Metastable layered perovskites containing interlayer transition metals can readily be obtained by simple ion exchange reactions on receptive hosts, such as those of the Dion-Jacobson and Ruddlesden-Popper structure types. In this work, we focused on adding to the library of layered perovskites by not only creating new compounds, but by also showing their ability to be further manipulated, and by studying the stability of the series through thermal behavior studies. The reactions with transition metal halides are particularly interesting since they often lead to novel architectures and magnetic behavior. On subsequent heat treatment, these exchange products typically decompose to thermodynamically more stable phases. The newly synthesized spin glass-like material, FeLa2Ti3O10, obtained by ion exchange of Li2La2Ti3O10 with FeCl2 at 350 °C, behaves differently. When heated to 700 °C, the compound undergoes a significant cell contraction (Δc ≈ -2.7 Å) with an increase in the oxidation state of iron present in the interlayer that not been observed before in such compounds. Efforts were also made to synthesize new series of compounds, here MSrTa2O7 (M= Co and Zn), with vacancies in the interlayer that could lead to future topochemical manipulations. The ability to vary the composition of different phases to form solid-solutions through atomic substitution at the A or B sites with ions of similar or different charge can lead to new structures as well an enhancement of the properties of the original compound or new ones. The synthesis and characterization of the new mixed A-cation containing layered perovskite RbLaNaNb3O10, where La3+ and Na+ share the same site in the perovskite slab and RbLaCaNb2MnO10 that exhibits an ordering of the B site with Mn in the center of the perovskite slab will be presented. Further topochemical manipulation of these phases via ion exchange reactions at low temperatures (< 500 °C), lead to the new series A’LaNaNb3O10 and A’LaCaNb2MnO10 where A’= H, Li, Na, K and CuCl.

京都大学 / 0048 / 新制・課程博士 / 博士(エネルギー科学) / 甲第20477号 / エネ博第346号 / 新制||エネ||69(附属図書館) / 京都大学大学院エネルギー科学研究科エネルギー社会・環境科学専攻 / (主査)教授 坂 志朗, 教授 髙部 圭司, 准教授 河本 晴雄 / 学位規則第4条第1項該当 / Doctor of Energy Science / Kyoto University / DGAM

京都大学 / 新制・課程博士 / 博士(工学) / 甲第24628号 / 工博第5134号 / 新制||工||1981(附属図書館) / 京都大学大学院工学研究科物質エネルギー化学専攻 / (主査)教授 陰山 洋, 教授 阿部 竜, 教授 藤原 哲晶 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

No description available.

O bagaço de cana-de-açúcar é um material rico em celulose, molécula que pode ser convertida enzimaticamente em glicose pela ação de celulases. No entanto, no bagaço, a celulose está associada à hemicelulose e à lignina, componentes que limitam a sua digestibilidade enzimática. A lignina é o principal limitante da hidrólise da celulose, reduzindo a acessibilidade das enzimas celulolíticas. A hemicelulose também limita a ação das celulases e a presença de ácidos hidroxicinâmicos, como ferúlico e cumárico contribuem para reduzir os níveis de hidrólise. Nesse trabalho, o bagaço de cana-de-açúcar foi tratado com clorito de sódio em meio ácido com o objetivo de remover seletivamente a lignina e produzir modelos com teores reduzidos desse componente. O bagaço de cana controle continha 22,8% de lignina e após a deslignificação por 4 horas, materiais com até 6,8% de lignina foram obtidos, mantendo quase inalteradas as frações celulósicas e hemicelulósicas. Os bagaços foram submetidos à hidrólise enzimática com celulases comerciais. Apenas 26% da celulose de bagaço controle foram convertidos em glicose e 11% da xilana foram convertidos em xilose, com 72 horas de hidrólise. Nos bagaços com menores teores de lignina, os valores de conversão de celulose e de xilana alcançaram 85% e 64% no mesmo tempo de hidrólise, respectivamente. Os resultados mostraram que existe uma correlação inversa entre o teor de lignina e a conversão de polissacarídeos e que a remoção de 60% da lignina confere níveis elevados de hidrólise. Para reduzir a inibição pelos produtos de hidrólise da celulose, foi adicionada &#946;-glicosidase ao meio reacional. Com a adição dessa enzima, os bagaços mais deslignificados alcançaram 100% de conversão de celulose e 82% de xilana, com 48 horas de hidrólise. No presente trabalho, o efeito da lignina não ligada aos polissacarídeos também foi avaliado. Para isso, lignina de bagaço de cana