The history of silicon technology showed that controlled doping was a key step for the realization of e®ective, stable and reproducible devices. When the conduction type was no longer determined by impurities but could be controlled by doping, the breakthrough of classical microelectronics became possible. Unlike inorganic semiconductors, organic dyes are up to now usually prepared in a nominally undoped form. However, controlled and stable doping is desirable in many organic-based devices as well. If we succeed in shifting the Fermi level towards the transport states, this could reduce ohmic losses, ease carrier injection from contacts and increase the built-in potential of Schottky- or pn-junctions.

The history of silicon technology showed that controlled doping was a key step for the realization of e®ective, stable and reproducible devices. When the conduction type was no longer determined by impurities but could be controlled by doping, the breakthrough of classical microelectronics became possible. Unlike inorganic semiconductors, organic dyes are up to now usually prepared in a nominally undoped form. However, controlled and stable doping is desirable in many organic-based devices as well. If we succeed in shifting the Fermi level towards the transport states, this could reduce ohmic losses, ease carrier injection from contacts and increase the built-in potential of Schottky- or pn-junctions.

abstract: Optoelectronic and microelectronic applications of germanium-based materials have received considerable research interest in recent years. A novel method for Ge on Si heteroepitaxy required for such applications was developed via molecular epitaxy of Ge5H12. Next, As(GeH3)3, As(SiH3)3, SbD3, S(GeH3)2 and S(SiH3)2 molecular sources were utilized in degenerate n-type doping of Ge. The epitaxial Ge films produced in this work incorporate donor atoms at concentrations above the thermodynamic equilibrium limits. The donors are nearly fully activated, and led to films with lowest resistivity values thus far reported. Band engineering of Ge was achieved by alloying with Sn. Epitaxy of the alloy layers was conducted on virtual Ge substrates, and made use of the germanium hydrides Ge2H6 and Ge3H8, and the Sn source SnD4. These films exhibit stronger emission than equivalent material deposited directly on Si, and the contributions from the direct and indirect edges can be separated. The indirect-direct crossover composition for Ge1-ySny alloys was determined by photoluminescence (PL). By n-type doping of the Ge1-ySny alloys via P(GeH3)3, P(SiH3)3 and As(SiH3)3, it was possible to enhance photoexcited emission by more than an order-of-magnitude. The above techniques for deposition of direct gap Ge1-ySny alloys and doping of Ge were combined with p-type doping methods for Ge1-ySny using B2H6 to fabricate pin heterostructure diodes with active layer compositions up to y=0.137. These represent the first direct gap light emitting diodes made from group IV materials. The effect of the single defected n-i¬ interface in a n-Ge/i-Ge1-ySny/p-Ge1-zSnz architecture on electroluminescence (EL) was studied. This led to lattice engineering of the n-type contact layer to produce diodes of n-Ge1-xSnx/i-Ge1-ySny/p-Ge1-zSnz architecture which are devoid of interface defects and therefore exhibit more efficient EL than the previous design. Finally, n-Ge1-ySny/p-Ge1-zSnz pn junction devices were synthesized with varying composition and doping parameters to investigate the effect of these properties on EL. / Dissertation/Thesis / Doctoral Dissertation Chemistry 2016

Le silicium (Si) et le germanium (Ge) sont les matériaux de base utilisés dans les circuits intégrés. Cependant, à cause de leur gap indirect, ces matériaux ne sont pas adaptés à la fabrication de dispositifs d'émission de lumière, comme les lasers ou diodes électroluminescentes. Comparé au Si, le Ge pur possède des propriétés optiques uniques, à température ambiante son gap direct est de seulement 140 meV au-delà du gap indirect tandis qu'il est supérieur à 2 eV dans le cas du Si. Compte tenu du coefficient de dilatation thermique du Ge, deux fois plus grand que celui du Si, une croissance de Ge sur Si à hautes températures suivie d'un refroidissement à température ambiante permet de générer une contrainte en tension dans le Ge. Cependant, l'existence d'un désaccord