京都大学 / 新制・課程博士 / 博士(エネルギー科学) / 甲第24922号 / エネ博第464号 / 新制||エネ||87(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 河本 晴雄, 教授 亀田 貴之, 准教授 南 英治 / 学位規則第4条第1項該当 / Doctor of Energy Science / Kyoto University / DFAM

Anthropogenic greenhouse gas emissions (GHG) contribute to global warming, and must be mitigated. With GHG mitigation as an overarching goal, this research aims to study the potential for newfound and abundant sources of natural gas to play a role as part of a GHG mitigation strategy. However, recent work suggests that methane leakage in the current natural gas system may inhibit end-use natural gas as a robust mitigation strategy, but that natural gas as a feedstock for other forms of energy, such as electricity generation or liquid fuels, may support natural-gas based mitigation efforts. Flaring of uneconomic natural gas, or outright loss of natural gas to the atmosphere results in greenhouse gas emissions that could be avoided and which today are very large in aggregate. A central part of this study is to look at a new technology for converting natural gas into methanol at a unit scale that is matched to the size of individual natural gas wells. The goal is to convert stranded or otherwise flared natural gas into a commercially valuable product and thereby avoid any unnecessary emission to the atmosphere. A major part of this study is to contribute to the development of a novel approach for converting natural gas into methanol and to assess the environmental impact (for better or for worse) of this new technology. This Ph.D. research contributes to the development of such a system and provides a comprehensive techno-economic and environmental assessment of this technology. Recognizing the distributed nature of methane leakage associated with the natural gas system, this work is also intended to advance previous research at the Lenfest Center for Sustainable Energy that aims to show that small, modular energy systems can be made economic. This thesis contributes to and analyzes the development of a small-scale gas-to-liquids (GTL) system aimed at addressing flared natural gas from gas and oil wells. This thesis includes system engineering around a design that converts natural gas to synthesis gas (syngas) in a reciprocating internal combustion engine and then converts the syngas into methanol in a small-scale reactor. With methanol as the product, this research aims to show that such a system can not only address current and future natural gas flaring regulation, but eventually can compete economically with historically large-scale, centralized methanol production infrastructure. If successful, such systems could contribute to a shift away from large, multi-billion dollar capital cost chemical plants towards smaller systems with shorter lifetimes that may decrease the time to transition to more sustainable forms of energy and chemical conversion technologies. This research also quantifies the potential for such a system to contribute to mitigating GHG emissions, not only by addressing flared gas in the near-term, but also supporting future natural gas infrastructure ideas that may help to redefine the way the current natural gas pipeline system is used. The introduction of new, small-scale, distributed energy and chemical conversion systems located closer to the point of extraction may contribute to reducing methane leakage throughout the natural gas distribution system by reducing the reliance and risks associated with the aging natural gas pipeline infrastructure. The outcome of this thesis will result in several areas for future work. From an economic perspective, factors that contribute to overall system cost, such as operation and maintenance (O&M) and capital cost multiplier (referred to as the Lang Factor for large-scale petro-chemical plants), are not yet known for novel systems such as the technology presented here. From a technical perspective, commercialization of small-scale, distributed chemical conversion systems may create a demand for economical compression and air-separation technologies at this scale that do not currently exist. Further, new business cases may arise aimed at utilizing small, remote sources of methane, such as biogas from agricultural and municipal waste. Finally, while methanol was selected as the end-product for this thesis, future applications of this technology may consider methane conversion to hydrogen, ammonia, and ethylene for example, challenging the orthodoxy in the chemical industry that “bigger is better.”

