noc20-me28-lec29

IIT Kanpur July 201
20 followers
Dec 3, 2023
Welcome to the 29th lecture in this open course on Diffusion in Multicomponent Solids.
In this lecture, I have explained three experimental techniques which are commonly used for measurements
of vacancy concentrations or the Enthalpy of Vacancy Formation.
These techniques include differential dilatometry, residual resistivity and positron annihilation.
In the last class, we discussed the energetic of vacancy formation and in this lecture,
we will go over some of the experimental techniques, which are used for measurement of vacancy
concentration or enthalpy of vacancy formation.
So, just quickly refresher.
Got the expression for mono-vacancy concentration, X1V as exponential delta S1V vibrational divided
by K times exponential minus delta H1V divided by KT where delta S1V vibrational denotes
the vibrational entropy change associated with the formation of a mono-vacancy and delta
H1V denotes the enthalpy of formation of a mono-vacancy.
K is Boltzmann constant and T is the temperature in Kelvin.
Now it should be noted, when we use this quantities delta H and delta S in terms of per mole,
then we need to replace K with the gas constant R. Now let us go over some of the techniques
which can be used for measurement of this various parameters.
X1V or delta H1V, the first of this technique is called differential dilatometry or DD in
short.
So, this technique relies upon the measurement of volume change that any metal piece undergoes
upon heating.
So when a metal piece is heated, it experiences expansion and this expansion has mainly two
contributions, the first one is purely from the thermal expansion and the second one is
because of the formation of vacancies.
As we have seen the vacancy concentration varies exponentially with temperature, which
is clear from this equation.
As the temperature increases, vacancy concentration increases exponentially and the way the vacancies
are formed, which we have seen in last class.
Introduction of a vacancy, gives rise to one additional lattice site into the crystal and
so as the temperature increases, to maintain the equilibrium, extra vacancies are formed
and so the extra lattice sites are added to the crystal structure.
This will give rise to the second contribution to the net volume change, which is caused
upon heating.
If we can measure this volume change and the individual contribution, we can relate them
to vacancy concentration and that is the principle here.
So, let us say if the piece of a metal is heated from a lower temperature T0 to a higher
temperature T. It undergoes the change in volume equal to
delta V. So, let us denote the volumes at T0 and T by V0 and V.
The lattice parameter at T0 and T are denoted
by A0 and A. Then let us consider, the volume per lattice site, which is denoted by omega.
So, omega 0 and omega, denote the volume per lattice site at the two respective temperatures
and let us suppose at T0, there were big N number of lattice sites
and as we heat it to T because of the extra vacancies which have to be created, to maintain
the equilibrium vacancy concentration at T, there are n vacancies which have been added.
So, small n is the extra vacancies at T over those at T0.
So, delta V is simply V minus V0 and at T0, volume per lattice site was omega 0.
There were N total number of lattice site and at T the volume per lattice site was omega
and there were total number of N plus n sites.
So we can write this as N plus n times omega minus big N times omega 0.
So if we take this capital N common, we get N times omega minus omega 0 plus small n times
omega and if we denote this omega minus omega 0 as delta omega, here delta V equal to N
delta omega plus small n times omega.
Now we have clearly got the two different parts.
So, delta omega is basically the change in volume per lattice site
and as you can understand, this is purely because of thermal expansion.
Now when we say that, we are also making one assumption here and that assumption is that
there is no change in volume per lattice site.
Because of creation of extra vacancies.
So, what we are assuming is that at any given temperature T, if a vacancy is created there
is one extra lattice site which is added and that extra lattice site also has the volume
omega or the creation of the vacancy does not change the volume per lattice site.
Now in practice there may be a little bit of change in volume because of creation of
the vacancies.
So to understand this, if we consider a lattice with a vacancy and as we have seen, the atoms
are held at their equilibrium positions because of the balanced interactions.
At equilibrium positions the interactions are balanced and so the atoms stay at their
equilibrium position, in fact they keep vibrating about the equilibrium positions.
But when a site is vacant, locally, the interactions are not balanced and so the surrounding atoms
will tend to fall into this vacant site.
So, that will give rise to some change in the volume because of creation of a vacancy.
So, this atomic rearrangement will cause some change in volume, but that change in volume
per lattice site will be negligible.
So we can neglect that.
So, now if we divide, both the side of this equation by the original volume V0 and delta
omega over and V0 is nothing but n times omega 0 plus n over N times omega over omega 0.
This N will get cancelled.
So, we can write, now again in this equation, the small n over big N, we can approximately
make it equal to X1V that is the equilibrium concentration of vacancies at temperature
T.
