A catalytic route to hydrogen production via thermochemical water splitting is highly desirable because it directly converts thermal energy into stored chemical energy in the form of hydrogen and oxygen. Recently, the Davis group at Caltech reported an innovative low-temperature (max 850°C) catalytic cycle for thermochemical water splitting based on sodium and manganese oxides (Xu, Bhawe and Davis, PNAS, 2012). The key steps are thought to be hydrogen evolution from a Na₂CO₃/MnO mixture, and oxygen evolution by thermal reduction of solids formed by Na⁺ extraction from NaMnO₂. Our work is aimed at understanding the fundamental chemical processes involved in the catalytic cycle, especially the hydrogen evolution from water. In this project, efforts are made to understand the interactions between the key components (Na, O₂, CO₂, and water) in the hydrogen evolution steps on a well-defined MnO(100) single crystal surface, utilizing x-ray photoelectron spectroscopy (XPS), low energy electron diffraction (LEED) and temperature programmed desorption (TPD). While some of the behavior of the catalytic system is observed with the model system developed in this work, hydrogen is only produced from water in the presence of metallic sodium, in contrast to the proposal of Xu et al. that water splitting occurs from the reaction of water with a mixture of Na₂CO₃ and MnO. These differences are discussed in light of the different operating conditions for the catalytic system and the surface science model developed in this work. / Ph. D.

The direct epitaxial growth of multilayer BN by atomic layer deposition is of critical significance forfo two-dimensional device applications. X-ray photoelectron spectroscopy (XPS) and low energy electron diffraction (LEED) demonstrate layer-by-layer BN epitaxy on two different substrates. One substrate was a monolayer of RuO2(110) formed on a Ru(0001) substrate, the other was an atomically clean Ni(111) single crystal. Growth was accomplished atomic layer deposition (ALD) cycles of BCl3/NH3 at 600 K substrate temperature and subsequent annealing in ultrahigh vacuum (UHV). This yielded stoichiometric BN layers, and an average BN film thickness linearly proportional to the number of BCl3/NH3 cycles. The BN(0001)/RuO2(110) interface had negligible charge transfer or band bending as indicated by XPS and LEED data indicate a 30° rotation between the coincident BN and oxide lattices. The atomic layer epitaxy of BN on an oxide surface suggests new routes to the direct growth and integration of graphene and BN with industrially important substrates, including Si(100). XPS and LEED indicated epitaxial deposition of h-BN(0001) on the Ni(111) single crystal by ALD, and subsequent epitaxially aligned graphene was deposited by chemical vapor deposition (CVD) of ethylene at 1000 K. Direct multilayer, in situ growth of h-BN on magnetic substrates such as Ni is important for spintronic device applications. Solid-state electrolytes (SSEs) are of significant interest for their promise as lithium-ion conducting materials but are prone to degradation due to lithium carbonate formation on the surface upon exposure to atmosphere, adversely impacting Li ion conduction. In situ XPS monitored changes in the composition of the SSE Li garnet (Li6.5La3Zr1.5Ta0.5O12, LLZTaO) upon annealing in UHV and upon Ar+ ion sputtering. Trends in core level spectra demonstrate that binding energy (BE) calibration of the Li 1s at 56.4 eV, yields a more consistent interpretation of results than the more commonly used standard of the adventitious C 1s at 284.8 eV. Annealing one ambient-exposed sample to >1000 K in UHV effectively reduced surface carbonate and oxygen, leaving significant amounts of carbon in lower oxidation states. A second ambient-exposed sample was subjected to 3 keV Ar+ ion sputtering at 500 K in UHV, which eliminated all surface carbon, and reduced the O 1s intensity and BE. These methods present alternative approaches to lithium carbonate removal than heating or polishing in inert atmospheres and are compatible with fundamental surface science studies. In particular, the data show that sputtering at mildly elevated temperatures yields facile elimination of carbonate and other forms of surface carbon. This is in contrast to annealing in either UHV or in noble gas environments, which result in carbonate reduction, but with significant remnant coverages of other forms of carbon.

