Helmholtz free energy

Are Mjaavatten
Mjaavatten Consulting, Oslo, Norway
[email protected]

November 2017

Introduction

This document describes the calculation of the molar Helmholtz fee energy f for a pure substance
and how to find thermodynamic properties from the partial derivatives of f. The description is
restricted to pure substances in gas form. Mixtures and phase behaviour are not included. This
forms the theoretical basis for my thermo Matlab class.

Pressure equation (Equation of state)

Ideal gas (ig)

𝑅𝑇
𝑝𝑖𝑔 = (1)
𝑣

𝜕𝑝𝑖𝑔 𝑅
( ) = (2)
𝜕𝑇 𝑣 𝑣

𝜕𝑝𝑖𝑔 𝑅𝑇 𝑝
( ) =− 2 = (3)
𝜕𝑣 𝑇 𝑣 𝑣

General cubic EOS

Many equations of state for real gases have the form:

𝑅𝑇 𝐹(𝑇)
𝑝𝑒𝑜𝑠 = − (4)
𝑣 − 𝑏 𝑣(𝑣 + 𝑏)

𝜕𝑝𝑒𝑜𝑠 𝑅 𝐹′(𝑇)
( ) = − (5)
𝜕𝑇 𝑣 𝑣 − 𝑏 𝑣(𝑣 + 𝑏)

𝜕𝑝𝑒𝑜𝑠 𝑅𝑇 2𝑣 + 𝑏
( ) =− + 𝐹(𝑇) (6)
𝜕𝑣 𝑇 (𝑣 − 𝑏)2 𝑣 2 (𝑣 + 𝑏)2

𝜕 2 𝑝𝑒𝑜𝑠 𝐹′′(𝑇)
( 2 ) =− (7)
𝜕𝑇 𝑣
𝑣(𝑣 + 𝑏)

𝜕 2 𝑝𝑒𝑜𝑠 𝑅 2𝑣 + 𝑏
( ) =− 2 + 2 𝐹′(𝑇) (8)
𝜕𝑇𝜕𝑣 𝑣 (𝑣 − 𝑏) 𝑣 (𝑣 + 𝑏)2

1
 𝜕 2 𝑝𝑒𝑜𝑠 2𝑅𝑇 𝑣(𝑣 + 𝑏) − (2𝑣 + 𝑏)2
( ) = + 2 𝐹(𝑇) (9)
𝜕𝑣 2 𝑣 (𝑣 − 𝑏)3 𝑣 3 (𝑣 + 𝑏)3

Redlich-Kwong (RK)
1
𝐹𝑅𝐾 (𝑇) = 𝑎𝑇 −2 ( 10 )

1 3
𝐹′𝑅𝐾 (𝑇) = − 𝑎𝑇 −2 ( 11 )
2

3 5
𝐹′′𝑅𝐾 (𝑇) = − 𝑎𝑇 −2 ( 12 )
4

So:
𝑅𝑇 𝑎 1
𝑝𝑅𝐾 = − 𝑇 −2 ( 13 )
𝑣 − 𝑏 𝑣(𝑣 + 𝑏)

Ref.: Reid, Prausnitz and Sherwood: The properties of gases and liquids, McGraw Hill, 1977

𝜕𝑝𝑅𝐾 𝑅 𝑎 3 1
( ) = + 𝑇 −2 ( 14 )
𝜕𝑇 𝑣 𝑣 − 𝑏 2 𝑣(𝑣 + 𝑏)

𝜕𝑝𝑅𝐾 𝑅𝑇 2𝑣 − 𝑏 1
−2
( ) =− −𝑎 2 𝑇 ( 15 )
𝜕𝑣 𝑇 (𝑣 − 𝑏) 𝑣 (𝑣 + 𝑏)2

Soave-Redlich-Kwong (SRK)
The original form of the SRK equations uses:

2
𝐹𝑆𝑅𝐾 (𝑇) = [1 + 𝑔(𝜔) (1 − √𝑇⁄𝑇𝑐 )] ( 16 )

