• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 3
  • 1
  • Tagged with
  • 4
  • 4
  • 3
  • 3
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • 2
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Développement de méthodes ex-situ de dopage de nanofils semiconducteurs IV / Development of ex-situ doping methods of group IV semiconductor nanowires

Fakhfakh, Mariam 31 January 2018 (has links)
L’objet de cette thèse est d’étudier le dopage ex-situ de nanofils semiconducteurs IV pour des applications en électronique, spintronique ou encore thermoélectricité. Deux techniques de dopage ont été explorées : l’implantation par faisceaux d’ions et le Spin-On-Doping (SOD).L’implantation d’ions Mn a été testée dans les nanofils de Ge avec l’objectif de synthétiser un matériau semi-conducteur ferromagnétique dilué. Une concentration en Mn de quelques pourcents peut être atteinte sans amorphisation du fil ni formation de précipités, ce qui est très encourageant. Lors d’expériences d’implantation réalisées in situ dans un microscope électronique en transmission, une forte exaltation de la pulvérisation sous irradiation électronique a été constatée.La technique SOD consiste à faire diffuser thermiquement les impuretés de type p ou n contenues une résine de type HSQ (Hydrogen silsesquioxane) qui enrobe les nanofils. Le recuit de la HSQ (dopée ou non) engendre une modification structurale des nanofils (bien que cette technique soit considérée comme non destructive). Lors du recuit, une transformation partielle de la phase diamant 3C vers la phase hexagonale 2H, a en effet été observée dans les nanofils de Si et de Ge, au-delà de 500 et 400°C respectivement. Les paramètres essentiels de la transformation de phase sont la contrainte de cisaillement résultant de la densification de la résine et le budget thermique. Les nanofils de Ge deviennent amorphes au-delà de 650°C, ce qui interdit en pratique leur dopage par SOD.Les caractérisations électriques ont été réalisées sur des nanofils de Si réalisés par gravure ionique réactive sur substrats orientés (111) et contactés en matrice ou individuellement. Pour le contactage de nanofils uniques en configuration NW-FET (nanowire field effect transistors), un procédé technologique basé sur la lithographie électronique a dû être développé. Les difficultés à surmonter étaient relatives à la faible longueur des nanofils. Diverses techniques de caractérisation ont été mises en œuvre (I-V en configuration verticale ou horizontale de type TLM (Transient Linear Measurement), SSRM (scanning spreading resistance microscopy), EBIC (electron beam induced current). Les mesures collectives concernent des ensembles de nanofils de type p enrobés dans une résine qu’elle soit dopante ou non. Pour observer un courant notable dans la structure, un recuit est nécessaire. Au-delà d’une température de recuit de 600°C, une polarisation négative du substrat induit un comportement conforme au mécanisme SCLC (space charge limited current) attendu pour des nanofils faiblement dopés enrobés dans une matrice isolante. En positif, on observe une caractéristique I(V) ohmique et une densité de courant jusqu’à 500 fois plus élevée dans les nanofils que dans le substrat. Ce comportement pourrait être dû à l’influence des états d’interface provenant de la technique de gravure. Cette hypothèse est confortée par le fait qu’après recuit à 900°C, le courant en direct s’explique en considérant dans les fils un dopage proche de celui du substrat, et surtout par l’observation en SSRM d’une couche conductrice interfaciale entre fils et HSQ. Elle permet aussi d’interpréter les mesures sous pointes faites sur les fils de type n. Le mode de transport SCLC a également été observé pour des nanofils individuels contactés sous pointe ou par lithographie. Ces mesures n’ont pas mis directement en évidence l’influence de la transformation de phase.Le dopage de type n ou p par SOD s’avère efficace après recuit à 900°C. Dans ce cas, les comportements observés, contacts ohmiques et jonctions p-n, peuvent être interprétés plus simplement en considérant des niveaux de dopage supérieurs à 3×10¹⁶ cm⁻³ en type p et 2×10¹⁶ cm⁻³ en type n. Ces valeurs déduites des résistivités mesurées sont sans doute très sous-estimées puisque la mobilité dans les fils est sans doute inférieure à celle du volume. / This thesis aims at studying the ex-situ doping of semiconducting nanowires (NWs) for applications in electronics, spintronics or thermoelectricity. Two widely used techniques have been envisaged: ion beam implantation and Spin-On-Doping (SOD).The ion beam implantation of Mn ions has been tested in Ge NWs in an attempt to form a 1D diluted magnetic semiconductor structure. A Mn concentration of few percents can be achieved without amorphization of the nanowire nor clustering, what is very promizing. During implantation done in situ in a transmission electron microscope, a strong enhancement of the sputtering under electron irradiation has been observed.The doping by SOD results from the thermal diffusion of p-type or n-type impurities contained in a HSQ (Hydrogen silsesquioxane) resist in which the NWs are embedded. The curing of the HSQ resist (doped or not) leads to a structural modification of nanowires (while SOD is generally assumed to be non-destructive). As a result of the annealing, a partial transformation of the 3C diamond phase towards the 2H hexagonal phase is observed in Si and Ge nanowires, above 500 et 400°C respectively. The main parameters of that phase transformation are the shear stress due to the HSQ densification and the thermal budget. Ge NWs are found to turn to amorphous above 650°C, what renders SOD practically unusable for Ge NWs. Two methods are currently used for the fabrication of nanowires, the VLS (vapor-liquid-solid) growth and reactive ion etching of (111) Si wafers. For practical reasons, etched NWs were used for the study of their electrical properties.The electrical characterizations were done on arrays of Si NWs embedded in a HSQ matrix or on single NWs. For contacting single NWs in the NW-FET(nanowire field effect transistors) configuration , a process based on electron beam lithography has been developed. The issues to be solved were related to the low length of NWs. Various measurement techniques were used: I-V in two tips or TLM (Transient Linear Measurement) arrangement, SSRM (scanning spreading resistance microscopy), EBIC (electron beam induced current). Collective measurements were done on arrays of p-type NWs embedded in a HSQ resist, doped or not. It was firstly observed that an annealing is needed to observe a noticeable current in the structure. Above an annealing temperature of 600°C, for a negative bias applied to the substrate, the observed behavior can be described by the SCLC (space charge limited current) mechanism expected for poorly doped NWs in an isolating matrix, while a positive bias applied to the substrate results in an ohmic characteristic and in a current density up to 500 times higher in the NWs than in the substrate. This unexpectedly high intensity in direct bias may be attributed to electrically active surface states resulting from the etching process. This hypothesis is conforted by the fact that an annealing at 900°C (without extra doping) the measured intensity can be explained by assuming the same doping level in NWS than in the substrate. In addition, an interfacial conductive between resist and nanowires can be observed by SSRM. These interfacial states can be also involved for understanding the measurements done on n-type NWS. The SCLC mechanism of transport has been also observed for single NWs contacted by tips or by lithographied contacts. These measurements were not able to evidence the effect of the phase transformation on the electrical properties.P-type and n-type doping by SOD becomes effective after annealing at 900°C. After doping, ohmic or rectifying behaviours on p-type substrate are observed as expected. That renders more easy the interpretation of results, by assuming doping levels in the NWs of 3×10¹⁶ cm⁻³ and 2×10¹⁶ cm⁻³ for p-type and n-type respectively. These values as deduced from resistivities are probably very underestimated as the mobilities in NW are probably much lower than in the bulk.
2

