Thermodynamics: 1. The Zeroth Law

Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

Thermodynamics

1. The zeroth law 1


2. The first law 3
2.1. The Joule experiment 3
2.2. Examples of reversible transformation 5
3. The second law 8
3.1. The Carnot theorem 10
3.2. The Carnot cycle 11
3.3. Entropy 14
3.3.1. Meaning of the macroscopic thermodynamic entropy 14
3.3.2. The direction of entropy 15
3.3.3. The second law in terms of entropy 17
3.4. The second law, microscopical point of view 17
3.5. Microscopic significance of the internal energy 21
3.6. Microscopic significance of work and heat 22
4. Unattainability of absolute zero: the third law 22
4.1. Atteinability of negative absolute temperatures 22
It is constructed as a mathematical theory, with several definitions and theorem, called laws.
It is constructed over the definitions of heath and temperature.

REMARK: work and heat are not properties belonging to bodies. Tey're ways of
transferring energy, in an ordered or disordered way.
Moreover, temperature is not related to the kinetic energy of particles (this is
valid only for perfect gas).

The laws of thermodinamic are 4. The zeroth law is the last one to be redacted.
1. The zeroth law
Before citing the zeroth law, we nedd several definitions.

Definition 1 A thermodinamic system is the observed physical system identified by the


set { xi }i I of variables used to describe it, called thermodinamic coordinates.

Once you choose the preferred thermodynamic system, the others are called surroundings
or environments. The universe is the maximal set of all thermodynamic systems
interacting with each other. There MUST be no other physical systems interacting with the
one in the universe.

Universe

Thermodynamic system

Environments

Definition 2 A system is called:


i) open if it can exchange both matter and energy;
ii) closed if it cannot exchange matter, but it can exchange energy;
iii) isolated if it cannot exchange matter nor energy.

By definition, the Universe is an isolated system.


Matter is intended as number of elementary constituents, so in a closed system, the number
of elementary constituents doesn't change.
EXAMPLE: a hydrostatic system is identified by the thermodynamic coordinates {P,V},
pression and volume.
Two thermodynamic system can be separated with two different types of wall:
1. adiabatic if they can independently assume and keep any value of thermodynamic
coordinates;
2. diathermic if they're not independent but they spontaneously change until they
reach a state in which they do not change anymore.

Definition 3 Two systems or part of a system separated by a diathermic wall are said to
be in thermal equilibrium when their thermodynamic coordinates have reached values
which are constant in time
Theorem 1 [The zeroth law] If a system A is found to be in thermal equilibrium with B
and B is found to be in thermal equilibrium with C, then it is also true that C will be found
to be in thermal equilibrium with A.

Thus we introduce a new physical quantity: temperature. All systems in thermal


equilibrium have the same temperature. This is an intensive quantity (intensive of the object,
not depending on how much of it one has).
2. The first law
There are two possible wy to heat or cool a body: work and thermal contact. Thinking in
terms of interaction, how do we imagine an exchange in temperatures? If two object have the
same mass and the same composition but two different temperature 𝜃 and 𝜃 the thermal
1 2

equilibrium is reached at the temperature


𝜃1 + 𝜃2
𝜃 eq = 2

Rearranging the equation we obtain:


m -
2 (𝜃 2 eq ) = m (𝜃 - 𝜃eq )
𝜃 1 1 (1)
taking into account also different masses. Something has been exchanged between the two
bodies. We call this quantity heat.
Definition 4 Heat (Q) is a way to describe the flow of temperature from a body to another.
We can rearrange equation (1) in order to predict the final thermal equilibrium for object of
different material, noticing that Q = -Q :
C m (𝜃 - 𝜃eq ) = C m (𝜃 - 𝜃eq )
1 2

2 2 2 1 1 1

Definition 5 [Calorie] When a gram of water increases or decreases its temperature by 1


degree, this is related to an amount of heat equal to one calorie.
Heat is measured using a calorimeter. If the temperature of a body changes, i associate it
to a new measurable quantity, heat. It's just a way to explain this exchange: it is not
temperature that passes from a body to another, nor the temperature multiplied by the mass.
We gave it the name of heat.
Is there a way to establish an equivalence between work and thermal contact, following the
idea that an exchange in temperature is due to this two quantity?
cicaioeiacio
2.1. The Joule experiment
There is a loss of total energy, due to the work made by friction:
1