𝑔(𝜔) is a given function of the acentric factor 𝜔. Tables in e.g. Reid, Prausnitz and Sherwood (1977)
specify 𝜔 for a number of substances. For the purpose of fitting parameters, we may write ( 16 ) as:

2
𝐹𝑆𝑅𝐾 (𝑇) = (𝑐 − 𝑑√𝑇) ( 17 )
It follows that:
𝑑 1
𝐹 ′ (𝑇) ≡ 𝐹𝑆𝑅𝐾 (𝑇) = (𝑑 − 𝑐𝑇 −2 )𝑑 ( 18 )
𝑑𝑇

𝑑2 𝑐𝑑 −3
𝐹 ′′ (𝑇) ≡ 2 𝐹𝑆𝑅𝐾 (𝑇) = 𝑇 2 ( 19 )
𝑑𝑇 2

The standard values of c and d are:

𝑐 = √𝑎(1 + 𝑔(𝜔))
𝑎
𝑑 = √ 𝑔(𝜔)
𝑇𝑐

Inserting this ( 16 ) and ( 17 )in ( 4 ), ( 5 ), and ( 6 ) we get

2
 2
𝑅𝑇 (𝑐 − 𝑑√𝑇)
𝑝𝑆𝑅𝐾 = − ( 20 )
𝑣−𝑏 𝑣(𝑣 + 𝑏)

𝑐
−𝑑
𝜕𝑝𝑆𝑅𝐾 𝑅
( ) = − √𝑇 𝑑 ( 21 )
𝜕𝑇 𝑣 𝑣 − 𝑏 𝑣(𝑣 + 𝑏)

𝜕𝑝𝑆𝑅𝐾 𝑅𝑇 2 2𝑣 + 𝑏
( ) =− + (𝑐 − 𝑑√𝑇) 2 ( 22 )
𝜕𝑣 𝑇 (𝑣 − 𝑏)2 𝑣 (𝑣 + 𝑏)2

Internal energy

The molar internal energy u as a function of molar entropy s and molar volume v is a fundamental
thermodynamic relation. This means the function 𝑢(𝑠, 𝑣), contains all thermodynamic information
about the fluid in question.

The differential of u is
𝜕𝑢 𝜕𝑢
𝑑𝑢 = ( ) 𝑑𝑠 + ( ) 𝑑𝑣 ( 23 )
𝜕𝑠 𝑣 𝜕𝑣 𝑠

Temperature and pressure are given by:

𝜕𝑢
𝑇= ( ) ( 24 )
𝜕𝑠 𝑣

𝜕𝑢
𝑝 = −( ) ( 25 )
𝜕𝑣 𝑠
Thus:
𝑑𝑢 = 𝑇𝑑𝑠 − 𝑝𝑑𝑣 ( 26 )

We want to use T and v as independent variables. Thus:

𝜕𝑠 𝜕𝑠
𝑑𝑢 = 𝑇𝑑𝑠 − 𝑝𝑑𝑣 = 𝑇 ( ) 𝑑𝑇 + [𝑇 ( ) − 𝑝] 𝑑𝑣 ( 27 )
𝜕𝑇 𝑣 𝜕𝑣 𝑇

We use the Maxwell relation

𝜕𝑠 𝜕𝑝
( ) =( ) ( 28 )
𝜕𝑣 𝑇 𝜕𝑇 𝑣
to get:
𝜕𝑠 𝜕𝑝
𝑑𝑢 = 𝑇 ( ) 𝑑𝑇 + [𝑇 ( ) − 𝑝] 𝑑𝑣 ( 29 )
𝜕𝑇 𝑣 𝜕𝑇 𝑣
We also have that

𝜕𝑢 𝜕𝑢
𝑑𝑢 = ( ) 𝑑𝑇 + ( ) 𝑑𝑣 ( 30 )
𝜕𝑇 𝑣 𝜕𝑣 𝑇

The heat capacity at constant volume is defined as:

𝜕𝑢
𝑐𝑣 ≡ ( ) ( 31 )
𝜕𝑇 𝑣

3
So:
𝜕𝑝
𝑑𝑢 = 𝑐𝑣 𝑑𝑇 + [𝑇 ( ) − 𝑝] 𝑑𝑣 ( 32 )
𝜕𝑇 𝑣

If subscript 0 denotes some reference state:

