Peterhouse, Cambridge

Preliminary work

First microscope with sweeping

In 1942, the physicist and Russian engineer Vladimir Zworykin, which worked in the laboratories of the Radio Corporation off America with Princeton with the the United States, published the details of the first electron microscope with sweeping which can analyze an opaque surface and not only to analyze one fine sample in transmission. An electronic gun with tungsten filament emitted electrons which were accelerated under a tension of: 10000 volts. The electronic optics of the apparatus was made up of three electrostatic reels, the reels of sweeping being placed between the first and the second lens. This system gave a very reduced image of the source of about 0.01 µm. Fact enough running at the beginning of the history of the MEB, the electronic gun was in bottom of the microscope so that the room of analysis can be with the good height for the manipulator. But this had an annoying consequence because the sample was thus likely to fall into the column from the microscope. This first MEB reached a resolution of about 50 Nm. But at that time, the electron microscope in transmission developed rather quickly and compared to performance of this last, the MEB caused much less passion and its development was thus slowed down.

Development of the electron microscope with sweeping

At the end of the Years 1940, Charles Oatley, then lecturer in the department of engineering of the Université of Cambridge to the the United Kingdom was interested in the field of electronic optics and decided to launch a research program on the electron microscope to sweeping, in complement of the work carried out on the electron microscope with transmission by Ellis Cosslett, also in Cambridge in the department of physics. One of the students of Charles Oatley, Ken Sander, started to work on a column for MEB by using electrostatic lenses but it had to stop one year afterwards because of the disease. It is Dennis McMullan which resumed this work in 1948. Charles Oatley and itself built their first MEB (called SEM1 for Scanning Electron Microscope 1 ) and in 1952, this instrument had reached a resolution of 50 Nm but what was most important was that it returned finally this narcotic effect of relief, characteristic of the modern MEB.

In 1960, the invention of a new detector by Thomas Eugene Everhart and RFM Thornley will accelerate the development of the electron microscope with sweeping: Detecting Everhart-Thornley. Extremely effective to collect the secondary electrons as well as the retrodiffused electrons, this detector will become very popular and will be found on almost each MEB.

Interaction electron-matter

See also: Interaction radiation-matter

In traditional optical microscopy, the visible Lumière reacts with the sample and the considered Photon S are analyzed by detectors or the human eye. In electronic microscopy, the beam of light is replaced by an electron beam primary educations which comes to strike the surface of the sample and the re-emitted photons are replaced by a whole spectrum of particles or radiations: secondary electrons, retrodiffused electrons, Auger electrons or x-rays. These various particles or radiations bring various types of information on the matter of which the sample is made up.

Secondary electrons

See also: secondary Electron

At the time of a shock between the electron S primary educations of the beam and the Atom S of the sample, a primary education electron can yield part of its energy to a not very dependant electron of the Bande of conduction of the atom, thus provocant a Ionization by ejection of this last. One calls secondary electron this electron ejected. These electrons generally have a weak energy (approximately 50 eV). Each primary education electron can create one or more secondary electrons.

From this weak energy, the secondary electrons are emitted in the surface layers close to surface. The electrons which can be collected by the detectors are often emitted with a depth lower than 10 nanometer S. Thanks to this weak kinetic energy, it is rather easy to deviate them with weak a potential difference. One can thus easily collect a great number of these electrons and obtain images of good quality with a good signal-to-noise report/ratio and a Résolution of about 40 Å (ångström) for a beam of 30 Å diameter. Since they come from the surface layers, the secondary electrons are very sensitive to the variations of the surface of the sample. The least variation will modify the quantity of collected electrons. These electrons thus make it possible to obtain information on the topography of the sample. On the other hand, they give little information on the contrast of phase.

Retrodiffused electrons

See also: Electron retrodiffused

The retrodiffused electrons ( back-scattered electrons in English) are electron S resulting from the interaction of the electrons of the primary education beam with cores of atoms of the sample and which reacted in a quasi elastic way with the atoms of the sample. The electrons are re-emitted in a direction close to their direction of origin with a weak loss of energy.

