Les plasmas radio-fréquences à couplage inductif en gaz halogénés simples (cl2/hbr/o2) sont fortement utilisés dans l'industrie des semi-conducteurs. Cependant, notre connaissance des processus réactionnels de ces plasmas est encore très partielle. De nombreux travaux de simulations (fluides, particulaires...), visant à améliorer celle-ci, ont été produits ces dernières décennies. Toutefois, trop peu de résultats expérimentaux sont disponibles dans la littérature afin de valider ou améliorer ces simulations. L'objectif de cette thèse est alors de produire un ensemble complet de résultats experimentaux. Nous nous focalisons essentiellement sur le cas des plasmas de cl2 et de o2 purs. Dans cette thèse, nous étendons les résultats expérimentaux déjà présents dans la littérature : densités de charges et de neutres, températures translationnelles. En particulier, les tendances en fonction de la pression, essentielles pour la simulation, sont soigneusement étudiées. Les vibrations moléculaires sont également étudiées à l'aide d'un montage innovant de spectroscopie d'absorption. Nous montrons que les simulations sont encore loin de représenter fidèlement les processus réactionnels des plasmas étudiés. Nous tentons de fournir à cet effet quelques pistes d'améliorations. Ce travail est la base nécessaire à l'amélioration continue des plasmas industriels utilisés pour la gravure de semi-conducteurs. / Radio-frequency inductively-coupled plasmas in simple halogen gases (cl2/hbr/o2) are widely used in the semi-conductor industry. However, our knowledge of these plasmas is still incomplete. To improve it, numerous simulation studies have been performed in the last decades. Unfortunately, experimental results to compare these studies are still scarce. The objectives of this thesis is to provide a comprehensive set of experimental results. We focused on the plasmas of pure o2 and cl2. In this thesis, we extend the already available experimental results : neutral nd charges densities, translational temperatures. In particular, the tendancies of these parameters as a function of the pressure are carefully studied. Molecular vibrations are studied as well using a new kind of absorption spectroscopy setup. We show that the recent simulations are still far from representing the reactional processes in the studied plasmas. We then try to provide some ideas of improvement. This work is the needed start to improve etching plasma industrial processes.

Made available in DSpace on 2016-06-02T19:22:49Z (GMT). No. of bitstreams: 1 90.pdf: 4644249 bytes, checksum: ebe3343d757ad085f8a5f1446b042caa (MD5) Previous issue date: 2003-02-28 / Universidade Federal de Minas Gerais / Deficit of water and sodium in the body is detected by receptors located in different parts of the body. These receptors or hormones signalize to specific areas in the brain that control renal responses and water and sodium intake. Among these areas are: organum vasculosum of the lamina terminalis (OVLT), subfornical organ (SFO), anteroventral third ventricle (AV3V) region, hypothalamus, amygdala, septal area (SA), nucleus of the solitary tract (NTS), area postrema (AP) and lateral parabrachial nucleus (LPBN). Besides the regulation of fluid and electrolytic balance, these areas are also involved in cardiovascular control. Angiotensin II (ANG II) is a peptide that induces water and sodium ingestion and participates in cardiovascular regulation. Other neurotransmitters, like serotonin, cholecystokinin and atrial natriuretic peptide, can inhibit water and sodium intake. Another important inhibitory mechanism for water and sodium intake is related to central α2-adrenergic/imidazoline receptors. Central injection of the anti-hypertensive drugs, moxonidine and clonidine (α2-adrenergic/imidazoline receptor agonists), reduces water and sodium intake in different protocols (water dehydration, 24 h sodium depletion and