𝑇 𝑣
𝜕𝑝
𝑢(𝑇, 𝑣) = 𝑢(𝑇0 , 𝑣0 ) + ∫ 𝑐𝑣 (𝑇, 𝑣0 )𝑑𝑇 + ∫ [𝑇 ( ) − 𝑝] 𝑑𝑣 ( 33 )
𝜕𝑇 𝑣
𝑇0 𝑣0

Ideal gas
From ( 1 ):
𝜕𝑝𝑖𝑔 𝑅
( ) = ( 34 )
𝜕𝑇 𝑣 𝑣
And:
𝜕𝑝𝑖𝑔 𝑅𝑇 𝑅𝑇
𝑇( ) − 𝑝𝑖𝑔 = − =0 ( 35 )
𝜕𝑇 𝑣 𝑣 𝑣
Inserted in ( 33 ) this gives:
𝑇 𝑣 𝑇
𝑅𝑇
𝑢𝑖𝑔 (𝑇, 𝑣) = 𝑢𝑖𝑔 (𝑇0 , 𝑣0 ) + ∫ 𝑐𝑣,𝑖𝑔 𝑑𝑇 + ∫ [ − 𝑝𝑖𝑔 ] 𝑑𝑣 = 𝑢0 + ∫ 𝑐𝑣,𝑖𝑔 𝑑𝑇 ( 36 )
𝑣
𝑇0 𝑣0 𝑇0

Note that 𝑢𝑖𝑔 is independent of v, so we can write:

𝑇

𝑢𝑖𝑔 (𝑇, 𝑣) = 𝑢𝑖𝑔 (𝑇) = 𝑢𝑖𝑔 (𝑇0 ) + ∫ 𝑐𝑣,𝑖𝑔 𝑑𝑇 ( 37 )

𝑇0
Cubic EOS
From ( 4 ) and ( 5 ) we get:
𝜕𝑝𝑐𝑢𝑏𝑖𝑐 𝐹(𝑇) − 𝑇𝐹 ′ (𝑇)
𝑇( ) − 𝑝𝑐𝑢𝑏𝑖𝑐 = ( 38 )
𝜕𝑇 𝑣 𝑣(𝑣 + 𝑏)

When 𝑣 → ∞ we have that 𝑝𝑐𝑢𝑏𝑖𝑐 (𝑇, 𝑣) → 𝑝𝑖𝑔 (𝑇, 𝑣), so the cubic model is equivalent to the ig model
in this limit. Stretching the mathematical stringency a little we may write:

𝑝𝑐𝑢𝑏𝑖𝑐 (𝑇0 , ∞) = 𝑝𝑖𝑔 (𝑇0 , ∞)

and
𝑢𝑐𝑢𝑏𝑖𝑐 (𝑇0 , ∞) = 𝑢𝑖𝑔 (𝑇0 , ∞) = 𝑢𝑖𝑔 (𝑇0 )
Integrating from ∞ to 𝑣0 :
𝑇 𝑣
𝜕𝑝𝑐𝑢𝑏𝑖𝑐
𝑢𝑐𝑢𝑏𝑖𝑐 (𝑇, 𝑣) = 𝑢𝑖𝑔 (𝑇0 ) + ∫ 𝑐𝑣 𝑑𝑇 + ∫ [𝑇 ( ) − 𝑝𝑐𝑢𝑏𝑖𝑐 ] 𝑑𝑣 ( 39 )
𝜕𝑇 𝑣
𝑇0 ∞
Using ( 38 ):
𝑣0 𝑣0
𝜕𝑝𝑐𝑢𝑏𝑖𝑐 𝐹(𝑇) − 𝑇𝐹 ′ (𝑇) 𝐹(𝑇) − 𝑇𝐹 ′ (𝑇) 𝑣
∫ [𝑇 ( ) − 𝑝𝑐𝑢𝑏𝑖𝑐 ] 𝑑𝑣 = ∫ 𝑑𝑣 = ln ( )
𝜕𝑇 𝑣 𝑣(𝑣 + 𝑏) 𝑏 𝑣+𝑏
∞ ∞

and finally:

