Silicium wird während des Ätzvorgangs von Solar-Wafern mit HF-HNO3-Mischungen hauptsächlich zu Hexafluorosilicat (SiF62-) umgewandelt, welches sich negativ auf den Ätzabtrag, die Reaktivität der Ätzlösung und die Oberflächeneigenschaften der Wafer auswirkt. Möglichkeiten, das Silicium als SiF62- abzutrennen, sollten in dieser Arbeit untersucht werden. Voraussetzung für Abtrennungsuntersuchungen war eine quantitative Bestimmung des SiF62- in Ätzlösungen mittels Ramanspektrometrie. Als Abreicherungs-möglichkeiten des Siliciums in Form von Hexafluorosilicat wurden einerseits das Ausfrieren als Hydrat der Hexafluorokieselsäure (H2SiF6•nH2O) und andererseits die Fällung als K2SiF6 untersucht. Die experimentellen Ergebnisse wurden jeweils gestützt durch thermodynamische Modellierungen: die Tieftemperaturphasendiagramme wurden durch eine modifizierte BET-Modellierung bestätigt, die Löslichkeiten durch Nutzung des SIT-Ansatzes. Demnach erwies sich die Ausfällung als K2SiF6 als ökonomisch günstigste Variante.:Inhaltsverzeichnis 1 Einleitung und Problemstellung 2 Auswertung der Literatur 2.1 Chemie des Siliciums in HF/HNO3-Ätzlösungen 2.1.1 Allgemeine Betrachtungen zum Ätzvorgang 2.1.2 Rolle des H2SiF6 bzw. SiF62- beim Ätzen 2.2 Speziation des Siliciums in fluoridhaltigen Lösungen 2.2.1 Schwingungsspektroskopie 2.2.2 Kernresonanzspektroskopie 2.2.3 Gleichgewichtskonstanten 2.3 Phasendiagramme im System HF-H2SiF6-HNO3-H2O bei tiefen Temperaturen 2.3.1 Binäre Systeme 2.3.2 Ternäre und höhere Systeme 2.3.3 Thermodynamische Modellierung konzentrierter Elektrolytlösungen mittels modifiziertem BET-Modell 2.4 Charakteristik kristalliner Hexafluorosilicate 2.4.1 Allgemein 2.4.2 Eigenschaften des K2SiF6 2.4.3 Struktur des K2SiF6 2.4.4 Thermisches Verhalten von K2SiF6 2.5 Thermodynamische Modellierung von Löslichkeiten 2.5.1 Specific Ion Interaction Theory 2.5.2 Pitzer-Modell 2.5.3 Modellierung nach Pitzer- Simonson- Clegg 2.5.4 (e)PC-SAFT EOS 2.5.5 Zusammenfassender Vergleich der Modelle 2.5.6 Vorliegende Daten als Grundlage einer Modellierung 2.5.7 Wahl eines geeigneten Löslichkeitsmodells 3 Experimentelle Untersuchungen zur Ramanspektroskopie wässriger Fluorosilicatlösungen 3.1 Messbedingungen und Messküvetten 3.2 Spektren im System HF-H2SiF6-HNO3-H2O 3.3 Quantitative Bestimmung von SiF62- und HNO3 4 Phasendiagramme im System HF-H2SiF6-HNO3-H2O bei tiefen Temperaturen (<273 K) 4.1 Versuchsaufbau zur thermischen Analyse 4.2 Durchführung und Auswertung von Temperatur-Zeit-Kurven am Beispiel des binären Systems HF-H2O 4.3 Ergebnisse 4.3.1 Binäre Systeme 4.3.2 Ternäre Systeme 4.3.3 Quaternäres System 4.4 Modellierung der Phasendiagramme mit dem BET-Modell 5 Löslichkeit von K2SiF6 in Ätzlösungen 5.1 Methodik und Analysentechniken 5.2 Ergebnisse in Lösungen des Systems HF-H2SiF6-HNO3-H2O-K2SiF6 5.3 K2SiF6-Löslichkeit in KF- bzw. KNO3-haltigen wässrigen Lösungen 5.3.1 Löslichkeitskurven 5.3.2 Struktur und Eigenschaften des K2SiF6∙KNO3-Doppelsalzes 5.4 Quantitative Beschreibung der Löslichkeit von K2SiF6 in HF-H2SiF6-HNO3-H2O 6 Beurteilung der Trennverfahren 7 Zusammenfassung Literaturverzeichnis Abbildungsverzeichnis Tabellenverzeichnis Symbole und Abkürzungen Anhang Eidesstattliche Versicherung Danksagung

Fe und Cu wurden als Schlüsselspezies für die Betrachtung von Oberflächenkontamination in der Prozessierung von c-Si-Solarzellen identifiziert. Studien mit gezielt aufgebrachten Metallkonzentrationen vor verschiedenen Passivierungs- und Diffusionsprozessschritten ergaben relativ hohe kritische Werte für Cu, außer bei thermischer Oxidation. Niedrige Werte wurden für beide Elemente vor Hochtemperaturschritten im n-Typ-Hocheffizienzprozess beobachtet, wobei sich die B-Diffusion als etwas weniger empfindlich darstellte. Temporäre Gettereffekte für Fe (in p-Typ-Si) und Cu (in n-Typ-Si) wurden beobachtet. Es zeigte sich, dass As-Cut-Wafer, unabhängig von der Sägeart (SiC-slurry oder Diamantdraht) sehr hohe Metallverunreinigungen (im Bereich 1*1011 – 5*1014 cm-2) in den Prozess einbringen. Das alkalische Ätzen verringert diese Menge kaum, was hohe Anforderungen an die anschließende Reinigung ergibt. Die Optimierung von HF/O3-Reinigungslösung für diese Flächen ergab beste Reinigungsergebnisse bei niedrigen HF-Konzentrationen, abhängig vom alkalischen Ätzschritt und anschließendem HCl/HF-Dip.

