compared to those of ferrihydrite which occurred over ≈ 2 pH units on a dry weight basis. Comamonas spp. can accumulate a larger amount of Cd2+ than ferrihydrite especially under lower pH conditions. Proton and Cd2+ adsorption onto Comamonas spp. cells over a wide sorbent/sorbate and pH range was reasonably well described by a four site non-electrostatic model. The acid-base and Cd2+ adsorption behaviour of Comamonas spp. in this work were within the range of studies of bacteria adsorption. Phthalic acid adsorption onto inactive Comamonas spp. was negligible over a pH range of 3 to 8 and became significant only at pH < 3 where H2Lp was fully protonated. This is consistent with the proposed mechanism for ligand adsorption onto bacterial surfaces which involved a balance between hydrophobic interaction and electrostatic repulsion. The presence of H2Lp decreased Cd2+ adsorption onto Comamonas spp. due to competition for Cd2+ between the bacterial cell surface and the formation of solution complexes of Cd2+. This was accurately modelled with the Cd-Lp solution species indicating that no significant surface ternary interaction occurred between Cd2+, phthalic acid and Comamonas spp.. Cadmium adsorption onto ferrihydrite-Comamonas spp. mixtures was slightly less than the simple additive predicted adsorption of ferrihydrite plus Comamonas spp.. This suggests there is a weak interaction between ferrihydrite and Comamonas spp. and this interaction could be modelled by including a generic reaction between the ferrihydrite and Comamonas spp. surface sites. Cadmium distribution in a system of inactive Comamonas spp.-ferrihydrite in the absence and presence of H2Lp could be predicted by combining the ferrihydrite and bacteria models with the inclusion of the ferrihydrite-bacteria interaction. The effects of H2Lp degradation on Cd2+ distribution were investigated in dynamic systems with live bacteria. Results showed that Cd2+ adsorption in these dynamic systems was reasonably estimated with the model parameters developed in the proceeding experiments though uncertainty exists in the dynamic process with regards to H2Lp biodegradation products and changes in the bacteria population. This thesis was therefore able to provide a better understanding of metal speciation in complex and heterogeneous realistic environments by experimentally examining and modelling metal speciation and distribution in various systems with increasing complexity. This helps to bridge the gap of quantitative description of metal speciation from simple laboratory experiment systems to real world systems, both natural and engineered.

Trace metal speciation in aquatic environments is inherently complex due to the large number of possible interactions with dissolved and particulate components. Adsorption onto iron oxyhydroxide and bacterial surfaces, as well as the formation of metal-ligand complexes can play important roles in controlling the fate and transport of trace metals in natural environments. The objective of this study is to describe and understand metal speciation and distribution in a complex biogeochemical system by incrementally increasing the complexity from simple binary systems to a dynamic quaternary system containing a trace metal, iron oxide and bacteria that are active and metabolizing an organic ligand. Copper, cadmium, and phthalic acid (H2Lp) adsorption onto ferrihydrite in binary systems was well reproduced using the diffuse layer model (DLM). The adsorption of H2Lp adsorption was analogous to that of inorganic diprotic acids in terms of the relationship between the adsorption constants and acidity constants. In ternary systems H2Lp caused Cu2+ or Cd2+ adsorption to be either enhanced (due to surface ternary complex formation) or inhibited (due to solution complex formation) depending on the conditions. The DLM could only describe the effect of H2Lp on metal ion sorption by including ternary complexes of the form ≡FeOHMLp (0), where ≡FeOH is a surface site and M is Cu or Cd. The relationship between binary metal adsorption constants and the ternary complex adsorption constants from this and previous studies suggest several properties of ternary complexes. First, ternary complex structures on both ferrihydrite and goethite are either the same or similar. Second, those cations having large adsorption constants also have large equilibrium constants for ternary complex formation. Third, ligands forming stronger solution complexes with cations will also form stronger surface ternary complexes but because of the strong solution complexes these ligands will not necessarily enhance cation adsorption. The bacterial strain Comamonas spp. was isolated from the activated sludge of a wastewater treatment plant. Comamonas spp. could effectively degrade H2Lp in the presence of Cd2+ and ferrihydrite and was therefore chosen to study the effect of H2Lp degradation on Cd2+ speciation. Proton, cadmium and H2Lp adsorption onto Comamonas spp. were measured. The Comamonas spp. titration curve is flatter than that of ferrihydrite, indicating a higher degree of site heterogeneity at the bacterial surface. Adsorption edges of Cd2+ adsorption onto Comamonas spp. occurred over about 4~5 pH units compared to those of ferrihydrite which occurred