noc19-ph02 Lecture 22-Electronic contribution to the specific heat of a solid Part-IV

IIT Kanpur July 201
20 followers
Dec 3, 2023
[Music]
[Music]
we continue with calculating now the
specific heat and by working through
this entire derivation we have reached a
point where the specific heat is written
in terms of the density of states at the
Fermi energy integrated over this
expression where now you have the
derivative of the Fermi Dirac
distribution so let us look at this
expression a bit more closely the
specific heat is the density of states
at the Fermi energy it's expected as I
have already discussed that only the
states which are very close to the Fermi
energy are the ones which get electrons
in these states are the ones which are
going to get excited around the Fermi
energy and go to unoccupied States as
you are supply a temperature T and that
is how the energy of the system will
rise and so they will contribute to the
specific heat so these states are
contributing to the specific heat
integral 0 to infinity ek minus EF D F D
by DT D ek and now let us evaluate this
expression the Fermi Dirac distribution
FD ek minus mu is 1 over 2 ek minus mu
ek minus mu by Kb T so therefore D F D
by DT okay here before we continue
further let us make a substitution that
you have a quantity which is let us say
let us call it as Zeta which is equal to
KBT
then D F D by DT is K B into D
F D by D Zeta okay is this is nothing
else but K B D so I have to multiply it
by KBT so this is KB into taking the
derivative of D by D Zeta of this
expression 1 by 1 plus e raised to ek
minus mu by Zeta and this would give me
a term KB into ek minus mu by Zeta
square times e raised to ek minus mu by
Zeta divided by 1 plus e raised to ek
minus mu by Zeta the whole square you
can just check it it's very easy to take
the derivative of this you will get this
will be the derivative of this
expression and then if you take the
derivative of this you will get this
term so you can now write and we will
take the integration now so now we can
put it back into the integral if we
substitute all of this here cv is equal
to KB into density of states at the
Fermi energy integral 0 to infinity ek
minus EF by Zeta the whole square into e
raised to ek minus mu divided by Zeta
divided by 1 plus e raised to ek minus
mu by Zeta the whole square d ek
so we have this expression okay and now
mu is of course the chemical potential
at finite temperature and now we can
make a substitution of this X equal to
ek minus EF by Zeta which implies that
DX is equal to de K by Zeta which is
implies that KBT into DX is equal to de
K because this Zeta is KBT now let us
look at this expression a bit longer and
what you see is that here is the
chemical potential mu and if you recall
that the chemical potential at any
finite temperature T is defined by the
density by the counting and that is GE k
FD ek minus mu D EK 0 to infinity ok and
if you actually calculate it again we
have calculated it for fine zero
temperature which is this counting has
to be equal to 0 to infinity G ek FD ek
minus mu at zero temperature ve K this
is at a finite temperature T not equal
to 0 and this is equal to 1 for energies
which are less than or equal to EF and
is equal to 0 for energies greater than
EF so by integrating this expression at
0 temperature which is equal to the
density we get the expression for the
fermi energy however if you don't do
this if you just consider this then you
can calculate the chemical potential at
a finite temperature and that chemical
potential at a finite temperature is
equal to the chemical
potential at zero temperature if you do
this detailed calculation which I will
not do here you can show that by solving
this expression again similar almost
along the lines with a little bit more
involved calculations you can show that
the chemical potential at a finite
temperature is the chemical potential at
zero temperature plus a term which is of
the order of T by TF the whole square so
the chemical potential mu T is equal to
the Fermi energy plus a term which is of
the order of T by DF the whole square
now if you recall that the Fermi
temperature of the material is like 10
raised to 4 Kelvin and the temperature
of the material can be of the order of
say 10 raised to 2 Kelvin that's
considering it that you've really heated
up the system to about 100 Kelvin or so
or 300 Kelvin and so on even then the T
by TF is just 10 raised to minus 2 of
course there are higher order
Corrections but this is the leading
order term you will get T by TF the
whole square and then there are higher
order Corrections but just the first
correction to the chemical potential at
a finite temperature is T by TF which is
minus 2 and so therefore this term at a
finite temperature is EF the Fermi
energy which is the chemical potential
at zero temperature plus a term which is
like of the order of 10 raise to minus 4
because it is T by DF the whole Square
and then you will have of course much
much smaller terms so already you have a
correction which is very small compared
to the Fermi energy and so although the
chemical potential because at 0
temperature if you look at the Fermi
Dirac distribution as a function of
energy it is like this so this is Fermi
energy
but if you look at the fermi-dirac
distribution at a finite temperature
then the distribution changes and the
particles start leaving lower energy
states and going up to higher energy
states
so although the fermi energy will be
sitting at somewhere at EF this is the
Fermi energy which is the original
