No description available.

Made available in DSpace on 2014-06-11T19:24:15Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-02-26Bitstream added on 2014-06-13T20:31:20Z : No. of bitstreams: 1 buchi_at_me_botfm.pdf: 494324 bytes, checksum: 8a45f84c2c6a750412659b4b78ff10f4 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Universidade Estadual Paulista (UNESP) / As serpentes do gênero Crotalus durissus terrificus possuem em seu veneno diversas substâncias, entre elas uma serinoprotease com atividade trombina símile denominada giroxina, a qual é capaz de coagular o fibrinogênio plasmástico, promovendo a formação de fibrina. O objetivo do estudo foi a purificação e caracterização estrutural da enzima giroxina de Crotalus durissus terrificus. Para tanto, foram utilizadas métodos cromatográficos de gel filtração em Sephadex G75 e afinidade em Benzamidina-Sepharose 6B para isolamento e purificação; eletroforéticos de SDSPAGE 12% reduzido e de seqüenciamento do N-terminal da enzima purificada para identificação e avaliação do grau de pureza; clonagem e expressão do cDNA da glândula venenífera por RT-PCR e testes de cristalização. A modelagem teórica molecular foi realizada a partir de ferramentas de bioinformática baseadas em análises comparativas de outras serinoproteases depositadas no NCBI (National Center for Biotechnology Information). O seqüenciamento N-terminal da giroxina purificada, produziu uma proteína de cadeia única com massa molecular aproximada de 30 KDa e sua seqüência completa de cDNA apresentou 714 pb os quais codificaram uma proteína completa contendo 238 aminoácidos. Foram obtidos cristais a partir das soluções nos 2 e 5 do Kit Crystal Screen® 2 e 1 respectivamente, verificando-se sua constituição protéica. Para os alinhamentos múltiplos da sequência molde giroxina símile B2.1 com outras seis serinoproteases derivadas de veneno de serpentes (SVSPs) indicou-se a preservação de resíduos de cisteína e de seus principais elementos de estruturais (α-hélices, β-barris e loops).Evidenciou-se a localização da tríade catalítica nos aminoácidos His57, Asp102, Ser198 e os sítios de atividade especifica S1 e S2 nos aminoácidos Thr193 e Gli215 respectivamente... / The venom of snakes Crotalus durissus terrificus genus have in their various substances, including a serinoproteases with thrombin-like activity called gyroxin, which clot the plasmatic fibrinogen and promote the fibrin formation. The objective of this study was purificate and characterize structurally the gyroxin enzyme from Crotalus durissus terrificus. For this, we used chromatographic gel filtration with Sephadex G75 and affinity benzamidine-Sepharose 6B for isolation and purification, SDS-PAGE 12% in reduce conditions for assessment purity, N-terminal sequencing of purified enzyme for identification and assessment purity; the cloning and expression of cDNA from venom gland by RT-PCR and crystallization tests. The theoretical molecular modeling was performed with bioinformatics tools based on comparative analysis of other serinoproteases deposited in the NCBI (National Center for Biotechnology Information). The N-terminal sequencing purified gyroxin protein produced a single chain with a molecular mass of approximately 30 kDa and its full-length cDNA had 714 bp which encoded a complete protein containing 238 amino acids. Crystals were obtained from the solutions 2 and 5 of the Crystal Screen Kit ® 2 and 1 respectively and was verificated a protein constitution. For multiple sequence alignments such gyroxin-like B2.1 with six other serinoproteases derived from snake venom (SVSPs) indicated the preservation of cysteine residues and its main structural elements (α helices, β-barrel and loops). Identificated the amino acids positions in the catalytic triad His57, Asp102, Ser198 and sites of specific activity in S1 and S2 amino acids Thr193 and Gli215 respectively. The recognition area of fibrinogen cleavage in SVSPs to template sequence gyroxin B2.1 was located in residues Arg60, Arg72, Gln75, Arg81, Arg82, Lis85, Glu86, Lis87... (Complete abstract click electronic access below)

