"Prism Spectrometers or Diffraction Grating Spectrometers:

Which One Should I choose?"

For spectral imaging applications on a microscope there are three major reasons to select a prism:

  1. If you need to capture very weak signals you cannot beat the 90% light transmission of a prism!
  2. The best efficiency you can expect from a DG is about 60% at just one wavelength - it is all downhill from there!
  3. Diffraction gratings (DG) actually Pollute the spectrum of a sample over the wavelength range from 365 to 920 nm! DGs just cannot help dumping useful signal into second order!

Diffraction Grating Facts
About second order pollution
All grating based instruments used for hyperspectral imaging work in first order. The problem is that superimposed second order contaminates first order producing "ghost" spectra, and elevated background noise.

Certain "spectral" features can appear "twice;" once in first order, and then again in second order. The only way to prevent this is to add "filters" that cut-out the blue and UV. To learn more about second order click here.

Second order diffraction hits you two ways:

  1. First order light is diluted when it is also dumped into second order
  2. When second order diffraction is superimposed onto first order it acts as a contaminant and has to be filtered out with even more loss of light!

Grating efficiency
Diffraction gratings have to be "blazed" (highest efficiency) at the blue end of the spectrum or there will be inadequate efficiency to capture wavelengths shorter than blaze. Efficiency drops with increasing wavelength.

Grating ghosts
If the grating is classically ruled it will produce "ghosts" of residual laser bleed-through or other strong emission.

Grating Non-Linearity
Contrary to the sales claims of some spectral instrument vendors; the wavelength dispersion of a grating is absolutely NOT linear.

Wavelength dispersion varies with the cosine of the angle of diffraction and path length. The wavelength range from 365 to 920 nm results in a significant change in the angle of diffraction; and with it the value of the cosine. In fact, it is so easy to demonstrate that it makes a very good lab in a physics 101 course.

The path length varies with wavelength across all CCD or linear array detectors. If the focal plane is also tilted, when needed to "flatten" the focal plain, then there will be increased non-linear wavelength dispersion.

Spectral resolution
To see a more detailed review of spectral resolution issues click here.

Wavelength calibration
To see how to perform wavelength calibration, and determine wavelength accuracy and view a videodescription click here.

To learn more about the physics of imaging spectrometer systems download
Imaging Spectrometer Fundamentals for Researchers in the Biosciences - a Tutorial

Now the Good News about the PARISS "Curved Prism"

The PARISS Imaging Spectrometer Offers
Spectral Integrity, High Resolution, and
Unbeatable Light Throughput

PARISS is a unique imaging spectrometer that delivers:

  • High light throughput (~90%) from ~420 nm through 1000 nm
  • An unbeatable wavelength range from 365 to 920 nm in a single shot!
  • High spectral resolution (~1.5 nm)
  • High spatial resolution due to the curved sides of the prism
  • Lowest possible scattered light for maximum signal to noise ratio

All prisms present inherently low scattered light because their surface area is orders of magnitude less than the very best diffraction grating!

The spectrum below is a perfect illustration of the excellence of the PARISS design. This spectrum was acquired with a single 10 ms PARISS acquisition using a Q-Imaging Retiga 2000R as the wavelength detector. The above spectrum is a tribute to both the light-transfer efficiency of the spectrometer and the camera (it is not electron multiplied, such as an EMCCD). We could argue that we observed EMCCD performance at a fraction of the price.

Figure 1: A Hg spectrum emitted by a wavelength calibration lamp and acquired by a PARISS Analytical Spectral Imaging System. The scan presents both outstanding spectral resolution (~1.5 nm FWHM at the 436 nm line) and spectral range. The PARISS system uses a prism as the wavelength dispersive element; consequently, there are no higher orders to pollute the spectrum. (Diffraction gratings commingle higher order diffraction)

Many imaging spectrometer systems find it either difficult, or impossible to capture any Ar lines above 650 nm. This is because Ar lines fade rapidly to zero as a Hg/Ar lamp warms up (<10 sec.). Although the fading Ar lines makes it hard to acquire in the red it is just as tough in the UV. This is because the Hg 365-436 nm lines only become bright after the Ar has faded. The net result is that only the most efficient spectrometers and detectors can hope to grab light at both 365 and 920 nm, simultaneously, in a single acquisition.

In other words, diffraction grating based instruments cannot hope to
accurately reproduce the spectrum shown in Figure 1!

Spectral resolution
To see a more detailed review of spectral resolution issues click here.

Wavelength calibration
To see how to perform wavelength calibration, and determine wavelength accuracy and view a video description click here.

New! After eleven incredible years read the PARISS update!

Contact LightForm to request a PARISS demonstration. Go to top

"There is nothing worse than a sharp image of a fuzzy concept." Ansel Adams
PARISS removes the fuzzies!______________________________________________________
© 1990-2008 LightForm, Inc all rights reserved

--------------------------------------------
LightForm, Inc.,
601 Route 206, Suite 26-479
Hillsborough NJ 08844
Tel: (908)281 9098
Email: jlerner@lightforminc.com