On a longtemps cru que les régions de l'espace séparant deux étoiles constituaient un milieu quasiment vide. En fait, il n'en est rien et on sait maintenant qu'elles renferment de vastes nuages de gaz essentiellement composés d'hydrogène atomique, de dihydrogène et de monoxyde de carbone CO. On a longtemps pensé également que, du fait des conditions physicochimiques très particulières qui prévalent dans ces nuages interstellaires (faibles températures, faibles densités de matière, champs de radiations élevés dus à  la présence d'étoiles proches) il était peu probable que des processus chimiques puissent s'y dérouler et des molécules y survivre. On sait désormais que ce n'est pas le cas et qu'une chimie complexe et riche peut se développer. A ce jour, près de deux cent espèces chimiques différentes ont été détectées dans les milieux interstellaire et circumstellaire. La physico-chimie menant à  la formation de nouvelles molécules, même les plus complexes, peut avoir lieu en phase gazeuse et aussi à la surface des grains interstellaires. C'est à ce second aspect que je me suis principalement intéressée.Ce travail de thèse de doctorat est une étude expérimentale de l'interaction et de la formation de molécules sur différents types de surfaces simulant les grains de poussière présents dans le milieu interstellaire (silicates et/ou diverses morphologies de glace d'eau qui les couvrent dans les nuages denses froids). Dans ce but, plusieurs expériences sont faites en utilisant le dispositif expérimental FORMOLISM qui réunit des techniques de l'ultravide, de la cryogénie, des jets atomiques et moléculaires ainsi que la spectrométrie de masse.Dans ce travail, j'étudie expérimentalement la formation de l'eau via l'hydrogénation de l'ozone dans des conditions similaires à celles des nuages moléculaires denses (couverture de glace d'eau sur des grains à 10 K). Nos résultats confirment les prédictions théoriques et montrent que l'ozone est bien une des voies possibles, très efficace pour la formation de l'eau, en plus de celles de O et O2. Ce résultat est interprété par l'absence de barrière pour cette réaction.Dans une autre série d'expériences je présente des résultats sur les énergies d'adsorption/désorption de deux isomères, l'éthanol (EtOH) et le diméthyl-éther (DME), sur une surface de silicates couverte ou non de glace d'eau, à basse température. Les résultats obtenus sont comparés aux calculs théoriques sur la stabilité thermodynamique de ces deux isomères, le plus stable (EtOH) interagissant de manière plus efficace avec la glace d'eau que l'isomère (DME) qui a une énergie d'adsorption plus faible. Ce résultat apparait lié au fait que l'éthanol a une plus forte liaison hydrogène avec la surface de la glace. Il peut permettre d'expliquer l'abondance préférentielle du DME observée dans certains milieux circumstellaires. / It has been long believed that regions of space between two stars were an almost empty environment. In fact, it is not the case and we know now that it contains vast gas clouds mainly containing atomic hydrogen, dihydrogen and carbon monoxide CO. It was long thought also that because of the very specific physicochemical conditions prevailing in these interstellar clouds (low temperature, low density material, high radiation fields due to the presence of nearby stars) it was unlikely that chemical processes could take place and molecules could survive. We know now that this is not the case and that a rich and complex chemistry can develop there. To date nearly two hundred different chemical species have been detected in the interstellar and circumstellar medium. The physical-chemistry leading to the formation of new molecules, even complex ones, can occur in the gas phase as well as on the surface of interstellar grains. I have been mainly interested in this second aspect.This PhD thesis is an experimental study of the interaction and of the formation of molecules on different surfaces simulating the dust grains present in the interstellar medium (silicates and/or various morphologies of water ice covering them in cold dense clouds). For this purpose, several experiments are performed using the FORMOLISM experimental setup, bringing together several techniques and methods (ultrahigh vacuum, cryogenics, atomic and molecular beams and mass spectrometry).In this work, I am studying the formation of water via the hydrogenation of ozone under conditions similar to those of dense molecular clouds (water ice covered grains at 10 K). Our results confirm theoretical predictions and show that ozone is one possible very efficient way to form water, in addition to the O and O2 pathways. This result can be interpreted by the existence of a barrier-free reaction.In another set of experiments, I present results on the energies of adsorption/desorption of two isomers, ethanol (EtOH) and dimethyl-ether (DME) on a surface of silicates covered or not with water ice, at low temperature. The results are compared with theoretical calculations on the thermodynamic stability of both isomers, the most stable (EtOH) interacting more efficiently with the ice water than the isomer (DME) which has a lower adsorption energy. This result appears related to the fact that ethanol has a stronger hydrogen bond with the surface of the ice. It may help to explain the observed preferential abundance of DME in some circumstellar medium.