Resumo: As serpentes do gênero Crotalus durissus terrificus possuem em seu veneno diversas substâncias, entre elas uma serinoprotease com atividade trombina símile denominada giroxina, a qual é capaz de coagular o fibrinogênio plasmástico, promovendo a formação de fibrina. O objetivo do estudo foi a purificação e caracterização estrutural da enzima giroxina de Crotalus durissus terrificus. Para tanto, foram utilizadas métodos cromatográficos de gel filtração em Sephadex G75 e afinidade em Benzamidina-Sepharose 6B para isolamento e purificação; eletroforéticos de SDSPAGE 12% reduzido e de seqüenciamento do N-terminal da enzima purificada para identificação e avaliação do grau de pureza; clonagem e expressão do cDNA da glândula venenífera por RT-PCR e testes de cristalização. A modelagem teórica molecular foi realizada a partir de ferramentas de bioinformática baseadas em análises comparativas de outras serinoproteases depositadas no NCBI (National Center for Biotechnology Information). O seqüenciamento N-terminal da giroxina purificada, produziu uma proteína de cadeia única com massa molecular aproximada de 30 KDa e sua seqüência completa de cDNA apresentou 714 pb os quais codificaram uma proteína completa contendo 238 aminoácidos. Foram obtidos cristais a partir das soluções nos 2 e 5 do Kit Crystal Screen® 2 e 1 respectivamente, verificando-se sua constituição protéica. Para os alinhamentos múltiplos da sequência molde giroxina símile B2.1 com outras seis serinoproteases derivadas de veneno de serpentes (SVSPs) indicou-se a preservação de resíduos de cisteína e de seus principais elementos de estruturais (α-hélices, β-barris e loops).Evidenciou-se a localização da tríade catalítica nos aminoácidos His57, Asp102, Ser198 e os sítios de atividade especifica S1 e S2 nos aminoácidos Thr193 e Gli215 respectivamente... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The venom of snakes Crotalus durissus terrificus genus have in their various substances, including a serinoproteases with thrombin-like activity called gyroxin, which clot the plasmatic fibrinogen and promote the fibrin formation. The objective of this study was purificate and characterize structurally the gyroxin enzyme from Crotalus durissus terrificus. For this, we used chromatographic gel filtration with Sephadex G75 and affinity benzamidine-Sepharose 6B for isolation and purification, SDS-PAGE 12% in reduce conditions for assessment purity, N-terminal sequencing of purified enzyme for identification and assessment purity; the cloning and expression of cDNA from venom gland by RT-PCR and crystallization tests. The theoretical molecular modeling was performed with bioinformatics tools based on comparative analysis of other serinoproteases deposited in the NCBI (National Center for Biotechnology Information). The N-terminal sequencing purified gyroxin protein produced a single chain with a molecular mass of approximately 30 kDa and its full-length cDNA had 714 bp which encoded a complete protein containing 238 amino acids. Crystals were obtained from the solutions 2 and 5 of the Crystal Screen Kit ® 2 and 1 respectively and was verificated a protein constitution. For multiple sequence alignments such gyroxin-like B2.1 with six other serinoproteases derived from snake venom (SVSPs) indicated the preservation of cysteine residues and its main structural elements (α helices, β-barrel and loops). Identificated the amino acids positions in the catalytic triad His57, Asp102, Ser198 and sites of specific activity in S1 and S2 amino acids Thr193 and Gli215 respectively. The recognition area of fibrinogen cleavage in SVSPs to template sequence gyroxin B2.1 was located in residues Arg60, Arg72, Gln75, Arg81, Arg82, Lis85, Glu86, Lis87... (Complete abstract click electronic access below) / Orientador: Benedito Barravieira / Coorientador: Marcos R. M. Fontes / Banca: Patrick Jack Spencer / Banca: Andreimar Martins Soares / Mestre

No description available.