Die strukturelle Charakterisierung im Nano- und Mikrometerbereich ist unerlässlich, da die Materialstruktur eng mit den makroskopischen Eigenschaften verbunden ist. Die vierdimensionale Rastertransmissionselektronenmikroskopie (4DSTEM) ist für diese Charakterisierung gut geeignet. Bei 4DSTEM wird ein konvergenter Elektronenstrahl über eine elektronentransparente Probe gerastert, während ein pixelierter Detektor an jeder Rasterposition ein Transmissionselektronenbeugungsbild aufnimmt. Die Analyse der in diesen Bildern kodierten Information erlaubt die Kartierung der Kristallorientierung, Kristallinität und weiteren Probeneigenschaften. 4DSTEM wird typischerweise in Rastertransmissionselektronenmikroskopen (STEMs) mit Elektronenergien von 60 keV bis 300 keV eingesetzt. Rasterelektronenmikroskope (SEMs) werden hingegen meist für die Oberflächenanalyse verwendet und nutzen Elektronenenergien bis 30 keV. 4DSTEM kann auch in SEMs integriert werden, indem ein pixelierter Detektor unter der elektronentransparenten Probe platziert wird. Während 4DSTEM in STEM eine sub-Ångström-Auflösung und die Durchstrahlung dickerer Proben ermöglicht, stellt 4DSTEM-in-SEM eine preiswertere und verfügbarere Alternative mit größerem Rastersichtfeld dar. Wir haben 4DSTEM-in-SEM in einem regulären SEM mithilfe einer szintillatorbasierten fasergekoppelten Kamera und einem Timepix3 hybriden Pixeldetektor (HPD) realisiert. Deren Kombination mit motorisierten Probentischen ermöglichte uns die Probe relativ zum Elektronenstrahl des SEMs zu bewegen und zu kippen, was die Probenkontrolle in STEMs widerspiegelt. Zur Demonstration des Potenzials von 4DSTEM-in-SEM, haben wir ein komplettes TEM-Netzchen sowie eine C60/MoS₂-van-der-Waals-Heterostruktur kartiert. Außerdem ermöglichte uns der datengesteuerte Modus des Timepix3 HPD eine Datenaufnahme mit mehr als 500k Bildern pro Sekunde, was der Verweilzeit von ringförmig integrierenden STEM-Detektoren ähnelt während das gesamte Beugungsbild aufgenommen wird. / Nano- and micro-scale structural characterization is essential for understanding, predicting, and optimizing the properties of materials since the material's structure is closely linked to its macroscopic properties. Four dimensional scanning transmission electron microscopy (4DSTEM) is well suited for this characterization. In 4DSTEM, a convergent electron beam is rastered over an electron-transparent sample, while a pixelated detector records a transmission diffraction pattern at each raster position. The encoded structural information in these patterns can be analyzed to map sample properties like crystal orientation, grain size, crystallinity, and more. 4DSTEM is typically employed in scanning transmission electron microscopes (STEMs) operating at electron energies of 60 keV to 300 keV. Scanning electron microscopes (SEMs) are widely used for surface morphology analysis operating at electron energies up to 30 keV. 4DSTEM can also be implemented in SEMs by placing a pixelated detector beneath the electron-transparent sample. While 4DSTEM in STEM achieves sub-Ångström spatial resolution and transmits through thicker samples, 4DSTEM-in-SEM offers a cost-effective and accessible alternative with a larger scan field of view while still transmitting through samples tens of nanometers thick. We implemented 4DSTEM-in-SEM using a standard SEM equipped with a custom scintillator-based fiber-coupled camera and a Timepix3 hybrid pixel detector (HPD). The combination of these detectors with motorized sample stages allowed us to move and tilt the sample relative to the SEM's electron beam mirroring the sample control in STEM. To demonstrate the potential of 4DSTEM-in-SEM, we mapped an entire TEM grid and analyzed a C60/MoS₂ van der Waals heterostructure. Additionally, the Timepix3 HPD's data-driven mode enabled acquisition speeds exceeding 500k frames per second, achieving dwell times comparable to annular integrating STEM detectors while recording the full diffraction patterns.

This work investigates the initial growth of iron and iron oxides on Ag(001). Surface structure and morphology of both post deposition annealed Fe films (in UHV and O2 atmosphere) as well as reactive grown iron oxide films will be analyzed in detail by low energy electron diffraction (LEED) and scanning tunneling microscopy (STM). The stoichiometry at the surface of the iron oxide films will be determined by X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES). The main focus of this work is to shed light on the question whether the growth of iron oxide films on Ag(001) is accompanied by the formation of strain reducing dislocation networks, or superstructures as found for other metal substrates in former studies. Here, we will distinguish between Fe films which were post deposition annealed in a thin O2 atmosphere and reactively grown iron oxide films.