foi adicionada ao cartão de celulose, e este foi submetido à hidrólise enzimática. Não houve diferença entre os níveis de hidrólise na presença e ausência de lignina, indicando que a lignina limita a hidrólise dos polissacarídeos quando ligada a eles. Foi utilizada a técnica de microespectrofotometria ultravioleta para detectar a lignina e os ácidos hidroxicinâmicos em fibras, vasos e células de parênquima das regiões de córtex e de medula, deslignificados e controle. Os espectros UV dos três tipos celulares mostraram bandas em 278 nm, associada à lignina, e em 315 nm, associada aos ácidos hidroxicinâmicos. Os vasos foram os tipos celulares mais lignificados, seguidos pelas fibras e células parenquimatosas. O tratamento com clorito causou uma rápida remoção dos ácidos hidroxicinâmicos das células de parênquima, enquanto que as fibras foram deslignificadas apenas com 4 horas de tratamento. As amostras de córtex e de medula também foram submetidas à hidrólise com celulases. A região de medula não tratada foi prontamente hidrolizada, alcançando 63% de conversão de celulose com 72 horas, enquanto que o córtex não tratado atingiu apenas 20%. O tratamento com clorito de sódio não aumentou os níveis de hidrólise da medula. No entanto, a remoção de lignina do córtex aumentou consideravelmente a conversão. / The sugarcane bagasse is a material rich in cellulose, wich can be enzymatically converted into glucose by the action of cellulases. However, in the bagasse, it is associated to hemicellulose and lignin, components that limit its enzymatic digestibility. Lignin is the main limitant of the enzymatic hydrolysis of cellulose, reducing the accessibility of cellulases. Hemicellulose also limits the cellulase action and the presence of the hydroxycinnamic acids, as ferulic and coumaric, contribute to reduce the hydrolysis levels. In this work, the sugarcane bagasse was treated with acid sodium chlorite, aiming to selective remove the lignin and produce models with lower contents of this component. The control bagasse had 22.8% of lignin and after a 4-hour delignification, materials with up to 6.8% of lignin were obtained, retaining the cellulosic and hemicellulosic fractions almost unchanged. The bagasses were submitted to the enzymatic hydrolysis with comercial cellulases. Only 26% of the cellulose of the control bagasse were converted into glucose and 11% of the xylan were converted into xylose, with 72 hours of hydrolysis. In the bagasses with lower lignin content, the cellulose and xylan conversion values reached 85% and 64% in the same time of hysylysis, respectively. The results showed that there is an inverse correlation between the lignin content and the enzymatic conversion of polysaccharides and that the removal of 60% of the lignin results in high levels of hydrolysis. To reduce the inhibition by the products of the cellulose hydrolysis, &#946;- glucosidase was added to the reaction medium. With this enzyme, the most delignified bagasses reached 100% of cellulose and 82% of xylan convertion, with 48 hours of hydrolysis. In this work, the effect of non-linked lignin was also analysed. For this, sugarcane bagasse lignin was added to cellulose card, and it was submitted to the enzymatic hydrolysis. There were no differences between the hydrolysis levels in the presence or absence of lignin, suggesting that lignin only limits hydrolysis when linked to the polysaccharides. The ultraviolet microspectrophotometry technique was used to detect lignin and hydroxycinnamic acids in fiber, vessels and parenchyma cells in rind and pith regions, delignified and control. The UV spectra of the three cell types showed bands in 278 nm, related to lignin, and 315 nm, related to the hydroxycynamic acids. The vessels were the most lignified cell type, followed by fibers and parenchyma cells. The chlorite treatment caused a rapid removal of hydroxycynnamic acids from parenchyma cell walls, while fibers where delignified only with a 4-hour treatment. The rind and pith samples were also submitted to enzymatic hydrolysis. The untreated pith region was promptly hydrolysed, reaching 63% of cellulose convertion in 72 hours, while untreated rind reached only 20%. The sodium chlorite treatment did not enhance the pith hydrolysis levels. However, the lignin removal in rind samples substantially enhanced the conversion.