Unrecognized metamorphic complexities can produce erroneous interpretations when using equilibrium thermodynamics and isotope geochronology. Universally employed methods for determining pressure-temperature conditions during regional metamorphism are based on the assumption of chemical equilibrium, and geochronology in metamorphic rocks can suffer from cryptic redistribution of isotopes. In this research, the scales of disequilibrium in regionally metamorphosed rocks and the effects of garnet resorption on Lu-Hf garnet ages were examined through numerical simulations of these processes. Concerning scales of disequilibrium, thirteen porphyroblastic datasets, previously measured using X-ray computed tomography, were examined by numerically simulating diffusion-controlled nucleation and growth of garnet while tracking chemical potential gradients to determine reaction affinity Ar (-[Delta]rG). Maximum nucleation rates are 10⁻¹³̇⁶-10⁻⁹̇⁸ nuclei cm⁻³ s⁻¹, interfacial energies are 0.004-0.14 J m⁻² assuming shape factors of 0.1-1.0, and Al intergranular diffusion (QD = 140 kJ/mol⁻¹) is 10⁻¹⁴̇⁴-10⁻¹¹̇¹ m² s⁻¹ at 600 °C. Limitations in determining crystallization kinetics arise from difficulties in constraining rock-specific properties (e.g., porosity and Al solubility). Ar at the time and location of nucleation is 0.4-5.9 kJ/mol⁻¹ of 12-oxygen garnet ([Delta]T = 4.0-62.0 °C) for the earliest nuclei, and 5.3-29.0 kJ/mol⁻¹ ([Delta]T = 50-125 °C) for nucleation at maximum Ar. The results demonstrate potential for delayed nucleation and metastability that can generate spurious interpretations. The timing of metamorphic events is also critical for understanding geologic history. In the Makhavinekh Lake Pluton aureole, Labrador, garnet resorption caused redistribution of Lu and loss of Hf from consumed rims, creating spuriously young ages. Garnet-ilmenite Lu-Hf geochronology using bulk separates yields apparent ages that young toward the contact from 1876 ± 21 Ma (4025 m) to 1396 ± 8 Ma (450 m). Toward the contact, garnet crystals are progressively more resorbed. Numerical modeling was used to test retention of Lu and loss of Hf during resorption as the dominant control on age. More resorption and Lu retention produce younger apparent ages (false ages). Application of the model to the aureole yields model ages from 1850 Ma to 1374 Ma, younging toward the contact. Thus, Lu-Hf geochronology applied to resorbed garnets requires careful examination of Lu zoning. / text

La croissance de la production d’énergie solaire a conduit au développement de l’industrie des cellules photovoltaïques à base de silicium. Pour obtenir un rendement photovoltaïque élevé, tout en limitant les pertes de matière première, la recherche s’oriente sur des procédés de croissance cristalline permettant d’obtenir des lingots monocristallins de section carrée. Le développement du procédé Kyropoulos au silicium permettrait de faire croître des lingots dans un environnement ayant de faibles gradients de température. Ces conditions limiteraient la formation de défauts cristallins et favoriserait le contrôle de la croissance de quatre facettes du silicium formant une section carrée. La combinaison de ce procédé à la solidification dirigée, permettrait de faire croître un lingot cubique. Cette thèse porte sur l’adaptation d’un four de solidification dirigée pour y permettre la croissance Kyropoulos de silicium. Pour cela, une installation expérimentale est développée et les phénomènes physiques liés au procédé de solidification sont étudiés. Ils concernent les échanges de chaleur et de matière ainsi qu’un bilan de forces en jeu entre les différentes phases (solide/liquide/gazeuse). Les lingots de silicium obtenus sont caractérisés. / The increase of solar energy production leads to the development of silicon photovoltaic cells industry. To combine a high photovoltaic efficiency and low material losses, research evolves crystalline processes to obtain a single crystal cubic ingot. The Kyropoulos process adapted to silicon could grow ingots in a low temperature region to limited crystalline defects and could allow the growth of silicon facets to get a square section ingot. The association of the Kyropoulos and the directional solidification processes could grow a cubic ingot. This PhD thesis reaches on the development of a directional solidification furnace to succeed the adaptation of the Kyropoulos process. An experimental setup is developed and physical phenomenon taking place during the process are studied. It is about heat and mass flow. Moreover, forces established between different phases (solid/liquid/gas) are also studied. Finally, silicon ingots crystallized are characterized.