This is because the temperature T0 is usually selected at a very low value and the temperature
T is at a very high value.
Because of which the vacancy concentration at T is much higher than that at T0.
So, the change in vacancy concentration is approximately equal to the new vacancy concentration
established at temperature T and so we can approximately take this equal to X1V and also
the change in volume per lattice site is also not significant.
So, omega is approximately equal to omega 0 or omega over omega 0 is approximately equal
to 1.
So if we rearrange.
We get X1V equal to delta V over V0 minus delta omega over omega 0.
Now the linear dimensions are easiest to measure.
So, we can relate this delta V by V0 or delta omega by omega 0 to the linear dimensions
and typically the relative change in volume is equal to approximately three times the
relative change in linear dimensions.
So, when we talk about the piece of metal, delta V by V0 is equal to 3 times delta l
over l0.
Where delta l is the change in length.
Because of heating from T0 to T and l0 is the original length of the metal piece at
T0.
What can be said about delta omega over omega 0.
So, the omega refers to the volume per lattice site and when it comes to the lattice, the
linear dimension, we can take as the lattice parameter.
So delta omega by omega 0 is equal to 3 times delta a by a0.
Where delta a is the change in lattice parameter, because of the heating from T0 to T, and a0
is the lattice parameter of the metal piece at T0.
So, we can write this as X1V is equal to 3 times delta l over l0 minus delta a over a0.
So, this gives us an idea about how we can go about measuring vacancy concentrations
by measuring the linear dimensions.
So, if you can measure delta l by l0 for the overall metal piece and delta a by a0 which
are the changes in lattice parameter.
Then we can measure the vacancy concentration at a given temperature T.
So, how do we go about this, obviously we need to make measurements of lattice parameters.
So, we need to use X-ray diffractometer and typically there are dedicated X-ray diffractometer
which are equipped with measurement of small changes in linear dimensions of the metal
piece.
At the same time the change in lattice parameter is tracked and they are related to the vacancy
concentration.
Of course we also need the XRD to be equipped with high temperature measurements.
So, that we can do the measurements at different temperatures.
This is an equilibrium measurement technique which means that at every temperature before
the measurement we equilibrate the sample so, that the equilibrium vacancy concentration
at that temperature is established and to make sure that we are measuring the equilibrium
vacancy concentrations, typically the measurements are done during heating as well as during
cooling.
Every time before the measurement at the given temperature the sample is equilibrated so
that the equilibrium vacancy concentration is established.
Now it looks very simple, but in fact the measurements are very critical.
We need very high accuracy and precision.
Why because the values of X1V, the equilibrium vacancy concentrations are typically very
small.
Even close to melting temperature, X1V is typically about 10 to power minus 4.
Which means if we expect about 10 percent accuracy, we need to measure delta l by l0
and delta a by a0 with accuracy of 10 to power minus 5.
That is when we have a very high vacancy concentration about 10 to power minus 4.
So, the measurements have to be done with very high accuracy.
Now one advantage of this method is that it gives the absolute concentration of vacancies,
X1V and since X1V is equal to exponential delta S1V vibrational by K times exponential
minus delta H1V by KT, if we do the measurements, at different temperatures, it also helps in
evaluating delta H1V as well as delta S1V.
So, if we take log on both sides, ln X1V should be equal to minus delta H1V by K times 1 by
T plus delta S1V vibrational by K.
So, if we plot the log of vacancy concentration as a function
of reciprocal of temperature.
We should be able to plot the data with a straight line.
The slope of it will yield the value of delta H1V.
So, the slope would be equal to minus delta H1V by K and the intercept will give the value
of delta S1V vibrational.
Now the intercept is at 1 by T equal to 0 which is not possible.
So, we have to just do the extrapolation of the data that we have obtained to the 1 by
T equal to 0 axis, to get to delta S1V vibrational, that is the vibrational entropy change due
to formation of a mono-vacancy.
There is one point has to be noted here, when we do this measurement, we are talking about
mono-vacancies.
But it should be noted that these measurements do not separate the mono-vacancies or di-vacancies
or other higher clusters of vacancies.
It just measures the vacancy concentration.
But typically we can take it as mono-vacancies.
Because the di-vacancy concentration will be much lesser as we have seen last class.
Even at near to melting temperature for platinum, the di-vacancy concentration was about three
orders of magnitude lower than the mono-vacancy concentration.
So, the error in mono-vacancy concentration involved is much lower.
Even if you assume that the entire vacancy concentration belongs to the mono-vacancies.
