<p>In the chlorate process, natural convection arises thanks to the hydrogen evolving cathode. This increases the mass transport of the different species in the chlorate electrolyte. There is a strong connection between mass transport and the kinetics of the electrode reactions. A better knowledge about these phenomena and their interactions is desirable in order to understand e.g. the reasons for deactivation of anode coatings and what process conditions give the longest lifetime and the highest efficiency.</p><p>One of the aims of his work was to understand how the chlorate process has to be run to avoid exceeding the critical anode potential (<em>E</em>_cr) in order to keep the potential losses low and to achieve a long lifetime of the DSAs. At <em>E</em>_cr anodic polarisation curves in chlorate electrolyte bend to higher Tafel slopes, causing increasing potential losses and accelerated ageing of the anode. Therefore the impact on the anode potential and on <em>E</em>_cr of different electrolyte parameters and electrolyte impurities was investigated. Additionally, the work aimed to investigate the impact of an addition of chromate on oxygen evolution and concentration profiles under conditions reminiscent of those in the chlorate process (high ionic strength, 70 °C, ruthenium based DSA, neutral pH), but without chloride in order to avoid hypochlorite formation. For this purpose a model, taking into account mass transport as well as potential- and concentration-dependent electrode reactions and homogeneous reactions was developed. Water oxidation is one of the side reactions considered to decrease the current efficiency in chlorate production. The results from the study increase the understanding of how a buffer/weak base affects a pH dependent electrode reaction in a pH neutral electrolyte in general. This could also throw light on the link between electrode reactions and homogeneous reactions in the chlorate process.</p><p>It was found that the mechanism for chloride oxidation is likely to be the same for potentials below <em>E</em>_cr as well as for potentials above <em>E</em>_cr. This was based on the fact that the apparent reaction order as well as α_a seem to be of the same values even if the anode potential exceeds<em> E</em>_cr. The reason for the higher slope of the polarisation curve above <em>E</em>_cr could then be a potential dependent deactivation of the active sites. Deactivation of active ruthenium sites could occur if ruthenium in a higher oxidation state were inactive for chloride oxidation.</p><p>Concentration gradients of H⁺, OH^-, CrO₄ ^2- and HCrO₄ ^-during oxygen evolution on a rotating disk electrode (RDE) were predicted by simulations. The pH dependent currents at varying potentials calculated by the model were verified in experiments. It was found that an important part of the chromate buffering effect at high current densities occurs in a thin (in the order of nanometers) reaction layer at the anode. From comparisons between the model and experiments a reaction for the chromate buffering has been proposed. Under conditions with bulk pH and chromate concentration similar to those in the chlorate process, the simulations show that the current density for oxygen evolution from OH^- would be approximately 0.1 kA m^-2, which corresponds to about 3% of the total current in chlorate production.</p>

In the chlorate process, natural convection arises thanks to the hydrogen evolving cathode. This increases the mass transport of the different species in the chlorate electrolyte. There is a strong connection between mass transport and the kinetics of the electrode reactions. A better knowledge about these phenomena and their interactions is desirable in order to understand e.g. the reasons for deactivation of anode coatings and what process conditions give the longest lifetime and the highest efficiency. One of the aims of his work was to understand how the chlorate process has to be run to avoid exceeding the critical anode potential (Ecr) in order to keep the potential losses low and to achieve a long lifetime of the DSAs. At Ecr anodic polarisation curves in chlorate electrolyte bend to higher Tafel slopes, causing increasing potential losses and accelerated ageing of the anode. Therefore the impact on the anode potential and on Ecr of different electrolyte parameters and electrolyte impurities was investigated. Additionally, the work aimed to investigate the impact of an addition of chromate on oxygen evolution and concentration profiles under conditions reminiscent of those in the chlorate process (high ionic strength, 70 °C, ruthenium based DSA, neutral pH), but without chloride in order to avoid hypochlorite formation. For this purpose a model, taking into account mass transport as well as potential- and concentration-dependent electrode reactions and homogeneous reactions was developed. Water