5.

Determination of Rate Equations from the Experimental

Data
The chemical engineer needs rate equation
for
designing of reactors
boosting the performance of a
reactor.
In previous chapters, we have discussed that
determination of the experiments
This leads to important rate equations
to becan only
carried onbe
found
1) to through
find the experimental
concentration data.
dependency of the
substance on time / residence time / at a fixed
temperature for determining the order of the reaction
2 to find the temperature dependency for
evaluating the rate constant to yield the complete
rate equation for some reaction parameters.
The experimental methods used in kinetics may be
classified into two categories,
i) Chemical

analysis
determine directly the concentration of one of
the
reacting substances or products against time
(or more specifically of residence time)
 it is difficult to determine the concentration C
at a definite time by any simple technique.
This is due to that the reaction cannot be
turned on and off like a stopcock.
ii) Physical Analysis
 the physical change during the course of
reaction that does not require removal
successive samples from the reaction mixture
is continuously measured.
 Theoretically, any property that undergoes
sufficient change may be used to follow the
  The advantages of application of physical
methods, as a rule are their measurement
rapidity.
 They also make it possible to obtain a large
number
 The of experimental
experimental data can
measurement in a given
often be time
interval. directly in reaction vessel itself. The
performed
vessel is distinct in two types of process, namely
batch and flow reactors.
Among physical methods the following may be
mentioned:
• Dilatometric methods:- based on the change
in volume due to reaction
•Optical methods:- such as polarimetry, refractometry,
colorimetric
•Electrical methods:- such as conductimetry,
potentiometry, polarography, and mass spectroscopy.
As a criteria to implement a physical method is that one
has to know the linear relationship between
physical properties and concentration,
i.e, concentration on the one hand and the electrical
conductivity, optical density, rotation of the plans of
polarization and the pressure of gaseous, on the other.
The most difficult part in kinetic is to find the rate after
having the experimental data.
Raw Data obtained from the batch or flow reactor
experiments are analyzed in one of the following
ways:
The differential method;
The integral method ;
The initial rate method;
The half-life method;
The least-square method
For our study and application purposes, we will limit the
above methods analyses in batch reactor systems.
 It is possible to carry
out only one
The differential experiment to find the
method order of reaction and
The integral method the rate constant
with respect to one of
the reactants
Require experiment
The initial rate method at many different
The half-life method initial conditions to
determine order of
reaction and the
The least-square rate
method
Used when three or more parameters
are involved such as α k, E, A etc.
For our study and application purposes, we will limit the
above methods analyses in batch reactor systems.
5.1 Differential Methods
In a differential method, a point rate is
measured while a small or “differential”
amount of reaction occurs, during which the
relevant reaction parameters (Ci, T, etc.)
change very little, and can be considered
constant.
 In this method, it requires differentiation of
the experimental data to obtain rate of
reaction

 Restriction for application for such

method is that it should only be
applied for irreversible reactions with
a single reactant
Differential Methods using BR
One reactant
Consider the decomposition of the following single
reactant
A   P (B)
The kinetic expression for such type of reaction is
given by 
r  r  k C 
A A A
Differential equation becomes
dC A
 k C
A
dt
 dC A 
ln   ln k   ln C A (4.2)
 dt 
we test the fitness of the experimental data
using equation (4.2), to find k and α
Two reactants
If there were two reactants A and B in reaction (A), and
the rate law were of the form


( r ) kC C 
A 4.3
A B
Here we have to obtain k, α, β using the
differentiation procedure
The procedure is similar to that for one reactant,
although there is an additional constant to
determine (β).
From equation (4.3) we have the differential equation

dC
ln( A ) ln k  lnC   lnC 4.3.1
dt A B
Like equation 4.2, this is a linear relation, although in
three-dimensional ln(-dCA /dt), - ln(-CA), and ln(-CB)
space. It is also linear with respect to the constants lnk,
α, β and hence their values can be obtained by linear
regression from an experiment which measures CA as a
function of t. Values of (-rA) can be generated from
these as a function of CA by differentiation, as
described above for the case of a single reactant. The
concentrations CA and CB are not independent but are
C  C stoichiometry
linked by the reaction C C
A A,0  B B,0
 
