0% found this document useful (0 votes)
91 views100 pages

ENCH 633 - 2017 - Week 2 3

The document discusses key concepts in chemical thermodynamics including: 1) Phase equilibria of pure species and the criteria for phase equilibrium at constant temperature and pressure based on Gibbs free energy. 2) The roles of energy and entropy in determining whether phase changes are spontaneous based on whether ΔH is less than or greater than TΔS. 3) The Clausius-Clapeyron equation relating the change in vapor pressure of a pure component with temperature based on enthalpy of vaporization. 4) Concepts of partial molar properties which describe how intensive properties vary with changing composition in a mixture at constant temperature and pressure. 5) The Gibbs-Duhem

Uploaded by

raja.mt
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
91 views100 pages

ENCH 633 - 2017 - Week 2 3

The document discusses key concepts in chemical thermodynamics including: 1) Phase equilibria of pure species and the criteria for phase equilibrium at constant temperature and pressure based on Gibbs free energy. 2) The roles of energy and entropy in determining whether phase changes are spontaneous based on whether ΔH is less than or greater than TΔS. 3) The Clausius-Clapeyron equation relating the change in vapor pressure of a pure component with temperature based on enthalpy of vaporization. 4) Concepts of partial molar properties which describe how intensive properties vary with changing composition in a mixture at constant temperature and pressure. 5) The Gibbs-Duhem

Uploaded by

raja.mt
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 100

ENCH-633: Chemical

Thermodynamics
Sathish Ponnurangam
Fall 2018

Week – 2-3:
(Reading: Chap-2,3 Molecular Thermodynamics Of Fluid-Phase Equilibria by Prausnitz et. al.,
Additional reading: Chap-6,7 Engineering and Chemical thermodynamics by Koretsky)
Phase equilibria of pure species

• How liquid water and vapor


equilibrates?
Find criteria for constant T, P
• Which combination has Suniv
maximum
• 𝒅𝑮 ≤ 𝟎
• If a non-PV work is
involved?
• 𝒅𝑮 ≤ 𝜹𝒘′ 2
Phase equilibria of pure species
• Total Gibbs free energy for pure species in two phases (,)
𝜷 𝜷
• 𝑮𝒊 = 𝒏𝜶𝒊 𝒈𝜶𝒊 + 𝒏𝒊 𝒈𝒊 ,
𝜷
• where 𝒏𝜶𝒊 𝒂𝒏𝒅 𝒏𝒊 are number of moles of i in  and  phases
𝜷 𝜷 𝜷 𝜷
• 𝒅𝑮𝒊 = 𝒏𝜶𝒊 𝒅𝒈𝜶𝒊 + 𝒈𝜶𝒊 𝒅𝒏𝜶𝒊 + 𝒏𝒊 𝒅𝒈𝒊 + 𝒈𝒊 𝒅𝒏𝒊 ≤ 𝟎
𝜷 𝜷
• 𝒅𝑮𝒊 = 𝒈𝜶𝒊 𝒅𝒏𝜶𝒊 + 𝒈𝒊 𝒅𝒏𝒊 ≤ 𝟎, for a given T and P, g is constant
𝜷
• Since it is a closed system, 𝒅𝒏𝜶𝒊 = −𝒅𝒏𝒊
𝜷
• (𝒈𝜶𝒊 −𝒈𝒊 )𝒅𝒏𝜶𝒊 ≤ 𝟎
• Consider water -steam at (90 C , 1 atm) and (110 C, 1 atm)
• Condition for phase equilibrium
𝜷
• 𝒈𝜶𝒊 = 𝒈𝒊 ; molar Gibbs free energy should be equal
• not the concentration
• water-steam at T=100 C and 1 atm

3
Roles of energy and entropy in phase equilibria

• Since Suniv >0, why not all solid atoms


sublimate?
• To view phase equilibria from
energetic vs. entropic effects
• ∆𝑮 = ∆𝑯 − 𝑻∆𝑺 ≤ 𝟎 ?
• Solid  gas
• What happens to H and S
• As bonds are broken, H and S
• If ∆𝑯 < 𝑻∆𝑺; spontaneous process
• If ∆𝑯 > 𝑻∆𝑺; non-spontaneous process
• If ∆𝑯 = 𝑻∆𝑺; Phase equilibrium
• Our life depends on precise balance
between these two terms
• Protein folding and denaturation
4
Clapeyron equation
• If a pure species is in equilibrium in two of its phases,
• T and P are not independent variables?
• Why it is true physically?
• Since at phase equilibrium, P and T are same for both phases, when T
, P has to 

• Psat(T)
𝜷
• 𝒈𝜶𝒊 = 𝒈𝒊

5
Clapeyron equation – Psat(T)
𝑷𝜶 = 𝑷𝜷
𝜷
𝑻𝜶 = 𝑻𝜷 • 𝒇𝒐𝒓 𝒑𝒐𝒊𝒏𝒕 𝟏: 𝒈𝜶𝒊 = 𝒈𝒊
𝜷
𝒈𝜶𝒊 = 𝒈𝒊 𝜷
• 𝒇𝒐𝒓 𝒑𝒐𝒊𝒏𝒕 𝟐: dP, dT and 𝒅𝒈𝜶𝒊 , 𝒅𝒈𝒊
𝜷 𝜷
• 𝒈𝜶𝒊 + 𝒅𝒈𝜶𝒊 = 𝒈𝒊 + 𝒅𝒈𝒊
𝜷
• 𝒅𝒈𝜶𝒊 = 𝒅𝒈𝒊
𝜷 𝜷
• −𝒔𝜶𝒊 𝒅𝑻 + 𝒗𝜶𝒊 𝒅𝑷 = −𝒔𝒊 𝒅𝑻 + 𝒗𝒊 𝒅𝑷

𝜷
𝒅𝑷 𝒔𝜶
𝒊 −𝒔𝒊
• = 𝜷
𝒅𝑻 𝒗𝜶
𝒊 −𝒗𝒊

𝜷 𝜷
• 𝒉𝜶𝒊 − 𝑻𝒔𝜶𝒊 = 𝒉𝒊 − 𝑻𝒔𝒊
𝜷
𝒅𝑷𝒔𝒂𝒕 𝒉𝜶
𝒊 −𝒉𝒊
• = 𝜷
𝒅𝑻 𝑻 (𝒗𝜶
𝒊 −𝒗𝒊 )
6
Vapor-liquid equilibrium: Clausius–Clapeyron equation
𝒅𝑷𝒔𝒂𝒕 𝒉𝒗𝒊 −𝒉𝒍𝒊
• =
𝒅𝑻 𝑻 (𝒗𝒗𝒊 −𝒗𝒍𝒊 )

• How to transform so that we get only P vs. T?