Pure rotational spectra of the rare isotopologues of titanium oxide, (TiO)-Ti-46, (TiO)-Ti-47, (TiO)-Ti-49, and (TiO)-Ti-50, have been recorded using a combination of Fourier transform millimeter-wave (FTmmW) and millimeter/submillimeter direct absorption techniques in the frequency range 62-538 GHz. This study is the first complete spectroscopic characterization of these species in their X-3 Delta(r) ground electronic states. The isotopologues were created by the reaction of N2O or O-2 and titanium vapor, produced either by laser ablation or in a Broida-type oven, and observed in the natural Ti isotopic abundances. Between 10 and 11 rotational transitions J + 1 <-> J were measured for each species, typically in all 3 spin-orbit ladders Omega-1,2, and 3. For (TiO)-Ti-47 and (TiO)-Ti-49, hyperfine structure was resolved, originating from the titanium-47 and titanium-49 nuclear spins of I = 5/2 and 7/2, respectively. For the Omega = 1 and 3 components, the hyperfine structure was found to follow a classic Lande pattern, while that for Omega = 2 appeared to be perturbed, likely a result of mixing with the nearby isoconfigurational a(1)Delta state. The spectra were analyzed with a case (a) Hamiltonian, and rotational, spin-orbit, and spin-spin parameters were determined for each species, as well as magnetic hyperfine and electric quadrupole constants for the two molecules with nuclear spins. The most abundant species, (TiO)-Ti-48, has been detected in circumstellar envelopes. These measurements will enable other titanium isotopologues to be studied at millimeter wavelengths, providing Ti isotope ratios that can test models of nucleosynthesis.

We report the second complete molecular line data release from the Census of High-and Medium-mass Protostars (CHaMP), a large-scale, unbiased, uniform mapping survey at sub-parsec resolution, of millimeter-wave line emission from 303 massive, dense molecular clumps in the Milky Way. This release is for all (CO)-C-12 J = 1 -> 0 emission associated with the dense gas, the first from Phase II of the survey, which includes (CO)-C-12, (CO)-C-13, and (CO)-O-18. The observed clump emission traced by both (CO)-C-12 and HCO+ (from Phase I) shows very similar morphology, indicating that, for dense molecular clouds and complexes of all sizes, parsec-scale clumps contain. similar to 75% of the mass, while only 25% of the mass lies in extended (>10 pc) or "low density" components in these same areas. The mass fraction of all gas above a density of 10(9) m(-3) is xi(9) greater than or similar to 50%. This suggests that parsec-scale clumps may be the basic building blocks of the molecular interstellar medium, rather than the standard GMC concept. Using (CO)-C-12 emission, we derive physical properties of these clumps in their entirety, and compare them to properties from HCO+, tracing their denser interiors. We compare the standard X-factor converting I (CO)-C-12 to N-H2 with alternative conversions, and show that only the latter give whole-clump properties that are physically consistent with those of their interiors. We infer that the clump population is systematically closer to virial equilibrium than when considering only their interiors, with perhaps half being long-lived (10s of Myr), pressure-confined entities that only terminally engage in vigorous massive star formation, supporting other evidence along these lines that was previously published.

