noc19-ch17-lec19

IIT Kanpur July 201
20 followers
Dec 3, 2023
In this lecture we will discuss the derivation of generalized thermodynamic potentials
Okay, welcome back, last class we went through mathematical properties of fundamental equations
particularly the Euler equation and Gibbs Duhem expressions or equations.
So today we are going to talk about generalized thermodynamic potential okay, start with this,
so we essentially are going to focus on generalized thermodynamic potential.
So why do we interested in this generalized thermodynamic potential, one of the reason
is that, so till date we have just discussed in the class only U and S, so U is a function
of S, V, n1 till nr and S is a function of U, V, so these are natural variables corresponding
to U and S okay, this essentially can yield all the thermodynamic properties of a system
whether it is a mechanical or thermal equation of state.
So this can yield all the thermal info about a system okay, so one can do this exercise
for given system, one can use this expression of U and there is a variety of different constraints
which we have looked into to come up with some equation of state thermal or mechanical
but the problem is that most of the experiment which we conduct are not under a constant
entropy or volume or constant internal energy or volume in the moles, we mostly conduct
those experiments at a constant temperature and pressure okay.
So that brings a question that what would be the appropriate thermodynamic function
when we have only temperature and pressure and moles for example constant okay, so usually
experiments are usually conducted by controlling T and P okay, so thus the question is that
it will be convenient if we can find appropriate thermodynamic function where the natural variable
is T and P okay, convenient to use natural variable T and P instead of let us say instead
of S and V okay, so this is something which we would like to find a thermodynamic function
where the natural variable is T, P, so thus the question is there is a function of T,
P, N1 and NR that contains the same complete thermodynamic function as in U and S, contains
the complete thermodynamic information okay.
So that is a very simple question which we are trying to raise that you know okay we
would like to find the appropriate thermodynamic function where the natural variables are T,
P, N which contains the same complete information as in U and S, so the question is how do we
obtain a function with different variable from the original fundamental relation okay,
so that is the question.
So let us consider a very naive approach okay, so let us say if you start with function U,
so you may have this U as a function of S and you have this function U of S okay, so
what we can consider very naively that we can have this S, U as a function of S, I am
ignoring the other terms and you take the derivative of S okay which is nothing but
basically T of S and then you invert T1, some function T1, so that T which essentially is
going to be S, so for example T if you consider T1S, so corresponding to this, this would
be your del U1 by del S and if you invert which you can consider let us say somehow
if you can invert and S is going to T1, some function T1 of T okay and if you plug in the
original, plug in to the original function okay, then you may obtain something like U1 tilde of T,
this is nothing but U1 of T1 by T okay. Now with this kind of very simple derivative
based inversion, the problem is that let us say if you have 2 function, U1 which you have
considered this and U2 let us say is also function of S which is nothing but a simple shift here,
in that case if you take this derivative inversion process which is a naive approach,
you are going to get same as U1, so if you try to take the derivative, obtain an inversion and
then you invert it back somehow, plug in the original, you are going to lose the information
that is what it says, if it is a simple approach like this, where if you consider 2 function which
is just a separated by a constant, then this would be the case where we are considering U1,
T as well as U2, so a simple approach like this is not going to work okay.
So we need to think in little more clever approach, so this kind of approach which
you have used here, here what is happening is information is getting lost in transformation
okay, so in order to obtain a suitable function will be making use of some mathematical approach
and that is called Legendre transformation, so I am going to now describe that okay,
so the idea behind is that let us consider this function which is Y, the function of X and you
have this, this is let us say the curve and what we are trying to do is now, we are trying to define
the slope at different points okay, so this okay, so we have now let us say P1, P2, P3,
this is a kind of slope and the corresponding intercept would be 5 okay, so this is kind of
a slope, so if you consider generic definition of slope here P would be something like this,
Y at that point minus the corresponding slope divided by X, corresponding X minus 0,
so this would be the raw definition of slope okay, in other word I can define phi as Y minus
P of X alright. So, now what we are defining this particular function phi is basically we
are saying this phi as a Legendre transformation or transform of Y okay, now the question is how
do you recover Y, so to recover Y is equal to Y of X given if phi is a function of phi P is given
okay, so at the end of the day the problem with the naive approach was that we want to invert it,
I want to return back to the original function, we will lose the information in this and that
kind of transformation, so we will be looking at how Legendre transformation works where we
would like to get back or recover the original function without losing it and this is something
we can see how it works here. So, if you want to recover this let us look at it,
so recall that the definition of slope, so is also that P is also D of Y by D of X okay,
that also the definition of this slope, so this is also there and as well as we can consider the
Y at that given point minus this intersect divided by the total X, so this is also the definition.
