No description available.

Fused deposition modeling (FDM) is gaining an ever increasing attention for its ability to fabricate complex geometry parts and prototypes at lower cost. The technology is striving to produce parts with high mechanical resistance that can withstand and perform under high stress environment. The adhesion strength between layers, transverse strength, is a limiting factor that need to be quantitatively evaluated to further understand and improve the bonding behavior of thermoplastic polymer in FDM. This interfacial adhesion is derived by the diffusion and penetration of polymer chains across the interface allowing the chain entanglement to form a bonding medium. This study investigates the bonding behaviour of polylactic acid (PLA) as a function of chain branching. The adhesion strength is quantitatively evaluated by developing and performing a peel test of a two-printed layer samples. It is possible to increase chain branching of PLA by bulk modification with epoxy chain extender. The modification of PLA was carried out using an internal batch mixer with four different concentrations of chain extender. The modified PLA was processed into print filament and characterized by parallel plate rheometry and DSC. It was found that the addition of chain extender increased molecular weight and degree of branching of PLA and in return the peel testing results reflected a significant increase in adhesion strength. Such improvement can be attributed to the long branched chains of PLA and its ability to create entanglements between layers. These findings can help in producing better PLA filaments to provide a higher stress resistance for FDM fabricated functional parts. / Thesis / Master of Applied Science (MASc) / Fused Deposition Modeling (FDM) is a recent popular method of plastic 3D printing technique, in which plastic filament is heated to a molten state to be then deposited through a layer-by-layer fashion to successfully fabricate parts. One of the drawbacks of that technology is the low bonding strength developed between layers as compared to strength along the length direction of layers. This study focuses on developing a testing methodology to evaluate the adhesion strength between layers and altering the material structure to maximize such strength. Four types of polylactic acid with different degrees of chain branching were successfully processed, printed and tested. Material with higher degree of branching yielded higher adhesion strength.

In dieser Arbeit wurde die alte Designvariante des Fabry-Pérot-Filters (FPI4) hinsichtlich Performance und Ausbeute untersucht. Zielstellung war eine Optimierung, welche die Probleme wie Sticking, Kurzschluss zwischen Schirm und Elektrode sowie Defekte bei den Vereinzelungsverfahren beseitigen sollten. Ferner wurden eine Chipbezeichnung und eine neue geteilte Elektrode anvisiert. Ergebnis dieser Optimierung war das neue Design FPI5. Weiterhin wurde neben der Auslegung des FPI5 für das adhäsive Bonden mit SU8 eine Designvariante für das direkte Bonden realisiert. Das direkte Bonden konnte bereits bei den FPI mit zwei beweglichen Reflektorträgern erfolgreich getestet werden und soll nun auf die Designvariante mit nur einem beweglichen Reflektorträger übertragen werden. Außerdem wurde eine Überprüfung der SU8-Bondfestigkeit durchgeführt. Dabei konnte nachgewiesen werden, dass die bis dato benutzten Bondparameter nicht das Optimum bezüglich der Bondfestigkeit darstellen, da in Testverbunden mit veränderten Bondparametern bis zu 50 % höhere Bondfestigkeiten erreicht wurden. / This work presents a new design type of the older well known tunable micromachined Fabry-Pérot filter (FPI4). Compared to the previous system the new design (FPI5) aimed to avoid the problems as sticking, short circuit between the shield and the electrode as well as the malfunction in dicing and separation. Furthermore an identification mark and a new separated electrode were to be improved. The conclusion of this optimization was the new Design FPI5. The whole geometry was in addition to the adhesive bonding with SU8 moreover designed for the direct bonding, which was successfully verified in the FPI design with two moveable reflector carriers. Additionally a verification of the bonding strength at different bonding parameters was done. This verification showed that a 50 % higher bonding strength could be achieved by using the new bonding parameters in comparison to the normally used bonding parameters.