New detections of HNC have been made toward 11 planetary nebulae (PNe), including K4-47, K3-58, K3-17, M3-28, and M4-14. These sources, which represent a wide range of ages and morphologies, had previously been observed in HCN by Schmidt & Ziurys. Measurements of the J = 1 -> 0 and J = 3 -> 2 transitions of HNC near 90 and 271 GHz were conducted using the new 12 m and the Sub-Millimeter Telescope of the Arizona Radio Observatory. HCN and HNC were also identified via their J = 1 -> 0 lines toward eight positions across the Helix Nebula (NGC 7293). Column densities for HNC, determined from radiative transfer modeling, were N-tot(HNC) similar to (0.06-4.0) x 10(13) cm(-2), corresponding to fractional abundances with respect to H-2 of f similar to (0.02-1.4) x 10(-7). The HCN and HNC column densities across the Helix were found to be N-tot (HCN) similar to (0.2-2.4). x. 10(12) cm(-2) and Ntot (HNC) similar to (0.07-1.6). x. 1012 cm(-2), with fractional abundances of (0.2-3.2) x 10(-7) and (0.09-2.2) x 10(-7). The [ HCN]/[ HNC] ratio varied between similar to 1-8 for all PNe, with [ HCN]/[ HNC] similar to 1-4 across the Helix. These values are greatly reduced from what has been found in asymptotic giant branch stars, where the ratio is typically > 100. Both the abundance of HNC and the [ HCN]/[ HNC] ratio do not appear to vary significantly with nebular age across a time span of similar to 10,000 years, in contrast to predictions of chemical models. The increase in HNC appears to arise in the proto-planetary stage, but becomes " frozen" once the PN phase is reached.

New laboratory measurements of hydrides have been carried out using terahertz direct absorption spectroscopy. Spin components of the N = 2 <- 1 transition of the free radical CrH (X (6)Sigma(+)) have been recorded in the range 730-734 GHz, as well as a new measurement of the J = 2 <- 1 line of AlH (X (1)Sigma(+)) near 755 GHz. Both species were created in an AC discharge of H-2, argon, and metal vapor. For CrH, the chromium source was Cr(CO)(6), while AlH was produced from Al(CH3)3. The J = 4.5 <- 3.5 and 3.5 <- 2.5 fine-structure components were recorded for CrH, each which consists of resolved proton hyperfine doublets. For AlH, the two main quadrupole components, F = 4.5 <- 3.5 and 3.5 <- 2.5, of the J = 2 <- 1 transition were observed as blended features. These data were analyzed with previous 1 <- 0 millimeter/submillimeter measurements with (6)Sigma and (1)Sigma Hamiltonians for chromium and aluminum hydrides, respectively, and rotational, fine-structure (CrH only), and hyperfine constants were derived. The new measurements have resulted in refined spectroscopic parameters for both species, as well as direct measurement of the respective 2 <- 1 rotational transitions. This work also resolves a 10 MHz discrepancy in the frequency of the AlH line. CrH and AlH have already been observed in the photospheres of stars via their electronic transitions. These data will facilitate their discovery at submillimeter/terahertz wavelengths in circumstellar envelopes and perhaps in diffuse clouds.

A wide range of observed luminosities of young forming stars conflicts with predictions of the standard star formation model, which features a constant accretion rate. To resolve this discrepancy, an episodic accretion model has been suggested. The focus of this dissertation is to test this model in low mass star formation. I present new observations of the CB130 region. The observed photometric data from Spitzer and ground-based telescopes are used to determine the luminosity, and radiative transfer modeling of dust and gas are used to characterize the envelope and disk. I compare molecular line observations to models to constrain the chemical characteristics and abundance variations. Based on the chemical model result and molecular line observations, the low luminosity of the embedded protostar is explained better as a quiescent stage between episodic accretion bursts rather than as the first hydrostatic core stage. I present CO₂ ice observations toward 19 low luminosity embedded protostars. About half of the sources have evidence for pure CO₂ ice, and six have significant double-peaked features, which are strong evidence of pure CO₂ ice. The presence of detectable amounts of pure CO₂ ice signify a higher past luminosity, consistent with the past high accretion. Using chemical evolution modeling, the episodic accretion scenario, in which mixed CO-CO₂ ice is converted to pure CO₂ ice during each high luminosity phase, explains the presence of