UW Madison Geoscience 777 Electron Probe Microanalysis Scanning

UW- Madison Geoscience 777 Electron Probe Microanalysis Scanning Electron Microscopy Electron Optical Column Updated 1/26/18

UW- Madison Geology 777 What’s the point? We need to create a focused column of electrons to impact our specimen, to create the signals we want to measure. This process is identical for both scanning electron microscope (SEM) and electron microprobe analyzer (EMPA). We use conventional terminology, from light optics, to describe many similar features here.

UW- Madison Geology 777 Generic EMP/SEM Electron gun Column/ Electron optics Optical microscope EDS detector SE, BSE detectors Vacuum pumps Scanning coils WDS spectrometers Faraday current measurement

UW- Madison Geology 777 Key parameters • Beam Energy: Source energy (leaving gun), landing energy (arriving at sample) in k. V/ke. V • Beam diameter (1 nm 1 –micron) (~simplification( • Beam current in p. A n. A m. A (electrons/second) • Beam brightness: current density/unit solid angle (FE source much brighter than W source) • Beam focus: lenses; last cross over (should be) exactly at sample surface • Beam astigmatism: correcting for non-round beam (unsynmetric magnetic fields; dirty apertures)

UW- Madison Geology 777 Key points • Source of electrons: various electron guns (thermoionic and field emission). We want stable current with small beam diameter. • Lenses are used to focus the beam and adjust the current • Beam can be either fixed (point for quant. analysis) or scanning (for images) • Current regulation and measurement essential (more for EP, less for some SEM work, more for other) • Electron Probe: Optical microscope essential to position sample (stage) height, Z axis (= X-ray focus) • Vacuum system essential

Electron Gun source

UW- Madison Geology 777 W, La. B 6, FE sources Most common in e-probes and many SEMs is the W filament, a thermionic type. A W wire is heated by ~2 amps of current, emitting electrons at ~2700 K – thermal energy permits electrons to overcome the work-function energy barrier of the material. $25 -$100@ Anothermionic source is La. B 6, also Ce. B 6), which has added benefits (“brighter”, smaller beam) but it is more expensive (each tip is 5 -10 x cost of W filament) and fragile (sensitive to vacuum problems). Both have very good (~1%) beam stability, compared the field emission (FE) guns, which are “brighter” and have much smaller beams (great for high resolution SEM images) but lower beam stability and require ultra high vacuum. And for FE, add ~$400 K to the price of the SEM or electron probe. Replacement emitter ~$8 -12, 000@

UW- Madison Geology 777 W filament: biased Wehnelt Cap First electron cross-over Current (~2 A) flows thru the thin W filament, releasing electrons by thermionic emission. There is an HV potential (E 0) between the filament (cathode) and the anode below it, e. g. 15 ke. V. The electrons are focused by the Wehnelt or grid cap, which has a negative potential (~ -400 V), producing the first electron cross over. Goldstein Fig 2. 4, p. 27

UW- Madison Geology 777 W filament: W filament is ~125 mm diameter wire, bent into hairpin, spotwelded to posts. W has low work function (4. 5 e. V) and high melting T (3643 K), permitting high working temperature. Accidental overheating will cause quick failure (top right). Under normal usage, the filament will slowly lose W, thinning down to ultimate failure–left, from our SX 51. With care/luck, a filament may last 6 -9 months, though 1 -2 month life is not uncommon. Top 3 images: Goldstein Fig 2. 8, p. 33

Emission current vs probe (Faraday cup) current This shows filament output on the S 3400: emission current IE, which flows from cathode to anode, is high (to 10 -4 A). However, what escapes through the hole in the anode and reaches down the column, is much lower, only 10 -8 A. Most SEMs can only read IE, lacking Faraday cups.

Saturation on the Hitachi S 3400 Since >99% of SEMs do not have Faraday cups, they provide “black box” saturation buttons. But that is not optimal for high resolution imaging, where you need a tight beam. Thus you set the instrument in “filament image” mode and optimize settings to get the tight spot (right), not the “donut” on the left. Goldstein, 2003, p. 32

UW- Madison Geology 777 Resolution Comparison-2 To right is another version, from a JEOL sales brochure. I have two comments: (1) “effective range for analysis by FE is wildly incorrect; no one show believe you can do epma at 100 p. A! Normal currents are 1020 n. A (2) Reed’s figure showing the cross over above 1 n. A needs explanation, maybe older cold FE? From Reed Above: from JEOL 8530 F brochure 2012

UW- Madison Geology 777 Resolution Comparison-1 For the ultimate in high resolution imaging, FE is tops -- if you have the money. Note the importance of reduced beam current (p. A not n. A)--good for imaging, though not good for EPMA. Side note: beam diameter in electron microprobes is incorrectly assumed to be represented by the diameter of the bright CL spot on fluorescent samples. It is not! From Reed

