Enhance the Contrast of Your Analysis

Having the highest resolving microscope is ineffective without contrast between the phases of your material. See how filtering your images in energy space can enhance image contrast to reveal a wealth of information about your material.

It is possible to improve the resolution and contrast of images plus diffraction patterns to generate unique contrast effects when using energy-filtering with a transmission electron microscope (TEM).

Contrast enhancement and resolution enhancement

Electrons that are scattered inelastically degrade the image contrast as a result of the chromatic aberration of the TEM objective lens. Only one electron energy can be precisely in focus and the resultant blur (\(d\)) at the specimen is given by the equation

\(d=C_{c}*\beta \left ( \Delta /E_{0} \right )\)


  • \(C_{c}\) = chromatic aberration constant
  • \(\beta\) = acceptance angle of the objective aperture
  • \(\Delta\) = range of energies contributing to the image
  • \(E_{0}\) = primary energy

Blur will be especially large for thick, low Z samples. For Z that is less than or equal to 12, the inelastic scattering quickly dominates over elastic scattering. Reducing the energy window with the energy filter will reduce the effects of the chromatic aberration.

Elastically scattered electrons only

When you only allow the elastically scattered electrons to contribute to the image (or diffraction pattern), this removes the inelastic fog or blur. For thin to medium-thick, unstained or lightly stained biological samples, use only the elastically scattered electrons to form the optimal image. As the sample becomes thicker, most of the electrons are inelasticity scattered and there is very little signal in the elastic image. In this case, use of the most probably loss electrons is recommended.

Elastically scattered electrons only
Unstained/osmicated sample (thickness = 0.4 µm, 100 keV).

Filtering is crucial for quantitative convergent beam electron diffraction (CBED) analysis.
Filtering is crucial for quantitative convergent beam electron diffraction (CBED) analysis.


Inelastic scattered electrons only

When only some of the inelastically scattered electrons contribute to the image, this will improve the contrast and resolution in images of very thick samples. This method can also be used to improve contrast in unstained or lightly stained biological samples.

Inelastic scattered electrons only
Image is of a HL 60 cell with a thickness of approximately 1 µm at 200 keV. The unfiltered image has very poor resolution due to chromatic aberration. While the elastic image has improved resolution, but very low intensity. The 120 eV loss image shows both good resolution and intensity, thus allowing fine structures to be identified.

Specimen thickness

Specimen thickness can be determined when you consider that each scattering event in the sample is independent and thus governed by Poisson statistics with a mean scattering length in the sample that is material dependent. By comparing the intensity of the zero­-loss peak (ZLP) to that of the entire spectrum, you can determine the probability that an electron can pass through the sample with no inelastic scattering events. This probability is given by:The three regimes of an energy loss spectrum

\(P_{o}=exp\left ( -t/\lambda \right )=I_{o}/I\)

and therefore the thickness

\(t=-\lambda ln\left (I _{o}/I \right )\)


  • \(\lambda\) =mean free path for inelastic scattering
  • \(I\) = total spectrum integral
  • \(I_{o}\) = ZLP integral

This is particularly easy for EFTEM where only two images are required, the elastic image, \(I_{o}\), and the unfiltered image, \(I\).

Low-loss contrast tuning

By imaging with a specific band of energies in the low-loss part of the spectrum, you can tune the image contrast to show specific material phases or features. This method is not typically quantitative, but it provides a method to rapidly survey the sample to enumerate and locate the phases or defects present. Further quantitative analysis can then determine the nature of these phases or defects.

Contrast tuned images
Contrast tuned images from a semiconductor sample. Left 8 ±2 eV EFTEM image. Right 19 ±2 eV EFTEM image. Different materials are highlighted in each energy range.