L 17-22 : Surface area and pore size

analysis

Prof. K.K.Pant
Department of Chemical Engineering
IIT Delhi.
kkpant@chemical.iitd.ac.in
 Metal Dispersion
• Metal Dispersion, D = No of Pt surface atoms / No of Total Pt
atoms
N total= from chemical composition
N surface is obtained by physical or chemical methods
Physical methods: Crystallite size from XRD, SEM/TEM
Chemical methods: Chemisorption of H2, CO, H2-O2 titration

PM distribution Profiles
a.Uniform
b.Egg shell
c.Egg white
d.Egg yolk

2
 PM distribution profiles
• Optimal dispersion depends on
– reaction kinetics and mode of catalyst
poisoning
– Attrition strength of catalyst
Egg shell favors:
– Reactions with positive order
• Fast reactions
-
Egg Yolk favors :Reactions with negative
order
Pore mouth poisoning : egg white or egg
yolk
Low attrition strength :egg white or egg yolk
3
 Factors affecting dispersion of PM
1. Concentration of PM

a. Low concentration – high dispersion

2. Presence of competing ions in impregnating solution

increases D.

a. Citric acid in H2PtCl6 impregnation on Al2O3 platforming)

4
 Factors influencing dispersion of PM
3. Functional groups on substrate surface for binding the PM
precursor – Point of zero charge (PZC) influences dispersion of
PM
Anions and neutral complexes disperse better on gamma Al2O3 at
pH<8

PZC gamma
alumina=8-9;
SiO2~3

5
Typical :Preparation of catalysts by Co-precipitation

6
7
 Physical and Mechanical Properties:
Definitions and Specifications
Property Definition/Specification

density
bulk density,  mass per unit vol. of bulk catalyst, i.e., bed, packed density
particle density,  mass per unit vol. of pellet; also called apparent density
solid density,  mass per unit vol. of solid; also called skeletal or true density

pore volume, Vpore

macropore volume volume of macropores per unit mass (dpore > 50 nm)
mesopore volume volume of mesopores per unit mass (dpore of 3-50 nm)
micropore volume volume of micropores per unit mass (dpore < 3 nm)

pore size and size distr.

macropores average size (diam.) and distribution for dpore > 50 nm
mesopores average size (diam.) and distribution for dpore of 3-50 nm
micropores average size (diam.) and distribution for dpore < 3 nm

surface area, Sint

meso and macropore SA SA of pores with diameters of 3-5,000 nm
micropore surface area SA of pores with diameters of less than 3 nm

catalyst particle size diameter and/or length of pellets/extrudates or hole size (pitch)
of monoliths

crushing strength
particle force necessary to crush particle in axial or radial direction
bulk displacement or percentage of fines versus hydraulic pressure

attrition percentage loss per time due to tumbling

9
 Determination of Surface Area

• Physisorb an inert gas such as argon or nitrogen and

determine how many molecules are needed to form a
complete monolayer

• For example, the N2 molecule occupies 0.162 nm2 at 77

K, the total surface area follows directly

• Although this sounds straightforward, in practice

molecules may adsorb beyond the monolayer to form
multilayers.
 Void Volume and Solid Density
• The void (pore) volume of a catalyst particle is
estimated by filling the pores with a liquid of
known density, followed by weighing of the
particle.

• The weight difference (before/after filling) is

due to liquid contained in the pores.

• Standard system is helium-mercury.

 Void Volume and Solid Density
• Finding the void fraction: VHg displaced  VHe displaced
Vg 
m catalyst sample
pore volume of particle
p 
total volume of particle m catalyst sample
m p Vg s 
 VHe displaced
m p Vg  m p 1 s 
Vgs  p  porosity

Vgs  1 m p  mass of particle
 Vgp Vg  void volume per gram particles
s  density of solid phase
p  density of porous particle
Physisorption

Different Adsorbates Used in Physisorption Studies

Adsorbate Boiling Point (K) Am (nm2/molecule)

N2 77.3 0.162
Ar 87.4 0.142
CO2 194.5 0.17
Kr 120.8 0.152
Langmuir Adsorption Isotherm
• fractional surface coverage,  = x/xm
• x is mass of gas adsorbed
• xm is mass of gas adsorbed at saturation (the unknown, same units as x )

