Techn. Universiẗat, Diss., 2005--Clausthal.

Zugl.: Braunschweig, Techn. Univ., Diss., 2009

Zugl.: München, Techn. Univ., Diss., 2008

Techn. Universiẗat, Diss., 2001--Chemnitz.

Nürnberg, Universiẗat, Diss., 2002--Erlangen.

Ziel der vorliegenden tierexperimentellen Studie war es, Unterschiede im Einheilverhalten der Werkstoffe Titan und VA-Stahl (316L) anhand der Matrixproteine Kollagen Typ I (C1), Kollagen Typ III (C3) und Fibronektin im implantatumgebenden Interface zu untersuchen und darzustellen. Hierzu wurden die Einheilkapseln der Implantate nach subkutaner, intramuskulärer und intraossärer Implantation nach den Bewertungskriterien Kapselqualität, Kapseldicke und Verteilungsmuster der Matrixproteine mittels konventioneller Mikroskopie und Konfokaler Laserscanning Mikroskopie (CLSM) analysiert. Nach subkutaner Implantation zeigten beide Werkstoffe in Übereinstimmung mit den Ergebnissen von SHANNON et al. (1997) vermehrt locker angeordnete, teils parallel orientierte Kollagenfasern mit erhöhtem Zellaufkommen an Fibroblasten und Makrophagen. Nach intramuskulärer Implantation jedoch fanden sich vorwiegend parallel angeordnete, teils dicht gepackte Kollagenfasern mit nur mäßig erhöhtem Zellaufkommen. Intramuskulär eingebrachte Implantate heilten in dünneren Kapseln ein, als subkutan eingebrachte Implantate. Es ergab sich keine Korrelation zu den ermittelten Kapselqualitäten. Dies erstaunt umso mehr, da unter der fortwährenden funktionellen Beanspruchung der intramuskulären Implantate im Bereich der Bauchmuskulatur gegenüber der statischeren Platzierung im subkutanen Rückenfett eine erhöhte Zell- und Matrixreaktion erwartet worden war. Im Lokalisationsvergleich zeigte sich intramuskulär für beide Werkstoffe ein erhöhtes Aufkommen an Fibronektin. Dies könnte auf die erhöhte Stoffwechselaktivität und funktionelle Belastung im Muskelgewebe zurückgeführt werden (ROSENGREN et al. 1994). Nach intraossärer Implantation konnten dünnere Kallusformationen für VA-Stahl gegenüber Titan in allen Proteinfluoreszenzen nachgewiesen werden. Die Qualität der Kallusformation und die histologische Kallusstruktur glichen sich mit zunehmender Implantationsdauer der regulären Knochenstruktur an. Die semiquantitativ beurteilte Verteilung der Matrixproteine mittels CLSM zeigte bei deutlichen Standardabweichungen für beide Werkstoffe erhöhte Fluoreszenz-Intensitäten nur in den implantatnahen Kapselanteilen. In den mittleren und den implantatfernen Kapselabschnitten waren für beide Werkstoffe inkonstant höhere Fluoreszenzwerte gegenüber den Vergleichskollektiven messbar. Der intraossäre Materialvergleich ergab implantatnahe und implantatferne Fluoreszenzmaxima für alle Matrixproteine, die mit zunehmender Implantationsdauer abfielen. Reproduzierbare, materialspezifische Unterschiede waren in Analogie zu BERGER-GORBET et al. (1996) nicht zu finden. In den mittleren Kallusabschnitten