The methylotrophic yeast Pichia pastoris is widely used for recombinant protein production due to its ability to grow to high cell densities as well as possession of an inducible methanol utilization pathway (MUT). The expression of genes encoding enzymes of the MUT pathway is very tightly regulated. These genes are turned on when methanol but not glucose is used as the sole carbon source. Thus, P. pastoris cells can be grown to high densities in glucose containing medium and expression of genes of MUT pathway can be turned on by changing the carbon source to methanol. This strategy is widely used for recombinant protein production wherein the gene of interest is cloned downstream of the methanol-inducible promoter of the gene encoding the first enzyme of the MUT pathway, alcohol oxidase I (AOXI). Despite production of a large number of recombinant proteins using the AOXI promoter, the mechanism of transcriptional activation of AOXI is not very well understood. It is only recently that a zinc finger protein known as Mxr1p (methanol expression regulator 1) was shown to play a key role in the regulation of AOXI as well as other genes of methanol utilization pathway (1) P. pastoris strains that do not express Mxr1p (mxr1) are unable to grow on peroxisomal substrates such as methanol and oleic acid. Methanol-inducible expression of genes involved in MUT pathway as well as those involved in peroxisome biogenesis (peroxins,) is severely impaired in mxr1 strains. While Mxr1p is constitutively expressed in cells cultured on glucose as well as methanol, it is cytosolic in glucose-grown cells, but nuclear in methanol-grown cells (1). The exact nucleotide sequence to which Mxr1p binds and regulates the expression of genes of MUT pathway is not known. The aim of this thesis is to map the Mxr1p binding sites in the promoters of methanol-inducible genes of P. pastoris. As a first step towards understanding the mechanism of transcriptional regulation of AOXI and other methanol inducible genes of P. pastoris by Mxr1p, the N-terminal region comprising of 150 amino acids, including the zinc finger DNA binding domain of Mxr1p was cloned into an E. coli expression vector and the recombinant protein was purified from E. coli cells. This recombinant protein (referred to as Mxr1p in this study) was used in an electrophoretic mobility shift assay (EMSA) to identify Mxr1p binding sites in the AOXI promoter. EMSA was carried out with sixteen different oligonucleotides spanning AOXI promoter region between -940 and -114 bp. Such studies led to the identification of six Mxr1p binding sites in AOXI promoter. Using a combination of DNase I footprinting as well as EMSA with chimeric double stranded oligonucleotides, the minimal Mxr1p binding site was identified as a 20 bp DNA sequence containing a core 5’CYCC 3’ sequence. Using methylation interference as well as extensive mutagenesis studies, nucleotides critical for Mxr1p binding were identified. Comparative analysis of Mxr1p binding sites identified in our study with the AOXI promoter deletion studies of Hartner et al (2) suggested that the Mxr1p binding sites identified in our study are likely to function as methanol-inducible enhancers in vivo, since deletion of AOXI promoter regions comprising Mxr1p binding sites results in a significant loss of methanol-inducible promoter activity. Thus, Mxr1p binding sites are likely to function as Mxr1p response elements (MXREs) in vivo. Mxr1p is considered to be the P. pastoris homologue of S. cerevisiae Adr1p (alcohol dehydrogenase II [ADH2] synthesis regulator). Adr1p is a key regulator of S. cerevisiae genes involved in the metabolism of glycerol, ethanol and oleic acid. The DNA binding domains of Adr1p and Mxr1p share 82% similarity and 70% identity. We therefore examined whether Mxr1p can bind to the Adr1p binding site of ADH2 promoter(ADH2-UAS1). Our studies indicate that Mxr1p does not bind to ADH2-UAS1. Interestingly, a single point mutation restores Mxr1p binding to ADH2-UAS1. Since Mxr1p is involved in the regulation of a number of genes including AOXI, we examined whether promoters of other Mxr1p-regulated genes also harbour MXREs similar to those identified in AOXI promoter. The promoters of genes encoding dihydroxyacetone synthase (DHAS) and peroxin 8 (PEX8) were chosen for this purpose. A detailed analysis of Mxr1p binding to these promoter sequences led to the conclusion that DHAS and PEX8 promoters also harbour Mxr1p binding sites similar to those of AOXI promoter. Based on these studies, we have derived a consensus sequence for Mxr1p binding. This study is the first report on detailed characterization of Mxr1p binding sites in three methanol-inducible promoters of P. pastoris and thus provides the molecular framework by which this transcription factor functions as a master regulator of genes involved in methanol utilization pathway of P. pastoris. Our study provides the blue print for mapping Mxr1p binding sites in the promoters of other Mxr1p-regulated genes.

