CRYSTALLIZATION

Prof. V V Basava Rao

Department of Chemical Engineering
University College of Technology(A)
Osmania University, Hyderabad-500 007
MECHANISM OF CRYSTALLIZATION
SUPERSATURATION Non equilibrium

NUCLEATION

CRYSTAL GROWTH

CRYSTAL Equilibrium
 How dos nucleation occur?

NUCLEATION

SPONTANEOUS ARTIFICIAL

• Agitation
• Mechanical shock
• Friction
• Electric/magnetic field
• Sonic/ultrasonic irradiation
• Etc.
Metastable zone width for KCl-water system
 HOMOGENEOUS NUCLEATION
The classical theory of nucleation assumes that clusters are
formed in solution by an addition mechanism:

A+A  A2
A2 + A  A3
A3 + A  A4
...........
An-1 + A  An (critical cluster)
that continues until a critical size is reached
(n = tens to thousands)
CLASSICAL THEORY OF NECLEATION

The overall excess free energy, G, between a small solid particle
of solute (assumed to be a sphere of radius r) and the solute in
the solution is defined as:

G  G S  GV (1)

where GS : surface excess free energy, i.e. the excess

free energy between the surface of the particles
and the bulk of the particle)
GV : volume excess free energy, i.e. the excess
free energy between a very large particle, r =
, and the solute in solution.
GS is a positive quantity and is proportional to r2. In a
supersaturated solution GS is a negative quantity and is
proportional to r3.
4 3
G  4r   r Gv
2
(2)
3

where Gv : free energy change of the transformation

per unit volume (a negative quantity)
 : surface energy, i.e. interfacial tension
between the developing crystalline surface and
the supersaturated solution in which it is
located
The two terms on the right-hand side of eq. (2) are of opposite
sign and depend differently on r, so G passes through a
maximum.
This maximum value, Gcrit, corresponds to the critical nucleus,
rc. For spherical cluster, it is obtained my maximizing eq. (2),
setting d(G)/dr = 0:
d  G 
 8r  4 r 2 Gv  0
dr
 2
rc  (3)
Gv
Where Gv is a negative quantity. From eqs. (2) and (3):

16 3 4  rc2 (4)

Gcrit  
3   Gv 
2
3
Particles smaller than rc will dissolve, or evaporate if the
particle is a liquid in a supersaturated vapor, because only in
this way can the particle achieve a reduction in its free energy.
Similarly, particles larger than rc will continue to grow.
Free energy diagram for nucleation
The rate of nucleation, B0, e.g. the number of nuclei formed
per unit time per unit volume, can be expressed in the form of
the Arrhenius type of expression:
 G 
B0  A exp    (5)
 kT 
where k : Boltzmann constant = 1.3805  10-23 J K-1
A : pre-exponential factor
(theoretical value = 1030 nuclei/cm3 s)

The basic Gibbs-Thomson relationship for a non-electrolyte:

2 v
ln S  (6)
kTr

where S : supersaturation ratio = c/c*

v : molecular volume
 2 v
From eq. (6): r 
kT ln S
Introducing the above equation into eq. (3) gives:

2 2 kT ln S kT ln S (7)
  Gv   
rc 2 v v

Introducing eq. (7) into eq. (4) gives:

16 3 16  3 v 2
Gcrit   (8)
3   Gv  3  kT ln S 
2 2

And from eq. (5) gives:

 16  3 v 2 
B0  A exp   3 3 2
(9)
 3 k T  ln S  
Effect of supersaturation on the nucleation rate
 Induction Time for Nucleation of Water Vapor

S Induction time
1.0 
2.0 1062 years
3.0 103 years
4.0 0.1 seconds
5.0 10-13 seconds

In the case of nucleation of water vapor, a ‘critical’ super-

saturation could be said to exist in the region of S ~ 4.0, but it is
also clear that nucleation would have occurred at any value of S
> 1 if sufficient time had been allowed to elapse.
• The rate equation predicts exponential growth once a critical
supersaturation is attained.
• In practice, an optimal temperature exists below which the
liquid is too viscous to nucleate and above which molecular
motions prevent crystal formation.
• The viscous effects can be incorporated into the rate
equation by taking into account the viscous free energy