4
 𝐹(𝑇) − 𝑇𝐹 ′ (𝑇) 𝑣
𝑢𝑐𝑢𝑏𝑖𝑐 (𝑇, 𝑣) = 𝑢𝑖𝑔 (𝑇) + ln ( ) ( 40 )
𝑏 𝑣+𝑏
′
𝑇𝐹 ′ (𝑇) 𝑣
𝑐𝑣 (𝑇, 𝑣) ≡ 𝑢 𝑇 = 𝑢 𝑇,𝑖𝑔 (𝑇) − ln ( ) ( 41 )
𝑏 𝑣+𝑏

𝐹(𝑇) − 𝑇𝐹 ′ (𝑇)
𝑢𝑣 = − ( 42 )
𝑣(𝑣 + 𝑏)

Entropy
From ( 26 ) and ( 29 ):

1 𝑝 𝑐𝑣 𝜕𝑝 𝑝 𝑝 𝑐𝑣 𝜕𝑝
𝑑𝑠 = 𝑑𝑢 + 𝑑𝑣 = 𝑑𝑇 + [( ) − + ] 𝑑𝑣 = 𝑑𝑇 + ( ) 𝑑𝑣 ( 43 )
𝑇 𝑇 𝑇 𝜕𝑇 𝑣 𝑇 𝑇 𝑇 𝜕𝑇 𝑣
So we have:
𝜕𝑠 𝑐𝑣
( ) = ( 44 )
𝜕𝑇 𝑣 𝑇

𝜕𝑠 𝜕𝑝
( ) =( ) ( 45 )
𝜕𝑣 𝑇 𝜕𝑇 𝑣
Integrating from (𝑇0 , 𝑣0 ) to (𝑇, 𝑣):

𝑇 𝑣
𝑐𝑣 𝜕𝑝
𝑠 = 𝑠(𝑇0 , 𝑣0 ) + ∫ 𝑑𝑇 + ∫ ( ) 𝑑𝑣 ( 46 )
𝑇 𝜕𝑇 𝑣
𝑇0 𝑣0

Ideal gas
𝑇 𝑣
𝑐𝑣 𝑅
𝑠𝑖𝑔 (𝑇, 𝑣) = 𝑠𝑖𝑔 (𝑇0 , 𝑣0 ) + ∫ 𝑑𝑇 + ∫ 𝑑𝑣
𝑇 𝑣
𝑇0 𝑣0
𝑇 ( 47 )
𝑐𝑣 𝑣
= 𝑠𝑖𝑔 (𝑇0 , 𝑣0 ) + ∫ 𝑑𝑇 + 𝑅 ln
𝑇 𝑣0
𝑇0

𝜕𝑠𝑖𝑔 𝑐𝑣,𝑖𝑔
( ) = ( 48 )
𝜕𝑇 𝑣 𝑇

𝜕𝑠𝑖𝑔 𝑅
( ) =− ( 49 )
𝜕𝑣 𝑇 𝑣
We shall also need:
𝜕 2 𝑠𝑖𝑔 1 𝜕𝑐𝑣,𝑖𝑔 𝑐𝑣,𝑖𝑔
( 2 ) = ( ) − 2 ( 50 )
𝜕𝑇 𝑣 𝑇 𝜕𝑇 𝑣 𝑇

If cv is constant:
𝑇 𝑣
𝑠𝑖𝑔 (𝑇, 𝑣) = 𝑠𝑖𝑔 (𝑇0 , 𝑣0 ) + 𝑐𝑣 ln + 𝑅 ln ( 51 )
𝑇0 𝑣0
If
𝑐𝑣,𝑖𝑔 (𝑇) = 𝑐0 − 𝑐1 (𝑇 − 𝑇0 ) + 𝑐2 (𝑇 − 𝑇0 )2 − 𝑐3 (𝑇 − 𝑇0 )3 ( 52 )

