Kinetics analysis - Differential, integral, half life, Arrhenius Expression

a. Differential - if data is of good quality, very good choice. Numerical

derivative amplifies the noise. Isothermal conditions data. This analysis is
meant to be rA vs CA, but usually constant volume batch data of simple
reactions, so -dCA/dt vs CA. Can help identify what the order is.

b. Integral method, smoothens noise, not good in identifying order, once order is
known, good in getting rate constant values. Assume rate expression and
dC A
evaluate.  k1C A, or k0 , or k2C A2 ... Bit more sophisticated like enzyme kinetics
dt
k1C A
rA 
1  k2C A2

c. Method of half lives - when we measure a chemical concentration, what if the

instrument has almost zero offset, but the reading is not calibrated? If 'linearity'
is ensured, we can still measure half concentration accurately (even if the
original concentration is unknown).
e.g. degradation of a dye in wastewater.

d. Initial rate - when we have reversible reactions, initial rate measurement is

important
Integral method example. Consider the following reaction
2 NOCl --> 2 NO + Cl2 conducted under constant volume and temperature
(200 °C). The concentration is measured and is given as follows. It is believed
to be second order reaction. Determine the rate constant.
Time (s) 0 200 300 500
C(M) 0.02 0.016 0.0145 0.012

Solution: In case of second order reaction, we can relate concentration of the

 1 1 
reactant with time as follows.     kt
 C A C A0 
 1 1 
Calculate    and determine the rate constant, as shown below
 C A C A0 

Time (s) 0 200 300 500 x 1000

C(M) 0.02 0.016 0.0145 0.012
-
1/CA -1/CA0 (M 0 12.5 18.97 33.3 y 64.77
1
)
 1 1 1 1
Fit   kt to get k = 0.06477 lit/mol/s. One can also try   kt and
C A C A0 C A C A0
we get CA0 = 0.02016 mol/lit and k = 0.0660 lit/mol/s

Arrhenius method. Usually we are given data as ‘k vs. temperature’, but in

reality, the data would be of the form rate = f(T,C) . We would have segregated
data at a few fixed temperatures, obtained rate constants and then analyzed the k
vs. temperature separately. It is better to analyze all the data together (nonlinear
regression). Sometimes it is, in fact, necessary.

Second Example:- Half life method along with Arrhenius expression.

Tetrahydrofuran decomposes to form methane, ethane, propane, hydrogen, CO
etc. The initial pressure, temperature (maintained throughout a given
experiment) and the half life values are given below.

T °C 569 530 560 550 539

Initial P 214 204 280 130 206
(mm Hg)
t½ (min) 14.5 67 17.3 39 47

PA N A
Here, we can assume that ideal gas law is obeyed. Then, CA   Hence,
RT V
initial P/T is proxy for initial concentration. Note that we don't have multiple
concentration data at a fixed temperature.

Assume order is not = 1

2n 1  1 1n 
E
t1  C A0 , k  k0 e RT
k  n  1
2
.

 
ln t 1
2
 2n1  1 
 ln 
  n  1
  ln  k0  

E1
R T
 1  n  ln  CA0 

 2n 1  1 
This is y = a + bx1 + Cx2. a  ln    ln  k0  , b = E/R, c = (1-n). We can
  n  1 k 0 
use analysis in Excel or Matlab or similar software (or do by hand), and we get
a = -29.76, b = 26666, c = -0.56. Order is roughly 1.5 , R2 = 0.999. I have used
Microsoft Excel ® Data Analysis --> Regression to get these.
The confidence intervals are: We are 95% confident that ‘a’ is between -28.8
and -30.8, ‘b’ is between 25.8K and 27.5K, and ‘c’ is between -0.49 and -0.63.

For first order reaction, t1 

2
ln  2 
k
and    ln ln 2  ln k   ER T1 ,
ln t 1
2
0

independent of initial concentration. This also fits reasonably well, we have

only 5 points. y = 27690 x -30.22, R2 = 0.948.
Initial
P (mm t½ t½ ln(t½
T °C Hg) (min) T P/T (min) 1/T (min))
569 214 14.5 842 0.254157 14.5 0.001188 2.674149
530 204 67 803 0.254047 67 0.001245 4.204693
560 280 17.3 833 0.336134 17.3 0.0012 2.850707
550 130 39 823 0.157959 39 0.001215 3.663562
539 206 47 812 0.253695 47 0.001232 3.850148

ln(t½ (min))
5
y = 27690x - 30.222
4 R² = 0.948
3

0
0.00118 0.00119 0.0012 0.00121 0.00122 0.00123 0.00124 0.00125

The confidence intervals are: We are 95% confident that the slope is between
15.7 K and 39.6K , and the intercept is between -15.8 and -44.7.

On comparing these two sets of results, we can conclude that the order is indeed
1.5.