These recovered electrons thus have a relatively high energy, going up to 30 KeV, and much more important than that of the secondary electrons. They can be emitted with more a great depth in the sample. The resolution reached with the retrodiffused electrons will be thus relatively weak, about the micrometer or the tenth of micrometer.

Moreover, these electrons are sensitive to the Atomic number Atome S constituting the sample. The heaviest atoms (those having a big number of protons) will réémetteront more electrons than the lighter atoms. This characteristic will be used for the analysis in retrodiffused electrons. The formed zones of atoms with a high atomic number will appear more brilliant than others, it is the contrast of phase. This method will be able to make it possible to measure the chemical homogeneity of a sample and will allow a qualitative analysis.

Auger electrons

See also: Electron Auger

When an atom is bombarded by a primary electron, an electron of a deep layer can be ejected and the atom enters an excited state. De-energizing can occur in two different ways: by emitting a photon X (radiative transition) or by emitting a electron Auger (Auger effect). At the time of de-energizing, an electron of a roadbase comes to fill the gap created by the electron initially ejected. During this transition, the outer-shell electron loses a certain quantity of energy which can be emitted in the form of photon X or can then be transmitted to an electron of a orbit more external and thus energy. This outer-shell electron is found in its ejected turn and can be recovered by a detector.

The Auger electrons have a very weak energy and are characteristic of the atom which emitted them. They thus make it possible to obtain information on the composition of the sample and more particularly of the surface of the sample like on the type of Chemical bond, in the measurement obviously where the MEB is equipped with a detector of electrons carrying out a discrimination in energy. In fact specialized MEB are equipped with analyzers in energy. One speaks then about “Auger analysis” or “Auger spectrometry”. The level of vacuum of the electron microscopes Auger must be much better than for the ordinary MEB, about 10^-10 Torr.

X-ray

See also: X-ray

The impact of a primary education electron with high energy can ionize an atom with an internal layer. De-energizing, the filling of the energy order of the electronic structure, occurs with emission of x-rays. the analysis as of these rays makes it possible to obtain information on the chemical nature of the atom.

Instrumentation

Electronic gun

See also: Electronic gun

The Electronic gun is one of the essential components of an electron microscope with sweeping. It is indeed the source of the beam of electron S which will come to sweep the surface of the sample. The quality of the images and the analytical precision that one can obtain with a MEB require that the electronic spot on the sample is at the same time fine, intense and stable. A strong intensity in the smallest possible spot requires a “brilliant” source. The intensity will be stable only if the emission of the source is also.

The principle of the electronic gun is to extract the electrons from a conducting material (which is an almost inexhaustible reserve) towards the vacuum where they are accelerated by an electric field. The electron beam thus obtained is treated by the electronic column which in fact a fine probe swept on the sample.

There exist two families of electronic gun according to the principle used to extract the electrons:

• the thermionic emission, with the Filament S of Tungsten and LaB₆ points;
• the field emission.

According to these distinctions and the operating process, the electronic guns have properties and characteristics different. There exist physical sizes to characterize them. The principal one is the Brillance but the lifespan is also very important, as well as stability. The maximum current available can also be taken into account, as well as energy dispersion.

Brightness of a source

One can define the brightness B of a source by the report/ratio of the current emitted by the source in the product of the surface of the source by the solid angle. In the general case, one can measure only the surface of a “virtual source” which is the zone from where the electrons seem to come. (Definition to be re-examined)

• the diameter of the virtual source D ;
• the emitted current I_e ;
• the half angle of opening α.

In the optical systems, the brightness, which is measured in A.m^-2.sr^-1 (amps per unit of area and solid angle), with the property to preserve itself when the energy of acceleration is constant. If energy varies, the brightness is proportional for him. To obtain an abundant signal of detection when the spot on the sample is very small, it is necessary that the brightness of the source is highest possible. In the literature, one trouvre very often brightness expressed in A.cm ^-2.sr^-1.