5
then
𝑇
𝑐𝑣 𝑇
∫ 𝑑𝑇 = [𝑐0 − 𝑐1 𝑇0 − 𝑐2 𝑇02 − 𝑐3 𝑇03 ] ln + [𝑐1 − 2𝑐2 𝑇0 + 3𝑐3 𝑇02 ](𝑇 − 𝑇0 )
𝑇 𝑇0
𝑇0 ( 53 )
𝑇2 − 𝑇02 𝑇3 − 𝑇03
+ [𝑐2 − 3𝑐3 𝑇0 ] + 𝑐3
2 3
Cubic EOS

As for internal energy, we have that

𝑠𝑒𝑜𝑠 (𝑇0 , ∞) = 𝑠𝑖𝑔 (𝑇0 , ∞) ( 54 )
So:
𝑣
𝜕(𝑝𝑒𝑜𝑠 − 𝑝𝑖𝑔 )
𝑠𝑒𝑜𝑠 (𝑇, 𝑣) − 𝑠𝑖𝑔 (𝑇, 𝑣) = ∫ ( ) 𝑑𝑣 ( 55 )
𝜕𝑇 𝑣
∞

𝜕(𝑝𝑒𝑜𝑠 − 𝑝𝑖𝑔 ) 𝑅 𝐹 ′(𝑇) 𝑅 𝑅𝑏 𝐹 ′ (𝑇)

= − − = −
𝜕𝑇 𝑣 − 𝑏 𝑣(𝑣 + 𝑏) 𝑣 𝑣(𝑣 − 𝑏) 𝑣(𝑣 + 𝑏)
𝑣
𝜕𝑝𝑒𝑜𝑠 − 𝑝𝑖𝑔 𝑣 − 𝑏 𝐹 ′ (𝑇) 𝑣 + 𝑏
∫( ) 𝑑𝑣 = 𝑅 ln + ln ( 56 )
𝜕𝑇 𝑣
𝑣 𝑏 𝑣
∞
Inserting in ( 46 ):
𝑣 − 𝑏 𝐹 ′ (𝑇) 𝑣 + 𝑏
𝑠𝑒𝑜𝑠 (𝑇, 𝑣) = 𝑠𝑖𝑔 (𝑇, 𝑣) + 𝑅 ln + ln ( 57 )
𝑣 𝑏 𝑣

𝜕𝑠𝑒𝑜𝑠 𝜕𝑠𝑖𝑔 𝐹′′ (𝑇) 𝑣 + 𝑏 𝑐𝑣,𝑒𝑜𝑠

= + ln = ( 58 )
𝜕𝑇 𝜕𝑇 𝑏 𝑣 𝑇

𝜕𝑠𝑒𝑜𝑠 𝜕𝑠𝑖𝑔 𝑅𝑏 𝐹 ′ (𝑇)

= + − ( 59 )
𝜕𝑣 𝜕𝑣 𝑣(𝑣 − 𝑏) 𝑣(𝑣 + 𝑏)

𝜕 2 𝑠𝑒𝑜𝑠 𝜕 2 𝑠𝑖𝑔 𝐹′′′ (𝑇) 𝑣 + 𝑏

= + ln ( 60 )
𝜕𝑇 2 𝜕𝑇 2 𝑏 𝑣

Helmholtz free energy

Just as the function 𝑢(𝑠, 𝑣), the function 𝑓(𝑇, 𝑣) for Helmholtz free energy is a fundamental
thermodynamic relation. If we know 𝑓(𝑇, 𝑣) we can calculate all thermodynamic properties.

𝑓(𝑇, 𝑣) = 𝑢(𝑇, 𝑣) − 𝑇𝑠(𝑇, 𝑣) ( 61 )

𝜕𝑓
𝑓𝑇 ≡ ( ) = −𝑠 ( 62 )
𝜕𝑇 𝑣

𝜕𝑓
𝑓𝑣 ≡ ( ) = −𝑝 ( 63 )
𝜕𝑣 𝑇

The higher derivatives 𝑓𝑇𝑇 , 𝑓𝑇𝑣 , 𝑓𝑣𝑣 , 𝑓𝑇𝑇𝑇 , 𝑓𝑇𝑇𝑣 , 𝑓𝑇𝑣𝑣 , and 𝑓𝑣𝑣𝑣 can be found from the expressions for
the derivatives of s and p.