In the presented thesis thin praseodymia films on non-passivated and passivated Si(111) substrates were investigated. The first part deals with PDA of praseodymia films with fluorite structure under UHV conditions in the temperature region from RT to 600°C. Here, a sophisticated model of the annealing process of praseodymia films is established. This is done by detailed analysis of XRD measurements using the kinematic diffraction theory in combination with the analysis of GIXRD, XRR and SPA-LEED measurements. It is shown that the untreated films, which are oxidized in 1 atm oxygen to obtain fluorite structure, do not exhibit pure PrO2 stoichiometry as it was assumed before. Instead, they decompose into two laterally coexisting species exhibiting a PrO2 and a Pr6O11. oxide phase, respectively. These species are laterally pinned to the lattice parameter of bulk Pr6O11. Homogeneous oxide films with Pr6O11 phase can be observed after annealing at 100°C and 150°C. Here, lateral strain caused by the pinning of the species is minimized and an increase of the crystallite sizes is determined. If higher annealing temperatures are applied, the film decomposes again into two coexisting species. Finally, after annealing at 300°C, a mixed crystalline film with both Pr2O3 and Pr2O3+Delta oxide phases is formed, where Delta denotes a considerable excess of oxygen within the sesquioxide phase. Again the lateral strain increases due to the tendency of praseodymia phases to increase their lattice parameters during oxygen loss combined with the lateral pinning. This is accompanied by a decrease of crystallite sizes, which are afterwards comparable to those of the untreated films. Further annealing at temperatures above 300°C does not significantly change the structure of the oxide film. However, the increase of the amorphous Pr-silicate interface between Si substrate and oxide at the expense of the crystalline oxide can be observed after annealing at higher temperatures. Furthermore, an increased mosaic spread of the crystallites occurs, which reduces the lateral strain caused by the oxygen loss. Nevertheless, the crystalline structure is stable against further annealing up to temperatures of 600°C. Transportation of the sample under ambient conditions after annealing at 200°C and 300°C leads to the formation of an additional crystalline structure at the surface which cannot be allocated to any praseodymia phase and may be explained by the contamination of the topmost crystalline layers with Pr-hydroxides. The results obtained from praseodymia films annealed in 1 atm nitrogen show that these films are good candidates to form homogeneous oxide films with pure cub-Pr2O3 structure by subsequent annealing in UHV. Here, a single oxide species is already observed after annealing at 300°C by SPA-LEED measurements which is in contrast to praseodymia films with fluorite structure where higher annealing temperatures (600°C) are necessary. In this case, negative effects like interface growth or increased defect density (mosaics, grain boundaries) can be minimized. Investigations on oxygen plasma-treated praseodymia films to obtain pure PrO2 stoichiometry are presented in the second part. Oxygen plasma-treated samples are compared with samples oxidized in 1 atm oxygen regarding the structure of the crystalline film. For this purpose, XRR and XRD measurements are performed to get structural information of the oxide film, which can be used to identify the corresponding oxide phases. Here, significantly smaller lattice constants of the crystalline oxide species can be observed after plasma treatment, which points to the incorporation of additional oxygen atoms. This verifies former studies, where a higher oxidation state of the oxide film was found by XPS measurements and it shows that plasma-treated films exhibit a higher oxidation state than films oxidized in 1 atm oxygen due to the availability of reactive atomic oxygen in the plasma. Furthermore, the Pr-silicate interface between crystalline film and Si substrate is not increased during plasma treatment. In the last part of the presented thesis, first results from the epitaxy of praseodymia films on Cl-passivated Si substrates are shown. The aim is to suppress the Pr-silicate formation during the growth process. Thus, praseodymia films are grown on passivated and non-passivated substrates to compare the crystallinity of both samples using XSW and LEED measurements. The structure of the oxide films on Cl-passivated Si is determined afterwards by XRR. It is shown that crystalline films with cub-Pr2O3 structure and several nanometer thickness can be successfully grown on Cl-passivated substrates. Here, the Pr-silicate interface layer are restricted to a single mono-layer. In contrast, the films grown on non-passivated substrates are completely amorphous containing Pr-silicates and Pr-silicides.