administration of ANG II). So, the goals of this study were: a) to study the effects of moxonidine injected into the cerebral ventricles (lateral ventricle - LV and 4th ventricle - 4th V), amygdaloid complex, central nucleus of amygdala (CNA), basal nucleus of amygdala (BNA) and lateral hypothalamus (LH) on 0.3 M NaCl intake induced by sodium depletion (treatment combining subcutaneous injection of furosemide + sodium deficient food for 24 h); b) to test the effects of moxonidine injected into LV, 4th V and LH on the pressor response produced by the injection of ANG II and carbachol (cholinergic agonist) into the LV; c) to investigate the participation of α2-adrenergic receptors on the effects of moxonidine, injected into LV, on sodium depletion-induced 0.3 M NaCl intake and ANG II-induced pressor response; d) using imunohistochemical technique, to detect c-fos protein in forebrain areas after moxonidine injection into LV in normovolemic rats, sodium depleted rats or rats that were treated with central injection of ANG II. Male Holtzman rats with a stainless steel guide-cannulas implanted into the cerebral ventricles: LV (volume of injection: 1-3 µl) and 4th V (volume of injection: 1 µl); unilaterally into LH (volume of injection: 0.5 µl) and bilaterally into the amygdaloid complex (volume of injection: 1 µl), CNA (volume of injection: 0.2-0.4 µl) and BNA (volume of injection: 0.2-0.4 µl) were used. Moxonidine (20 nmol) injected into LV reduced sodium depletion-induced 0.3 M NaCl intake during all the period of the experiment (120 min), while moxonidine injected into 4th V, reduced 0.3 M NaCl intake only in the first 60 min. Bilateral injections of moxonidine (5, 10 and 20 nmol/1 µl) into amygdaloid complex and BNA (20 nmol/0.4 µl), but not into CNA, reduced sodium depletion-induced 0.3 M NaCl intake. Unilateral injection of moxonidine into LH did not change sodium depletion-induced 0.3 M NaCl intake. These results show that the activation of α2-adrenergic/imidazoline receptors produced by the injection of moxonidine into LV, 4th V and amygdaloid complex (especially into the BNA), but not into LH and CNA, reduce hypertonic NaCl intake. To investigate the role of the α2-adrenergic receptors on the inhibitory effect of moxonidine on 0.3 M NaCl intake, specific α2-adrenergic receptor antagonists, such as RX 821002, yohimbine and SKF 86466, were combined with moxonidine. The results show that icv injection of RX 821002 (40 and 80 nmol) and SKF 86466 (640 nmol) abolished the inhibitory effect of moxonidine (20 nmol) on 0.3 M NaCl intake during all the period of the experiment, while yohimbine (320 nmol) abolished the antinatriorexigenic effect of moxonidine only in the last hour of the experiment (60 to 120 min). These results suggest the involvement of central α2-adrenergic receptors on the inhibitory effect of moxonidine on sodium depletion-induced 0.3 M NaCl intake. Besides, we observed an increase on sodium depletion-induced 0.3 M NaCl intake following the treatment with RX 821002 (40 nmol) and yohimbine (320 nmol) alone, that suggests a possible tonic function to the central α2-adrenergic receptors on the control of hypertonic NaCl intake. The injection of moxonidine alone (respectively, 20 and 80 nmol) into LH and VL did not change mean arterial pressure (MAP) and heart rate (HR), while moxonidine administered into 4th V produced hypotension and bradycardia. The 80 nmol dose of moxonidine injected into LV reduced the pressor response produced by central injections of ANG II (50 ng) and carbachol (4 nmol). Moxonidine (20 nmol) injected into LH and 4th V reduced the ANG II-induced pressor response, but not carbachol-induced pressor response. So, it was demonstrated that central injection of moxonidine reduces the pressor responses produced by angiotensinergic (mainly) and cholinergic activation (in a minor degree). The injection of yohimbine (320 nmol) into the