定义：
光束合成的一种方法，采用波长敏感的合束器。

光谱合束（也称为波长合束或简单的称为波长合成，或非相干合束）是通过光束合成实现功率缩放领域中的一种技术。目的是将一些高功率的激光光束合成在一起得到一个光束，具有非常高的功率并且尽量保持光束质量，因此亮度提高。另一种可用的方法为相干光束合成。

光谱合束原理
光谱合束的基本原理是将几束无交叠光谱的光束在偏振敏感的合束器中合成得到一束光束。合束器可以是棱镜和衍射光栅等，由于入射光束中不同波长的成分产生的偏移不同，最后它们在同一个方向上传播。还可以利用光谱中各成分随波长变化的透射率，例如二色性反射镜或体布拉格光栅。

将很多同一类型的不同发射器的输出合成时，每一发射器（激光器）的辐射带宽应该仅是增益带宽的很小部分，因为带宽变大会降低光束质量。另外，合束器需要具有很强的色散（例如，角色散），各个发射器的波长需要足够稳定。

图1：采用衍射光栅进行光谱合束的示意图。每一波长都对应一个衍射角，发射器自动根据其位置调整到相应的波长。
有些方案中发射器是独立的调谐到某一波长，以对应的角度输出到合束器（例如，衍射光栅）。但是如果能够改进这一方案，使发射器自动根据其空间位置调整辐射波长，将是非常有效的。如图1，MIT Lincoin实验室已经制作了这种装置。
这一原理首先是应用到激光器二极管，以二极管阵列的形式，它同样适用于光纤激光器。二极管激光器很大的优势是可以直接进行电泵浦且效率很高，然而光纤系统单位发射器的功率更高，并且发射器的增益带宽更大，因此可以得到更高的功率。
合束器中与波长有关的热效应会限制功率缩放程度。尤其是如果采用透射光栅，存在很严重的研制，而反射光栅则至少可工作在100 kW以下，因为它吸收较少的功率并且可以从背侧冷却。

结论
与相干合束合成相比，光谱合束不需要合成的光束具有时间相干性。这会消除很多技术上的困难，并且在高功率时也更容易实现稳定工作。另外，输出光的多波长特性是不可避免的，有些应用中没有太大影响，但是需要窄带输出时则需要消除不需要的波长。

Definition: a class of methods for beam combining, based on wavelength-sensitive beam combiners

Alternative term: wavelength beam combining

More general term: beam combining

The term spectral beam combining (also called wavelength beam combining or simply wavelength combining, or incoherent beam combining) denotes a class of techniques within the more general area of power scaling by beam combining. The goal is to combine several high-power laser beams so as to obtain a single beam not only with correspondingly higher power but also with more or less preserved beam quality and thus with increased brightness. An alternative class of techniques, discussed in a separate article, is coherent beam combining.

Lasers sources involving spectral beam combining can be considered as multi-line lasers.

Principle of Spectral Beam Combining

The general principle of spectral beam combining is to have several beams with non-overlapping optical spectra and combine them at some kind of wavelength-sensitive beam combiner. Examples of such combiners are prisms and diffraction gratings, which can deflect incident beams according to their wavelengths so that subsequently these all propagate in the same direction. Other approaches rely on optical components with wavelength-dependent transmission, such as dichroic mirrors or volume Bragg gratings.

For combining the outputs of a large number of different emitters of the same type, each emitter (laser) needs to have an emission bandwidth which is only a small fraction of the gain bandwidth, because a larger bandwidth would lead to a reduced beam quality. Furthermore, the beam combiner must have a sufficiently strong dispersion (e.g. angular dispersion), and the wavelengths of all emitters must be sufficiently stable.

Figure 1: Setup for wavelength beam combination with a diffraction grating. Each wavelength has its own diffraction angle at the grating, and the emitters automatically tune to the wavelengths corresponding to their location. For clarity, only the beams of the outer emitters are shown.

Some approaches build on emitters which are independently tuned to some wavelength, and their outputs are aligned to reach the beam combiner (e.g. a diffraction grating) at the corresponding angle. It can be advantageous, however, to modify the scheme such that each emitter automatically adjusts its wavelength according to its spatial position. This is achieved e.g. with a setup developed at MIT Lincoln Laboratory [1, 5], as shown in Figure 1.

This principle was first applied to laser diodes in the form of diode arrays [1], but is also suitable for fiber lasers [3]. Whereas diode lasers have the advantages of being directly electrically pumped and of being very efficient, fiber-based systems can have a higher power per emitter, and the emitters can have a broader gain bandwidth, so that much higher powers are possible.

Of course, power scalability can be limited, e.g. by thermal effects on the wavelength-sensitive beam combiner. For transmission gratings, in particular, this introduces severe limitations, whereas reflection gratings could work at power levels of at least 100 kW, because they absorb less power and can be cooled from the back side.

General Remarks

Compared with coherent beam combining, the technique of spectral beam combining has the great advantage of not requiring mutual temporal coherence of the combined beams. This eliminates some important technical challenges and makes it much easier to obtain stable operation at high power levels. On the other hand, the multi-wavelength nature of the output is then unavoidable, which may not matter for some applications while excluding others where a narrow-bandwidth output is required.

Bibliography