de maille de 4,2% entre deux matériaux conduit à une croissance Stranski-Krastanov avec la formation des films rugueux et contenant de forte densité des dislocations. Nous avons mis en évidence l'existence d'une fenêtre de température de croissance, permettant de supprimer la croissance tridimensionnelle de Ge/Si. En combinant la croissance à haute température à des recuits thermiques par cycles, une contrainte de 0,30% a pu être obtenue. Le dopage de type n a été effectué en utilisant la décomposition de GaP, ce qui produit des molécules P2 ayant un coefficient de collage plus grand par rapport à celui des molécules P4. En particulier, en mettant en oeuvre la technique du co-dopage en utilisant le phosphore et l'antimoine, nous avons mis en évidence une augmentation de l'émission du gap direct du Ge à environ 150 fois, ce qui constitue l'un des meilleurs résultats obtenus jusqu'à présent. / Silicon (Si) and germanium (Ge) are the main materials used as active layers in microelectronic devices. However, due to their indirect band gap, they are not suitable for the fabrication of light emitting devices, such as lasers or electroluminescent diodes. Compared to Si, pure Ge displays unique optical properties, its direct bandgap is only 140 meV above the indirect one. As Ge has a thermal expansion coefficient twice larger than that of Si, tensile strain can be induced in the Ge layers when growing Ge on Si at high temperatures and subsequent cooling down to room temperature. However, due to the existence of a misfit as high as 4.2 % between two materials, the Ge growth on Si proceeds via the Stranski-Krastanov mode and the epitaxial Ge films exhibits a rough surface and a high density of dislocations. We have evidenced the existence of a narrow substrate temperature window, allowing suppressing the three-dimensional growth of Ge on Si. By combining high-temperature growth with cyclic annealing, we obtained a tensile strain up to 0.30 %. The n-doping in Ge was carried out using the decomposition of GaP to produce the P2 molecules, which have a higher sticking coefficient than the P4 molecules. In particular, by implementing a co-doping technique using phosphorus and antimony, we have evidenced an intensity enhancement of about 150 times of the Ge direct band gap emission. This result represents as one of the best results obtained up to now.

Dans le cadre de ce travail de thèse, nous avons étudié une approche permettant de réaliser les composants d'émission de la lumière basés sur les couches epitaxiées de Ge contraint en tension et fortement dopé de type n. Afin de créer de contrainte en tension dans les films épitaxiés de Ge, nous avons investi deux méthodes : faire croître du Ge sur InGaAs ayant un paramètre de maille plus grand que celui de Ge, et faire croître du Ge sur Si, en prenant l'avantage du coefficient de dilatation thermique du Ge, qui est deux fois plus grand que celui du Si. Concernant la croissance de Ge sur les substrats Si, nous avons étudié deux orientations cristallines, <001> and <111>, afin de pouvoir comparer la valeur de contrainte en tension obtenue et aussi la densité des dislocations émergeantes. Le dopage de type n dans le Ge a été effectué en utilisant le phosphore et l'antimoine. Nous avons montré que quand le dopage est effectué à des températures relativement basses et suivi d'un recuit thermique rapide, de concentration d'électrons électriquement activés de ~ 4x10^19 cm-3, a pu être obtenue. Cette valeur représente l'un des meilleurs résultats expérimentalement obtenus jusqu'à présent. Des mesures de recombinaison radiative par spectroscopie de photoluminescence effectuées à température ambiante ont mis en évidence une augmentation de l'émission du gap direct de Ge d'environ 150 fois. Finalement, nous avons étudié les effets de la barrière de diffusion sur l'efficacité de dopage pendant les recuits thermiques. Une comparaison sur l'efficacité de trois barrières de diffusion, Al2O3, HfO2 and Si3N4, sera présentée et discutée. / During my thesis, we studied approaches to achieve light-emitting devices based on tensile strained and highly n-doped Ge epitaxial films. In order to create an elastic tensile strain in the epitaxial Ge films, we have investigated two methods: The epitaxial growth of Ge on InGaAs buffer layers that have a larger lattice constant, and the epitaxial growth