O bagaço de cana-de-açúcar é um material rico em celulose, molécula que pode ser convertida enzimaticamente em glicose pela ação de celulases. No entanto, no bagaço, a celulose está associada à hemicelulose e à lignina, componentes que limitam a sua digestibilidade enzimática. A lignina é o principal limitante da hidrólise da celulose, reduzindo a acessibilidade das enzimas celulolíticas. A hemicelulose também limita a ação das celulases e a presença de ácidos hidroxicinâmicos, como ferúlico e cumárico contribuem para reduzir os níveis de hidrólise. Nesse trabalho, o bagaço de cana-de-açúcar foi tratado com clorito de sódio em meio ácido com o objetivo de remover seletivamente a lignina e produzir modelos com teores reduzidos desse componente. O bagaço de cana controle continha 22,8% de lignina e após a deslignificação por 4 horas, materiais com até 6,8% de lignina foram obtidos, mantendo quase inalteradas as frações celulósicas e hemicelulósicas. Os bagaços foram submetidos à hidrólise enzimática com celulases comerciais. Apenas 26% da celulose de bagaço controle foram convertidos em glicose e 11% da xilana foram convertidos em xilose, com 72 horas de hidrólise. Nos bagaços com menores teores de lignina, os valores de conversão de celulose e de xilana alcançaram 85% e 64% no mesmo tempo de hidrólise, respectivamente. Os resultados mostraram que existe uma correlação inversa entre o teor de lignina e a conversão de polissacarídeos e que a remoção de 60% da lignina confere níveis elevados de hidrólise. Para reduzir a inibição pelos produtos de hidrólise da celulose, foi adicionada &#946;-glicosidase ao meio reacional. Com a adição dessa enzima, os bagaços mais deslignificados alcançaram 100% de conversão de celulose e 82% de xilana, com 48 horas de hidrólise. No presente trabalho, o efeito da lignina não ligada aos polissacarídeos também foi avaliado. Para isso, lignina de bagaço de cana foi adicionada ao cartão de celulose, e este foi submetido à hidrólise enzimática. Não houve diferença entre os níveis de hidrólise na presença e ausência de lignina, indicando que a lignina limita a hidrólise dos polissacarídeos quando ligada a eles. Foi utilizada a técnica de microespectrofotometria ultravioleta para detectar a lignina e os ácidos hidroxicinâmicos em fibras, vasos e células de parênquima das regiões de córtex e de medula, deslignificados e controle. Os espectros UV dos três tipos celulares mostraram bandas em 278 nm, associada à lignina, e em 315 nm, associada aos ácidos hidroxicinâmicos. Os vasos foram os tipos celulares mais lignificados, seguidos pelas fibras e células parenquimatosas. O tratamento com clorito causou uma rápida remoção dos ácidos hidroxicinâmicos das células de parênquima, enquanto que as fibras foram deslignificadas apenas com 4 horas de tratamento. As amostras de córtex e de medula também foram submetidas à hidrólise com celulases. A região de medula não tratada foi prontamente hidrolizada, alcançando 63% de conversão de celulose com 72 horas, enquanto que o córtex não tratado atingiu apenas 20%. O tratamento com clorito de sódio não aumentou os níveis de hidrólise da medula. No entanto, a remoção de lignina do córtex aumentou consideravelmente a conversão. / The sugarcane bagasse is a material rich in cellulose, wich can be enzymatically converted into glucose by the action of cellulases. However, in the bagasse, it is associated to hemicellulose and lignin, components that limit its enzymatic digestibility. Lignin is the main limitant of the enzymatic hydrolysis of cellulose, reducing the accessibility of cellulases. Hemicellulose also limits the cellulase action and the presence of the hydroxycinnamic acids, as ferulic and coumaric, contribute to reduce the hydrolysis levels. In this work, the sugarcane bagasse was treated with acid sodium chlorite, aiming to selective remove the lignin and produce models with lower contents of this component. The control bagasse had 22.8% of lignin and after a 4-hour delignification, materials with up to 6.8% of lignin were obtained, retaining the cellulosic and hemicellulosic fractions almost unchanged. The