PhD – Environmental Fluid Mechanics – Physics of the Atmosphere

University of Trieste – International Center for Theoretical Physics

The van der Waals equation of state

by Dario B. Giaiotti and Fulvio Stel (1)

Regional Meteorological Observatory, via Oberdan, 18/A I-33040 Visco (UD) - ITALY

Abstract

This lecture deals with a more general form of the equation of state, called van der Waals
equation, which gives a better description of reality both under the conceptual and numerical
point of view. The distinction from gas and vapour naturally springs out from this equation as
well as the existence of an unstable region of the P-V diagram, nor foreseen neither described by
the classical ideal gas law. Even if this unstable region is not correctly described by the van der
Waals equation under the analytical point of view, with added assumptions (conceptual patch) it
is possible to give at least a qualitative description of reality. Last but not least the van der Waals
equation foreseen the existence of negative pressures. These negative pressures exist in nature
(even if they are unstable) and are fundamental for some biotic mechanisms

How to obtain the van der Waals (1873)

An equation of state is a relationship between the thermodynamic variables that fully define the
state of a system. If the system is air (that is essentially a mixture of nitrogen and oxygen) or
water vapour, four thermodynamic variables are enough to determine the state of the system and
they can be a combination of pressure, volume, mass (or number of moles) and temperature. In
the atmospheric physics the equation of state usually used is the well known

p V =n RT

where p is pressure n is the number of moles, T is the absolute temperature and R is the ideal gas
constant (R = 8.3143 J mol-1 K-1).

Another form, even more used in meteorology, is the following

p= ρ Ra T

where ρ is the ideal gas density and Ra is the gas constant (i.e., the ideal gas constant R divided
by the molar weight of the specific gas) that, for dry air, is Ra = 287.04 J K-1 .
Both the above equations have essentially the same physical meaning and the same physical
troubles.

The first of these troubles is the fact that when temperature decreases, keeping constant pressure
and number of moles, the volume occupied by the gas tends to zero as well. Of course this
cannot be physically true because, being material, the gas constituents cannot disappear. This
means that both the above equations of state cannot be used at extremely low temperatures.
However, this trouble can be easily solved in an heuristic way using a slightly different form for
the equation of state, that is

p V −V 0 =n R T
The advantage of this form of the equation of state is that when T tends toward zero at constant
pressure and number of moles the gas volume tends toward V0 . Then the gas constituents (atoms
or molecules) do not disappear, i.e., they still occupy the portion of space V0 . If V0 is enough
small for T ≫0 the modified equation of state has almost the same behaviour of the previous
forms.

Another trouble of the ideal gas law, neither solved by the above modified form, is that when the
number of moles increases at constant volume and temperature, it is quite natural to assume that
the gas pressure might not increase as expected (i.e., linearly with the number of moles) because
of the mutual attraction between the gas constituents. In other words the pressure exerted by the
gas on an virtual container is lower than that exerted by an ideal gas. This can be taken into
account modifying the above equation of state in the following form

n RT
p= − pn
V −V 0

The quantity p n is function of the number of atoms and/or molecules per unit volume as well as
from the average force exerted by two atoms and/or molecules. By simple inductive description
of the interaction between the gas constituents described in the below figure, assuming as φ the
average force exerted by the gas constituents within a radius of influence λ , it s possible to give
an analytic description of the term p n , that is

F N 2φ
p n= = 
S 2 S

where N is the number of gas constituents contained in the sphere of λ diameter and F is the
average force exerted on the surface S that divides the sphere of influence in two equivalent
hemispheres.

Figure: Inductive determination of the average force F exerted over the surface S with the increasing
of the number of gas constituents N into the sphere of influence.
Introducing the Avogadro number A and the total gas volume V, the number of gas constituents
N inside the sphere of influence becomes

nA 4 λ 3
N= ⋅ π 
V 3 2

and the above expression for the pressure correction assumes the form

[ ]
2
nA 4 λ 3
⋅ π 
V 3 2 φ .
p n=
2 λ 2
π 
2

Now, defining the molar specific volume as v m=V /n A , with some algebra we can write the
pressure correction as

π λ4
p n= φ
36 v 2m

which, even through this simple analytical approach, underlines the strong dependence of the
correction factor by the radius of influence (fourth power of λ) and by the constituents density
(second power of 1/v m ). If density is low as well as the radius of influence, no matter the
strength of the average force φ , the correction term becomes negligible. That is the reason why
many gases at standard atmospheric conditions can be considered as ideal.

Using the above pressure correction term in the modified version of the ideal gas law we obtain
the following “real” gas law, usually called van der Waals equation of state,

RT b
p= − 2
 v m−a v m

where a and b are two positive defined constants that depends from the particular gas or vapour
we are dealing with. In particular the a constant is related to the volume occupied by the gas
constituents while b is essentially related to the radius of influence for the intra-molecular
interaction.

