Spectral purity expresses the degree of monochromaticity of a radiation, and has been early identified as one of the main remarkable characteristics of lasers even before the experimental demonstration of the first laser source. However, since no generic standard definition is available, complementary aspects have to be considered in order to assess spectral purity of a laser, mostly depending on the approach and on the application aimed with a laser set-up.
Linewidth is probably the most usual way of quantifying purity. It is commonly defined as the full width at half maximum (FWHM) of the laser line (Figure 1). This definition is both intuitive and a heritage of the pioneering paper of Schawlow and Townes, which give the finite limit of a laser linewidth due to quantum noise. Comparison of an actual laser linewidth with this theoretical limit is an interesting metric, but is essentially relevant for ultra-stable single mode continuous wave (CW) lasers. In most cases technical noises and drifts will significantly damp the linewidth to (much) higher values than the Schawlow Townes limit from the kHz to the MHz range for stabilized CW lasers like laser diodes, and from typically 1 MHz to 1 GHz for single frequency pulsed lasers depending on the pulse duration, and potential spectral broadening effects like frequency chirps.
Figure 1: Definition of linewidth as the full width at half maximum (FWHM) in the case of a single frequency laser (only one cavity mode exists). Two generic line shapes are represented, with the same linewidth and same total integrated power in the limit of the spectral window represented. For an equivalent total power, the Gaussian line shape “looks” broader in the vicinity of the half maximum but narrower in the wings than the Lorentzian line shape.
However, in the general cases the linewidth is not sufficient and has to be complemented with additional information on the actual energy distribution over the whole spectrum, even for single-frequency lasers. For example in high resolution spectroscopy deconvolution of the laser line shape enables to reach higher spectroscopic retrieval resolution. Moreover, the terminology linewidth is still often (loosely) used to describe multi-mode lasers. When a linewidth figure is given for a multi-mode laser, it would generally account for the envelope of the multi-mode spectrum, as it could be measured with an optical spectrum analyzer with a resolution lower than the free spectral range of the laser modes (Figure 2).
Figure 2: Schematic representation of a multi-mode laser spectrum. Each independent mode has its own linewidth (as defined in Figure 1 for a single mode laser). The whole modal structure could be defined as the linewidth, line shape, and relative power of each mode taken separately, which would draw a detailed but not straightforward metric. For the sake of simplicity only an equivalent linewidth, corresponding to the full width at half maximum of the envelope, is most frequently considered.
To characterize spectral purity while taking into account the multi-mode structure of a laser, the first approach is to characterize the side mode suppression ratio (SMSR), defined as the power ratio of the main lasing longitudinal mode to the next strongest mode (Figure 3). For laser systems with optimized selective cavity losses such as low power laser diodes, the SMSR reaches extremely high values in the 60 dB range, and has to be measured with high dynamic apparatus such as spectrometers or monochromators. The second approach is to characterize the ratio between the optical energy integrated in a fixed narrow spectral interval Dn with the total energy of the source as illustrated (Figure 3). This type of metric for spectral purity is more generic and more precise than SMSR as it quantifies a spectral density inside a narrow band Dn. It is thus very frequently used in spectrometry, since it can be adapted to the resolution and dynamic that is specifically required for the application. For example, in the frame of future spaceborne lidar missions for greenhouse gases concentration measurements (e.g. MERLIN mission for CH4), the purity of the laser source typically has to be > 99.9 % in Dn=1 GHz.
Figure 3: Definition of side mode suppression ratio (SMSR), and spectral purity as the ratio of integrated power inside a narrow band of width Dn to the total laser power.
Reference: A. L. Schawlow and C. H. Townes, “Infrared and optical masers”, Phys. Rev. 112 (6), 1940 (1958)