How Many Electrons Do You Need For An EBSD Pattern?

Matt Nowell, EBSD Product Manager, EDAX

I always liked the commercial that asked,” How many licks does it take to get to the center of a Tootsie Pop?”. I like contests where you estimate the number of M&Ms in a jar. Taking the concept away from delicious treats and moving towards something more technical, I’ve also enjoyed looking at the number of grains we need to measure with EBSD to get a good idea of the texture of a material.

Recently I’ve been working with our new Clarity™ Direct Electron Detector for EBSD. It’s the first commercial EBSD direct detector and will be launching soon. Traditionally, EBSD patterns are captured when the diffracted electrons strike a phosphor screen, where energy is converted into light photons, which are focused through a lens onto an imaging sensor, where the light photons are then converted back to electrons. However, a direct electron detector is just that, it captures the diffracted electrons directly. This allows us to count the electrons in an EBSD pattern directly.

EBSD pattern collected with Clarity™ with an average of 5,000 electrons per pixel.

Figure 1. EBSD pattern collected with Clarity™ with an average of 5,000 electrons per pixel.

Take the EBSD pattern collected from a nickel superalloy using the Clarity™ shown in Figure 1. For an EBSD pattern like this, remember that it has been background corrected to flat-field the image and improve the contrast. This is because the actual live EBSD pattern does not have a uniform intensity across the sensor, as shown in Figure 2. In this example, a background collected while imaging many grains was collected and subtracted from the live signal to produce the image in Figure 1. The background image has the spatial information for a specific orientation removed, while retaining the overall intensity gradient that is a function of the material of interest and the sample geometry. Note that the Clarity™ uses four direct electron detectors that are coupled together. The cross-hair image visible in Figure 2 shows the location of the seams between the detectors. These can be masked out of the image if desired but are quickly minimized with this background correction.

EBSD pattern from Figure 1 prior to background correction.

Figure 2. EBSD pattern from Figure 1 prior to background correction.

For Figure 1, a pixel at the center of the signal intensity contained approximately 10,000 electrons, and the average counts for all pixels was approximately 5,000 electrons. After background subtraction, I drew a line across the image, and the intensity profile across this line is shown in Figure 3. This profile shows that the final processed EBSD pattern has a dynamic range of about 1,700 electrons.

Line profile across the EBSD pattern in Figure 1 showing the dynamic range of the EBSD signal.

Figure 3. Line profile across the EBSD pattern in Figure 1 showing the dynamic range of the EBSD signal.

EBSD pattern with an average of 10 electrons per pixel.

Figure 4. EBSD pattern with an average of 10 electrons per pixel.

Now seeing that I could count the number of electrons in an EBSD pattern, I wanted to know how many I needed to get a usable EBSD pattern. I could decrease the exposure time, decrease the beam current, or do both. In this case, I continually decreased the exposure time to find where the EBSD pattern indexing started to fail. Figure 4 shows an EBSD pattern where the maximum number of electrons is 20 and the average number of electrons is 10. Even with this small amount of a signal, I was still able to index it with a confidence index of 0.92 and a fit of 0.6°, which indicates a good orientation solution. Talk about doing a lot with a little. This performance is enabled by the single electron sensitivity and zero readout noise of the detector, which makes this camera very exciting for low beam dose applications for beam-sensitive materials. I look forward to sharing more later.

Indexing solution for the pattern in Figure 4 with a confidence index of 0.92.

Figure 5. Indexing solution for the pattern in Figure 4 with a confidence index of 0.92.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.