SPECTROSCOPY & STELLAR SPECTRA

 

 

 

Spectrographs

 

Prism

Why a slit?

 

Without one, you would get overlapping images at every wavelength it emitted light at. This is a “flash” spectrum of the Sun during a solar eclipse. Here the Moon is thankfully limiting how much of the Sun we see. But notice that in some cases, you can see an entire silhouette! If the Moon were not there, we would have overlapping Suns at every wavelength of light emitted.

Gratings

 

Light hitting a diffraction grating produces a number of individual spectra For a plain flat reflection or transmission grating, the brightest one is white, and little if any color is seen, This is the zeroth-order spectrum. On either side one can see a nice rainbow of color in the first-order spectra. Further out, one sees second-order spectra which are more dispersed than the first-order, but fainter, and so on.

 

The angle where each spectrum is sent is given by

 

where m is the order, d is the separation of the “apertures” (slits/grooves), and λ is the wavelength. The more closely spaced the grooves are (grooves/mm), the broader the spectrum in each order. NOTE that light with a wavelength of 440 nm in the second order (m=2) is sent to the same place as 880 nm light in the first order (m=1)!!! To keep spectra from overlapping, one often uses order blocking filters.

 

 

Placing most of the light in the zeroth-order, where the wavelengths aren’t separated wastes most of the light. By tilting the facets on a reflection grating, a process called blazing, one can push the maximum into the m=1 order, the m=2 order, etc. A grating blazed for the 1^st order will have one of its 1^st order spectra get the lion’s share of the light.

One often has to worry about the spectral purity, that is, what range of wavelengths Δλ are actually hitting a certain portion of your detector.

 

Finally, how well-separated the wavelengths are is given by the spectral resolution

 

 where N is the total number of grooves in the grating. Basically, Δλ is the fineness of wavelength that you can “see” clearly.

 

 

Some “interesting” compact designs:

 

 

 

 

 

 

Echelle Spectrographs

 

Echelle gratings are interesting beasts. They often operate in high orders (m=80-100), producing numerous overlapping spectra. These can be separated using a ‘cross-disperser”, another grating with grooves perpendicular to the echelle grating (running parallel to the echelle’s spectra).

 

What if there are many stars in a field that need to be observed. Must you do each individually? NO!!

 

Integral Field Spectroscopy

 

Here the field of view is covered with “lenslets” connected to optical fibers, or an array of mirrors in order to make an “image slicer”, each of which sends the light to a spectrograph. This is a recent development, and an improvement over a similar concept of placing individual “pickoff mirrors” and fibers  in the focal plane of the telescope:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Now, a swath of mirrors basically can cover the entire field! Nothing is missed!

 

They can even be ganged together!

 

In this way a series of spectra, one for each object, can be produced.

 

 

Okay, so now let’s look at the actual process of the classification of stellar spectra .

PDF - See: http://spiff.rit.edu/classes/phys440/lectures/spec_class/spec_class.html