  16  3 v 2 Gvisc 
B0  A exp  3 3   (10)
 3 k T  ln S  2
kT 

• This was observed by Tamman (1925) for several organic

salts, and Mullin and Leci (1969) for the spontaneous
nucleation of citric acid solutions and is shown in the
following figure.
 Spontaneous nucleation in supercooled citric acid solution:
(A) 4.6 kg of citric acid monohydrate/kg of "free" water (T= 62 °C)
and (B) 7.0 kg/kg (T= 85 °C).
An empirical approach to the nucleation process is described by
Nielsen (1964), expressing a relationship between the induction
period, tind (the time interval between mixing two reacting
solutions and the appearance of crystals) and the initial
concentration, c, of the supersaturation solution:

t ind  k c 1 p
(11)

where k : constant
p : number of molecules in a critical nucleus
 HETEROGENEOUS NUCLEATION
• A foreign substance present in a supersaturated solution is
generally known to reduce the energy required for nucleation.
• Nucleation in a heterogeneous system generally occurs at a
lower supersaturation than a homogeneous system.
• Partial attraction is possible in a case where the foreign
substance and the crystal have almost identical atomic
arrangement.
• It was shown (Preckshot and Brown 1952) that the energy for
nucleus formation was reduced only if the difference in iso-
morphism between the crystal and the foreign particle was
<15%.
• For differences >15%, the energy requirements were similar to
that for a homogeneous system.
 SECONDARY NUCLEATION
• Secondary nucleation results from the presence of crystals in
the supersaturated solution.
• These parent crystals have a catalyzing effect on the
nucleation phenomena, and thus, nucleation occurs at a
lower supersaturation than needed for spontaneous
nucleation.
• Although several investigations of secondary nucleation exist,
the mechanisms and kinetics are poorly understood.
Strickland-Constable (1968) described several
possible mechanisms of secondary nucleation:

 ‘Initial’ breeding (crystalline dust swept off a newly

introduced seed crystal)
 ‘Needle’ breeding (the detachment of weak out-growths)
 ‘Polycrystalline’ breeding (the fragmentation of weak
polycrystalline mass)
 ‘Collision’ breeding ( a complex process resulting from the
interaction of crystals with one another or with parts of the
crystalline vessel)
FACTORS AFFECTING SECONDARY NUCLEATION

The rate of secondary nucleation is governed by three processes:

(1) the generation of secondary nuclei on or near a solid phase;
(2) removal of the clusters; and
(3) growth to form a new solid phase

Several factors influence these processes:

(1) the supersaturation,
(2) the rate of cooling,
(3) the degree of agitation, and
(4) the presence of impurities.
Supersaturation

• The degree of supersaturation is the critical parameter

controlling the rate of nucleation.
• The size of the critical nucleus decreases with increasing
super-saturation, thus, the probability of the nuclei surviving
to form crystals is higher.
• In general, nucleation rates are enhanced with increasing
super-saturation. However, the nucleation exponent is found
to be lower than that for primary nucleation
Temperature
• The role of temperature in the production of secondary
nuclei is not fully understood.
• For several systems, the nucleation rate declined with
increasing temperature for a given supersaturation.
• A few contradictory results exist - Genck and Larson (1972)
found a decrease in nucleation rate with increasing
temperature for a potassium nitrate system and increasing
rates with increasing temperature for a potassium chloride
system.
• It was shown by Nyvlt (1981) and others that the nucleation
order is not sensitive to temperature variations.
 Stirring
• Stirring the solution leads to lower nucleation rates.
• However, Sikdar and Randolph (1976) found that the
nucleation rate increased with the degree of agitation for
smaller crystals of magnesium sulfate (8-10 m) — the
nucleation rates were independent of the degree of agitation
for larger crystals.
• The results of Melia and Moffit (1964a, 1964b) on the
secondary nucleation of potassium chloride are shown in the
following figure; they found that the nucleation rate increases
with supersaturation, and the degree of supercooling and
agitation
Dependence of number of secondary nuclei produced on stirrer speed
and supercooling in secondary nucleation of potassium chloride
Hardness of the contact material

