AMERICAN JOURNAL OF SCIENTIFIC AND INDUSTRIAL RESEARCH

© 2013, Science Huβ, http://www.scihub.org/AJSIR

ISSN: 2153-649X, doi:10.5251/ajsir.2013.4.3.285.287

Application of matlab ordinary differential equation function solver

(ode45) in modelling and simulation of batch reaction kinetics
1
Anyigor, C.M, and 1Afiukwa, J.N
1
Department of Industrial Chemistry, Ebonyi State University, P.M.B. 053 Abakaliki, Ebonyi
State, Nigeria. Email: chumaco_15@yahoo.com
ABSTRACT

This study presents the simulation of modelled chemical reaction in a batch reactor using the
Matlab ordinary differential equation function solver (ODE45). Modelling and optimisation of
actual batch reaction rate constants under isothermal condition was implemented on a set of
experimental data. The data was simulated using hypothetical isothermal rate constants,
1 2 with the initial concentration profile of the reaction: and
respectively between reaction time of minutes. The Process model developed was
employed to optimise the rate constant (K) at which maximum yield of the desired product was
achieved. Results showed maximum product yield, ‘D’ at 1= with a reaction by-product, ‘E’ at
2 = within 40 minutes reaction time. This study is intended to benefit industrial operations
which aim at achieving optimum yield of a reaction process in a record time.

Keywords: Reaction Kinetics, Simulation, Modelling, Matlab and Process control.

INTRODUCTION In this study, the use of process model of a

The focus of an industrialist is how to achieve hypothetical batch reaction of the type: →
maximum yield of the desired product in a reasonable ….. (1) and → ….. (2), was implemented
time frame. Chemical Kinetics is concerned with the through the application of Matlab ordinary differential
velocity or time rate of reaction and the mechanisms equation (ODE45) function solver in order to optimise
of the reaction. The rate constant determines how the rate constants (kmax) under isothermal condition.
quantitative the yield of the reaction is. Therefore Certain relevant assumptions were made and
efforts aimed at optimising the rate constant would hypothetical K-values grouped as
enhance yield maximally. Computer aided respectively were
mathematical application is a growing trend used in investigated on the hypothetical reaction equations to
modelling and simulation of reaction kinetics to find the optimum K (kmax).
achieve such a desired optimum yield (Ullah and
Wolkenhauer 2011). In many chemical processes of METHODOLOGY:
transformation of reactants into products, only the The initial conditions of the concentration species
disappearance of the reactants and appearance of used were;
the final products may be observed (Sharma and respectively and the following assumptions
Pathania 2006). But for batch reactions occurring in associated with batch reactions occurring in a reactor
steps, intermediate products may be formed which were made: (i) that the batch reactor is a
may not be detected because they are promptly used homogeneous mixture, (ii) operate isothermally under
up in the subsequent step to yield the main product. closed system, (iii) has constant volume and (iv)
The rate of reaction generally decreases with time occur at the rates, and for the hypothetical
but the coefficient of determination (rate constant) reactions 1 and 2 respectively. The material balance
remains unchanged throughout the reaction. The rate of the batch reactor was obtained from the general
depends on the active concentration of the reactants equation:
and is greatly influenced by temperature among other
factors such as the nature and concentration of the Rate of Change =
reacting species (Higham 2007).
Where for a batch reactor, ; then
 Am. J. Sci. Ind. Res., 2013, 4(3): 285-287

product of the net reaction i.e. → . Thus

The numerical differential equations of the reaction
rates were defined based on the material balance
and integrated using the ODE45 solver function of
Matlab (Howard 2007). The developed differential
models of the reaction rates for each of the reaction
parameters (A, B, C, D and E) were shown:
(3)
(4)
(5)
(6)
(7)
Where were the concentration
species, were the reactants, was the
intermediate product whereas were the final
products occurring in the batch reactor. and
were the isothermal kinetic reaction rate coefficient
for reaction 1 and 2 respectively.
Two Matlab files were created. The main file (m-file)
of Matlab code of the initial condition of concentration
species put in vector form was used to define the
initial conditions and to call up the second Matlab file
of codes (function file), ‘[t,y] = ode45(@reactor1,[0
40],y)’. The function file was employed by integrating
the ODEs of the reaction model (Appendices ).
Simulation was implemented and the results of
concentration against time were plotted. The
simulation using the ODE45 function solver of Matlab
solved the integration of the established reaction
reactant was totally consumed as indicated with its
concentration dropping to zero from unity (Fig.3).
Based on the reaction models implementation
(Appendices ) and the integrated numerical
differential equations of the reaction rates (Equations
) within limits of minutes, the results of
3
the plot of concentration (mol/dm ) of the species
involved in the reaction against time (mins) were
shown in Figs 1, 2, and 3. The batch reactor was
allowed to progress for minutes to attain a
constant concentration profile of species indicated as
legends in the figures. Maximum value of the main
3
product, ‘ ’, (0.83mol/dm ) was realised between
minutes at an assumed values of
0.5 and 0.1 respectively (Fig.3). The concentration of
3
‘D’ rose from 0.0 and levelled up at 0.83 mol/dm
from minutes. After this time limit, it is
unlikely to have more yields since one of the
reactants seem to have been totally consumed. This
value was much higher than was obtained at
tried at 1.0 and 2.0 (Fig 1) and 0.2 and 0.4
(Fig. 2) respectively. The result of Fig. 1 is
unthinkable, meaning that positive integer values of
the rate constants may not yield reasonable product.
The relationship between Figs. 2 and 3 indicated that
there should be an optimum K-value at which
maximum yield of the desired product can be
achieved. This optimum rate constant was attained at
referring to equation 1. The
becomes useful if the side reaction giving the
by- product ‘ ’ was desired. Negative values of K at -
1 and -2 were also evaluated but the result was
negative implying that rate constant cannot be
negative.
1.2
ConcA
1 ConcB
ConcC
CONCENTRATION (mol/dm3)

