Scanning Electron Microscopy

The scanning electron microscope has an electron beam that scans across the sample at variable energy levels from 1 kiloelectron volt (KeV) to 30 KeV (Leute 1987, Parkes 1986: 185, Goodhew et al. 2001: 122). Since the wavelength of an electron beam is smaller than that of light, which is used by conventional microscopes, it gives better depth of field and resolution, which can almost give the sample a three-dimensional appearance (Goodhew et al. 2001: 2-3, Williams 1994: 159). This means that the ‘hills’ and the ‘valleys’ of a sample can be in focus at the same time, which is important when working with a porous material such as faience (Freestone 1985: 67-68). The magnification of the SEM can go up to x200,000 but it is normally used within the range of x10 to x5,000 (Caple 2006: 195).

The electron beam is controlled and produced by an electron gun assembly, which requires careful alignment of several components including the electron gun, the condenser lenses, the scan coils and the detectors (Figure 5.1) (Goodhew et al. 2001: 3). The electron beam passes through the anode hole, and then through a series of magnetic ‘lenses’ (actually coils). These lenses de-magnify the beam to a specific diameter (spot size), which illuminates a corresponding area of the sample being analysed (Tite 1972: 246). The condenser lenses usually control the beam diameter, and this is particularly important because it affects the resolution of the microscope as it deflects the beam over the surface of the sample, during the scan (Chescoe and Goodhew 1990: 16, Tite 1981: 200). The samples have to be conductive, and are often covered in a thin layer of carbon to limit charging, which can happen when the sample is bombarded with electrons and can cause interference. Also the whole system needs to be under vacuum, which is necessary

for the electrons not to be scattered by the gas molecules in the air (Goodhew et al. 2001: 20, Leute 1987: 123).

The interaction of the beam with the samples results in the emission of several different signals, such as secondary electrons, backscatter electrons, and x- rays (Goodhew et al. 2001: 125). The SEM is primarily an imaging method, the images are greyscale, and have good depth of field (Ponting 2004: 166). There are two primary imaging modes: secondary electron image (SEI) and backscattered election image (BEI) (Potts 1995: 380, Tite 1972: 246). The distinction between the two types of electrons is the angle of incidence of the primary electron beam. Both types of electron are counted through a scintillator, which sends the electrical signal produced to the display as an image (Chescoe and Goodhew 1990: 1-2, Goodhew et al. 2001: 130).

Figure 5.1: The SEM (adapted from Chescoe and Goodhew 1990).

5.2.1.1 SECONDARY ELECTRON IMAGING

Secondary electron imaging (SEI) consists of the secondary electrons ejected from the atoms of the sample as a result of primary electrons colliding with these atoms (Caple 2006: 197, Parkes 1986: 186, Tite 1972: 247). It is these

electrons that form the secondary electron reading and image. Those electrons with lower energy than the primary electrons escape the surface of the sample, and it is the overall number of these secondary electrons that are counted to form the image, providing a detailed picture of the microstructure of the surface (Parkes 1986: 186). The number depends also on the surface topography and the angle of the primary beam to the surface of the sample (Leute 1987: 123). This method of imaging occurs when the secondary electrons collide with a phosphor or scintillator, which then emits light. The light is converted into photons by a scintillator-photomultiplier system, also known as the Everhart-Thornely detector (Chescoe and Goodhew 1990: 20, Goodhew et al. 2001: 129). The photomultiplier then converts the photons into electron pulses, which are then amplified by the cathode ray tube (CRT; see Figure 5.1).

This system is usually built into the SEM and is the most commonly used imaging mode, because it has excellent depth of field for microstructure analysis (Goodhew et al. 2001: 129). SEI is utilised to study the surface texture (microtopography) of the sample, and, as stated earlier, this is dependent on the angle between the specimen surface and the beam. The images that are generated are in greyscale, with varying shades of grey, without any sharp boundaries, and are reflections of the angle of the surface to the detector (Freestone 1985: 67-68). Alternatively, backscatter electron images provide variations in the greyscale based, on the atomic number of the elements present (Henderson 2000: 18).

5.2.1.2 BACKSCATTER ELECTRON IMAGING

When the electron beam hits the surface of the sample, the primary electrons are deflected by the charge on the atoms, and bounce back. These are the so-called ‘backscattered electrons’ (Caple 2006: 197, Goodhew et al. 2001: 75, Parkes 1986: 185). The backscatter electrons are reflective of the atomic number of the atoms that they interact with; heavy elements deflect better than lighter elements. The number of electrons increases with the atomic number of the element present.

The backscatter electron information is compiled by a scintillator detector (e.g. Robinson et al. 2004), attached to the SEM, to form an image based on the number of the backscattered electrons and the ways in which they vary over the surface analysed (Ponting 2004: 168). The higher the atomic number of the sample area, the lighter/brighter it appears in the image (with variation in greyscale; see Goodhew et al. 2001: 75-76). Copper is almost white in greyscale, whereas silica is grey, and the voids appear black. The detector is normally located to one side of the

sample because the backscattered electrons travel in straight lines (Henderson 2000: 18). The detector is often retractable because it can restrict the working distance of the SEM and interferes with the detection of the x-rays (Goodhew et al. 2001: 130).

This method of imaging is very useful when it comes to faience analysis, and is complimentary to SEI and EDS. BEI encompasses both of the other signal types received from the SEM (SEI and EDS). It provides an image of the microstructure of the sample, and it also provides an indication of the chemical composition of the surface of the material (Henderson 2000: 19, Freestone 1985: 68, Ponting 2004: 168, Tite 1972: 248). The latter is very useful for studying faience, as it depicts the differences between the glaze, interparticle and core layers of faience, based on the atomic numbers of the elements present. Since copper has a higher atomic number than silica, the copper-rich glaze appears brighter than the silica. However, this is only an image, which is useful for determining glazing method and viewing the microstructure of the sample. EDS analysis still needs to be conducted in order to determine the specific elements that are present, and in what quantities.