bagasses were submitted to the enzymatic hydrolysis with comercial cellulases. Only 26% of the cellulose of the control bagasse were converted into glucose and 11% of the xylan were converted into xylose, with 72 hours of hydrolysis. In the bagasses with lower lignin content, the cellulose and xylan conversion values reached 85% and 64% in the same time of hysylysis, respectively. The results showed that there is an inverse correlation between the lignin content and the enzymatic conversion of polysaccharides and that the removal of 60% of the lignin results in high levels of hydrolysis. To reduce the inhibition by the products of the cellulose hydrolysis, &#946;- glucosidase was added to the reaction medium. With this enzyme, the most delignified bagasses reached 100% of cellulose and 82% of xylan convertion, with 48 hours of hydrolysis. In this work, the effect of non-linked lignin was also analysed. For this, sugarcane bagasse lignin was added to cellulose card, and it was submitted to the enzymatic hydrolysis. There were no differences between the hydrolysis levels in the presence or absence of lignin, suggesting that lignin only limits hydrolysis when linked to the polysaccharides. The ultraviolet microspectrophotometry technique was used to detect lignin and hydroxycinnamic acids in fiber, vessels and parenchyma cells in rind and pith regions, delignified and control. The UV spectra of the three cell types showed bands in 278 nm, related to lignin, and 315 nm, related to the hydroxycynamic acids. The vessels were the most lignified cell type, followed by fibers and parenchyma cells. The chlorite treatment caused a rapid removal of hydroxycynnamic acids from parenchyma cell walls, while fibers where delignified only with a 4-hour treatment. The rind and pith samples were also submitted to enzymatic hydrolysis. The untreated pith region was promptly hydrolysed, reaching 63% of cellulose convertion in 72 hours, while untreated rind reached only 20%. The sodium chlorite treatment did not enhance the pith hydrolysis levels. However, the lignin removal in rind samples substantially enhanced the conversion.

Im Rahmen der vorliegenden Arbeit ist es gelungen, ein Siliciumsubnitrid der Summenformel Si2N chemisch und strukturell zu charakterisieren. Die braune, metastabile Verbindung ist über eine topotaktische Reaktion aus der Zintl-Phase Calciumdisilicid und Ammoniumbromid im molaren Verhältnis 1:2 zugänglich. Charakteristisch für das Siliciumsubnitrid sind eine schichtartige Morphologie und perfekte Spaltbarkeit parallel zu den Schichten. Thermoanalytische und röntgenografische Untersuchungen belegen, dass die Reaktion im Temperaturbereich 165°- 350°C abläuft. Beim Einsatz pulverförmiger Eduktmischungen wird das Siliciumsubnitrid als röntgenamorphes Pulver im mikroskopischen Gemisch mit kristallinem CaBr2 erhalten. Die Abtrennung des Calciumbromids gelingt bisher nur unvollständig. Mit spektroskopischen Methoden (NMR, IR und Raman) konnten die Baueinheiten im Siliciumsubnitrid identifiziert werden. Stickstoff liegt im Siliciumsubnitrid in Form einer trigonalen [N(Si)3]-Koordination vor. Über REDOR-NMR-Experimente wurde belegt, dass das Subnitrid zwei unterschiedliche Siliciumspezies, eine [Si(Si4)]- und eine [Si(Si,N3)]-Umgebung, enthält. Ergebnisse elektronenmikroskopischer Untersuchungen (REM, TEM) sind mit der Bildung des Siliciumsubnitrids über eine topotaktische Reaktion im Sinne des Erhalts schichtartiger Strukturverbände aus dem Calciumdisilicid vereinbar. Im Transmissionselektronenmikroskop werden extrem dünne Schichten beobachtet. Beugungsexperimente (SAD) an partiell geordneten Bereichen zeigen den Erhalt eines hexagonalen Reflexmusters. Die topotaktische Reaktion wurde auch mit CaSi2-Einkristallen durchgeführt, deren Oberflächen durch Aufsublimation mit Ammoniumbromid belegt waren. Die Untersuchung von Edukt und Produkt mit Röntgen-Einkristallbeugungsmethoden deutet auf weitgehende Erhaltung einer periodischen Struktur während der topotaktischen Reaktion hin. Die neue Verbindung kann durch elektrisch neutrale Schichtpakete von etwa 650 pm beschrieben werden.