The ideal gas law can be considered as a special case of the van der Waals equation of state, in
fact when both a and b go to zero the van del Waald equation becomes the ideal gas one. Apart
from this the above equation can give some useful theoretical hints on the behaviour of gases and
vapours.

Critical temperatures (1861)

The first interesting theoretical aspect is that in the p-V (or equivalently p-vm) diagram the “real”
gas isotherms are no more hyperbolas, but they can have minima and maxima as can be seen by
simply deriving the van der Waals equation with respect to molar specific volume and imposing
null derivatives, obtaining.

∂p
=−
RT
2
b
=0 ∂2 p RT b
and =2 −6 4 =0 .
∂v m v m−a 2 3
vm 2
∂ vm v m−a 3
vm
But it is much more useful to go a step further in the theoretical approach and consider the two
null derivatives together with the van der Waals equation ad a closed system of equations that
can be solved. In particular what is usually done is to take the Van der Waals equation in
conjunction with its null-imposed derivatives to obtain a closed system of three equations in the
three variables p, vm and T. The solutions of this system are

b
p= 2
≡ p c , T = 8b ≡T c and v =3 a≡v c .
27 a 27 R a

These solutions are called the “critical values” of the system for reasons that will appear clear
(hopefully) later.

These three pressure, temperature and specific molar value can be used to express the free
variables of the van der Waals equation in their “natural” form, that is with their typical units.
This is done simply retrieving the parameters a, b and R (remember that a and b different for
different gases and vapours) as function of the pressure, volume and temperature critical values,
that is

3 pc 1 27
R= , a= v c and b= p .
8 vc T c 3 64 c

Substituting the above values into the van der Waals equation we obtain a much simpler and
general form of the equation of state, which is called law of corresponding states, that is

T
8
p T c 3
= − 2
pc v
3 −1  v 
vc vc

always written in the more elegant form

8T r 3
pr= − 2
3 v r −1 v r

This equation can represent, in principle, the relationship between pressure, temperature and
molar specific volume for all the substances when measured in units of specific pressure,
temperature and molar specific volume. Just to have a look to the specific variables for real
substances you can use the table below.

Substance Tc (°C) pc (hPa)

N2 -122 34037

O2 -119 50346

H2O 374 220834

CO2 31 733000

Of course when Tc and pc are fixed, its molar specific value (density) is fixed as well by the
simple relationship
pc 3 R 3
= = R ρ
T c 8 v c 8 gas c

What can be noticed from the above table is that critical pressures have high values, far beyond
the classical atmospheric values (1013 hPa). On the contrary, critical temperatures show a
different behaviour for carbon dioxide and water and for oxygen. But what is the meaning of the
critical temperature? It can be shown by quite simple algebra (compute the derivative of the law
of corresponding states) the mathematical constraint to admit minima and maxima is that of
T r 1 , that is T T c . This means that for typical atmospheric conditions only water and carbon
dioxide may have minima and maxima. Just to have a naive description of what said, above take
a look to the figure below. But what does it mean for the van der Waals equation (i.e., the law of
corresponding states) the possibility to have a minimum or maximum? This means that between
the minimum and maximum pressures decreases with molar specific volume (i.e., with
increasing density). This situation is completely new when compared for the behaviour of ideal
gases, for which pressure increases with the decreasing of volume. This behaviour is essentially
unstable, in fact in this range of the PV diagram small compressions (expansions) of a system
initially in equilibrium with the environment grow. This non linear behaviour stands for an
extremely important process that takes place in atmosphere for some substances that is the
change of phase.

20
17.5 Tr = 0.6
Tr = 0.8
15
Tr = 1.0
12.5 Tr = 1.2
10 Tr = 2.0

7.5
pr

5
2.5
0
-2.5
-5
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 3
vr

Real isotherms (1873)

The behaviour of real isotherms is quite different for molar specific volumes (densities) in the
unstable range, in particular what empirically observed is that for decreasing molar specific
volume (i.e., increasing of density) “real” isotherms reach a point where pressure does not
increase any more and stays constant up to a specific volume (density) where pressures starts to
(quickly) grow again. Even this behaviour does not springs out naturally from the van der Waals
equation it is possible to use its form at least to find the point in which reality starts to differ from
theory. This was done by James Maxwell a few years after the publication of the van der Waals
work with semi-heuristic considerations. In fact, if pressure remains constant after a certain point
(during phase transition) and if the van der Waals equation gives a real description away from
phase transition, we can use it to find at least the value of the pressure where phase transition
takes place. In particular we are pretty sure that the total energy should remain constant (first law
of thermodynamics), then for a cyclic process (see figure below) that moves from the point A
(beginning of phase transition) to the point E (ending of phase transition) along a real isotherm
and comes back to the initial point A along a van der Waals isotherm, we should have

∮ dq=∮ du∮ p dv .