• In general, it was found that a harder material is more

effective in enhancing the nucleation rates.
• For example, it was found that a polyethylene stirrer reduced
the nucleation rates by a factor of 4-10, depending on the
agitation
• Crystal hardness also affects nucleation behavior — a hard,
smooth crystal is less effective.
• Irregular crystals with some roughness are generally more
active.
Effect of agitator speed on secondary nucleation rate
for steel and plastic impellers.
Impurity

• It is well known that a small amount of impurity can

profoundly affect the nucleation rate, however, it is
impossible to predict the effect prior.
• The presence of additives can either enhance or inhibit the
solubility of a substance.
• Enhanced solubilities would lead to lower supersaturations
and lower growth rates.
• The effect of impurities is complex and unpredictable.
 NUCLEATION KINETICS
 A general theory for the prediction of nucleation rates does not
exist.
 Several correlations based on the power law model have been
found to explain most of the experimental data satisfactorily.
 The power law is given by:

B  k N C n (no/m3.s) (12)

 This form is valid if the adsorption layer mechanism is the

source of nuclei.
 The nucleation rate in this case is independent of the
suspension concentration.
 In the industrial crystallizer, most of the nuclei are generated
by contact with the crystallizer environment.
 The nucleation rate in this case is a function of the degree of
agitation, the suspension density, and the supersaturation.

B  k N' W i M Tj  C n  (13)
where W : agitation rate (rpm)
MT : suspension density (mass of crystals per volume
of solution).
 In some situations an equation that does not include the effect
of agitation is used

B  k N" M Tj  C n  (14)

In this case k N" may vary with the agitation rate.

 The kinetics for secondary nucleation can be measured either
by measuring the width of the metastable zone, the induction
time, or by counting the number of nuclei formed.
 One of the methods for the determination of nucleation rates
is by measuring the maximum possible supercooling that can
be obtained in a saturated solution when it is cooled at
different rates (metastable zone width measurement).
 The polythermal experiment (proposed by Nyvlt 1968) is
carried out in a jacketed crystallizer cooled by a circulating
water/ethylene-glycol bath accurate to ± 0.1 °C.
 The temperature can be increased or decreased at a constant
rate by a programmed controller.
 The crystallizer is fitted with an accurate thermometer ±0.1
°C to read the solution temperatures.
 A schematic of the apparatus is shown in the following figure.
Schematic diagram of apparatus for
measurement of nucleation rates.
 Approximately 200 ml of saturated solution is placed in a
crystallizer and allowed to equilibrate thermally.
 The solution is stirred at a constant rate and cooled slowly
until a number of small crystals are formed.
 The temperature of the solution is then raised at a very slow
rate until the last crystal disappears.
 This temperature is denoted as the saturation temperature, Ts.
 The solution is then heated to a temperature 1° above Ts and
maintained for 30 min.
 The solution is now cooled at a constant rate (r1) and the
temperature at which the first crystal appears is noted (T1).
 The difference between this temperature and the saturation
temperature is denoted as AT1max for the cooling rate r1.
 The experiment is repeated for two different cooling rates.
The nucleation rate at the metastable limit can be approximated as

