On-lattice coalescence and annihilation of immobile reactants in loopless lattices and beyond

E. Abad
DOI: https://doi.org/10.1103/PhysRevE.70.031110
2005-01-11
Abstract:We study the behavior of the chemical reactions $A+A\to A+S$ and $A+A\to S+S$ (where the reactive species $A$ and the inert species $S$ are both assumed to be immobile) embedded on Bethe lattices of arbitrary coordination number $z$ and on a two-dimensional (2D) square lattice. For the Bethe lattice case, exact solutions for the coverage in the $A$ species in terms of the initial condition are obtained. In particular, our results hold for the important case of an infinite one-dimensional (1D) lattice ($z=2$). The method is based on an expansion in terms of conditional probabilities which exploits a Markovian property of these systems. Along the same lines, an approximate solution for the case of a 2D square lattice is developed. The effect of dilution in a random initial condition is discussed in detail, both for the lattice coverage and for the spatial distribution of reactants.
Statistical Mechanics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to understand the behavior of chemical reactions \( A + A \rightarrow A + S \) and \( A + A \rightarrow S + S \) of immobile reactants in acyclic lattices (such as Bethe lattices) and two - dimensional square lattices. Specifically, the research objectives include: 1. **Exact and approximate solutions**: For Bethe lattices, the author obtained an exact solution regarding the coverage by the method of conditional probability expansion. These results are especially applicable to infinite one - dimensional lattices (\( z = 2 \)). For two - dimensional square lattices, the author developed an approximate solution method. 2. **Influence of initial conditions**: The influence of dilution effects on lattice coverage and the spatial distribution of reactants under random initial conditions was studied. In particular, how a partially filled lattice affects the subsequent dynamics and steady - state of the system was explored. 3. **Reaction - induced spatial self - organization**: The non - uniform fluctuations caused by internal chemical noise during the reaction process were analyzed, and these fluctuations are particularly significant in low - dimensional or under specific geometric constraints. How these fluctuations affect macroscopic variables such as global concentration or coverage was studied. 4. **Application of the conditional probability method**: The conditional probability method was used to estimate lattice coverage, and this method is very effective when dealing with immobile reactants. This method is exact in 1D and branching media (such as Bethe lattices) and can be used as a starting point for density expansion in other regular lattices. 5. **Dynamic behavior and correlation functions**: The reaction - induced dynamic behavior of spatial distribution was studied, especially the specific influence of different initial conditions on correlation functions. In particular, in the annihilation reaction (AR) model, several types of crossover behaviors were discovered. ### Summary of main formulas - **Evolution equation**: \[ \frac{dP_k}{d\tau} = -(k - 1)P_k-\nu\left(\frac{zk - 2k + 2}{2}\right)P_{k + 1}, \quad k = 1,2,\ldots \] where \(\nu = 1\) for the coagulation reaction (CR) and \(\nu = 2\) for the annihilation reaction (AR). - **Exact solution for coverage**: \[ Q_s(\tau)=\theta(\tau)=p\left[1+\frac{z - 2}{2\nu p}(1 - e^{-\tau})\right]^{-\frac{z}{z - 2}} \] - **Asymptotic coverage**: \[ \theta_S=\theta(\infty)=p\left[1+\frac{z - 2}{2\nu p}\right]^{-\frac{z}{z - 2}} \] - **Approximate solution for two - dimensional square lattices**: \[ Q_s(\tau)=\frac{p}{(1+\nu p(1 - e^{-\tau}))^2} \] \[ Q_{ss'}(\tau)=\frac{p e^{-\tau}}{1+\nu p(1 - e^{-\tau})} \] ### Conclusion By the conditional probability method, the author successfully described the behavior of immobile reactants in different lattice structures and revealed the important influence of initial conditions on the system dynamics and steady - state. This method not only provides exact solutions but also provides an effective tool for approximate solutions in complex lattice structures.