Figure. An ideal van der Waals isotherm (black line) and a real isotherm (red line). A
represents the beginning of phase transition, E the end of phase transition, the segment ACE
represents the real isotherm, the line ABCDE the van der Waals one.

Since the initial and final point are the same we are sure that

∮ du=0 and ∮ ds=0

where s is the system entropy (reversible process). Using the definition of entropy, remembering
that temperature is constant (isotherm), we are sure that

1 1
∮ ds=∮ T dq= T ∮ dq=0
then the first law of thermodynamics states that

∮ p dv=∫ABCA pdv∫CDEC pdv=0 then ∫ABCA pdv =−∫CDEC pdv

These simple considerations give us an easy rule to graphically evaluate (not easy task) or
numerically evaluate the value of pressure that makes equal the two integrals.

Negative pressures?

One interesting aspect that surely readers already noticed is the existence of negative pressures.
This is another interesting aspect of van der Waals equation. Far beyond to be a failure of the van
der Waals equation these are an important predictive aspect. To understand, or at least to create
an habit in our minds to this concept, we can think to negative pressures as “tension” or “tensile“
pressures instead of “compression” pressures. This can be easily understood remembering
recalling the stress tensor, which needs, to be defined, a surface and a force. If forces and
surfaces have opposite verses pressure is negative. Experimentally it is possible to obtain
extremely low (high in absolute value) negative pressures up to 300 000 hPa for water.
Significantly low levels of negative pressure are currently occurring in trees. In fact, if you just
try to compute the hydrostatic pressure at the rots of the trees exerted by the vertical amount of
water that goes from the bottom up to the top of the leaves, you will find unsuspectedly high
values. In spite of these high pressure values water reaches even the upper part of the tallest
sequoia. This is possible thanks to the negative pressure. Negative pressures can be reached, as
well foreseen by the van der Waals equation, only during phase transition, then negative
pressures are unstable. Nevertheless this unstable situation is maintained by trees through
transpiration. In other words they transpire to maintain a net flux from the roots up to the leaves,
bringing nutrients with the fluid as well. This is the reason why plants do not need an earth to
survive.

Do we really need van der Waals?

The answer to this question can be given only fixing correctly the problem. We can compare
ideal gas law with van der Waals law only fixing two of the tree variables. So far we choosed to
fix temperature and pressure to determine what happens to density (molar specific volume).

Above critical temperature we do not care about the difference between ideal and van der Waals
aeriforms.

At the critical temperature we already found that

pc 3 R 3
= = R ρ
T c 8 v c 8 gas c

Since we know that for ideal gases

p
=R gas ρc
T

we can easily obtain the following relationship, remembering that pressure and temperature are
the same.

ρc p c p T c 8 8
= = =
ρ T c T pc 3 3

Then at critical pressure and temperature van der Waals density is higher than ideal one. Under
the naïve point of view we can say that van der Waals gas needs an higher density (higher
frequency of bumps) to counteract the intra molecular attraction to reach the same pressure and
temperature.

For lower temperatures we can use the already obtained relationships

1 27
a= v c and b= p .
3 64 c

into the van der Waals equation, that is

 [ ]
R T 9 v
p= − Tc c
vm v 8 vm
1− c
3v m

if we are far from the critical point (case already studied) we can consider v m≫ v c and
collecting T we obtain

p~
RT
vm
1 c −
[ 
v 1 9 Tc
vm 3 8 T ]
substituting specific molar volume with density ρ=M /v m where M is the molar weight we
obtain

p~R gas T ρ 1
[  ρ 1 9 Tc
−
ρc 3 8 T ]
then

p 1
Rgas T ρ
~ 1
[ 
ρ 1 9 Tc
−
ρc 3 8 T ]
that is

ρideal
ρ
~ 1
[ 
ρ 1 9 Tc
−
ρc 3 8 T ]
If, with a further approximation, we substitute densities with pressures (e.g., p ~ ρ for both ideal
and van der Waals) we obtain

ρideal
ρ
~ 1
[ 
p 1 9 Tc
−
pc 3 8 T ]
This equation gives the van der Waals density departures from ideal gas density and, using the
above critical values (see the table) you can decide if for the peculiar range of temperatures and
pressure you are in an ideal or van der Waals situation.

Choosing the standard atmospheric partial pressures and 0 Celsius as a referring temperature you
will obtain the following values

0 ºC C02 N2 O2 H2O
ρ_vdwaals/ρ_ideal 1.0011 1.0009 1.0013 1.0001

That is, even if van der Waals gases and vapours are slightly denser than the ideal case the
mistake committed is of the order of the 0.001 or lower.
References and notes

If we do not pay sufficient attention to the problem of physical reality , we can get traped in
paradoxes
(D. Ruelle – Chance and Chaos)

Atmospheric Thermodynamics. C. Bohren and B. Albrecht. Oxford Univ. Press. 1998

A hort Course in cloud Physics. R. R. Rogers and M. K. Yau. Butterworth heinemann. 1996