So, with this I can say D of Y is nothing but P DX, now you can differentiate this expression,
so this will become D of Y, D of Y minus P DX minus X DP okay, this essentially is going to
be 0, so this is going to be minus X DP okay, so this is going to be 0, so you can recover this
by a simple definition that minus X is nothing but D phi by okay. So, let us start to summarize
here these two approaches or this particular approach that in the case of Y given as Y of X,
the definition we have to make use of is the definition of the slope, this is the
tangential transformation minus PX plus Y and now here we were going to eliminate X and Y
to yield phi is equal to phi of P okay. Now, in the case of inverse to recover the function,
we are going to use phi is equal to phi of P, the definition is minus of X is D of phi by DP
and Y now can be written as X P plus phi okay, so here we are going to eliminate P and phi to
recover or to yield Y is equal to Y of X okay. So, this is a latitudinal transformation which
we are interested in okay, so let us now try to extend this for multiple variables,
it is not just dependent only on X but many other variables, so that is what we can now
generalize this. So, generalization okay, so what we can consider now is Y as a function of X0,
X1, X3 and the derivative here can be written as P of K is equal to del Y by del X of K.
Now, this essentially means that this is a slope corresponding to one of the variable,
now if you consider the Legendre transformation or the function phi,
now here the phi would be P1 till P of T and this is nothing but this is a family of tangent
hyperplane and the definition of phi would be Y okay, minus summation of all the of the slope
multiplied by X of K okay, so this is by definition here, so if you take the derivative here D of phi,
then you can show that this is nothing but minus summation XK DP K okay, so this is something
which is there, so if you look at the definition of D phi okay, so this is your D phi, D phi is
nothing but minus X multiplied by DP right, where X is nothing but the partial, nothing but the
derivative in this case, when you are considering many variables, the X of K would be nothing but
del phi by del PK okay. So, again you have to pay attention to the details, what we have done
is that we would like to generalize this expression, so where the derivatives of this
function is at a given with respect to a specific XK, where rest of the variables are constant is
PK and now we are defining this Legendre transformation function phi, which is a
function of all these slopes and that is why we are calling this as a family of tangent hyperplane,
the definition of phi would be Y minus summation PK XK, now we are extending it for the one just
variable, it was simply Y minus PX but now we are having many independent variables,
so we are just saying summation PK XK, where XK is nothing but the partial derivatives of phi
with respect to PK okay. So, this is basically Legendre transformation, so that means what we
have done is Y, XY is equal to X1 XT, now this we are getting a Legendre transform function phi P1
PT, so we have to eliminate phi and PKs okay, so that is something which we have to, if you want
to do here, we have to eliminate Y and XKs okay, whereas in the case of inverse is going to be phi
and PKs. Now, you may have situation where only not all the variables you would like to change,
only let us say M plus 2 are independent variable and in that case, your function let us say is Y
as X0, X1, XM plus 1, XM plus 2 and so forth till XM. So, if you are interested let us say in some
situation, you might like to just take a Legendre transformation which I am saying as LT of only
M plus 2 independent variable, so that means you would like to only change okay or M plus 2 would
be okay, M plus 2 would be only this okay. So, if you are interested in only this expect
or to take the Legendre transformation, then your PK is going to be del Y XK, X of course J is not
equal to K that is what all the definition should be because others are constant, so essentially when
you are doing this, you are saying that XJ is constant but this is not equal to K and similarly,
here as well, so this is by definition but then K varies from 0 to M plus 1 for this specific case
where because we are only interested to take Legendre transformation of only M plus 2 independent
variable okay. So, in that case what would be Legendre transformation function, so this can
be written as phi is equal to Y minus summation K is equal to 0 M plus 1 PK XK okay and therefore,
your phi is going to be phi P0 P1 till P M plus 1, this is Legendre transformation of only M plus
1 variable and then you have this X M plus 2 till X M plus 1 okay.
So, we have been using a generic symbol such as Y but we can now interpret this Y as a some kind
of energy function such as U or S, so let us now try to use this Legendre transformation concept
for transforming the variables from our basic fundamental relation expressions of U to something
which we desire as far as the independent variables are concerned. So, let us look at it,
so if we interpret this function, now let us say we consider the case of energy okay,
consider energy expression or okay or the expression of fundamental equation okay,
so U is U as S1 you know and V or N1 till N sum R okay.
So, this is nothing but if you look at its Y is equal to function of Y, X1 and so forth right,
so here basically X is this, V is this, these are the natural variables for basically the U. So,
what we want is to derive a corresponding to intensive variable, so what we want is
to derive an expression or function, derive a function corresponding to intensive variable
T or P or mu okay, so this is what we want to do that.