Le collage direct est maintenant utilisé par un nombre croissant d'applications en microélectronique (Elaboration de SOI, technologie imager Back Side Illumination, technologies 3D...). C'est une technique d'assemblage permettant de coller deux surfaces sans apport de matière adhésive. Principalement utilisée pour le collage de wafers, elle vient en complément de techniques telles que l'épitaxie ou le dépôt de couches minces. Ce collage s'effectue sous certaines conditions : il faut que les surfaces soient suffisamment propres, planes et lisses pour qu'il y ait une adhésion spontanée à température et pression ambiante. Enfin, un traitement thermique est appliqué pour augmenter l'énergie d'adhérence. Pendant le processus de fabrication, il peut apparaître des défauts de collage qui sont essentiellement dus à un piégeage de particules. Ces défauts se présentent sous la forme de bulles d'air. Finalement, les défauts de collage et l'énergie de collage sont les deux caractéristiques à partir desquelles est déduite une qualité de collage.Aujourd'hui, la technique utilisée pour la mesure d'énergie de collage est le clivage au coin. C'est une technique qui consiste dans un premier temps à séparer partiellement deux wafers par une lame, et dans un second temps, à calculer l'énergie de collage à partir d'une équation comportementale qui intègre la longueur de décollement. Mis à part le fait qu'elle permette la mesure d'énergie seulement sur quelques points, il se trouve que c'est une technique destructive. Un contrôle non destructif serait très intéressant pour l'industrie microélectronique et spécialement pour les lignes d'inspection. De plus, les procédés de fabrication microélectronique n'étant pas uniforme, avoir la possibilité d'obtenir une cartographie d'énergie de collage serait un atout majeur. A ce jour, aucune technique respectant ces deux exigences n'est connue. L'objectif de cette étude est d'utiliser la microscopie acoustique pour mesurer l'énergie de collage.Dans cette étude, un modèle inspiré de la "méthode des matrices hybrides" a été développé afin de modéliser des collages de différentes qualités. Le résultat de la modélisation montrera que le coefficient de réflexion acoustique de la structure collée est influencé par la qualité d'interface. En se plaçant dans des conditions précises, une méthode expérimentale est alors réalisée pour la mesure de la qualité d'interface. En parallèle, des wafers de Silicium réalisés par collage direct ont été spécialement conçus pour valider la méthode. Sur ce principe, des cartographies bidimensionnelles d'énergie de collage sont réalisées.Dans un second temps, la technique est améliorée afin d'augmenter la résolution latérale. Pour cela, un transducteur ayant une lentille est utilisé pour focalisé les ondes ultrasonores en points du collage. Une étude théorique est tout d'abord menée en utilisant le modèle du "spectre angulaire" afin de simuler la diffraction par la lentille. Enfin, des cartographies expérimentales confirmeront la faisabilité de mesures d'énergie de collage hautes résolutions. / Direct bonding is used for many applications in microelectronics (SOI Silicon-On-Insulator technology, imager back side illumination technology, 3D technology...). It is a processes that consists in an assembly of two surfaces without any adhesive material. It is primarily used to bond silicon wafers and it is complementary with other microelectronics technique such as epitaxy, thin film deposition... Bonding requires special wafer surface conditions and preparations. The surfaces have to be clean, flat and smooth to obtain a spontaneous adhesion at ambient temperature and atmospheric pressure. A heat treatment is applied to increase the adherence energy. During the manufacturing process, bonding defects may appear which are due to trapping of particles. These bonding defects are essentially formed of air. Finally, bonding defects and bonding energy are the two main characteristics from which is deduced the bonding quality.Nowadays, the main technique that is used to measure the direct bonding energy is the double cantilever beam (DCB). The method consists in firstly partially separating the two wafers by a blade, and secondly calculating the bonding energy from an equation that integrates the debonding lenght. The major disadvantage of this technique is its destructiveness. Furthermore it is only possible to make measurements on few points.Thus a non-destructive characterisation could be very interesting especially for an industrial in-line inspection. Moreover, having the possibility to obtain a mapping of the bonding energy could lead to interesting development. Up to know, no technique can reach the both requirements. The aim of this work is to use the acoustic microscopy to measure the direct bonding energy.In this study, a model based on "hybrid matrix method" has been developed to model bonding with different qualities. The results of the modelling show that the acoustic reflection coefficient of the bonded structure is influenced by the quality of the interface. From these results, an experimental method is proposed to perform quality of the interface measurements from the reflection coefficients acquired under normal incidence. In parallel, silicon wafers have been bonded to validate the method. Finally, once the method validated, two-dimensional mappings of the interface quality are realised.Secondly, the technique is improved to increase the lateral resolution. For this, a transducer having a lens is used to focus the ultrasonic waves on the bonded structure. A theoretical study is conducted using the model of the "angular spectrum" to simulate the diffraction lens. Finally, experimental mapping confirm the feasibility of measuring bonding energy of high resolutions.

