Numerical Solution of a Reaction-Diffusion System with Fast Reversible Reaction by Using Adomian’s Decomposition Method and He’s Variational Iteration Method

In this paper, the approximate solution of a reaction-diffusion system with fast reversible reaction is obtained by using Adomian decomposition method (ADM) and variational iteration method (VIM) which are two powerful methods that were recently developed. The VIM requires the evaluation of the Lagrange multiplier


Introduction:
Non-linear phenomena, that appear in many areas of scientific fields such as solid state physics, plasma physics, fluid dynamics, mathematical biology and chemical kinetics can be modeled by partial differential equation.A broad class of analytical solutions methods and numerical solutions methods were used in handle these problems [10].
In the 1980's, Adomian [2][3][4] introduced a new powerful method which provides an efficient means for the analytic and numerical solution of differential equations.It is free from rounding off errors since, it does not involve discretization and is computationally inexpensive.
Recently, the variational iteration method, which was first proposed by He [14][15][16][17] in 1997, has many merits and advantages over the Adomian method.A comparative study between the two methods was conducted by Wazwaz [24].
The main advantage of the two methods is that it can be applied directly for all types of differential and integral equations, homogeneous or inhomogeneous.Another important advantage is that the methods are capable of greatly reducing the size of computational work, while still maintaining high accuracy of the numerical solution [24].
The reaction diffusion equations (RDEs) have recently attracted considerable attention, partly due to their occurrence in many fields of science, in physics as well as in chemistry or biology, partly due to their interesting features and rich variety of properties of their solutions [27].
Eymard et al. [13] studied the numerical solution of the reaction-diffusion system with fast reversible chemical reaction of type mA nB  by using the finite volume method.
The present paper has been organized as follows: Section 2 will deals with the analysis of the methods applied to non-linear system of partial differential equations.In Sections 3 and 4, we will apply the ADM and VIM, respectively, to solve the reactiondiffusion system.An illustrative example with numerical results will be given in Section 5.
In this work, we consider a reversible chemical reaction between mobile species A and B , that takes place inside a bounded region    , we have the reactiondiffusion system of partial differential equations [8,13]: where, 0 T  and  is a bounded set of  , with the initial conditions For a reversible reaction , the rate functions are of the form , where 1 k and 2 k are rate constants, a and b are diffusion coefficients and k is the chemical kinetics factor (for further details see [11,12]).
Bothe and Hilhorst [8] exploited a natural Lyapunov functional and used compactness arguments to prove that where, w is the unique weak solution of the nonlinear diffusion problem and The identities in (3) can be explained as follows: the first one states that the system is in chemical equilibrium, while the second one defines w as the quantity that is conserved under the chemical reaction.Given a function w , the system (3) can be uniquely solved for ( , ) u v if A r and B r are strictly increasing with ( ) ( ) is well-defined and strictly increasing.Under these assumptions, u is the unique which gives the explicit representation of u and v [13]     u f w  and ( ) v g w  . Here,

Analysis of the methods
We first consider the system of partial differential equations written in an operator form , , with initial data , 0 , , 0 .
where t L is considered, without loss of generality, a first order partial differential operator, 1 R and 2 R are linear operators, 1 N and 2 N are nonlinear operators, and 1 g and 2 g are inhomogeneous terms.

Analysis of the ADM
Applying the inverse operator 1 t L  to the system (6) and using the initial data (7) yields , .
The Adomian decomposition method assumes a series that the unknown functions ( , ) u x t and ( , ) v x t can be expressed by an infinite series of the form [23]   , and the non-linear operators   where, ( , ) n u x t and   , , 0 n v x t n  are the components of ( , ) u x t and ( , ) v x t that will be elegantly determined, and n A and , 0 n B n  are Adomian polynomials that can be generated from the general algorithm [10] Substituting ( 9) and ( 10) into (8) gives , , Following Adomian analysis, the nonlinear system ( 6) is transformed into a set of recursive relations given by and The schemes ( 13) and ( 14) can easily determine the components ( , ) n u x t and ( , ) n v x t .It is, in principle, possible to calculate more components in the decomposition series to enhance the approximation.Consequently, can recursively determine every term of the series and, hence the solutions ( , ) u x t and ( , ) v x t are readily obtained.It is interesting to note that we obtained the solutions by using the initial conditions only.