Direct methanol fuel cells (DMFC) are being investigated as a portable energy conversion device for military and commercial applications. DMFCs offer the potential to efficiently extract electricity from a dense liquid fuel. However, improvements in materials properties and lowering the cost of the electrocatalysts used in a DMFC are necessary for commercialization of the technology. The cathode electrocatalyst is a critical issue in DMFC because the state-of-the-art catalyst, platinum, is very expensive and rare, and its performance is diminished by methanol that crosses over from the anode to the cathode through the Nafion membrane. This thesis investigates the addition of platinum to a palladium-cobalt nanoalloy electrocatalyst supported on carbon black in order to improve catalyst activity for the oxygen reduction reaction (ORR) and catalyst stability against dissolution in acidic environment without significantly reducing the methanol-tolerance of the catalyst. Platinum was added to the palladium-cobalt nanoalloy catalyst using two synthesis methods. In the first method, platinum was directly alloyed with palladium and cobalt using a polyol reduction method, followed by heat treatment in a reducing atmosphere to form catalysts with 11 and 22 atom % platinum. In the second method, platinum was added to a palladium-cobalt alloy by galvanic displacement reaction to form catalysts with 10 and 22 atom % platinum. The palladium cobalt alloy was synthesized using a polyol method, followed by heat treatment in a reducing atmosphere to alloy the nanoparticles before the Pt displacement. It was found that both methods significantly improve catalyst activity and stability, with the displaced catalysts showing a higher activity than the corresponding alloy catalyst. However the alloy catalysts showed similar resistance to dissolution as the displaced catalysts, and the alloyed catalysts were more tolerant to methanol. The displaced catalyst with 22 atom % platinum (8 wt. % Pt overall) performed similar to a 20 wt. % commercial platinum catalyst in both RDE and single cell DMFC tests. The 10 and 22 atom % Pt displaced catalysts and 22 atom % Pt alloyed all showed higher Pt mass specific activities than a commercial Pt catalyst. / text

Die Verringerung von CO2-Emissionen und die Schaffung eines Kohlenstoffkreislaufs sind Gegenstand vieler Forschungsarbeiten. Mittels Absorption wird CO2 in einem Lösungsmittel (vorrangig Alkoholamine) aus verschiedenen Gasmischungen abgetrennt und anschließend zu Wertstoffen umgesetzt. In der vorliegenden Arbeit war ein kombiniertes Verfahren aus CO2-Absorption und direktem Umsatz von CO2 zu Methanol im Amin zu untersuchen. Die Einsparung des Desorptionsschritts und die Möglichkeit zur Reduzierung der Reaktionstemperatur des exothermen Hydrierprozesses sind wesentliche Vorteile dieser Verfahrensweise. Um dieses Ziel zu erreichen, wurde die Performance verschiedener Amine untersucht und die Reaktionsparameter optimiert. Zur Verhinderung auftretender Nebenreaktionen fällt der Suche nach einem neuartigen Feststoffkatalysator besondere Bedeutung zu. Hierbei konnte ein Katalysatorsystem identifiziert werden, mit dessen Einsatz eine deutlich höhere Methanolausbeute als mit dem Standardkatalysator sowie eine Einschränkung der Nebenreaktionen erreicht wurde.

The carbonylation of methanol to acetic acid is one of the most important applications in homogeneous catalysis. The first chapter comprises a review on the mechanistic studies into the catalytic cycle of the ‘Monsanto process’ and includes some of the most prominent studies into the use of phosphines in the rhodium-catalysed carbonylation of methanol. The second chapter of this thesis reports on an investigation into the application of rhodium complexes containing several C4 bridged diphosphines, namely BINAP, dppb, dppx and dcpb as catalysts for hydrogen tolerant methanol carbonylation. An investigation into the structure, reactivity and stability of pre-catalysts and catalyst resting states of these complexes has also been carried out. The origin of this hydrogen tolerance is explained based on the differing reactivities of the Rh acetyls with hydrogen gas, and by considering the structure of the complexes. In the third chapter I report on an investigation into how electronic properties and coordination mode affect the elimination of phosphonium salts from rhodium complexes. The stability of a range of monodentate, bidentate and tridentate rhodium-phosphine complexes was tested. I also report on the formation of a novel bidentate complex containing a partially quaternised TRIPHOS ligand and investigate the mechanism of formation using 13CH3I. Strong evidence is also presented supporting a dissociative mechanism as the means of phosphine loss from the rhodium centre. In the final chapters I report an investigation into the stability of rhodium-aminophosphine ligand complexes and into increasing the solubility of potential rhodium pre-catalysts through the use of amine-containing phosphine ligands.