Syfte: Syftet med föreliggande in-vitro studie är att undersöka olika kompositbaserade reparationssystem och dess bindningsstyrka till porslin för metallkeramiska konstruktioner. Material och metod: Metallstavar tillverkades via CAD/CAM i CoCr (5.0x5.0x15.0mm) kortsidan förbehandlades genom sandblästring (aluminumoxid, 250 µm), ångblästring, oxidation, sandblästring och ångblästring. Fältspatporslin applicerades enligt följande: opak, dentin, emalj, till höjden av 1.2 ± 0.2 mm. Porslinsytorna slipades med sandpapper, 600-, 800-, 1200-kornstorlek kiselkarbid. Fem grupper (n=15) gav totalt 75 provkroppar. På vardera provkropp applicerades adhesiv samt kompositmaterial enligt fabrikanternas anvisningar (Kuraray, Kulzer, Voco, Ivoclar) inklusive kontrollgrupp som etsades med fluorvätesyra (Ultradent). Samtliga provkroppar termocyklades (5000 cykler, mellan 5°C och 55°C) före/efter applicering av kompositmaterial. Skjuvkraftstest utfördes på alla provkroppar med belastningshastighet 0.5 mm/min. Frakturanalys utfördes där brottytorna delades in i adhesiv-, kohesiv- eller mixfraktur. Resultaten analyserades med one-way ANOVA, Tukey´s test med signifikansnivån α=0.05. Resultat: Fluorvätesyra påvisade en signifikant högre bindningsstyrka 11.32 ± 4.28  MPa i jämförelse med övriga grupper (p<0.01). Övriga gruppers medelvärde var mellan 2.61 ± 1.36 till 3.96 ± 2.85 MPa. Det fanns ingen signifikant skillnad mellan de övriga grupperna. Samtliga grupper fick adhesiva frakturer, men Voco-gruppen uppvisade även mixfrakturer.  Slutsats: Samtliga reparationssystem visar på låg bindningsstyrka och baserat på resultaten kan det inte rekommenderas. Förbehandling av porslinsyta med fluorvätesyra i kombination med silan resulterar i högst bindningsstyrka till komposit. Värdet är trots allt lågt för en god långtidsprognos, men det enda som kan rekommenderas. Det faktum att fluorvätesyra har giftiga egenskaper vore det lämpligt att fortsätta utvecklingen av reparationssystem för att hitta ett lämpligt alternativ. / Purpose: The purpose of this study is to investigate different repair systems and their bond strength to porcelain for metal-ceramic constructions. Material and method: Metal were manufactured by CAD / CAM in CoCr (5.0x5.0x15.0mm), was pretreated with sandblasting (alumina, 250 µm), steam blasting, oxidation, sandblasting and steam blasting. Feldspar porcelain was applied according to the following steps: opaque, dentin, enamel, selfglaze to a final porcelain height of 1.2 ± 0.2 mm. The porcelain surfaces were groud with silicon carbide sandpaper, 600-, 800-, 1200-grain size. Five groups (n = 15) in a total of 75 specimens. Adhesives and composite materials were applied to each group according to the instructions of the various manufacturers  (Kuraray, Kulzer, Voco, Ivoclar, Ultradent, Etching with hydrofluoric acid). All specimens were thermocycled (5000 cycles) in two water baths between 5 ° C and 55 ° C before and after application of composite material. Shear bond strength test was performed on all specimens with a load speed of 0.5 mm / min and the bond strength was reported in MPa. Fracture analysis was performed where the fracture surfaces were divided into adhesive, cohesive or mixed fracture. The results were analyzed with one-way ANOVA, Tukey´s test with significance level α= 0.05. Result: The result clearly showed that hydrofluoric acid had a significantly higher bond strength of 11.32 ± 4.28 MPa compared to the other groups (p <0.01). The mean value of the other groups was between 2.61 ± 1.36 to 3.96 ± 2.85 MPa, with  no significant difference between the groups. All groups had adhesive fractures, but the Voco-group also showed mixed fractures. Conclusion: The study concludes that all repair systems have a low bond strength and based on the results can’t be recommended. Pretreatment of the porcelain surface with hydrofluoric acid in combination with silan results in the highest bond strength to composite, the bond strength is still too low for long-term use, but the only one that can be recommended. However hydrofluoric acid is highly toxic which require further research in the field in order to develop reparation systems with suitable properties.