konnten reproduzierbare Fluoreszenzunterschiede nur bei Detektion von Kollagen Typ I (C1) in allen Zeitintervallen gesehen werden. Im Vergleich zur Literatur konnte die von VIROLAINEN et al. (1997) beschriebene biphasische Proteinanhäufung, wie auch ein wechselndes Proteinaufkommen (LINDHOLM et al. 1996) nach intraossärer Implantation nicht nachvollzogen werden. Ergänzende Beobachtungen der hier vorgestellten Studie verdeutlichen, dass die lokale, intraossäre Anreicherung von Matrixproteinen, unabhängig von Implantatinsertion oder gar Werkstoffeigenschaften, nach jeglicher Traumatisierung von Knochengewebe den knöchernen Reparationsprozess begleitet. Unter dem Aspekt der Restitutio ad Integrum von Knochenwunden können diese Beobachtungen auf das implantatnahe und das implantatferne Restitutionszentrum übertragen werden. Die Aktivität dieser Restitutionszentren hält bis zum Abschluss der knöchernen Remodellierung über 12 Wochen hinaus an. Dies deckt sich mit Aussagen von STEFLIK et al. (1998), wonach der periimplantäre Knochenumbau langfristig dynamisch bestehen bleibt. Um der zunehmenden Verbreitung nicht nur dentaler Implantate gerecht zu werden, muss auch zukünftig ein besseres Verständnis der Komunikationswege zwischen Implantaten und Biosystemen gewonnen werden. Dies bedeutet für die Herstellung und Weiterentwicklung von Implantaten, dass nicht nur die Werkstoff- und Oberflächenauswahl wichtig ist, sondern auch die funktionell erforderliche Oberflächenstrukturierung auf die gewünschte Wechselwirkung mit Bestandteilen der EZM und den Zellen angepasst sein sollte (THULL 2005). Die CLSM kann hierbei aufgrund der Möglichkeit der 3-dimensionalen in-situ-Darstellung des Implantatinterface biologisch-strukturelle und molekularbiologisch-immunologische Fragestellungen beantworten. / Aim of the study was to investigate the differences in woundhealing of extracellular matrixproteins in the surrounding tissue of metallic implants. Therefore the proteins collagen type I, type III and fibronectin were detected by confocal laserscanning microscopy after subcutaneous, intramuscular and enossal insertion of titanium and stainless-steel implants. Intramuscular periimplant capsules showed denser orientated kollagen fibres than subcutaneous capsules. No correlation could be shown between the capsuleformation and the implant material. After enossal implantation stainless-steel implants healed in thinner capsules than titanium implants. The semiquantitative evaluation of the proteindistribution by confocal laserscanning microscopy showed higher intensity closer to the implant-interface in all localizations and materials. Reproducable differences or correlations to the implantmetrial could not be seen. Better understanding and further investigations of material surface and tissue interactions are necessary to elaborate highly developed implants with perfectioned surfaces.