Chemical Bath Deposition Of Group Ii-vi Semiconductor Thin Films For Solar Cells Applications

Khallaf, Hani 01 January 2009 (has links)
Chemical bath deposition (CBD) is the analog in liquid phase of the well-known chemical vapor deposition technique in the vapor phase. In CBD, deposition of thin films takes place from aqueous solutions at low temperatures by a chemical reaction between dissolved precursors, with the help of a complexing agent. Among all techniques used to grow Group II-VI semiconductors, CBD has the advantage of being a simple, low temperature, and inexpensive large-area deposition technique. So far, its contribution in thin film solar cells industry has been mainly limited to growing n-type CdS and/or ZnS window layers for CdTe-based and CIGS-based solar cells. In this work we first optimize the CBD process of CdS using nitrilotriacetic acid and hydrazine as complexing agents as an alternative to ammonia. We then study the effect of the cadmium precursor on the optical/electrical properties, as well as crystal structure, morphology, and composition of CBD-CdS films. A better understanding of the CBD process of CdS as a whole has been achieved and high quality CBD-CdS films have been obtained. Next, we investigate in-situ doping of CBD-CdS with group III elements, such as B, Al, In, and Ga. The objective is to show that CBD is capable of not only growing CdS but also of doping it to reduce its resistivity and, as a result, facilitate its use in solar cells as well as other optoelectronic device fabrication. A four orders of magnitude drop of film resistivity has been achieved without a significant change in film bandgap, structure, or morphology. Finally, we test the possibility of using CBD to grow transparent conducting oxide (TCO) films, such as Al-doped ZnO films and cadmium stannate films. First, we study CBD of ZnO and later in-situ doping of ZnO using Al. High quality ZnO thin films have been grown using CBD with the help of four different complexing agents. Post heat treatment in argon ambient helped reduce resistivity of CBD-ZnO undoped films to ~ 10-1 Ω-cm. In-situ doping of such films using Al shows promising results. Such films could be an alternative to indium tin oxide (ITO) layers that are commonly used as TCO layers for solar cells. Another approach is to use CBD to grow CdO and SnO2 thin films, with the goal of obtaining Cd2SnO4 by later annealing of these two layers. Cadmium stannate is another TCO candidate that could replace ITO in the near future. We have succeeded in growing CBD-CdO thin films using three different complexing agents. Undoped CBD-CdO films with a resistivity as low as 1.01 x10-2 Ω-cm and a carrier density as high as 2.59 x 1020 cm-3 have been obtained. SnO2 films have been successfully grown using CBD. Fabrication of Cadmium stannate thin films using CBD is investigated. In summary, our objective to expand the use of CBD beyond just growing CdS and ZnS, and to test the possibility of using it for in-situ doping of group II-VI semiconductors as well as TCO layers fabrication proved to be successful. We believe that this may have a significant impact on solar cells as well as other optoelectronic devices fabrication industry, due to the simplicity and the cost-effectiveness of CBD.
3