BpA x m Bp
A 
1 BpA x 1 Bp
which rearranges to,

1  1 m  1 m
x x Bp x
so measure x for differing p and plot 1/x vs1/p

Indian Institute of Technology Delhi 14

1. Langmuir Isotherm:
Kp N 2 v p 1 p
N     
2
1  Kp N 2 vm v Kvm vm

p/v Slope = 1/vm

15
 2. BET Isotherm:
p 1 (c  1) p
  P0 = vapor pressure / Satn pressure
v( p0  p) vm c cvm p0

p/[v(p0-p)] Slope = (c-1)/cvm

1/cvm

p/p0

vm = 1/(slope + Intercept)

16
• Convert vm to no. of molecules
•  = area covered by one molecule
2/3
v N   M 
S   m 0    1.09 
 22400   N0 

• For Nitrogen:
 = 0.808 g/cc at -195.8 0C
 = 16.2x10-16 cm2 = 16.2 (A0)2

S  4.35  10 4 vm vm is in CC at STP

Specific Surface area = S/W cm2/gm

17
• If α = area of adsorbed molecule
M = molecular weight (same units as x)
S = surface area (same units as a)

S   x / M    N avo    
m

Note: Langmuir approach works best for

chemisorption - area is the “reactive” surface.
“Total” area can be determined using
physisorption and an applicable isotherm (BET
eq.)
Indian Institute of Technology Delhi 18
 Void Volume and Solid Density
• Finding the void fraction: VHg displaced  VHe displaced
Vg 
m catalyst sample
pore volume of particle
p 
total volume of particle m catalyst sample
m p Vg s 
 VHe displaced
m p Vg  m p 1 s 
Vgs  p  porosity

Vgs  1 m p  mass of particle
 Vgp Vg  void volume per gram particles
s  density of solid phase
p  density of porous particle
Experimental Methods of Estimating
Pore Volume and Diameter
Two methods are used : one is based on gas
adsorption is suitable for the estimation of
pore sizes in the range 15-200A0 .

Other is based on the volume of mercury

which can be forced under pressure into the
pores of a solid and is suitable in the pore
size range 100-105 A0.

20
 Pore-Volume Distribution
• Mercury-penetration method
– Hg porosimetry : based on the intrusion of mercury into a
porous structure under stringently controlled pressures.
– From pressure versus intrusion data, volume and size
distributions may be generated.
– Hg does not wet most substances and will not spontaneously
penetrate pores by capillary action, it must be forced into
the pores by the application of external pressure.
– The required pressure is inversely proportional to the size of
the pores, only slight pressure being required to intrude Hg
into large macropores, whereas much greater pressures are
required to force Hg into micropores.
 Pore-Volume Distribution
• Equate force of surface tension with the force
of the applied pressure
Fp  F
a p  2a cos 
2
(1)

 2 cos 
a   140°
p
8.75  105
a Å  
 
Hg
p lb in 2 (3)

σ is surface tension of Hg = 0.425 N/m

 Pore-Volume Distribution

• Micromeritics
AutoPore IV 9500

– Able to measure pore diameters

from 0.003 to 360 µm

– Can achieve pressures to 33,000

psi or 60,000 psi
 Gas Adsorption Method:
The gas adsorption method of estimating pore
volume and diameter : based upon the fact
that gas condenses to liquid in narrow pores
at pressure less than the saturated vapour
pressure of the adsorbate .
(Smaller pores fill first, vaporisation will take
place from larger pores first)
24
Equation relating the lowering of the vapour
pressure above a cylindrical column of liquid
contained in a capillary to the radius, r of the
capillary :
=> By equating the work done in enlarging a
spherical drop of liquid to the work done in
adding molecules to the interior of the drop.

25
 Gas adsorption method

The work done in enlarging the surface area of a drop of

liquid is = σ ΔS , (4)
where σ is the surface tension and ΔS the change in
surface area (equivalent to 8πrΔr for a spherical drop of
original radius r).
•The work done in adding molecules to the drop interior
=( µ0 - µ) Δn , (5)
where µ0 is the Chemical Potential of the vapour over a plane
liquid surface , µ the corresponding potential over the curved
surface and Δn the increase in the number of moles of liquid26.
 Thus

 
8 πr Δr = μ - μ Δn ...................... 6 
0

=>The increase in volume of the drop is given by

V Δn= 4 π r 2 Δr .....................  7 
where V is the molar volume of the liquid .
From eqs (6) and (7),
2 V
μo -μ = r ..................8 
Using the standard thermodynamic relations,

27
 μ = μ0 + RT ln Po
0
μ = μ0 + RT ln P
Where Po is the standard vapour pressure and
µ0 is the standard chemical potential at unit
Pressure. Substitution gives the Kelvin equation.