Solution-processed organic field effect transistors (OFETs) have emerged in recent years as promising contenders to be part of electronic and optoelectronic circuits owing to their compatibility with low-cost high throughput roll-to-roll manufacturing technology. The stringent performance requirements for OFETs in terms of carrier mobility, switching speed, turn-on voltage and uniformity over large areas require the performance of single crystal-based OFETs, but these suffer from major scale-up challenges. To achieve device performance approaching that of single crystals with scalable, high throughput and industry-compatible solution coating of OFETs requires understanding and ultimately controlling the crystallization of organic semiconductors (OSCs), and producing very low defect-density thin films. In this thesis, we develop an understanding of the process-structure-property-performance relationship in OSCs that bring fresh insights into the nature of solution crystallization and lead to novel ways to control OSC crystallization, and finally help achieve fabrication of high-performance OFETs by scalable, high throughput and industry-compatible blade coating method. We probe the solution crystallization of OSCs by employing a suite of ex & in situ characterization techniques. This leads us to an important finding that OSC molecules aggregate to form a dense amorphous intermediate state and nucleation happens from this intermediate state during blade coating under a wide window of coating conditions. This phenomenon resembles the so-called two-step nucleation model. Two-step nucleation mediates the crystallization of a wide range of natural and synthetic products ranging from soft materials, such as proteins, biominerals, colloids and pharmaceutical molecules, to inorganic compounds. We go on to show that this nucleation mechanism is generally applicable to achieve formation of high-quality polycrystalline films in a variety of small molecule OSCs and their polymer blends. This phenomenon results in highly textured and well-connected domains, which exhibit reduced interfacial and bulk trap-state densities, helping raise the carrier mobility by one to two orders of magnitude in OFETs in comparison to direct nucleation. We extend the understanding developed for solution crystallization of various acenes and thiophene-based small molecule OSCs to the high-performance benzothieno-benzothiophene (BTBT) based small molecule OSCs. On this end, we develop protocols to fabricate high-quality thin films of BTBT based OSCs by blade coating at industrially compatible coating speeds (>100 mms-1). These films show massive single-domains with very few apparent defects when crystallized via multiple liquid-crystalline phases in two-step nucleation conditions, resulting in an average carrier mobility of ~10 cm2V-1s-1. To sum up, this thesis develops an understanding of OSC solution crystallization and efficient protocols to control polycrystalline thin film quality for high-performance OFETs. These protocols involve a combination of two-step nucleation pathway, solvent mixtures, polymer blends and device-manufacturing conditions. Our efforts enable to realize high-performance OFETs based on high-quality polycrystalline OSC thin films at industry-compatible conditions.

Ces dernières années, une attention particulière a été portée aux polymères biodégradables et biocompatibles, notamment du point de vue écologique. Le développement de biopolymères pour des applications d'emballage alimentaire implique des exigences industrielles spécifiques telles qu’un bas prix ainsi que de bonnes propriétés mécaniques, thermiques et barrières. Le Polylactide (PLA) a attiré beaucoup d’attention car il est produit à partir de ressources naturelles renouvelables, et en raison de sa capacité de mise en œuvre et de ses bonnes propriétés mécaniques. Pour que le PLA puisse être développé à grande échelle pour des applications industrielles dans le domaine de l’emballage, ses propriétés barrières doivent être améliorées. La cristallisation est une méthode très utilisée pour améliorer les propriétés barrières mais n'est pas suffisante