The demand for higher efficiency and cleaner power sources increases daily. The Direct Methanol Fuel Cells (DMFC) is one of those power sources that produces reliable electrical energy at high efficiencies and very low pollution levels. Remote telecommunication sites need power sources that can deliver reliable power. This dissertation informs the reader about the working principles of the DMFC and the materials it consists of. A good amount of theoretical background is also given on the DMFC, especially on the Membrane Electrode Assembly (MEA). Different membranes as well as their properties are discussed. Results from other researchers on DMFCs are also captured. A DMFC stack including a test rig, was built. The DMFC stack consisted of five single DMFC cells. Each cell contained an MEA, Gas Diffusion Layers (GDLS), highly corrosive resistant metal support grids, bipolar flow field plates and end plates. The DMFC stack was operated and tested in a test rig. The test rig held the air blower which supplied the cathode with the required oxidant (air), and the methanol solution tank plus its liquid pump. The liquid pump circulated the methanol solution through the anode side of the stack. It was observed that the DMFC is very susceptible to corrosion, especially if the methanol solution becomes conductive owing to solubility of C02 in it. Methanol itself is a corrosive substance. However the results obtained from the experiments clearly indicate that the DMFC can be implemented as an electrical power source for telecommunications.

Die breite Anwendung von Modellsystemen, um heterogene katalytische Prozesse zu verstehen, basiert darauf, die Lücke der strukturellen Komplexität zu überbrücken zwischen heutigen technischen Katalysatoren, bestehend aus einem Metalloxid sowie einem darauf geträgerten Metall, sowie kristallinen Metallen und planaren Metall/Oxid-Systemen, welche dazu benutzt werden, Struktur-Reaktivitäts-Beziehungen mittels einer Fülle von Surface Science-Methoden zu untersuchen. In der vorliegenden Arbeit liegt das Hauptaugenmerk auf so genannten Vanadiumoxid-‚Monolagen’-Katalysatoren, die insbesondere für Oxidationsreaktionen von Methanol eingeführt wurden. Mittels eines ‚bottom-up’-Ansatzes wurden Silica-geträgerte Vanadiumoxid-Modellkatalysatoren untersucht. Durch Kombination einer Reihe experimenteller Techniken wurde die Oberfläche von Mo(112), die als Substrat für den Silica-Film diente, im Detail untersucht und die atomare Struktur des Silica-Films wurde ermittelt. Adsorption von Wasser und das Wachstum von Vanadiumoxid-Nanopartikeln auf dem Silica-Film und schließlich die Reaktivität von Vanadiumoxid/Silica-Systemen gegenüber Methanol wurden untersucht. Im Gegensatz zu früher vorgeschlagenen Modellen sollte eine Sauerstoff-induzierte p(2×3)-Überstruktur, die sich auf einer Mo(112)-Oberfläche ausbilded, angenommen werden als ein eindimensionales Oberflächenoxid, bei dem sich Mo=O-Gruppen bevorzugt entlang der [-1-11]-Richtung der Mo(112)-Oberfläche ausbilden. Monolagen-Silica-Filme, die auf Mo(112) gewachsen wurden, bestehen aus einem zweidimensionalen Netz von SiO4-Tetraedern. In Abhängigkeit der Bedingungen, unter denen der Film präpariert wurde, kann die Struktur durch zusätzlich auf dem Mo-Substrat adsorbierte Sauerstoff-Atome verändert werden. Die Defekt-Struktur schließt Antiphasen-Domänengrenzen ein, die durch eine Verschiebung um die halbe Gitterkonstante entlang der [-110]-Richtung gebildet werden, und eine geringe Dichte von Punkt-Defekten, die höchstwahrscheinlich Silizium-Fehlstellen darstellen. Wasser dissoziiert nicht auf dem Monolagen-Silica-Film. Eine Wasser-Struktur, die geordnet bezüglich des Silica-Films ist, wurde bei 140 K beobachtet, was der guten Übereinstimmung der Gitterkonstanten von Silica-Film und hexagonalem Eis geschuldet ist. Amorphe Lagen festen Wassers, die die Oberfläche bei 100 