6
Enthalpy
The enthalpy is given by:

ℎ = 𝑢 + 𝑝𝑣 = 𝑓 + 𝑝𝑣 + 𝑇𝑠 ( 64 )

𝜕ℎ 𝜕𝑓 𝜕𝑝 𝜕𝑠
( ) = ( ) + ( ) 𝑣 + 𝑠 + 𝑇 ( ) = −𝑓𝑇𝑣 𝑣 − 𝑓𝑇𝑇 𝑇 ( 65 )
𝜕𝑇 𝑣 𝜕𝑇 𝑣 𝜕𝑇 𝑣 𝜕𝑇 𝑣

𝜕ℎ 𝜕𝑓 𝜕𝑝 𝜕𝑠
( ) = ( ) + ( ) 𝑣 + 𝑝 + 𝑇 ( ) = −𝑓𝑣𝑣 𝑣 − 𝑓𝑇𝑣 𝑇 ( 66 )
𝜕𝑣 𝑇 𝜕𝑣 𝑇 𝜕𝑣 𝑇 𝜕𝑣 𝑇

Thermodynamic properties from Helmholtz free energy

Fundamental relationship:

𝑑𝑓 = −𝑠𝑑𝑇 − 𝑝𝑑𝑣 ( 67 )
So

𝜕𝑓
𝑠=− ( 68 )
𝜕𝑇

𝜕𝑓
𝑝=− ( 69 )
𝜕𝑣

Heat capacities
𝑑𝑄 𝜕𝑠 𝜕2𝑓
𝑐𝑣 ≡ ( ) = 𝑇 ( ) = −𝑇 ( 2 ) ( 70 )
𝑑𝑇 𝑣 𝜕𝑇 𝑣 𝜕𝑇 𝑣

𝑑𝑄 𝜕𝑠
𝑐𝑝 ≡ ( ) = 𝑇 ( ) ( 71 )
𝑑𝑇 𝑝 𝜕𝑇 𝑝
From
𝜕𝑠 𝜕𝑠 𝜕𝑠 𝜕𝑠 𝜕𝑣 𝜕𝑣
𝑑𝑠 = ( ) 𝑑𝑇 + ( ) 𝑑𝑣 = ( ) 𝑑𝑇 + ( ) [( ) 𝑑𝑇 + ( ) 𝑑𝑝] ( 72 )
𝜕𝑇 𝑣 𝜕𝑣 𝑇 𝜕𝑇 𝑣 𝜕𝑣 𝑇 𝜕𝑇 𝑝 𝜕𝑝 𝑇
we get
𝜕𝑠 𝜕𝑠 𝜕𝑠 𝜕𝑣
( ) = ( ) +( ) ( ) ( 73 )
𝜕𝑇 𝑝 𝜕𝑇 𝑣 𝜕𝑣 𝑇 𝜕𝑇 𝑝
From
𝜕𝑝 𝜕𝑝
𝑑𝑝 = ( ) 𝑑𝑇 + ( ) 𝑑𝑣 ( 74 )
𝜕𝑇 𝑣 𝜕𝑣 𝑇
we get
𝜕𝑝
( )
𝜕𝑣 𝜕𝑇 𝑣
( ) =− ( 75 )
𝜕𝑇 𝑝 𝜕𝑝
( )
𝜕𝑣 𝑇
So:

7
 𝜕𝑝
( )
𝜕𝑠 𝜕𝑠 𝜕𝑠 𝜕𝑇 𝑣
𝑐𝑝 = 𝑇 ( ) = 𝑇 ( ) − 𝑇 ( ) ( 76 )
𝜕𝑇 𝑝 𝜕𝑇 𝑣 𝜕𝑝
𝜕𝑣 𝑇 ( )
𝜕𝑣 𝑇

𝜕 2𝑓 2
𝜕 2𝑓
𝜕𝑣𝜕𝑇 = 𝑇 𝑓𝑇𝑣
𝑐𝑝 − 𝑐𝑣 = 𝑇 ( 77 )
𝜕𝑇𝜕𝑣 𝜕 2 𝑓 𝑓𝑣𝑣
𝜕𝑣 2

For ideal gas we get:

𝜕𝑝 2 𝑅2
2 ( ) 𝑇
𝑓𝑇𝑣 𝜕𝑇 𝑣 2
𝑐𝑝 − 𝑐𝑣 = 𝑇 = −𝑇 = 𝑝𝑣 = 𝑅 ( 78 )
𝑓𝑣𝑣 𝜕𝑝
( ) 𝑣
𝜕𝑣 𝑇
as expected.

