Instruments that give the spectrum of the radial ion from an object are called SPECTROGRAPHS. The most important part of an optical spectrograph is the element that breaks up the light into its constituent colours. This can be done simply by use of a series of colour filters, hut such an approach is not an efficient way of using the light from faint objects because scarce photons are absorbed by the filters. However, to examine the two dimensional colour structure or spectrum of an object it is necessary to make a series of observations in different colour bands using filters. The filters commonly used are of two main types. Coloured glasses or plastic films generally exclude the light above or below a certain wave¬length and are used to provide broad hand filtering (e.g. UBV FILTERS) of the light. For work in narrower bands (less than 25 nm wide) INTERFERENCE FILTERS are used. These consist of layers of substances with different refractive indices, arranged so that their optical thickness matches the wavelength of the light to be trans¬mitted through the filter. Radiation of different wavelengths in reflected by the many layers (sometimes as many as 17), In this way very narrow bandwidth (<1nm) tilters can be constructed which transmit radiation only within this narrow band. Spectrographs are usually constructed with either a prism or a diffraction grating to disperse the light and so give its spectrum. When light crosses an air-glass boundary its path is deflected (refracted) by an amount which depends on the wave-length or colour of the light. Blue light has a shorter wavelength and is refracted more than red light. This dependence of refractivity on wavelengths means that a prism is able to break up white light into a spectrum. This spectrum is visually similar to a rainbow. DIFFRACTION GRATINGS also produce a spectrum although they work in quite a different way. REFLECTION GRATINGS consist of a flat mirror on which are etched a large number of parallel grooves, often as many as several hundred per centimetre. The light is only reflected from the shiny parts between the scratches, as if it had emerged from a series of thin parallel slits. The light waves reflected by these strips of mirror interfere with one another so that in any one direction all the reflected waves cancel out except for a few particular wavelengths. TRANSMISSION DIFFRACTION GRATINGS work in accordance with the same prin-ciples as a reflection grating, except that the scratches are marked on a sheet of glass. When designing a spectrograph we must remember that it is much easier to interpret the spectrum if we make sure that the light is parallel before it enters the prism or hits the diffraction grating. The light from the star or galaxy to be analysed is focused onto the slit of the spectrograph. This ensures that we know exactly from which part of the object the light in the spectrum is coming. After passing through the slit, the light diverges and must be made parallel (COLLIMATED) by a lens before it is DISPERSED by the diffraction grating. After dispersion the light of various wavelengths is in the form of parallel beams each moving in a slightly different direction. All the light of one colour is gathered and focused by the camera lens to give a clear, focused spectrum of the object. An objective prism is sometimes placed in front of the primary mirror of a telescope. Objective prisms look rather like thin sheets of glass - which, indeed, they are, but their sides are not quite parallel. Because of this they act like a prism, deflecting the light of a star by an amount dependent on the colour of the light from the star. If we take a direct photograph of a star field through an objective prism, the image of each star is slightly elongated. This is because the light from the stars has been spread out into tiny spectra. These spectra are too small to reveal much detail, but they can help with the crude classifications of stellar or galaxy type. The great advantage of objective prism work is that it is possible to record the spectra of a large number of objects simultaneously, whereas a conventional spectrograph will normally be used to observe only one object at a time. A small portion of a plate taken with an objective prism With bright objects it is possible to use grating spectrographs of much higher dispersion. These spread the light from the object further; consequently we can investigate much finer details of the spectrum of the star. With the highest dispersions, however, the spectrum becomes inconveniently long, requiring massive spectro¬graphs and detectors with a very large sensitive area. One arrangement for producing a high-dispersion spectrum which is still com¬pact is the ECHELLE. We have seen how it is that interference effects produce the dispersed beam with a diffraction grating. In the same way that an interferometer has many interference fringes, a diffrac¬tion grating produces many dispersed beams, each in a different direction. By varying the details of the way a diffraction grating is made and is mounted in the spectrograph we can arrange to concentrate much of the light into only a few beams. In conventional grating spectrographs it is usual to concentrate the light into one or two of the first few beams. In the echelle grating configuration however. The starlight is concentrated into many overlapping higher-order beams. In order to separate the different beams they are all passed to another dispersing element (a prism or a diffraction crating) which has its direction of dispersion nearly perpendicular to the first. The light in overlapping beams is of different wave¬length, so that a prism will deflect the light paths differently for each beam. It is possible to arrange that the various beams lie one above the other. The particular format of the echelle spectrum is more convenient than that of conventional high-dispersion spectrographs since it is better suited to the modern detectors described in the next section which usually have sensitive areas that are circular. It is also possible to cover a wider range of wavelengths in a single exposure. The FOURIER SPECTROGRAPH is used increasingly for measuring the spectra of objects, particularly in the near infrared (700 mn-2 Jim wavelength). The way that these work is more difficult to understand. We start by thinking about the basic properties of the light to be analysed. A spectral line has a certain width set by the physical conditions within the object itself. Within that line the light consists of a stream of photons. Two consecutive photons from the line will generally be different, and that difference on average will be greater if the spectral line is wider. If we can find a way of measuring the average difference between two photons we can get information on the width of the spectral line with which they are associated. There is, indeed, a device which can analyse spectral lines in this way: it is called a MICHELSON INTERFEROMETER (named after its inventor, the American physicist, A.A. Michelson). The principle on which it works is identical to the simple radio interferometer. The light from the object is split into two parts that are then directed along paths of different lengths. They are then brought together again in order that the photons can interfere I with one another; i.e. add or subtract, depending on their relative^ phase. As the path-length difference is changed, those photons that are adequately similar will interfere to produce a rapidly fluctuating output signal as their relative phases change. Relatively dissimilar photons will gradually stop interfering with one another as the path-length difference is increased, and will produce a constant output signal. If we record the total output signal from such an interferometer as a function of path-length difference we will find that we have all the information which we need to reconstruct the original spectrum. This is done in a computer using a mathematical function known as the Fourier transform - hence the name of the spectrograph. The big advantages of this method, advantages which justify the considerable complexity of the device are (1) that all the light from the object can be used, since an entrance slit is not needed and (2) that a single detector element is needed rather than a multi-element detector such as a photographic plate. This is of particular importance in the infrared where multi-element detectors are extremely difficult to fabricate.

Comments are closed.