K homogen bedecken, wurden als reaktive Lagen für Vanadiumoxid-Partikel benutzt, um die ‚Nasschemie’ nachzubilden, wie sie in der Präparation technischer Katalysatoren zum Einsatz kommt. Die Ergebnisse verdeutlichen, dass die Eis-Lagen die Bildung von hydratisierten Vanadiumoxid-Nanopartikeln, welche teilweise von V=O und V-OH-Gruppen terminiert werden, begünstigen. Die Dehydratisierung geschieht oberhalb 500 K, wobei eine V-terminierte Oberfläche entsteht. Methanol dissoziiert auf dehydratisierten Vanadiumoxid-Partikeln, und Methoxy-Spezies sind auf der Oberfläche stabil bis 500 K, allerdings nur in der Gegenwart von V-Plätzen. Die Produktion von Formaldehyd, die bei etwa 550 K stattfindet, ist stark abhängig von der Struktur der Oberfläche der Vanadiumoxid-Partikel und weist ein Maximum bei einem spezifischen Verhältnis zwischen V- und V=O-Oberflächenplätzen auf. Die hier vorgestellten Ergebnisse könnten unser Verständnis von katalytischen Reaktionen auf molekularer Ebene bedeutend vorantreiben. / The widespread use of model systems for understanding the heterogeneous catalytic processes is based on bridging the structural complexity gap between present generation of supported metal and metal oxide technical catalysts and crystalline metal and planar metal/oxide systems, which are utilized to investigate structure-reactivity relationships by a large variety of surface science techniques. In this thesis, we focused on a concept of so-called ''monolayer'' vanadium oxide catalysts, which have been introduced particularly for methanol oxidation reactions. Following a bottom-up approach, silica supported vanadium oxide model catalysts were investigated. Combining a number of experimental techniques, the surface of Mo(112) used as a substrate for the silica films was characterized in detail and the atomic structure of the silica film was determined. Adsorption of water and growth of vanadium oxide nanoparticles on the silica films, and finally the reactivity of vanadium oxide/silica systems towards methanol were studied. In contrast to the previously suggested models, an oxygen induced p(2×3) superstructure formed on a Mo(112) surface should be considered as one dimensional surface oxide where Mo=O groups are formed preferentially along the [-1-11] direction of the Mo(112) surface. Monolayer silica films grown on Mo(112) surfaces are composed of two-dimensional network of SiO4 tetrahedra. Depending on the film preparation conditions, the structure can be altered by additional oxygen atoms adsorbed on the Mo substrate. The defect structure includes antiphase domain boundaries which form by a half-lattice shift along the [-110] direction and a low density of point defects, most probably silicon vacancies. Water does not dissociate on the monolayer silica film. An ordered structure of water with respect to silica film was observed at 140 K owing to good lattice matching between the silica film and hexagonal ice. Amorphous solid water layers homogenously covering the surface at 100 K were used as reactive layers for vanadium oxide particles in order to mimic ''wet chemistry'' used in preparation of technical catalysts. The results revealed that ice layer assisted the formation of hydrated vanadium oxide nanoparticles partially terminated by V=O and V-OH groups. The dehydration takes place above 500 K, thus exposing V-terminated surface. Methanol dissociates on dehydrated vanadium oxide particles and methoxy species are stable on the surface up to 500 K only in the presence of vanadium terminated surface sites. Formaldehyde production which takes place at ~550 K is strongly affected by the surface structure of the vanadium oxide particles and exhibits a maximum at specific ratio between V- and V=O sites on the surface. The results presented may have a strong impact on our understanding of the catalytic reactions at the molecular level.