location of the chemical potential at
zero temperature the actual chemical
potential will be slightly higher this
is the chemical potential at finite
temperature and this chemical potential
at finite temperature is the chemical
potential at zero temperature plus a
term which is of the order of 10 raise
to minus 4 and higher order Corrections
but you can see I have exaggerated the
shift in the chemical potential with
respect to 0 temperature this is at a
finite temperature and this is the zero
temperature this is the chemical
potential at zero temperature at a Fermi
energy so while this was the original
location when it was at zero temperature
because of a redistribution of particles
around the Fermi energy the chemical
potential just moves around a little bit
and this is the amount of motion so for
all practical purposes you can assume
that the mutti is of the order of Fermi
energy so in our calculation to simplify
the calculations a bit further when we
are looking at the specific heat if you
go back to this expression this chemical
potential which is occurring here this
chemical potential which is occurring
here we will replace this with the Fermi
energy we will approximate these with
the Fermi energy and therefore we can
rewrite this expression as equal to if
you go back to this expression then we
will rewrite this expression as KB g EF
integral 0 to infinity ek minus EF by
Zeta the whole square e raised to ek
minus EF by Zeta divided by 1 plus
e raise to e K minus EF by Zeta the
whole squared e e K and now X is e K
minus EF by Zeta you replace this by
this de K will be equal to KB into T DX
which I have already shown you
furthermore now you have the limits of
integration if EK equal to 0 x is equal
to minus EF by Zeta for ek equal to
infinity X is equal to s to infinity so
the specific heat can now be written as
KB square into T the density of states
at the Fermi energy integral of minus EF
by Zeta to infinity ok X square e raised
to X divided by 1 plus e raised to X the
whole square DX so now we are coming to
some point where we can now have an
integral ok and this if you look at the
lower limit of integral it is going from
a minus value to a plus value and what
is important to see out here is that
this negative value if you take it to
minus infinity it does not change things
too much because EF by Zeta is of the
order of EF by KBT which is of the order
of T by TF this is the Fermi temperature
divided by the temperature ok this is
far far greater than 1 so e raised to X
by 1 plus e raised to X the whole square
ok for minus
EF by Zeta this is a very large value so
the minus value of this would all
approximately be of the order of one
plus a very small number the whole
square where this number is much much
less than one which is of the order of e
raised to X so therefore you're and this
itself tends to 0 so already this number
is very large - EF by Zeta the lower
limit of integration is already a very
large number and it is almost making the
contribution from this integral so
therefore if you take X square into e
raised to X by 1 plus e raised to X the
whole square this term goes to 0 as X
tends to minus EF by Zeta because EF by
Zeta is EF by KBT which is of the order
of TF by T which is much much greater
than 1 it's of the order of in fact this
term is of the order of 10 raised to
plus 2 and so if you go to a - value of
this these exponential terms become very
small and tend to 0 and so this term
tends to 0 if you are going towards the
negative limit so already this term is 0
so instead of going to this value if I
take it to minus infinity I don't make
any this lower limit which is EF by Zeta
if I take it towards minus infinity the
lower limit by taking it to minus
infinity I don't do any error in this
integral so instead of keeping the
limits of integration from minus EF to
Zeta I will change it to minus infinity
and I don't make any errors by doing
that because I already see that at minus
EF by Zeta the integral is already
contributing to only zeros and so if I
take it to minus infinity even larger
values it will make it even for the
zeros so in this negative direction I
can take it to minus infinity so KBT in
to gef integral minus infinity to play
infinity x square e raised to X divided
by 1 plus e raised to X the whole square
DX and it turns out that this is a
standard integral whose value is PI
square by 3 this is a standard integral
in mathematics were integral from minus
infinity to plus infinity X square e
raised to X divided by 1 plus e raised
to X the whole square DX this definite
integral has a well-defined value which
is PI square by 3 and so now my specific
heat of the electron gas finally does
starts appearing which is PI square by 3
GE F the density of states at the Fermi
into KB square T this is my expression
for the density of states of the
electron gas this is my expression for
the specific heat of the electron gas
this is my expression for the specific
heat of electron gas as obtained from
some of Phil's model the first things
that you see from this is that the
specific heat C V is proportional to
temperature just as it was expected and
was seen in the experiments that at very
low temperatures when you remove the
contributions which are coming from the
vibrating ionic lattice what remains is
only the specific heat of the electrons
and if you look at the behavior
temperature dependence of the specific
heat of the electrons they are linearly
dependent on temperature and that comes
directly from the Summerfield's model so
this is one of the major successes of