Esta pesquisa in vitro avaliou a influência do aquecimento de diferentes agentes de acoplamento na resistência de união de um cimento resinoso (Panavia F2.0®/Kuraray Noritake. Japão) e uma zircônia (Y-TZP - YZ ). Inicialmente, placas de zircônia foram silicatizadas com o sistema Rocatec (3M / ESPE). Foram aplicados os agentes de acoplamento químico: Single Bond Universal (3M/ESPE); Monobond N (Ivoclar Vivadent) e Clearfil Ceramic Primer (Kuraray Co); após, foram aquecidos em um forno resistivo em diferentes temperaturas (750C e 1000C) por 5 min, antes da cimentação resinosa de um cilindro de resina composta previamente confeccionado (diam. 0,8 mm). Os grupos controle de cada agente de acoplamento foram mantidos à temperatura ambiente (240C). Metade dos corpos-de-prova foi submetida imediatamente ao teste de microcisalhamento (0,5 mm/min) realizado com auxílio de um fio ortodôntico e a outra metade envelhecida por meio de a 2.000 ciclos térmicos (50C e 550C), seguido de armazenamento em água destilada a 370 C por 2 meses, antes do ensaio mecânico. A análise de variância de três fatores (ANOVA) e o teste de Tukey (p?0,05) mostraram que o aquecimento dos agentes químicos de acoplamento produziu maiores valores de resistência de união entre o cimento e a zircônia. A resistência de união imediata foi superior para o Monobond N (30 MPa) e Single Bond Universal (32,5 MPa). Após o envelhecimento, houve redução da resistência de união em todos os grupos e o grupo Monobond N (16,9 MPa) apresentou resistência de união superior à dos demais. A associação de um tratamento térmico simples a um agente de acoplamento é válida para produzir resistência de união superior entre o cimento resinoso e a zircônia. / This in vitro study evaluated the influence of the heating different coupling agents on microshear bond strength between zirconia (Y-TZP-YZ ) and a resin cement (Panavia F2.0® / Kuraray Noritake, Japan). Initially, the zirconia blocks were silicoated with the Rocatec system (3M / ESPE). After, the chemical coupling agents were applied: Single Bond Universal (3M / ESPE); Monobond N (Ivoclar Vivadent) and Clearfil Ceramic Primer (Kuraray Co). They were then heated in a resistive furnace at different temperatures (750C and 1000C) for 5 min prior to resin cementation of a preformed resin composite cilinder (0.8 mm diameter). Control groups were maintained at room temperature (240C). Half of the specimens were immediately submitted to the microshear test (0.5 mm/min) performed with an orthodontic wire and the other half was subjected to 2,000 thermal cycles (50C and 550C), followed by storage in distilled water at 370C for 2 months before the microshear test. 3-way ANOVA and Tukey\'s test (p<0.05) showed that the heating of the chemical coupling agents produced higher bond strength between the ceramic and resin cement. Immediate bond strength was superior for Monobond N (30 MPa) and Single Bond Universal (32,5 MPa). After aging, there was reduction of bond strength in all groups and the Monobond N group (16,9 MPa) was superior to the others. The combination of a simple heat treatment with a coupling agent is valid to produce greater bond strength.

L’électronisation de l’industrie a mené à l’augmentation de la pollution électromagnétique pouvant être néfaste pour les systèmes électroniques sensibles et les êtres vivants. L’un des moyens pour limiter la propagation des champs électromagnétiques est l’utilisation d’un blindage. L’étude s’est ainsi focalisée sur l’élaboration à froid d’un composite bimétallique pour le blindage magnétique. Le composite étudié allie les propriétés physiques de l’aluminium et de l’acier via le trilame Al8011/AcierDC01/Al8011. Le trilame est élaboré par colaminage à température ambiante. Il s’est avéré que la qualité de l’adhérence des interfaces Al/Acier et l’architecture du trilame dépendent fortement des paramètres du colaminage. Une préparation minutieuse des