Fabrication of hemispherical structures for application in hemispherical resonator gyro-scopes (HRG) is an integral part of modern sensing systems, especially in relation to space navigation. First, it is important for these structures to be as symmetric as possible in order to accurately track both in-plane and out-of-plane acceleration that occurs in fast moving satellites and space crafts. Next, they need to be larger for easier application in current mm scale systems and to maintain a lower noise floor and high quality factor. The work in this paper introduces a methodology for the analyzation of the micromachining process for larger symmetric hemispherical shell resonators (HSR). This is in order to increase their size while maintaining symmetry through isotropic etching using HNA and the pop-up ring mask design. The implementation of the pop-up ring mask allows for symmetric etching of<111> silicon and larger MEMS structures at a low cost while giving more design control to the user in comparison to alternative designs such as the pinhole. The investigation of how hemispheric structures are affected based on the adjustment of the pop-up ring design serves to both create larger symmetric HSRs and create a better model for future designs and applications. During this investigation, a range of design tests were done to create the hemispherical resonator molds in order to gauge the effectiveness of the pop-up ring changes. These results were then used to develop a method for achieving the desired larger symmetric HSRs.

Les cellules solaires à base de nanofils de silicium offrent une alternative intéressante pour la réalisation de panneaux photovoltaïques à haut rendement et à faible coût. Elles bénéficient notamment des excellentes propriétés optiques des nanofils qui forment une surface à très faible réflectivité tout en piégeant efficacement la lumière. Dans cette thèse, nous utilisons et améliorons une méthode de gravure chimique peu coûteuse et industrialisable pour la fabrication de forêts de nanofils de silicium. En adaptant la mouillabilité du substrat et des nanofils, nous avons remédié au problème d'agglomération inhérent à cette méthode lorsqu’on veut obtenir des forêts denses et désordonnées de nanofils. En combinant cette méthode de gravure chimique à la lithographie assistée par nanosphères, nous avons pu fabriquer des réseaux ordonnés de nanofils avec un contrôle précis des propriétés géométriques (diametre des nanofils et distance entre eux). Les propriétés optiques de ces réseaux ont été étudiées théoriquement et expérimentalement afin d'identifier les configurations optimales. Nous avons ensuite fabriqué des cellules solaires à partir de ces différents types de nanofils et deux types de structures. Le premier type, des cellules solaires HIT (Hétérojonction avec couche mince Intrinsèque) à base de nanofils de silicium, a été fabriqué par RF-PECVD. L'optimisation des conditions de dépôt plasma nous a permis d'obtenir des cellules solaires hautement performantes: rendements de 12,9% et facteurs de forme au-delà de 80%. Le second type, des cellules solaires hybrides, est basé sur la combinaison d'une couche organique et des nanofils de silicium. La caractérisation des cellules fabriquées montre des rendements prometteurs. Enfin, nous présentons des résultats préliminaires pour transférer ces concepts à une technologie couches minces. / Silicon Nanowire (SiNW) based solar cells offer an interesting choice towards low-cost and highly efficient solar cells. Indeed solar cells based on SiNWs benefit from their outstanding optical properties such as extreme light trapping and very low reflectance. In this research project, we have fabricated disordered SiNWs using a low-cost top-down approach named the Metal-Assisted-Chemical-Etching process (MACE). The MACE process was first optimized to reduce the strong agglomeration observed at the top-end of the SiNWs by tuning the wettability properties of both the initial substrate and the SiNWs surface. By combining the MACE process with the nanosphere lithography, we have also produced ordered SiNW arrays with an accurate control over the pitch, diameter and length. The optical properties of these SiNW arrays were then investigated both theoretically and experimentally in order to identify the geometrical configuration giving the best optical performance. Disordered and ordered SiNW arrays have been integrated into two types of solar cells: heterojunction with intrinsic thin layer (HIT) and hybrid devices. SiNW based HIT devices were fabricated by RF-PECVD and the optimization of the process conditions has allowed us to reach efficiency as high as 12.9% with excellent fill factor above 80%. Hybrid solar cells based on the combination of SiNWs with an organic layer have also been studied and characterized. The possible transfer of this concept to the thin film technology is finally explored.