LV abolished the inhibitory effect of moxonidine (80 nmol), also injected into LV, on the pressor response produced by icv injection of ANG II, suggesting that moxonidine acting through central α2-adrenergic receptors inhibits ANG IIinduced pressor response. The injection of moxonidine into LV in normovolemic and satiated rats induced the expression of c-fos protein in the following areas: OVLT, ipslateral lateral septal area (ipsLSA), ventral median preoptic nucleus (vMPN), paraventrivular nucleus (PVN) and supraoptic nucleus (SON). These areas are involved in the fluid and electrolytic balance and cardiovascular regulation. ANG II injected into LV produced c-fos expression in the following areas: ipsLSA, dorsal median preoptic nucleus (dMPN), PVN and SHN and reduced cfos expression in contLSA (contra lateral lateral septal area). Previous injection of moxonidine did not change the c-fos protein expression induced by central injection of ANG II. Separated treatments with moxonidine and ANG II produce c-fos expression in similar areas, so it is difficult to know which treatment is responsible to c-fos protein expression observed after the combination of the two treatments. Maybe, moxonidine could be inhibiting the c-fos expression induced by ANG II and the c-fos expression noted after the combined treatment could be produced only by moxonidine. In sodium depleted rats, icv injection of moxonidine induced an increase on c-fos expression in ipsLSA and dMPN, and a decrease in OVLT, suggesting that changes in the activity of these areas could be responsible to the inhibitory effect of moxonidine on 0.3 M NaCl intake. In summary, the results showed: a) moxonidine injected into LV, 4th V, amygdaloid complex and BNA, but not into LH and CNA, inhibits sodium depletion-induced 0.3 M NaCl intake; b) RX 821002, yohimbine and SKF 86466 (specific α2-adrenergic receptor antagonists) abolished the inhibitory effect of moxonidine on 0.3 M NaCl intake, suggesting that the inhibitory effect of moxonidine is mediated through α2-adrenergic receptors. RX 821002 and yohimbine increased sodium depletioninduced 0.3 M NaCl intake, suggesting a possible tonic role of α2-adrenergic receptors on the inhibition of NaCl intake; c) moxonidine injected into LV, 4th V and LH reduced the pressor response produced by central angiotensinergic activation, while moxonidine injected only into LV was able to reduce the pressor effect of carbachol. Therefore central injection of moxonidine can inhibit mainly the ANG II-induced pressor response and only partially carbachol-induced pressor response. The reduction on ANG II-induced pressor response produced by moxonidine injected into LV was abolished by the pre treatment with yohimbine, suggesting the involvement of central α2-adrenergic receptors on the inhibitory effect of moxonidine on the pressor response produced by angiotensinergic activation; d) in normovolemic and satiated rats, moxonidine injected into LV induced c-fos expression in several cerebral areas: OVLT, ipsLSA, vMPN, PVN and SON. ANG II (50 ng) injected into LV increased c-fos expression in the following areas: ipsLSA, dMPN, PVN and SHN and reduced c-fos expression in contLSA. The icv injection of moxonidine did not change de c-fos expression induced by ANG II. e) in sodium depleted animals, the icv injection of moxonidine induced an increase in c-fos expression in ipsLSA and dMPN and a reduction in OVLT, suggesting that changes in the activity of these areas could be responsible to the inhibitory effect of moxonidine on 0.3 M NaCl intake in this protocol. / Em situações em que a água e/ou sódio estão em falta no organismo, receptores localizados em diversas partes do organismo ou hormônios produzidos sinalizam para algumas regiões específicas do cérebro, que uma vez ativadas, desencadeiam respostas renais e/ou o comportamento de busca pela água e sódio. Entre essas áreas destacam-se: o órgão vasculoso da lâmina terminal (OVLT), o órgão subfornical (OSF), a região anteroventral do terceiro ventrículo (AV3V), o hipotálamo, a amígdala, a área septal (AS), o núcleo do trato solitário (NTS), a área postrema (AP) e o núcleo parabraquial lateral (NPBL). Além de participarem do controle do equilíbrio hidroeletrolítico, essas áreas também estão envolvidas com o controle cardiovascular. A angiotensina II (ANG II) é um peptídeo que ativa a ingestão de água e de sódio, além de participar da regulação cardiovascular. Outros neurotransmissores podem inibir a ingestão de água e sódio, como serotonina, colecistocinina e peptídeo natriurético atrial. Um mecanismo inibitório da ingestão de água e de sódio também muito estudado está relacionado com receptores adrenérgicos α2 e imidazólicos centrais. A moxonidina, assim como a clonidina, agonistas de receptores adrenérgicos α2 e imidazólicos, são drogas anti-hipertensivas que administradas centralmente reduzem a ingestão de água e de sódio induzida por diferentes protocolos (privação hídrica, depleção de sódio de 24 h, administração de ANG II). Portanto, foram objetivos deste trabalho: a) estudar os efeitos da injeção de moxonidina nos ventrículos cerebrais (ventrículo lateral VL e 4º ventrículo 4º V), no complexo amigdalóide, no núcleo central da amígdala (NCA), no núcleo basal da amígdala (NBA) e no hipotálamo lateral (HL) sobre a ingestão de NaCl 0,3 M induzida por depleção de sódio (tratamento com o diurético furosemide + dieta deficiente de sódio por 24 h); b) verificar os efeitos da moxonidina injetada no VL, 4º V e HL sobre as respostas pressoras produzidas pela injeção de ANG II ou carbacol no VL; c) investigar o papel dos receptores adrenérgicos α2 nos efeitos da moxonidina injetada no VL sobre a ingestão de NaCl 0,3 M induzida por 24 h de depleção de sódio e respostas pressoras produzidas pela ANG II injetada centralmente; d) analisar a ativação de áreas prosencefálicas após a administração de moxonidina no VL em ratos normovolêmicos, depletados de sódio ou tratados com ANG II centralmente com a utilização da marcação por imuno-histoquímica da proteína c-fos. Para tanto, foram utilizados ratos com cânulas de aço inoxidável implantadas nos ventrículos cerebrais: VL (volume de injeção de 1 a 3 µl) e 4º V (volume de injeção 1 µl); unilateralmente no HL (volume de injeção 0,5 µl); e bilateralmente no complexo amigdalóide (volume de injeção de 1 µl), NCA (volume de injeção de 0,2 - 0,4 µl) e NBA (volume de injeção de 0,2 - 0,4 µl). A injeção de moxonidina (20 nmol) no VL promoveu a redução da ingestão de NaCl 0,3 M induzida por 24 h de depleção de sódio durante todo o experimento, enquanto que moxonidina injetada no 4º V reduziu a ingestão de NaCl 0,3 M apenas nos 60 min de experimento. Injeções bilaterais de moxonidina no complexo amigdalóide (5, 10 e 20 nmol/1 µl) e no NBA (20 nmol/0,4 µl) reduziram a ingestão de NaCl 0,3 M, enquanto que injeções bilaterais dessa droga no NCA não alteraram a ingestão de NaCl 0,3 M no protocolo utilizado. Moxonidina (20 nmol/0,5 µl) injetada unilateralmente no HL não afetou a ingestão de NaCl 0,3 M induzida por depleção de sódio. Esses resultados mostram que a inibição da ingestão de NaCl hipertônico decorre da ativação de receptores adrenérgicos α2/imidazólicos produzida pela injeção de moxonidina no VL, 4º V e complexo amigdalóide (em especial no NBA), mas não no HL e NCA. Para se aprofundar no estudo do papel dos receptores adrenérgicos α2 no efeito inibitório da moxonidina na ingestão de NaCl 0,3 M foram utilizados diferentes antagonistas específicos de receptores adrenérgicos α2, como o RX 821002, ioimbina e SKF 86466 em associação com a moxonidina. Os resultados mostraram que a injeção icv de RX 821002 (40 e 80 nmol) e SKF 86466 (640 nmol) aboliram o efeito inibitório da moxonidina (20 nmol) sobre a ingestão de NaCl 0,3 M durante todo o período experimental, enquanto, a