Metal silicon systems have a wide range of applications, ranging from the use
in electronic industry, as superconductors, protective coatings and as high temperature
structural materials. Mo- and Nb-based silicides have emerged as suitable high
temperature materials and extensive studies are being conducted make it suitable for
various applications. Because of very good strength to density ratio, Nb-based
silicides have attracted maximum attention. This is basically a mixture of Nb solid
solution and Nb5Si3 intermetallic compound. A very small amount of NbCr2 Laves
phase could also be present because of Cr addition. Incorporation of other alloying
elements, which are mainly partitioned to these phases, helps to achieve a property
balance like, high temperature strength, high fracture toughness, high creep and
oxidation resistance.
The knowledge on diffusion parameters is useful to understand many physical
and mechanical properties. In this thesis, diffusion couple technique is used in
different temperature ranges to study the growth kinetics and diffusion of the phases
in an interdiffusion zone in binary silicides, Nb/Si, Mo/Si and V/Si, binary solid
solutions, Nb/Mo, Nb/Ti, Nb/Zr and ternary silicides, Nb-Mo/Si, Nb-Ti/Si, Nb-Zr/Si.
The parabolic growth constant, the integrated diffusion coefficients and the
tracer diffusion coefficients are calculated from the experimental results obtained in
this study and also from the results already available in the literature on the binary
silicides. The activation energy for growth kinetics and the diffusion coefficients are
also calculated to gain knowledge on the diffusion mechanism. The atomic
mechanism of the diffusing species in all the phases of Nb and Mo silicide are
discussed with the help of crystal structure and possible defects present. Also, a
detailed analysis is done on the growth mechanism of the phases in Nb/Si and Mo/Si
systems.
In the Nb/Si system, Si is found to have higher diffusion rate in both the NbSi2
and Nb5Si3 phases. The number of nearest neighbour Si bonds is higher than nearest
neighbour Nb bonds and hence one may predict high concentration of Nb antisites to
be present in the NbSi2 phase. The growth mechanism analysis following the physico
chemical approach explains the absence of the Kirkendall plane in the Nb5Si3 phase
and duplex morphology in the NbSi2 phase in the Nb/Si couple.
In the Mo/Si system, Si diffusion is faster than Mo in all the three phases. In
the MoSi2 phase, Mo is practically immobile due to the absence of vacancies on the
Mo sublattice. Similar defect structure is expected in the Mo5Si3 and Mo3Si phases
also with additional Si antisite defects to assist Si diffusion. The growth mechanism
analysis explains the absence of the Kirkendall plane in the Mo5Si3 and Mo3Si phases and continuous columnar grains in the MoSi2 phase in the Mo/Si couple. In the V/Si system, the activation energy for integrated diffusion coefficient of the VSi2 phase is found to be reasonably lower than the other phases which could happen because of very high concentration of defects, and/or because of contribution from the grain boundary diffusion as it shows the presence of columnar grains. Problems associated with the analysis done in literature are also discussed. A diffusion study is performed in different temperature ranges for the three
binary metallic solid solution systems to determine the interdiffusion coefficients over the entire composition range using the relation developed by Wagner. The change in
activation energy for interdiffusion with composition is also determined. It is found
that activation energy for interdiffusion in Nb/Mo system is much higher than that for
Nb/Ti and Nb/Zr system. Further the impurity diffusion coefficients of the species are
determined and compared with the available data in literature. It is found that the
activation energy for the impurity diffusion of Nb in Ti, Zr and Mo is higher than that
of Ti, Zr and Mo in Nb.
Interdiffusion study is done in the ternary silicides with the aim to examine the role of alloying additions, such as, Ti, Mo and Zr on the growth kinetics and diffusion
behaviour of the phases in the Nb/Si system. The average interdiffusion (or integrated) coefficients are calculated when possible. The reaction and dissociation of the species at the interfaces are considered to understand the growth mechanism of the
phases. An attempt is made to understand the change in diffusion mechanism because
of the presence of third element. It is found that none of the alloying elements
participate in the diffusion process although they do alter the growth kinetics and diffusion rate in both the phases, NbSi2 and Nb5Si3. It is also found that Nb becomes
immobile in the NbSi2 phase in the presence of the alloying elements. Mo reduces the
growth of both the phases while Ti addition does not cause any change in the growth
but affects the diffusivity. Zr addition also reduces growth of the Nb5Si3 phase. It
however complicates the interdiffusion zone in the Nb(Zr)/Si couple, which limits to
qualitative study only.
The Growth and consumption rate of the end members become very significant in many practical applications. Hence, relations for the growth and consumption rate in systems with finite end member thickness is developed
considering single and double phase layer in the interdiffusion zone. Two different
methodologies are used, the diffusion based and the physico-chemical approach to
develop the same relations. We have shown that the diffusion based approach is rather
straightforward; however, the physico-chemical approach is much more versatile than
the other method. It is found that the position of the marker plane becomes vague in
the second stage of the interdiffusion process in such a system, where two phases
grow simultaneously.
Identifer | oai:union.ndltd.org:IISc/oai:etd.ncsi.iisc.ernet.in:2005/2107 |
Date | 05 1900 |
Creators | Prasad, Soma |
Contributors | Paul, Aloke |
Source Sets | India Institute of Science |
Language | en_US |
Detected Language | English |
Type | Thesis |
Relation | G24919 |
Page generated in 0.0026 seconds