• Assumption 1: molar volume of liquid negligible,
• 𝒗𝒍𝒊 ≪ 𝒗𝒗𝒊  𝒗𝒍𝒊 = 𝟎
𝒅𝑷𝒔𝒂𝒕 ∆𝒉𝒗𝒂𝒑,𝒊
• =
𝒅𝑻 𝑻 𝒗𝒗𝒊
𝑹𝑻
• Assumption 2: Ideal gas, 𝒗𝒗𝒊 =
𝑷𝒔𝒂𝒕

𝒅𝑷𝒔𝒂𝒕 ∆𝒉𝒗𝒂𝒑,𝒊 𝑷𝒔𝒂𝒕 ∆𝒉𝒗𝒂𝒑,𝒊


• = =
𝒅𝑻 𝑻 𝒗𝒗𝒊 𝑹𝑻𝟐
𝒅𝑷𝒔𝒂𝒕 ∆𝒉𝒗𝒂𝒑,𝒊 𝒅𝑷𝒔𝒂𝒕 ∆𝒉𝒗𝒂𝒑,𝒊
• = 𝒅𝑻; = 𝒅𝑻
𝑷𝒔𝒂𝒕 𝑹𝑻𝟐 𝑷𝒔𝒂𝒕 𝑹𝑻𝟐 7
Vapor-liquid equilibrium: Clausius–Clapeyron equation

𝒅𝑷𝒔𝒂𝒕 ∆𝒉𝒗𝒂𝒑,𝒊
• = 𝒅𝑻
𝑷𝒔𝒂𝒕 𝑹𝑻𝟐

• Assumption 3: hvap,i is independent of T


𝒅𝑷𝒔𝒂𝒕 𝒅𝑻
• = ∆𝒉𝒗𝒂𝒑,𝒊
𝑷𝒔𝒂𝒕 𝑹𝑻𝟐

𝒔𝒂𝒕 1
• ln 𝑷 = −∆𝒉𝒗𝒂𝒑,𝒊 +𝑪
𝑹𝑻
Or
𝑷𝒔𝒂𝒕 ∆𝒉𝒗𝒂𝒑,𝒊 1 1
• ln 𝟐𝒔𝒂𝒕 =− −
𝑷𝟏 𝑹 𝑻𝟐 𝑻𝟏 8
Vapor-liquid equilibrium: Clausius–Clapeyron equation
𝒔𝒂𝒕 1
• ln 𝑷 = −∆𝒉𝒗𝒂𝒑,𝒊 +𝑪
𝑹𝑻

• Assumptions cancel total error, 𝒅𝑷𝒔𝒂𝒕


𝒅𝑻
=
𝒉𝒗𝒊 −𝒉𝒍𝒊
𝑻 (𝒗𝒗𝒊 −𝒗𝒍𝒊 )
• Let us assume T
• Assumption 3: hvap ↓ (bonds in liquid become weaker)
• Assumption 1: molar volume of liquid negligible,
• Real 𝒗𝒍𝒊 
𝑅𝑇
• Assumption 2: Ideal gas, 𝑣𝑖𝑣 = 𝑃𝑠𝑎𝑡
• Real gas Psat < ideal gas Psat; real 𝒗𝒗𝒊 ↓
• Assumption 1 and 2 leads to denominator ↓
https://en.wikipedia.org/wiki/Enthalpy_of_vaporization

• Most accurate and commonly reported


functional form is Antoine equation
B
• ln 𝑷𝒔𝒂𝒕 = 𝑨 −
𝑪+𝑻
9
Thermodynamics of mixtures

• Any state function of a pure substance depends on molecular PE


and KE (any two state variable can fix this, e.g., P and T)
• Only i-i interactions are present
• Mixtures: i-i, j-j, and i-j interactions are important
• Depends on the nature of components and their quantity

Need to develop formal thermodynamic equations for mixtures 10


Partial molar properties
• For mixtures, we want to establish how state
properties vary as a function of number of moles
of each species along with T, P

11
Partial molar properties

• Defined always at constant T, P and nj≠i

• Partial molar property defined with number of


moles NOT mole fraction
12
Partial molar properties

• Partial molar property

13
Partial molar properties

Partial molar properties: contribution of given species to


the properties of a mixture, at a constant T,P and nj≠i

14
• Determine C by
• physical argument or
• Euler’s theorem on homogenous functions

• Extensive property K depends on total number of


moles
• K=nTk
• As nT  0, K=0  as ni  0, K=0  C=0

15
• Extensive total solution property, K is sum of partial molar property
adjusted by its moles
• Intensive property, k, is weighted average of partial molar property

16
𝑲𝒊 − 𝒌𝒊
17
Gibbs-Duhem equation
• Most useful equation relating partial molar
properties of different species

18
Gibbs-Duhem equation

19
Summary of nomenclature

20
Summary of nomenclature

21
Property changes of mixing
• Δ𝐾𝑚𝑖𝑥 = 𝐾 − ∑𝑛𝑖 𝑘𝑖
• e.g., Δ𝑉𝑚𝑖𝑥 = 𝑉 − ∑𝑛𝑖 𝑣𝑖 , Δ𝐻𝑚𝑖𝑥 = 𝐻 − ∑𝑛𝑖 ℎ𝑖
• Δ𝐾𝑚𝑖𝑥 = ∑𝑛𝑖 𝐾𝑖 − ∑𝑛𝑖 𝑘𝑖 = ∑𝑛𝑖 (𝐾𝑖 − 𝑘𝑖 )
• Δ𝑉𝑚𝑖𝑥 = ∑𝑛𝑖 (𝑉𝑖 − 𝑣𝑖 ), Δ𝐻𝑚𝑖𝑥 = ∑𝑛𝑖 (𝐻𝑖 − ℎ𝑖 )

• Δ𝑘𝑚𝑖𝑥 = 𝑘 − ∑𝑥𝑖 𝑘𝑖
Δ𝐾𝑚𝑖𝑥 𝐾 𝑛𝑖
• Δ𝑘𝑚𝑖𝑥 = = − ∑ 𝑘𝑖 = 𝑘 − ∑𝑥𝑖 𝑘𝑖
𝑛𝑇 𝑛𝑇 𝑛𝑇
• Δ𝑣𝑚𝑖𝑥 = 𝑣 − ∑𝑥𝑖 𝑣𝑖 , Δℎ𝑚𝑖𝑥 = ℎ − ∑𝑥𝑖 ℎ𝑖
Enthalpy of mixing
• At constant T, P Δℎ𝑚𝑖𝑥 > 0, < 0, = 0?
• What is enthalpy?
• Adding sulfuric acid heats up the solution?
• Why water cools when a salt is dissolved?
• Why add salt to thaw ice?

Milo Koretsky
Enthalpy of solution
• Enthalpy change per mole of solute, Δℎ𝑠
Entropy of mixing
• Mixing ideal gases at constant T, P?
• Example
• Let gases ‘a’ and ‘b’ mix with each other
• Δ𝑆𝑚𝑖𝑥 = Δ𝑆𝑎𝐼 + Δ𝑆𝑏𝐼 + Δ𝑆 𝐼𝐼
P, T 𝒑𝑶𝟐 , 𝑻
P, T
Isothermal Isothermal mixing
expansion

𝚫𝑺𝑰𝑶𝟐
𝚫𝑺𝑰𝑰 = 𝟎
𝒑𝑵𝟐 , 𝑻
P, T
𝚫𝑺𝑰𝑵𝟐

Isothermal
expansion
• Step 1: Constant T and volume expansion
𝑝𝑎 𝑝𝑏
𝐼
• Δ𝑆 = Δ𝑆𝑎𝐼 + Δ𝑆𝑏𝐼 = 𝑛𝑎 −𝑅𝑙𝑛 + 𝑛𝑏 −𝑅𝑙𝑛
𝑃 𝑃