of Ge on Si, by which we take benefit of the thermal expansion coefficient of Ge which is twice greater than that of Si. Concerning the growth of Ge on Si substrates, we have studied two crystalline orientations, <001> and <111>, in order to compare the values of the accumulated tensile strain and also the density of threading dislocations. The n-type doping in Ge was performed using a co-doping technique with phosphorus (P2 molecule) and antimony (Sb). We demonstrated that the dopants sticking coefficient leads to dopant incorporation in the Ge film larger than their solid solubility, which generally increases with increasing substrate temperature. As a result, when the doping is carried out at relatively low temperatures and followed by rapid thermal annealing, electrically activated electron concentration of 4x1019 cm-3 was demonstrated. This value is one of the best results obtained experimentally so far. The radiative recombination, at RT, measured by photoluminescence spectroscopy showed an increase in the direct gap emission of Ge of about 150 times. Finally, we studied the effects of diffusion barrier on the doping concentration during the thermal annealing. A comparison between the advantages of three diffusion barriers, Al2O3, HfO2 and Si3N4, will be presented and discussed.

Carbon based nanostructures, such as fullerenes, carbon nanotubes and graphene showed a high potential for a vast of electronic and energy applications. However, properties of such materials in pristine forms can be insufficient to satisfy diverse specific demands, and tailoring their intrinsic properties is of increasing importance. In this work, different types of single-walled carbon nanotubes (SWCNTs) with controlled semiconducting fractions are p-/n-type doped by chemical doping in an attempt to tailor physical properties of the SWCNTs for the use in flexible thermoelectric (TE) devices and thermoplastic polymer-based conducting composites. Several p-/n-type doping schemes and an electronic type separation strategy have been developed to fulfill the task. A complete solution for efficient and scalable production of doped SWCNTs for the fabrication of flexible thermoelectric components is developed in this work. For p-type doping, a combined experimental and theoretical work demonstrates that boron atomic doping is an efficient way to simultaneously improve Seebeck coefficient (S) and electrical conductivity (σ) of SWCNT films, showing an increased thermoelectric power factor (S2σ) up to 255 μW/mK2 by a factor of 2.5 comparing to the pristine SWCNTs. For n-type doping, treatment of SWCNTs with potassium oxide and crown ether solution lead to a negative Seebeck coefficient of -30 μV/K and a promising S2σ up to 50 μW/mK2. A gel chromatography method has been developed to separate large-diameter (1.2-1.8nm) SWCNTs by electronic properties and to increase the purity of the sorted semiconducting carbon nanotubes (sc-SWCNTs) up to 95%. Effects of p-/n-type doping induced by different plasma treatments on the thermoelectric properties have been investigated for thin films made of sorted sc-SWCNTs. The high-purity sc-SWCNTs show significantly improved S of 125 μV/K. As the effects of p-type doping, air plasma treatments with proper duration (40s) lead to the increase of S, σ and thus S2σ up to 190 μW/mK2. The n-type doping for the SWCNT films have been performed via ammonia plasma treatment, and a negative S value of -80 μV/K has been achieved in air at 110oC, which is one of the best values ever reported for n-type carbon nanotube films. A flexible thermoelectric module was fabricated by printing ink made of the prepared boron doped SWCNTs and an organic solvent as an example for producing efficient all-carbon thermoelectric generators. At a temperature difference ΔT=60 K, the output voltage reaches 20 mV and the power output of 400 nW is obtained, although no “n”-legs are used in this module. At last, a work has been done on the development of melt mixed composites as TE materials, in which polypropylene is used as the matrix and boron-doped SWCNTs are used as conducting fillers. A percolation threshold lower than 0.25wt. % and a maximum conductivity up to 125 S/m at 5wt. % of SWCNT load have been achieved. The maximum conductivity is more than two times higher than that of the composites made with pristine SWCNTs as fillers.