Thermionic emission: Tungsten filament and LaB₆ points

Materials such as the Tungsten and the hexaborure of Lanthane (LaB₆) are used because of their weak Output, i.e. of energy necessary to extract a electron from the Cathode. In practice, this energy is brought in the form of thermal energy by heating cathode to a sufficiently high temperature so that a certain quantity of electrons acquires sufficient energy to cross the Barrière of potential which maintains them in the solid. The electrons which crossed this barrier of potential find in the vacuum where they are then accelerated by a Electric field.

In practice, one can use a tungsten filament, formed like a hairpin, which one heats by Joule effect, as in an electric bulb. The filament is thus brought up to an higher temperature with: 2200 °C, typically 2700 °C.

Cathodes in LaB₆ must be heated at a temperature less low but the manufacturing technique of cathode is a little more complicated because LaB₆ cannot be formed in filament. In fact, one hangs a point of monocrystal of LaB₆ to a carbon filament. The Cristal of lanthanum hexaborure is carried in the neighborhoods of 1500 °C to allow the emission of electrons. This cathode requires a vacuum more thorough than for a tungsten filament (about 10^-6 with 10^-7 Torr against 10^-5). Cathodes in hexaborure of Cerium (CeB₆) have very close properties.

The tungsten filament brought up to a temperature of 2700 ° a.c. a typical brightness of 10⁵ Has (cmsr) for a tension of acceleration of 20 kilovolts. It has, at this temperature, one lifespan between 40 and 100 hours. The diameter of the virtual source is about 40 µm

LaB₆ cathode carried at a temperature of 1500 ° a.c. a typical brightness of 10 A.cm ^-2.sr^-1 for one lifespan between 500 and: 1000 hours. The diameter of the virtual source is about 15 µm.

Guns with emission of field

The principle of a gun with emission of field is to use a metal cathode in the shape of very fine point and to apply a tension about: 2000 with: 7000 Volt S enters the point and the anode. One produces thus, by “effect of point”, a very intense electric field, about 10⁷ V.cm ^-1, at the end of cathode. The electrons are then extracted from the point by Tunnel effect. There exist two types of guns to emission of field (English FEG for Field Emission Gun ):

• the emission of cold field (English CFE). The point remains with room temperature:
• the emission of field assisted thermically (English TFE). The point is then brought up to a typical temperature of 1800 °K.

The large advantage of the guns with emission of field is a theoretical brightness which can be hundred times more important than that of LaB₆ cathodes. The second type of gun (assisted thermically) is used more and more, because it makes it possible for a sacrifice in very modest brightness to better control the stability of the emission. The current available is also higher. With a gun with emission of cold field, the current available on the sample is indeed never higher than 1 Na, whereas with the thermal assistance, it can approach the 100 Na.

Another large difference between the guns with emission of field and the guns thermionic is that the virtual source is much smaller. That comes owing to the fact that all the trajectories are normal on the surface of the point, which is a sphere from approximately 1 µm. The trajectories thus seem to come from a point. Thus very high brightnesses are obtained: 10⁸ A.cm ^-2.sr^-1 for the cold cathodes and 10⁷ A.cm ^-2.sr^-1 for cathodes with emission of field heated. On the sample, the brightness is always degraded ! colspan=" 2" | Thermionic emission ! colspan=" 2" | Emission of field |- ! Materials ! Tungsten ! LaB₆ ! S-FEG ! C-FEG |- | Brightness (A.cm^-2sr^-1) | 10⁵ | 10⁶ | 10⁷ | 10⁸ |- | Temperature (°C) | 1.700-2 400 | 1.500 | 1.500 | ambient |- | Diameter of point (Nm) | 50.000 | 10.000 | 100-200 | 20-30 |- | Cut source (Nanomètre) | 30.000-100 000 | 5.000-50 000 | 15-30 | < 5 |- | Current of emission (µA) | 100-200 | 50 | 50 | 10 |- | Lifespan (hour) | 40-100 | 200-1 000 | > 1.000 | > 1.000 |- | Minimal vacuum (Pa) | 10^-2 | 10^-4 | 10^-6 | 10^-8 |- | Short-term stability (%RMS) | <1 | <1 | <1 | 4-6 --> |}