dans le cas du PLA. De nouvelles stratégies sont étudiées pour obtenir des effets plus importants. L'une d'entre elles consiste à confiner géométriquement le polymère jusqu’à l’échelle nano en utilisant le procédé de co-extrusion multicouche combiné éventuellement avec un recuit. Cette technologie respecte l'environnement et a déjà prouvé son efficacité pour améliorer les propriétés barrières aux gaz dans le cas d'autres polymères.Cette étude propose d'abord le développement de films PLA avec des structures cristallines différentes dans le but d'optimiser les conditions de cristallisation pour obtenir de meilleures propriétés barrières à l'oxygène. Parmi les différentes structures cristallines obtenues, la perméabilité est la meilleure lorsque le PLA a été cristallisé rapidement à partir de l’état vitreux pour atteindre un degré de cristallinité élevé et un bon découplage de la phase amorphe et cristalline. Ensuite, le PLLA a été confiné par deux polymères amorphes, le Polystyrène et le Polycarbonate. Nous avons montré que le polymère confineur peut influencer la structure cristalline et la mobilité de la phase amorphe du PLLA, influençant ainsi sa perméabilité. / In recent years, much attention has been focused on biodegradable and biocompatible polymers, particularly from an ecological viewpoint. The development of biopolymers for food-packaging applications implies ecofriendly character to specific industrial requirements as low-cost and good mechanical, thermal and barrier properties. Polylactide (PLA) has been attracting great attention, because it can be obtained from renewable sources, and due to its good process ability and mechanical properties. As one of the major challenges for high performance PLA packaging at a large scale is the improvement of its gas barrier properties, the tailoring of the PLA microstructure. Crystallization is a method used to improve barrier properties but is not sufficient in the case of PLA. New strategies are actually studied to obtained stronger effects. One of them consists in the geometrical confinement of the polymer at the molecular scale using the layer-multiplying co-extrusion process combined eventually with annealing processes to create nanometric thickness layers.This technology is environmentally friendly and has already proved its efficiency to improve the gas barrier properties in case of other polymers. This study first proposes the development of PLA films with different structures crystalline with the aim of optimize the crystallization conditions to get better oxygen barrier properties. Among the different crystalline structures obtained, permeability was better when PLA was rapidly crystallized from glass to reach a high crystallinity degree and decoupling of the amorphous and crystalline phase. Then, PLLA was confined by two amorphous polymers, polystyrene and polycarbonate, and its crystals structure and amorphous mobility was changed. We showed that the confiner polymer could influence PLLA confinement, both in the crystalline phase and in the amorphous phase, thus influencing its permeability.

Nós apresentamos ajustes multicromáticos de isócronas ao diagrama cor-magnitude do aglomerado globular NGC 6366, baseados em dados fotométricos que obtivemos com o Southern Astrophysical Research (SOAR) Telescope e de arquivos do Hubble Space Telescope (HST) ACS/WFC. Nós corrigimos os dados por avermelhamento diferencial e calculamos a linha de tendência média dos diagramas cor-magnitude. Nós comparamos os dados com modelos de isócronas estelares de Dartmouth Stellar Evolution Database e PAdova and TRieste Stellar Evolution Code, ambos com a difusão microscópica começando na sequência principal. Considerando todas as determinações prévias de metalicidade desse aglomerado, nós testamos de metalicidades [Fe 111]=-1.00 até [Fe 111]=-0.50, e idades de 9 a 13 Ganos. Depois de determinar a extinção total para seletiva somente com estrelas pertencentes ao aglomerado, Rv=3.06 + 0.14, nós encontramos os parâmetros para esse aglomerado com sendo avermelhamento E (13 — V)= 0.69 ± 0.02(int) +0.04(ext), módulo de distância (m — M)v= 15.02 ± 0.07(int) ±0.13(ext), Idade= 