Die Mikrostruktur von Zirkonoxid–Aluminiumoxid Keramiken wurde im Rasterelektronenmikroskop (REM) untersucht und mittels quantitativer Bildanalyse weiter charakterisiert. Die so erhaltenen spezifischen morphologischen Kennwerte wurden mit denen, die an dreidimensionalen Modellstrukturen äquivalent gewonnen wurden, verglichen. Es wurden modifizierte Voronoistrukturen benutzt, um die beteiligten Phasen in repräsentativen Volumenelementen (RVE) auf Voxelbasis zu erzeugen. Poren wurden an den Ecken und Kanten dieser Strukturen nachträglich hinzugefüg. Nachdem alle relevanten Kennwerte der Modellstrukturen an die realen keramischen Mikrostrukturen angepasst wurden, musste das RVE für die Finite Element Simulationen (FES) geeignet vernetzt werden. Eine einfache Übernahme der Voxelstrukturen in hexaedrische Elemente führt zu sehr langen Rechenzeiten, und die erforderliche Genauigkeit der FES konnte nicht erreicht werden. Deshalb wurde zunächst eine adaptive Oberflächenvernetzung ausgehend von einem generally classed marching tetrahedra Algorithmus erzeugt. Dabei wurde besonderer Wert auf die Beibehaltung der zuvor angepassten Kennwerte gelegt. Um die Rechenzeiten zu verkürzen ohne die Genauigkeit der FES zu beeinträchtigen, wurden die Oberflächenvernetzungen dergestalt vereinfacht, dass eine hohe Auflösung an den Ecken und Kanten der Strukturen erhalten blieb, während sie an flachen Korngrenzen stark verringert wurde. Auf Basis dieser Oberflächenvernetzung wurden Volumenvernetzungen, inklusive der Abbildung der Korngrenzen durch Volumenelemente, erzeugt und für die FES benutzt. Dazu wurde ein FE-Modell zur Simulation der Impedanzspektren aufgestellt und validiert. Um das makroskopische elektrische Verhalten der polykristallinen Keramiken zu simulieren, mussten zunächst die elektrischen Eigenschaften der beteiligten Einzelphasen gemessen werden. Dazu wurde eine Anlage zur Impedanzspektroskopie bis 1000 °C aufgebaut und verwendet. Durch weitere Auswertung der experimentellen Daten unter besonderer Berücksichtigung der Korngrenzeffekte wurden die individuellen Phaseneigenschaften erhalten. Die Zusammensetzung der Mischkeramiken reichte von purem Zirkonoxid (3YSZ) bis zu purem Aluminiumoxid. Es wurde eine sehr gute Übereinstimmung zwischen den experimentellen und simulierten Werten bezüglich der betrachteten elektrischen, mechanischen und thermischen Eigenschaften erreicht. Die FES wurden verwendet, um die Einflüsse verschiedener mikrostruktureller Parameter, wie Porosität, Korngröße und Komposition, auf das makroskopische Materialverhalten näher zu untersuchen. / The microstructures of zirconia–alumina ceramics are investigated by scanning electron microscopy (SEM) and further characterised by quantitative image analysis. This leads to specific morphological parameters which are compared with the same parameters derived from three-dimensional model structures generated in voxel-based representative volume elements (RVE). Modified Voronoi clusters are employed to represent alumina and zirconia phases. Pores are added at the grain corners and edges respectively. After adjusting all the relevant morphological parameters of the model to the real ceramics’ microstructure, the RVE has to be meshed for finite element simulations (FES). Hexahedral elements which simply use the voxel structure did not lead to sufficient accuracy of the FES. As a first step, an adapted surface tessellation is generated, using a generally classed marching tetrahedra method. Special care is taken to preserve the topology as well as the individual volumes and interfaces of the model. In terms of processing time and accuracy of the FES it is very important to simplify the initially generated surface mesh in a manner that preserves detailed resolution at corners and along edges, while decimating the number of surface elements in flat regions, i.e. at the grain boundaries. From the surface mesh an adequate tetrahedral volume tessellation, including solid elements representing the grain boundaries, is created, which is used for the FES. Therefore, a FE-model for the simulation of impedance spectra has been established and validated. To simulate the macroscopic electrical behaviour of polycrystalline ceramics, the electrical properties of the individual constituting phases need to be measured. This is done by impedance spectroscopy up to 1000 °C. Further analysis of the experimental data with special respect to the effect of the grain boundaries is carried out to obtain the individual phases’ properties. The sample composition was varied from pure zirconia to pure alumina. A very good agreement between experimental and simulated data was achieved in terms of electrical, thermal and mechanical properties. The FES were employed to scrutinize the effects of systematically varying microstructural properties, such as porosity, grain size and composition, on the macroscopic material behaviour.