ioimbina (320 nmol) aboliu o efeito antinatriorexigênico da moxonidina apenas nos períodos finais do experimento (60 e 120 min). Esses dados demonstram o envolvimento dos receptores adrenérgicos α2 centrais no papel inibitório da moxonidina sobre a ingestão de NaCl 0,3 M induzida por depleção de sódio. Além disso, foi observado um aumento da ingestão de NaCl 0,3 M após o tratamento apenas com os antagonistas RX 821002 (40 nmol) e ioimbina (320 nmol), sugerindo um possível papel tônico dos receptores adrenérgicos α2 centrais no controle da ingestão de NaCl hipertônico. Apenas a administração de moxonidina (respectivamente, 20 e 80 nmol) no HL e VL não promoveu alterações na pressão arterial média (PAM) e na frequência cardíaca (FC). Por outro lado, quando administrada no 4º V, a moxonidina (20 nmol) promoveu hipotensão e bradicardia. A administração de moxonidina no VL na dose de 80 nmol reduziu as respostas pressoras produzidas pela injeção central de ANG II (50 ng) e de carbacol (4 nmol). Moxonidina (20 nmol) injetada no HL e no 4º V reduziu as respostas pressoras produzidas pela injeção central de ANG II, mas não as do carbacol. Assim, demonstrou-se que a injeção central de moxonidina inibe respostas pressoras produzidas pela ativação angiotensinérgica (principalmente) e colinérgica (em menor grau). A administração de ioimbina (320 nmol) no VL aboliu o efeito inibitório da moxonidina (80 nmol) no VL sobre a resposta pressora à ANG II icv, sugerindo que a moxonidina atuaria em receptores adrenérgicos α2 centrais inibindo a resposta pressora à ANG II. Em animais normovolêmicos e saciados, a injeção de moxonidina no VL promoveu a expressão da proteína c-fos nas seguintes áreas: OVLT, área septal lateral ipsilateral (ASLips), núcleo preóptico mediano ventral (NPMv), núcleo paraventricular (NPV) e núcleo supaóptico (NSO), que são estruturas envolvidas no controle do equilíbrio hidroeletrolítico e regulação cardiovascular. O tratamento com ANG II (50 ng) icv aumentou a expressão de c-fos nas seguintes áreas: ASLips, núcleo preóptico mediano dorsal (NPMd), NPV e núcleo septo-hipotalâmico (NSH) e reduziu a expressão de c-fos na ASLcont (área septal lateral contralateral). O tratamento prévio com moxonidina não alterou a expressão de c-fos produzida pela ANG II central. Como os dois tratamentos isoladamente aumentam a expressão de c-fos em áreas cerebrais semelhantes, é difícil saber qual tratamento é o responsável pela expressão de c-fos após a combinação dos dois tratamentos. Talvez, a moxonidina esteja inibindo a expressão de c-fos desencadeada pela ANG II e a expressão de cfos que se observa após os dois tratamentos seja efeito apenas da moxonidina. Em animais depletados de sódio, verificou-se que houve aumento da marcação para c-fos no NPMd e na ASLips e redução no OVLT após a injeção de moxonidina, sugerindo que modificação na atividade dessas áreas possa estar relacionada ao efeito inibitório da moxonidina sobre a ingestão de NaCl 0,3 M. Em resumo, os resultados mostraram que: a) injeção de moxonidina no VL, 4º V, complexo amigdalóide e NBA, mas não no HL e NCA, inibe a ingestão de NaCl 0,3 M induzida pelo protocolo de depleção de sódio; b) RX 821002, ioimbina e SKF 86466, antagonistas específicos de receptores adrenérgicos α2, aboliram o efeito inibitório da moxonidina sobre a ingestão de NaCl 0,3 M, sugerindo que este efeito inibitório da moxonidina seria mediado pelos receptores adrenérgicos α2 centrais. RX 821002 (40 nmol) e ioimbina (320 nmol) aumentaram a ingestão de NaCl 0,3 M induzida pela depleção de sódio, sugerindo um possível papel tônico dos receptores adrenérgicos α2 centrais na inibição da ingestão de NaCl; c) moxonidina injetada no VL, 4º V e HL reduziu as respostas pressoras decorrente da ativação angiotensinérgica central, enquanto que somente a dose de 80 nmol de moxonidina injetada no VL foi capaz de reduzir o efeito pressor do