over ≈ 2 pH units on a dry weight basis. Comamonas spp. can accumulate a larger amount of Cd2+ than ferrihydrite especially under lower pH conditions. Proton and Cd2+ adsorption onto Comamonas spp. cells over a wide sorbent/sorbate and pH range was reasonably well described by a four site non-electrostatic model. The acid-base and Cd2+ adsorption behaviour of Comamonas spp. in this work were within the range of studies of bacteria adsorption. Phthalic acid adsorption onto inactive Comamonas spp. was negligible over a pH range of 3 to 8 and became significant only at pH < 3 where H2Lp was fully protonated. This is consistent with the proposed mechanism for ligand adsorption onto bacterial surfaces which involved a balance between hydrophobic interaction and electrostatic repulsion. The presence of H2Lp decreased Cd2+ adsorption onto Comamonas spp. due to competition for Cd2+ between the bacterial cell surface and the formation of solution complexes of Cd2+. This was accurately modelled with the Cd-Lp solution species indicating that no significant surface ternary interaction occurred between Cd2+, phthalic acid and Comamonas spp.. Cadmium adsorption onto ferrihydrite-Comamonas spp. mixtures was slightly less than the simple additive predicted adsorption of ferrihydrite plus Comamonas spp.. This suggests there is a weak interaction between ferrihydrite and Comamonas spp. and this interaction could be modelled by including a generic reaction between the ferrihydrite and Comamonas spp. surface sites. Cadmium distribution in a system of inactive Comamonas spp.-ferrihydrite in the absence and presence of H2Lp could be predicted by combining the ferrihydrite and bacteria models with the inclusion of the ferrihydrite-bacteria interaction. The effects of H2Lp degradation on Cd2+ distribution were investigated in dynamic systems with live bacteria. Results showed that Cd2+ adsorption in these dynamic systems was reasonably estimated with the model parameters developed in the proceeding experiments though uncertainty exists in the dynamic process with regards to H2Lp biodegradation products and changes in the bacteria population. This thesis was therefore able to provide a better understanding of metal speciation in complex and heterogeneous realistic environments by experimentally examining and modelling metal speciation and distribution in various systems with increasing complexity. This helps to bridge the gap of quantitative description of metal speciation from simple laboratory experiment systems to real world systems, both natural and engineered.

Trace metal speciation in aquatic environments is inherently complex due to the large number of possible interactions with dissolved and particulate components. Adsorption onto iron oxyhydroxide and bacterial surfaces, as well as the formation of metal-ligand complexes can play important roles in controlling the fate and transport of trace metals in natural environments. The objective of this study is to describe and understand metal speciation and distribution in a complex biogeochemical system by incrementally increasing the complexity from simple binary systems to a dynamic quaternary system containing a trace metal, iron oxide and bacteria that are active and metabolizing an organic ligand. Copper, cadmium, and phthalic acid (H2Lp) adsorption onto ferrihydrite in binary systems was well reproduced using the diffuse layer model (DLM). The adsorption of H2Lp adsorption was analogous to that of inorganic diprotic acids in terms of the relationship between the adsorption constants and acidity constants. In ternary systems H2Lp caused Cu2+ or Cd2+ adsorption to be either enhanced (due to surface ternary complex formation) or inhibited (due to solution complex formation) depending on the conditions. The DLM could only describe the effect of H2Lp on metal ion sorption by including ternary complexes of the form ≡FeOHMLp (0), where ≡FeOH is a surface site and M is Cu or Cd. The relationship between binary metal adsorption constants and the ternary complex adsorption constants from this and previous studies suggest several properties of ternary complexes. First, ternary complex structures on both ferrihydrite and goethite are either the same or similar. Second, those cations having large adsorption constants also have large equilibrium constants for ternary complex formation. Third, ligands forming stronger solution complexes with cations will also form stronger surface ternary complexes but because of the strong solution complexes these ligands will not necessarily enhance cation adsorption. The bacterial strain Comamonas spp. was isolated from the activated sludge of a wastewater treatment plant. Comamonas spp. could effectively degrade H2Lp in the presence of Cd2+ and ferrihydrite and was therefore chosen to study the effect of H2Lp degradation on Cd2+ speciation. Proton, cadmium and H2Lp adsorption onto Comamonas spp. were measured. The Comamonas spp. titration curve is flatter than that of ferrihydrite, indicating a higher degree of site heterogeneity at the bacterial surface. Adsorption edges of Cd2+ adsorption onto Comamonas spp. occurred over about 4~5 pH units compared to those of ferrihydrite which occurred over ≈ 2 pH units on a dry weight basis. Comamonas