2
mv mgh
2
<

It is measured an increase in temperature by the thermometer. This work is measured in


Joule, the amount of heat is measured in calorie, and it is measured as the amount of heat
necessary to reduce the temperature of the water from the heated state to the initial state.
We want to keep the parallelism with mechanics, so the work is negative if it is done by
some external environment towards the physical system, positive in the other direction. In
thermodynamics, work is positive if temperature increases, negative if it decreases. An
important observation is that a fixed quantity of work and heat produces the same effect:
If you wanto to have the same effect of 1 cal you need 4.186 J of work.
This is called the mechanical equivalence of heat and work. In other terms:
Q-W = 0 (2)
∮𝛿 Q- W 𝛿 = 0
B
∫ 𝛿Q- 𝛿W = depends only on A and B
A
It means: property of the system in the state B minus property of the system in the state A,
so it is the variation of the same property of the system in two different states. The equation
(2) is due to the fact that the sign is contained in Q and W, and they are choosen in two
opposite way. It is wrong to say that heat converts into work and viceversa.
This is called a function of state, and this lead us to introduce the first law:
Theorem 2 [First law of thermodynamics] The quantity Q-W is an internal property
of the physical system, it is an energy, so it is called internal energy and so

Q-W U B -U A U (3)
Q- W dU
= ( ) ( ) = 𝛥

𝛿 𝛿 =

You can change internal energy either by thermal contact, which is called heat exchange,
or by work made on the system

The symbol 𝛿 is due to the fact that it is not a mathematical differential, it means "a small
amount of exchange".
Work is considered negative when it is don on the sistem, which explains the negative sign in
the formula (3).
For hydrostatic systems, F = Pext A, and the work is W = F ⋅ d = Pext (A ⋅ d) , for
d the height of the fluid, Pext the pressure exerced on the system, so:
W Pext V
W PextdV
= 𝛥

𝛿 =

If the transformation is quasi-static (or reversible) then Pext = Pint .

2.2. Examples of reversible transformation


For perfect gas measures in n numbers of moles, we will consider three transformation for a
reversible process: at constant volume, at constant pressure and at constant temperature.
• We say that a transformation is isochoric if it happen at constant volume.
P Pf
Pi
V

𝛥 U Q nCV
= = 𝛥𝜃 is a function of state, so it doesn't depend on the path.
• We say that a transformation is isobaric if it happen at constant pressure.
P
Vi Vf
V

The amount of work done in this transformation is W = P𝛥V, combined with


PV = nR𝜃 for ideal gas, we obtain W = nR𝛥𝜃, which depends only on the initial and
final states, so it is a function of state. We use again Q = nCp 𝛥𝜃, which allows us to find

nCV nCp -nR


Cp -CV R
𝛥𝜃 = 𝛥𝜃 𝛥𝜃

• We say that a transformation is isothermal if it happen at constant temperature.

P nRV=
𝜃

All the amount of heat exchanged will be absorbed by the gas and it is used to increase
the volume, in such a way it will mantain the temperature constant. It means 𝛥U = 0,
but
it doesn't mean that the heat exchange is zero!

U Q-W
Q W
𝛥 = = 0

as expected, so:

𝛿W pdV nRV dV
= =
𝜃
W
f nR𝜃
dV nR ln Vf Q
= ∫
i V Vi = 𝜃 =

• We say that a transformation is adiabatic if it happen at Q = 0.


So dU = -𝛿W, and rearranging:
nCV d - nRV dV
𝜃 =
𝜃

d
𝜃
𝜃
= - CRV dVV Cp -C
=
V dV
(

CV V - - dV
V
)
= (𝛾 1)


C
with 𝛾 = p , so integrating:
CV
-
f V f
ln 𝜃i = ln Vi = 0
𝜃
𝛾 1

𝜃f V
f = 𝜃i Vi
- 𝛾 1 - 𝛾 1
(4)
𝜃V
- = constant
𝛾 1

PV
𝜃 =
nR
PV 𝛾
= constant
The relation will lead to a steeper graph.

It is a very low process...


• Adiabatic free expansion of the ideal gas means an expansion with Q = 0, which
is against nothing, so Pext = 0.
Gas Vacuum

If we open the wall, the gas will distribute uniformly in the whole space. We wait until
a final equilibrium state, in which the gas will be expansed without a heat exchange
(the system is isolated). What is the amount of work?