carbacol (4 nmol), mostrando que a injeção central de moxonidina é capaz de inibir principalmente a resposta pressora da ANG II (50 ng) e em menor grau, o efeito pressor do carbacol. A redução do efeito pressor da ANG II promovido pela injeção de moxonidina (80 nmol) no VL foi abolido pela injeção icv de ioimbina (320 nmol), sugerindo uma participação dos receptores adrenérgicos α2 centrais no efeito inibitório da moxonidina sobre a resposta pressora produzida pela ativação angiotensinérgica; d) em ratos normovolêmicos e saciados, a injeção de moxonidina (20 nmol) no VL promoveu expressão da proteína c-fos em várias áreas cerebrais: OVLT, ASLips, NPMv, NPV e NSO. A injeção de ANG II (50 ng) no VL aumentou a expressão de c-fos na ASLips, NPMd, NPV e NSH e reduziu na ASLcont. A injeção icv de moxonidina não alterou a expressão de c-fos produzida pela associação veículo + ANG II. e) em animais depletados de sódio, a injeção icv de moxonidina (20 nmol) aumentou a marcação para a proteína c-fos na ASLips e NPMd e reduziu no OVLT, sugerindo que modificação na atividade dessas áreas possa ser responsável pelo efeito inibitório da moxonidina sobre a ingestão de NaCl 0,3 M.

Une étape cruciale dans la fabrication des MEMS de haute fréquence est la gravure par plasma de la couche mince d’AlN de structure colonnaire agissant comme matériau piézoélectrique. Réalisé en collaboration étroite avec les chercheurs de Teledyne Dalsa, ce mémoire de maîtrise vise à mieux comprendre les mécanismes physico-chimiques gouvernant la cinétique ainsi que la formation de dommages lors de la gravure de l’AlN dans des plasmas Ar/Cl2/BCl3. Dans un premier temps, nous avons effectué une étude de l’influence des conditions opératoires d’un plasma à couplage inductif sur la densité des principales espèces actives de la gravure, à savoir, les ions positifs et les atomes de Cl. Ces mesures ont ensuite été corrélées aux caractéristiques de gravure, en particulier la vitesse de gravure, la rugosité de surface et les propriétés chimiques de la couche mince. Dans les plasmas Ar/Cl2, nos travaux ont notamment mis en évidence l’effet inhibiteur de l’AlO, un composé formé au cours de la croissance de l’AlN par pulvérisation magnétron réactive et non issu des interactions plasmas-parois ou encore de l’incorporation d’humidité dans la structure colonnaire de l’AlN. En présence de faibles traces de BCl3 dans le plasma Ar/Cl2, nous avons observé une amélioration significative du rendement de gravure de l’AlN dû à la formation de composés volatils BOCl. Par ailleurs, selon nos travaux, il y aurait deux niveaux de rugosité post-gravure : une plus faible rugosité produite par la présence d’AlO dans les plasmas Ar/Cl2 et indépendante de la vitesse de gravure ainsi qu’une plus importante rugosité due à la désorption préférentielle de l’Al dans les plasmas Ar/Cl2/BCl3 et augmentant linéairement avec la vitesse de gravure. / A crucial step in the fabrication of high-frequency MEMS is the etching of the columnar AlN thin film acting as the piezoelectric material. Realized in close collaboration with researchers from Teledyne Dalsa, the objective of this master thesis is to better understand the physico-chemical mechanisms driving the etching kinetics and damage formation dynamics during etching of AlN in Ar/Cl2/BCl3 plasmas. In the first set of experiments, we have studied the influence of the operating parameters of an inductively coupled plasma on the number density of the main etching species in such plasmas, namely positive ions and Cl atoms. These measurements were then correlated with the etching characteristics, in particular the etching rate, the surface roughness, and the chemical properties of the AlN layer after etching. In Ar/Cl2 plasmas, our work has highlighted the inhibition effect of AlO, a compound formed during the AlN growth by reactive magnetron sputtering and not from plasma-wall interactions