pure CO₂ ice, the total amount of CO₂ ice, and the observed residual C18O gas. I used CARMA to observe a sample of embedded protostars that spans the full range of protostellar luminosities, especially lower luminosity sources. The standard model predicts the disk mass increases steadily while the episodic accretion model predicts no clear relationship between disk mass and bolometric temperature. Masses of six detected disks spread out regardless of bolometric temperature. With the pure CO₂ ice detection, I can explain disk masses of the source in the context of episodic mass accretion. I conclude that episodic mass accretion provides a good explanation of the low luminosity of protostars, molecular line strength, pure CO₂ ice detection, total CO₂ ice amount and spread of disk masses. / text

Studies of carbon's evolutionary path have been conducted via millimeter and submillimeter observations of circumstellar envelopes (CSEs), planetary nebulae (PNe), molecular clouds and comets. The 12C/13C isotope ratio was measured in Galactic molecular clouds using the CN isotopologs. A gradient of 12CN/13CN was determined to be 12C/13C = 6.01 DGC +12.28, where DGC is distance from the Galactic center. The results of CN are in agreement with those of CO and H2CO indicating a true ratio not influenced by fractionation effects or isotope-selective photodissociation. The 12C/13C isotope ratios in the envelopes of various types of stars were also measured from both CO and CN isotopologs. Such objects as carbon and oxygen-rich asymptotic giant branch (AGB) stars, supergiants, planetary nebulae, and S-type stars were observed. Results from this study indicate 12C/13C values for supergiants ~ 10 and AGB stars 12C/13C ~ 20- 76. Theory would suggest a lower ratio for objects undergoing third dredge-up, though this is seemingly not the case. Multiple carbon-bearing species including CO, HCN, HNC, CN, CS, and HCO+ have also been observed towards the oxygen-rich supergiant, VY CMa. This object has recently revealed a unique chemistry where carbon is not solely contained in CO, and may play a more important role in the chemical network of oxygen-rich circumstellar envelopes. Additionally, observations of species with carbon- carbon bonds, such as CCH and c-C3H2, have been conducted towards evolved planetary nebulae, such as the Helix and Ring nebulae. There is a close agreement in the inventories of species found in PNe and diffuse clouds, suggesting a potential molecular precursor to the interstellar medium. Observations of carbon-bearing species (H2CO and CO) in comets C/1995 O1 (Hale-Bopp), C/2001 Q4 (NEAT), and C/2002 T7 (LINEAR) have been conducted. Formaldehyde is known to have an extended distribution in these objects, likely arising from silicate-organic grains. Evidence of cometary fragmentation was also obtained for H2CO in comet T7 LINEAR as well as for HNC and HCO+ in Hale-Bopp. Such events could contribute to planetary distribution of organics.

Millimeter wave studies of molecules in circumstellar envelopes and a planetary nebula have been conducted. Using the Submillimeter Telescope (SMT) of the Arizona Radio Observatory (ARO) on Mt. Graham, a comparative spectral survey from 215-285 GHz was carried out of the carbon-rich asymptotic giant branch star IRC +10216 and the oxygen-rich supergiant VY Canis Majoris. A total of 858 emission lines were observed in both objects, arising from 40 different molecules. In VY Canis Majoris, AlO, AlOH, and PO were detected for the first time in interstellar space. In IRC +10216, PH3 was detected for the first time beyond the solar system, and C3O, and CH2NH were found for the first time in a circumstellar envelope. Additionally, in the evolved planetary nebula, the Helix, H2CO, C2H, and cyclic-C3H2 were observed using the SMT and the Kitt Peak 12 m telescopes. The presence of these three molecules in the Helix suggests that relatively complex chemistry occurs in planetary nebulae, despite the harsh ultraviolet field. Overall, the research on molecules in circumstellar and planetary nebulae furthers our understanding of the nature of the material that is fed back into the interstellar medium from evolved stars. Besides telescope work, laboratory research was also conducted - the rotational spectrum of ZnCl was measured and its bond length and rotational constants were determined. Lastly, in partial fulfillment of a graduate certificate in entrepreneurial chemistry, the commercial applications of terahertz spectroscopy were explored through literature research.