Chemical Vapor Depositionof Si and SiGe Films for High-Speed Bipolar Transistors

Pejnefors, Johan January 2001 (has links)
This thesis deals with the main aspects in chemical vapordeposition (CVD) of silicon (Si) and silicon-germanium (Si1-xGex) films for high-speed bipolar transistors.In situdoping of polycrystalline silicon (poly-Si)using phosphine (PH3) and disilane (Si2H6) in a low-pressure CVD reactor was investigated toestablish a poly-Si emitter fabrication process. The growthkinetics and P incorporation was studied for amorphous Si filmgrowth. Hydrogen (H) incorporated in the as-deposited films wasrelated to growth kinetics and the energy for H2desorption was extracted. Film properties such asresistivity, mobility, carrier concentration and grain growthwere studied after crystallization using either furnaceannealing or rapid thermal annealing (RTA). In order tointegrate an epitaxial base, non-selective epitaxial growth(NSEG) of Si and SiGe in a lamp-heated single-waferreduced-pressure CVD reactor was examined. The growth kineticsfor Si epitaxy and poly-Si deposition showed a differentdependence on the deposition conditions i.e. temperature andpressure. The growth rate difference was mainly due to growthkinetics rather than wafer surface emissivity effects. However,it was observed that the growth rate for Si epitaxy and poly-Sideposition was varying during growth and the time-dependencewas attributed to wafer surface emissivity variations. A modelto describe the emissivity effects was established, taking intoconsideration kinetics and the reactor heating mechanisms suchas heat absorption, emission andconduction. Growth ratevariations in opening of different sizes (local loading) andfor different oxide surface coverage (global loading) wereinvestigated. No local loading effects were observed, whileglobal loading effects were attributed to chemical as well astemperature effects. Finally, misfit dislocations formed in theSiGe epitaxy during NSEG were found to originate from theinterface between the epitaxial and polycrystalline regions.The dislocations tended to propagate across the activearea. <b>Keywords:</b>chemical vapor deposition (CVD), bipolarjunction transistor (BJT), heterojunction bipolar transistor(HBT), silicon-germanium (SiGe), epitaxy, poly-Si emitter,in situdoping, non-selective epitaxy (NSEG), loadingeffect, emissivity effect
4

Chemical Vapor Depositionof Si and SiGe Films for High-Speed Bipolar Transistors

Pejnefors, Johan January 2001 (has links)
<p>This thesis deals with the main aspects in chemical vapordeposition (CVD) of silicon (Si) and silicon-germanium (Si<sub>1-x</sub>Ge<sub>x</sub>) films for high-speed bipolar transistors.<i>In situ</i>doping of polycrystalline silicon (poly-Si)using phosphine (PH<sub>3</sub>) and disilane (Si<sub>2</sub>H<sub>6</sub>) in a low-pressure CVD reactor was investigated toestablish a poly-Si emitter fabrication process. The growthkinetics and P incorporation was studied for amorphous Si filmgrowth. Hydrogen (H) incorporated in the as-deposited films wasrelated to growth kinetics and the energy for H<sub>2</sub>desorption was extracted. Film properties such asresistivity, mobility, carrier concentration and grain growthwere studied after crystallization using either furnaceannealing or rapid thermal annealing (RTA). In order tointegrate an epitaxial base, non-selective epitaxial growth(NSEG) of Si and SiGe in a lamp-heated single-waferreduced-pressure CVD reactor was examined. The growth kineticsfor Si epitaxy and poly-Si deposition showed a differentdependence on the deposition conditions i.e. temperature andpressure. The growth rate difference was mainly due to growthkinetics rather than wafer surface emissivity effects. However,it was observed that the growth rate for Si epitaxy and poly-Sideposition was varying during growth and the time-dependencewas attributed to wafer surface emissivity variations. A modelto describe the emissivity effects was established, taking intoconsideration kinetics and the reactor heating mechanisms suchas heat absorption, emission andconduction. Growth ratevariations in opening of different sizes (local loading) andfor different oxide surface coverage (global loading) wereinvestigated. No local loading effects were observed, whileglobal loading effects were attributed to chemical as well astemperature effects. Finally, misfit dislocations formed in theSiGe epitaxy during NSEG were found to originate from theinterface between the epitaxial and polycrystalline regions.The dislocations tended to propagate across the activearea.</p><p><b>Keywords:</b>chemical vapor deposition (CVD), bipolarjunction transistor (BJT), heterojunction bipolar transistor(HBT), silicon-germanium (SiGe), epitaxy, poly-Si emitter,<i>in situ</i>doping, non-selective epitaxy (NSEG), loadingeffect, emissivity effect</p>

Page generated in 0.0719 seconds