Physics 301 18-Sep-2002 3-1

Pressure
Last lecture, we considered two systems with entropy as a function of internal energy,
volume and number of particles,
(U, V, N) =
1
(U
1
, V
1
, N
1
) +
2
(U
2
, V
2
, N
2
) .
We allowed them to exchange internal energy (that is, they were placed in thermal con-
tact), and by requiring that the entropy be a maximum, we were able to show that the
temperature is
1

=
_

U
_
V,N
.
Suppose we continue to consider our two systems, and ask what happens if we allow
them to exchange volume as well as energy? (Were placing them in mechanical as well as
thermal contact.) Again, the total entropy must be a maximum with respect to exchanges
of energy and exchanges of volume. Working through similar mathematics, we nd an
expression for the change in total entropy and insist that it be zero (so the entropy is
maximum) at equilibrium,
0 =
=

1
U
1
U
1
+

2
U
2
U
2
+

1
V
1
V
1
+

2
V
2
V
2
=
_

1
U
1

2
U
2
_
U
1
+
_

1
V
1

2
V
2
_
V
1
,
from which we infer that at equilibrium,

1
U
1
=

2
U
2
,
which we already knew, and

1
V
1
=

2
V
2
.
This last equation is new, and it must have something to do with the pressure. Why?
Because, once the temperatures are the same, two systems exchange volume only if one
system can push harder and expand while the other contracts. We dene the pressure:
p =
_

V
_
U,N
.
We will see later that this denition agrees with the conventional denition of pressure as
force per unit area.
Copyright c 2002, Princeton University Physics Department, Edward J. Groth
Physics 301 18-Sep-2002 3-2
Chemical Potential
Well, theres one variable left, guess what were going to do now! Suppose we allow
the two systems to exchange particles as well as energy and volume. Again, we want to
maximize the entropy with respect to changes in all the independent variables and this
leads to,
0 =
=

1
U
1
U
1
+

2
U
2
U
2
+

1
V
1
V
1
+

2
V
2
V
2
+

1
N
1
N
1
+

2
N
2
N
2
=
_

1
U
1

2
U
2
_
U
1
+
_

1
V
1

2
V
2
_
V
1
+
_

1
N
1

2
N
2
_
N
1
.
So, when the systems can exchange particles as well as energy and volume,

1
N
1
=

2
N
2
.
The fact that these derivatives must be equal in equilibrium allows us to dene yet another
quantity, , the chemical potential
=
_

N
_
U,V
.
If two systems are allowed to exchange particles and the chemical potentials are unequal,
there will be a net ow of particles until the chemical potentials are equal. Like temperature
and pressure, chemical potential is an intensive quantity. Unlike temperature and pressure,
you probably have not come across chemical potential in your elementary thermodynamics.
You can think of it very much like a potential energy per particle. Systems with high
chemical potential want to send particles to a system with low potential energy per particle.
Note that we can write a change in the entropy of a system, specied in terms of U, V ,
and N as
d =
1

dU +
p

dV

dN ,
or rearranging,
dU = d p dV + dN .
Which is the conservation of energy (rst law of thermodynamics) written for a system
which can absorb energy in the form of heat, which can do mechanical pV work, and which
can change its energy by changing the number of particles.
Copyright c 2002, Princeton University Physics Department, Edward J. Groth
Physics 301 18-Sep-2002 3-3
Probability
Here, we will introduce some basic concepts of probability. To start with, one imagines
some experiment or other process in which several possible outcomes may occur. The
possible outcomes are known, but not denite. For example, tossing a die leads to one
of the 6 numbers 1, 2, 3, 4, 5, 6 turning up, but which number will occur is not known
in advance. Presumably, a set of elementary outcomes can be dened and all possible
outcomes can be specied by saying which elementary outcomes must occur. For example,
the tossing of the die resulting in an even number would be made up of the elementary
events: the toss is 2 or the toss is 4 or the toss is 6. A set of elementary events is such
that one and only one event can occur in any repetition of the experiment. For example,
the events (1) the toss results in a prime number and (2) the toss gives an even number
could not both be part of a set of elementary events, because if the number 2 comes up,
both events have occurred!
One imagines that a very large number of tosses of the die take place. Furthermore,
in each toss, an attempt is made to ensure that there is no memory of the previous toss.
(This is another way of saying successive tosses are independent.) Then the probability
of an event is just the fraction of times it occurs in this large set of experiments, that
is, n
e
/N, where n
e
is the number of times event e occurs and N is the total number of
experiments. In principle, we should take the limit as the number of trials goes to .
From this denition it is easy to see that the probabilities of a set of elementary events
must satisfy
p
i
0 ,
and