Les progrès récents des nanotechnologies permettent d'envisager de nouvelles générations de diodes laser. L'objectif de cette thèse est d'étudier l'apport des cristaux photoniques bidimensionnels pour explorer la faisabilité d'un réseau monolithique de diodes laser tout cristal photonique émettant au voisinage de 2,3 µm en filière GaSb. Chaque laser doit répondre à un certain nombre de critères : une émission monomode à une longueur d'onde stable et précise, une émission fine spectralement avec une puissance de sortie élevée, une longueur d'onde accordable, présentant aucun saut de mode sur la gamme d'accordabilité, un fonctionnement à température ambiante, sous pompage électrique et en régime continu. Ces critères répondent au cahier des charges de la technique de spectroscopie d'absorption à diodes laser accordables multiplexées (MTDLAS : Multiplexed Tunable Diode Laser Absorption Spectroscopy). La technique de MTDLAS est ici mise en œuvre pour les applications de détection de gaz dans le moyen infra-rouge (MIR), soit la gamme de longueur d'onde allant de 2 à 5 µm. Cette gamme de longueur d'onde présente plusieurs fenêtres de transparence (autour de 2,3 µm et de 3,4 à 4 µm) où l'absorption par la vapeur d'eau et le dioxyde de carbone est très faible. L'existence de ces fenêtres est mise à profit pour la détection de molécules gazeuses de l'atmosphère, telles que le monoxyde de carbone ou le méthane. Pour mes travaux de thèse, la longueur d'onde d'émission laser retenue est de 2,3 µm. Cette longueur d'onde correspond à la fenêtre de transparence pour la détection notamment du CH4, du CO et du HF. Ainsi, ce mémoire présente la conception de diodes laser tout cristal photonique, et le développement d'un procédé de fabrication de ces diodes lasers, qui a mené à la réalisation d'une série de composants. Nous montrons qu'une déformation de la maille photonique, associée à une optimisation de la largeur du guide, permet d'obtenir un fonctionnement monomode stable. Se basant sur ce principe, plusieurs géométries de mailles de cristaux photoniques ont été étudiées. Puis, nous nous attachons à lever le verrou technologique de la gravure profonde à fort rapport d'aspect dans les alliages AlGaAsSb. Pour cela, nous présentons des études paramétriques de gravure, conduisant à la mise au point d'un procédé optimisé de gravure profonde. Des profondeurs de gravure de 3,4 µm pour des trous de 370 nm de diamètre, soit un rapport d'aspect de plus de 9, ont ainsi pu être atteintes. Cette étape critique de gravure a été insérée dans un procédé technologique de fabrication de diodes laser tout cristal photonique, que nous avons mis au point. Cela a mené à la réalisation d'une série de composants. / Recent advances in nanotechnology allow considering new generations of laser diodes. The purpose of this thesis is to study the contribution of two-dimensional photonic crystals to design and fabricate a monolithic array of laser diodes emitting near 2.3 µm in GaSb system. Each laser in the array has to respond to stringent criteria : a stable and precisely predefined single-mode emission wavelength, high output power, tunable wavelength with no mode hopping over the tunability range. Moreover, such device should operate at room temperature, under electrical pumping and continuous regime. These criteria respond to the specifications of the technique of Multiplexed Tunable Diode Laser Absorption Spectroscopy (MTDLAS). The MTDLAS technique is here implemented for gas sensing applications in the Mid-wavelength infrared (MidIR), corresponding to the wavelength range from 2 to 5 microns. This wavelength range contains two transparency windows (around 2.3 µm and from 3.4 to 4 µm), where the absorption by water vapor and carbon dioxide is reduced. The existence of these windows is harnessed for the detection of gaseous molecules in the atmosphere, such as carbon monoxide or methane. For my thesis work, the chosen wavelength of laser emission is 2.3 µm. This wavelength corresponds to a transparency window allowing detection of CH4, CO and HF. This dissertation presents the design of all photonic crystal laser diode, and the development of a manufacturing process of such device, which led to the production of a set of components. We show that the engineering of the photonic lattice combined with an optimization of the width of the laser waveguide provide stable, single-mode emission operation. Based on this principle, several geometries for photonic crystal lattice were studied. Then we tackle the technological challenge of deep etching with high aspect ratio in AlGaAsSb alloys. For this, we present parametric studies of etching, leading to the development of an optimized process for deep etching. We succeed to etch holes of 370 nm diameter and 3.4 µm depths, corresponding at an aspect ratio in excess of 9. We have developed a technological manufacturing process of all photonic crystal laser diodes, where this critical etching step has been successful inserted. This led to the realization of a set of components.