11 ± 1.15 Ganos. Os modelos evolucionários falham em reproduzir a sequência de baixa Tef em diagramas cor-magnitude de multi-bandas, indicando que eles ainda tem uma descrição incompleta. Nós encontramos que as isócronas de Dartmouth Stellar Evolution Database ajustam melhor o ramo das sub gigantes e baixa sequência principal do que as isócronas de PAdova and TRieste Stellar Evolution Code. O uso de anãs brancas de aglomerados globulares em vez de anãs brancas do campo no estudo da formação, propriedades físicas e evolução estelar é justificado por vários fatores. Dentre eles o fato de que as anãs brancas mais frias de aglomerados globulares velhos são remanescentes das estrelas mais velhas que se formaram durante a formação do halo, ou seja, trazem informação da época em que a Galáxia estava se formando. Utilizamos as tabelas fotométricas de dados obtidos com o HST/ACS por Richer et al. (2008) e Bedin et al. (2009) dos aglomerados globulares NGC 6397 e M 4, respectivamente, que alcançam a volta para o azul da sequência de esfriamento das anãs brancas. Com a análise da sequência de esfriamento desses aglomerados, demonstramos que a cristalização de anãs brancas acontece através da liberação de calor latente. Demonstramos também, pela primeira vez, o problema dos modelos de sequência de esfriamento de anãs brancas na volta para o azul e propormos que uma possível solução é, além de considerar o efeito da absorção induzida por colisão, é também necessário considerar que na atmosfera de anãs brancas mais frias de aglomerados globulares há uma mistura de H e He, em vez de uma atmosfera de II puro. / We present multichromatic isochrone fits to the colour-magnitude data of the globular cluster NGC 6366, based on photometric data obtained with Southern Astrophysical Research (SOAR) Telescope and archive data from Hubble Space Telescope (HST) ACS/WFC. We corrected the photometric data for differential reddening and calculated the mean ridge fine of the colour magnitude diagrams. We compared the isochrones of Dartmouth Stellar Evolution Database and PAdova and TRieste Stellar Evolution Code, both with microscopic diffusion starting on the main sequence. Bracketing all previous determinations of this cluster, we tested metallicities from [FelH]=- 1.00 to [Fe/H]=-0.50, and ages from 9 to 13 Gyrs. After determining the total to selective extinction ratio only from stars belonging to this cluster, obtaining Rv=3.06 + 0.14, we found the parameters for this cluster to be reddening E(B — V)=0.69 ± 0.02(int) ±0.04(ext), distance modulus (m — M)v= 15.02 ± 0.07(int) ±0.13(ext), Age= 11 ± 1.15 Gyr. Evolutionary models fail to reproduce the low-T eff main sequence in multi-band colour magnitude diagrams, indicating that they still have an incomplete physical description. We found that the Dartmouth Stellar Evolution Database isochrones fit better the sub giant branch and low main sequence than the PAdova and TRieste Stellar Evolution Code. The use of white dwarf stars from globular clusters instead of field white dwarf stars in the study of formation, physical properties and stellar evolution is justified by several factors. Among them is the fact that the coolest white dwarf stars in old globular clusters are remnants of the oldest stars that formed during the formation of the Galactic halo, i. e., bring us information about the time that the Galaxy was still forming. We use the photometric data obtained with the HST/ACS by Richer et al. (2008) and Bedin et al. (2009) of the globular clusters NGC 6397 and M 4, respectively, that reach the blue turn of the white dwarf cooling sequence.With the analysis of the cooling sequence of these clusters, we demonstrated that the crystallization of white dwarfs occurs with the release of latent heat. Also, we demonstrate, for the first time, the problem of the blue turn in the white dwarf cooling sequence and propose that a possible solution is, in addition to considering the effect of collision-induced absorption, it is necessary to consider that the atmosphere of cooler white dwarf stars is composed by a mixture of H and He, rather than an atmosphere of pure H.

