198
14 Spectrography
14.2 The Spectrograph
In order to obtain spectra, we must disperse the light from the observed object such
that photons of the same wavelength fall in the same location in our imaging device.
There are two ways of doing this: by refraction using a prism and by diffraction
using a diffraction grating. There is a third device that uses both methods, known as
a grism. The grism is a prism with a grating etched on its surface.
It is also possible to undertake limited spectrography using narrowband filters.
As these are tuned to a single spectral line, such as [OIII] or [SII], it is possible to
measure the intensity of this line, although not its width.
The diffraction grating is a reflective or transmissive surface etched with regular
lines, often over a thousand lines per millimetre. The lines in effect turn the source
into multiple sources, one per line. These constructively and destructively interfere,
creating multiple rainbow patterns on the detector.
Equation (14.1) shows the diffraction equation, where d is the line spacing in the
grating, θ is the emergence angle, n is an integer known as the order, and λ is the
wavelength:
d(sin θ) = nλ.
(14.1)
As we can see when n = 0, the dispersion angle θ becomes zero. The 0th-order
image, therefore, is not a strictly a spectrum, but rather the image of the object. To
either side of the object is a series of repeated spectra, becoming fainter as we move
away from the 0th order. The first of these spectra, on either side of the 0th order,
are the first-order spectra, which are typically the spectra that will be analysed. For
reflection gratings, (14.1) becomes
d(sin θ i + sin θ e ) = nλ,
(14.2)
where the term sin θ is replaced by the sum of the sines of the incident angle θ i and
the angle of refraction, θ e .
Figure 14.1 shows the layout of a traditional long-slit spectrometer as found at
many small observatories. Light enters the spectrograph through a slit that is designed
both to improve the spectral resolution and to isolate the source. Often, the size of
the slit can be adjusted. Once inside the spectrograph, the light is collimated, so that
the incident angle is consistent. The reflection grating disperses the light depending
on wavelength. Typically, the grating can be rotated so that the area of the spectrum
of interest will fall on the centre of the CCD. Likewise, some gratings are double
backed with the option of a high- or low-resolution grating. The dispersed light
passes through the camera’s optical system and forms a spectrum on the CCD. Most
spectrographs will have a slit illuminator to allow the alignment of the source and
slit. They may also have a calibrated low-intensity light source such as a small argon
or neon bulb. These will produce a strong reference line in the spectra for calibration
purposes. Of course, if you are observing in a light-polluted area, the sodium doublet
lines are ideal for calibration.
14 Spectrography
14.2 The Spectrograph
In order to obtain spectra, we must disperse the light from the observed object such
that photons of the same wavelength fall in the same location in our imaging device.
There are two ways of doing this: by refraction using a prism and by diffraction
using a diffraction grating. There is a third device that uses both methods, known as
a grism. The grism is a prism with a grating etched on its surface.
It is also possible to undertake limited spectrography using narrowband filters.
As these are tuned to a single spectral line, such as [OIII] or [SII], it is possible to
measure the intensity of this line, although not its width.
The diffraction grating is a reflective or transmissive surface etched with regular
lines, often over a thousand lines per millimetre. The lines in effect turn the source
into multiple sources, one per line. These constructively and destructively interfere,
creating multiple rainbow patterns on the detector.
Equation (14.1) shows the diffraction equation, where d is the line spacing in the
grating, θ is the emergence angle, n is an integer known as the order, and λ is the
wavelength:
d(sin θ) = nλ.
(14.1)
As we can see when n = 0, the dispersion angle θ becomes zero. The 0th-order
image, therefore, is not a strictly a spectrum, but rather the image of the object. To
either side of the object is a series of repeated spectra, becoming fainter as we move
away from the 0th order. The first of these spectra, on either side of the 0th order,
are the first-order spectra, which are typically the spectra that will be analysed. For
reflection gratings, (14.1) becomes
d(sin θ i + sin θ e ) = nλ,
(14.2)
where the term sin θ is replaced by the sum of the sines of the incident angle θ i and
the angle of refraction, θ e .
Figure 14.1 shows the layout of a traditional long-slit spectrometer as found at
many small observatories. Light enters the spectrograph through a slit that is designed
both to improve the spectral resolution and to isolate the source. Often, the size of
the slit can be adjusted. Once inside the spectrograph, the light is collimated, so that
the incident angle is consistent. The reflection grating disperses the light depending
on wavelength. Typically, the grating can be rotated so that the area of the spectrum
of interest will fall on the centre of the CCD. Likewise, some gratings are double
backed with the option of a high- or low-resolution grating. The dispersed light
passes through the camera’s optical system and forms a spectrum on the CCD. Most
spectrographs will have a slit illuminator to allow the alignment of the source and
slit. They may also have a calibrated low-intensity light source such as a small argon
or neon bulb. These will produce a strong reference line in the spectra for calibration
purposes. Of course, if you are observing in a light-polluted area, the sodium doublet
lines are ideal for calibration.