A B
Values of CB can thus be calculated from measured
values of CA. Alternatively, k, α, and β can be
obtained directly from equation (4.3.1) by nonlinear
Generally, in the case of many reactants involved in
the reaction, we can use the so called method of
excess
As an example

A + B P

 
r  k C A C B
A (4.3.1)

analyzed by applying differential method

incorporating isolation method i.e.
reaction runs excess in B
i.e concentration of B remains essentially
unchanged during the course of reaction compared
to concentration of A then equation (4.2.1) becomes
 dC A
 k ' C
A
dt
 k C  and k '
k ' kCB is known as pseudo
B ,0 reaction rate constant

 dC A 
ln   ln k ' ln C A
. dt 


ere, it is possible to determine k’ and directly.

After determining 
we can further process the reaction

with excess of concentration of A to determine
Procedure to carry out differential method:
• Measure concentration versus time :
data
•Arrange the differential equation in a
natural logarithmic form dC A 
Ex ln   ln k  ln C A
dt  
• Differentiate the CA versus t data either
numerically or graphically.
• Plot the natural logarithmic form of the rate
equation dC A 

ln 
• Observe a straight line graph of 
versus ln C A  dt 
•Obtain the slope of the graph
 dC A 
ln 
It is to be noted that for the first-order case, 
 dt 
versus ln C A should yield a straight line with a slope of 1.
  dC A 
ln  
 dt 
Slope=order of reaction

ln k

lnC A
  dC A can be determined from Differentiate the CA
( )
dt versus t data using
Graphical and
Numerical methods including
polynomial,
finite difference,
 non-linear least square
analysis etc.
Differential Methods using PFR
As in the case of a BR, a PFR can be operated in
both a differential way to obtain kinetics data

Consider steady-state operation for a system reacting

according to
A Products

The system is not necessarily of constant density,

and to emphasize this, we write the material balance for
calculating (-rA) in the form

dX
( r )  F A (4.5)
A A,0 dV
Depending on the method of analysis for species A, XA
may be calculated from CA, together with the flow rates,
FV and FA, by equations
F F
X  A,0 A
A F
A,0
F F F
 F  A A,0 C  A
A F
 
A A V, f

By varying CA,0 at the inlet, and/or by varying flow

rate, in a series of experiments, each at steady-state
at the same T; one can measure (-rA) as a function of
CA at the given T to obtain values of k and α in the
rate law, in the same manner as described for a BR.
Differential Methods using CSTR
No need to differentiate CA data
Consider a constant-density reaction with one reactant
A Products
One experiment at steady-state generates one
point value of (-rA) for the conditions (CA,0 ,FV,0, T)
chosen. This value is given by the material balance
obtained
C C
( r )  A,0 A (C  C ) F V
A t A,0 A V ,0

To determine the form of the rate law, values of (-

rA) as a function of CA may be obtained from a
series of such experiments operated at various
conditions. For a given reactor (V) operated at a given T
conditions are changed by varying either C or F .
For a rate law given by (-rA) = kCnA , the
parameter-estimation procedure is the
same as that in the differential method for a
BR in the use of equation 4.2.1 (linearized
form of the rate law) to determine k and n.
The use of a CSTR generates point (-rA) data
directly without the need to differentiate CA
data (unlike the differential method with a
BR)
If there is more than one reactant, with a rate law given
by

( r ) kC C 
A A B

the procedure to determine (-rA) is similar to that

for one reactant, and the kinetics parameters are
obtained by use of equation 4.3.1, the linearized form of
5.2. Integral methods
The integral method follows a trial and error
approach to find reaction order
General procedure
generate values of CA as a function of t,

