Examining a High Intensity Electron Gun Used Cathode with Optical Microscopy

We make it a habit to examine samples sent to us with optical microscopy even if the requested analysis is x-ray photoelectron spectroscopy (XPS), thermal analysis, electrochemistry, or FTIR spectroscopy.  While optical microscopy is not the newest of laboratory techniques, it provides relatively easily obtained valuable information.  It allows us to document the appearance of a sample so that other materials characterization results can be matched with certain material appearances over time.  The next similar sample sent by a customer for analysis may not be as similar as thought.  One of most frequent things we learn is whether there are two or more chemical or structural phases present.  There may be phases present which optical microscopy will not reveal, but it does very often make it clear that multiple phases are present which the eye may not have perceived or perceived in such detail.  This is very important to know when interpreting the quantitative elemental composition determined by XPS or EDX (Energy Dispersive X-ray analysis) or various phase change features in a thermogravimetry, differential scanning calorimetry (DSC), or thermomechanical analysis.

In the optical microscopy images below, we will examine a cathode which has seen 25 months of service in a high intensity electron gun used in an x-ray source.  This tungsten cathode consists of a thin circular foil mounted on three legs and doped with cesium to lower its work function.  A filament consisting of a serpentine etched circular tungsten foil sits behind it and produces a flood of electrons which are accelerated at 12 KeV into the back side of this cathode.  The most common failure mode is that the tungsten filament separates at a grain boundary between grains, which often grow with usage until a single grain traverses the entire current path and the grain boundaries become the strong electron scatterers or the high resistance points until failure occurs at a grain boundary.  Another failure mode has been that one of the spot welds holding the three legs of the cathode fails and the cathode shorts out to the extractor lens of the electron gun.  In the case of this cathode/filament assembly, the failure was the rupture of the filament.

The cathode shown below was imaged with an inspection microscope with illumination at an angle of about 30 degrees relative to the surface normal.  This provides an image much like what the eye sees commonly, though at a much lower magnification.

This surface is highly reflective, so much of the detail is lost.  But, we do get a sense that many irregularities and perhaps chemical phases or surface roughness result mostly near the three legs of the cathode.  If we place the illuminating light at an angle of about 65 degrees to the surface normal, we get a better image which makes it clear that there is great variability of surface condition on this used cathode.  These are also very much enlarged grains with usage etched grain boundaries.

Looking at the cathode at higher magnification we see striation patterns more clearly:

Are these bright lines due to chemical phases or are they ridge lines of raised surface areas?  Let us make use of the metallographic microscope with light normal to the surface of the cathode.

This microscope does not offer the depth of field of the inspection microscope, so we cannot see the three legs of the cathode any more.  But, we do see that there is a great variation in surface reflectivity and this may create a great variability in local surface electron emission in use.  We can make use of a Nomarski interference phase contrast feature of this microscope to differentiate areas of about the same surface height from those of different heights.  This mode inserts a prism at the objective lens and uses two polarized films to create images similar to those below:

The areas of a given color are at the same distance from the objective lens, so we can learn more about the surface topography now.  If the material being imaged were partially transparent to visible light, we would have to be prepared for possible phase contrast due to materials of different index of refraction, which would probably be due to different chemical phases.  This tungsten is to reflective for that to be of concern here.  The presence of the colored patterns over a given grain do mean that there is surface topography on that grain.  Let us look at a higher magnification at parts of this cathode surface:

At twice the prior magnification:

The upper image of these last two was a relatively non-reflective area near a leg and the very surface rough grain in the lower left corner of the lower image was also near a leg.  Note the very wide grain boundary on the right side of the large blue-green grain in the lower image also.  Such a wide grain boundary in the filament would likely become a point of separation and failure.  The surface roughening is probably due to one of these reasons:

• Impurities from the cooler legs migrate to the cathode surface near them and promote the surface ridge lines and other roughened topography.
• Grains near the legs are slightly cooler and impurities incident upon the surface have a higher probability of sticking on the surface and promote surface roughening.
• The surface roughening is a function of a certain range of temperatures due to a phase change of the tungsten crystal structure which occurs in a certain temperature window.  Tungsten is usually said to have the bodied centered cubic (bcc), which is its thermodynamically stable phase.  It also has a beta phase of simple cubic structure which grows as rods under oblique incidence sputtering of tungsten onto a substrate.  Perhaps some impurities may stabilize the simple cubic structure within a certain temperature window.

The last image above seems to show a large single grain whose top left corner is breaking away, perhaps due to the spread of an impurity from the highly roughened grain in the lower left corner.  It would be interesting to perform XPS and EDX analysis of this cathode to gain further understanding of the mechanisms for the roughening of some grain surfaces.  Would we detect impurities responsible for this?

Nomarski phase contrast interference sure can produce some colorful and beautiful images!  They are also useful.