World Wide Journal of Multidiscip linary Research and Development

WWJMRD 2017; 3(12): 216-220

www.wwjmrd.com
International Journal
Peer Reviewed Journal Numerical Methods for Ordinary Differential
Refereed Journal
Indexed Journal Equations Systems with Small Parameter with
UGC Approved Journal
Impact Factor MJIF: 4.25 Applications in Kinetics Chemistry
e-ISSN: 2454-6615

Ilea Mihai Ilea Mihai, Turnea Marius, Rotariu Mariana, Arotaritei Dragos,
University of Medicine and Gheorghita Andrei
Pharmacy “Grigore T. Popa”,
Department of Medical
Biosciences, Iasi, ROMANIA Abstract
Mathematical modelling is a more and more frequently used working tool, located, in this particular
Turnea Marius case, at the intersection between two basic sciences: mathematics and chemistry. We can wonder: can
University of Medicine and mathematical models offer predictions on the evolution of a chemical reaction? .Chemical reactions
Pharmacy “Grigore T. Popa”, of a wide variety can be modelled with coupled differential equations. These describe the time
Department of Medical evolution of the concentrations of the various chemical species: reactants, intermediaries, catalysts,
Biosciences, Iasi, ROMANIA and products. Such problems are quite simple to set up and solve with Matlab. The Matlab
functions can numerically integrate the differential equations that arise. The resulting concentrations
Rotariu Mariana
can be plotted as a function of time and also be used to accurately compute the expected
University of Medicine and
Pharmacy “Grigore T. Popa”,
concentration of the molecular species. The use of MATLAB in this article illustrates the important
Department of Medical role of informatics in research in chemical reactions modelling. An important advantage of using a
Biosciences, Iasi, ROMANIA numerical method is that the mathematical representation of the chemical processes allows
transparency and correctness in relation to chemical assumptions, thus allowing us to test our
Arotaritei Dragos understanding on the chemical reactions by comparing the graphs obtained with other chemical
University of Medicine and models observed.
Pharmacy “Grigore T. Popa”,
Department of Medical Keywords: Chemical kinetics Reaction, Numerical Methods, Matlab, Small Parameter
Biosciences, Iasi, ROMANIA
Introduction
Gheorghita Andrei
Polytechnic University of Chemical kinetics, a topic in several chemistry courses, illustrates the connection between
Bucharest, Splaiul mathematics and chemistry .Differential equations systems with small parameter are of basic
Independentei 313, importance in kinetics chemistry because many biological laws and relations appear
Bucharest, ROMANIA mathematically in the form of a ordinary differential equation systems. The kinetic equations
which describe the performance of a chemical kinetics are governed by linear differential
equations and a general analytical solution was successfully derived using differential
equations theory. When modelling the kinetics of a composite chemical reaction, it is
necessary to take into account all available information on the reaction mechanism. When we
draw a more complex reaction mechanism, each and every arrow corresponds to a term in the
differential equation, and the sign of the term depends on the direction of the arrow
(O’Malley : 1991) . Differential equations systems derived from chemical kinetics must often
be solved numerically. The following article is a detailed numerical method on how to use
Matlab to numerically integrate sets of differential equations with small parameter forms of
rate laws, and plot the solution as time-dependent concentrations. A critical evaluation of
methods for the numerical solution of differential systems with small parameter suggests that
the availability of more than one method is an important element in the numerical analyst’s
arsenal. Instead of allowing the program to integrate over the largest time increment
Correspondence: consistent with the error parameters, we force it to compute and store the values of the
Ilea Mihai
variables each  , where  is a small number (Linß & Stynes: 2009).
University of Medicine and
Pharmacy “Grigore T. Popa”,
Department of Medical
Biosciences, Iasi, ROMANIA
~ 216 ~
World Wide Journal of Multidisciplinary Research and Development

Materials and methods From the algorithm of Euler's method we obtain:

Consider three chemical reactions involving three
s 1 s 1 d s 1v
substances: D v(t )  T s 1
 O(T s  2 )
dt (4)
 X 
k1
2Y u (t ), v(t ) that approximate
 Suppose there are functions
2Y  X
k2
(1) the functions u(t  T ) and v(t  T ) :

Y 
k
Z
u (t  T ) u (u (t ), v (t ))
Given the reactions represented in (1), how can we use 
numerical integration to solve for the concentrations of
v(t  T )  v(u (t ), v(t )) (5)
X , Y and Z for all combinations of rate constants k1 , k 2 , We build the algorithm of the form:
and k? tn  to  nT
A macroscopic model of this reaction assumes that the 
concentrations and reaction constants are well-defined. ui 1  u (ui , vi )  u (ti 1 ), i  0, s