No description available.

Transmission Electron Microscopy (TEM) on light element materials and soft matters is problematic due to electron irradiation damage and low contrast. In this doctoral thesis techniques were developed to address some of those issues and successfully characterize these materials at high resolution. These techniques were demonstrated on graphene flakes, DNA/magnetic beads and a number of water containing biomaterials. The details of these studies are given below. A TEM based method was presented for thickness characterization of graphene flakes. For the thickness characterization, the dynamical theory of electron diffraction is used to obtain an analytical expression for the intensity of the transmitted electron beam as a function of thickness. From JEMS simulations (experiments) the absorption constant λ in a low symmetry orientation was found to be ~ 208 nm (225 ± 9 nm). When compared to standard techniques for thickness determination of graphene/graphite, the method has the advantage of being relatively simple, fast and requiring only the acquisition of bright-field (BF) images. Using the proposed method, it is possible to measure the thickness change due to one monolayer of graphene if the flake has uniform thickness over a larger area. A real-space TEM study on magnetic bead-DNA coil interaction was conducted and a statistical analysis of the number of beads attached to the DNA-coils was performed. The average number of beads per DNA coil was calculated around 6 and slightly above 2 for samples with 40 nm and 130 nm beads, respectively. These results are in good agreement with magnetic measurements. In addition, the TEM analysis supported an earlier hypothesis that 40 nm beads are preferably attached interior of the DNA-coils while 130 nm beads closer to the exterior of the coils. A focused ion-beam in-situ lift-out technique for hydrated biological specimens was developed for cryo-TEM. The technique was demonstrated on frozen Aspergillus niger spores which were frozen with liquid nitrogen to preserve their cellular structures. A thin lamella was prepared, lifted out and welded to a TEM grid. Once the lamella was thinned to electron transparency, the grid was cryogenically transferred to the TEM using a cryo-transfer bath. The structure of the cells was revealed by BF imaging. Also, a series of energy filtered images was acquired and C, N and Mn elemental maps were produced. Furthermore, 3 Å lattice fringes of the underlying Al support were successfully resolved by high resolution imaging, confirming that the technique has the potential to extract structural information down to the atomic scale. The experimental protocol is ready now to be employed on a large variety of samples e.g. soft/hard matter interfaces.

Gegenstand dieser Arbeit war: - die Optimierung und Modellierung des CVD-Wachstums von Fe-gefüllten CNTs aus Ferrocen, - die Auswahl geeigneter Schichtsysteme für das orientierte Wachstum Fe-gefüllter CNTs, - eine umfassende Charakterisierung der Nanostrukturen und deren Bezug zu den Wachstumsparametern, - die Formulierung eines allgemeingültigen Wachstumsmodels. Es wurde eine Anlage zur thermisch induzierten chemischen Gasphasenabscheidung bei Atmosphärendruck verwendet. Im Mittelpunkt der Syntheseexperimente standen Fe-gefüllte MWCNTs. Als Precursoren dienten Ferrocen und Cyclopentadienyl-eisen-dicarbonyl-dimer. Für die Darstellung von CNT-Ensembles mit idealerweise paralleler Ausrichtung der Einzelindividuen kamen thermisch oxidierte Si-Substrate (Schichtdicke des Oxid: 1 µm) zum Einsatz. Das Wachstum der CNTs wurde überwiegend als cokatalysierter Prozess durchgeführt, d.h. neben dem Fe aus dem Precursor dienten dünne Metallschichten (Fe, Co, oder Ni, Schichtdicke ≤ 10 nm), die auf den Substraten deponiert waren, als Katalysatorreservoir. Zunächst ging es darum den CVD-Prozess hinsichtlich tubularer CNTs mit senkrechter Vorzugsorientierung zur Substratoberfläche, einer guten Kristallinität der Hülle, sowie einem hohen Füllungsanteil der ferromagnetischen α-Fe-Phase zu überprüfen. Generell ließ sich die Abscheidung gefüllter CNTs für mittlere Substrattemperaturen im Bereich von 1013 – 1200 K durchführen. Die