0.8 ConcD
models of the batch reactor system from initial time to ConcE
the final time conditions of minutes. The 0.6
integral time limits adopted in this work was realised
by trial and error until a constant concentration profile 0.4

was obtained 0.2

RESULTS AND DISCUSSION 0

Application of Matlab ODE45 solver function on -0.2

hypothetical bimolecular reaction kinetics to optimise 0 5 10 15 20 25 30 35 40
TIME (minutes)
the rate constants was investigated under isothermal Fig. 1: Concentration Profile of the Batch Reactor after
condition. In the hypothetical bimolecular reaction: reaction time of 40 minutes with rate constant values
→ ; served as an intermediate product of and 2.0.
which instantly reacted with to yield as a by-

286
 Am. J. Sci. Ind. Res., 2013, 4(3): 285-287

optimised time frame of minutes under

1
ConcA constant temperature. The order of the reaction can
ConcB
ConcC
be estimated from the reaction rate constant equation
0.8 which gave the Kmax.
ConcD
CONCENTRATION (mol/dm3)

ConcE
RECOMMENDATION
0.6
The Matlab ODE45 function solver can be tried on
0.4 complex reactions such as reversible ;
consecutive ( → → and parallel ( → → ).
0.2 It is also recommended for chemical, biological and
biochemical reaction systems for batch reactor
0 system analysis and product yield maximization.
0 5 10 15 20 25 30 35 40
TIME (minutes)
Acknowledgement
Fig. 2: Concentration Profile of the Batch Reactor after The authors acknowledge with thanks Dr. Mark Willis,
reaction time of 40 minutes with rate constant values Senior Lecturer, Applied Process Control Newcastle
of and 0.4. University UK, for his invaluable instructional
mentoring that inspired this line of thought.
1
REFERENCE
0.8 Denisov, T. E., O. M. Sarkisov, et al. (2003). Chemical
Kinetics: Fundamentals and New Developments.
CONCENTRATION (mol/dm3)

Amsterdam, Netherland, ELSERVIER SCIENCE

0.6 B. V.
ConcA
ConcB Higham, D. J. (2007). Modelling and Simulation Chemical
0.4 ConcC Reactions. Available at:
ConcD https://www.google.com.ng/webhpsource=search_
ConcE app#hl=en&gs_rn=9&gs_ri=psy-
0.2 ab&tok=odWmG3ID0NXRDUlJlKW6Ag&cp=59&g
s_id=72&xhr=t&q=chemical+kinetics+reaction+rat
e+modelling+with+Matlab+ODE45&es_nrs=true&
0
0 5 10 15 20 25 30 35 40 pf=p&sclient=psyab&oq=chemical+kinetics+reacti
TIME (minutes)
on+rate+model
Howard, P. (2007). Solving ODE in MATLAB. Avaliable at:
Fig. 3: Concentration Profile of the Batch Reactor after
www.cs.rice.edu/../matode.pdf (Accessed August
reaction time of 40 minutes with a rate constant values
2012): Pages 1 - 22.
of and .
Sharma, L. R. and M. S. Pathania (2006). PRINCIPLES
CONCLUSION OF PHYSICAL CHEMISTRY. Jalandhar, India,
Simulation of kinetic process models using Matlab VISHAL PUBLISHING CO.
functions solver, ODE45 programme proved to be a Ullah, M. and O. Wolkenhauer (2011). Stochastic
reliable finger print for obtaining dependable results Approaches for Systems Biology. Available at:
of Industrial Chemical Processes. The results http://www.springer.com/978-1-4614-0477-4,
obtained in this study showed that maximum yield of Springer. 290: p. 73.
3
the product ‘D’ (0.83mol/dm ) was obtained when the
coefficient of determinant, of the reaction was
. This maximum result was achieved within an

287