Pext W
U
= 0→ = 0

nCV U
𝛥 = 0

𝛥𝜃 = 𝛥 = 0 → 𝛥𝜃 = 0

The intermediate state are not equilibrium state, we know nothing. This is an
irreversible transformation, and it is the most irreversible transformation: nothing
changes, you know only the initial and the final states, nothing in between.

3. The second law


The zeroth law tells us that if we put together two bodies with two different temperature
TH > TC tend to reach a thermal equilibrium: the hot body will become colder, the cool
body becomes hotter. We describe this process heat flow. If we want the inverse process, we
need a machine, it's not a spontaneous process. This leads us to the formulation of Clausius
of the second law of thermodynamics. Work can be lost spontaneously, and the result of this
is heat. This conversion is also called dissipation. The entire conversion of work into heat is
possible! The inverse process, so that a body which is cooling entirely releasing energy in
term of positive work is not observed, unless we have a part of transfer of heat into a body
which has a lower temperature.
The machine is working cyclically, so it returns to the same state several time. If you apply
the forst law, 𝛥U = 0, so W = Q = QH + QC with QC > 0, QH < 0. It results that

| QH W QC
| = | |+| |

This bodies are called sinks or reservoirs.


The process of transfer heat from a hot body to a cooler one is spontaneously possible.
In the reversed case W > 0, QC < 0, QH > 0 . So this heat exchange produces work!
Much more desirable...

Definition 6 The efficiency of a heat machine is


W
𝜂 =
QH
| |

The efficiency of a refrigerator is


| QC |
𝜔 =
W
| |

Theorem 3 [Clausius statement] It is impossible to devise a cyclic process that


transfers heat from a body at a given temperature to a body at a higher temperature.

Theorem 4 [Second law, Kelvin-Planck formulation] It is impossible to devise a


cyclically operating device, the sole effect of which is to absorb energy in the form of heat
from a single thermal reservoir an to deliver an equivalent amount of work.

The meaning of the second statement is that you cannot convert all heat released from a hot
body into work, you always need two bodies, another colder to receive a small amount of heat.
This means that it cannot happen that 𝜂 = 1 or 𝜔 = +∞, becouse |QC |, |W | can never be
0, and

W QH - QC
= | | | | (5)

The second law is valid also for reversible transformation! For every physical process, |QC |
and |W| can never be zero.
Figure 1: Second law of thermodynamics
The only transformations are in the reservoirs.

Theorem 5 The two statements are equivalent.

PROOF. Suppose is valid the first statement, and suppose that a cold body realeases |QC |
into a hot body, and there is a cyclic machine that doesn't violate the second statement
through which the cold body releases a quantity of heat of |QC | to the cold body. In
definitive, the heat transferred by the cold body is 0...
The converse: create a machine that violates the second statement, so the cold body releases
|Q | to a machine that converses it entirely into work spend by another machine to pass |Q |
1 2

from the same cold body to a hotter body. The total amount of work is 0, the heat extracted
by the cold body is |Q | + |Q | and it is the same received by the hot body with 0 work.
1 2

This violates the first statement. □

3.1. The Carnot theorem


A little part of energy coming from the cold body will be lost, so converted into (not usable?)
work.
What if we want to produce work? The aim of scientists is to maximize the theoretical
efficiency of a thermal engine: it must be less than one by the second law of
thermodynamics.
Theorem 6 [Carnot theorem] The efficiency of the ideal reversible machine is the same
independenlty from the machine (from the substances used to realize the cyclic process).
If the machine is irreversible (real), the efficiency is lower.
𝜂 rev = constant

𝜂 irr ⩽ 𝜂 rev

PROOF. Suppose that there are two reversible engines with two different efficiency, for
example 𝜂 < 𝜂 , this means that
2 1

W W | | | |
𝜂2 =
QH QH
| |
<
| |
= 𝜂1

This means |Q H | > |Q H |, |Q C | > | | Q C . Now let flow this quantity between two
2 1

reservoirs, so |Q H | = |W| + |Q C | and the same for the other engines, but in the reverse
2 1 2 1

2 2

way. Consider the whole system:


W = |Q H | - |Q C | = |Q H | - |Q C |
2 2 1 1

in which the work produced by one system is used to power the second (the work is zero).
You're violating the second law of thermodynamics (the clausius statement). This means
𝜂 ⩾ 𝜂 . Use the same argument with swapped engine.
2 1 □

3.2. The Carnot cycle

Definition 7 It is called a Carnot cycle any cyclic reversible transformation between


two reservoirs at no more than two temperatures.