 dC * 
B    r1 (15)
 dT 
The maximum supersaturation is given by

 dC * 
C    Tmax (16)
 dT 

Combining Eqs. (15), (16), and (12), simplifying, and taking

logarithms, yields

dC
     
log  r1  log k N m log Tmax  m  1 log (17)
dT

where m is used in place of n to signify an apparent nucleation

order.
The dependence of cooling rate on the maximum attainable
supercooling in aqueous sodium chromate solutions
 The rate of nucleation can also be determined by observing
the time elapsed between the creation of supersaturation and
the formation of a new phase.
 This time interval is defined as the induction time and is a
function of the solution temperature and supersaturation.
 The formation of a new phase can be detected in several
different ways—for example, by the appearance of crystals or
by changes in properties (turbidity, refractive index) of the
solution.
 The induction time, tind is the sum of the time needed for
reaching steady-state nucleation, ttr; the nucleation time, tn;
and the time required for the critical nucleus to grow to a
detectable size, tg.
t ind  t tr  t n  t g
(18)
 It can be shown (Sohnel and Mullin 1988) that the transient
period (ttr) is unimportant in aqueous solutions of moderate
supersaturations and viscosities.
 In certain special cases however, at very low supersaturations
the transient period cannot be ignored (Packter 1974) and the
following analysis is not applicable.
 Recently, Kubota et al. (1986) has suggested a method to take
into account the transient period.
 If the transient period can be ignored, the induction time is a
function of the nucleation and growth times. Three cases
emerge:
1. tn >> tg
2. tn ~ tg
3. tn << tg
If tn >> tg , the induction time is inversely proportional to the
steady-state nucleation rate and

 F  3 v 2 
t ind  A exp  2 (19)
  kT   ln S  
3

where  is the wetting angle that is 1 for homogeneous

nucleation and is <1 for heterogeneous nucleation, and F is the
shape factor ratio.

If tn ~ tg , the induction time for nucleation followed by diffusion

growth is given by
v2 3  F 3 v 2 
t ind  exp  2 (20)
 5 kT   ln S  
15 3
2 Dx

where x is the solute mole fraction

If tn << tg , the induction times for various cases are given by
Sohnel and Mullin (1988). For mononuclear growth, the
induction time in this case is given by

d3  F 2v 4 3 
t ind  exp  2 (21)
6 Ds rc   kT  2
 ln S  

where Ds is the surface diffusion coefficient and rc is the critical

radius of the nuclei.

In general, the induction time is given by the expression

t ind  KS 0 n (22)
Induction period as a function of supersaturation for CaCO3
precipitation at 25 °C showing regions of homogeneous
and heterogeneous nucleation.
 Example 1
Yield
Yield of
of aa Crystallization
Crystallization Process
Process
AA salt
salt solution
solution weighing
weighing 10000
10000 kg kg with
with 30
30 wt
wt % % Na
Na22CO
CO33 isis cooled
cooled
to
to 293
293 KK (20
(20 °C).
°C). The
The salt
salt crystallizes
crystallizes as
as the
the decahydrate.
decahydrate. What What
will
will be
be the
the yield
yield of
of Na
Na22CO
CO33•10H
•10H22O O crystals
crystals ifif the
the solubility
solubility isis 21.5
21.5
kg
kg anhydrous
anhydrous NaNa22CO
CO33/100
/100 kgkg of
of total
total water?
water? Do Do this
this for
for the
the
following
following cases:
cases:

(a)
(a) Assume
Assume that
that no
no water
water isis evaporated.
evaporated.
(b)
(b) Assume
Assume that
that 3%
3% of
of the
the total
total weight
weight of
of the
the solution
solution isis lost
lost by
by
evaporation
evaporation of
of water
water inin cooling.
cooling.

FIGURE3.
FIGURE3. Process
Process flow
flow
for
for crystallization
crystallization
 Example 1
Solution
The molecular weights are 106.0 for Na2CO3, 180.2 for
10H20, and 286.2 for Na2CO3 •10H2O. The process
flow diagram is shown in Fig. 3, with W being kg H2O
evaporated, S kg solution (mother liquor), and C kg
crystals of Na2CO3 •10H2O.Making a material balance
around the dashed line box for water for part
(a), where W = 0.

(13)

where (180.2)/(286.2) is wt fraction of water in the crystals.

 Example 1
Making a balance for Na2CO3,
(14)

Solving the two equations simultaneously, C = 6370 kg of Na2CO3

•10H2O crystals and S = 3630 kg solution.
For part (b), W = 0.03(10000) = 300 kg H2O. Equation (13)
becomes