Optical column

Thermionic columns for gun with emission

The function of the electronic column is to produce on the surface of the sample an image of the sufficiently reduced virtual source so that the spot obtained is enough fine to analyze the sample with the required resolution, in the range from the 0,5 to 20 Nm. The column must also contain means to sweep the beam.

As the sources of the guns with thermionic emission have a typical diameter of 20 µm, the reduction of the electronic column must be of at least: 20000, produced by three stages comprising each one a magnetic Lens (See figure above).

The electronic column must also comprise a diaphragm of limitation of opening, because the magnetic lenses should be used only in their central part to have aberrations smaller than the required resolution. The resulting astigmatism, for example of defect of sphericity of the lenses can be compensated by a “stigmator”, but the spherical aberration and the chromatic aberration cannot be corrected.

The sweeping of the spot on the sample results from magnetic fields according to the two transverse directions, X and Y, produced by reels of deflection which are traversed by electric currents. These reels of deflection are located right before the last lens.

When the energy of impact is weak, and that there is an electric field of deceleration close to the sample, the installation of the detector of secondary electrons in space between the last lens and the sample poses problems more and more. A solution then consists in laying out the detector inside the column. Indeed, the electric field which slows down the primary education electrons, accelerates the secondary ions. In English, this type of arrangement is known under the name of in-lens detector or Through-Tea-Lens detector (detecting TTL). In French, one could say “detecting in the column”.

Detector of secondary electrons

See also: Detecting Everhart-Thornley

The detector of secondary electrons or Détecteur Everhart-Thornley was developed with an aim of improving the system of collection used originally by Vladimir Zworykin and which consisted of a phosphorescent screen Photomultiplicateur. In 1960, two students of Charles Oatley, Thomas Eugene Everhart and RFM Thornley, had the idea to add a guide of light between this phosphorescent screen and this Photomultiplicateur. This guide allowed a coupling between the scintillator and the photomultiplier, which improved the performances largely. Invented more than one half-century ago, this detector is today that most frequently used.

A Everhart-Thornley detector is composed of a Scintillateur which emits photons under the impact of electrons with high energy. These photons are collected by a guide of light and are transported towards a Photomultiplicateur for detection. The scintillator is carried to a tension of several kilovolts in order to communicate energy with the detected secondary electrons - it is acted in fact of a process of amplification. So that this potential does not disturb the primary electrons, it is necessary to lay out a grid, left Faraday screen room, to armor the scintillator. In the normal functioning, the grid is polarized with some + 200 Volt S compared to the sample in order to create on the surface of the sample a sufficient to drain the secondary electrons, but enough weak electric field not to create aberrations on the incidental beam.

The polarization of the scintillator to a high tension and the strong electric field which results from it are incompatible with a MEB with weak vacuum: It would then occur an ionization of the atmosphere of the room of consecutive observation to the effect Paschen.

Polarized with 250 volts compared to the sample, (see diagram of left), the grid attracts most of the secondary electrons emitted by the sample under the impact of the primary education electron beam. It is because the electric field generated by the Faraday screen room is strongly dissymmetrical that one can obtain an effect of relief.

When the grid is negatively polarized, typically with - 50 volts (see diagram of right-hand side), the detector pushes back the main part of the secondary electrons whose initial energy is often lower than 10 EV. The Everhart-Thornley detector then becomes a detector of retrodiffused electrons.