the sama fields model the second success
is that if I look at this expression a
little bit more carefully if you recall
that G of e K was equal to 3 by 2 N by
EF divided by e by EF raised to half
okay because the density of states goes
as this and now if I look at the density
at EK equal to EF the density of states
is at the Fermi energy is 3 by 2 n by EF
because this becomes equal to 1 so the
density of states at the Fermi energy is
nothing else but 3 by 2 n by EF and if I
now replace this in this expression if I
replace the behavior of the density of
states at the Fermi energy I will get
the specific heat is equal to PI square
by 3 into 3 by 2 n kb into KBT by EF I
am only really rearranging terms this is
the expression 3 by 2 n KB where 1 KB I
have taken and joined it with this and
this is KBT by EF now this is nothing
else but PI square by 3 3 by 2 n KB by V
n is the number of electrons and V is
the volume KBT
by EF what is this this is nothing else
but three by two are the specific heat
of the electron gas per unit volume from
kinetic theory this is the specific heat
of the electron gas per unit volume from
drudes model or which uses the electron
gas model as kinetic theory we're using
the kinetic theory of gases you know
that it is going to be three by two R
for any electrons in this gas it will be
three by two R so you see that that
expression has returned back but the
specific heat of the electron gas is not
three by two R it has that term but it
is smaller than 3 by 2 R by an amount
which is this so the specific heat of
the electron gas is PI square by 3 into
3 by 2 R by V KBT divided by KBT F as
the Fermi energy is written as KB times
the Fermi temperature and T by TF is of
the order of 10 raise to minus 2 if you
look at PI square by 3 into if you put
in numbers for example for copper if you
put in the temperature and you calculate
the Fermi temperature which we have
calculated earlier this is about 80,000
for copper and temperature you take it
as about 300 Kelvin and if you put in PI
square by 3 you will find that the
specific heat of the electron gas is
compared to the value which is obtained
from kinetic theory
this ratio into pi square by 3 is of the
order of 0.01 it will bring down the
value by 0.01 compared to the kinetic
theory of gases so from the kinetic
theory of gases if you just were to
measure the specific heat of the
electron gas that value would be 3 by 2
R at some temperature T however if you
were to measure the specific heat of the
electron gas it would be linearly
temperature dependent and at this
temperature the actual specific heat
would be 0.01 times the 3 by 2 R value
this is a hypothetical value or a
theoretical value which is they obtained
by considering the gas classical having
a Maxwell Boltzmann distribution and it
will give you 3 by 2 R but if you use
the Summerfield's model and consider the
gas as quantum mechanical then it shows
that it is far more lower than 3 by 2 R
at this temperature okay so there is a
significant suppression with respect to
the classical theory because of using
the quantum mechanical nature of the
electron gas so this was a very major
success of the sama fields model because
now you could explain why is the
specific heat of the electron gas so
much lower than compared to what you
know from kinetic theory and also you
get the temperature dependence very
neatly this CV you can write it as PI
square by 3 density of states at the
Fermi energy KB Square t ok
this is your specific heat okay and this
term this constant C V can be written as
gamma times T the specific heat of the
electron gas can be written as some
constant into gamma times T where gamma
is called as the sama Feld's constant
gamma is the Summerfield's constant
where gamma is PI square by three
density of states at the Fermi energy
into KB square this has a very specific
value and this of course you can write
it as PI square by 3 into 3 by 2 density
divided by the Fermi energy into KB
square this Fermi energy you know is H
cross square by 2 m into 3 PI square N
raised to 2 by 3 we have already seen
this as the expression for the Fermi
energy so if you put that all of that in
here then you will get term which is
this term sama Phil's constant will be
given as pi square KB square n raised to
1 by 3 divided by H cross square 3 PI
square raised to 2 by 3 into m where m
is the mass of the electron in the solid
so the sama feels constant depends on
the density as well as the mass of the
electrons inside the solid this is a
function of
density and mass of electrons in the
solid so what you can do is that you can
actually calculate the sama fields
constant okay you can find out how much
is the value of the sama fields constant
for different metals because you know
that gamma is some function of the
density of electrons and the mass of the
electrons and for mass you just use the
bare mass of electrons whatever is the
bare mass of the free electrons you can
use that to calculate the sum of fields
constant and so for different materials
you can find out the value of the
Summerfield's constant milli joules per
mole per degree Kelvin square in these
units if you find out the sama Fields
constant you can determine it and
calculate it for different materials for
sodium the value which you measure is
one point three eight so as I had said
if you go to low temperature and you
measure the specific heat as a function
of temperature then at very low
temperature there is a linear dependence
of the specific heat so if you are very
far away from the room temperature at