UW- Madison Geology 777

UW- Madison Geology 777 Some electron units/values Brightness is a measure of the current emitted/unit area of source/unit solid area of beam (not used in daily activities) High voltage and Current - Analogies Baseball HV: speed of the ball curr: size of the ball Water through hose HV: water pressure curr: size of the stream of water

UW- Madison Geology 777 Key points • Source of electrons: various electron guns (thermoionic and field emission). We want stable current with small beam diameter. • Lenses are used to adjust the current and focus the beam Bem can be either fixed (point for quant. analysis) or scanning (for images) • Current regulation and measurement essential (more for EP, less for some SEM work, more for other) • Electron Probe: Optical microscope essential to position sample (stage) height, Z axis (= X-ray focus) • Vacuum system essential

Column: focusing the electrons Simple iron electromagnet: a current through a coil induces a magnetic field, which causes a response in the direction of electrons passing through the field. Rotationally symmetric electron lens: beam electrons are focused, as they are imparted with radial forces by the magnetic field, causing them to curve toward the optic axis and cross it. (Goldstein et al, 1992, p. 44)

UW- Madison Geology 777 Producing minimum beam diameter Similar geometry to light optics (though inverted: reducing image size): (opposite of magnification, here showing the “object” size shrunk) Light Optics Electron Optics 1/f = 1/p + 1/q (f = focal distance) d 0 is the demagnified gun (filament) crossover-typically 10 -50 um, then after first condenser lens, it is further demagnified to crossover d 1. After C 2 and objective lens, the final spot is 1 nm-1 um. (Goldstein et al, 1992, p. 49)

UW- Madison Geology 777 Producing minimum beam diameter Source: W filament Similar geometry to light optics (though inverted: reducing image size): Beam on sample: spot frozen in mid-scan across sample below Light Optics Electron Optics 1/f = 1/p + 1/q (f = focal distance) d 0 is the demagnified gun (filament) crossover-typically 10 -50 um, then after first condenser lens, it is further demagnified to crossover d 1. After C 2 and objective lens, the final spot is 1 nm-1 um. (Goldstein et al, 1992, p. 49)

UW- Madison Geology 777 Condenser & Objective Lenses: working distance WD WD Note: we cannot change the working distance on the SX 51; however, this is a critical adjustable parameter on the SEM. Left: shorter working distance (~q 2), greater convergence (a 2): smaller depth of field, smaller spot (d 2), thus higher spatial resolution. Right: longer WD, smaller convergence: larger depth of field, larger spot, decreased resolution. (Goldstein et al, 1992, p. 51)

UW- Madison Geology 777 Condenser Lenses: adjusting beam current Probe current (Faraday cup current, e. g. 20 n. A) is adjusted by increasing or decreasing the strength of the condenser lens(es): a) weaker condenser lens gives smaller convergence a 1 so more electrons go thru the aperture. Thus higher current with larger probe (d 2) and decreased spatial resolution (a 2). b) is converse case, for low current situation. With quant. EPMA we are shooting for high currents, so the left case holds. For SEM work, it depends: for CL and EBSD, the same holds, but for high resolution SE imaging, the right case holds, i. e. drop the current as low as you can get away with. (Goldstein et al, 1992, p. 52)

UW- Madison Geology 777 Key points • Source of electrons: various electron guns (thermionic and field emission). We want stable current with small beam diameter. • Lenses are used to adjust the current and focus the beam • Beam can be either fixed (point for quant. analysis) or scanning (for images) • Current regulation and measurement essential (more for EP, less for some SEM work, more for other) • Electron Probe: Optical microscope essential to position sample (stage) height, Z axis (= X-ray focus) • Vacuum system essential

UW- Madison Geology 777 Scanning Coils The primary mission of the electron microprobe is to focus the beam on a spot and measure X-rays there. However, it was early recognized that being able to scan (deflect) the beam had two advantages: X-rays could be produced without moving the stage, and electron images could be used to both identify spots for quantification, and for documentation (e. g. BSE images of multiphase samples). Later, with the development of the SEM as a separate tool, scanning was essential. Scanning requires 1) deflection coils and 2) display system (CRT) with preferably 3) digital capture ability. Reed 1993, Fig 2. 3, p. 18

UW- Madison Geology 777 Scanning --> 2 D Image The electron probe, a fine point (<1 um) is rapidly scanned across the sample, and the signal from each (x, y) coordinate is mapped onto the screen or a file. Goldstein et al, 3 rd Edition, Fig 4. 4

UW- Madison Geology 777 Key points • Source of electrons: various electron guns (thermionic and field emission). We want stable current with small beam diameter. • Lenses are used to adjust the current and focus the beam • Beam can be either fixed (point for quant. analysis) or scanning (for images) • Current regulation and measurement essential (more for EP, less for some SEM work, more for other) • Electron Probe: Optical microscope essential to position sample (stage) height, Z axis (= X-ray focus) • Vacuum system essential

UW- Madison Geology 777 Probe current: monitoring and stabilization EPMA requires precise measurement of X-ray counts. X-ray count intensity is a function of many things, but here we focus on electron dosage. If we get 100 counts for 10 n. A of probe (or beam or Faraday) current, then we get 200 counts for 20 n. A, etc. Therefore, it is essential that we 1) measure precisely the electron dosage for each and every measurement, and 2) attempt to minimize any drift in electron dosage over the period of our analytical session. The first relates to monitoring, the second to beam regulation.

