AI 中文总结
该研究提出利用有限壁环形波导的轨道l-量子比特，通过约束狄拉克流实现无场横向AB相位门，可执行R_z(2δ_l)操作并具有相应灵敏度与相位门磁场参数。
AI 中文摘要
自由空间阿哈罗诺夫-玻姆（AB）贝塞尔模式已知携带依赖通量的方位角薛定谔概率流和动力学轨道角动量。我们将该响应扩展到核心排除的有限壁环形波导的自旋分辨守恒狄拉克流，证明约束会将其幸存的轨道分量转化为横向阿哈罗诺夫-玻姆（TAB）传播相位，适用于质心路径在矢量势A上投影为零的直传播模式。径向梯度流在自旋反转时反向；保留完整的倏逝尾将其作为边界贡献闭合，留下与自旋无关的轨道相位Δφ_ln∝lΦL_int⟨ρ⁻²⟩_ln/v_z。相反旋向的模式|±l⟩形成同路径量子比特，实现带差分读出和共模相位抑制的R_z(2δ_l)，而直接一阶串扰需要角谐m=±2l。对于a=20nm、R=30nm、E_z=10meV、|l|=10的100μm段，灵敏度为0.1885mrad/mG，R_z(π)出现在16.67G处。因此有限壁约束将本征方位角狄拉克流转化为导向无场相位操作。
英文摘要
The Aharonov--Bohm (AB) effect is usually read out through phase differences associated with spatially distinct electron paths. We show that confined orbital modes provide a same-path alternative: a core-confined magnetic flux writes opposite propagation phases on the co-propagating modes $|\pm l\rangle$ of a straight annular electron guide while the transported electron-wave support remains field free. In a spin-resolved Dirac treatment, the phase is carried by the overlap of the field-free vector potential $A_ϕ$ with the mode's azimuthal conserved-current texture. The spin-dependent radial-gradient current becomes a boundary term that cancels when the complete finite-wall evanescent tail is retained, leaving the spin-independent orbital phase $Δϕ_{ln}\propto lΦL_{\rm int}\langleρ^{-2}\rangle_{ln}/v_z$. The matched $|\pm l\rangle$ modes therefore realize a same-path $R_z(2δ_l)$ gate, with differential internal-mode readout and common-mode phase rejection. For $a=75\,\mathrm{nm}$, $R=95\,\mathrm{nm}$, $L_{\rm int}=1\,\mathrm{mm}$, $E_z=10\,\mathrm{meV}$, and $|l|=10$, the gate angle is $2.315\,\mathrm{rad/G}$ and $R_z(π)$ occurs at $1.357\,\mathrm{G}$. Finite-barrier, mode-spacing, disorder-mismatch, and readout-visibility checks quantify the main implementation constraints. More broadly, the result connects a mode-resolved AB energy shift to a measurable propagation operation and shows how the spatially distributed conserved current of a Dirac wave can become an operational quantum-control resource.
Comments10 pages, 2 figures, 1 table. Revised version with updated device-scale parameters, numerical benchmarks, and minor improvements to the presentation. The main theoretical results are unchanged