• Step 2: Constant T & P


• Δ𝑆 𝐼𝐼 = 0
𝑝𝑎 𝑝𝑏
• Δ𝑆 = −𝑅 𝑛𝑎 𝑙𝑛 + 𝑛𝑏 𝑙𝑛
𝑃 𝑃

𝑝𝑎 𝑝𝑏 𝑝𝑖
• Δ𝑠 = −𝑅 𝑦𝑎 𝑙𝑛 + 𝑦𝑏 𝑙𝑛 = −R ∑𝑚
𝑖=1 𝑦𝑖 ln = Δ𝑠𝑚𝑖𝑥
𝑃 𝑃 𝑃
Regular solution vs. Ideal solution
• Ideal solution: Enthalpy of mixing = 0 and entropy of
mixing = −R ∑𝑚 𝑖=1 𝑥𝑖 ln 𝑥𝑖
• Regular solution: Ideal gas mixing is assumed for
liquid mixture (fully random mixing) but enthalpy of
mixing non-zero (but non-specific interactions)
• Which liquid mixtures can be approximated as regular
mixture?
• Liquids with vdW interactions and simple fluids
Multicomponent multi-phase system
• Include composition changes across phases
in fundamental energy equation?
• 𝒅𝑼 = 𝜹𝒒 + 𝜹𝒘
• 𝒅𝑼 = 𝑻𝒅𝑺 + 𝜹𝒘
• 𝛅𝐰 = 𝒅𝒓𝒊𝒗𝒊𝒏𝒈 𝒇𝒐𝒓𝒄𝒆 × 𝒓𝒆𝒔𝒖𝒍𝒕𝒊𝒏𝒈 𝒂𝒄𝒕𝒊𝒐𝒏
• e.g., -PdV, 𝒄𝒉𝒆𝒎𝒊𝒄𝒂𝒍 𝒅𝒓𝒊𝒗𝒊𝒏𝒈 𝒇𝒐𝒓𝒄𝒆 × 𝒅𝒏𝒊

• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + ∑𝒊 𝒄𝒉𝒆𝒎𝒊𝒄𝒂𝒍 𝒅𝒓𝒊𝒗𝒊𝒏𝒈 𝒇𝒐𝒓𝒄𝒆 × 𝒅𝒏𝒊


• Chemical driving force each component

29
Other examples of work and generalized energy equation

• Surface forces?
• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + 𝜸𝒅𝑨
• Rubber band?
• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + 𝒇𝒅𝑳
• Electrical forces?
• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + 𝚿𝒅𝑸
• Magnetic forces?
• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + 𝑩𝒅𝑰
• Generalized equation:
• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + 𝒇𝒅𝑳 + 𝜸𝒅𝑨 + 𝚿𝒅𝑸 + 𝑩𝒅𝑰 + ∑𝒊 ℱ𝒊 𝒅𝑿𝒊
30
Chemical processes T,P as a controlled variable

• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + ∑𝒊 𝒄𝒉𝒆𝒎𝒊𝒄𝒂𝒍 𝒅𝒓𝒊𝒗𝒊𝒏𝒈 𝒇𝒐𝒓𝒄𝒆 × 𝒅𝒏𝒊

• 𝑮 = 𝑼 − 𝑻𝑺 + 𝑷𝑽
• 𝒅𝑮 = 𝒅𝑼 − 𝑻𝒅𝑺 − 𝑺𝒅𝑻 + 𝑷𝒅𝑽 + 𝑽𝒅𝑷

• 𝒅𝑮 = −𝑺𝒅𝑻 + 𝑽𝒅𝑷 + ∑𝒊 𝒄𝒉𝒆𝒎𝒊𝒄𝒂𝒍 𝒅𝒓𝒊𝒗𝒊𝒏𝒈 𝒇𝒐𝒓𝒄𝒆 × 𝒅𝒏𝒊


Let us call the driving force for chemical components to escape
as chemical potential, 𝛍𝒊
• 𝒅𝑮 = −𝑺𝒅𝑻 + 𝑽𝒅𝑷 + ∑𝒊 𝛍𝒊 𝒅𝒏𝒊
• Extended form of Gibbs equation
31
Chemical processes T,P as a controlled variable

• 𝒅𝑮 = −𝑺𝒅𝑻 + 𝑽𝒅𝑷 + ∑𝒊 𝛍𝒊 𝒅𝒏𝒊


• Include composition in dG
• 𝑮 = 𝒇(𝑻, 𝑷, 𝒏𝟏 , 𝒏𝟐 , … )

𝝏𝑮 𝝏𝑮 𝝏𝑮
• 𝒅𝑮 = 𝒅𝑻 + 𝒅𝑷 + ∑𝒊 𝒅𝒏𝒊
𝝏𝑻 𝑷,𝒏𝑖 𝝏𝑷 𝑻,𝒏𝑖 𝝏𝒏𝒊 𝑻,𝑷,𝒏
𝒋≠𝒊

𝝏𝑮 𝝏𝑮 𝝏𝑮
• −𝑺 = ;𝑽 = ; 𝛍𝒊 = = 𝐺𝑖
𝝏𝑻 𝑷,𝒏𝑖 𝝏𝑷 𝑻,𝒏𝑖 𝝏𝒏𝒊 𝑻,𝑷,𝒏
𝒋≠𝒊

Applying Maxwell relations?


𝝏(𝑮𝒊 /𝑻) 𝟏 𝝏 𝑮𝒊 𝑮 𝒊 𝑻 𝑺𝒊 − 𝑮 𝒊 𝑯𝒊
𝝏 𝑮𝒊 𝝏 𝑮𝒊 = − = = −
−𝑺𝒊 = 𝑽𝒊 = 𝝏𝑻 𝑻 𝝏𝑻 𝑻𝟐 𝑻𝟐 𝑻𝟐
𝝏𝑻 𝑷,𝒏𝑖
𝝏𝑃 𝑇,𝒏𝑖
𝑷,𝒏𝑖 𝑷,𝒏𝑖
32
• 𝒅𝑼 = 𝑻𝒅𝑺 − 𝑷𝒅𝑽 + ∑𝒊 𝛍𝒊 𝒅𝒏𝒊
• 𝒅𝑯 = 𝑻𝒅𝑺 + 𝑽𝒅𝑷 + ∑𝒊 𝝁𝒊 𝒅𝒏𝒊
• 𝒅𝑨 = −𝑺𝒅𝑻 − 𝑷𝒅𝑽 + ∑𝒊 𝝁𝒊 𝒅𝒏𝒊
• 𝒅𝑮 = −𝑺𝒅𝑻 + 𝑽𝒅𝑷 + ∑𝒊 𝝁𝒊 𝒅𝒏𝒊
• U=f(S,V,n1,n2…); H=f(S,P,n1,n2…); A=f(T,V,n1,n2…)
𝝏𝑼 𝝏𝑼 𝝏𝑼
• 𝒅𝑼 = 𝒅𝑺 + 𝒅𝑽 + ∑𝒊 𝒅𝒏𝒊
𝝏𝑺 𝑽,𝒏 𝝏𝑽 𝑺,𝒏 𝝏𝒏𝒊 𝑺,𝑽,𝒏
𝒋≠𝒊