Germanium (Ge) is the most compatible semiconductor material with silicon-based complementary metal-oxide semiconductor technology, which has higher electron and hole mobility than Si, leading to enhanced device performance. In addition, semiconductor nanowires (NWs) have attracted significant attention as promising candidates for next-generation nanoscale devices. Due to their unique geometry and physical properties, NWs show excellent optical and electrical properties such as quantum size effects, enhanced light absorption, and high biological and chemical sensitivity. Furthermore, high response to light irradiation is one of the most significant properties of semiconductor NWs, which makes them excellent candidates for photodetectors. Hence, Ge NWs are promising high-mobility nanostructures for optoelectronic devices. Despite constant improvement in the performance of single NW-based devices, determining their electrical properties remains challenging. Here, a symmetric six-contact Hall bar configuration is developed for top-down fabricated highly doped Ge NWs with different widths down to 30 nm, which simultaneously facilitates Hall effect and four-probe resistance measurements. Furthermore, accurate control of doping and fabrication of metal contacts on n-type doped Ge NWs with low resistance and linear characteristics remain significant challenges in Ge-based devices. Therefore, a combined approach is reported to fabricate Ohmic contacts on n-type doped Ge NWs using ion implantation and rear-side flash lamp annealing. This approach allows the fabrication of axial p–n junctions along the single NWs with different widths. The fabricated devices demonstrated rectifying characteristics in dark conditions. The photoresponse of the axial p–n junction photodetectors was investigated under three different illumination wavelengths of 637 nm, 785 nm, and 1550 nm. Moreover, the fabricated axial p–n junction photodetector demonstrated a high-frequency response up to 1 MHz at zero bias.

L’oxyde de zinc ZnO, est un semiconducteur II-VI très prometteur pour les applications en opto-électronique dans le domaine UV, notamment pour la réalisation de dispositifs électroluminescents (LED). Les potentialités majeures du ZnO pour ces applications résident notamment dans sa forte liaison excitonique (60 meV), sa large bande interdite directe (3.4 eV), la disponibilité de substrats massifs de grand diamètre ainsi que la possibilité de réaliser des croissances épitaxiales de très bonne qualité en couches minces ou nano structurées (nanofils). Néanmoins, le développement de ces applications est entravé par la difficulté de doper le matériau de type p. L'impureté permettant d'obtenir une conductivité électrique associée à des porteurs de charges positifs (trous), et donc la réalisation de jonctions pn à base de ZnO, n'a pas encore été réellement identifiée. C'est pourquoi une des étapes préliminaires et nécessaires à l'obtention d'un dopage de type p fiable et efficace, réside dans la compréhension du dopage résiduel de type n, ainsi que des phénomènes de compensation et de passivation qui sont mis en jeu au sein du matériau. La maîtrise de la nature des contacts (ohmique ou Schottky) sur différentes surfaces d'échantillons de ZnO nous a permis dans ce but de mettre en œuvre à la fois des mesures de transport (résistivité et effet Hall) et des mesures capacitives (capacité-tension C(V), Deep Level Transient Spectroscopy (DLTS) et Spectroscopie d'admittance).Dans un premier temps, nous avons donc cherché à comprendre de manière approfondie les propriétés électriques du ZnO massif. Nous avons ainsi étudié le rôle des défauts profonds et peu profonds sur la conductivité des échantillons, aux travers de différents échantillons massifs obtenus par synthèse hydrothermale ou par croissance chimique en phase vapeur. Nous avons également étudié l'impact de la température de recuits post-croissance, sur les propriétés de transport des échantillons. A la lumière des résultats obtenus sur le dopage résiduel de type n des échantillons de ZnO massifs, nous avons ensuite procédé à différents essais de dopage de type p du ZnO par implantation ionique d'azote et par diffusion en ampoule scellée d’arsenic. L'impureté azote a été choisie dans le cadre d'une substitution simple de l'oxygène qui devrait permettre de créer des niveaux accepteurs dans la bande interdite du ZnO. Nous avons également étudié l'impureté arsenic, qui selon un modèle théorique peut former un complexe qui permet