optimale Wachstumstemperatur lag bei ≈ 1103 K. Mit den beiden Precursoren - Ferrocen und Cyclopentadienyl-eisen-dicarbonyl-dimer ließen sich Fe-gefüllte CNTs in guter Qualität darstellen. Letztere Verbindung verringerte die Abscheidung von amorphem Kohlenstoff auf der CNT-Oberfläche, barg allerdings die Nachteile einer Sauerstoffkontamination und höherer Verdampfungs-temperaturen in sich. Aus der Vielzahl von Experimenten konnte abgeleitet werden, dass die Haupteinflussgrößen für den Innen- und Außendurchmesser der CNTs die Katalysatorschicht auf dem Substrat, die Synthesetemperatur und der Precursormassenstrom sind. Höhere Temperaturen und/oder ein Mehrangebot an Precursor äußerten sich stets in größeren Durchmessern. Zusätzliche Metallschichten auf den oxidierten Si-Substraten erlaubten eine gezielte Durchmesservariation. Beispielsweise zeigte sich an Substraten mit 2 nm Fe bzw. 2 nm Ni, dass sich die mittleren CNT-Außendurchmesser gegenüber dem auf unbeschichteten Substraten (34 nm) zu 44 nm bzw. 30 nm verändern lässt. Mit Al-Zwischenschichten konnten sogar Durchschnittswerte für den CNT-Außendurchmesser von 18 nm erzielt werden. Durch Röntgenstrukturuntersuchungen und Mössbaueranalysen an CNT-Ensembles wurde α-Fe als Hauptbestandteil der Füllung identifiziert. Auf den hohen Anteilen der α-Fe-Phase beruhte auch das magnetische Verhalten der Nanodrähte. Ein Beleg für die Schlüsselrolle des Systems Fe-C während des Wachstumsprozesses war die Phase Fe3C, mit orthorhombischer Struktur. Weniger häufig ließ sich γ-Fe nachweisen. Darüber hinaus konnten sämtliche CNT-Füllungen mittels SAED und HRTEM als Einkristalle charakterisiert werden. Die innerhalb der CNTs eingeschlossenen Fe- oder Fe-C-Nanodrähte wiesen außerdem keine kristallographische(n) Vorzugsrichtung(en) gegenüber den CNT-Wänden auf. Anhand der experimentellen Befunde war es möglich ein phänomenologisches Wachstumsmodell vorzuschlagen, welche eine Erweiterung des VLS-Mechanismus darstellt. Das in der vorliegenden Arbeit vorgestellte Modell greift das base-growth-Konzept auf und favorisiert die Akkumulation von Katalysatormaterial über die geöffneten Enden der CNTs. Eine genauere kinetische und thermodynamische Beschreibung war aufgrund der im Nanometerbereich nur schwer zugänglichen Stoffdaten nicht möglich.

It is an aim of this work to provide insight into the energetic influence on the ordering of molecular thin films on crystalline substrates. Here, the term substrate either refers to inorganic crystal surfaces or highly ordered layers of another organic molecular species. In order to calculate the total interface potential of extended molecular domains, a new calculation technique (GRID technique) is developed in the first part of this work. Compared to the standard approach, this method accelerates the potential calculation drastically (times 10000). The other parts of the thesis are dedicated to the comparison of experimental results (obtained by scanning tunneling microscopy and electron diffraction) to the optimal layer structure as predicted by optimization calculations. Potential calculations which are performed for the system perylenetetracarboxylicdianhydride (PTCDA) on graphite demonstrate that point-on-line coincident structures correspond to energetically favorable alignments of the molecular lattice with respect to the substrate lattice. The capability of the GRID technique to predict the optimal layer structure is demonstrated for the system peri-hexabenzocoronene (HBC) on graphite. The organic-organic heteroepitaxy system PTCDA on HBC on graphite is investigated in order to clarify to which extent the ordering mechanism there differs from that of the organic-inorganic heteroepitaxy system PTCDA on graphite. As a result of this investigation, a new type of epitaxy, i.e., substrate induced ordering is found. This new epitaxy type is governed by the inner structure of the substrate lattice unit cell. Here, the substrate surface is a layer of organic molecules itself, hence the substrate surface unit cell does indeed exhibit a complex inner structure. A generalized classification scheme for epitaxial growth incorporating this new type of epitaxy is proposed. In the last