定义：光能量的量子。

当一束很弱的光束进入一个很灵敏的光电探测器，可以发现能量是以小束的形式传播的，而不是连续的。这可以解释成光束包含了一束束的能量，称为量子或者光量子。光子能量为h ν = h c / λ，即普朗克常数h与光频率ν的乘积。在二十世纪，Max Planch处理热辐射问题中就采用了光包含能量束的观点，Albert Einstein在研究光电效应时也利用了这一观点。光子则是在1926年物理化学家 Gilbert N. Lewis定义的。 
将光子看做光的粒子可以帮助理解许多量子现象，但是如果不了解其限制因素则很容易得出错误的结论。现代量子光学给出了非常自洽的、但是不是很简洁的对光的本性的描述。这里将光子看做量子化电磁场的元激发。这一理论框架下光子的性质比较特别，不能采用简单的粒子图像或者纯波动图像来解释。 

光子的一些关键性质 

1. 光的传播（在自由空间或者波导中）实际就是波的传播。某时刻某一点的量子力学场振幅是光所有可能路径的叠加。叠加可能会干涉相长或者相消，这是光学干涉效应的基础。纯粒子图像很难与实验观察相符，例如，在传统的双缝干涉实验中，每一个粒子都会通过其中一条缝，而与另一条缝不相干；但是这无法解释只有当堵住一条缝之后，有些粒子才能到达某一位置，而两条缝都打开时就不能到达该位置的现象（相消干涉）。 
2. 当光与原子或者其它粒子发生相互作用时，只有是光子能量h ν整数倍的能量才能够被光场吸收或者释放。这可以简单的理解为只能吸收或辐射一定数目的光子。只有当作用的粒子（例如原子）能够接收这些能量时，相互作用过程才会发生，即它们量子力学能级间的能级差对应光子能量，或者是光子能量整数倍（参阅双光子吸收）。纯的波动图像将这些能量限制看做共振效应，但是无法解释能量的量子化现象。 
3. 在灵敏的光电探测器中，能量量子化特别明显，可以进行光子计数，即能够探测单光子吸收效应。这在很多的科学和技术领域都有应用。 
4. 光子的静止质量为0，因此不能使其变慢或者静止。存在所谓的慢光现象，是由于电磁场与物质之间强烈的相互作用引起的，因此只有在介质中才存在这种现象。这时并不仅仅是电磁场激发，纯的光子图像不能给出很合理的解释。 
5. 由于光子的泊松特性，多个光子倾向于激发相同模式的辐射场。例如，受激辐射过程中（激光器中很重要）就能发现，还可以在热激发辐射后的能量谱中看到（黑体辐射）。 
6. 光子可以以纠缠态存在，即不同光子的有些性质（例如，偏振）0是相干的，尽管只有当进行测量时这些性质才具有特定的值。由于对不同光子的测量可以发生在不同的位置，这似乎意味着超光速传输信息是可能的（Einstein-Podolsky-Rosen佯谬），但是实际上是不会发生的。 

量子理论不仅能应用到可见光中，它还可以应用到任意的电磁波现象中。但是在射频技术中，量子效应并不像在光学和激光器技术中这么重要。这是因为射电波的光子能量与其室温下的热能相比很小，而光学现象中则刚好相反。 
在激光物理中，常见的现象是光子在介质中传播，例如，透明晶体或者玻璃，以及激光增益介质。这时严格来说称为光子就不合适了，因为电磁波会与介质发生相互作用，因此传播的是准粒子，有时称为极化激元，类似于电磁场的激发态与极化介质耦合的产物。

Definition: quanta of light energy

When a weak light beam hits a sensitive photodetector, energy is found to be delivered in the form of small bunches, rather than continuously. This can be interpreted such that the light beam consists of small bunches of energy, called photons or light quanta (German 'Lichtquanten' = portions of light). The photon energy is h ν = h c / λ, i.e. the product of Planck's constant h and the optical frequency ν, and is also related to the vacuum wavelength λ.

The idea that light consists of such energy bunches had already been used early in the 20^th century by Max Planck in the context of thermal radiation, and by Albert Einstein when investigating the photoelectric effect. The term photon, however, was coined only in 1926 by the physical chemist Gilbert N. Lewis [1].

