The auxiliary collimating and focusing optics that modify the wavefronts incident on and diffracted by a grating, as well as the angular configuration in which it is used, is often called its mount. Grating mounts are a class of spectrometer, a term that usually refers to any spectroscopic instrument, regardless of whether it scans wavelengths individually or entire spectra simultaneously, or whether it employs a prism or grating. For this discussion we consider grating spectrometers only.

A monochromator is a spectrometer that images a single wavelength or wavelength band at a time onto an exit slit; the spectrum is scanned by the relative motion of the entrance and/or exit optics (usually slits) with respect to the grating. A spectrograph is a spectrometer that images a range of wave-lengths simultaneously, either onto photographic film or a series of detector elements, or through several exit slits (sometimes called a polychromator). The defining characteristic of a spectrograph is that an entire section of the spectrum is recorded at once.

A plane grating is one whose surface is flat. Plane gratings are normally used in collimated incident light, which is dispersed by wavelength but is not focused. Plane grating mounts generally require auxiliary optics, such as lenses or mirrors, to collect and focus the energy. Some simplified plane grating mounts illuminate the grating with converging light, though the focal properties of the system will then depend on wavelength. For simplicity, only plane reflection grating mounts are discussed below, though each mount may have a transmission grating analogue.

This design involves a classical plane grating illuminated by collimated light. The incident light is usually diverging from a source or slit, and collimated by a concave mirror (the collimator), and the diffracted light is focused by a second concave mirror (the camera); see Figure 6-1. Ideally, since the grating is planar and classical, and used in collimated incident light, no aberrations should be introduced into the diffracted wavefronts. In practice, since spherical mirrors are often used, aberrations are contributed by their use off-axis.³⁹

Figure 6-1.   The Czerny-Turner mount. The plane grating provides dispersion and the concave mirrors provide focusing.

Like all monochromator mounts, the wavelengths are imaged individually. The spectrum is scanned by rotating the grating; this moves the grating normal relative to the incident and diffracted beams, which (by Eq. (2-1)) changes the wavelength diffracted toward the second mirror. Since the light incident on and diffracted by the grating is collimated, the spectrum remains at focus at the exit slit for each wavelength, since only the grating can introduce wavelength-dependent focusing properties.

Aberrations^† caused by the auxiliary mirrors include astigmatism and spherical aberration (each of which is contributed additively by the mirrors); as with all concave mirror geometries, astigmatism increases as the angle of reflection increases. Coma, though generally present, can be eliminated at one wavelength through proper choice of the angles of reflection at the mirrors; due to the anamorphic (wavelength-dependent) tangential magnification of the grating, the images of the other wavelengths experience higher-order coma (which becomes troublesome only in special systems).

This design is a special case of a Czerny-Turner mount in which a single relatively large concave mirror serves as both the collimator and the camera (Figure 6-2). Its use is limited, since stray light and aberrations are difficult to control - the latter effect being a consequence of the relatively few degrees of freedom in design (compared with a Czerny-Turner monochromator). This can be seen by recognizing that the Ebert-Fastie monochromator is a special case of the Czerny-Turner monochromator in which both concave mirror radii are the same, and for which their centers of curvature coincide. However, an advantage that the Ebert-Fastie mount provides is the avoidance of relative misalignment of the two mirrors.

Figure 6-2.   The Ebert-Fastie mount. A single concave mirror replaces the two concave mirrors found in Czerny-Turner mounts.

In this mount (see Figure 6-3), a plane grating is illuminated by converging light. Usually light diverging from an entrance slit (or fiber) is rendered converging by off-axis reflection from a concave mirror (which introduces aberrations, so the light incident on the grating is not composed of perfectly spherical converging wavefronts). The grating diffracts the light, which converges toward the exit slit; the spectrum is scanned by rotating the grating to bring different wavelengths into focus at or near the exit slit. Often the angles of reflection (from the primary mirror), incidence and diffraction are small (measured from the appropriate surface normals), which keeps aberrations (especially off-axis astigmatism) to a minimum.

Since the incident light is not collimated, the grating introduces wavelength-dependent aberrations into the diffracted wavefronts (see Chapter 7). Consequently the spectrum cannot remain in focus at a fixed exit slit when the grating is rotated (unless this rotation is about an axis displaced from the central groove of the grating⁴²). For low-resolution applications, the Monk-Gillieson mount enjoys a certain amount of popularity, since it represents the simplest and least expensive spectrometric system imaginable.