11 Conclusion 11.1 Glaze compositions Glazes from tiles of imposing Islamic buildings and some tableware glazes of the medieval epoch in Central Asia, the Middle East, Asia Minor, and North Africa are analysed regarding their main composition and colouring agents. Three major production recipes can be distinguished, i.e. alkali glazes, alkali lead glazes, and lead glazes. In the work of Tite (2011), Islamic glazes from Egypt, Iran, Iraq, and Syria are subdivided into four groups of composition, being partly consistent with those of this work. The alkali lime glazes with <2 wt% PbO correspond to the alkali glazes, but with higher content of CaO. The second and third group of low lead alkali and lead alkali glazes (2-10 wt% PbO and 10-35 wt% PbO) can be subsumed to the alkali lead group described here. Tite´s high lead group has PbO contents >35 wt% and is comparable to the lead glazes (>30 wt% PbO) of this study. The lead and the alkali oxides serve as a flux for the lowering the melting point. In the interaction of ceramic body and glaze, primarily an influence from Si, Al, and K is observed in the line scans from the cross section of ceramic and glaze. However, the input of ceramic material doesn’t seem to be critical for the classification of glazes according to their alkali and alkali lead compositions. In every epoch and locality, except of the Ilkhanate dynasty in Iran, lead glaze samples can be verified. This is also observed in previous investigations e.g. from medieval Iraq, Jordan and Iran (McCarthy, 1996; Al-Saad, 2002; Holakooei et al., 2014). In the Moroccan and Bulgarian glazes, lead seems to be the only important flux. In part, the lead flux is supplemented by additional alkali contents. The lack of alkali and alkali lead glazes in Bulgarian and Moroccan glazes (assuming that the Ottoman alkali lead glazes are imported tableware) seems to affect the regions with Roman-influenced history and with geographical distance to the Near East alkali flux tradition. For the alkali lead glazes and alkali glazes, the overall characteristic is sodium dominated, although the absolute soda values are in part surprisingly low. Samples from Bukhara, Takht-i-Suleiman and the Turkish localities have the highest, but still moderate Na2O values up to 15 wt%, compared to other analyses from e.g. India (Gill & Rehren, 2011). The source of the alkali flux is either mineral natron or plant ash. The source can be determined regarding the MgO values, limited to 1.3 wt% in mineral natron and exceeding 2.0 wt% in the case of plant ashes. In the samples of the present study, the K2O component is not suitable for the indication of the flux-relevant alkali source due to its broad scattering. The P2O5 contents are also enhanced in the plant ash compositions but the data set is not sufficient for statistical evaluation. An influence of the ceramic body on the glaze composition is observed only for SiO2, Al2O3, and K2O in quartz frit ceramics with slight K-feldspar content. The earliest Uzbek tableware glazes from the 10th-11th century (Seljuq period) were generally produced using a lead flux. The same applies to part of the Uzbek tile glazes which were produced between the 13th and 16th century. In Iran, glazes from the 12th century (Khwarezmid period) are lead glazes, but also alkali-fluxed glazes with mineral natron characteristics can be found. Although the production of lead-rich glazes was established from the 8th-9th century on in Iraq, Syria, and Egypt (Henshaw, 2010; Tite et al., 2011), alkali glazes are found in almost all regions except of Bulgaria and Morocco. Plant ash-fluxed alkali glazes are found in 13th century glazes from Takht-i-Suleiman. The plant ash flux technology is assumed to be continuously used in Mesopotamia, Iran, and Central Asia (Sayre & Smith, 1974; Henderson, 2009), but it could be shown that a parallel use of mineral natron parallel existed in the alkali glaze production from the 12th-15th century from Uzbekistan to Afghanistan. Mineral natron characteristics are also reported by Mason (2004) for Syrian and Iranian alkali glazes on lustre ware of the 8th-14th century. Tile glazes with partly mineral natron compositions are found in the Mughal architectural glazes from the 14th- 17th century from India (Gill et al., 2014). Alkali and alkali lead tile glazes from Samarkand from the 13th century (Mongolian period) have mineral natron flux characteristics, but samples from the 15th century (Timurid period) show plant ash signature. Alkali fluxed Uzbek glazes from Bukhara from the 16th century (Sheibanid dynasty) are also made by plant ash flux and are subdivided into two groups with high and low sodium oxide content. The Afghan alkali glazes have sodium oxide contents similar to the sodium-poor Uzbek subgroup, which points to a possible exchange of glaze makers or glaze making technology from Uzbekistan and Afghanistan in the 15th-17th century. Regarding the extensive exchange of Timurid craftsmen in Central Asia, this option seems to be even more likely (Golombek, 1996). One sample from the 15th century from Afghanistan with mineral natron reveals that this material was parallel used in these centuries. Concerning the colouring of the glazes, it has to be distinguished between pigments and colouring ions which are incorporated in the glassy matrix. The colouring agents for translucent glazes are cations of various transition metals. As ions, Co2+ (blue), Cu2+ (green in a lead rich matrix), Fe3+ (brown/black), Mn4+ (brown/black) and Mn3+ (violet) are determined by EPMA. For opaque yellow, white, and turquoise glazes, different pigments were used. The crystalline