In other word: two isothermal transformation at a fixed temperature, two adiabatic


transformations without exchanging heat. The Carnot theorem says that the efficiency is
independent from the reversible machine, so we make the calculation in a simple case! This
ideal case is the ideal gas in a Carnot cycle.
If the temperatures are fixed, the system will follow an adiabatic transformation to reach the
equilibrium:
P a
TH b
Adiabatic transformation
TC c d
V

W |QH - QC
- QQHC
| | | | |
𝜂 =
QH
| |
=
QH | |
= 1
| |

We have already calculated (use for perfect gas instead of T):


𝜃

QH nR H VVab
= 𝜃 ln
QC C | | 𝜃
ln
Vc
Vd
QC nR C VVdc QH H Vb
⇒ = ⋅
𝜃
| | = 𝜃 ln ln
Va

Using now (4)

𝜃H Vb
- 𝛾 1
= 𝜃 C VC
𝛾
Vb Vc
𝜃H Va
- 𝛾 1
= 𝜃C V
d
𝛾

Va Vd =

Vb - Vc
ln
Va Vd = 1

𝜃C
riv = 1 - 𝜃 (6)
𝜂
H
It is a very important result, showing that every reversible machine has this efficiency. To
calculate this, you have to use the thermal scale of perfect gas!
You see that if 𝜃C approaches the absoulte zero, then the efficiency would become 1 (which is
impossible).

30/11/21
Since the efficiency is independent from the substance (see (6)), we can give a new definition
of temperature. Before, we need to choose a substance and a property changing with
temperature, and we associated a number to this property that is changing. It contains a
number of arbitrary choiches.
We can do something better of the scale of temperature for the ideal gas. It contains the
absolute zero for gas, that is the zero also for all other substances! Tis is becouse the
efficiency is the same for all bodies. I am extending the meaning of the scale of temperature
for gas ⟹ absolute scale of temeprature. I create a Carnot cycle to study the temperature of
a body, using water at the triple point 273.16 K as T and I define the new temperature as
Q K. The scale is the same as the celcius degrees scale, but the zero is
2

T 1 = 273.16
Q
| 1|

shifted to - K. The absolute zero is an actual zero! you cannot pass through it, it's
| 2|

273.15

effectively the bottom of the scale.


Now we can say that C
T QC , and the sign of QC is negative if measured by the
| |

TH QH
=

T Q
system, so C - C . Tis amounts to say that, if the process is reversible,
TH QH =

QC QH
TC TH+ = 0

In the case of an irreversible cycle, the efficiency


𝜂 = 1- QQHC is less than the reversible
| |

case becouse if we take the same amount of heat from the col body, QC irr QC rev , so
| | > | |

the work is lower! To see this consider (5). In this case,

QC QH
T⏡⏣⏣
⏠⏣⏣
C TH

+ ⩽ 0

<0

and it will be equal to zero only in the reversible case. Generalizing, ∑ QTii ⩽ 0 , and for a
continuous cycle:
Theorem 7 [Clausius' inequality] For continuous cycles, ∮
𝛿 Q ,where the
equality holds only for reversible cycles.
T ⩽ 0

3.3. Entropy
B
Remember from Joul's experiment ∮𝛿Q- 𝛿W = 0 ⟹ ∫A 𝛿Q- 𝛿W = U(B) -U(A) .
But we have made a choice on the name of U, it's not an absolute choice. But this leads us to
say that
B 𝛿Q

A T rev = S(B) -S(A)
since it is 0 if the cycle is reversible. This new function S is called entropy. This is the
macroscopic definition of entropy and we measure it in [J / K] .
For the irreversible case, we close the cycle with a reversible transformation and obtain
𝛿Q B 𝛿Q A 𝛿Q

T irr = ∫A T irr + ∫B T rev ⩽ 0
B 𝛿Q B 𝛿Q

A T irr A T rev = S(B) -S(A)
⩽ ∫

If the process occurs in an isolated physical system, ∫A


B Q
𝛿
, and so
T irr = 0

S B -S A
( ) ( ) ⩾ 0 .

