'Double-slit' and multiple-path Interference studies from Rb excited and ionized by high-resolution laser radiation.

2002年，利用电子在探测器上产生干扰的双缝实验被《物理世界》读者评选为“物理学中最美丽的5个实验”之一。2013年的最新实验表明，在下一个电子发射之前被探测到的单个电子，当信号随着时间的推移而增加时，也会产生干扰图案。这令人信服地表明，正如理查德·费曼（Richard Feynman）在20世纪60年代初所预测的那样，单个电子同时具有波状和粒子状的特征。费曼认为这样的实验是不可能完成的，然而，从那时起，技术的进步使这成为可能。由于我们不知道电子穿过哪个狭缝，就产生了干涉图样。我们给电子分配一个波函数，然后狭缝定义了波从源到探测器的两种可能路径。狭缝之外的波阵面然后在探测器处重新组合，它们之和的平方给出了探测到电子的概率。如果一个波的波峰遇到另一个波的波谷，波就会相互抵消，在那个位置检测到电子的概率为零。相反，如果两个波峰或两个波谷到达一个点，那么就有最大的电子被探测到的可能性。因此，在探测器上产生干涉图案，这取决于波在任何给定点上如何重新组合。在曼彻斯特，我们最近在单个原子中发明了一种新型的“双缝”实验，其中“缝”被激光激发的原子状态1和2所取代。激发状态1的激光束也电离状态2，而激光激发状态2电离状态1。然后有两种电离途径，我们不知道哪一种途径产生被探测到的光电子。我们必须再次添加每个路径的波函数来确定结果，从而导致干扰。状态（狭缝）可以被打开或关闭（有效地打开或关闭单个狭缝）通过选择性调谐和调谐激光，这使我们能够确定干涉模式。在本提案中将要进行的新实验中，我们将通过选择不同的激发态和使用不同的激光偏振来更详细地探索这一过程。我们在德国的合作者从理论上预测，这将对随后的模式产生巨大的变化。我们将进一步探讨的预测是，注入第三束激光可以选择性地控制干涉。这个新想法在传统的双缝实验中没有类似之处，可能会在其他必须操纵波函数的领域（例如量子计算）中找到应用。没有理由说明为什么这些过程必须局限于单个原子&这项工作的第二个方面将探索如何将激光激发和电离应用于原子阵列。我们将首先在磁光阱中将原子冷却到接近绝对零度，然后使用驻波激光创建受激原子的周期性阵列。然后，原子将被第二束激光电离，使出现的光电子的德布罗意波长在尺寸上与阵列的尺寸相当。干涉将再次发生，但是现在的总和是来自所有产生光电子的位置的波。由此产生的产率取决于单个原子，以及它们在阵列中的位置。这与衍射光栅对光的作用类似，不过现在的波是针对电子而不是光子的。通过改变激光器的特性，我们可以以不同的方式“塑造”光栅，这将改变产生的电子分布。我们合作者的初始模型支持这些想法，需要实验来测试和完善模型。这项工作可以应用于表面的电子衍射研究和控制粒子加速器中的电子注入。

The double-slit experiment using electrons to produce interference at a detector was voted as one of the 5 'most beautiful experiments in physics' by Physics World readers in 2002. Recent experiments in 2013 demonstrated that SINGLE electrons that were detected before the next electron was emitted also produce an interference pattern when the signal builds up over time. This convincingly shows individual electrons have both wave-like & particle-like character, as predicted by Richard Feynman in the early 1960's. Feynman thought such experiments would never be done, however advances in technology since then have now made this possible. The interference pattern arises since we do not know which slit the electron passes through. We assign a wavefunction to the electron, & the slits then define 2 possible pathways for the wave to travel from source to detector. The wavefronts beyond the slits then recombine at the detector, & the square of their sum gives the probability an electron is detected. If the peak of one wave meets the trough of another, the waves cancel & there is zero probability an electron will be detected at that position. By contrast, if two peaks or two troughs arrive at a point, there is then maximum probability an electron will be detected. An interference pattern is hence produced across the detector, which depends on how the waves recombine at any given point.In Manchester we recently invented a new type of 'double-slit' experiment in a single atom, where the 'slits' are replaced by atomic states 1 & 2 excited by lasers. The laser beam that excites state 1 also ionizes state 2, whereas the laser exciting state 2 ionizes state 1. There are then 2 pathways to ionization, & we do not know which was taken to produce the detected photoelectron. We again have to add the wavefunctions from each path to determine the outcome, leading to interference. The states (slits) can be turned on or off (effectively opening or closing individual slits) by selectively tuning & detuning the lasers & this allows us to determine the interference pattern.In the new experiments to be carried out in this proposal we will explore this process in much greater detail, by selecting different excited states & by using different laser polarizations. Our collaborators in Germany theoretically predict this will produce large changes to the ensuing pattern. A further prediction we will explore is that injection of a third laser beam can selectively control the interference. This new idea has no analogy in a conventional double-slit experiment & may find application in other areas where wavefunctions must be manipulated (e.g. quantum computing).There is no reason why these processes must be confined to single atoms & the second facet of this work will explore how laser excitation & ionization can be applied to arrays of atoms. We will first cool the atoms to close to absolute zero in a magneto-optical trap, before creating a periodic array of excited atoms using a standing-wave laser. The atoms will then be ionized by a second laser, set so that the de Broglie wavelength of the emerging photoelectrons is comparable in size to the dimensions of the array. Interference will once again occur, however now the summation is for waves from ALL sites from which the photoelectrons are born. The resulting yield then depends on both the individual atoms, as well as their position in the array. This is expected to be similar to the effect a diffraction grating has on light, however now the waves are for electrons rather than photons. By altering the properties of the lasers we can 'shape' the grating in different ways, which will change the electron distribution that is produced. Initial models from our collaborators support these ideas & experiments are needed to test & refine the models. This work could find application in electron diffraction studies of surfaces & for controlling the injection of electrons into particle accelerators.