Les conditions menant à la formation des étoiles massives sont toujours étudiées mais un scénario de leur évolution a été avancé : lors de l’effondrement d’un coeur froid pré-stellaire sous l’effet de la gravité, le milieu se réchauffe et entre ainsi dans la phase de coeur chaud moléculaire (CCM). La proto-étoile centrale en formation accrète de la matière, augmentant sa masse et sa luminosité, et finalement devient suffisamment évoluée pour émettre des photons UV qui irradient l’entourage de l’étoile formant ainsi une région HII hypercompact (HC), puis une région HII ultracompact (UC). À ce stade, une région de photo-dissociation (PDR) se forme entre la région HII et le coeur moléculaire. La composition chimique du milieu nous permet de connaître les processus physiques ayant lieu pendant les différentes phases de la formation des étoiles. De plus, la chimie nous permet également de déterminer le stade de l’évolution d’un objet astrophysique par l’utilisation de codes chimiques incluant l’évolution temporelle de la température et du champ de rayonnement. Jusqu’à présent, peu d’études ont examiné les PDRs internes et cela a été uniquement en présence d’une cavité formée par un écoulement (appelé ici outflow) de matière depuis les pôles de la proto-étoile vers le milieu environnant. La connaissance de ces régions uniques autour des régions HII hypercompact et ultracompact restent donc à approfondir. Ma thèse de doctorat se concentre sur l’évolution spatio-temporelle de la chimie dans les régions HII hypercompact et ultracompact avec des PDRs internes aussi bien que dans les coeurs chauds moléculaires. Le but de cette étude est, premièrement, de comprendre l’impact et les effets sur la chimie du champ de rayonnement, en général très fort dans ces régions. Deuxièmement, le but est d’étudier l’émission de diverses espèces spécifiques aux régions HII HC/UC et de comparer cette émission à celle des CCMs, où le champ de rayonnement UV n’a pas d’influence car il est immédiatement atténué par le milieu. En fin de compte nous voulons déterminer l’âge d’une région donnée en utilisant la chimie associée au transfert radiatif. Pour étudier ces stades transitoires de la formation des étoiles massives, nous utilisons le code astrochimique Saptarsy optimisé et amélioré pendant cette thèse de doctorat. Saptarsy est un code gaz-grain calculant l’évolution spatio-temporelle d’abondances relatives. Il est basé sur l’approche des équations des taux de réactions et utilise le réseau chimique OSU (Université de l’État de l’Ohio) mis à jour. De plus, Saptarsy est couplé au code de transfert radiatif RADMC-3D via un programme, basé sur le langage Python, nommé Pandora. Ceci est fait afin d’obtenir des spectres synthétiques directement comparables avec des observations en utilisant l’évolution spatio-temporelle détaillée des abondances chimiques.En plus de la comparaison entre un modèle de région HII HC/UC avec un modèle de CCM, nous obtenons des modèles pour des tailles différentes de régions HII, pour plusieurs densités au front d’ionisation et pour deux profils de densité. Nous étudions les abondances qui dépendent de manière critique des conditions initiales et nous explorons aussi l’importance de l’émission venant de l’enveloppe pour diverses espèces chimiques. Nous constatons que parmi la douzaine d’espèces que nous avons étudiées seulement quatre d’entre elles sont spécifiques à la phase de région HII ou à la phase de coeur chaud. Ces espèces sont C+ et O pour la première phase et CH3OH et H218O pour la deuxième phase. Cependant, un plus grand nombre d’espèces pourrait être utilisées pour étudier et identifier ces phases. / Conditions leading to the formation of high-mass stars are still under investigation but an evolutionary scenario has been proposed: As a cold pre-stellar core collapses under gravitational force, the medium warms up and enters the hot molecular core (HMC) phase. The forming central proto-star accretes materials, increasing its mass and luminosity and eventually it becomes sufficiently evolved to emit UV photons which irradiate the surrounding environment forming a hyper compact (HC) and then a ultracompact (UC) HII region. At this stage, a very dense and very thin internal photon-dominated region (PDR) forms between the HII region and the molecular core.Information on the chemistry allows to trace the physical processes occurring in these different phases of star formation. Therefore, chemistry also allows the determination of the evolutionary stage of astrophysical objects through the use of chemical models including the time