Table of Contents
Fetching ...

Extended-range percolation in five dimensions

Zhipeng Xun, Dapeng Hao, Robert M. Ziff

TL;DR

This work investigates bond and site percolation on the five-dimensional sc(5) lattice with extended neighborhoods, using a high-statistics single-cluster growth Monte Carlo approach to determine the universal exponents $τ$ and $Ω$ and precise $p_c$ thresholds for up to the 7th nearest neighbor. The authors establish a self-consistent method to extract $p_c$, $τ$, and $Ω$ from $s^{τ-2}P_{ geq s}$ and $Q(s)$ plots, finding $τ=2.4177(3)$ and $Ω=0.27(2)$, in agreement with recent five-loop RG predictions. For bond percolation, $p_c$ values confirm Bethe-lattice scaling with finite-$z$ corrections $zp_c-1 \sim z^{-0.88}$ or $(3+\ln z)/z$; for site percolation, $p_c$ follows $p_c \approx c/(z+b)$ with $c=1.722(7)$ and $b=1$, approaching the continuum limit $zp_c \to 32 η_c = 1.742(2)$ as $z$ grows. These results bridge high-dimensional percolation theory with RG and continuum percolation, and quantify finite-size and neighborhood-shape effects in five dimensions.

Abstract

Percolation on a five-dimensional simple hypercubic (sc(5)) lattice with extended neighborhoods is investigated by means of extensive Monte Carlo simulations, using an effective single-cluster growth algorithm. The critical exponents, including $τ$ and $Ω$, the asymptotic behavior of the threshold $p_c$ and its dependence on coordination number $z$ are investigated. Using the bond and site percolation thresholds $p_c = 0.11817145(3)$ and $0.14079633(4)$ respectively given by Mertens and Moore [Phys. Rev. E 98, 022120 (2018)], we find critical exponents of $τ= 2.4177(3)$, $Ω= 0.27(2)$ through a self-consistent process. The value of $τ$ compares favorably with a recent five-loop renormalization predictions $2.4175(2)$ by Borinsky et al. [Phys. Rev. D 103, 116024 (2021)], the value 2.4180(6) that follows from the work of Zhang et al. [Physica A 580, 126124 (2021)], and the measurement of $2.419(1)$ by Mertens and Moore. We also confirmed the bond threshold, finding $p_c = 0.11817150(5)$. sc(5) lattices with extended neighborhoods up to 7th nearest neighbors are studied for both bond and site percolation. Employing the values of $τ$ and $Ω$ mentioned above, thresholds are found to high precision. For bond percolation, the asymptotic value of $zp_c$ tends to Bethe-lattice behavior ($z p_c \sim 1$), and the finite-$z$ correction is found to be consistent with both and $zp_{c} - 1 \sim a_1 z^{-0.88}$ and $zp_{c} - 1 \sim a_0(3 + \ln z)/z$. For site percolation, the asymptotic analysis is close to the predicted behavior $zp_c \sim 32η_c = 1.742(2)$ for large $z$, where $η_c = 0.05443(7)$ is the continuum percolation threshold of five-dimensional hyperspheres given by Torquato and Jiao [J. Chem. Phys 137, 074106 (2015)]; finite-$z$ corrections are accounted for by taking $p_c \approx c/(z + b)$ with $c=1.722(7)$ and $b=1$.

Extended-range percolation in five dimensions

TL;DR

This work investigates bond and site percolation on the five-dimensional sc(5) lattice with extended neighborhoods, using a high-statistics single-cluster growth Monte Carlo approach to determine the universal exponents and and precise thresholds for up to the 7th nearest neighbor. The authors establish a self-consistent method to extract , , and from and plots, finding and , in agreement with recent five-loop RG predictions. For bond percolation, values confirm Bethe-lattice scaling with finite- corrections or ; for site percolation, follows with and , approaching the continuum limit as grows. These results bridge high-dimensional percolation theory with RG and continuum percolation, and quantify finite-size and neighborhood-shape effects in five dimensions.