Theorem 8 In an isolated physical system 𝛥Q = 0, and 𝛥S ⩾ 0, and it's equal to zero


only if the case of a reversible transformation.

The universal entropy difference is positive in the case of an irreversible transformation, zero
for a reversible transformation. The universe entropy increas es in every irreversible
transformation.

3.3.1. Meaning of the macroscopic thermodynamic entropy


If the universe is composed by a machine and two reservoirs, since the machine is doing a
cycle, its difference in entropy is zero, independently by the nature of the process. Since the
Q
cold sink isn't changing its temperature, its difference in entropy is C , while for the hot
TC
QH
reservoir is - . Notice that Wrev -Wirr = |QC |irr - |QC |rev is the amount of work lost
TH
by the irreversible process (since QH rev QH irr ). And
=

TC 𝛥Suniverse irr = |QC |irr - QTHH TC = |QC |irr - |QC |rev


, ,

3.3.2. The direction of entropy


Suppose the system is changing from the state A to the state B, thermodynamic equilibria.
𝛥S A → B is the same for reversible and irreversible transformation (it's a state function). The
point of the difference of this preocesses is what is happening in the surroundings, in the
environment. For the reversible process, 𝛥Stot = 0, so 𝛥Srev = - 𝛥Ssurroundings , and for
the irreversible process 𝛥Stot > 0, so 𝛥Srev can be negative, provided that 𝛥Ssurroundings
is positive and greater than the absolute value of the difference in entropy of the
thermodynamic system.

Figure 2: p-v and T-S diagram for the Carnot cycle

We define dS =
Q dU
T rev T rev V constant , since Q dU for constant volume. So
𝛿
= 𝛿 =

dU
1
(7)
T Q rev V constant
=
𝛿
,
In the general case,
S Q ∫ nCV dT nR dV
f 𝛿
𝛥 =
i T rev

T V = +

𝛥 S nCV TTfi nR VVfi


= ln + ln

• for isochoric transformation V is constant:


𝛥S = nC V ln
Tf
Ti
• for isothermal transformation T is constant:
𝛥S = nR ln
Vf
Vi
• for isobaric transformation P is constant:
𝛥S = nC P ln
Tf
Ti
• for adiabatic reversible transformation:
𝛥 S = 0

01/12/21
In an isothermal process, Ti = Tf , while for the volume, it has increased with the law
Pf / Pi = Vi / Vf . Now we want to calculate 𝛥Sf i in an adiabatic transformation:

f 𝛿Q
𝛥S f i = ∫i T

rev

P
isot
her
mal

Vi Vf V

We are sure that the initial and final position will lie on an isothermal tranformation, also
if it isn't. The adiabatic free expansion is an irreversible process, while the isothermal is a
reversible transformation. So their difference is not in the initial and final state, it's the way
the system relates with the surroundings. In the process where a gas expands,
𝛥 Ssystem = 0, becouse the system is isolated, so it isn't exchanging heat nor work.
Substitute the vacuum with a piston, so the expansion is very low, and a source of heat to
compensate the loss of heat and lead to a stable temperature.

Piston
Gas

The difference between reversible and irreversible is all in the universe! Is it isolated? Does it
wxchange heat with others?

3.3.3. The second law in terms of entropy


Theorem 9 The entropy in the universe increases in the course of any spontaneous
process.

This means that 𝛥S > 0 in any spontaneous process. If the Kelvin statement wouldn't
hold, the system will have 𝛥S < 0, becouse there is only loss of heat, which is converted
entirely into work. If the clausius' statement would be false, we still would have 𝛥S < 0,
becouse TC < TH and the same amount of heat, so |𝛥SC | > |𝛥SH | but the first is
negative...
Spontaneous processes leads to a tendency to thermal equilibrium.
In the final state, Ti = Tf . If the process involves an increasing entropy, the final state
(thermal equilibrium state) will have the maximum amount of entropy of the universe.

3.4. The second law, microscopical point of view


Let's consider an adiabatic free expansion. The number of molecules N is the same.
Figure 3: Adiabatic free expansion

The probability to find a molecule decrases in time. The initial state is asymmetric and
ordered. The final state is symmetric, uniform and disordered. Pay attention to the
terms you use!