evolution of the temperature and radiation field. So far, few studies have investigated internal PDRs and only in the presence of outflows cavities. Thus, these unique regions around HC/UCHII regions remain to be examined thoroughly.My PhD thesis focuses on the spatio-temporal chemical evolution in HC/UC HII regions with internal PDRs as well as in HMCs. The purpose of this study is first to understand the impact and effects of the radiation field, usually very strong in these regions, on the chemistry. Secondly, the goal is to study the emission of various tracers of HC/UCHII regions and compare it with HMCs models, where the UV radiation field does not impact the region as it is immediately attenuated by the medium. Ultimately we want to determine the age of a given region using chemistry in combination with radiative transfer. To investigate these transient phases of massive star formation, we use the astrochemical code Saptarsy optimized and improved during this PhD thesis. Saptarsy is a gas-grain code computing the spatio-temporal evolution of relative abundances. It is based on the rate equation approach and uses an updated Ohio State University (OSU) chemical network. Moreover, Saptarsy works along with the radiative transfer code RADMC-3D via a Python based program named Pandora. This is done in order to obtain synthetic spectra directly comparable to observations using the detailed spatio-temporal evolution of species abundances.In addition to comparing a HC/UCHII region to a HMC model, we obtain models for different sizes of HII regions, for various densities at the ionization front and for two different density profiles. We investigate the critical dependance of the abundances on the initial conditions and we also explore the importance of the emission coming from the envelope for various species. We find that among the dozen of molecules and atoms we have studied only four of them trace the UC/HCHII region phase or the HMC phase. They are C+ and O for the first and CH3OH and H218O for the second phase. However, more species could be studied to probe and identify these phases.

Dans cette thèse nous avons étudié l’Acetyle cyanide (H3CCOCN noté AC), une molécule d’intérêt astrochimique. Nous avons présenté les résultats des mesures effectuées au Schncrotron SOLEIL où nous avons mesuré les spectres des photoélectrons, des photoions en coïncidence (PEPICO). Afin d’interpréter ces spectres, ces mesures ont été suivies part des calculs ab-initio tres sophistiqués.De plus, nous avons exploré les surfaces d’énergie potentielle (SEPs) de l'acétyl-cyanide et de son cation en utilisant les méthodes standard CCSD(T) et celles explicitement corréllées CCSD(T)-F12 et des méthodes plus précises comme la méthode CCSD(T)/CBS+CV. Un ensemble de paramètres spectroscopiques précis (de vibrations et de rotations) est calculé. Les calculs effecués incluaient les tautomères énoliques qui peuvaient être formés à des énergies d'excitation couvertes étudiées dans cette thèse. Nous avons déduit les rapports de branchement des trois ions qui peuvaient être formés (H3CCOCN+, H3CCO+ et H2CCO+) en fonction de l'énergie d'excitation du photon.Cette étude nous a permis de déduire des données spectroscopiques de haute résolution qui sont indispensables pour une interprétation des observations astrophysiques . Nos calculs ont montré que la décomposition unimoléulaire de H3CCOCN+ se est très complexe / In this thesis we studied the acetyl cyanide molecule (H3CCOCN denoted AC) given the astrochemistry interest. We presented the results of measurements performed in the Synchrotron SOLEIL where we measured the photoelectron photoion coincidence (PEPICO) spectra and the SPES spectra. We also performed very sophisticated ab initio calculations in order to assign these spectra.In addition, we explored the potential energy surface (PES) of acetyl cyanide and its cation using standard (CCSD(T)) and explicitly correlated (CCSD(T)-F12) methods and more precise methods as CCSD(T)/CBS+CV. A set of specific spectroscopic parameters (vibration and rotation) was also calculated. The performed calculations included enolic tautomers which can be formed within the covered excitation energies used studied in this thesis. We deduce the branching ratios for the ions which can be formed (H3CCOCN+, H3CCO+ and H2CCO+) as a function of the photon excitation energy.This study allowed us to provide deduce a high resolution spectroscopic data which are essential for the interpretation of astrophysical observations. Our calculations reveal unimolecular decomposition H3CCOCN+ is very complex