• Guess the reaction order and integrate

the mole balance equation
•Calculate the resulting function for the data and
plot this against function of time.
•If the resulting plot is linear, the reaction order is
probably guessed correctly.
•If the plot is not linear, guess another order and
repeat the procedure
The integral method is mostly used for known order
of reaction and it is best used to evaluate the rate
constant at different temperatures in order to
determine the activation energy.
Integral methods analysis using BR
For a reaction involving a single reactant A with a rate law
given
( r ) kC n
A A 4.6
we obtain, using the material balance as

dC
A  kdt
n 4.6.1
C
A
Integration of this between the limits of CA,O at t = 0, and CA at t
results in

4.6.2
Equation 4.6.2 implies that a plot of CA1-n versus t is a
straight line with slope and intercept indicated in
Figure. Since such a linear relation is readily identified,
this method is commonly used to determine values of
both n and k; however, since n is unknown initially,
a value must first be assumed to calculate values of the
ordinate.
As noted in equation 4.6.2, the form given there is
not applicable to a first-order rate law, this is due
to that n = 1, equation 4.6.2 becomes indeterminate (kt = 0/0).
In such case, we return to equation 4.6, which then
integrates to
dC 4.7
- A  kdt
C
A
which then integrates to
C C e kt 4.7.1
A A,0
or linearization becomes
lnC lnC  kt 4.7.2
A A,0
a linear relation for a first-order reaction is obtained from
a plot
If the rate law involves more than one species, as

( r ) kC  C  C 
A A B C
 the same general test procedure may be used, but
the integrated result depends on the form of the
rate law.
Consider the reaction  A B   P
A B
with the rate law
r kC C
A B
From the rate law and the material-balance equation, the
equation to be integrated is

dC
- A  kdt 4.8
C C
A B
As we know, CB is not independent of CA, but is
related to through stoichiometry, to which we add
the extent of reaction to emphasize that there is only
one composition variable: CA
 C C C C
A A,0  B B,0  4.9
  V
A B
Equation 4.9 may then be used to eliminate both CA and
CB from equation 4.8, which becomesd k  dt
(C a )(C  b ) a
A,0 B,0
Where
 
a A ; b B
V V
Integration by the method of partial fractions become

4.10
 Thus, ln CA / CB is a linear function oft, with the
intercept and slope as indicated, for this form of
rate law. The slope of this line gives the value of
k, if the other quantities are known.
ntegral methods analysis using PFR
It is possible to obtain values of kinetics parameters by
means of such a reactor from the material balance
equation rearranged as
dX
V  A 4.11

F (-r )
A,0 A
If the rate law (for (-rA)) is such that the integral can be
evaluated analytically, then it is only necessary to
make measurements (of CA or XA) at the inlet or out
let.
Thus, if the rate law is given by equation 4.6, integration
of the right side of equation 4.11 results in an expression
of the form
g(XA) / k ,where g(XA) is in terms of the order n,
values of which can be assumed by trial, and k is
unknown. The left side of equation 4.11 for a given
reactor (V) can be varied by changing FA,0, and g(XA)
is a linear function of V/FA,0 with slope k, if the
correct valuelaw
If the rate of nisis such
used that the integral in equation
4.11 cannot be evaluated analytically, it is
necessary to make measurements from samples at
several points along the length of the reactor, and
use these in a numerical or graphical procedure
with equation 4.11
 Exercise
If the gas-phase reaction

A B +C
is first-order with respect to A, show how the value or the
rate constant k can be obtained from
measurements of CA (or XA) at the inlet and outlet
of a PFR ans CSTR operated isothermally at T, and
at (essentially) constant P
Solution
The rate law is
( r ) kC
A A
dX
V  A
and CA and XA are related by
 4.11
F (-r )
A,0 A
 F F (1 - X )
C  A  A,0 A
A F F
V, f V, f
If we assume ideal-gas behavior, and that only A is
present in the feed, the dependence is given
F F (1  X )
V, f V,0 A
Substitution of the above equations for (-rA) , CA, and FV,f
in equation 4.11 results in
X (1  X )
  
 

