Division of Physics and Applied Physics, School of Physical and Mathematical Sciences, Nanyang Technological University, 637371, Singapore.
Nature. 2013 Jan 24;493(7433):504-8. doi: 10.1038/nature11721.
Optical irradiation accompanied by spontaneous anti-Stokes emission can lead to cooling of matter, in a phenomenon known as laser cooling, or optical refrigeration, which was proposed by Pringsheim in 1929. In gaseous matter, an extremely low temperature can be obtained in diluted atomic gases by Doppler cooling, and laser cooling of ultradense gas has been demonstrated by collisional redistribution of radiation. In solid-state materials, laser cooling is achieved by the annihilation of phonons, which are quanta of lattice vibrations, during anti-Stokes luminescence. Since the first experimental demonstration in glasses doped with rare-earth metals, considerable progress has been made, particularly in ytterbium-doped glasses or crystals: recently a record was set of cooling to about 110 kelvin from the ambient temperature, surpassing the thermoelectric Peltier cooler. It would be interesting to realize laser cooling in semiconductors, in which excitonic resonances dominate, rather than in systems doped with rare-earth metals, where atomic resonances dominate. However, so far no net cooling in semiconductors has been achieved despite much experimental and theoretical work, mainly on group-III-V gallium arsenide quantum wells. Here we report a net cooling by about 40 kelvin in a semiconductor using group-II-VI cadmium sulphide nanoribbons, or nanobelts, starting from 290 kelvin. We use a pump laser with a wavelength of 514 nanometres, and obtain an estimated cooling efficiency of about 1.3 per cent and an estimated cooling power of 180 microwatts. At 100 kelvin, 532-nm pumping leads to a net cooling of about 15 kelvin with a cooling efficiency of about 2.0 per cent. We attribute the net laser cooling in cadmium sulphide nanobelts to strong coupling between excitons and longitudinal optical phonons (LOPs), which allows the resonant annihilation of multiple LOPs in luminescence up-conversion processes, high external quantum efficiency and negligible background absorption. Our findings suggest that, alternatively, group-II-VI semiconductors with strong exciton-LOP coupling could be harnessed to achieve laser cooling and open the way to optical refrigeration based on semiconductors.
自发反斯托克斯发射伴随的光辐照可导致物质冷却，这一现象被称为激光冷却或光制冷，它是由普林舍姆于 1929 年提出的。在气态物质中，通过多普勒冷却可以在稀释的原子气体中获得极低的温度，并且通过辐射的碰撞再分布已经证明了超密气体的激光冷却。在固态材料中，激光冷却是通过反斯托克斯发光期间声子的湮灭来实现的，声子是晶格振动的量子。自从在掺杂稀土金属的玻璃中首次进行实验演示以来，已经取得了相当大的进展，特别是在掺镱玻璃或晶体中：最近，从环境温度冷却到约 110 开尔文的记录被打破，超过了热电珀耳帖冷却器。在半导体中实现激光冷却会很有趣，其中激子共振占主导地位，而不是在掺杂稀土金属的系统中，其中原子共振占主导地位。然而，尽管进行了大量的实验和理论工作，主要是在 III-V 族砷化镓量子阱上，迄今为止，尽管在半导体中尚未实现净冷却。在这里，我们报告了在半导体中使用 II-VI 族硫化镉纳米带（或纳米带）从 290 开尔文开始实现约 40 开尔文的净冷却。我们使用波长为 514 纳米的泵浦激光，并获得了约 1.3%的估计冷却效率和约 180 微瓦的估计冷却功率。在 100 开尔文下，532 纳米的泵浦导致净冷却约 15 开尔文，冷却效率约为 2.0%。我们将硫化镉纳米带中的净激光冷却归因于激子和纵光学声子（LOP）之间的强耦合，这允许在上转换发光过程中多个 LOP 的共振湮灭，高外量子效率和可忽略的背景吸收。我们的发现表明，替代地，具有强激子-LOP 耦合的 II-VI 族半导体可以被利用来实现激光冷却，并为基于半导体的光制冷开辟道路。