This thesis is comprised of three parts. The first part concerns the investigation of the topochemical reduction of LaSrNiRuO₆ in order to prepare LaSrNiRuO₄ via anion deintercalation. The second part discusses the oxide-for-hydride anion exchanges performed in SrV_1-xTi_xO₃, and the resulting SrV_1-xTi_xO_2-yH_1+y reduction products. Finally, the results from redox-neutral topochemical cation exchange reactions conducted in the three-dimensional perovskite structure of NaTaO₃ are presented along with the characterisation of a novel product of composition Ni_0.5TaO₃. The topochemical reduction of LaSrNiRuO₆ using CaH2 was carried out to produce a novel extended oxide phase with composition LaSrNiRuO₄. This phase is composed of sheets of apex-linked Ni¹⁺O₄ and Ru²⁺O₄ squares in a checkerboard ordered arrangement. To the best of our knowledge, this material is the first example of a B-cation ordered infinite-layer oxide phase. The low oxidation states of the transition-metal cations are confirmed by DFT calculations from which a spin moment S = &frac12; is determined for the nickel while the ruthenium centres adopt an intermediate-spin S = 1 configuration. LaSrNiRuO4 behaves paramagnetically at room temperature. However, upon cooling (T &LT; 250 K) a phase transition is observed in which the nickel spins interact ferromagnetically, while the ruthenium cations appear to undergo a change in spin configuration to a diamagnetic spin state. A possible explanation is given for this observation based on an ordered arrangement of local Jahn-Teller distortions. While investigating the preparation of LaSrNiRuO₄, it was observed that different samples of the LaSrNiRuO₆ starting materials exhibited markedly different reactivity. The observed differing reactivity is inconsistent with the crystal structure and composition of the LaSrNiRuO₆ samples, from which all the materials are identical. Careful investigation of the X-ray diffraction data collected from the LaSrNiRuO6 materials revealed that the reactivity of the samples is a consequence of the microstructure. By quenching or slow-cooling the materials during their synthesis, the size of the crystalline domains formed is affected and this in turn is observed to define the extent to which the topochemical deintercalation of oxide anions takes place. A mechanism to explain this effect is presented in which the greater 'plasticity' of small crystalline domains helps to limit the influence of lattice strain during the reaction. Similar with the observations for the LaSrNiRuO₆ phases, it was found that the reactivity of SrV_0.95Ti_0.05O₃ samples towards topochemical oxide-for-hydride exchange is also determined by the characteristics of the starting materials. The cooling rate can lead to phase segregation in SrV_0.95Ti_0.05O₃ samples which in turn affects the reduction behaviour. A modification of the energy profile for the oxide-for-hydride exchange in SrV_1-xTi_xO₃ phases is proposed on the basis of the electronic configuration that the transition-metal cations adopt upon reduction (d²,V³⁺ and d¹,Ti³⁺). Finally, topochemical exchange reactions can also be carried out between cations in complex transition metal oxides when the mobility of the species to be exchanged is sufficiently greater with respect to the host lattice. The preparation of Ni_0.5TaO₃ from exchange of Na⁺ by Ni²⁺ in NaTaO3 represents a synthetic approach not yet widely explored in the long-standing challenge that the preparation of magnetoelectric multiferroic materials represents. The topochemical reactions studied in this work highlight the possibility of directing and modifying the product phases, by tuning features of the reagents. This is in contrast with the limited control available in thermodynamic processes.