<p>   The arrival of Si transistors and integrated circuit technology more than half a century ago made vacuum electronic technology almost extinct. Today, there are only a few niche applications for vacuum electronics. The main issues with this technology are its high voltage requirement and high-power consumption, difficult and costly fabrication technology, lack of integration capability, and poor reliability characteristics. Some of these issues may be addressed by going to nm scale fabrication that did not exist 60 years ago. Other problems such as reliability and lack of integration capability require alternative solutions to what has been proposed so far. Vacuum is the ultimate conduction media allowing electrons to reach the speed of light without any scattering. Consequently, a vacuum transistor, if designed correctly, can achieve THz frequency performance, while delivering Watt-level powers. No semiconductor technology can compete with vacuum technology to deliver such performance. </p> <p>In this work, novel methods for implementing nanoscale field emitter arrays used in vacuum electronics are proposed. Gated and ungated field emitters are fabricated with self-assembly technology and electron beam lithography. Different anisotropic dry etching recipes are developed to achieve emitters with different sharpness and aspect ratios. Our methods lead to field emitter array operation under low voltages (less than 20 V) and high current densities (around 50 A/cm2) using self-assembly and soft film anode-cathode isolator, and field emitter devices with ~4.5 A/cm2 current density with a turn-on voltage less than 50 V using electron beam lithography and oxide anode-cathode isolator. </p> <p>Making reliable field emitter devices is challenging. Due to Joule heating, ion bombardment, and geometrical variations for each tip in the field emitter arrays, emission current becomes nonuniform across the array. Sharper tips emit at a higher rate and eventually, the heat generated at the tip deforms the tips leading to electron emission at a lower rate. With ultra-low doped emitters, the current of each tip is limited to a few nano-amperes leading to a negligible current fluctuation at the tips. </p> <p>Our fabricated ultra-low doped devices with both self-assembly and electron beam lithography techniques presented constant emission current with almost no change over 24 hours of continuous operation. Such excellent reliability characteristics in vacuum field emitter devices have not been demonstrated to date.</p> <p>The screening effect in close-packed field emitter arrays which occurs by nearby conductive or semiconductive objects is thoroughly investigated and different solutions are proposed to reduce this effect between the emitters. Simulation studies using Sentaurus TCAD, MATLAB, and COMSOL Multiphysics simulators facilitated the design and optimization of gated and ungated field emitter arrays. These studies included the effect of sharpness, the distance between neighboring emitters, enclosing the emitters by a Si block around the emitters as well as anode-cathode separation on the electrical characterization of field emitter arrays. </p> <p>The optimum location and operating voltages which lead to a maximum gate control and emitter current density are also studied for gated field emitter arrays. Instead of individually gating each field emitter, it was found that controlling the emission of a sub-array with a metallic all-around gate is more efficient and it leads to higher current densities. Guided by simulations, gated field emitter arrays with 5×5 and 2×2 sub-arrays are developed. In terms of strength of the grid control (transconductance), turn-on voltage, maximum emission current, and field intensification factor, the device with the 2×2 sub-array was superior to the one with the 5×5 sub-array. The VFET with 5×5 sub-arrays achieved a higher current density due to a larger number of field emitters packed per active emission area. Finally, plans to further improve the technology and transitioning into the fabrication of vacuum integrated circuits are discussed.</p> <p>  </p>