Po 2 V
In = 9
P rRT
Which may be applied to the liquid inside a
capillary of circular cross-section .
If the angle of wetting between solid and liquid is θ
then the component of the surface tension is ϭ cos θ
and the right hand side of Eq(9) is modified by the
factor cos θ .
28
 Kelvin Equation
r (pore radius) = t + 2VCos  /(RT (ln p/p0))

Pore Size Distribution

t Adsorbed layer
dm dp
Cylindrical pore

Ink-bottle pore

Pore with shape of interstice

between close-packed particles
 Total pore volume and PSD
Po 2 V
In = 9
P rkRT

2 V
r =
P RTln Po
+rk 10 
P
1
 5  3
t = 4.3   11
 ln Po 
 P
Adsorbed layer thickness (A0)

Kelvin radius rk=(rp-t ) (A0)= 9.52 (ln P0/P) -1

30
 P 2σVcosθ
ln( =) -
P rRT
0

For Liq. N2 , σ = 8.85 dyne/cm, V= 35cm3/mol,

(0.00156 cm3 liq/cm3 vapor at NTP)
By relating the relative pressure and the pore radius
the pore size distribution of the catalyst is determined
for pore size below 20nm.
31
 Kelvin Equation
P 2σVcosθ
t-Method ln( =) -
P rRT
 BET 0
– only valid in small pressure interval
– interpretation not very easy
 thickness (t) of adsorbed layer can be calculated

0.354 nm

 plot of t versus p for non-porous materials is the same (has been

checked experimentally)
 t-plot helps in interpretation
Dynamic method for estimating Surface Area
(N2 adsorbed, He does not adsorbed).
Thermostat thermal conductivity cell
Flow Control
Wave
N2 To soap flim
Flow meter

Rotameter

Sample Tube

Cold Trap Liquid N2 bath

33
34
Type A hysteresis loop and
possible pore structures

35
 BET Isotherm
 Modification of Langmuir isotherm
 Both monolayer and multilayer adsorption
 Layers of adsorbed molecules divided in:
– First layer with heat of adsorption Had,1
– Second and subsequent layers with Had,2 = Hcond

p 1 C 1 p  H  H cond 
   0 C  exp  ad 
BET isotherm:
0

nad p  p nmC nmC p  RT 

 BET equation does not fit entire adsorption isotherm

– different mechanisms play a role at low and at high p
The assumptions of BET isotherm are:

•Gas adsorbs on a flat, uniform surface of the solid with a

uniform heat of adsorption due to vander Waals forces
between the gas and the solid.

•There is no lateral interaction between the adsorbed

molecules.

•After the surface has become partially covered by adsorbed

gas molecules, additional gas can adsorb either on the
remaining free surface or on top of the already adsorbed
layer.

•The adsorption of the second and subsequent layers occurs

with a heat of adsorption equal to the heat of liquefaction of
the gas
37
 To derive the BET isotherm the surface is divided into regions covered
by i adsorbate layers; each region is characterized by a fractional
coverage θi.

• divide the surface into areas that are uncovered

(fraction Ѳ0), covered by a single monolayer (Ѳ 1), two
monolayers (Ѳ 2), or by i layers (Ѳ i).
Ѳ0, Ѳ1, ..., Ѳn = Surface area covered (/cm2) by 0, 1, ...,
n layers of adsorbed molecules

At Equilibrium Ѳ0, Ѳ1, Ѳ2 must remain constant =>

On bare surface
Rate of Evaporation from First Layer = Rate
of Condensation onto Bare Surface
K-1Ѳ1 = k0 P Ѳ0 (1)

&
39
For I Layer:
Rate of Condensation on the Bare Surface +
Rate of Evaporation from the second layer
= Rate of Condensation on the 1st Layer +
Rate of Evaporation from the First Layer,
k0PѲ0 + k-2Ѳ2 = k1PѲ1 + k-1Ѳ1 (2)

 From 1 and 2 , k-2Ѳ2 = k1PѲ1

 In general k-iѲi = ki-1PѲi-1 and k2=k3=k4---

40
BET Isotherm
 BET Isotherm
At equilibrium we have Ist Layer K-1Ѳ1 = k1 P Ѳ0
II layer, k1PѲ0 + k-2Ѳ2 = k2PѲ1 + k-1Ѳ1 ==> k-2Ѳ2 = k2PѲ1

(2)
In general k-iѲi = ki P Ѳi-1  Ѳi / Ѳi-1 = ki/k-i P

• Where F is the incoming flux per site F = P/[No√(2πmkBT)], and Eid is the desorption
energy from layer i.

• Note that, due to the change in substrate from the first to the second layer, there
may be a difference between E1d and E2d.

