发表机构
A. P. Shah Institute of Technology（A.P.沙阿理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究构造了一族静态自旋网络哈密顿量，利用对称性约化和逆谱方法，实现了从局域双激发态到迪克态的完美态转移，并给出了符号化的存在性证明。
AI 中文摘要
我构造了一族与时间无关、激发数守恒的自旋哈密顿量，对于每个系统尺寸 $N \ge 4$，该哈密顿量能够实现从局域双激发态到对称双激发迪克态的完美态转移。该哈密顿量具有物理形式 $H = \sum_{i<j} J_{ij} (\sigma_i^+ \sigma_j^- + \sigma_j^+ \sigma_i^-) + \sum_i \epsilon_i n_i$，其中耦合为实数，并且在有限时间内满足 $e^{-iHt} |110\cdots0\rangle = e^{i\phi} |D_N^{(2)}\rangle$。该构造利用了作用于初始未占据自旋上的 $S_{N-2}$ 置换对称性，将动力学约化到一个四维不变子空间。要求 $(|\psi_0\rangle + |D_N^{(2)}\rangle)/2$ 为零本征向量，这以闭合形式确定了在位能，并留下三个自由耦合参数。剩余的逆谱问题归结为两个关于两个无量纲耦合比的多项式方程；消去一个比率得到一个六次倒数多项式，通过代换 $z = x + x^{-1}$ 可将其转化为三次方程。对于谱族 $(-n,-1,1)$，其中 $n$ 为奇数，对首项系数的显式分解以及在 $z=-2$ 处的边界评估表明，对于每个 $N \ge 4$，可以选择足够大的奇数 $n$，使得存在满足 $z < -2$ 的实根；子结式论证为该根提供了到原始系统的实提升。存在性证明完全是符号化的，不依赖数值优化；直接传播仅用作独立验证。该结果是完美态转移的一个受约束类比：与一般实态问题（其中无约束的实对称矩阵就足够了）不同，这里要求哈密顿量必须来自激发数守恒的自旋网络形式。
英文摘要
I construct a family of time-independent, excitation-preserving spin Hamiltonians realising perfect state transfer from a localised two-excitation state to the symmetric two-excitation Dicke state, for every $N\ge4$. The Hamiltonian has the physical form $H=\sum_{i<j}J_{ij}(σ_i^+σ_j^-+σ_j^+σ_i^-)+\sum_iε_i n_i$ with real couplings, and satisfies $e^{-iHt}|110\cdots0\rangle=e^{iϕ}|D_N^{(2)}\rangle$ at a finite time. An $S_{N-2}$ permutation symmetry on the unoccupied spins reduces the dynamics to a four-dimensional invariant subspace. Requiring $(|ψ_0\rangle+|D_N^{(2)}\rangle)/2$ to be a zero eigenvector fixes the on-site energies in closed form and leaves three coupling parameters free. The inverse spectral problem then becomes two polynomial equations in two coupling ratios; eliminating one gives a degree-six reciprocal polynomial, which $z=x+x^{-1}$ converts to a cubic. For the spectral family $(-n,-1,1)$ with odd $n$, factorising the cubic's leading coefficient and evaluating it at $z=-2$ shows that some odd $n$ always produces a real root below $-2$, which a subresultant lifts back to the original system. The existence argument is symbolic and uses no numerical optimisation. Since that coefficient contains no odd powers of $n$, the required $n$ comes with an explicit threshold, not an asymptotic guarantee. I also cost the construction: couplings grow as $N^{1/2}$ and the on-site range as $N^{3/2}$, the transfer time stays within a factor 2.3-3.0 of the Mandelstam-Tamm limit at every size, and the fidelity is sensitive to systematic drift of the spectator-spectator coupling class but tolerant of independent bond disorder, which self-averages. This is a constrained analogue of perfect state transfer: for general real states an unconstrained real symmetric matrix suffices, whereas here the Hamiltonian must have excitation-preserving spin-network form.
Comments12 pages, 7 figures