pigments are investigated by a µ XRD2 device with the result of SnO2, SiO2, and PbSiO4 as whitening agents. PbSiO4 and Pb2Sn2O6 are found in the yellow glaze, from which only the lead tin oxide causes the yellow colour. In the black glazes, different Cr-rich pigments, Cu-Cr-Mn-oxides and iron containing clinopyroxenes are found, even in samples of the same period and region. Cr-rich particles are also detected in two turquoise Afghan glazes from the 15th and 16th century. The use of the ions of Fe, Cu, Co, Cr, and Mn seems to be widely common in the Islamic glazes and corresponds to the described colouring agents in e.g. the study of Tite (2011). The use of opacifying SnO2 particles is widespread as it is reported from different Islamic glazes from Iraq, Iran, Egypt, and Syria (Henshaw, 2010; O´Kane, 2011; Tite, 2011). The colouring agents are known already from former, e.g. Egyptian, Roman and pre-islamic periods, but especially SnO2 pigments became increasingly widespread in the Islamic glazing tradition. The use of yellow and black pigments instead varies already within the buildings from Bukhara from Cr crystals and clino-pyroxenes in the mosque Khoja Zainuddin to a Cu-Cr-Mn-oxide in the madrassa Mir-i Arab of the same epoch. Regarding the matrix compositions connected with the colouring, a certain assignment within the different locations and epochs can be seen. It is noticeable that e.g. the content of lead in turquoise glazes in Uzbekistan is in the range of 0.0-9.2 wt% Pb, whereas blue glazes are mostly alkali ones with PbO contents <2.0 wt%. The turquoise glazes show, that this restriction is not influenced by any defaults of availability and processability. The assumption of common addition of lead and tin to the glaze, which is already described for Iranian glazes of the 13th century (Allan et al., 1973) cannot be confirmed by correlations of tin and lead oxide in the compositions. 11.2 Portable XRF measurement With the p-XRF, semi-quantitative information about the major element compositions is generated. The depth of the detectable signals depends on the analysed sample setup. The p-XRF data are collected with the XL3 Hybrid device of the company Analyticon Instruments. In the comparison of p-XRF results of the “mining” program from Uzbek glazes with EPMA results, the same major composition groups can be distinguished. The Moroccan glazes, all lead rich, are measured with the “mining” as well as with the “soil” program, revealing a better performance in the “mining” measurements. The deviations are nevertheless high, because of the high lead contents, which make the calculation of matrix correction difficult. The measurement of the colouring oxides MnO2, CoO, and CuO is satisfying with the internal calibration of the device and even improved with the “mining” program measurement, if compared to the results of the “soil” program. The measurements of glaze imitations lead to better results than that of bulk glass. This can be attributed to the smoother surface texture. In spite of the accuracy limits in the measurements of particular elements in glazes, the classification of flux composition into three groups could be confirmed with the p XRF analysis. The measurement precision is therefore sufficient for the semi-quantitative analysis of the flux characteristic of glazes. Especially for the on-site measurement of large sample quantities on historical buildings, the device is a suitable tool. 11.3 Restoration material The ORMOCER® fulfils the requirements of stability, reversibility, and transparency, which are imposed to a modern restoration material. As pigments, historically coloured glass, cobalt blue, Egyptian blue, lead tin yellow, manganese violet, iron oxide, copper oxide, and cassiterite were used. The metal compounds have higher colour intensities than the pigments of coloured glass. It has to be considered that the proportion of ORMOCER® in the batch must be high enough (70 vol%) to guarantee the ORMOCER® properties of weathering and mechanical stability. The adhesion properties of the ORMOCER® and the homogeneity of the mixture are the best in a fraction of max. 30 vol% particles per ORMOCER®. With integrated particles, the ORMOCER® G materials show homogeneous coatings, whereas the particles in the ORMCOER® E show more agglomeration. In the sedimentation and weathering experiments, the use of an ultrasonic finger in combination with a roller mill is favourable compared to the treatment with bead grinding mill. The treatments with ultrasonic finger and roller mill result in less sedimentation and better adhesion of the dispersions. The treatment of the dispersions in the bead grinding mill does not result in sufficient adhesion, certainly due to the sedimentation behaviour and a congregation of particles on the bottom of the coating. The modification of dispersed nano-particles by 3-methacryl-oxypropyltrimethoxysilan leads to a further homogenization in the sedimentation tests. It is therefore approved for the use in coloured glaze supplements. In weathered coatings of nano-particle compounds, the surface modification shows certainly no enhancement of stability. The treatment of pigmented coatings with an additional layer of pure ORMOCER® results in a bright and transparent appearing, which is closer to the original optical appearance of the glaze. A long-time test application on a historical building will be the next step to validate the suitability of the restoration material. / Analyse historischer Islamischer Glasuren und die Entwicklung eines Restaurierungsmaterials

Zugl.: Aachen, Techn. Hochsch., Diss., 2007.

Techn. Univ., Diss.--Dresden, 2007.