d'obtenir un niveau accepteur plus proche de la bande de valence que le niveau. Outres les études réalisées sur les échantillons de ZnO massif et les essais de dopage de type p, nous avons également étudié les propriétés électriques d'échantillons de ZnO monocristallins sous forme de couches minces obtenues par croissance en phase vapeur d’organométalliques, dopées intentionnellement ou non. Les corrélations entres les mesures SIMS et C(V) nous ont permis notamment de mettre en évidence une diffusion et un rôle très importante de l'aluminium sur les propriétés électriques des couches minces de ZnO épitaxiées sur substrat saphir.Dans le cadre de cette thèse nous avons réussi à clarifier les mécanismes du dopage de type n, intentionnel ou non intentionnel, dans le ZnO monocristallin. Nous avons également identifié les impuretés et les paramètres de croissance importants permettant d'obtenir un dopage résiduel de type n le plus faible possible dans les couches épitaxiées. Cette maitrise du dopage résiduel de type n est une étape préliminaire indispensable aux études de dopage de type p car elle permet de minimiser la compensation des accepteurs introduits intentionnellement. Cette approche du dopage sur des couches minces de ZnO dont le dopage résiduel de type n est très faible apparait comme une voie très prometteuse pour surmonter les problèmes d'obtention du dopage de type p. / Zinc oxide (ZnO) is a II-VI semiconductor which appears as a very promising material for UV opto-electronic applications, in particular for the production of light emitting devices (LED). For these applications, ZnO presents strong advantages as a high exciton binding energy (60 meV ), a wide direct band gap (3.4 eV), the availability of large diameter bulk substrates for homoepitaxial growth of high quality thin films or nanostructures. However, the development of these applications is hampered by the difficulty to dope ZnO p-type. The impurity leading to an electrical conductivity associated with positive charge carriers (holes), and therefore the production of ZnO pn junctions have not yet been really identified.In this thesis we have studied the physical mechanisms that govern the electrical properties of ZnO single crystal and epilayers. The control of contacts (ohmic or Schottky) on different ZnO surfaces allowed us to carry out both transport measurements (resistivity and Hall effect) and capacitance measurements (C(V), Deep Level Transient Spectroscopy (DLTS) and admittance spectroscopy).At first, we have studied the role of deep and shallow defects on the n-type conductivity of bulk ZnO samples obtained by Hydrothermal synthesis (HT) or by Chemical Vapor Transport (CVT). We also investigated the impact of post-growth annealing at high temperature under oxygen atmospheres on the transport properties of samples. Thanks to the previous results on the residual n-type doping, we have reported on several attempts to obtain p-type ZnO. We have discussed the potential of different candidates for the achievement of p-type doping and present our tentative experiments to try and demonstrate the reality, the ability and the stability of p-type doping by nitrogen implantation and arsenic diffusion. The nitrogen impurity has been chosen for oxygen substitution, which should allow the creation of acceptor levels in the ZnO band gap. We also studied arsenic as a potential p-type dopant, according to a model whereby arsenic substitutes for oxygen and, if associated with two zinc vacancies, forms a complex with a shallower ionization energy than in the case of direct oxygen substitution.In addition to the studies on bulk ZnO samples and attempts on p-type doping, we have also studied the electrical properties of thin film ZnO samples obtained by Metal Organic Vapor Phase Epitaxy, either intentionally or unintentionally doped. Correlations between SIMS and C(V) measurements allowed us to highlight especially the importance of aluminum as a residual impurity in epitaxial layers grown on sapphire substrates.In this thesis we have clarified intentional or unintentional n-type doping mechanisms in ZnO single crystal samples. We have also identified impurities and growth parameters responsible for the residual n-type doping. This understanding is a crucial and preliminary step for understanding the doping mechanisms at stake in this material and is also necessary to achieve stable p-type conductivity, which is still the main challenge for the realization of optoelectronic devices based on ZnO.