spp. can accumulate a larger amount of Cd2+ than ferrihydrite especially under lower pH conditions. Proton and Cd2+ adsorption onto Comamonas spp. cells over a wide sorbent/sorbate and pH range was reasonably well described by a four site non-electrostatic model. The acid-base and Cd2+ adsorption behaviour of Comamonas spp. in this work were within the range of studies of bacteria adsorption. Phthalic acid adsorption onto inactive Comamonas spp. was negligible over a pH range of 3 to 8 and became significant only at pH < 3 where H2Lp was fully protonated. This is consistent with the proposed mechanism for ligand adsorption onto bacterial surfaces which involved a balance between hydrophobic interaction and electrostatic repulsion. The presence of H2Lp decreased Cd2+ adsorption onto Comamonas spp. due to competition for Cd2+ between the bacterial cell surface and the formation of solution complexes of Cd2+. This was accurately modelled with the Cd-Lp solution species indicating that no significant surface ternary interaction occurred between Cd2+, phthalic acid and Comamonas spp.. Cadmium adsorption onto ferrihydrite-Comamonas spp. mixtures was slightly less than the simple additive predicted adsorption of ferrihydrite plus Comamonas spp.. This suggests there is a weak interaction between ferrihydrite and Comamonas spp. and this interaction could be modelled by including a generic reaction between the ferrihydrite and Comamonas spp. surface sites. Cadmium distribution in a system of inactive Comamonas spp.-ferrihydrite in the absence and presence of H2Lp could be predicted by combining the ferrihydrite and bacteria models with the inclusion of the ferrihydrite-bacteria interaction. The effects of H2Lp degradation on Cd2+ distribution were investigated in dynamic systems with live bacteria. Results showed that Cd2+ adsorption in these dynamic systems was reasonably estimated with the model parameters developed in the proceeding experiments though uncertainty exists in the dynamic process with regards to H2Lp biodegradation products and changes in the bacteria population. This thesis was therefore able to provide a better understanding of metal speciation in complex and heterogeneous realistic environments by experimentally examining and modelling metal speciation and distribution in various systems with increasing complexity. This helps to bridge the gap of quantitative description of metal speciation from simple laboratory experiment systems to real world systems, both natural and engineered.

Iron is often present in leach liquors produced in chemical and hydrometallurgical processes. It is known that voluminous iron precipitates with high impurity values are formed if the conditions during its formation are not controlled well. These products are also often difficult to treat in downstream processes. This study therefore focused on the determination of product quality parameters for the production of good quality iron precipitates from zinc-rich solutions. Special attention was given to the quality of metastable phases such as ferrihydrite and schwertmannite formed at elevated temperatures and in the pH range 1.5 to 3.5 in a continuous crystallizer. These phases are produced over a range of supersaturation levels with the best quality products formed at lower supersaturation. It was shown that most industrial processes are operated well above the metastability limit at relatively high supersaturation. However, stagewise precipitation of iron, even above the metastability limit, yielded better quality products. It was also shown that localized supersaturation levels could be controlled through changes in the micro and macromixing environments. The three-zone model approach was used to improve the quality of ferrihydrite and schwertmannite precipitates. Changes in the reactor design and the position of reagent feed points also impacted on the quality of the precipitates. Control over the localized supersaturation not only ensures the production of good quality nuclei, but also impacts on particle growth, which is required to make downstream processing of precipitates possible. In precipitation processes, growth mainly takes place through agglomeration as the rate of molecular growth is generally low. The final quality of iron precipitates is greatly influenced by the quality of the agglomerates formed during iron precipitation. A Hadamard matrix was used to indicate the relative importance of the most relevant operating parameters for the formation of good quality iron precipitates. / Thesis (PhD (Metallugical Engineering))--University of Pretoria, 2007. / Materials Science and Metallurgical Engineering / unrestricted