very low temperatures you will get this
linear temperature dependence of the
specific heat which comes from the
electrons inside the solid and from the
slope of this the slope is equal to the
sum of its constant gamma so you can not
only calculate this is the calculated
value by putting in the density of
electrons which you can of course
measure using the hall if you recall
that the hall coefficient RH is equal to
plus minus 1 by n e so you can measure
if you measure the Hall effect through
the Hall effect if you measure the Hall
coefficient you can determine the
density or there is also the expression
for the density and you put in the mass
of the electron which is the bare mass
of the electron you will calculate the
sum of fields constant and you can
derive it you can get it for all these
different metals from an experiment you
can also determine the sum of fields
constant so this is theoretically
calculated and the experimentally
obtained value is one point zero nine
four so there is reasonably good match
next if you go to copper it is about
theoretically calculated value is zero
point six nine five the experimental
value is zero point five zero five
slightly different okay if you go to cop
gold zero point two nine seven two nine
and zero point six four two roughly they
are similar aluminum one point three
five is the value you should
theoretically get if you experimentally
measure it you will get a value of zero
point nine one two this is now starting
to become significantly different there
are materials like lithium where one it
is six one point six three and it is
point seven four nine more than 50%
difference there is a significant
difference as you between the
theoretical value and the experimentally
measured value strontium three point six
and it is one point seven nine zero if
you recall that the Summerfield's
constant is some constant n raised to 1
over three into m it's some constant so
if you take the ratio of observed value
which you are getting from an experiment
to the theoretical value assuming that n
is the same the density is the same
whether you are using it for an
experiment or for theoretical
calculation the density is not different
the difference between the observed and
theoretical value is the mass of the
electron which is actually present
inside the solid divided by
the actual free electron mass the free
electron mass the mass of an individual
electron is M II that is what you are
using in theory okay so the ratio of the
observed value to the theoretical value
so if you take the ratio of this to this
you will get the ratio of the observed
to the gamma Theory which is nothing
else but the actual mass of the electron
divided by the mass of the free electron
and that value for sodium is about one
point two six it's about one point three
eight one point one four one point four
eight two point one eight two point zero
so you can see that certain materials
the as if the electron inside the solid
the electron inside the solid becomes
heavy because the mass of the actual
mass of the electron is approximately
twice the mass of the free electron so
compared to the mass of the free
electron it looks like the M actual
which is what you get in an experiment
is almost twice in some metals is almost
twice that of the mass of the free
electron so this is an unusual thing
which comes up the moment you start
understanding where from is a specific
heat of the electrons coming up it leads
to another new issue you see that the
mass of the electron is no longer that
of a free electron mass but sometimes it
is higher and these are some of the
drawbacks which are not explained by the
Summerfield's theory in fact there are
certain class of materials in which
which are certain class of compounds
which are compounds like uranium based
compounds cerium aluminum compounds or
solicits like seee
CU 2 SI 2
in this if you do the same experiment of
measuring the specific heat and from the
slope measure the Summerfield's constant
and take the ratio of theory observed to
the theory you can show that the actual
mass of the electron in these metals can
be as large as thousand times the mass
of the free electron so here the
electron really has become heavy
compared to the free electron mass the
actual electron inside the solid has
become as if it's a very heavy particle
okay the electron inside the solid is no
longer behaving like a free electron
moving around freely with a mass M E but
it's mass has increased its inertia has
increased and these go by a very famous
name which are called as heavy fermions
these are not explained within the
sommerfeld's model so this is something
which is outside the purview of the
Summerfield's model but yet the
Summerfield's model actually lends a way
to discover such materials and these new
phenomenas where actually the electron
seems to be much heavier than what it
would be if it was free so we will
discuss next a few of these drawbacks
which come out from the Summerfield's
model although the Summerfield's model
has been very successful in explaining
the behavior of the specific heat which
comes out from the theory of the quantum
mechanical nature of the electrons it
has its drawbacks and this is one of the
drawbacks that the electron mass does
not behave like a the electron does not
behave really like a free electron its
mass is very different could be very
different from that of a free electron
and sometimes in some materials they can
be orders of magnitude larger than the
free electrons we will continue this in
the neck
[Music]
[Music]
[Music]