Im Rahmen der vorliegenden Arbeit ist es gelungen, ein Siliciumsubnitrid der Summenformel Si2N chemisch und strukturell zu charakterisieren. Die braune, metastabile Verbindung ist über eine topotaktische Reaktion aus der Zintl-Phase Calciumdisilicid und Ammoniumbromid im molaren Verhältnis 1:2 zugänglich. Charakteristisch für das Siliciumsubnitrid sind eine schichtartige Morphologie und perfekte Spaltbarkeit parallel zu den Schichten. Thermoanalytische und röntgenografische Untersuchungen belegen, dass die Reaktion im Temperaturbereich 165°- 350°C abläuft. Beim Einsatz pulverförmiger Eduktmischungen wird das Siliciumsubnitrid als röntgenamorphes Pulver im mikroskopischen Gemisch mit kristallinem CaBr2 erhalten. Die Abtrennung des Calciumbromids gelingt bisher nur unvollständig. Mit spektroskopischen Methoden (NMR, IR und Raman) konnten die Baueinheiten im Siliciumsubnitrid identifiziert werden. Stickstoff liegt im Siliciumsubnitrid in Form einer trigonalen [N(Si)3]-Koordination vor. Über REDOR-NMR-Experimente wurde belegt, dass das Subnitrid zwei unterschiedliche Siliciumspezies, eine [Si(Si4)]- und eine [Si(Si,N3)]-Umgebung, enthält. Ergebnisse elektronenmikroskopischer Untersuchungen (REM, TEM) sind mit der Bildung des Siliciumsubnitrids über eine topotaktische Reaktion im Sinne des Erhalts schichtartiger Strukturverbände aus dem Calciumdisilicid vereinbar. Im Transmissionselektronenmikroskop werden extrem dünne Schichten beobachtet. Beugungsexperimente (SAD) an partiell geordneten Bereichen zeigen den Erhalt eines hexagonalen Reflexmusters. Die topotaktische Reaktion wurde auch mit CaSi2-Einkristallen durchgeführt, deren Oberflächen durch Aufsublimation mit Ammoniumbromid belegt waren. Die Untersuchung von Edukt und Produkt mit Röntgen-Einkristallbeugungsmethoden deutet auf weitgehende Erhaltung einer periodischen Struktur während der topotaktischen Reaktion hin. Die neue Verbindung kann durch elektrisch neutrale Schichtpakete von etwa 650 pm beschrieben werden.

Maßgeschneiderte Eigenschaften von Funktionsmaterialien sind ein fundamentaler Aspekt für die Technologien unserer Gesellschaft und deren Weiterentwicklung. In diesem Zusammenhang bilden die Modifizierung bestehender Synthesestrategien und die Entwicklung neuartiger Synthesewege die grundlegende Voraussetzung für Innovation. Der Zugang zu den benötigten Materialien wird in den bis dato angewandten Synthesemethoden häufig durch die thermodynamische Stabilität einer Verbindung begrenzt. Zielstellung der vorliegenden Arbeit ist es, eine Strategie zur postsynthetischen Umwandlung und Modifizierung bereits vorhandener komplex strukturierter Feststoffe anzuwenden, durch die es gelingt, Zugang zu weiteren Materialien zu erhalten. Als Feststoffprekursoren wurden hierfür verschiedene ternäre und quaternäre, bismutreiche Subhalogenide gewählt, die bei niedrigen Temperaturen bis 70 °C mit dem Reduktionsmittel n-Butyllithium (nBuLi) zur Reaktion gebracht wurden, um diese in topochemischen Reaktionen zu neuen intermetallischen Phasen umzuwandeln. Die Bismutsubiodide Bi12Ni4I3, Bi8Ni8SI2 und Bi28Ni25I5 enthalten intermetallische Stäbe, deren Querschnitte nur vier bis elf Atome umfassen, was effektiven Durchmessern von ca. 0,8 bis 1,2 nm entspricht. Zudem befinden sich Iodidionen in den Kristallstrukturen, die die metallischen Stäbe voneinander separieren. Die reduktiven Behandlungen dieser Feststoffprekursoren führten jeweils zur quantitativen Deinterkalation der Iodidionen und dadurch zur Zusammenlagerung der metallischen Stäbe zu kompakten Stabpackungen. In Pseudomorphosen wurden zum einen Kristalle erhalten, die eine Vielzahl parallel angeordneter Bi3Ni-Faserbündel enthielten, zum anderen bildeten sich die bisher unbekannten, kristallinen Phasen Bi8Ni8S und Bi28Ni25. Während bei den Umwandlungen die strukturellen