So let us look into some results from the literature.
Here I am showing the results from the classic work reported by Simmons and Balluffi on the
measurement of equilibrium vacancy concentration in gold using the differential dilatometry
technique.
This plot is a plot of delta l by l or delta a by a, versus temperature and this is in
the smaller range of temperature, from room temperature to 600 degree Celsius.
Now they have reported the measurement accuracy of about 10 to power minus 5, in the measurements
of delta l by l or delta a by a and in this temperature range, up to 600 degree centigrade,
the changes in vacancy concentration is not significant.
So, the changes in this delta l by l or delta a by a, were not in the detectable limit and
so you can see all these measurements overlap in this temperature range.
But at higher temperature range, 900 to 1050 degree centigrade you can see the difference
between delta l by l and delta a by a is detectable.
Delta l by l is of course higher than delta a by a and the difference is increasing with
increasing temperature.
So, then plotted, in this temperature range 900 to 1050 degree centigrade, X1V versus
1 by T and they could get the enthalpy of vacancy formation and vibrational entropy
of vacancy formation and the values they obtained were delta H1V of 0.94 electron volts and
delta S1V equal to 1kb where kb is the Boltzmann constant.
With this parameters if we estimate the vacancy concentration at 900 degree centigrade, it
is about 2.5 times 10 to power minus 4 at 1000 degree centigrade it is about 5.2 10
to power minus 4 and at 1,064 degree Celsius which is the melting temperature of gold,
it comes out to be around 7.9 times 10 to power minus 4 site fractions.
So, this was about the differential dilatometry.
The next technique that we are going to talk about is the residual resistivity measurement.
As the name suggest, this technique relies upon measuring the resistivity of a sample
and relating it to the vacancy concentration.
So, the resistivity of a metal basically has different contributions.
One is of course from the phonons or thermal vibrations of atoms and other one is from
the defects which are present in the crystal lattice of the metal.
So, the first contribution is rho T and the defects include vacancies, grain boundaries,
dislocations, impurity atoms and other defects.
So, since we are interested in vacancy measurement.
So, let us separate the vacancy contribution and the contribution due to other defects.
So, let us call the contribution from the vacancy to be rho v and rho d is because of
the other defects, including grain boundary, dislocations, etc.
Now if we can measure the resistivity of the same piece of metal in two different conditions,
such that the rho T and rho d is constant in the two conditions.
But only rho v changes, then all the change in the resistivity is only because of the
change in vacancy concentration.
So, we basically want to make delta rho equal to delta rho v and this is proportional to
delta X1V.
So, how do we go about this, so we have to first take a piece of metal and equilibrate
it at a lower temperature and measure its resistivity.
So, the first step is, measure rho 0 at T0, then heat the metal piece to a high temperature
T, equilibrate at T. So, that the equilibrium vacancy concentration at temperature T is
achieved and then quench, from T to T0.
Now if the grain boundary and dislocation density is not affected, then essentially,
we get back the sample at temperature T0 with quenched in vacancies from higher temperature
T. So, whenever we are doing heating or cooling, as the temperature changes, system tries to
achieve the equilibrium vacancy concentration and this is done by diffusion of vacancies
to sources or sinks.
If the vacancies have to be created there are sources.
So, this grain boundaries dislocation and surfaces act as sources.
So, for example if there is a dislocation and if a vacancy has to be created, an atom
from the crystal will jump into the dislocation and the vacancy from the dislocation will
make a reverse jump and will be introduced into the crystal and this is the way a vacancy
is created.
Similarly, grain boundary or surfaces can act as sources of vacancies.
When we heat the sample to a higher temperature.
Now when the equilibration starts at higher temperature, you can understand the vacancy
concentration near the defects would be very large.
But the equilibrium wants that, the vacancy concentration should be uniformly distributed
throughout the bulk of the crystal.
So, the vacancies then have to diffuse from there to the other parts of the lattice and
so the equilibration will take some time.
Similarly, when the temperature is reduced, the extra vacancies, the vacancies which are
in excess of the equilibrium concentration at that temperature have to be gotten rid
of and that occurs by diffusion of these vacancies towards the defects.
Now the same defects would now act as sinks.
That means that the creation or annihilation of vacancies, requires time.
Because it is driven by diffusion.
By quenching from higher temperature T, we are not giving enough time to vacancies for
annihilation and so the excess vacancy concentration, will be retained at the lower temperature.
And the fourth step here is to measure rho again at T0.
But the difference between step one and step four is that, in step one this was the resistivity
of well annealed sample.
Step four, the resistivity is of quenched sample.