Nós apresentamos ajustes multicromáticos de isócronas ao diagrama cor-magnitude do aglomerado globular NGC 6366, baseados em dados fotométricos que obtivemos com o Southern Astrophysical Research (SOAR) Telescope e de arquivos do Hubble Space Telescope (HST) ACS/WFC. Nós corrigimos os dados por avermelhamento diferencial e calculamos a linha de tendência média dos diagramas cor-magnitude. Nós comparamos os dados com modelos de isócronas estelares de Dartmouth Stellar Evolution Database e PAdova and TRieste Stellar Evolution Code, ambos com a difusão microscópica começando na sequência principal. Considerando todas as determinações prévias de metalicidade desse aglomerado, nós testamos de metalicidades [Fe 111]=-1.00 até [Fe 111]=-0.50, e idades de 9 a 13 Ganos. Depois de determinar a extinção total para seletiva somente com estrelas pertencentes ao aglomerado, Rv=3.06 + 0.14, nós encontramos os parâmetros para esse aglomerado com sendo avermelhamento E (13 — V)= 0.69 ± 0.02(int) +0.04(ext), módulo de distância (m — M)v= 15.02 ± 0.07(int) ±0.13(ext), Idade= 11 ± 1.15 Ganos. Os modelos evolucionários falham em reproduzir a sequência de baixa Tef em diagramas cor-magnitude de multi-bandas, indicando que eles ainda tem uma descrição incompleta. Nós encontramos que as isócronas de Dartmouth Stellar Evolution Database ajustam melhor o ramo das sub gigantes e baixa sequência principal do que as isócronas de PAdova and TRieste Stellar Evolution Code. O uso de anãs brancas de aglomerados globulares em vez de anãs brancas do campo no estudo da formação, propriedades físicas e evolução estelar é justificado por vários fatores. Dentre eles o fato de que as anãs brancas mais frias de aglomerados globulares velhos são remanescentes das estrelas mais velhas que se formaram durante a formação do halo, ou seja, trazem informação da época em que a Galáxia estava se formando. Utilizamos as tabelas fotométricas de dados obtidos com o HST/ACS por Richer et al. (2008) e Bedin et al. (2009) dos aglomerados globulares NGC 6397 e M 4, respectivamente, que alcançam a volta para o azul da sequência de esfriamento das anãs brancas. Com a análise da sequência de esfriamento desses aglomerados, demonstramos que a cristalização de anãs brancas acontece através da liberação de calor latente. Demonstramos também, pela primeira vez, o problema dos modelos de sequência de esfriamento de anãs brancas na volta para o azul e propormos que uma possível solução é, além de considerar o efeito da absorção induzida por colisão, é também necessário considerar que na atmosfera de anãs brancas mais frias de aglomerados globulares há uma mistura de H e He, em vez de uma atmosfera de II puro. / We present multichromatic isochrone fits to the colour-magnitude data of the globular cluster NGC 6366, based on photometric data obtained with Southern Astrophysical Research (SOAR) Telescope and archive data from Hubble Space Telescope (HST) ACS/WFC. We corrected the photometric data for differential reddening and calculated the mean ridge fine of the colour magnitude diagrams. We compared the isochrones of Dartmouth Stellar Evolution Database and PAdova and TRieste Stellar Evolution Code, both with microscopic diffusion starting on the main sequence. Bracketing all previous determinations of this cluster, we tested metallicities from [FelH]=- 1.00 to [Fe/H]=-0.50, and ages from 9 to 13 Gyrs. After determining the total to selective extinction ratio only from stars belonging to this cluster, obtaining Rv=3.06 + 0.14, we found the parameters for this cluster to be reddening E(B — V)=0.69 ± 0.02(int) ±0.04(ext), distance modulus (m — M)v= 15.02 ± 0.07(int) ±0.13(ext), Age= 11 ± 1.15 Gyr. Evolutionary models fail to reproduce the low-T eff main sequence in multi-band colour magnitude diagrams, indicating that they still have an incomplete physical description. We found that the Dartmouth Stellar Evolution Database isochrones fit better the sub giant branch and low main sequence than the PAdova and TRieste Stellar Evolution Code. The use of white dwarf stars from globular clusters instead of field white dwarf stars in the study of formation, physical properties and stellar evolution is justified by several factors. Among them is the fact that the coolest white dwarf stars in old globular clusters are remnants of the oldest stars that formed during the formation of the Galactic halo, i. e., bring us information about the time that the Galaxy was still forming. We use the photometric data obtained with the HST/ACS by Richer et al. (2008) and Bedin et al. (2009) of the globular clusters NGC 6397 and M 4, respectively, that reach the blue turn of the white dwarf cooling sequence.With the analysis of the cooling sequence of these clusters, we demonstrated that the crystallization of white dwarfs occurs with the release of latent heat. Also, we demonstrate, for the first time, the problem of the blue turn in the white dwarf cooling sequence and propose that a possible solution is, in addition to considering the effect of collision-induced absorption, it is necessary to consider that the atmosphere of cooler white dwarf stars is composed by a mixture of H and He, rather than an atmosphere of pure H.