vi 1  v(ui , vi )  v(ti 1 )
Furthermore, the concentrations vary deterministically and
continuously (Vasil’eva & Butuzov : 1995 ) .Assume that: (6)
1 v0 it under the
1. k1  k 2  , the forward and reverse reactions The algorithm consists in determining
 u
involving substance X and Y are fast conditions it 0 is given so that the relationship (3) is true
1 has the form (Miller & Shishkin : 1996 ) :
u0 v0 is to be determined
2. k <<  , the last reaction is slow.  The value is specified ,
The system of equations for concentration x and y has  Consider the number of iterations: l  1
the form: v0(1)  v0
 We consider initialization
 dx
  x  y2 u0 , v0( l ) , we get the values: v1(l ) , v2( l ) ,..., vs( l)1 (7)
 dt  For
 (2)
  (1) s D s 1v0(l )
 dy 
 2 x  2 y 2  ky We introduce the notation:
 dt  If  it is small, the iteration is convergent
If we denote the concentrations of substances X and Y by
 v0( s 1)  v0( s )  
x and y . Instantaneous rate is the slope of a  if not small :
concentration versus time plot and is shown by the  Increasing the value of l and returning to step four.
dx dy
differential equations: and . The speed of a For s0 we get: (8)
dt dt
reaction (1) is expressed as the change in concentration of a For s 1 obtain:
  v2  2v1  v0 (9)
reactant or product over a certain amount of time.
To determine the numerical algorithm (Roos & Tobiska: Results & Discussion
2008), we will use the Euler method on a well-specified Using Euler's method for n steps and fixed pitch p (
range 0,T1  . T  np ), we have implemented the values of functions u
 s 1
 D v(t )  D v(t  T )  D v(t )
s s

 0 and v from relations (7).

 D v(t )  v(t )
 (3)

Table.1: For
  0,1 , p   , n  1 , we got the estimates
100
s Soluţia iterată y Soluţia asimptotică y
0 0.500000000 0.500000000
1 0.502038375 0.50211400
2 0.502021875 0.502016233
3 0.502021813 -
4 0.502021813 -

~ 217 ~
World Wide Journal of Multidisciplinary Research and Development

Fig.1: Simulations of system solutions (1) for   0,1.

Table.2: For
  0,1 , p   , n  4 , we got the estimates
10
s Soluţia iterată y Soluţia asimptotică y
0 0.497775080 0.500000000
1 0.502056821 0.502014000
2 0.502024335 0.502016233
3 0.502024087 -
4 0.502024017 -

Fig. 2: Comparison of system solutions (1) for

  0,01 , in connection with changing the domain. Whatever small  we take, from a
transparent and exact solution of the simple model at   0 ,we abruptly face a far more complicated solution of a model that is just a
little bit better .

Table .3: For

  0,01 , p   , n  1 ,we got the estimates
100
s Soluţia iterată y Soluţia asimptotică y
0 0.500000000 0.500000000
1 0.500200614 0.500201400
2 0.500200422 0.50200412

~ 218 ~
World Wide Journal of Multidisciplinary Research and Development

Fig.3: Simulation of system solutions (1) for   0,001 . It is noticeable that for T quite large, we will have a rapid convergence. The
smallness of  is relative to the size of the solution domain.

Fig. 4: 3D simulation of the "quick" solutions of the system (1) for different parameter  . Of the 6 graphs, there is no significant difference
in the behavior of the solution graphs.

~ 219 ~
World Wide Journal of Multidisciplinary Research and Development

Conclusions
Chemical kinetics deals with chemistry experiments and
interprets them in terms of differential equations systems
with small parameters. Chemistry has difficulty with
developing their own conceptual framework of what a
differential equation with small parameters really
represents. We studied a system of differential equations
with small parameter for chemical reactions. Solutions to
ordinary differential equations cannot be determined
uniquely without some outside condition typically an initial
value or a boundary value. We used numerical methods to
find approximate solutions to the system. All the equations
system includes small parameter  . We compared types of
solutions graphically and observed that the solutions get
closer to each other as   0 .This property of the
solutions suggests that the numerical method gives a very
accurate approximation to the solution of our model.
Simulation of the system of differential equations
containing a small parameter depending on various values
of  (0    1) parameter can generate new work
hypotheses work .We note that this work highlights how a
simple mathematical model with small parameter can be
used to predict the biologically relevant behaviors of a
chemical reaction. We hope that the numerical simulations
from our model can help to build a more efficient path to
chemical discovery. Hopefully our article is a useful
pedagogic tool for understanding the equations of chemical
kinetics.

References
1. O’Malley Jr., R. E. Singular perturbation methods for
ordinary differential equations, 1991, Springer-Verlag,
New York.
2. Linß T., Stynes M., Numerical solution of systems of
singularly perturbed differential equations ,2009 ,
Comput. Methods Appl. Math., 9: 165–191.
3. Vasil’eva A.B , Butuzov V.F. , Kalachev L.V., The
Boundary Function Method for Singular Perturbation
Problems, 1995 ,SIAM, Philadelphia, PA .
4. Roos H.G., Stynes M. ,Tobiska L., Robust Numerical
Methods for Singularly Perturbed Differential
Equations, 2008 ,Springer Series in Computational
Mathematics, vol. 24. Springer, Berlin, 2nd edition .
5. Miller J. J. H, O’Riordan E. , Shishkin G. I., Fitted
Numerical Methods for Singular Perturbation
Problems, 1996, World-Scientific, Singapore.

~ 220 ~