or from the incorporation of moisture in the columnar nanostructure of AlN. In presence of small amounts of BCl3 in the Ar/Cl2 plasma, we have observed a significant increase of the etching yield of AlN due to the formation of volatile BOCl compounds. Furthermore, our work has demonstrated that there are two levels of roughness following etching: a lower roughness produced by the presence of AlO in Ar/Cl2 plasmas which is independent of the etching rate and a larger roughness due to preferential desorption of Al in Ar/Cl2/BCl3 plasmas which increases linearly with the etching rate.

Une étape cruciale dans la fabrication des MEMS de haute fréquence est la gravure par plasma de la couche mince d’AlN de structure colonnaire agissant comme matériau piézoélectrique. Réalisé en collaboration étroite avec les chercheurs de Teledyne Dalsa, ce mémoire de maîtrise vise à mieux comprendre les mécanismes physico-chimiques gouvernant la cinétique ainsi que la formation de dommages lors de la gravure de l’AlN dans des plasmas Ar/Cl2/BCl3. Dans un premier temps, nous avons effectué une étude de l’influence des conditions opératoires d’un plasma à couplage inductif sur la densité des principales espèces actives de la gravure, à savoir, les ions positifs et les atomes de Cl. Ces mesures ont ensuite été corrélées aux caractéristiques de gravure, en particulier la vitesse de gravure, la rugosité de surface et les propriétés chimiques de la couche mince. Dans les plasmas Ar/Cl2, nos travaux ont notamment mis en évidence l’effet inhibiteur de l’AlO, un composé formé au cours de la croissance de l’AlN par pulvérisation magnétron réactive et non issu des interactions plasmas-parois ou encore de l’incorporation d’humidité dans la structure colonnaire de l’AlN. En présence de faibles traces de BCl3 dans le plasma Ar/Cl2, nous avons observé une amélioration significative du rendement de gravure de l’AlN dû à la formation de composés volatils BOCl. Par ailleurs, selon nos travaux, il y aurait deux niveaux de rugosité post-gravure : une plus faible rugosité produite par la présence d’AlO dans les plasmas Ar/Cl2 et indépendante de la vitesse de gravure ainsi qu’une plus importante rugosité due à la désorption préférentielle de l’Al dans les plasmas Ar/Cl2/BCl3 et augmentant linéairement avec la vitesse de gravure. / A crucial step in the fabrication of high-frequency MEMS is the etching of the columnar AlN thin film acting as the piezoelectric material. Realized in close collaboration with researchers from Teledyne Dalsa, the objective of this master thesis is to better understand the physico-chemical mechanisms driving the etching kinetics and damage formation dynamics during etching of AlN in Ar/Cl2/BCl3 plasmas. In the first set of experiments, we have studied the influence of the operating parameters of an inductively coupled plasma on the number density of the main etching species in such plasmas, namely positive ions and Cl atoms. These measurements were then correlated with the etching characteristics, in particular the etching rate, the surface roughness, and the chemical properties of the AlN layer after etching. In Ar/Cl2 plasmas, our work has highlighted the inhibition effect of AlO, a compound formed during the AlN growth by reactive magnetron sputtering and not from plasma-wall interactions or from the incorporation of moisture in the columnar nanostructure of AlN. In presence of small amounts of BCl3 in the Ar/Cl2 plasma, we have observed a significant increase of the etching yield of AlN due to the formation of volatile BOCl compounds. Furthermore, our work has demonstrated that there are two levels of roughness following etching: a lower roughness produced by the presence of AlO in Ar/Cl2 plasmas which is independent of the etching rate and a larger roughness due to preferential desorption of Al in Ar/Cl2/BCl3 plasmas which increases linearly with the etching rate.