Indiana University-Purdue University Indianapolis (IUPUI) / While most sealants available are resin-based, glass ionomer-based cements can be used as sealants, with the advantage of being more tolerant to moisture during placement and of releasing fluoride. The objective of this study was to evaluate the influence of different fissure conditioning techniques on penetration and microleakage of glass ionomer (GI) and resin-modified glass ionomer cements (RMGI) used as sealants. Clinically sound extracted human molars were distributed into nine experimental groups (n = 15 each). Group 1 (control) was sealed with resin-based sealant (Delton) following clinically accepted techniques. Groups 2 through 6 were sealed with RMGI (Vitremer) after having the fissure conditioned with either polyacrylic acid (RMGI-control), 35-percent H3PO4, low viscosity 35-percent H3PO4 with a surfactant, self-etch conditioner, or 35-percent H3PO4 followed by self-etch conditioner. Groups 7 through 9 were sealed with GI sealant (Fuji Triage) after having the fissures conditioned with either polyacrylic acid (GI-control), 35- percent H3PO4 or low viscosity 35-percent H3PO4 with a surfactant. After aging through thermocycling (2500 cycles), specimens were incubated in methylene blue for four hours and sectioned at multiple locations. Digital images were obtained using a digital stereomicroscope, and microleakage was determined by scoring the dye penetration along the enamel-sealant interface. The penetration of the material was determined by calculating the percentage of the total length of the fissure penetrated by the material. Results: The use of self etch-conditioner significantly increased RMGI penetration, while surface conditioning with 35-percent phosphoric acid with surfactant significantly decreased microleakage of GI. The resin-based sealant placed after 35-percent phosphoric acid surface conditioning showed the best penetration and the least level of microleakage. In conclusion, results from this study suggest that the placement of glass ionomer-based sealants can be enhanced by modifying current conditioning methods.

Indiana University-Purdue University Indianapolis (IUPUI) / Traditional resin luting agents generally have mechanical properties that are superior to the newer so-called “universal” self-etching/self-adhesive resin luting agents. However, recent reports indicate that some properties of these new luting agents have been improved, approaching those of the traditional etch and rinse resin luting agents. The objective of this study was to test some mechanical properties of four of these self-etching/self-adhesive resin luting agents [Maxcem Elite (ME), Multilink Automix (MA), RelyX Unicem (RU), SmartCem 2 (SC)] and compare them to a traditional etch and rinse resin luting agent [RelyX ARC (RA)] and a resin-modified glass ionomer luting cement [Fuji Plus (FP)], both of which have much longer histories of clinical success. By comparing the properties of the newer cements to the standards, it may be possible to determine how clinically successful the newer cements may be. The mechanical properties tested were flexural strength (FS) and shear bond strength (SBS). The FS test included making beams of each material, storing them in water for periods of time (24 hours and 90 days) and then performing a three-point bending test on a universal testing machine. The 90 day groups were thermocycled. The SBS test involved preparing human molar specimens, making flat dentin surfaces. Composite cylinders were fabricated, luted to the dentin surfaces with each of the materials tested, stored in water for periods of time (24 hours or 90 days), and then a knife edge shear test was performed on a universal testing machine. The 90 day groups were thermocycled. A Weibull-distribution survival analysis was performed. The results revealed significant differences in the FS of all materials tested at 24 hours. After 90 days and thermocycling, only SC and RA were not significantly different. At both time periods, FP had the lowest and MA the highest FS. The SBS results showed MA, RA, and RU to have the highest bond strengths; SC and ME the lowest at 24 hours. After 90 days and thermocycling, RA had significantly higher bond strength than all other groups; ME, FP and SC had the lowest. The self-etching/self-adhesive resin luting agents all performed at least as well as FP, with the exception of SC (SBS 24 hour). They did not all perform as well as RA, with the exception of SC (FS 90 day), MA (SBS 24 hour, FS 24 hour and 90 day), and RU (SBS 24 hour). The newer luting agents should expect to have clinical success, regarding flexural strength and shear bond strength, at least as good as resin-modified glass ionomer luting cements and approach the level of traditional etch and rinse resin luting agents.