𝝏𝑯 𝝏𝑯 𝝏𝑯
• 𝒅𝑯 = 𝒅𝑺 + 𝒅𝑷 + ∑𝒊 𝒅𝒏𝒊
𝝏𝑺 𝑷,𝒏 𝝏𝑷 𝑺,𝒏 𝝏𝒏𝒊 𝑺,𝑷,𝒏
𝒋≠𝒊

𝝏𝑨 𝝏𝑨 𝝏𝑨
• 𝒅𝑨 = 𝒅𝑻 + 𝒅𝑽 + ∑𝒊 𝒅𝒏𝒊
𝝏𝑻 𝑽,𝒏 𝝏𝑽 𝑻,𝒏 𝝏𝒏𝒊 𝑻,𝑽,𝒏
𝒋≠𝒊

𝝏𝑮 𝝏𝑼 𝝏𝑯 𝝏𝑨
• 𝛍𝒊 = = = =
𝝏𝒏𝒊 𝑻,𝑷,𝒏 𝝏𝒏𝒊 𝑺,𝑽,𝒏 𝝏𝒏𝒊 𝑺,𝑷,𝒏 𝝏𝒏𝒊 𝑻,𝑽,𝒏
𝒋≠𝒊 𝒋≠𝒊 𝒋≠𝒊 𝒋≠𝒊
33
Relationship among partial molar quantities
𝑯 = 𝑼 + 𝑷𝑽
𝑯 𝒊 = 𝑼 𝒊 + 𝑷 𝑽𝒊
𝑨 𝒊 = 𝑼 𝒊 − 𝑻𝑺 𝒊
𝑮𝒊 = 𝑯𝒊 − 𝑻𝑺𝒊

𝝏𝑯 𝝏𝑼 𝝏𝑽 𝝏𝑷
= +𝑷 +𝐕
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊

𝑨𝒕 𝒄𝒐𝒏𝒔𝒕𝒂𝒏𝒕 𝑻, 𝑷, 𝒏𝒋≠𝒊

𝝏𝑯 𝝏𝑼 𝝏𝑽
= +𝑷
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊
34
Multi-component phase equilibria
• phase equilibria
• 𝒅𝑮 = 𝟎 = 𝒅𝑮𝜶 + 𝒅𝑮𝜷
𝝏𝑮
• 𝒅𝑮𝜶 = −𝑺𝜶 𝒅𝑻 + 𝑽𝜶 𝒅𝑷 + ∑𝒊 𝛍𝜶𝒊 𝒅𝒏𝜶𝒊 , where, 𝛍𝒊 =
𝝏𝒏𝜶
𝒊 𝑻,𝑷,𝒏𝜶
𝒋≠𝒊

𝜷 𝜷 𝝏𝑮
• 𝒅𝑮 = −𝑺 𝒅𝑻 + 𝑽 𝒅𝑷 + ∑𝒊 𝛍𝒊 𝒅𝒏𝒊 , where, 𝛍𝒊 =
𝜷 𝜷 𝜷
𝜷
𝝏𝒏𝒊 𝜷
𝑻,𝑷,𝒏𝒋≠𝒊
• at constant T, P
𝜷 𝜷
• 𝒅𝑮 = 𝒅𝑮𝜶 + 𝒅𝑮𝜷 = ∑𝒊 𝛍𝜶𝒊 𝒅𝒏𝜶𝒊 + ∑𝒊 𝛍𝒊 𝒅𝒏𝒊 = 𝟎
𝜷
• 𝒅𝒏𝜶𝒊 = −𝒅𝒏𝒊
𝜷
• ∑𝒊(𝛍𝜶𝒊 −𝛍𝒊 )𝒅𝒏𝜶𝒊 = 𝟎
𝜷
• 𝛍𝜶𝒊 = 𝛍𝒊 , for each species across all phases
35
Phase equilibria - Fugacity
𝝏𝑮
𝛍𝒊 =
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊

• Escaping tendencies of the species are equal across the phases


• partial molar Gibbs free energy should be equal
• a quantity that balances enthalpic and entropic components of energy are
equal for each species across phases
• Molecularly, pressure is also related to molecular potential and
kinetic energy of molecules (so is Gibbs free energy).
• Why not use a pressure like term to characterize the escaping
tendencies!

• G.N. Lewis first recognized and defined this quantity, fugacity


36
Fugacity
(∑𝒊 𝒏𝒊 )𝑹𝑻
• for ideal gas, 𝝏𝝁𝒊 𝝏𝑽 𝝏 𝑷 𝑅𝑇
𝑽𝒊 = 𝑽𝒊 = = =
𝝏𝑃 𝑇,𝒏𝑖 𝝏𝑛𝑖 𝝏𝑛𝑖 𝑃
𝑇,𝑃,𝑛𝑗≠i 𝑇,𝑃,𝑛𝑗≠i

𝝏𝝁𝒊 𝑅𝑇
• 𝑽𝒊 = =
𝝏𝑃 𝑇,𝒏𝑖 𝑃
𝑹𝑻 𝑷
• 𝒅𝝁𝒊 = 𝒅𝑷; 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝑷 𝑷𝟎

• Since energies have no absolute values, choose appropriate and


convenient reference state
• Reference chemical potential, i0: at reference pressure P0 and at
system T (ideal gas)

37
For real gases - Fugacity
𝑷
• for ideal gas, 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝑷𝟎
• For mixtures,
• 𝒑𝒊 = 𝒚𝒊 𝑷; 𝒑𝟎𝒊 = 𝒚𝒊 𝑷𝟎
𝒑𝒊
• 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln , in terms of partial pressure for ideal gases
𝒑𝟎𝒊

• G.N. Lewis, defined a new term fugacity for real gases, 𝒇,


𝒇𝒊
• 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝒇𝟎𝒊
• Fugacity of real gas is similar to partial pressure of ideal gas

• Reference state for fugacity should be same as for chemical potential


• Full definition of fugacity
𝒇𝒊
• 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝒇𝟎𝒊

𝒇𝒊
• lim ≡𝟏
𝑷→𝟎 𝒑𝒊

38
• Fugacity coefficient,
𝒇𝒊 𝒇𝒊
• 𝝋𝒊 = = (Note: partial pressure of the system; not reference state)
𝒑𝒊 𝒚𝒊 𝑷𝒔𝒚𝒔
• Fugacity coefficient compares fugacity with ideal partial pressure under
identical conditions

• if i = 1;
• No net intermolecular interactions: attractive and repulsive forces balance each
other
• if i < 1;
• corrected pressure or tendency to escape is lower than in the case of ideal gas
• if i > 1;
• corrected pressure or tendency to escape is higher than in the case of ideal gas

39
Fugacity and Fugacity Coefficient of Pure Gases
𝒇𝒊
• 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝒇𝟎𝒊

𝒇𝒊
• lim ≡𝟏
𝑷→𝟎 𝒑𝒊
• For gases, choose a reference state of very low pressure where
gases behave ideally

Choice reference state differs for gases vs. condensed phases (liquids and solids) 40
𝒇𝒊 and 𝝋𝒊 for gas mixtures
Fugacity and coefficients are functions of composition
𝒇𝒊
𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝒇𝟎𝒊

Reference pressure is at low values such that gases are ideal


gas; Temperature and number of moles same as that of system

𝒇𝒊 𝒇𝒊 𝒇𝒊
𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln = 𝑹𝑻 ln = 𝑹𝑻 ln
𝒇𝟎𝒊 𝒑𝟎𝒊 𝑦𝑖 𝑃𝑙𝑜𝑤
𝒑𝟎𝒊 = 𝒚𝒊 𝑷𝟎 = 𝒚𝒊 𝑷𝒍𝒐𝒘

41
Fugacity in liquid phase
• In gas phase,
𝒇𝒊 𝒇𝒊 𝒇𝒊
• 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln = 𝑹𝑻 ln = 𝑹𝑻 ln
𝒇𝟎𝒊 𝒑𝟎𝒊 𝒚𝒊 𝑷𝒍𝒐𝒘

• 𝒑𝟎𝒊 = 𝒚𝒊 𝑷𝟎 = 𝒚𝒊 𝑷𝒍𝒐𝒘
𝒇𝒊 𝒇𝒊
• 𝝋𝒊 = =
𝒑𝒊 𝒚𝒊 𝑷𝒔𝒚𝒔

• There is a problem to define reference state for liquid?