<p>In this thesis we have performed core-electron spectroscopy studies of gas phase molecular systems starting with smaller diatomic, continuing with triatomic and extending our research to more complex polyatomic ones. We can subdivide the results presented here into two categories: the first one focusing on electronic fine structure and effect of the chemical bonds on molecular core-levels and the other one dealing with nuclear dynamics induced by creation of a core hole. In our research we have mostly used synchrotron radiation based techniques such as X-ray Photoelectron (XPS), X-ray Absorption (XAS), normal and Resonant Auger (AES and RAS, respectively) and Energy-Selected Auger Electron PhotoIon COincidence (ES-AEPICO) spectroscopies.</p><p>We have demonstrated that resonant Auger spectroscopy can be used to aid interpretation of the features observed in XAS for Rydberg structures in the case of Cl₂ and C1s⁻¹π*¹ states of allene molecules. The combined use of high-resolution spectroscopy with <i>ab initio</i> calculations can help the interpretation of strongly overlapped spectral features and disentangle their complex profiles. This approach enabled us to determine the differences in the lifetimes for core-hole 2p sublevels of Cl₂ which are caused by the presence of the chemical bond. We have shown that contribution in terms of the Mulliken population of valence molecular orbitals is a determining factor for resonant enhancement of different final states and fragmentation patterns reached after resonant Auger decays in N₂O.</p><p>We have also performed a systematic study of the dependence of the C1s resonant Auger kinetic energies on the presence of different substituents in CH₃X compounds. For the first time we have studied possible isomerization reaction induced by core excitation of acetylacetone. We could observe a new spectral feature in the resonant Auger decay spectra which we interpreted as a signature of core-excitation-induced keto-enol tautomerism.</p>

In this thesis we have performed core-electron spectroscopy studies of gas phase molecular systems starting with smaller diatomic, continuing with triatomic and extending our research to more complex polyatomic ones. We can subdivide the results presented here into two categories: the first one focusing on electronic fine structure and effect of the chemical bonds on molecular core-levels and the other one dealing with nuclear dynamics induced by creation of a core hole. In our research we have mostly used synchrotron radiation based techniques such as X-ray Photoelectron (XPS), X-ray Absorption (XAS), normal and Resonant Auger (AES and RAS, respectively) and Energy-Selected Auger Electron PhotoIon COincidence (ES-AEPICO) spectroscopies. We have demonstrated that resonant Auger spectroscopy can be used to aid interpretation of the features observed in XAS for Rydberg structures in the case of Cl2 and C1s−1π*1 states of allene molecules. The combined use of high-resolution spectroscopy with ab initio calculations can help the interpretation of strongly overlapped spectral features and disentangle their complex profiles. This approach enabled us to determine the differences in the lifetimes for core-hole 2p sublevels of Cl2 which are caused by the presence of the chemical bond. We have shown that contribution in terms of the Mulliken population of valence molecular orbitals is a determining factor for resonant enhancement of different final states and fragmentation patterns reached after resonant Auger decays in N2O. We have also performed a systematic study of the dependence of the C1s resonant Auger kinetic energies on the presence of different substituents in CH3X compounds. For the first time we have studied possible isomerization reaction induced by core excitation of acetylacetone. We could observe a new spectral feature in the resonant Auger decay spectra which we interpreted as a signature of core-excitation-induced keto-enol tautomerism.