For the light which enters through its input slit the monochromator is a sort of traffic agent. The main component of a monochromator is the reflecting grating. There are also some other components which are usually arranged in a Czerney-Turner mounting, like that shown in the picture. The function of the monochromator is to isolate at the exit slit a narrow domain of wavelengths from the incident light. The width of this domain is called bandwidth of the monochromator.
The wavelength of the radiation leaving the exit slit is controlled by a simple rotation of the grating about its vertical axis. In the modern instruments this rotation is performed by a stepper motor, controlled by a computer. In reality the grating doesn't rotate as much as the figure shows, the figure only underlines the selection principle. There are two concave spherical mirrors in the monochromator which are used first to achieve a parallel rendering of the input beam of light and the second to condense the selected wavelengths through the exit slit.
What ? Are you asking me why do we need a hollow cathode lamp if we use the monochromator anyway ? Everybody asks this question, so I have the answer: the bandwidth of the monochromator is much larger than the width of an atomic line. Suppose that the sample absorbs 50% of the radiation having the appropriate wavelength and look at the pictures to see the difference.
The first picture shows the absorption in the hollow cathode case and the second presents the case of a lamp having continuous spectrum. The problem is that in the second case the detector won't make any difference between the light detected in the presence and in the absence of the sample.
Click here to continue.