(15)
Equation (14) does not change, since no salt is in the W stream.
Solving Eqs. (14) and (15)simultaneously, C = 6630 kg of Na2CO3
•10H2O crystals and S = 3070 kg solution.
 Yields and Heat and Material Balances in
Crystallization
Heat effects and heat balances in crystallization
• When a compound whose solubility increases as temperature
increases dissolves, there is an absorption of heat, called the heat of
solution – occurs when the solubility decreases as the temperature
increases
• At equilibrium the heat of crystallization is equal to the negative of
the heat of solution at the same concentration in solution.
• The enthalpy H1 of the entering solution at the initial temperature is
read off the chart, where H1 is kJ for the total feed. The enthalpy H2 of
the final mixture of crystals and mother liquor at the final
temperature is also read off. If some evaporation occurs, the enthalpy
Hv of the water vapor is obtained from the steam tables. Then the
total heat absorbed q in kJ is

(16)
• If q is positive, heat must be added to the system. If it is negative,
heat is evolved or given off.
 Example 2
Heat Balance in Crystallization
A feed solution of 2268 kg at 327.6 K (54.4 °C)
containing 48.2 kg MgSO4/100 kg total water is cooled to
293.2 K (20°C), where MgSO4•7H2O crystals are
removed. The solubility of the salt is 35.5 kg MgSO4/100
kg total water. The average heat capacity of the
feed solution can be assumed as 2.93 kJ/kg• K.
The heat of solution at 291.2 K (18 °C) is -13.31 x 103
kJ/kg mol MgSO4•7H2O. Calculate the yield of
crystals and make a heat balance to determine the total
heat absorbed, q, assuming that no water is vaporized.
 Example 2
Solution
Making a water balance and a balance for MgSO4
using equations similar to (13) and (14) in
Example 1, C = 616.9 kg MgSO4•7H2O crystals
and S = 1651.1 kg solution.
To make a heat balance, a datum of 293.2 K
(20°C) will be used. The molecular weight
of MgSO4•7H20 is 246.49. The enthalpy of the
feed is H1:
 Example 2
The heat of solution is -(13.31x103)/246.49 = -54.0 kJ/kg crystals.
Then the heat of crystallization is -(-54.0) = +54.0 kJ/kg crystals,
or 54.0(616.9) = 33312 kJ. This assumes that the value at 291.2 K
is the same as at 293.2 K. The total heat absorbed, q,
is

Since q is negative, heat is given off and must be removed.

 Equipment for Crystallization
Tank Crystallization
• Hot saturated solutions are allowed to cool in open tanks
• After a period of time, the mother liquor is drained and the crystals
removed
• Nucleation and the size of crystals are difficult to control
• Labor cost are very high
• Has limited application; used to produce certain fine chemical and
pharmaceutical products.
 Equipment for Crystallization
Scraped surface crystallizers
• One type of scraped surface
crystallizer is the Swenson-Walker
crystallizer, which consists of an
open trough 0.6 m wide with a
semicircular bottom having a cooling
jacket inside.
• Slow-speed spiral agitator rotates
and suspends the growing crystals
on turning.
• Blades pass close to the wall and
break off any deposits of crystals on
the cooled wall.
• Used in crystallizing ice cream and
plasticizing margarine.
 Equipment for Crystallization
Circulating-liquid evaporated-crstallizer
• Supersaturation is generated by evaporation.
• Circulating liquid is drawn by the screw pump down
inside the tube side of condensing steam heater
• Heated liquid then flows into the vapor space,
where flash evaporation occurs, giving some
supersaturation.
• The vapor leaving is condensed.
• The supersaturated liquid flow down the downflow
tube and then up through the bed fluidized and
agitated crystals, which are growing in size.
• The living saturated liquid then goes back as a
recycle stream to the heater, where it is joined by
the entering feed.
• The larger crystals settle out and a slurry of crystals
and mother liquor is withdrawn as product.
• Also called Oslo crystallizer.
 Equipment for Crystallization
Circulating-magma vacuum
crystallizer
• The magma or suspension of crystals is
circulated out the main body through a
circulating pipe by a screw pump
• Magma flows through a heater, where its
temperature is raised 2-6 K.
• The heated liquor then mixes with body
slurry and boiling occurs at liquid surface
• This cause supersaturation in the swirling
liquid near the surface, which results in
deposits on the swirling suspended crystals
until they leave again via the circulating pipe
• The vapors leave through the top
• A steam-jet ejector provides the vacuum
 n k Yo u
Th a