Now if the density of dislocations and the grain boundary area remains constant, the
only contribution to the change in resistivity is going to be from the change in vacancy
concentration.
Because we are doing the measurements at the same temperature rho T is also same and then
we can write this.
Delta rho is equal to some constant A times delta X1V.
So, how do we keep the grain boundary and dislocations constant.
That is why it is important to start with a well annealed sample.
The sample has to be first equilibrated at a high temperature.
So, that the grain growth takes place, and the grain boundaries are well stabilized.
There should not be any further grain growth or elimination of smaller grains during the
subsequent high temperature heating.
Preferably a large grain sample is used.
So, that there would not be much significant in the grain boundary area.
Similarly, the dislocations density will be minimized after annealing.
But during quenching dislocation may be introduced.
So, we have to take care that the dislocations are not introduced during quenching.
So, we can try to avoid the introductions of new dislocation by first quenching it to
a lower temperature.
Let us say to a 100 degree centigrade and then from there to the lower temperature T0.
Now this lower temperature T0 is typically sub-zero temperature, minus 70, minus 80 degree
Celsius.
Because at that temperature, the contribution from thermal vibrations is going to be very
low.
So, the change in delta rho V is going to be prominent.
Now this delta X1V, again the difference between T and T0 is going to be very large.
So, delta X1V can be approximately taken as equal to X1V at the quenching temperature
T and so we can write, delta rho is equal to A exponential delta S1V by K times exponential
minus delta H1V by KT and remember this T here, belongs to the quenching temperature.
Remember this T here denotes actually the quenching temperature.
Now with this technique, as you can see we can measure, delta rho from different quenching
temperatures and plot it, plot the logarithm of delta rho versus 1 by T, again we will
get a straight line, whose slope gives us the value of delta H1V.
So, with this technique we do not get the absolute concentration of vacancies.
Because there is this constant A, which we do not know.
But by determining delta rho at different temperatures and plotting it against 1 by
T we can get to the enthalpy of vacancy formation delta H1V.
Again since A is not known, so from the intercept we cannot actually determine delta S1V.
So, this is a simple technique, it does not give the absolute vacancy concentration.
But we can get delta H1V based upon residual resistivity measurement.
So, let us look into some of the literature data.
I have shown the results from couple of papers here.
The first one is from Bourassa and Lengeler, who measured the vacancy concentration by
residual resistivity measurement in pure copper and they used the quenching temperature in
the range of 750 to 1050 degree centigrade.
You can see a nice straight line fitting is obtained and they measured delta H1V to be
1.27 electron volt for copper.
With this if we determine X1V at the melting temperature of copper it comes out to be around
1.4 times 10 to power minus 4.
If we take delta S1V to be 2kb.
Again, delta S1V cannot be measured by this technique, this is just what I have taken
from other data.
Another example is from the work of Kloske and Kauffman.
They measured the residual resistivity in the temperature range of 500 to 950 degree
centigrade.
This was the quenching temperature in the gold alloy containing 1.2 atomic percent silver
and they obtained in this alloy delta H1V to be 1.01 electron volt.
Again, if we calculate the vacancy concentration at 1064 degree Celsius which is the melting
point of pure gold, it comes out to be around 4.3 times 10 to power minus 4 site fractions.
So, this was the second technique for vacancy measurement.
The third one that I will talk briefly here, is based upon positron annihilation.
So, some of the radioactive isotopes, emit positrons, while they are decaying.
These positrons are nothing but the anti-particles.
They have the same mass as electrons and they also have the same magnitude of charge as
that of an electron.
But the charge has the opposite sign of electron, these are positively charged particle.
So, in a sense they are anti-electron and when the positrons combine with electrons,
they emit two gamma rays.
So, what essentially is done in this technique is the piece of metal is kept close to a radioactive
isotope.
The most commonly used radioactive isotope is sodium 22.
This emits gamma rays of energy about 1.12 mega-electron volts and positrons.
So, the positrons will go into the piece of metal and there they will combine with the
free electrons of the metal and emit gamma rays of the energy about 511 kilo-electron
volts.
Now when these positrons go into a vacant site of the metal piece.
There, there is an absence of core electron in the vacancies and so it will take time
for the positron to combine with the electron.
So, we say that the positrons are trapped in the vacant sites and more the number of
vacancies, more positrons will be trapped and then the average lifetime of the positrons
will increase with the vacancy concentration and this is the principle which is used in
measuring the vacancy concentration from the lifetime measurement of positrons.
So, this was the third technique, so we will stop here today, thank you.