Indiana University-Purdue University Indianapolis (IUPUI) / Background: Numerous studies have reported the use of hydrofluoric (HF) acid as one of the most effective methods for the achievement of a durable bond between glass-based ceramics and resin cements. Nevertheless, there is little information available regarding the potential deleterious effect on the ceramic mechanical strength. Objectives: (1) to investigate the effect of HF acid etching regimens on the biaxial flexural strength of a low-fusing nanofluorapatite glass-ceramic (IPS e.max ZirPress, Ivoclar Vivadent), (2) to study the ability of an unfilled resin (UR) to restore the initial (i.e., before etching) mechanical strength, and (3) to evaluate the effect of HF acid etching on the ceramic surface morphology before and after UR treatment via scanning electron microscopy (SEM). Methods: One hundred and forty-four disc-shaped (15 ± 1 mm in diameter and 0.8 ± 0.1 mm in thickness) IPS e.max ZirPress specimens were allocated into 12 groups, as follows: G1-control (no etching), G2-30 s, G3-60 s, G4-90 s, G5-120 s, G6- 60 + 60 s. Meanwhile, groups (G7- G12) were treated in the same fashion as G1-G6, but followed by silane and UR applications. Surface morphology evaluation of non-etched and etched IPS e.max ZirPress (G1-G12) was carried out by scanning electron microscopy (SEM). The flexural strength was determined by biaxial testing as described in ISO 6872. Statistics were performed using two-way ANOVA and the Sidak multiple comparisons (α = 0.05). In addition, the Weibull statistics were estimated. Results: A significant effect of etching time (p=0.0290) on biaxial flexural strength was observed. Indeed, G4 led to a significantly (p=0.0392) higher flexural strength than G1. Correspondingly, G10 revealed a considerably higher flexural strength than G7 (p=0.0392). Furthermore, biaxial flexural strength was significantly higher for G7 – G12 than for G1 – G6 (p<0.0001). For G1 – G6, G4 showed the highest Weibull characteristic strength while the lowest Weibull characteristic strength was seen in G6. In G7 – G12, the highest Weibull characteristic strength was presented in G10 whereas G7 had the lowest. Finally, the SEM data revealed that the HF acid etching affected the surface of IPS e.max ZirPress by generating pores and irregularities and more importantly that the UR was able to penetrate into the ceramic microstructure. Conclusion: Within the limitations of this study, HF acid etching time did not show a damaging effect on the biaxial flexural strength of the IPS e.max ZirPress ceramic. Moreover, the ceramic biaxial flexural strength could be enhanced after UR treatment.

This dissertation presents the fabrication and analysis of in-fiber devices based on elliptical core D-shaped optical fiber. Devices created inside optical fibers are attractive for a variety of reasons including low loss, high efficiency, self-alignment, light weight, multiplexibility, and resistance to electromagnetic interference. This work details how D-fiber can be used as a platform for a variety of devices and describes the creation and performance of two of these devices: an in-fiber polymer waveguide and a surface relief fiber Bragg grating. In D-fiber the core is very close to the flat side of the ‘D’ shape. This proximity allows access to the fields in the fiber core by removal of the cladding above the core. The D-fiber we use also has an elliptical core, allowing for the creation of polarimetric devices. This work describes two different etch processes using hydrofluoric acid (HF) to remove the fiber cladding and core. For the creation of devices in the fiber core, the core is partially removed and replaced with another material possessing the required optical properties. For devices which interact with the evanescent field, cladding removal is terminated before acid breaches the core. Etching fibers prepares them for use in the creation of in-fiber devices. Materials are placed into the groove left when the core of a fiber is partially removed to form a hybrid waveguide in which light is guided by both the leftover core and the inserted material. These in-fiber polymer waveguides have insertion loss less than 2 dB and can potentially be the basis for a number of electro-optic devices or sensors. A polarimetric temperature sensor demonstrates the feasibility of the core replacement method. This work also describes the creation of a surface relief fiber Bragg gratings (SR-FBGs) in the cladding above the core of the fiber. Because it is etched into the surface topography of the fiber, a SR-FBG can operate at much higher temperatures than a standard FBG, up to at least 1100 degrees Celsius. The performance of a SR-FBG is demonstrated in temperature sensing at high temperatures, and as a strain sensor.