Abstract

Percolation on a five-dimensional simple hypercubic (sc(5)) lattice with extended neighborhoods is investigated by means of extensive Monte Carlo simulations, using an effective single-cluster growth algorithm. The critical exponents, including and , the asymptotic behavior of the threshold and its dependence on coordination number are investigated. Using the bond and site percolation thresholds and respectively given by Mertens and Moore [Phys. Rev. E 98, 022120 (2018)], we find critical exponents of , through a self-consistent process. The value of compares favorably with a recent five-loop renormalization predictions by Borinsky et al. [Phys. Rev. D 103, 116024 (2021)], the value 2.4180(6) that follows from the work of Zhang et al. [Physica A 580, 126124 (2021)], and the measurement of by Mertens and Moore. We also confirmed the bond threshold, finding . sc(5) lattices with extended neighborhoods up to 7th nearest neighbors are studied for both bond and site percolation. Employing the values of and mentioned above, thresholds are found to high precision. For bond percolation, the asymptotic value of tends to Bethe-lattice behavior (), and the finite- correction is found to be consistent with both and and . For site percolation, the asymptotic analysis is close to the predicted behavior for large , where is the continuum percolation threshold of five-dimensional hyperspheres given by Torquato and Jiao [J. Chem. Phys 137, 074106 (2015)]; finite- corrections are accounted for by taking with and .
Paper Structure (7 sections, 10 equations, 7 figures, 5 tables)

This paper contains 7 sections, 10 equations, 7 figures, 5 tables.

Figures (7)

  • Figure 1: Plots of $s^{\tau-2}P_{\geq s}$ with $\tau = 2.4178$ (a) and $Q(s)$ (b) versus $s^{-\Omega}$ with $\Omega = 0.28$ for bond percolation of the sc(5) lattice under different values of $p$ (a), and for $p = 0.11817145$ (b). In (a), we added the linear function $0.024 s^{-\Omega}$ in order to make the slope on the right nearly horizontal and allow us to expand the vertical scale and separate the curves. For $p \ne p_c$, the data in (a) behaves as $0.921+C_1(p-p_c)(s^{-\Omega})^{-\sigma/\Omega}$ with $\sigma/\Omega\approx -1.83$. Data fitting shows that these values of $p_c$, $\tau$ and $\Omega$ are self-consistent.
  • Figure 2: Plots of $s^{\tau-2}P_{\geq s}$ with $\tau = 2.4176$ (a) and $Q(s)$ (b) versus $s^{-\Omega}$ with $\Omega = 0.21$ and $\Omega = 0.27$ for site percolation on the sc(5) lattice under $p = 0.14079633$. Clearly, better linear behavior obtains with $\Omega=0.27$ than $\Omega = 0.21$; dashed lines are included as guides to the eye. Data fitting shows that these values of $p_c$, $\tau$ and $\Omega=0.27$ are self-consistent.
  • Figure 3: Plot of $s^{\tau-2}P_{\geq s}$ versus $s^{-\Omega}$ with $\tau = 2.4177$ and $\Omega = 0.27$ for bond percolation of the sc(5)-1,2 (a), sc(5)-1,2,3 (b), sc(5)-1,2,3,4 (c), sc(5)-1,...,5 (d), sc(5)-1,...,6 (e), and sc(5)-1,...,7 (f) lattices under different values of $p$.
  • Figure 4: Plot of $zp_c$ versus $z^{-x}$ and versus $(3 + \ln z)/z$ for bond percolation on the sc(5)-$1,...,n$ lattices with compact nearest neighborhoods for $n = 3,4,5,6,7$. If we pick $x=0.88$, we get an intercept close to 1, as well as a good linear fit. This data can also evidently be fitted with $zp_c -1 \sim a_1(C + \ln z)/z$ with $C=3$.
  • Figure 5: Plots of $s^{\tau-2}P_{\geq s}$ versus $s^{-\Omega}$ with $\tau = 2.4177$ and $\Omega = 0.27$ for site percolation on the lattices sc(5)-1,2 (a), sc(5)-1,2,3 (b), sc(5)-1,2,3,4 (c), sc(5)-1,...,5 (d), sc(5)-1,...,6 (e), and sc(5)-1,...,7 (f) under different values of $p$.
  • ...and 2 more figures