oxidation is one of the side reactions considered to decrease the current efficiency in chlorate production. The results from the study increase the understanding of how a buffer/weak base affects a pH dependent electrode reaction in a pH neutral electrolyte in general. This could also throw light on the link between electrode reactions and homogeneous reactions in the chlorate process. It was found that the mechanism for chloride oxidation is likely to be the same for potentials below Ecr as well as for potentials above Ecr. This was based on the fact that the apparent reaction order as well as αa seem to be of the same values even if the anode potential exceeds Ecr. The reason for the higher slope of the polarisation curve above Ecr could then be a potential dependent deactivation of the active sites. Deactivation of active ruthenium sites could occur if ruthenium in a higher oxidation state were inactive for chloride oxidation. Concentration gradients of H+, OH-, CrO4 2- and HCrO4 - during oxygen evolution on a rotating disk electrode (RDE) were predicted by simulations. The pH dependent currents at varying potentials calculated by the model were verified in experiments. It was found that an important part of the chromate buffering effect at high current densities occurs in a thin (in the order of nanometers) reaction layer at the anode. From comparisons between the model and experiments a reaction for the chromate buffering has been proposed. Under conditions with bulk pH and chromate concentration similar to those in the chlorate process, the simulations show that the current density for oxygen evolution from OH- would be approximately 0.1 kA m-2, which corresponds to about 3% of the total current in chlorate production. / QC 20101122

The chlorate process is very energy intensive and a major part of the production costs are for electrical energy. Since the electricity prices are constantly increasing and may also vary periodically, the chlorate plants may be forced to adjust their production rate to the price at each moment in order to minimise their costs. Variation of current load requires increased knowledge regarding the electrode behaviour in a wide current range. In this thesis, the aim was to study the impact of the electrolyte on the electrode reactions in order to reduce the energy consumption. The work has mainly been experimental and additionally mathematical modelling has been carried out. A wide current range has been investigated in order to increase the understanding of the phenomena and to obtain results useful for low-load operation during the periods of high electricity cost. To operate the anode as energy efficiently as possible, the anode potential should not exceed the critical potential (Ecr), where the slope of the anodic polarisation curve increases, most likely due to ruthenium(VIII)-formation, and where the side reaction of oxygen evolution increases. In this work, the influence of different electrolyte parameters on Ecr has been studied. It was shown that a higher chloride concentration and an increased temperature lowered Ecr, which was expected to increase the risk of exceeding Ecr. However, this was not observed due to a simultaneous favouring of the chloride oxidation. Hence it was concluded that the electrolyte parameters should be optimised so that the lowest possible anode potential is obtained, which would enable higher current densities without exceeding Ecr. A further conclusion is that the increased slope of the polarisation curve at Ecr was possibly related to the lower activity for chloride oxidation on ruthenium oxidised to ruthenium(VIII). At full-load operation, the cathode potential was shown to be rather independent of the electrolyte composition despite a large variation of electrolyte parameters. The cathode composition appears to be more critical than the electrolyte composition when aiming at reducing the energy consumption. A strategy to increase the cathode activity could be to in situ apply a catalytic film onto the electrode surface. Therefore, Y(III) was added to a chloride electrolyte in order to form a yttrium hydroxide film on the alkaline cathode surface during hydrogen evolution. The yttrium-hydroxide film activated reduction of water (hydrogen evolution) and hindered hypochlorite reduction, proton reduction and nitrate reduction. The inhibiting properties are important for the prevention of side reactions, which currently are avoided by reducing Cr(VI) of the electrolyte on the cathode, producing an inhibiting chromium-hydroxide film. The studies on Y(III) increase the expectations for finding alternatives to the toxic Cr(VI). The addition of chromate to the chlorate electrolyte gives a high cathodic current efficiency and chromate has buffering properties in the electrolyte. The role of the buffer has been investigated for the oxygen evolution from water (one