Definition 8 The thermodynamic probability 𝛺 is the number of ways in which the


particles can be distributed among all the possible energy states for the particle.

We have a discrete finite number of particles N and states. Each particle will be defined
to be in a certain state. You place the particles in a certain state of energy. Take an energy
level and say how much particles have this value of energy; then calculate alll the possible
permutations (it is a mathematical intuition!). If the internal energy is associated to a
thermal equilibrium state (so to the temperature), I have the restriction that the total
internal energy must be U. The equilibrium distribution is the distribution that maximize
the number of way I can permutate the energy states. In fact, in the previous experiment, the
final equilibrium state corresponds to the maximum disorder.
Maximum S [J / K] corresponds to thermal equilibrium, which corresponds to
maximum 𝛺, so there must be a relationship between entropy and thermodynamic
probability, that is microscopical. How to relate entropy to a number of probability? It must
eb S = f(𝛺), but which kind of function? Let's compare entropy property to the probability
property. Suppose we understand the absolute meaning of entropy, not only the difference in
it (𝛥S), and take two different physical systems with SA and SB . it's an extensive property
of the system. The total entropy is S = SA + SB = f(𝛺) = f(𝛺A ) + f(𝛺B ) . In
probability, 𝛺 = 𝛺A · 𝛺B , so
f(𝛺A · 𝛺B ) = f(𝛺A ) + f(𝛺B )
And the only function with this property is the logarithm, so
S ⏠⏣⏣
const
=
⏡⏣⏣

· ln 𝛺

J
K
Remember that PV = nR𝜃 =
N R Nk k R J the Boltzmann constant.
NA NA K
𝜃 = 𝜃, =

We will demonstate that this constant is in fact k.


PROOF. Consider an adiabatic free expansion
𝛥S = nR ln
Vf = k ln Vf (8)
Vi Vi
calculated from a macroscopical point of view. Let's consider now the microscopical point of
view. We will make an initial hypothesis: we suppose that each molecule occupies a finite
volume Vn . It will be a contradiction in the case of perfect gas, but ion the end it will
disappear. So the probability for each molecule is V / Vn . In the final state, the number of
possible state is higher (Vf / Vn ) than the initial state (Vi / Vn ). The number of particles is
N, so the number of ways I can put this particles in a certain number of state is
N N
𝛺f =
V f V i
Vn , 𝛺i = Vn
And, since
𝛥S = const(ln 𝛺 f - ln 𝛺 i )
N
= const ⋅ ln
𝛺f
= const ⋅ ln
V f
𝛺i Vi
And by (8), we conclude that this const is in fact k. So it holds

S k= ln 𝛺 (9)

07/12/21
This is called the Bolzmann identity. (9) refers to the entropy of the whole system. This
definition is based on the identification of 𝛥S = ∫𝛿Q / Trev . Then we were able to relate
the entropy to the so called thermodynamic probability represented by 𝛺, and we also
considered a constant that connects the number to the physical quantity (k = R / NA ) using
the behavior of the ideal gas.
Now, this definition of entropy opens the way to the statystical thermodynamic. You have
to recognize the existence of a finite number of energy states. We were able to identify
the probability with the entropy.
Now, given a certain temperature (a state of thermal equilibrium), which is the distribution
of states that corresponds to this temperature? Thermal equilibrium corresponds to masimum
entropy, that is maximum distribution. So calculate 𝛺 and find the maximum of this function
with the constraint of the internal energy given by the thermal equilibrium and of the
number
of particles that must be fixed. We will number the energy levels (𝜖i ) and add the
degeneracy gi and count how many particles ni occupy every state, but you have the
restriction that
i

ni = N and the restriction on the total energy.
In the non degenerate case 𝛺 =
0n !n !n ! … nN !
1 2
i

N! , with ni𝜖i = E. How is this
calculation done? It can be demonstrated (see the file, Lagrange multiplier...) that this result
depends on a parameter 𝛽:

ni e - i P i 𝛽𝜖

N ∑j e - j
=
𝛽𝜖
= (𝜖 )

It can be seen as a probability distribution, called the Boltzmann distributionB or


partition function. Notice that 𝛽 depends on the temperature. We need to find the exact
relationship between this parameter and temperature. It can be demonstrated (see moodle)
that:

1
𝛽 =
kT
- kTi
e = P(𝜖i).
𝜖

and so j
∑j e- kT
𝜖

Keeping constant the number of particles, if you decrease temperature, this leads to increase
the number of particles that occupy the lowest energy levels and decreasing those which
occupy the highest value of energy, and viceversa. If T → 0, the particles will be all in one
level, if T → ∞, the distribution will be uniform among all the possible states.