Preparation of the sample

Metal samples

Biological samples

The first stage is a stage of Fixation which aims at killing the cells all while endeavouring to preserve the structures of them so that one can observe the sample in a state as close as possible to the alive state. The second stage consists in extracting from the sample the elements intended for the observation. It is not rare to be interested only in one body or a precise element of spéciment, for example the surface of a eye, a élytre, a scale or a Poil of an insect. It is thus necessary often to isolate this part before preparing it for the observation. There exist several techniques to extract these parts. Simplest being a manual dissection or the Dissolution of the soft parts and pulpits.

A requirement with all the biological samples but more particularly samples is cleanliness. The surface of the biological sample to study must contain less impurities possible, to allow a perfect clearness even with important growths. For that, there exist three principal techniques: manual, mechanical or chemical cleaning.

The samples must be absolutely dry and not comprise any water trace. Indeed, the pressure in the room of observation is very low and the water molecules contained in the sample would be likely to destroy the cells while evaporating or to pollute the room of observation. There exist also various methods to reach that point according to the nature of the biological sample: air drying, by skirting of the critical point or chemical dehydration.

Once cleaned, dried, made conducting, the sample is ready to be gone up on the slide is placed in the room of observation.

Various types of imageries

Imagery in secondary electrons

The detection of the secondary electrons is the traditional mode of observation of the morphology of surface. The collected secondary electrons come from a narrow volume (approximately 10 Nm). In fact, the zone of réémission makes about the same diameter as the beam. The resolution of the microscope is thus the diameter of the beam, that is to say approximately 10 Nm. A grid placed in front of the detector of electrons, positively polarized (200-400 V), attracts the electrons. In this manner, the majority of the secondary electrons is detected whereas the retrodiffused electrons, which have more raised energy, are almost not deviated by the electric field produced by the grid of the collector. The quantity of produced secondary electrons does not depend on the chemical nature of the sample, but of the angle of incidence of the primary education beam with surface: the more the incidence is shaving, the more excited volume is large, therefore more the production of secondary electrons is important, from where a topographic effect of Contraste (a slope appears more “luminous” that a dish). This effect is reinforced by the fact that the detector is located on the side; the electrons coming from the faces located “back” at the detector are reflected by surface and thus arrive in more minor amount at the detector, creating an effect of shade.

Imagery in retrodiffused electrons

Imagery in electron diffraction retrodiffused

For detailed articles, to see Electron diffraction retrodiffused and Theory of diffraction on a crystal

Coupled to a sensor CCC, detector EBSD is composed of a phosphorescent screen which is directly in the room of analysis of the microscope. The sample is tilted in direction of the detector and the angle compared to the electron beam primary educations is about 70 °. When the electrons come to strike the surface of the sample, they penetrate it on a certain depth and are diffracted by the crystallographic plane according to an angle $\backslash \; theta\_B$ whose value is given by the Loi of Bragg:

$d\_\; \{HKL\}$ represents the Distance interréticulaire, $\backslash \; lambda$ the Wavelength and the Integer $n$ the order of diffraction.

Diffraction is done on 360 ° and each diffracting plan creates a “cone of diffraction” whose top is at the point of impact of the electron beam primary educations. There thus exists as many cones of diffraction of diffracting plans. Spacing between these various cones is, via the law of Bragg, connected to the distance between the crystalline plans.

The slope of the sample and the position of the phosphorescent screen are such as these cones come to strike the screen. The electrons make scintillate the phosphorescent screen and can be detected by camera CCC. On the screen, these portions of truncated cones appear in the shape of lines. The stereotype of diffraction which one obtains is a superposition of alternate dark bands with bands of stronger intensity than one calls lines of Kikuchi . These lines, their various points of intersection and their spacings, can be, by knowing the distance from the screen to the sample, converted into angles and one can thus determine the cell parameters.

With this method and because of great slope of the sample, the space resolution is very asymmetrical: about 1 µm laterally but about 50 to 70 µm longitudinally.

Imagery while running of sample

Apart from this example of the junctions p-n, the imagery while running of electrons is particularly adapted to locate defects (for example a specific Défaut) of a Crystal lattice which appear then in the shape of points or lines blacks, a heterogeneity of doping.

