Laboratoire Kastler-Brossel, ENS, CNRS, Université Pierre et Marie Curie - Paris 6, 24 rue Lhomond, 75005 Paris, France.
Nature. 2011 Jul 13;475(7355):210-3. doi: 10.1038/nature10225.
A measurement necessarily changes the quantum state being measured, a phenomenon known as back-action. Real measurements, however, almost always cause a much stronger back-action than is required by the laws of quantum mechanics. Quantum non-demolition measurements have been devised that keep the additional back-action entirely within observables other than the one being measured. However, this back-action on other observables often imposes its own constraints. In particular, free-space optical detection methods for single atoms and ions (such as the shelving technique, a sensitive and well-developed method) inevitably require spontaneous scattering, even in the dispersive regime. This causes irreversible energy exchange (heating), which is a limitation in atom-based quantum information processing, where it obviates straightforward reuse of the qubit. No such energy exchange is required by quantum mechanics. Here we experimentally demonstrate optical detection of an atomic qubit with significantly less than one spontaneous scattering event. We measure the transmission and reflection of an optical cavity containing the atom. In addition to the qubit detection itself, we quantitatively measure how much spontaneous scattering has occurred. This allows us to relate the information gained to the amount of spontaneous emission, and we obtain a detection error below 10 per cent while scattering less than 0.2 photons on average. Furthermore, we perform a quantum Zeno-type experiment to quantify the measurement back-action, and find that every incident photon leads to an almost complete state collapse. Together, these results constitute a full experimental characterization of a quantum measurement in the 'energy exchange-free' regime below a single spontaneous emission event. Besides its fundamental interest, this approach could significantly simplify proposed neutral-atom quantum computation schemes, and may enable sensitive detection of molecules and atoms lacking closed transitions.
测量必然会改变被测量的量子态，这一现象被称为反作用。然而，实际测量几乎总是会导致比量子力学定律所要求的更强的反作用。已经设计出了量子非破坏测量，使额外的反作用完全包含在被测量以外的可观测量中。然而，这种对其他可观测量的反作用往往会施加自己的约束。特别是，用于单原子和离子的自由空间光学检测方法（如货架技术，一种敏感和成熟的方法）不可避免地需要自发散射，即使在分散区域也是如此。这会导致不可逆的能量交换（加热），这在基于原子的量子信息处理中是一个限制，因为它排除了量子位的直接重复使用。量子力学不需要这种能量交换。在这里，我们通过实验证明了具有明显少于一个自发散射事件的原子量子位的光学检测。我们测量包含原子的光学腔的透射和反射。除了量子位检测本身，我们还定量测量了自发散射的发生量。这使我们能够将获得的信息与自发发射的量联系起来，并且在平均散射小于 0.2 个光子的情况下，我们获得了低于 10%的检测误差。此外，我们进行了量子 Zeno 型实验来量化测量的反作用，并发现每个入射光子都会导致几乎完全的状态崩溃。这些结果共同构成了在单个自发发射事件以下的“无能量交换”区域中量子测量的完整实验表征。除了其基本兴趣之外，这种方法可以大大简化提议的中性原子量子计算方案，并可能实现对缺乏闭合跃迁的分子和原子的敏感检测。