• Intermolecular interactions are always present

For liquids,
𝒇𝒊 𝒇𝒊
• 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln = 𝑹𝑻 ln
𝒇𝟎𝒊 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊
• 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 = 𝒙𝒊 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 (𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 is independent of concentration, obviously defined at xi=1)
• for L/R, 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 = 𝒇𝒊
• 𝒇𝒐𝒓 𝑯𝒆𝒏𝒓𝒚′ 𝒔 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 =ℋ
𝒇𝒊 𝒇𝒊
• But, 𝜸𝒊 = = 𝒐𝒓 𝒇𝒊 = 𝛾𝑖 𝑥𝑖 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊
𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 𝒙𝒊 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊
42
Ideal solution
• Two ways to define a ideal solution to use as a reference
state
• Consider a binary mixture (a,b), reference state for species a, if
• xa1: a-a interactions dominate - Lewis/Randall rule
• xa0: a-b interactions dominate - Henry’s law
• Compare real interactions of ‘a’ (a-a and a-b) with ideal solution
(a-a or a-b)

Henry’s Lewis/Randall
𝒇𝒊 In reference state fugacity of liquid is linearly
𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln 43
𝒇𝑖𝑑𝑒𝑎𝑙
𝒊
proportional to the mole fraction
Reference state: Ideal solution

• 𝒇𝒊 = 𝛾𝑖 𝒇𝑖𝑑𝑒𝑎𝑙
𝒊 = 𝛾 𝑥 𝒇𝒊𝒅𝒆𝒂𝒍
𝑖 𝑖 𝒊
• What does this graph say?
• Like or unlike interactions are stronger?
• Like interactions (a-a) is stronger than (a-
b)
• 𝒇𝒊 (a-a &a-b) > 𝒇𝑖𝑑𝑒𝑎𝑙
𝒊 (a-a, Randall/Lewis)
• a >1
• 𝒇𝒊 (a-a &a-b) < 𝒇𝑖𝑑𝑒𝑎𝑙
𝒊 (a-b, Henry’s)
• aHenry’s <1

Henry’s Lewis/Randall 44
Redefining criteria for equilibria

45
Redefining criteria for equilibria

• For vapor-liquid equilibrium

46
Fugacity of gas vs liquid phases
Gases Liquids
Fugacity 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝒇𝒊
= 𝒇𝒊 𝒇𝒊 𝒇𝒊
𝒇𝟎𝒊 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln = 𝑹𝑻 ln 𝑜𝑟𝑹𝑻 ln
𝒇𝒊 𝒇𝒊 𝒇𝟎𝒊 𝒇𝒊 𝓗𝒊
𝑹𝑻 ln = 𝑹𝑻 ln and
𝒑𝟎𝒊 𝑦𝑖 𝑃𝑙𝑜𝑤
𝒇𝒊
lim ≡𝟏
𝑷→𝟎 𝒑𝒊
Reference state Ideal gas: 𝒇𝟎𝒊 = pi0=xiPlow ideal soln: P0=Psys; T0=Tsys ; pure fluids: 𝒇𝟎𝒊 = 𝒇𝒊 𝒐𝒓 𝓗
At P0=Plow; T0=Tsys; ni0=ni,sys Lewis/Randall(xa1) :𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 = 𝒇𝒊 = 𝒑𝒖𝒓𝒆 𝒔𝒑𝒆𝒄𝒊𝒆𝒔 𝒇𝒖𝒈𝒂𝒄𝒊𝒕𝒚
Henry’s (xa1) :𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 = 𝓗
Deviation from Fugacity coefficient: Activity coefficient: 𝛾𝑖 =
𝒇𝒊
=
𝒇𝒊 𝒇
= 𝒙 𝒇𝒊 𝒐𝒓
𝒇𝒊
non-ideality 𝒇𝒊 𝒇𝒊 𝒇𝒊 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 𝒙𝒊 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 𝒊 𝒊 𝒙𝒊 𝓗𝒊
𝝋= = = 𝒇𝒊 𝒇𝒊
𝒑𝒊,𝒔𝒚𝒔 𝒚𝒊 𝑷𝒔𝒚𝒔 𝒚𝒊 𝑷𝑳𝒐𝒘 Activity 𝒂𝒊 = ; 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln = 𝑅𝑇 ln 𝑎𝑖 = 𝑅𝑇 ln 𝑥𝑖 𝛾𝑖
𝒇𝟎𝒊 𝒇𝟎𝒊

Fugacity
determination

47
Activity coefficient vs. fugacity coefficient
Activity
Activity is more useful when we look at
Compares real fugacity with ideal
solution fugacity (same T, P and
chemical equilibrium
concentration)
𝒇𝒊 Ref. state - ideal soln: P0=Psys; T0=Tsys ;
𝒂𝒊 = pure fluids: 𝒇𝟎𝒊 = 𝒇𝒊 𝒐𝒓 𝓗
𝒇𝟎𝒊

𝒇𝒊 𝒇𝒊 𝒇𝒊
𝒂𝒊 = = or
𝒇𝟎𝒊 𝒇𝒊 𝓗

𝒇𝒊 𝒇𝒊
𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 𝐥𝐧 = 𝑹𝑻𝒍𝒏𝒂𝒊 = 𝑹𝑻 𝐥𝐧 = 𝑹𝑻𝒍𝒏(𝒙𝒊 𝜸𝒊 )
𝒇𝟎𝒊 𝒇𝒊 𝒐𝒓 𝓗
48
Determining fugacity of pure real gas

• Tables
• EOS
• Generalized correlations

49
Tables

Saturated steam at 1 atm is at 100 0C

Reference state: what is lowest pressure at


100 0C that can be read from steam tables
50
51
EOS

vdW EOS in virial form

52
Generalized correlations
If z(0) and z(1) can be converted in to fugacity coefficient, then

ln ln ln

54
𝒇𝒊 and 𝝋𝒊 for gas mixtures: EOS
How to incorporate EOS in the following equation?