Elementary chemical imagery by spectrometry of x-rays

See also: dispersive Analysis in wavelength, dispersive Analysis in energy

The energy of the X-rays emitted at the time of the de-energizing of the atoms depends of their chemical nature (they are the characteristic lines). By analyzing the spectrum of x-rays, one can have an ultimate analysis, i.e. to know which types of atoms are present. The beam sweeping the screen, one can even draw up a chemical cartography, with however a resolution much lower than the image in secondary electrons (about 3 μm).

The analysis can be done by dispersion wavelength (WDS, wavelength dispersive spectroscopy ), it is the principle of the Microsonde of Castaing invented in 1951 by Raymond Castaing, or dispersion of energy (EDS, energy dispersive spectroscopy ). The technique using the wavelengths is more precise and allows quantitative analyzes whereas that using energy is faster and less expensive.

In dispersion of energy the detection of photons X is carried out by a detector made up of a crystal diode of Silicium doped in Lithium on the surface.

This crystal is maintained at the temperature of the Azote liquid to minimize the electronic noise, and thus to improve the resolution in energy and thus the spectral resolution. The detector is protected by a window in Béryllium to avoid its icing at the time of a contact with the ambient air.

Vacuum measurement partial

Analysis X in this mode remains possible.

Applications

Micro-electronics, technology of the semiconductors and microfabrication

The marketing microsopes electronic with sweeping is about contemporary take-off of the industry of the Semi-conducteur S. It is in this sphere of activity that the MEB was spread most massively, being recognized like an invaluable tool in the development of the manufactoring processes of the devices whose characteristic element, the grid of transistor passed from a typical width of some micrometers at the end of the Années 1960 to less than 100 nanometers to XXIe century. Not only the MEB made it possible to see beyond the limits of the optical microscope, but the vision in relief proved very practical for the assistance with the microfabrication where it is often important to control the verticality of the layers deposited or the engraved layers. See, for example, on the figure opposite, an image of MEB of a reason for photorésine engraved.

Very popular in the research laboratories and development, the MEB also became a very widespread tool in the production units manufacture, as a tool for industrial control. The room of analysis must then be able to accept silicon wafers (in English, wafer ) whole, i.e. whose diameter is, in 2006, of 200 mm or 300 Misters One even gave a name particular to the apparatuses which carries out dimensional check, i.e., which checks the width of a line. In English, one calls them CD-SEM . These apparatuses are entirely automated: they strictly speaking do not produce images: the calculator of control brings a reason for test exactly on the beam axis which is then swept in only one direction. The signal of the detector of secondary electrons is recorded and analyzed for générérer the measured width. If this one is in-outside gauge given, alarm is given, and it silicon wafer, is regarded as bad can rejected.

Another application of the MEB in the production units of semiconductors is the characterization of microparticles which contaminate the surface of the sections: the final goal is to identify the cause of the contamination in order to remove it. The particle whose size can vary from 100 Nm with 1 µm was detected by a machine of specialized inspection which communicates the coordinates of the particles with the MEB of analysis. This one is then used at the same time in the imagery mode, to produce an image of the particle with high magnification and in Microsonde of Castaing, which implies that the MEB is equipped with a spectrometer X. the image can help with the identification of the particle, but it is especially the chemical characterization resulting from the analysis in wavelength of the x-rays which will give a track making it possible to go back to the cause of the contamination.

The electronic probe of a MEB can be used not to observe, but to write and manufacture. It is then about Lithographie to electron beam

Science of materials

To obtain certain figures of diffraction (peudo-Kikuchi, Kossel), one is brought to pervert the system of sweeping of the instrument: instead of generating a sweeping in rectangular mode, one excites reels of deflection in order to make swivel the beam of several degrees around a fixed point of the sample. The generated image is then a figure of diffraction corresponding to a zone of the sample of some micrometers.