Charakteristiken der intermetallischen Teilstrukturen der Bismutsubiodide auf die reduzierten Phasen übertragen werden, ändern sich die elektronischen Situationen mit der Variation der Elektronenzahl. Dies lässt sich besonders gut am Beispiel der Umwandlung des Bismutsubiodids Bi28Ni25I5 in die reduzierte Phase Bi28Ni25 verdeutlichen. Die elektronische Struktur ändert sich durch die Reduktion kaum, sodass die zusätzlichen Elektronen im intermetallischen Teil antibindende Zustände füllen. Das intermetallische Bindungssystem verhält sich dabei wie ein strukturell rigides Elektronenreservoir und toleriert die Änderung der Elektronenzahl bei der topochemischen Umwandlung zu Bi28Ni25. Mit der elektronisch ungünstigen Situation geht die Metastabilität der reduzierten intermetallischen Phase einher. Die reduktive Behandlung des Bismutsubiodids Bi13Pt3I7 führte nicht nur zur selektiven topochemischen Deinterkalation von Iodidionen sondern zusätzlich zum Ausbau von Bismutatomen, wodurch die in Bi13Pt3I7 vorhandenen Iodidobismutatschichten in Iodidschichten umgewandelt werden. Die intermetallischen Schichten der Ausgangsverbindung bleiben erhalten und nähern sich an, sodass das bis dato unbekannte Bismutsubiodid Bi12Pt3I5 resultiert. Das topotaktische Fortbestehen der intermetallischen Schichten zeigt sich dabei an intermediär gebildeten Kompositkristallen aus Mutter- und Tochterverbindung. Durch den Abbau der isolierenden Iodidobismutat¬schichten erfolgen die elektronische Kopplung der intermetallischen Schichten und der Übergang des zwei-dimensionalen Metalls Bi13Pt3I7 in das dreidimensionale Metall Bi12Pt3I5. Die topochemische Reaktion wird durch eine Reaktionstemperatur von 45 °C limitiert: Bei erhöhter Reaktionstemperatur bis 70 °C tritt eine Umstrukturierung unter weiterem Iod- und Bismutausbau auf, und die metastabile, binäre Phase Bi2Pt(hP9) wird aufgebaut. Die dichte Kristallstruktur des erstmals dargestellten Bismutsubchlorids Bi12Rh3Cl2 baut sich aus einem intermetallischen [Bi4Rh]-Netzwerk auf, in dessen Kanäle Chloridionen eingeschlossen sind. Im Zuge der Niedertemperaturreaktion mit nBuLi erfolgt ein unerwarteter quantitativer Austausch der Chloridionen gegen Bismutatome, der die Kristalle des Subchlorids in Kristalle der binären Verbindung Bi14Rh3 überführt. Die kristallchemische Analyse zeigte, dass den [RhBi8/2]-Antiprismen des [Bi4Rh]-Netzwerks die Funktion von Scharnieren zukommt, welche eine Aufweitung des intermetallischen Netzwerks ermöglichen. So entstehen breite Diffusionspfade, und es resultiert ein dreidimensionales Transportsystem für den enormen Massetransport durch den Kristall. Bei der Austauschreaktion werden die zuvor unabhängig voneinander leitenden intermetallischen Stränge kantenverknüpfter [RhBi8/2]-Würfel elektrisch kontaktiert. Die physikalischen Eigenschaften ändern sich dabei maßgeblich: Aus dem eingeschränkten Metall Bi12Rh3Cl2 entsteht der metastabile Supraleiter Bi14Rh3. Mit zunehmender Kenntnis über die Strategien zur postsynthetischen Umwandlung und Modifizierung komplexer Strukturen können diese grundsätzlich dazu beitragen, Materialien mit technologisch relevanten Eigenschaften darzustellen. Insbesondere Phasen, die nur bei hohen Temperaturen thermodynamische Stabilität erlangen oder sogar unter allen Bedingungen metastabil vorliegen, werden durch die geschickte Wahl der Synthesestrategie zugänglich. Möglicherweise werden mit dem wachsenden Wissen zu neuartigen Synthesestrategien die chemischen und physikalischen Eigenschaften eines Materials auf diesem Weg gezielt veränderbar. Insbesondere die herausragenden Stabilitäten der nanoskaligen, intermetallischen Stäbe werfen zudem die Frage auf, ob diese durch die Reaktion mit oberflächenaktiven Reagenzien vereinzelt werden können, um neuartige nanoskalige Leiter herzustellen.