𝝏𝝁𝒊
= 𝑽𝒊
𝝏𝑷 𝑻,𝒏𝒊

If the EOS is volume-explicit [V=V(P)], then integrating above will give 𝒇𝒊 and 𝝋𝒊

Most EoS are pressure explicit [P=P(V)]: we can apply cyclic rule
𝝏𝑽
𝑽𝒊 =
𝝏𝒏𝒊 𝑻,𝑷,𝒏𝒋≠𝒊

Keeping three parameters, T, P, ni as variables (𝑛𝑗≠𝑖 𝑖𝑠 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡)

𝝏𝑽 𝝏𝑷 𝝏𝒏𝒊
= −𝟏
𝝏𝒏𝒊 𝑻,𝑷 𝝏𝑽 𝑻,𝒏𝒊 𝝏𝑷 𝑻,𝑽 55
𝝏𝑽 𝝏𝑷 𝟏 𝝏𝑽 𝝏𝑷 𝝏𝑷
=− =−
𝝏𝒏𝒊 𝝏𝑽 𝝏𝒏𝒊
𝑻,𝑷 𝑻,𝒏𝒊 𝝏𝒏𝒊 𝝏𝑽 𝝏𝒏𝒊
𝝏𝑷 𝑻,𝑽
𝑻,𝑷 𝑻,𝒏𝒊 𝑻,𝑽

Now, keeping T and ni are constant,


𝝏𝑽 𝒅𝑷 𝝏𝑷 𝝏𝑷
=− 𝑽𝒊 𝒅𝑷 = − 𝒅𝑽
𝝏𝒏𝒊 𝑻,𝑷
𝒅𝑽 𝝏𝒏𝒊 𝑻,𝑽 𝝏𝒏𝒊 𝑻,𝑽

Substituting
𝒗 𝑷 𝑷
𝒇𝒊 𝝏𝑷
𝝁𝒊 − 𝝁𝒐𝒊 = 𝑹𝑻 𝒍𝒏 = 𝑽𝒊 𝒅𝑷 = − 𝒅𝑽
𝒚𝒊 𝑷𝑳𝒐𝒘 𝑷𝑳𝒐𝒘 𝑷𝑳𝒐𝒘 𝝏𝒏𝒊 𝑻,𝑽

For example,

56
Generalized correlations

log 𝝋 = log 𝝋(𝟎) + 𝝎𝒎𝒊𝒙 log 𝝋(𝟏)

57
𝒇𝒊 and 𝝋𝒊 for gas mixtures
• Fugacity of a species in a gas mixture depends on composition
(i-j and i-i interactions)
• Three level of treatments for determining fugacity
• Full rigor, mixing rules needed

• All interactions are the same (i interacts only with i, no


mixing rules needed)

• Ideal gas, no interactions

58
When is Lewis fugacity rule good?
• Consider the expression for fugacity of ‘a’ in
a binary mixture obeying vdW EOS and
mixing rules

𝒗
• Under what conditions, 𝒇𝒂 = 𝒚𝒂 𝒇𝒗𝒂 ?

59
Ideal solution: Macroscopic
• At macroscopic level, ideal solution obeys following

𝝁𝒊 = 𝑹𝑻 ln 𝒇𝑖𝑑𝑒𝑎𝑙
𝑖
𝒈𝒊 = 𝑹𝑻 ln 𝑓𝑖𝑖𝑑𝑒𝑎𝑙

60
Calculation of pure species fugacity, fi, (function of T,P) for Lewis/Randall
reference state
To obtain Lewis/Randall reference state, we need pure species fugacity

If the P is closer to Pisat then, above


expression holds
 Fugacity at Psat needs to be corrected
for accurate fugacity at P,
 We need fi as a function of P,
 Use thermodynamic relations to obtain this
function

61
Pure species fugacity, fi (function of T,P)
For pure species, 𝒅𝒈𝒊 = −𝒔𝒊 𝒅𝑻 + 𝒗𝒊 𝒅𝑷
𝝏𝒈𝒊 𝝏 ln𝒇𝒊
Therefore, = 𝒗𝒊 = 𝑹𝑻 (since, 𝒈𝒊 = 𝑹𝑻ln𝒇𝒊 )
𝝏𝑷 𝑻 𝝏𝑷 𝑻

𝒗𝒊 𝝏 ln𝒇𝒊 𝒗𝒊
= 𝒅𝑷 = 𝒅(ln𝒇𝒊 )
𝑹𝑻 𝝏𝑷 𝑻 𝑹𝑻

𝒇𝒍𝒊 𝒇𝒍𝒊 𝒇𝒍𝒊 𝑷 𝒗𝒊


𝒇𝒍,𝒔𝒂𝒕
𝒅(ln𝒇𝒊 ) = 𝒍𝒏 = 𝒍𝒏 = 𝑷𝒔𝒂𝒕
𝒅𝑷
𝒊 𝒇𝒍,𝒔𝒂𝒕
𝒊
𝝋𝒔𝒂𝒕
𝒊 𝑷𝒊
𝒔𝒂𝒕
𝒊 𝑹𝑻

62
Calculation of pure species fugacity, fi

Exponential term: Pressure correction for


fugacity: Poynting correction
 Usually negligible below 100 bar
 What approximation can be done to simply this equation?
 Liquids can be assumed to be incompressible, molar volume constant

 At low P and Pisat, gas behaves ideally, fugacity coefficient and exponential factor
equal to 1, then

63
Psat from Antoine equation can be used to obtain Lewis/Randall reference fugacity
Henry’s constant vs. P and T
• Start with how chemical potential varies with P and T
• Vapor-liquid equilibrium:
𝜷 𝜷
𝝁𝜶𝒊 𝝁 𝒅
𝝁𝜶
𝒊
=𝒅
𝝁𝒊 Exact diff
= 𝒊 𝑻 𝑻
𝑻 𝑻 Maxwell

• This assume ideal gas for vapor phase


• Henry’s law valid at infinite dilution, xi  0  𝒇𝒍𝒊 = 𝒙𝒊 ℋ
𝒇𝒍𝒊
𝝏(𝝁𝒐𝒊 + 𝑹𝑻 𝒍𝒏 𝒐
𝝏𝝁𝒍𝒊 𝒇𝒊 𝝏(𝑹𝑻 𝒍𝒏 𝒇𝒍𝒊 𝝏(𝑹𝑻 𝒍𝒏 𝒙𝒊 𝓗 𝝏𝒍𝒏 𝒙𝒊 𝝏𝒍𝒏 𝓗 𝐑𝐓
= = = = 𝑹𝑻 + =
𝝏𝒙𝒊 𝝏𝒙𝒊 𝑻,𝑷 𝝏𝒙𝒊 𝝏𝒙𝒊 𝑻,𝑷
𝝏𝒙𝒊 𝑻,𝑷
𝝏𝒙𝒊 𝑻,𝑷
𝐱𝐢
𝑻,𝑷 𝑻,𝑷

64
Thermodynamic relations between i
• Purpose is to directly obtain all activity coefficients in a mixture from an
expression for Gibbs free energy
• Gibbs-Duhem allowed inter-relationship between chemical potential, I

• Can activity coefficients be related using this equation?