UW- Madison Geology 777 Probe current monitoring Electron beam intensity must be measured, to be able to relate each measurement to those before and after (i. e. to the standards and other unknowns). This is done with a Faraday cup, where the beam is focused tightly within the center of a small aperture over a drilled out piece of graphite (or metal painted with carbon). Current flowing out is measured. Why graphite? Because it absorbs almost all of the incident electrons, with no backscattered electrons lost. Goldstein et al 1992, Fig 2. 25, p. 65

UW- Madison Geology 777 Probe current regulation Optimally, the beam current should remain as constant as possible, particularly over the duration of each measurement (depends upon number of elements, etc, but most are 45120 seconds). On Cameca electron probes, this is accomplished in a feedback loop with the condenser lenses, where a beam regulation aperture measures the electrons captured on a well defined area (red area on bottom aperture), where larger aperture above it provides ‘shading’ and eliminates ‘excess’ electrons (green). Reed 1993, Fig 4. 12, p. 47

UW- Madison Geology 777 Key points • Source of electrons: various electron guns (thermionic and field emission). We want stable current with small beam diameter. • Lenses are used to adjust the current and focus the beam • Beam can be either fixed (point for quant. analysis) or scanning (for images) • Current regulation and measurement essential (more for EP, less for some SEM work, more for other) • Electron Probe: Optical microscope essential to position sample (stage Z) for X-ray focus • Vacuum system essential

UW- Madison Geology 777 Optical Microscope An essential part of an electron microprobe is an optical microscope. The reason is that we need to consistently verify that all standards and specimens sit at the precise same height (Z position). This is because they must all be in “X-ray spectrometer focus”, which shortly you will find described as the “Rowland circle”. Mounting of specimens relative to an absolute height is problematic, for a variety of reasons -- difficult to mount samples perfectly flat, and the fact that we use different holders and shuttles manufactured to different tolerances, together with different screw tightening by operators.

UW- Madison Geology 777 Back to the first point: • Source of electrons: various electron guns (thermionic and field emission). We want stable current with small beam diameter to get nice sharp images at high magnification.

Beam Diameter and Imaging Resolution The electron microprobe and the SEM have significantly different beam diameters and imaging resolutions, because (1) The probe’s main job is cranking out x-ray counts, and to optimize that, you need lots of beam current (tens of n. A, up to hundreds for trace elements). Also, if operating at 15 -20 k. V, the interaction volume is ~2 -3 microns, so “beam size” in EPMA does not mean analytical spatial resolution*. But (2) The SEM’s goal is to produce sharp images, and you can utilize several features to do that: (a) Introduce small apertures to tighten up the beam diameter (150, 80, 50, 30 microns) (b) Turn down the probe current (p. A) which minimizes scattering (c) Go to high saturation so filament image is tight and centered (d) Go to the shortest working distance possible (e. g. 6 mm) (e) Play with k. V, to find best setting (could be high, could be low) * Exception: now with FE-EPMA, with tighter beams and operation at low ke. V, we do worry!

Resolution Tests One common SEM resolution is defined as “point to point” resolution and is the smallest separation of adjacent particles that can be detected in an image. The manufacturers “cheat” by using optimal images, sputtered gold balls on carbon (image to right), where there is good secondary electron generation and high contrast. It is optimized in being a 3 D image where the secondary electrons show surface well. A flat polished rock thin section would not show such fine scale resolution. Also note that a 3 nm (30 A) resolution does not mean you can see 3 nm gold balls: most of the balls are ~50 nm in size. Another resolution test is scanning across a very sharp edge (e. g. razor blade) and determining the distance between 85 -90% to 15 -10% of the intensity drops off.

UW- Madison Geology 777 Key points • Source of electrons: various electron guns (thermionic and field emission). We want stable current with small beam diameter. • Lenses are used to adjust the current and focus the beam • Beam can be either fixed (point for quant. analysis) or scanning (for images) • Current regulation and measurement essential (more for EP, less for some SEM work, more for other) • Electron Probe: Optical microscope essential to position sample (stage) height, Z axis (= X-ray focus) • Vacuum system essential