Figure 6-3.   The Monk-Gillieson mount. A plane grating is used in converging light.

A grating used in the Littrow or autocollimating configuration diffracts light of wavelength l back along the incident light direction (Figure 6-4). In a Littrow monochromator, the spectrum is scanned by rotating the grating; this reorients the grating normal, so the angles of incidence a and diffraction b change (even though a = b for all l). The same auxiliary optics can be used as both collimator and camera, since the diffracted rays retrace the incident rays. Usually the entrance slit and exit slit (or image plane) will be offset slightly along the direction parallel to the grooves so that they do not coincide; this will generally introduce out-of-plane aberrations. True Littrow monochromators are quite popular in laser tuning applications (see Chapter 13).

Figure 6-4.   The Littrow monochromator mount. The entrance and exit slits are slightly above and below the dispersion plane, respectively; they are shown separated for clarity.

Two monochromator mounts used in series form a double monochromator. The exit slit of the first monochromator usually serves as the entrance slit for the second monochromator (see Figure 6-5), though some systems have been designed without an intermediate slit. Stray light in a double monochromator with an intermediate slit is much lower than in a single monochromator: it is approximately the product of ratios of stray light intensity to parent line intensity for each single monochromator.

A double monochromator may be designed to have either additive dispersion or subtractive dispersion.

• In the case of additive dispersion, the reciprocal linear dispersion of the entire system is the sum of the reciprocal linear dispersions of each monochromator: that is, the spectrum that is dispersed by the first monochromator is further dispersion in passing through the second monochromator.

Figure 6-5.   A double monochromator mount.

• In the case of subtractive dispersion, the entire system is designed so that the spectral dispersion at the exit slit of the second monochromator is essentially zero; in this case, the dispersion of the double monochromator system is that of the first monochromator. A subtractive-dispersion monochromator has the property that the light leaving its exit slit is spectrally uniform: the homogeneous combination of all wavelengths is transmitted through the intermediate slit, instead of a spectrum of continuous varying wavelength as seen in single monochromators and additive-dispersion double monochromators. Such instruments have found use in fluorescence and luminescence excitation.⁴⁵

A triple monochromator mount consists of three monochromators in series. These mounts are used only when the demands to reduce instrumental stray light are extraordinarily severe (e.g., Raman spectroscopy).⁴⁶

The vast majority of monochromator mounts are of the constant deviation variety: the grating is rotated to bring different wavelengths into focus at the (stationary) exit slit. This mount has the practical advantage of requiring a single rotation stage and no other moving parts, but it has the disadvantage of being "on blaze" at only one wavelength – at other wavelengths, the incidence and diffraction angles do not satisfy the blaze condition

where q_B is the facet angle.

An alternative design that may be considered is the constant-scan monochromator), so called because in the grating equation

it is the scan angle f rather than the half-deviation angle K that remains fixed. In this mount, the bisector of the entrance and exit arms must remain at a constant angle to the grating normal as the wavelengths are scanned; the angle 2K = a - b  between the two arms must expand and contract to change wavelength (see Figure 2-9).

Constant-scan plane grating monochromators have been designed⁴⁷ but have not been widely adopted, probably due to the complexity of the required mechanisms for the precise movement of the slits. Hunter described a constant-scan monochromator for the vacuum ultraviolet in which the entrance and exit slits moved along the Rowland circle (see Section 7.2 below).⁴⁸ The imaging properties of the constant-scan monochromator with fixed entrance and exit arms have not been fully explored, but since each wavelength remains on blaze, there may be applications where this design proves advantageous. [As noted in Section 2.8, though, the efficiency will drop as |2K| increases, i.e. as the monochromator is used off-Littrow.]

The plane grating monochromator mounts described in Section 6.2 have an exit slit through which a narrow spectral region passes; the center wavelength of this spectra region is changed by rotating the grating. Alternatively, a wide spectral region can be imaged at once by leaving the grating fixed and using a series of exits slits (or an array of detector elements) in a focal plane. Such optical systems are call spectrographs.

Often the imaging properties of a plane grating spectrograph (with no auxiliary optics) are acceptable over only a portion of the spectrum of interest, which requires the use of additional lenses or mirrors to provide additional focusing power to render the focal curve as close to the line (or curve) represented by the slits or detector array.