chapter, the structure of the first layers of titanylphthalocyanine (TiOPc) on Au(111) is investigated and compared to potential optimization calculations. The correspondence of experimental and theoretical results provides evidence that the GRID technique can, in principle, also be applied to molecular layers on metal surfaces. / Das Ziel der vorliegenden Arbeit ist es, Einblicke in die energetischen Einflüsse, die zur Ausbildung der Schichtstruktur organischer Moleküle auf kristallinen Substraten führen, zu geben. Diese Substrate sind entweder Oberflächen anorganische Kristalle oder selbst hochgeordnete Molekülschichten. Um das totale Grenzflächenpotential ausgedehnter Moleküldomänen berechnen zu können, wird im ersten Teil der Arbeit eine neue Berechnungsmethode (GRID Technik) vorgestellt. Im Vergleich mit herkömmlichen Berechnungsmethoden auf der Basis molekülmechanischer Kraftfelder ist diese neue Methode daher um ein Vielfaches schneller (Faktor 100000). Die folgenden Teile der Arbeit sind dem Vergleich experimenteller Ergebnisse (Rastertunnelmikroskopie und Elektronenbeugung) mit, durch Potentialoptimierungsrechnungen als energetisch günstig vorhergesagten, Schichtstrukturen gewidmet. So kann für das System Perylentetracarbonsäuredianhydrid (PTCDA) auf Graphit mittels Potentialberechnungen nachgewiesen werden, daß die experimentell gefundenen ?Point-on-line koinzidenten? Strukturen energetisch günstige Anordnungen des Molekülgitters bezüglich des Substratgitters darstellen. Die Eignung der neuen Berechnungsmethode zur Vorhersage der günstigsten Adsorbatgitterstruktur für ein gegebenes System aus Molekül und Substrat, wird anhand des Systems peri-Hexabenzocoronen (HBC) auf Graphit demonstriert. Das organisch-organische Heteroepitaxiesystem PTCDA auf HBC auf Graphit wird untersucht, um zu klären, inwieweit sich die dafür gültigen Ordnungsmechanismen von denen unterscheiden, die für das Wachstum des organisch-anorganischen Heteroepitaxiesystems PTCDA auf Graphit verantwortlich sind. Dabei gelingt es, eine bisher nicht klassifizierte Art von Epitaxie, d.h. substratinduzierter Ordnung, nachzuweisen. Dieser neue Epitaxietyp ist bedingt durch die innere Struktur einer Substrateinheitszelle - das Substrat ist ja hier selbst eine Schicht geordneter Moleküle, die natürlich eine innere Struktur aufweisen. Im folgenden wird ein verallgemeinertes Klassifizierungssystem für Epitaxietypen abgeleitet, welches den neuen Epitaxietyp beinhaltet. Im letzten Kapitel wird die Struktur von der ersten Lagen von Titanylphthalocyanin (TiOPc) auf Au(111) experimentell untersucht und mit entsprechenden Potentialoptimierungsrechnungen verglichen. Die Übereinstimmung von experimentellen und theoretischen Ergebnissen zeigt, daß die GRID Technik, zumindest prinzipiell, auch für Molekülschichten auf Metallsubstraten anwendbar ist.

Corrosion protection performance of a completely soluble polyaniline-dodecylbenzenesulfonic acid salt (PANI-DBSA) on C45 steel has been studied with electrochemical impedance and potentiodynamic measurements. Chloroform is the most suitable solvent to process the pristine PANI-DBSA because of negligible interaction of the solvent with the polyaniline (PANI) backbone. An anodic shift in the corrosion potential (<img src="http://scitation.aip.org/stockgif3/Dgr.gif" alt="Delta" align="bottom" border="0"><i>E</i>=~70&nbsp;&nbsp;mV), a decrease in the corrosion current and a significant increase in the charge transfer resistance indicate a significant anti-corrosion performance of the soluble PANI deposited on the protected steel surface. Corrosion protection follows the mechanism of formation of a passive oxide layer on the surface of C45 steel. In situ UV-Vis spectroscopy was used to investigate the differences in permeability of aqueous anions into PANI-DBSA. Preliminary results of electron diffraction studies show that PANI-DBSA possesses an orthorhombic type of crystal structure. An increase in the feed ratio of DBSA to aniline increases the tendency of aggregation of spherical particles of PANI obvious in transmission electron microscopy. PANI-DBSA slowly loses its electrochemical activity in acid free electrolyte without undergoing degradation.