0 0

=0

Gibbs-Duhem eq with activity coefficients


65
For a binary mixture

If a vs. xa is known, slope of b with xa is constrained

xb0, Henry’s
xb1, L/R ideal ref
ideal ref

lna based on Lewis/Randall lnb based on Lewis/Randall


reference state reference state
66
Fugacity of gas vs liquid phases
Gases Liquids
Fugacity 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln
𝒇𝒊
= 𝒇𝒊 𝒇𝒊 𝒇𝒊
𝒇𝟎𝒊 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln = 𝑹𝑻 ln 𝑜𝑟𝑹𝑻 ln
𝒇𝒊 𝒇𝒊 𝒇𝟎𝒊 𝒇𝒊 𝓗𝒊
𝑹𝑻 ln = 𝑹𝑻 ln and
𝒑𝟎𝒊 𝑦𝑖 𝑃𝑙𝑜𝑤
𝒇𝒊
lim ≡𝟏
𝑷→𝟎 𝒑𝒊
Reference state Ideal gas: 𝒇𝟎𝒊 = pi0=xiPlow ideal soln: P0=Psys; T0=Tsys ; pure fluids: 𝒇𝟎𝒊 = 𝒇𝒊 𝒐𝒓 𝓗
At P0=Plow; T0=Tsys; ni0=ni,sys Lewis/Randall(xa1) :𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 = 𝒇𝒊 = 𝒑𝒖𝒓𝒆 𝒔𝒑𝒆𝒄𝒊𝒆𝒔 𝒇𝒖𝒈𝒂𝒄𝒊𝒕𝒚
Henry’s (xa1) :𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 = 𝓗
Deviation from Fugacity coefficient: Activity coefficient: 𝛾𝑖 =
𝒇𝒊
=
𝒇𝒊 𝒇
= 𝒙 𝒇𝒊 𝒐𝒓
𝒇𝒊
non-ideality 𝒇𝒊 𝒇𝒊 𝒇𝒊 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 𝒙𝒊 𝒇𝒊𝒅𝒆𝒂𝒍
𝒊 𝒊 𝒊 𝒙𝒊 𝓗𝒊
𝝋= = = 𝒇𝒊 𝒇
𝒑𝒊,𝒔𝒚𝒔 𝒚𝒊 𝑷𝒔𝒚𝒔 𝒚𝒊 𝑷𝑳𝒐𝒘 Activity 𝒂𝒊 = ; 𝝁𝒊 − 𝝁𝟎𝒊 = 𝑹𝑻 ln 𝟎𝒊 = 𝑅𝑇 ln 𝑎𝑖 = 𝑅𝑇 ln 𝑥𝑖 𝛾𝑖
𝒇𝟎𝒊 𝒇𝒊

Fugacity Fugacity of pure species and mixture gases: Activity coefficient of reference states:
determination Tables, EOS, and generalized correlations a) Pure species: Pisat and Poynting correction (for P)
𝑙 𝑣𝑖𝑙
𝑠𝑎𝑡
𝑓𝑖 = 𝜑𝑖 𝑃 𝑒𝑥𝑝 (𝑃 − 𝑃𝑖𝑠𝑎𝑡 )
𝑅𝑇
𝑃 𝑉𝑖∞
b) Henry’s constant, ℋ𝑖𝑎𝑡 𝑃 = ℋ𝑖𝑎𝑡 1 𝑏𝑎𝑟 𝑒𝑥𝑝 1 𝑏𝑎𝑟 𝑅𝑇
𝑑𝑃

67
• Fugacities of condensed phase mixture can be
calculated from volumetric properties as well
• Not very practical as we need volumetric data at constant T
and xi to integrate from zero density (I.G. density) to
condensed phase density through 2-phase region

• Alternative method: Define excess function


(deviation from ideal solution)
• Activities can be determined from Gibbs excess function
Determining i using Gibbs excess energy, gE

• Using excess Gibbs energy, activity coefficient of all species in a


mixture can be obtained from a single quantitative expression

• Fugacity    (gE vs. x)


69
Excess properties
• Using excess Gibbs energy, activity coefficient
of all species in a mixture can be obtained from
a single quantitative expression
• Excess property 𝝏𝑮𝑬
• = −𝑺𝑬
𝝏𝑻 𝑷,𝒙
• 𝐺 𝐸 = 𝐺 𝑇, 𝑃, 𝑥 − 𝐺 𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥
• 𝐴𝐸 = 𝐴 𝑇, 𝑃, 𝑥 − 𝐴𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥 𝝏
𝑮𝑬
𝑻 𝑯𝑬
• 𝐻𝐸 = 𝐻 𝑇, 𝑃, 𝑥 − 𝐻𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥 • = − 𝟐
𝝏𝑻 𝑷,𝒙 𝑻

• 𝑈 𝐸 = 𝑈 𝑇, 𝑃, 𝑥 − 𝑈 𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥 𝝏𝑮𝑬
• = 𝑽𝑬
𝝏𝑷 𝑻,𝒙
70
Molar and partial molar excess functions
• 𝑔𝐸 = 𝑔 𝑇, 𝑃, 𝑥 − 𝑔𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥
• 𝑎𝐸 = 𝑎 𝑇, 𝑃, 𝑥 − 𝑎𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥
• ℎ𝐸 = ℎ 𝑇, 𝑃, 𝑥 − ℎ𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥
• 𝑢𝐸 = 𝑢 𝑇, 𝑃, 𝑥 − 𝑢𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥
• 𝑣 𝐸 = 𝑣 𝑇, 𝑃, 𝑥 − 𝑣 𝑖𝑑𝑒𝑎𝑙 𝑇, 𝑃, 𝑥

Partial molar excess property


𝝏𝑲 𝑬 𝝏 𝑲 − 𝑲𝒊𝒅𝒆𝒂𝒍
𝑲 = 𝚺𝐢 𝒏𝒊 𝑲𝒊
𝑲𝑬 = = = 𝑲𝒊 − 𝑲𝒊𝒅𝒆𝒂𝒍
𝒊 • 𝑲𝑬 = 𝚺𝐢 𝒏𝒊 𝑲𝑬𝒊
𝝏𝒏𝒊 𝑷,𝑻,𝒏𝒋 ≠𝒊
𝝏𝒏𝒊 𝑷,𝑻,𝒏𝒋 ≠𝒊
• 𝑮𝑬 = 𝚺𝐢 𝒏𝒊 𝑮𝑬𝒊
• 𝑮𝑬 = 𝑮𝒊 − 𝑮𝒊𝒅𝒆𝒂𝒍 • 𝑨𝑬 = 𝚺𝐢 𝒏𝒊 𝑨𝑬𝒊
𝒊
• 𝑯𝑬 = 𝚺𝐢 𝒏𝒊 𝑯𝑬𝒊
• 𝑨𝑬 = 𝑨𝒊 − 𝑨𝒊𝒅𝒆𝒂𝒍
𝒊 • 𝑼𝑬 = 𝚺𝐢 𝒏𝒊 𝑼𝑬𝒊
• 𝑯𝑬 = 𝑯𝒊 − 𝑯𝒊𝒅𝒆𝒂𝒍
𝒊 • 𝑽𝑬 = 𝚺𝐢 𝒏𝒊 𝑽𝑬𝒊
• 𝑼𝑬 = 𝑼𝒊 − 𝑼𝒊𝒅𝒆𝒂𝒍
𝒊
• 𝑽𝑬 = 𝑽𝒊 − 𝑽𝒊𝒅𝒆𝒂𝒍
𝒊
Excess properties as property change of mixing

Excess property = property change of (real-ideal) mixing

• Class I properties (u,h,v):

• Class II properties (g,a,s):

72
Activity coefficients from excess Gibbs energy

or

Obtain all activity coefficient by taking appropriate


partial derivatives of above equation 73
• A is experimental fitting parameter (depends on T and P but not on xa or xb)

74
Asymmetric models for gE Two-Suffix Margules - symmetric

Intermolecular interactions: Intermolecular interactions: NOT


similar in type and magnitude similar in type and magnitude

𝑮𝑬𝒂 = 𝑨𝒙𝟐𝒃 = 𝑨 𝟏 − 𝒙𝒂 𝟐 = 𝑹𝑻𝒍𝒏𝜸𝒂

𝑨 𝟏−𝒙𝒂 𝟐
𝒍𝒏𝜸𝒂 = 𝑹𝑻

𝑨𝒙𝟐𝒂
𝒍𝒏𝜸𝒃 = 𝑹𝑻

Two-Suffix Margules good Two-Suffix Margules NOT good 75


Two-Suffix Margules Equation
• Simplest non-ideal model for gE vs. x
• Similar size and strongly similar interacting molecules
• Consider binary solution with Lewis/Randall reference state