3.5. Microscopic significance of the internal energy


We know that the absolute internal energy is E = ∑ ni i, so if we take the average,

𝜖

E N
/ = ni i N . It depends on the temperature, and this allows us to say that
𝜖 /

temperature is related to internal energy. If we change physical system, energy states might
change: the sequence depends on the physical system. Also if the temperature is the same, it
doesn't mean that the total energy is the same! The temperature says only how to distribute
the atoms in that specific sequence of energy level, that change from physical system to
physical system.
Two physical system at the same temperature do not necessarily have the same average
internal energy becouse their energy levels are different.
Internal energy doesn't mean kinetic energy, temperature is not a measure of the kinetic
energy of a system of particles: this is true only for the ideal gas. In the pther case, we also
have potential energy due to the interactions of the particles of the system (configurational
energy).
Suppose now we have a continuous distribution of energy levels.

- kTE
P E dE e dE
- kTE
( ) =


0
e dE
This is the probability distribution of occupation of an energy level in a continuous
distribution of the value of E. This is consideren in non quantum physics. Each state
containes both the kinetic part and the configurational part of the energy. Calculating the
average internal energy we get:

- kTE

Ee dE
E kT

0

- kTE
< > = =


0
e dE
3.6. Microscopic significance of work and heat
Work means that the system exchanges energy with the surroundings making use of the
uniform motion and uniform potential energy in an external field: it is an order
way. Heat makes use of the internal energy random motion and internal
configurational energy of the atoms. From a mathematical point of view, the result is that
in the case of work there is a uniform shift in the energy levels, while the number of particles
in a state remains the same. It means that temperature changes. Heating means that the
sequence of energy levels remains the same, while the number of particles per state changes.
• Work:∑ ni d i ;
• Heat:∑ dni i .
( 𝜖 )

( ) 𝜖

4. Unattainability of absolute zero: the third law


Remind that the efficiency of a refigerator is the amount of heat that you can extract from
the cold sink divided by the work you have to do to extract this heat.

𝜔 = | QC W QC QH - QC
|/| | = | | / (| | | QH - QC QC
|) = 1 / (| | | TH TC -
|) / | | = 1/( / 1)

It means that in a practical refigerator you cannot obtain TC , otherwise this shall be
= 0

infinite.

Theorem 10 No finite sequence of cyclic processes can succeed in cooling a body to


absolute zero

So 𝛽 is more natural than the absolute temperature, becouse it would be infinite if T = 0.

14/12/21
From the microscopical point of view we have a discrete number of particles and find the
possible distribution on the discrete energy states in the thermal equilibrium, which
corresponds to the maximum of the distribution.

4.1. Atteinability of negative absolute temperatures


Consider a physical system with only two energy states: a ground state E and a second state
above its energy, E. Notice that, from the quantum theory, such a system exists. Notice that
0
nE E n N N number of total particles. Take nnE e- E kT , and a consider
( )+ 0( 0) = =
𝛥 /

T : all the particles are in the ground state. If T , then nE n N and it's
0

= 0 →∞ = 0 = /2

a uniform distribution.

E E
T T T
E E
= ∞ > 0 = 0

0 0

We can give a meaning of the attainability of negative absolute temperature. Take it as an


hypothesis: we would have something like this:
nE = e kTE|𝛥 |

n
| |

If T - , nE n N , but if T
→ ∞ = 0 = /2 →0
- this means that n 0 = 0, nE N .
=

E E
T - T T
E E
= ∞ < 0
= 0
0 0

The behavior of 𝛽(E) is more informative of what is happening (see the slides for graphs):

Negative values of absolute temperature has only a microscopical meaning, but negative
temperatures are "higher" than the positive ones! This is becouse the particles occupy higher
energy states. We would have a flow of heat from the negative temperatures to higher
temperatures. You can obtain such a state, but this can be reached only in a restricted finite
time.
This is the end...?

You might also like