tôles et de leurs surfaces concomitantes se révèle être tout aussi importante que le colaminage en lui-même. De plus, au cours du colaminage, la tôle d’acier s’est montrée sensible à des instabilités plastiques qui amènent par la suite à sa striction et à sa fragmentation. Ces instabilités plastiques favorisent l’adhérence grâce à des soudages Al/Al mais n’en sont pas un prérequis. Au niveau de l’efficacité de blindage, le trilame s’est révélé être particulièrement intéressant puisque, grâce à sa composition et à sa structuration, il est capable d’atténuer aussi bien les champs magnétiques basses fréquences (< 1 kHz) que les champs magnétiques de plus hautes fréquences (> 1 kHz). Dans une étude à iso-masse, le trilame a présenté une meilleure efficacité de blindage que les tôles d’Al, de Cu et d’acier. Cependant, la fragmentation de l’acier dans le trilame s’est révélée être néfaste pour le blindage magnétique, nécessitant de faire alors un compromis entre tenue mécanique et efficacité de blindage. / The growth of electronic devices has led to an increase in electromagnetic pollution that can be harmful to sensitive electronic systems and living beings. One of the means of limiting the propagation of electromagnetic fields is the use of shielding. Then, the study focused on the elaboration of a bimetallic composite for magnetic shielding. The studied composite combines the physical properties of aluminum and steel via the Al8011/SteelDC01/Al8011 trilayer. The trilayer is produced by cold roll bonding (CRB). It has been found that the adherence quality of the Al/Steel interfaces and the architecture of the composite heavily depend on the CRB parameters. Careful preparation of the sheet and their concomitant surfaces is just as important as CRB itself. Furthermore, during CRB, the steel sheet was sensitive to plastic instabilities which subsequently led to its necking and fragmentation. These plastic instabilities promote adherence through Al/Al welds but are not a prerequisite. In terms of shielding effectiveness, the bimetallic composite has proved to be particularly interesting. Thanks to its composition and its structuring, it can attenuate both low (< 1 kHz) and high frequency (> 1 kHz) magnetic fields. In an iso-mass study, the composite showed a higher shielding effectiveness than Al, Cu and steel sheets. However, the steel fragmentation in the composite proved to be detrimental to magnetic shielding, then requiring a compromise between mechanical strength and shielding effectiveness.

In dieser Arbeit wurde die alte Designvariante des Fabry-Pérot-Filters (FPI4) hinsichtlich Performance und Ausbeute untersucht. Zielstellung war eine Optimierung, welche die Probleme wie Sticking, Kurzschluss zwischen Schirm und Elektrode sowie Defekte bei den Vereinzelungsverfahren beseitigen sollten. Ferner wurden eine Chipbezeichnung und eine neue geteilte Elektrode anvisiert. Ergebnis dieser Optimierung war das neue Design FPI5. Weiterhin wurde neben der Auslegung des FPI5 für das adhäsive Bonden mit SU8 eine Designvariante für das direkte Bonden realisiert. Das direkte Bonden konnte bereits bei den FPI mit zwei beweglichen Reflektorträgern erfolgreich getestet werden und soll nun auf die Designvariante mit nur einem beweglichen Reflektorträger übertragen werden. Außerdem wurde eine Überprüfung der SU8-Bondfestigkeit durchgeführt. Dabei konnte nachgewiesen werden, dass die bis dato benutzten Bondparameter nicht das Optimum bezüglich der Bondfestigkeit darstellen, da in Testverbunden mit veränderten Bondparametern bis zu 50 % höhere Bondfestigkeiten erreicht wurden.:ABKÜRZUNGS- UND FORMELVERZEICHNIS VI 1 EINLEITUNG 8 1.1 Motivation 8 1.2 Kapitelübersicht 11 2 DAS FABRY-PEROT-INTERFEROMETER 12 3 DAS ALTE DESIGN FPI4 15 3.1 FPI4 15 3.2 Optimierungspunkte und Schwachstellen 17 4 DAS NEUE DESIGN FPI5 19 4.1 Verhindern von Sticking 20 4.2 Isolator zwischen Schirm und Elektrode 22 4.3 