So, let us look at it and essentially what you are going to see that all the derivation has
simply this particular kind of transformation where we are using Legendre function or
transformation okay, in energy or entropy and essentially we are going to get the thermodynamic
function okay, so we are going to apply this thing and we are going to get thermodynamic function
which we call potential functions for different kind of independent variables and that is something
which you would like to see that. So, let us take a case and let us say okay,
so let us take a case first where we are going to look at U, S, V, N, N1 till N sum let us say T
and then we would like to get a function such as it is T, V and N, so essentially what we are
trying to do is we are trying to change one of the variable this to T okay. Now, if you look at it
the way we have done it in the expression here that if you are interested in let us say M plus
1 variable to be Legendre transformed, then essentially we multiply the slope corresponding
to the kth variable multiplied by the variable X. So, this is what we are going to do that because
here it is only one variable S to be converted to T, so essentially our function should be something
like this the variable S, so we do not know what is S at this point, so let us first come up with
the slope S or slope P. So, the P here is nothing but the derivative of del U by del S at a function
of V and all Ni's are constant okay, so this is the slope right by definition and this value if
you look at the basic thermodynamic function in differential form is dU is TdS minus PdV plus
summation mui dNi, so for the case of V, Ni is constant dU by dS is nothing but T okay.
So, what we have is now a thermodynamic Legendre transformation function which is nothing but
U minus TS and this is what we define, so this is phi now, this is what we are defining is in
order to differentiate this with different Legendre function which we are going to get,
so this is this function is now we call it A okay, so this is nothing but is Helmholtz potential,
sometimes this is also called Helmholtz free energy, so this free energy concept will come
little later but I am going to just write it down okay. So, now you have a function A which is
nothing but U minus TS and remember now, is A is a function with T, V, N1, C and T right,
so if you take a differentiation of this A in a differential form, this is going to be dA is
equal to dU minus T dS minus S dT right. Now, you can use here the basic definition
of dU, this will be T dS minus P dV minus S dT plus summation mui dNi, so there is another term
here minus T dS okay, so this get cancelled this and this, so what you have is minus P dV minus S
dT plus summation mui dNi, so this is the differential expression of Helmholtz free
energy potential okay, that is something which became. So, now we have a function which basically
has a natural variable T, V, N and that is what A is okay, the Helmholtz free energy.
Now, this is one thermodynamic function okay, now what about others, so let us consider another
function, we are saying that well, in this case we wanted to have Legendre transformation from
for the variable S to something we call it T because we want to change the S to the derivative
of dU by dS and that is nothing but T but what about this, where let us consider U as a function
of S, V okay and we want this to change only, the volume has to be changed in a way which is nothing
but the derivative of U with respect to V but which is nothing but if you look at here,
the derivative of U with respect to V and others are other variables are constant is nothing but
minus P, we want to change this to P here okay. So, if you want to change this, what is the
natural variable, so let us look at it, so P is the pressure here, so what we want to do is,
we want to consider the function mu minus the slope multiplied by this variable V,
the slope is nothing but del U by del V and the rest S and ni's are constant okay and this is
nothing but minus of P okay, so you have a function U minus of minus P, V which is nothing
but U plus PV okay, so this is your Legendre function in order to change the variable SV,
N1 and so forth to SP and this we call it enthalpy, so this is your enthalpy.
So, enthalpy has a natural variable SP, N1 and NT, now if you want to show that this you written,
one of the important thing with the Legendre transformation is to obtain the original
information using the inverse approach, so that means the transformation back to the original
does not lose any information, so we should try and check whether we are getting the same,
if you redo this thing here and here, if I try this thing to get S and V back okay,
using the same approach here okay, so do we get this U okay, so we can try that.
Now, dh here is going to be dU plus PdV plus VdP okay and dU we know, this is nothing but TdS minus
PdV plus summation mu i dni plus this here PdV plus VdP, now this gets cancelled, so you have
an expression TdS plus VdP plus summation mu i dni, so if you look at the derivative of this
function, so because we want to change P to its derivative, so that means we are talking about
del H by del P and S and rest are constant, then this is nothing but if you look at this expression
okay, then this is nothing but your V okay, right. So, we have to have a function now,
H minus V times P and this is nothing but if you look at the original expression here by definition,
H minus PV is nothing but U, so essentially you are getting back the original expression,
so that means we are not losing any information, this is a transformation, so it makes perfect you
know value to our mathematical means to convert from one variables to another variables okay.
So, same thing we could have done that for A also, so you can get back A minus del A by del
T at constant V and N multiplied by T and this is going to be A minus del A by del T is nothing
but minus of S and T and this is by definition is A plus TS and this is nothing but if you look
at the definition of A, here A plus TS is nothing but U, so you could do the same exercise for the
Helmholtz free energy as well okay. Now, I can take it further for the case where we would like
to change not just one variable but then another variable, so I will but that can be done in the
next class, I will take a break here and then we will continue this exercise for obtaining the
thermodynamic function appropriate of various natural variables are TP okay and then we will
try to summarize this exercise okay, so I will see you in the next class.