possible anodic side reaction), as well as cathodic hydrogen evolution. Models have been developed for these systems to increase the understanding of the interaction between buffer, electrode reactions and mass transport; the results have been verified experimentally. The chromate buffer increased the limiting current significantly for the cathodic H+ reduction and the cathodic overpotential was reduced drastically at currents lower than the limited current. A too low overpotential could result in the cathodic protection being lost. The presence of chromate buffer increased the limiting current for the oxygen evolution from OH-. The modelling of these systems revealed that the homogeneous reactions connected to the electrode reactions were not in equilibrium at the electrode surface. Further, a good resolution of the interface at the electrode surface was crucial since the, for the electrode reactions, important buffering takes place in an nm-thick reaction layer. / Framställning av klorat är mycket energiintensiv och kräver stora mängder elenergi. Stigande elpriser, som dessutom ofta varierar under dygnet eller säsongsvis, gör att man vill reducera onödiga förluster samt ibland försöka anpassa produktionen så att man när elpriset är högt minskar den, för att sedan öka produktionen igen då elpriset sjunker. Denna flexibla drift kräver ny kunskap om hur elektroderna beter sig i ett större strömintervall än vad som tidigare varit av intresse. Målet med detta arbete var att, med fokus på elektrolytens betydelse, identifiera möjliga förbättringar för kloratprocessen och därmed minska energiförbrukningen. Studierna har i huvudsak varit experimentella men även matematisk modellering har använts. Ett brett strömintervall har undersökts för att bättre förstå fenomenen och för att även kunna använda resultaten då höga elpriser gör att man vill köra processen vid lägre laster än normalt. För att driften av anoden ska vara så energieffektiv som möjligt bör anodpotentialen inte överskrida den kritiska potentialen (Ecr), där den anodiska polarisationskurvan får en högre lutning (troligtvis pga Ru(VIII)-bildning) och bireaktionen syrgasutveckling ökar. I detta arbete har påverkan av olika elektrolytparametrar på Ecr undersökts. Det visade sig att en ökad kloridkoncentration och ökad temperatur sänkte Ecr. Trots att detta borde göra att Ecr lättare överskrids, blev inte detta fallet eftersom kloridoxidationen samtidigt gynnades. Slutsatsen blir därför att elektrolytparametrarna bör optimeras så att lägsta möjliga anodpotential uppnås, vilket då även gör att strömtätheten kan ökas utan att Ecr överskrids. Slutsatsen är vidare att polarisationskurvans högre lutning vid Ecr kan ha att göra med att rutenium oxiderat till rutenium(VIII) har lägre aktivitet för kloridoxidation. Vid full last visade sig katodens potential vara relativt oberoende av elektrolytsammansättningen trots att denna varierades kraftigt. Katodens sammansättning verkar vara viktigare att ta hänsyn till än elektrolytens för kunna åstadkomma en större energibesparing. Ett alternativ till att öka katodens aktivitet skulle vara att in-situ belägga elektrodytan med en katalytisk film. Försök gjordes att sätta till Y(III) till kloridelektrolyt för att under vätgasutveckling fälla ut en yttriumhydroxidfilm på den alkaliska katodytan. Yttriumhydroxidfilmen aktiverade vattenreduktion (vätgasutveckling) och inhiberade hypokloritreduktion, protonreduktion och nitratreduktion. De inhiberande egenskaperna är viktiga för att förhindra bireaktioner, vilka idag hindras av att Cr(VI) i elektrolyten reduceras på katoden och bildar en hindrande kromhydroxidfilm. Försöken med Y(III) visar att det finns goda möjligheter att hitta alternativ till det miljöfarliga Cr(VI). Kromattillsatsen i kloratelektrolyt ger förutom ett högt katodiskt strömutbyte även en buffrande effekt till elektrolyten. Effekten av buffert har undersökts för en av de anodiska bireaktionerna, syrgasutveckling ur vatten, samt för vätgasutvecklingen på katoden. Dessa system har modellerats för att bättre förstå samspelet mellan buffert, elektrodreaktioner och materietransport och resultaten har verifierats experimentellt. Kromatbufferten ökade gränsströmmen för katodisk H+-reduktion betydligt och katodöverpotentialen sjönk kraftigt vid lägre strömmar än gränsströmmen. Detta kan vara ett problem om överpotentialen sjunker så lågt att elektroden inte är katodiskt skyddad. För syrgasutvecklingen ökade närvaron av kromatbuffert gränsströmmen för syrgasutveckling ur OH-. Modellering av dessa system visar att de homogena reaktioner som var kopplade till elektrodreaktionerna inte var i jämvikt vid elektrodytan. Vidare visade det sig vara mycket viktigt med en bra upplösning av gränsskiktet vid elektrodytan, då den buffring som är viktig för elektrodreaktionerna sker i ett mycket tunt reaktionsskikt (nanometertjockt). / <p>QC 20100901</p>