Speed of sound
𝜕𝑝
𝑐2 = ( ) ( 79 )
𝜕𝜌 𝑠

𝑀𝑤
𝜌= ( 80 )
𝑣

𝜕𝑝 𝜕𝑝 𝑑𝑣 𝑣 2 𝜕𝑝
𝑐2 = ( ) = ( ) =− ( ) ( 81 )
𝜕𝜌 𝑠 𝜕𝑣 𝑠 𝑑𝜌 𝑀𝑤 𝜕𝑣 𝑠
Now, expressing differentials in (T, v) and (S,v) coordinates:

𝜕𝑝 𝜕𝑝
𝑑𝑝 = ( ) 𝑑𝑣 + ( ) 𝑑𝑇
𝜕𝑣 𝑇 𝜕𝑇 𝑣

𝜕𝑇 𝜕𝑇
𝑑𝑇 = ( ) 𝑑𝑣 + ( ) 𝑑𝑠
𝜕𝑣 𝑠 𝜕𝑠 𝑣

𝜕𝑝 𝜕𝑝 𝜕𝑇 𝜕𝑇
𝑑𝑝 = ( ) 𝑑𝑣 + ( ) {( ) 𝑑𝑣 + ( ) 𝑑𝑠}
𝜕𝑣 𝑇 𝜕𝑇 𝑣 𝜕𝑣 𝑠 𝜕𝑠 𝑣
Setting ds= 0:
𝜕𝑝 𝜕𝑝 𝜕𝑝 𝜕𝑇
( ) =( ) +( ) ( )
𝜕𝑣 𝑠 𝜕𝑣 𝑇 𝜕𝑇 𝑣 𝜕𝑣 𝑠
𝜕𝑇
Now consider the term ( ) . When s is constant:
𝜕𝑣 𝑠
𝜕𝑠 𝜕𝑠
𝑑𝑠 = ( ) 𝑑𝑣 + ( ) 𝑑𝑇 = 0
𝜕𝑣 𝑇 𝜕𝑇 𝑣
So:
𝜕𝑠
𝜕𝑇 ( )
𝜕𝑣 𝑇
( ) =−
𝜕𝑣 𝑠 𝜕𝑠
( )
𝜕𝑇 𝑣
So we get:

8
 𝜕𝑠
𝜕𝑝 𝜕𝑝 𝜕𝑝 (𝜕𝑣 ) 𝑇 2
𝑓𝑇𝑣
( ) =( ) −( ) = −𝑓𝑣𝑣 +
𝜕𝑣 𝑠 𝜕𝑣 𝑇 𝜕𝑇 𝑣 ( 𝜕𝑠 ) 𝑓𝑇𝑇
𝜕𝑇 𝑣

The speed of sound is thus given by:

2
𝑣 2 𝜕𝑝 𝑣2 𝑓𝑇𝑣
𝑐2 = − ( ) = (𝑓𝑣𝑣 − ) ( 82 )
𝑀 𝜕𝑣 𝑠 𝑀𝑤 𝑓𝑇𝑇

Inserting the expressions for ideal gas we get:

𝑅
𝜕𝑝 𝑝 𝑅 𝑣 𝑝 𝑅2𝑇 𝑅 𝑝𝑣 𝑅𝑇 𝑅𝑇 𝑅 𝑅𝑇 𝑐𝑣 + 𝑅
( ) =− − 𝑐 =− − = − ( + ) = − (1 + ) = −
𝜕𝑣 𝑠 𝑣 𝑣 𝑣 𝑣 𝑐𝑣 𝑣 2 𝑣2 𝑅 𝑐𝑣 𝑣2 𝑐𝑣 𝑣 2 𝑐𝑣
𝑇

