Here - PMOD/WRC

for smelting under vacuum are presented in Table 1 for
Ti#1 and in Table 2 for Ti#2.
Table 1. Impurity content in Ti#1 and TiC-C, ppm
Impurity Al Cr Fe Cu Ni
Ti#1 40 10 9 5 2
TiC-C 0.04 0.8 0.5 0.1 0.08
Table 2.
Impurity content in Ti#2 and TiC-C, ppm
Impurity Al K Si Ca Fe Cr Cu Mn
Ti#2 700 300 100 100 60 30 30 20
TiC-C 0.04 0.05 0.04 0.06 0.08 0.7 0.1 0.05
Discussion
In case evaporation of atoms from molten alloy goes in
“molecular” mode (e.g. under vacuum or in inert gas flux),
velocity of evaporation obeys Langmuir-Knudsen law
U i ~p i 0 T (1), where U i - velocity of evaporation of
i-element (including T i and C), p i 0 – equilibrium pressure
of saturated vapour. Vapour pressure of carbon is so low
that it can be neglected. For most metallic impurities
dissolved in titanium inequality p i 0 > p Ti 0 (2) is correct.
Thus purification of initial titanium is thermodynamically
favorable even in absence of carbon. Kinetics also can be
in no way the limiting factor at melting temperature of
titanium. Nevertheless, metallic impurities do not leave
molten titanium because they chemically interact with Ti
atoms forming strong inter-metallic compounds [7]. In
molten state, inter-metallic compounds undergo some
structural changes but remain chemical bounds strong
enough. Purification of titanium from Fe, Si, Al atoms
goes especially hard. According to thermodynamics of
perfect solutions Fe atoms should have been evaporated
from molten solution with titanium. But experiments
showed that TiFe compound melts keeping strict
stoichiometry.
It is carbon whose presence is of critical importance for
purification of TiC-C. Having strong affinity to titanium
carbon bounds all the existing Ti atoms and blocks
formation of inter-metallic compounds. Staying in form of
free atoms metallic impurities evaporate from molten
eutectic TiC-C according to (1), (2) more intensively than
Ti and C, and purification occurs. Compound TiAl may
react with carbon forming aluminium carbide, which is
fugitive and also evaporates from molten TiC-C.
On the other hand, the possibility of dissolving some
amount of impurities in adjoining layer of graphite crucible
must be verified. It is related to Fe atoms first of all.
Depending on distribution coefficient in complex system
“eutectic alloy-graphite crucible” impurity atoms,
dissolved in inner crucible walls, may appear in eutectic on
melting plateau.
At room temperature titanium (especially titanium
powder) willingly absorbs H, O, N atoms not determined
by mass-spectroscopy method. Hydrogen is totally
removed from Ti at temperatures over 700 0 C [7]. So there
is no hydrogen in titanium-carbon mixture at carbide
synthesizing temperatures. That is why we see no problem
in using both titanium hydride and titanium powder for
preparing TiC-C eutectics. Quality of melting plateau
usually was better if we used titanium hydride. This
empirical fact has no explanation in terms of eutectic
purity.
At heating Ti and O atoms interact forming a number
of oxides Ti 3 O, TiO, Ti 2 O 3 , TiO 2 . Atoms of Ti and N form
high-melting compound -TiN with variable stoichiometry
which solidus line extends up to 3290 0 C [7]. Carbon
reduces Ti from titanium oxides, e.g. for mono-oxide of
titanium the following reaction goes:
TiO + 2C = TiC + CO (gas) (3)
Direct reaction (3) becomes thermodynamically favorable
at about 1600 0 C when there can be no kinetic limitations.
Therefore, oxygen is removed from TiC-C system in form
of CO gas prior to eutectic melting. It concerns both
melting under vacuum and in argon flux; in the former
case CO is removed due to pumping-out, in the latter -
carried away by argon flux.
Similar reaction goes with participation of TiN
compound:
2TiN + 2C = 2TiC + N 2 (gas) (4)
Direct reaction (4) becomes thermodynamically favorable
at about 1800 0 C. So despite the fact, that TiN has higher
melting temperature than TiC-C, nitrogen evaporates prior
to eutectic melting.
We see that presence of carbon again is of critical
importance for elimination of O, N from eutectic TiC-C.
Conclusion
It is at the stage of metal-carbon eutectics preparation,
that most impurities contained in initial metal are
eliminated. Therefore demands to initial purity of metal for
preparing eutectics are not as tough as for pure substances.
For the reason of complexity of melting mechanism we
suggest that eutectic systems should be less sensitive to
impurity than uniform metals. If this is the case, demands
to final purity of eutectics can be not so tough as well.
Above certain purity level structural and kinetic factors can
become more significant in terms of deviation from
equilibrium phase transition in M-C and MC-C eutectics.
References
[1] M.K. Sakharov, B.B. Khlevnoy, V.I. Sapritsky, M.L.
Samoylov, S.A. Ogarev, “Development and investigation of
high-temperature fixed point based on TiC-C eutectic”,
TEMPMEKO 2004, in progress
[2] J. Ancsin, “Equilibrium melting curves of relatively pure and
doped silver samples”, Metrologia, 38, 2001, pp. 229-235
[3] J. Ancsin, “Impurity dependence of Cu-Ag eutectic melting
temperatures”, TEMPMEKO 2004, in progress
[4] P. Bloembergen, Y. Yamada, N. Sasajima, S. Torizuka, N.
Yoshida, N. Yamamoto, “On the effect of impurities on the
melting curve of the eutectic system Fe-C”, 7, AIP Conference
proceeding, 2003, pp. 261-266
[5] A.R. Ubbelohde “The Molten State of Matter”, J. Wiley,
London, 1978
[6] N. Sasajima, Y. Yamada, P. Bloembergen, Y. Ono,
“Dependence of iron-carbon eutectic melting on pre-freezing
rate and annealing conditions”, TEMPMEKO 2004, in
progress
[7] N.P. Liakishev, “Phase-equilibrium Diagrams of Binary
Systems”, Moscow, 1999
280