Analysis of the VIM
According to VIM, we can construct correction functionals for equations of the system (6) as follows [7,25]: where 1  and 2  are general Lagrange multipliers, which can be identified optimally via the variational theory, the subscript n denotes the n th order approximation, n u and n v are considered as restricted variations, i.e., 0 , of the solutions ( , ) u x t and ( , ) v x t will be readily obtained by using selected functions 0 u and 0 v .Consequently, the solutions are given by [25]  .

Application of ADM to reaction-diffusion system:
In this section, we apply the ADM to solve the reaction-diffusion system with initial data The components of Adomian polynomials can be obtained from (11).To clarify that, we have two cases:


The same clarification can be shown for computing n B .

Application of VIM to reaction-diffusion system:
To solve the system (17) by means of VIM, we construct correction functionals which read where n u and n v are considered as restricted variations, i.e., 0 Its stationary conditions can be obtained as follows: The Lagrange multipliers, therefore, can be identified as 1 2 1      , and the iteration formulas are given by An illustrative example of an application of the two methods will be presented in the next section.

Numerical experiment:
We choose the reaction of the reversible dimerization of o-phenylenedioxy− dimethylsilane (2, 2-dimethyl-1, 2, 3-benzodioxasilole) which has been studied by 1 H NMR spectroscopy (for further details see [19]).Since the reaction is of the type 2A B  , the reaction terms take the form [13]     and the initial conditions 0 u and 0 v are defined as follows: Using Adomian decomposition method, the equations ( 18) and ( 19) become Hence, the recursive relations can be expressed as The solutions are obtained by using the initial conditions only.Consequently, the pair of zeroth components is given by 1 50 cos for 0.03, 0.07 4 7 0 for 0.07,0.1 and the pair of first components read The numerical results will be shown later in Tables 1-3 and Figs.1-2.Now, we solve the system (22) by the means of variational iteration method.The iteration formulas (21) become We start with the initial approximations given by ( 27) and (28), with the above iteration formulas, we found the pair of first components , for 0,0.03 , , for 0.03, 0.07 , for 0.07, 0.1 , 50 , sin 0.03 .4 7 We found the approximations 3 ( , ) u x t and 3 ( , ) v x t by the means of VIM.We also found the approximation w in terms of ( , )

Conclusions
In this paper, the Adomian decomposition method (ADM) and variational iteration method (VIM) were implemented to solve reaction-diffusion system which describes a reversible chemical reaction.The VIM reduces the volume of calculations by not requiring the Adomian polynomials while ADM requires the use of Adomian polynomials for non-linear terms, and this needs more work.However, the results show that the two methods are very simple and effective for non-linear problems.In our work, we used Matlab codes to obtain the solutions from the two algorithms.

2 ,
N u v can be decomposed by the infinite series of the so-called Adomian polynomials

Fig. 1 .
Fig.1.(a) The numerical results for ( , ) u x t obtained by ADM with 1 k  , (b) The numerical results for ( , ) u x t obtained by VIM with 1 k  , (c) The numerical results for ( , ) v x t obtained by ADM with 1 k  , (d) The numerical results for ( , ) v x t obtained by VIM with 1 k  , (e) The numerical results for ( , ) w x t obtained by ADM

Fig. 2 .
Fig.2.(a) The numerical results for ( , ) u x t obtained by ADM with 16 k  , (b) The numerical results for ( , ) u x t obtained by VIM with 16 k  , (c) The numerical results for , ) v x t obtained by ADM with 16 k  , (d) The numerical results for ( , ) v x t obtained by VIM with 16 k  , (e) The numerical results for ( , ) w x t obtained by ADM with 16 k  , (f) The numerical results for ( , ) w x t obtained by VIM with 16 k  .

Table 1 :
obtained by ADM and VIM for different values of t and k .The comparison shows that the two solutions obtained are in excellent agreement.Comparison of the approximate values of  

Table 2 :
Comparison of the approximate values of  

Table 3 :
Comparison of the approximate values of  