V 1 A  X 2ln(1 X ) 
A dX   A

A 
 A k
F k 0 (1  X )
V,0 A
Thus, for given V, T, and P, if FV,f is varied to obtain
several values of XA at the outlet, the expression
 X 2ln(1  X is
 
 ) a linear function of V/FV,f, with slope k


A A 


EXPERIMENTAL METHOD to DETERMINE RATE

PARAMETERS in CSTR

Consider a constant-density reaction with one reactant

A B +C
One experiment at steady-state generates one point
value of (-rA) for the conditions (CA, FV,0, T) chosen. This
value is given by the material balance obtained

C C
( r )  A,0 A (C  C ) F V
A t A,0 A V ,0
Exercise
How would the procedure described above have to be
modified if density were not constant?
Solution
If density is not constant, the volumetric flow rate inlet
and outlet, FV,0 and FV,f. respectively, are not the same.

As a consequence, (-rA), for each experiment at steady-

state conditions, is calculated from the material balance
in the form
C F C F
( r )  A,0 V ,0 A V , f
A V
Apart from this, the procedure is the same as
described above for cases of one or more than one
5.3 Initial rate methods
In this method, initial rate can be found, i.e.
a series of experiments is conducted at different
initial concentration,
 The initial rate can be found by differentiating
the data and extrapolating to zero time.

In other words, (say) the initial rates of reaction are

found by plotting the concentration-Vs-time curve and
finding the slopes at t = 0 as shown in Figure

dc
Slope = initial rate 
= dt
C

t
5.4 Method of half-
The general equation for half-life as defined
lives
t  2n 1 1
1 k C n 1(n  1) (4.12)
2 A,0

Equations 4.12 lead to the same conclusion that

t C n 1 m, cons tan t( for all n )

1 A,0
2

This may be used as a test to establish the value of

n, by trial, from a series of experiments carried out
to measure t1/2 for different values of CA,0. The value
of k can then be calculated from the value of n
obtained, from equation 4.12.
Taking the natural log of both sides, we get

lnt ln 2n 1 1(1 n)lnC (4.12.1)

1 (n  1)k A,0
2

The equation can be used in linear form (ln t1/2

versus ln CA,0) for testing similar to that described
in the previous section.
The slope of a plot as a function of
ln t
of 1 initial
2
is equal to 1 minus the
concentrati ln C A,0
reaction
on
Order n
 ln t
1
2 Slope = 1- n

ln C A,0
Example for a first-order rxn: t1/2 is
independent of CA,0. A series of experiments
carried out with different values of CA,0 would
thus all give the same value of t1/2 if the
 5.5. Least-square methods
 The method of least-squares can give the
order of the reaction with respect to all
the components at once.

 That is why this method is effectively used,

when the rate law depends on the
concentration of more than one
species, frequency factor, activation
energy etc., especially if it is not possible
to use the method of excess.

 The Least-Square method requires having a

Let us series ofthe
consider measurements
reaction, aA  bB   P
nd the rate expression is given by,
dC A a b
  r A kC AC B
dt
Taking the natural log of both sides of the equation,
we get
ln( rA ) ln k  a ln C A  b ln C B
Which is a general form of
Y c  aX1  bX 2 (a)
Where
 dC A 
Y ln  ; c ln k ; X1 ln C A ; X 2 ln C B
 dt 
we now carry out N experimental runs, for the jth run
,
- the equation can be written as,
Y j c  a X1, j  b X 2, j
For N runs, it will be 1,2 -------------- N. For which the
equations can be written as,

N (7.7) N N
 Y j Nc  a  X1 j  b  X 2 j
j 1 j 1 j 1
N N N 2 N
 X1 j Y j c  X1 j  a  X1 j  b  X1 j X 2 j
j 1 j 1 j 1 j 1

N N N N 2
 X 2 jY j c  X 2 j  a  X1 j X 2 j  b  X 2 j
j 1 j 1 j 1 j 1
These equations are three linear with three unknowns,
and can be solved for c, a and b which are rate
constants and order of reaction of A and B.