i
p
i
= 1 ,
where p
i
is the probability of event i and i is an index that ranges over the possible
elementary events.
The above denition is intuitive, but gives the sense of a process occuring in time.
That is, we throw the same die over and over again and keep track of what happens.
Instead, we can imagine a very large number of dice. Each die has been prepared, as nearly
as possible, to be identical. Each die is shaken (randomized) and tossed independently.
Again, the probability of an event is the fraction of the total number of trials in which
the event occurs. This collection of identically prepared systems and identically performed
trials is called an ensemble and averages that we calculate with this construction are called
ensemble averages.
You are probably thinking that for the die, the probabilities of each of the six elemen-
tary events 1, 2, 3, 4, 5, 6 must be 1/6. Well, they could be, but its not necessary! Youve
heard of loaded dice, right? All thats really necessary is that each p
i
be non-negative
and that their sum be 1. Probability theory itself makes no statement about the values of
Copyright c 2002, Princeton University Physics Department, Edward J. Groth
Physics 301 18-Sep-2002 3-4
the probabilities. The values must come from somewhere else. In general, we just assign
probabilities to the elementary events. Often we will appeal to symmetry or other argu-
ments to assign the probabilities. For example, since a die is a symmetric cube, no face
can be distinguished (mechanically) from any other face and we can plausibly argue that
the probabilities should be equal.
Aside: well, the dots have to be painted on, so the die isnt perfectly symmetric.
Presumably, dierences in the amount and pattern of paint on each face make a negligible
dierence in the mechanical properties of the die (such as cm, moment of inertia, etc.) so
its a very good approximation to regard the die as symmetric. However, some dice have
rather large indentations for each dot. Ive occasionally wondered if this might make a
detectable dierence in the probabilities.
In our discussion of the entropy, we postulated that a system is equally likely to be
in any microscopic state consistent with the constraints. This amounts to assigning the
probabilities and is basically an appeal to symmetry in the same way that assigning equal
probabilities to each face of a die is an appeal to symmetry!
Averages
Assuming there is some numeric value associated with each elementary event, we can
calculate its average value just by adding up all the values and dividing by the total number
of trialsexactly what you think of as an average. So, if event i produces the value y
i
,
then its average value is
y =
1
N
(y
1
+ y
1
+ y
1
. .
n
1
times
+y
2
+ y
2
+ y
2
. .
n
2
times
+ )
=
1
N
(n
1
y
1
+n
2
y
2
+ )
=

i
y
i
p
i
.
Quantities like y, whose value varies across an ensemble, are called random variables.
After the average, we will often be most interested in the variance (often called the
square of the standard deviation.) This is just the average value of the square of the
deviation from the average.
var(y) =
2
y
=