Mining and mineral processing generate a wide range of As-rich minerals, including scorodite (FeAsO4•2H2O), and arsenical ferrihydrite, which are common secondary minerals found in mine tailings. Scorodite and arsenical ferrihydrite are relatively stable under a wide range of physico-chemical conditions which makes them suitable arsenic sinks in mining environments. However, bacteria can reduce these minerals and release arsenic into the aqueous environment. Two dissimilatory iron and arsenic reducing bacteria, Shewanella sp. ANA-3 and Shewanella putrefaciens CN32, were used to investigate their effects on the reductive dissolution of scorodite and arsenical 6-line ferrihydrite in a chemically defined medium containing low phosphate concentrations representative of the natural environment. Analysis of the aqueous phase of all biotic reduced samples found mainly As(III), the more toxic form of As, while very little As(V) was reduced in the abiotic samples. Solid state analysis of the scorodite biotic post-reduction minerals identified scorodite, biogenic Fe(II)-As(III) compounds, parasymplesite and tooeleite, while in the biotic reduced arsenical six-line ferrihydrite, biogenic Fe(II)-As(III) compounds, hematite, akaganeite and unconfirmed magnetite were identified as secondary reduction products. Results from this research add to the body of literature on As and Fe biogeochemistry and provide very useful information for future assessments of the long term stability of As-rich minerals. L’activité minière et la transformation du minerai génèrent divers minéraux riches en arsenic, tels la scorodite (FeAsO4•2H2O) et la ferrihydrite riche en arsenic, lesquels sont des minéraux secondaires communs des résidus miniers. Comme la scorodite et la ferrihydrite riche en arsenic sont relativement stables sous une grande gamme de conditions physico-chimiques, ces minéraux peuvent potentiellement être utilisés pour stocker de façon permanente l’arsenic dans les environnements miniers. Cependant, certaines bactéries peuvent réduire ces minéraux, ce qui entraine la solubilisation de l’arsenic. Deux bactéries capables de réduire l’arsenic et le fer, soit Shewanella sp. ANA-3 et Shewanella putrefaciens CN32, ont été utilisées afin de déterminer leurs effets sur la réduction microbienne de la scorodite et de la ferrihydrite riche en As dans un milieu de culture contenant de faibles concentrations de phosphate. Les analyses de la phase aqueuse ont démontré que dans tous les systèmes biotiques, As(V) a été réduit en As(III), alors que dans les systèmes contrôles abiotiques, peu de As(V) a été réduit. L’analyse des minéraux secondaires présents à la fin réduction dans les systèmes biotiques contenant de la scorodite indique que la scorodite est encore présente, ainsi que des composés organiques riches en Fe(II) et As(III), de la parasymplésite et de la tooéleite, alors que dans les systèmes biotiques contenant de la ferrihydrite riche en As, des composés riches en Fe(II) et en As(III), de l’hématite, de l’akaganéite et de la magnétique ont été identifiés comme minéraux secondaires. Les résultats de cette étude enrichissent la littérature sur le cycle biogéochimique du Fe et de As et fournissent de l’information importante pour l’évaluation de la stabilité à long terme de minéraux riches en As.

The redox cycling of iron is intimately linked to the cycling of C, S, N, P as well as the speciation, mobility, and bioavailability of various toxic contaminants in soils and sediments. Within these environments, the cycling of iron is catalytically driven by iron-oxidizing (FeOB) and iron reducing bacteria (FeRB) which mediate the formation, transformation, and dissolution of various iron-bearing minerals. Under oxic conditions, FeOB promote the formation of iron oxides on or in close proximity of their cell walls and extracellular polymeric substances, and such composite, termed biogenic iron oxides (BIOS), offers highly reactive heterogenous sites that efficiently immobilize trace metals and contaminants alike. However, under reducing conditions, FeRB mediate the reductive dissolution of BIOS and in turn lead to the remobilization of associated contaminants. Conversely, contaminants may become immobilized by secondary iron minerals that form from the metabolic activity of FeRB. Therefore, determining the factors that influence the reactivity of BIOS, as well as the formation of secondary iron minerals is of critical importance to develop a better understanding of the geochemical cycling of iron and in turn the transport of contaminants in the environment. This thesis investigated (1) the impact of simulated diagenesis (ageing for ~5 years at 4ºC) on the mineral stability and reactivity of BIOS towards reduction by Shewanella putrefaciens CN32, (2) the effects of phosphate at an environmentally relevant (10µM) and excess (3.9mM) concentration on the rates and extent of microbial reduction of synthetic 2-line ferrihydrite and BIOS, as well as the formation of secondary iron minerals, and (3) the impact of sterilization by γ-irradiation on the mineral stability and reactivity of BIOS. It was found that simulated diagenesis did not affect the mineralogical composition of BIOS but significantly lowered the reactivity of BIOS towards microbial reduction. The concentration of phosphate was found to have contrasting effects on the rates of reduction of ferrihydrite and BIOS, but in general, excess concentration of phosphate enhanced the extent of Fe(III) reduction. The formation of a specific secondary iron mineral was also found to depend on the concentration of phosphate, as well as, in the case for BIOS, the presence of intermixed cell derived organic matter. γ-irradiation did not alter the mineralogy and reactivity of BIOS towards microbial reduction, and it was concluded to be a suitable technique to sterilize BIOS.