In the shell-model framework, valence-space Hamiltonians connecting multiple major-oscillator shells are of key interest for investigating the physics of neutron-rich nuclei, which have been the subject of intense experimental activity for decades. Here we present an extension of the ab initio valence-space in-medium similarity renormalization group which allows the derivation of such Hamiltonians nonperturbatively. Starting from initial two- and three-nucleon forces from chiral effective field theory, we then calculate properties of nuclei in the important island-of-inversion region above oxygen, so far unexplored with ab initio methods. Our results in the neon and magnesium isotopes indicate the importance of neutron excitation from the $sd$ to $pf$ shells and ground states dominated by intruder configurations around $N=20$, consistent with the conclusions from phenomenological studies. We also benchmark the excitation spectrum of $^{16}$O with coupled-cluster theory, finding generally good agreement, and discuss implications for ground state energies and charge radii in oxygen and calcium isotopes. Finally we outline the proper procedure for treating the long-standing issue of center-of-mass contamination, and show that with a particular choice of valence space, these spurious states can be removed successfully.

中文翻译：

---

从头算多壳价空间哈密顿量和反演岛

在壳模型框架中，连接多个主振荡器壳的价空间哈密顿量对于研究富中子核的物理学具有重要意义，这几十年来一直是激烈实验活动的主题。在这里，我们提出了 ab initio valence-space in-medium 相似性重整化群的扩展，它允许非微扰地推导此类哈密顿量。从手征有效场理论的初始两核和三核力开始，我们然后计算氧上方重要的反转岛区域中原子核的性质，到目前为止，还没有用从头算方法进行探索。我们在氖和镁同位素中的结果表明中子激发从 $sd$ 到 $pf$ 壳和由 $N=20$ 附近的入侵者配置主导的基态的重要性，与现象学研究的结论一致。我们还使用耦合簇理论对 $^{16}$O 的激发光谱进行了基准测试，发现总体上一致，并讨论了氧和钙同位素中基态能量和电荷半径的影响。最后，我们概述了处理长期存在的质心污染问题的正确程序，并表明通过特定的价空间选择，可以成功去除这些虚假状态。