Abdichtfolie für den Sägeprozess 22 4.4 Chipbezeichnung 23 4.5 Elektrodenform 23 4.6 Zusammenfassung des Redesigns 25 5 PROZESSTECHNISCHE REALISIERUNG 27 5.1 FPI5 – Direktbond 27 5.1.1 Oberteil 28 5.1.2 Unterteil 30 5.2 FPI5 – SU8 31 5.2.1 Oberteil 32 5.2.2 Unterteil 34 5.3 Selektive Schichtabscheidung der optischen Schichten 36 6 CHARAKTERISIERUNG DER MUSTER 37 6.1 FPI5 – Direktbond 37 6.1.1 Oberteil 38 6.1.2 Unterteil 40 6.1.3 Beschichtung der optischen Schichten in Jena 42 6.1.4 Zusammenfassung 44 6.2 FPI5 – SU8 47 6.2.1 Oberteil 48 6.2.2 Unterteil 49 6.2.3 Beschichtung der optischen Schichten in Jena 52 6.2.4 Zusammenfassung 54 7 BONDEN DER FPI 57 7.1 Vorversuche zur SU8-Bondfestigkeit 57 7.1.1 Realisierung der Bondmuster 59 7.1.2 Charakterisierung der Bondmuster 60 7.1.3 Vereinzelungstests der Bondmuster 63 7.1.4 Bondfestigkeitsüberprüfung der Bondmuster mit abgerundeten Bondrahmen 64 7.1.5 Bondfestigkeitsüberprüfung der Bondmuster ohne abgerundeten Bondrahmen 67 7.2 Direktbonden der FPI 73 7.2.1 Ablauf des Direktbondens 73 7.2.2 Charaktersierung der Bondmuster 76 7.3 SU8-Bonden der FPI 78 7.3.1 Präparation der SU8-Rahmenstruktur 78 7.3.2 Ablauf des SU8-Bondens 80 7.3.2 Charakterisierung der Bondmuster 81 8 ZUSAMMENFASSUNG UND AUSBLICK 83 LITERATURVERZEICHNIS 85 ABBILDUNGSVERZEICHNIS 87 TABELLENVERZEICHNIS 88 ANLAGENVERZEICHNIS 90 / This work presents a new design type of the older well known tunable micromachined Fabry-Pérot filter (FPI4). Compared to the previous system the new design (FPI5) aimed to avoid the problems as sticking, short circuit between the shield and the electrode as well as the malfunction in dicing and separation. Furthermore an identification mark and a new separated electrode were to be improved. The conclusion of this optimization was the new Design FPI5. The whole geometry was in addition to the adhesive bonding with SU8 moreover designed for the direct bonding, which was successfully verified in the FPI design with two moveable reflector carriers. Additionally a verification of the bonding strength at different bonding parameters was done. This verification showed that a 50 % higher bonding strength could be achieved by using the new bonding parameters in comparison to the normally used bonding parameters.:ABKÜRZUNGS- UND FORMELVERZEICHNIS VI 1 EINLEITUNG 8 1.1 Motivation 8 1.2 Kapitelübersicht 11 2 DAS FABRY-PEROT-INTERFEROMETER 12 3 DAS ALTE DESIGN FPI4 15 3.1 FPI4 15 3.2 Optimierungspunkte und Schwachstellen 17 4 DAS NEUE DESIGN FPI5 19 4.1 Verhindern von Sticking 20 4.2 Isolator zwischen Schirm und Elektrode 22 4.3 Abdichtfolie für den Sägeprozess 22 4.4 Chipbezeichnung 23 4.5 Elektrodenform 23 4.6 Zusammenfassung des Redesigns 25 5 PROZESSTECHNISCHE REALISIERUNG 27 5.1 FPI5 – Direktbond 27 5.1.1 Oberteil 28 5.1.2 Unterteil 30 5.2 FPI5 – SU8 31 5.2.1 Oberteil 32 5.2.2 Unterteil 34 5.3 Selektive Schichtabscheidung der optischen Schichten 36 6 CHARAKTERISIERUNG DER MUSTER 37 6.1 FPI5 – Direktbond 37 6.1.1 Oberteil 38 6.1.2 Unterteil 40 6.1.3 Beschichtung der optischen Schichten in Jena 42 6.1.4 Zusammenfassung 44 6.2 FPI5 – SU8 47 6.2.1 Oberteil 48 6.2.2 Unterteil 49 6.2.3 Beschichtung der optischen Schichten in Jena 52 6.2.4 Zusammenfassung 54 7 BONDEN DER FPI 57 7.1 Vorversuche zur SU8-Bondfestigkeit 57 7.1.1 Realisierung der Bondmuster 59 7.1.2 Charakterisierung der Bondmuster 60 7.1.3 Vereinzelungstests der Bondmuster 63 7.1.4 Bondfestigkeitsüberprüfung der Bondmuster mit abgerundeten Bondrahmen 64 7.1.5 Bondfestigkeitsüberprüfung der Bondmuster ohne abgerundeten Bondrahmen 67 7.2 Direktbonden der FPI 73 7.2.1 Ablauf des Direktbondens 73 7.2.2 Charaktersierung der Bondmuster 76 7.3 SU8-Bonden der FPI 78 7.3.1 Präparation der SU8-Rahmenstruktur 78 7.3.2 Ablauf des SU8-Bondens 80 7.3.2 Charakterisierung der Bondmuster 81 8 ZUSAMMENFASSUNG UND AUSBLICK 83 LITERATURVERZEICHNIS 85 ABBILDUNGSVERZEICHNIS 87 TABELLENVERZEICHNIS 88 ANLAGENVERZEICHNIS 90