𝜕𝑝 𝑅𝑇 𝑐𝑝 𝑅𝑇𝛾
( ) =− 2 =− 2
𝜕𝑣 𝑠 𝑣 𝑐𝑣 𝑣

𝛾𝑅𝑇
𝑐=√ ( 83 )
𝑀𝑤

Joule-Thompson coefficient
The Joule-Thompson coefficient is defined as:
𝜕𝑇
𝜇𝐽𝑇 = ( ) ( 84 )
𝜕𝑝 ℎ

𝜕ℎ 𝜕ℎ
𝑑ℎ = ( ) 𝑑𝑝 + ( ) 𝑑𝑇
𝜕𝑇 𝑝 𝜕𝑝 𝑇
For constant h:
𝜕ℎ 𝜕𝑝 𝜕ℎ
0=( ) ( ) +( )
𝜕𝑇 𝑝 𝜕𝑇 𝐻 𝜕𝑝 𝑇
So:
𝜕ℎ 𝜕ℎ
( ) ( )
𝜕𝑇 𝜕𝑝 𝑇 𝜕𝑝 𝑇
𝜇𝐽𝑇 =( ) =− =−
𝜕𝑝 𝐻 𝜕ℎ 𝑐𝑝
( )
𝜕𝑇 𝑝
We also have:
𝜕𝑠 𝜕𝑠
𝑑ℎ = 𝑇𝑑𝑠 + 𝑣𝑑𝑝 = 𝑇 ( ) 𝑑𝑇 + [𝑇 ( ) + 𝑣] 𝑑𝑝
𝜕𝑇 𝑝 𝜕𝑝 𝑇
So, if dT = 0:
𝜕ℎ 𝜕𝑠 𝜕𝑣
( ) = 𝑇 ( ) + 𝑣 = −𝑇 ( ) + 𝑣
𝜕𝑝 𝑇 𝜕𝑝 𝑇 𝜕𝑇 𝑝
We have used the Maxwell relation:
𝜕𝑠 𝜕𝑣
( ) = −( )
𝜕𝑝 𝑇 𝜕𝑇 𝑝

𝜕𝑝 𝜕𝑝
𝑑𝑝 = ( ) 𝑑𝑇 + ( ) 𝑑𝑣
𝜕𝑇 𝑣 𝜕𝑣 𝑇

9
 𝜕𝑝 𝜕𝑝 𝜕𝑣
0= ( ) +( ) ( )
𝜕𝑇 𝑣 𝜕𝑣 𝑇 𝜕𝑇 𝑝

𝜕𝑝 𝜕𝑝
−( ) −( )
𝜕𝑣 𝜕𝑇 𝑣 𝜕𝑇 𝑣
( ) = =
𝜕𝑇 𝑝 𝜕𝑝 𝜕𝑝
( ) ( )
𝜕𝑣 𝑇 𝜕𝑣 𝑇

𝜕𝑝
( )
𝜕𝑇 𝑣
−𝑇 −𝑣 𝜕𝑝
𝜕ℎ 𝜕𝑣 𝜕𝑝
( ) 𝑇 ( ) − 𝑣 ( ) ( )
𝜕𝑝 𝑇 𝜕𝑇 𝑝 𝜕𝑣 𝑇 1 𝜕𝑇 𝑣
𝜇𝐽𝑇 =− = = = − (𝑇 + 𝑣)
𝑐𝑝 𝑐𝑝 𝑐𝑝 𝑐𝑝 𝜕𝑝
( )
𝜕𝑣 𝑇

1 𝑓𝑇𝑣
𝜇𝐽𝑇 = − (𝑇 + 𝑣)
𝑐𝑝 𝑓𝑣𝑣 ( 85 )

For ideal gas:

𝑅𝑇
1
𝜇𝐽𝑇,𝑖𝑔 = − (𝑇 𝑣 + 𝑣) = 0
𝑐𝑝 𝑅𝑇
− 2
𝑣

Reference

My approach to thermodynamics is taken from the book “Thermodynamics and an introduction to

thermostatistics” by Herbert B. Callen (Wiley, 1985) which I highly recommend.

10