Le premier problème qui se pose naturellement lors de l'étude des algèbres de Lie nilpotentes est la classification de celles-ci en petite dimension. La classification des algèbres de Lie nilpotentes complexes a été complétée jusqu'en dimension 7. Pour les dimensions inférieures ou égales à 6, il n'existe, sauf isomorphismes, qu'un nombre fini d'algèbres de Lie nilpotentes complexes. Ancochea a classé les algèbres de Lie nilpotentes complexes en dimension 7 selon leur suite caractéristique. On obtient ainsi, une liste plus étendue qui contient des familles d'algèbres de Lie non isomorphes entre elles.On envisage alors d'étudier les algèbres de Lie nilpotentes selon leur nilindice, en commençant par celles qui ont un nilindice maximal, c'est-à-dire , les algèbres de Lie filiformes. Dès 1970. Vergne a initié l'étude des algèbres de Lie filiformes. Elle a montré que sur un corps ayant une infinité d'éléments, il n'existe, sauf isomorphismes, que deux algèbres de Lie filiformes naturellement graduées de dimension paire 2n, nommées L2n et Q2n, et une seule en dimension impaire 2n + 1, appelée L2n+ avec n E N.Plus récemment, Snobl et Winternitz ont déterminé les algèbres de Lie ayant comme nilradical l'algèbre Ln, sur le corps des complexes et des réels. Afin de compléter cette classification à toutes les algèbres de Lie filiformes naturellement graduées, nous avons procéder de même avec les algèbres Q2n,. Nous démontrons ensuite que si une algèbre de Lie indécomposable de dimension finie possède un nilradical filiforme alors elle est forcément résoluble. Les algèbres de Lie filiformes ne présentent donc aucun intérêt dans l'étude des algèbres de Lie non résolubles.Ce résultat n'est plus vrai pour les algèbres de Lie quasi-filiformes dont leur nilradical est abaissé d'une unité par rapport aux filiformes. En effet, en cherchant toutes les algèbres de Lie dont le nilradical est quasi-filiforme naturellement gradué, on a trouvé des algèbres de Lie non résolubles ayant un nilradical quasi-filiforme.Ce même contre-exemple, révèle aussi des différences entre la notion de rigidité dans R et dans C. La classification des algèbres de Lie rigides complexes ayant été déjà faite jusqu'à dimension 8, on est alors amené à trouver cette classification dans le cas réel.Par ailleurs, on a déterminé les algèbres de Lie quasi-filiformes ayant un tore non nul, on obtient une liste beaucoup plus riche que pour le cas filiforme. Cette liste nous permet de prouver la complétude des algèbres de Lie quasi-filiformes. Rappelons que toutes les algèbres de Lie filiformes sont aussi complètes.Finalement, on s'intéresse à l'existence de structures complexes associées aux algèbres de Lie filiformes et quasi-filiformes. Goze et Remm ont démontré que les algèbres filiformes n'admettaient pas ce type de structure. Depuis une approche différente, nous allons redémontrer ce résultat et nous allons voir qu'il existe par contre des algèbres de Lie quasi-filiformes munies d'une structure complexe, mais seulement en dimension 4 et 6. / The first problem which arises naturally in the study of the nilpotenttie algebras is their classification in small dimension. The classification of nilpotent complex Lie algebras was completed until dimension 7. For dimensions lower or equal to 6, there is, except isomotphisms, a finite number of nilpotent complex Lie algebras. In dimension 7, Ancochea classified the nilpotent complex Lie algebras according to their characteristic sequence and he obtains a more extensive list which contains families of non isomorphic Lie algebras.We intend then to study the nilpotent Lie algebras according to their nilindex by beginning with those which have a maximal nilindex. also called filiform Lie algebras. From 1970. Vergne started the study of the filiform Lie algebras. She showed that on a field having an infinity of elements. there are, except isomorphisme, only two naturally graded Lie algebras of even dimension 2n, named L2n, and Q2n,. and there is only one in odd dimension 2n+1, called L2n+1.More recently, Snobl and Winternitz determined the complex and real Lie algebras having the algebra L„ as nilradieal. To generalize this classification to all filiform naturally graded Lie algebra_ we have proceed in a similar wav with the algebra Q2n,. Moreover, we prove that indecomposable Lie algebras with filiform nilradieal are necessarily solvable. Thus, the filiform Lie algebra are irrelevant in the study of the non solvable Lie algebras.This result is not truc for the quasi-filiform Lie algebras. Let us recall that the nilindex of quasi-filiform Lie algebras is, by definition, lowered by a unit with regard to the filiform. Indeed, by looking for all the Lie algebras having a quasifiliform naturally graded nilradieal, we found non solvable Lie algebras having a quasi-filiform nilradical.The same counterexample also reveals differences between the notion of rigidity in R and in C. The classification of complex rigid Lie algebras having been already made until dimension 8, we are then brought to find this classification in the real case.Besides, we determined the quasi-filiform Lie algebras admitting a tonus of derivations, we obtain a list much richer than for the filiform case. This list allows us to prove that all quasi-fi liform Lie algebras are complete. Let us remind that all the filiform Lie algebras are also complete.Finally, we are interested in the existence of complex structures associated to the filiform and quasi-filiform Lie algebras Goze and Remm proved that the filiform algebras did not admit this type of structure. Since a different approach, we are going to re-demonstrate this result and we see that there are, on the other hand, quasi-filiform Lie algebras provided with a complex structure, but only in dimension 4 and 6.

The bilayer disc is a flat, lipid aggregate structure in the nanometre regime. It is composed of a bilayer of amphiphilic molecules with micelle-forming amphiphilic molecules supporting the rim, which prevent disc fusion and self-closure. Stable discs have been found in lipid mixtures containing polyethylene glycol (PEG)-lipids as a rim-stabilizing component. One of the aims of the work described in this thesis was to increase the fundamental knowledge and understanding of the systems in which these discs are formed. Other micelle-forming surfactants apart from PEG-lipids were also explored to see if they could be used to stabilize the disc aggregate structure. Due to the similarities of these lipid discs with natural membranes it was hypothesized that they could be used as models for biological membranes. It was demonstrated that discs are formed in PEG-lipid/lipid systems when the lipid mixture contains components that reduce the spontaneous curvature and increase the monolayer bending rigidity. Discoidal structures are furthermore preferred when the lipids are in the gel phase, probably due to a combination of high bending rigidity and reduced PEG-lipid/lipid miscibility. The disc size could be varied by changing the PEG-lipid concentration. The size and size homogeneity of the discs could also be varied by changing the preparation path. Generally, the preferences of certain lipid systems to form discs remained when the PEG-lipid was replaced by more conventional surfactants. However, discs prepared in PEG-lipid/lipid systems are more useful as model membranes because of their relatively large size and good temperature, dilution and long-term stability. Data obtained with isothermal titration calorimetry and drug partition chromatography indicate that these bilayer discs may serve as an attractive and sometimes superior alternative to liposomes in studies of drug-membrane interactions.

No description available.

No description available.