• However, for i = 1 and higher we consider desorption essentially as sublimation

from a multilayer of gas, and hence E2d = Ei>2d and k1≠ k2 = k3 = ... = k∞.
• Suppose there are N0 sites on the surface,
then the number of atoms adsorbed is Na
∞
Na = N0 ∑ i θi (1)
i=0
∞
• Where we have the usual sum rule ∑ θi=1
i=0

• Assumming that this surface at

temperature T is in equilibrium with a gas
then the adsorption rate equals the
desorption rate.
At equilibrium
 Eid  Eid1
dθi
 Fθi1  e θi  Fθi  e
kB T kB T
θi1
dt
 Fθi1  ki θi  Fθi  ki1θi1  0
(2)
F is the incoming flux per site
F = P/[No√(2πmkBT)], and Eid is the desorption
energy for layer i.

For i = 1 and higher we consider desorption

essentially as sublimation from a multilayer of
gas, and hence E2d = Ei>2d and k1≠ k2 = k3 = ...
= k∞.
• Writing the rate equations for the adsorption
desorption equilibrium for each layer, we
obtain

dθ0 F
= 0 =- Fθ0 +k1θ1 ⇒ θ1 = θ0
dt k1
dθ1 F 2 θ0
=0=Fθ0 - k1θ1 +k 2θ2 - Fθ1 ⇒ θ2 = θ1 =F
dt k1 k1k 2
dθi F θ0
=0= Fθi-1 -k iθi +k i+1θi+1 -Fθi ⇒ θi+1 = θi =Fi+1

k1  k 2 
i
dt ki
(3)
 • The total coverage, θt, is
F
i i
Na ∞ ∞
 F  k2 k2 ∞
 F  k2 k
θt = =  iθi =  i   θ0 = θ0  i   = θ0 2
No i=1 i=1  k 2  k1 k1 i=1  k 2  k1 1- F
k2

(4)
• where we have used
∞
x
 ix i
=
1- x 
2
i=1 (5)
• The total coverage is expressed as a
function of rate constants and the incoming
flux, Now to eliminate θ0,
 Total surface area of the catalyst
i
∞ ∞
k2
∞
 F 
 θi =1= θ0 +  θi = θ0 + 
k
θ0 
k

i=0 i=1 i=1 1  2 
  F  
 i
    
k ∞
 F  k
1+ k 2  2  
= θ0 1+ 2     = θ
 k i=1  k  
0
 k1  F 
 1 2   1-  

  k2  

(6)
• Using,
∞ ∞
1 x

i=1
x =  x -1=
i

i=0
i

1- x
-1=
1- x
(7)
=> θ0 becomes,
-1
  F  
   
 k2  k2  
θ0 = 1+ -1
 k1  F 
 1-  

  k2  
 (8)
• Substitution in the Eq. (4) for θt leads to
k2 F p
C
k1 k 2 p0
θt = =
 F k2 F   F   p p  p 
 1- +   1-   1- + C   1- 
 k 2 k1 k 2   k 2   p 0 p0   p0 
• (9)
• where C = k2/k1 is the ratio of desorption
rate constant for the second and first layer
and,
p F p
= = X (10)
p0 k 2 k 2N0 2πmk BT
 p F p
= = X
p0 k 2 k 2N0 2πmkBT

• X , is the ratio of adsorption and desorption rate from the

second layer and higher

• For (P/P0)→1 the total coverage θt→∞ meaning that the

adsorbate condenses into multilayer on the surface.

• At saturation pressure of gas an infinite numbers of

adsorbate layers builds up on surface.
  pv a 
Na  
 RT  = va
θt = =
N0  pv m  vm
 
 RT  (11)
• By substituting this in the above expression
for θt and by rearranging the equation we
find

p 1  c -1 p P
= + ≡I +S
vm p0 -p  cvm cvm p0 P0 (12)
  pv a 
Na  
 RT  = va
θt = =
N0  pv m  vm (11)
 
 RT 

• By substituting this in the above

expression for θt and by rearranging the
equation we find

p 1  c -1 p P
= + ≡I +S (12)
vm p0 -p  cvm cvm p0 P0
• Plotting P/[v(P0 – P)] versus P/P0 yields a
straight line with slope S = (c – 1)/ c vm,
crossing the y-axis at I= 1/ c vm
• The volume adsorbed in the first monolayer
is found as vm = 1/(S+ I)
• The volume vm can be converted into the
number of molecules adsorbed by N0 =
Pvm/RT and if area each molecule occupies
( α) then the total area, Sg= N0 α ,