(y y)
2
_
,
where
y
is the standard deviation in y, not the entropy! The standard deviation is a
measure of the spread about the average.
Copyright c 2002, Princeton University Physics Department, Edward J. Groth
Physics 301 18-Sep-2002 3-5
Probabilities for Continuous Variables.
Rather than giving one of a nite (or innite) number of discrete outcomes, an ex-
periment might result in the measurement of a random variable which is continuously
distributed over some nite (or innite) range. In this case we deal with a probability
density rather than discrete probabilities. For example, we might make a measurement of
a continuous variable x. Then the probability that the measurement falls in a small range
dx around the value x is
Prob(x < result < x + dx) = p(x) dx ,
where p(x) is the probability density. Just as for discrete probabilities, the probability
density must satisfy
p(x) 0 ,
and
_
allowed range of x
p(x) dx = 1 .
We can simply dene p(x) = 0 when x is outside the allowed range, so the normalization
becomes
_
+

p(x) dx = 1 .
The average of any function of x, y(x) is dened by
y =
_
+

y(x)p(x) dx .
Copyright c 2002, Princeton University Physics Department, Edward J. Groth
Physics 301 18-Sep-2002 3-6
The Binomial Distribution
As an example of working with probabilities, we consider the binomial distribution.
We have N trials or N copies of similar systems. Each trial or system has two possible
outcomes or states. We can call these heads or tails (if the experiment is tossing a coin),
spin up or spin down (for spin 1/2 systems), etc. We suppose that each trial or system is
independent and we suppose the probability of heads in one trial or spin up in one system
is p and the probability of tails or spin down is 1 p = q. (Lets just call these up and
down, Im getting tired of all these words!)
To completely specify the state of the system, we would have to say which of the N
systems are up and which are down. Since there are 2 states for each of the N systems,
the total number of states is 2
N
. The probability that a particular state occurs depends on
the number of ups and downs in that state. In particular, the probability of a particular
state with n up spins and N n down spins is
Prob(single state with n up spins) = p
n
q
Nn
.
Usually, we are not interested in a single state with n up spins, but we are interested in all
the states that have n up spins. We need to know how many there are. There is 1 state
with no up spins. There are N dierent ways we have exactly one of the N spins up and
N 1 down. There are N(N 1)/2 ways to have two spins up. In general, there are
_
N
n
_
dierent states with n up spins. These states are distinct, so the probability of getting any
state with n up spins is just the sum of the probabilities of the individual states. So
Prob(any state with n up spins) =
_
N
n
_
p
n
q
Nn
.
Note that
1 = (p + q)
N
=
N

n=0
_
N
n
_
p
n
q
Nn
,
and the probabilities are properly normalized.
To illustrate a trick for computing average values, suppose that when there are n up
spins, a measurement of the variable y produces n. What are the mean and variance of y?
To calculate the mean, we want to perform the sum,
y =
N

n=0
n
_
N
n
_
p
n
q
Nn
.
Consider the binomial expansion
(p +q)
N
=
N

n=0
_
N
n
_
p
n
q
Nn
,
Copyright c 2002, Princeton University Physics Department, Edward J. Groth
Physics 301 18-Sep-2002 3-7
and observe that if we treat (for the moment) p and q as independent mathematical
variables and we dierentiate both sides of this expression with respect to p (keeping q
xed), we get
N(p + q)
N1
=
N

n=0
n
_
N
n
_
p
n1
q
Nn
.
The RHS is almost what we wantits missing one power of p. No problem, just multiply
by p,
Np(p + q)
N1
=
N

n=0
n
_
N
n
_
p
n
q
Nn
.
This is true for any (positive) values of p and q. Now specialize to the case where p+q = 1.
Then
Np =
N

n=0
n
_
N
n
_
p
n
q
Nn
= y .
A similar calculation gives
var(y) = Npq .
The fractional spread about the mean is proportional to N
1/2
. This is typical;
as the number of particles grows, the fractional deviations from the mean of physical
quantities decreases in proportion to N
1/2
. So with N
0
numbers of particles, fractional
uctuations in physical quantities are 10
12
. This is extremely small. Even though the
macroscopic parameters in statistical mechanics are random variables, their uctuations
are so small that they can usually be ignored. We speak of the energy of a system and
write down a single value, even though the energy of a system in thermal contact with a
heat bath is properly a random variable which uctuates continuously.
Copyright c 2002, Princeton University Physics Department, Edward J. Groth