• Two constraints
• Obeys Gibbs-Duhem equation

• Simplest non-ideal model that satisfies both these


conditions,

76
Does two-suffix Margules obey Gibbs-Duhem equation?

xa+xb=1

77
Two-suffix Margules in terms of mixing equations

∆𝒈𝒎𝒊𝒙 = 𝒈 − 𝒙𝒊 𝒈𝒊

78
• 1: Pure species summation
• 2: decrease in g due to mixing
of ideal solution
• 3: decrease in g due to non-
ideal mixing (gE)
• if A < 0; (a-a, b-b) < or > a-b?
• Axb2 = RT lna; Axa2 = RT lnb
• a < 1 for L/R reference state
• Unlike interactions are stronger than
like
• If A is positive: mixing is
entropically favored and
separation energetically favored
79
• When A is large positive?
• a-a and b-b interactions much stronger than a-b.
• The gE flips the total g to get a maximum and two minimums

The system phase separates at the two minima


compositions to minimize g ( a-rich & b-rich phases)

80
 aHenry’s with respect to Henry’s reference

• a with respect to L/R reference

𝒇∞
 At infinite dilution (xa0), a Henry’s= 𝟏 𝐚𝐧𝐝 𝓗 = 𝒂
𝒙𝒂

81
Raoult’s Law
𝒚𝒊 𝝋𝒗𝒊 𝑷 = 𝒙𝒊 𝜸𝒍𝒊 𝒇𝟎𝒊
• Ideal gas, ideal solution, L/R reference state
K-value of a species

• Consider a binary system, (P and T


dependence)

82
83
Vapor-Liquid Equilibrium, VLE: Bubble and dew point
• Vapor-liquid equilibria: very important for optimization of processes in petroleum
industry (eg., extraction, distillation)
• General rule: index 1 or a is for most volatile component (low B.P or high V.P)

 Consider propane-pentane binary mixture at 5 bar

B.P of ‘2’
 Bubble temperature: When a liquid mixture is heated,
first vapor bubble appears
Td
 y1 different than x1?
 Vapor phase richer in volatile component
 Addition of more heat increases temperature (unlike
pure species)
 Dew temperature: When a vapor mixture is cooled, first
Tb
y1 liquid drop appears
 Bubble curve and dew curves
B.P of ‘1’

84
Phase diagrams

85
Non-ideal liquids
• a-a and b-b interactions are stronger/weaker than
a-b interactions, Raoult’s law does not apply

86
We want to solve for P, xa, xb

xa+xb=1

Psat? ?
Antoine equation Using vdW EOS

sat?
Using vdW EOS ?
Three suffix Margules

87
Since we do not know P, xa, xb, we set the initial values of 𝝋𝒊 𝒂𝒏𝒅 𝜸𝒊 = 𝟏 88
Isothermal VLE diagrams
 Consider propane-pentane binary mixture at 300 K
 Which are bubble and dew curves?
 Bottom curve is dew and top bubble

Pvap of ‘1’

Pb

Pvap of ‘2’
Pd

89
Positive and negative deviations from Raoult's Law

90
Ethane(1) and benzene(2) VLE curves at 10 oC and 25 oC

 Why dew curves stay close to P=0 and x(1) = 1


axes?
 Due to large difference in volatility (Pvap (ethane)
>> Pvap(benzene)
 Horizontal line (more benzene in the liquid phase)
 Need very high pressures to complete compressing
ethane
 Bubble and dew pressures increase with T
 Vapor pressure increases with T

91
Retrograde condensation
 For T > critical Ta, and T < Tpc
 V-L have identical composition and properties
 Above critical point only one homogeneous
phase
 Consider same line for T=175 oC, no bubble
point but two dew points!
 But there is a second dew point (liquid phase
goes through maximum)
 Retrograde condensation
 Consider T=235 0C, no bubble or dew points
 No transition line from vapor mixture into
homogeneous supercritical phase

92
Phase diagrams vs. Gibbs energy (g)
 Equilibrium state =
minimum G
 Subcooled liquid: gliq < gvap
for all xa
 Superheated vapor: gliq >
gvap for all xa

 When Tboil,a < T < Tboil,b


 Why gvap is more +ve at low
xa ?
 High amounts of less
volatile component

93
Azeotropes (boiling without change)
• When deviations from Raoult’s is large, Px and Py can have same extrema in
pressure - extrema point is called azeotrope
• Large deviations – unlike interactions very different from like interactions
• Extrema occur at exactly same composition, xi=yi, pinch point bad for
separations
• Maxima occur when deviations from ideality is such that Psys exceeds Psat of
lighter component
• Also when Psat are closer together

94
Minimum boiling

Maximum boiling

95
Solubility of gases in liquids
• Can Lewis/Randall reference state be used for activity
coefficient of O2 dissolved in water (at 25 oC and 1 atm)?
• O2 is supercritical state at 25 oC, 1 atm and hence no pure liquid
species at this T and P  Lewis/Randall application is questionable
• Henry’s reference state is well suited and convenient
• Generally, if the T >> Tc of the species, use Henry’s reference state
• If H is lower, unlike interaction > like interaction ?

96
VLE using Henry’s Law reference state (ideal vapor phase and ideal solute)
𝑯𝒆𝒏𝒓𝒚′ 𝒔
𝒚𝒂 𝝋𝒗𝒂 𝑷 = 𝒙𝒂 𝜸𝒂 ℋ𝒂 𝒚𝒃 𝝋𝒗𝒃 𝑷 = 𝒙𝒊 𝜸𝒍𝒃 𝒇𝟎𝒃

Consider binary mixture (O2 in water),


• If gas mixture is ideal
• a=1, b=1
• a is sparingly soluble in liquid, xa0; aHenry’s =1;
• For b, which reference is convenient when xa0?
• Lewis/Randall, b=1

This approximation (aHenry’s =1) works well up to xa = 0.03


97
VLE using Henry’s Law reference state for non-ideal solute and ideal vapor

• Consider binary mixture and xa is large enough that


number of a-a interactions increase 
• aHenry’s ≠ 1, b ≠ 1
𝑯𝒆𝒏𝒓𝒚′ 𝒔
𝒚𝒂 𝑷 = 𝒙𝒂 𝜸𝒂 ℋ𝒂
𝒚𝒃 𝑷 = 𝒙𝒊 𝜸𝒍𝒃 𝒇𝟎𝒃

98
VLE using Henry’s Law reference state for ideal solute and non-ideal vapor

• Consider binary mixture, What happens at high P but xa is small?


• vapor phase is not ideal but in the liquid mixture, a-b and b-b interactions for a
and b, respectively are ideal 
• a≠1b≠1

At high P, concentration of less volatile ‘b’ in gas mixture is negligible : yb  0, ya  1

Therefore, 𝝋𝒂 = 𝝋𝒂

How to obtain at Henry’s constant of ‘a’ and pure species fugacity of b at high P?

yaaP ^
99
Fugacity in the solid phase
• Pure solids/liquids: by definition ideal since all
intermolecular interactions are the same

• Solid solutions (substitutions or interstitials):

https://pubs.acs.org/doi/pdf/10.1021/ie801179a 100

You might also like