Although a 'naïve' interpretation of photons as particles of light gives a useful picture for the intuitive understanding of many quantum phenomena, it can be seriously misleading to apply it without understanding its limitations. A consistent and very powerful, but certainly not simple description of the nature of light is achieved by modern quantum optics. Here, photons are seen as the elementary excitations of the electromagnetic quantum field. This theory attributes fairly strange properties to photons, which cannot be reconciled either with a simple particle picture or with a pure wave picture, but accurately match a wide range of observations.

Some Key Properties of Photons

The delivery of energy is quantized.

• When light interacts with atoms or other particles, only amounts of energy which are integer multiples of the photon energy h ν can be transferred to or from the light field. This can be easily interpreted as absorption or emission of some number of photons, and would thus so far be compatible with a simple particle picture of photons. Such processes are possible only if the involved atoms, ions or molecules are able to accept such amounts of energy, i.e., only if they have quantum-mechanical energy levels with an energy difference corresponding to the photon energy, or in some cases some integer multiple of it (→ two-photon absorption). A pure wave picture could explain these energy constraints as resonance effects, but can not explain the quantization of exchanged energy.
• The energy quantization is also apparent in the interaction with sensitive photodetectors, which allow for photon counting, i.e. to register single photon absorption events. This finds applications in various areas of science and technology.

Particularly destructive interference is not compatible with a simple particle model.

• The propagation of light (e.g. in free space or in a waveguide) is that of a wave field. The quantum-mechanical field amplitude arising at some point in space and time is the superposition of contributions which correspond to different possible paths for light. These contributions can constructively or destructively interfere with each other, and this is the basis of the well-known optical interference effects. The likelihood or rate of photon detection at a certain location is proportional to the modulus squared of the quantum-mechanical field amplitude, and may thus suppressed if different field contributions cancel each other, leading to a weak overall field. A pure particle picture is hard to reconcile with such phenomena. For example, in the classical double-slit experiment an ordinary particle would have to go through one of the two slits, and the other slit would be irrelevant; it could not be explained why particles can reach certain locations behind the double slit only when one of the slits is blocked, but not when both are open (destructive interference).
• Photons have zero rest mass, and can therefore not be slowed or brought to rest. There are phenomena of “slow light”, but these occur only for light in media, where the electromagnetic field strongly interacts with matter.
• Photons can carry angular momentum of two different forms: resulting from their spin (s = 1) and also as an orbital angular momentum related to the corresponding electric field profile.
• Due to their boson nature, photons obey Bose–Einstein statistics (in contrast to Fermi–Dirac statistics for electrons). Multiple photons “like” to populate the same mode of the radiation field. This can be seen e.g. in the process of stimulated emission (and is thus also very important for lasers), but also in the energy spectrum of thermally excited radiation (black body radiation).
• Photons can occur in entangled states, where certain properties (e.g. polarization) are correlated between different photons, even though these properties acquire definite values only when a measurement is performed. As measurements on the different photons can occur at different places, this seemed to imply the possibility of superluminal transmission of information (Einstein–Podolsky–Rosen paradox), but a closer inspection shows that in reality this is not the case.

Of course, quantum theory can be applied to any kind of electromagnetic wave phenomena, not only to visible light. However, quantum effects are not as important e.g. in the field of radio technology, as in optics and laser technology. This is because the photon energy of radio waves is very tiny compared with the thermal energy k_B T at room temperature, whereas the opposite is true for optical phenomena.

Photons in Laser Physics

Various phenomena in laser physics, such as stimulated emission as the basis for light amplification in laser gain media, are often explained based on photons. Nevertheless, much of laser physics can be described with purely classical pictures, not involving photons; for example, light amplification can be described with rate equation models involving excitation numbers and classical light wave amplitudes or intensities.

Discrete photons “explain” intensity fluctuations of light, but comprehensive models are substantially more sophisticated than that simple picture.

A stronger involvement of photons, however, is natural in the context of laser noise, or more generally noise in optics. For example, the high-frequency intensity noise of lasers is usually at the shot noise level, and the magnitude of that is compatible with a simple model of completely randomly arriving (uncorrelated) photons. Lower (sub-shot noise) intensity noise may be observed for squeezed states of light, as can be interpreted as photons becoming correlated, such that they arrive in a more regular fashion. The calculation of squeezing and many other phenomena is again usually based not on a simple particle picture, but either on full-blown quantum mechanics or on a simplified semiclassical model involving amplitude and phase fluctuations which have to obey certain rules.

Bibliography