Intracellular protein sorting plays a key role in the regulation of cellular metabolism, gene expression, signal transduction and a number of other cellular processes. Proteins targeted to specific cellular compartments contain organelle-specific targeting sequences which interact with various components of the import machinery that are often evolutionarily conserved. For example, proteins targeted to peroxisomes interact with specific receptor proteins through unique peroxisomal targeting signals (PTS) which results in their import into peroxisomal matrix or insertion into peroxisomal membrane. Peroxisomal protein import has been studied in a number of species and several conserved PTS and receptor proteins have been identified. In our study, we report the unexpected finding that cytochrome c (cyt c), which lacks a canonical PTS, is targeted to peroxisomes of the methylotrophic yeast, Pichia pastoris. This is a unique feature of P. pastoris and is not observed in other yeast species such as the conventional yeast, Saccharomyces cerevisiae or other methylotrophic yeasts such as Hansenula polymorpha. Using S. cerevisiae cyc1 null mutant strain as a surrogate model, we demonstrate that P. pastoris cytochrome c (PpCyt c) is targeted to S. cerevisiae peroxisomes indicating that peroxisomal targeting is a unique and inherent property of PpCyt c and the machinery required for this is conserved in S. cerevisiae as well. We further demonstrate that Ppcyt c targeted to the fatty acid-induced peroxisomes of S. cerevisiae is a hemoprotein with covalently attached heme suggesting that PpCyt c synthesized in cytosol is first targeted to mitochondria where heme is added to the apoprotein by cytochrome c heme lyase and the holoprotein is then re-targeted to peroxisomes through an unknown mechanism. Proteins imported into peroxisomes carry specific peroxisomal targeting signals (PTS) known as PTS1 and PTS2. PTS1 is a tripeptide sequence (SKL) at the carboxy terminus of peroxisomal matrix proteins. To investigate whether the carboxy terminus of PpCyt c contain PTS1 or PTS1-like sequences, we made GFP fusion proteins with PpCyt c carboxy terminal amino acids (GFP-ATK, GFP-LAKATK) and examined their ability to localize to peroxisomes. Neither of these two proteins is targeted to peroxisomes indicating that PTS1-like sequences are not involved in peroxisomal targeting of Ppcyt c. Two receptors known as Pex5 and Pex7 are known to be involved in peroxisomal protein import and we therefore examined PpCyt c import into peroxisomes of P. pastoris strains lacking pex5 and pex7. Peroxisomal import of PpCyt c is abolished in pex5 but not pex7 mutant strain indicating that PpCyt c is imported into peroxisomes by a pex5-dependent but PTS1independent pathway. Since we observed significant amino acid differences between PpCyt c and S. cerevisiae cytochrome c (ScCyt c) in their carboxy-and amino-termini, we interchanged these amino acids between PpCyt c and ScCyt c and examined their subcellular localization. Such studies revealed that swapping the N-terminal or C-terminal amino acids of PpCyt c with those of S. cerevisaie cytochrome c (ScCyt c) abolishes peroxisomal localization of PpCyt c. Thus, both N-and C-terminal amino acids of PpCyt c are essential for its import into peroxisomes. Interestingly, in a number of fungal species, the N-and C-terminal amino acid sequences of cytochrome c are identical to those of PpCyt c indicating that peroxisomal targeting of cytochrome c may be observed in other yeast species as well. S. cerevisiae cells expressing PpCyt c exhibit several unique biochemical properties. S. cerevisiae cells expressing PpCyt c grow more rapidly than those expressing ScCyt c when cultured on media containing oleic acid as the sole carbon source and uptake of C-oleic acid from the medium as well as its assimilation into neutral lipids is quantitatively higher in the former. Surprisingly, the phenotype of S. cerevisiae cells expressing PpCyt c is dramatically altered such that the kinetics of growth on fatty acid containing media as well as lipid profile appear to be identical to those of P. pastoris rather than S. cerevisiae. Thus peroxisomal targeting of cytochrome c dramatically alters the kinetics of growth of S. cerevisiae cells in fatty acid containing media as well as the lipid metabolism raising several interesting questions on the molecular mechanisms involved in the alteration of phenotype of S. cerevisiae. It is likely that peroxisomal targeting of cytochrome c results in quantitative as well as qualitative changes in fatty acid metabolism and this opens up new vistas for the bioconversion of fatty acids into value-added lipid products by metabolic engineering. Based on these studies, we propose a new role for cytochrome c in peroxisomal fatty acid metabolism. Our study demonstrates that evolutionarily conserved proteins such as cytochrome c can acquire unique, species-specific functions that may be of great physiological significance to that organism.

Aramide and arylene ether multiblock copolymers of (AB)n-type with various degrees of sulfonation have been prepared for use in direct methanol fuel cells. / Aramid- und Arylethersulfon-Multiblockcopolymere vom Typ (AB)n mit unterschiedlichem Sulfonierungsgrad wurden hergestellt und hinsichtlich ihrer Eignung als Polymerelektrolyte in der Direkt-Methanol-Brennstoffzelle getestet.