an instrument that is used to observe and photograph greatly magnified images of objects and that employs a beam of electrons accelerated to high energies under high-vacuum conditions instead of a beam of light. The magnification of an electron microscope may be as high as 106. The electrons are accelerated to energies of 30–100 kiloelectron volts or higher.
The physical foundations of particle-beam optical instruments were established in 1834—that is, nearly 100 years before the development of the electron microscope—by W. R. Hamilton, who discovered analogies between the transmission of light beams in optically inhomogeneous media and the trajectories of particles in fields of force. The feasibility of developing the electron microscope became evident after the hypothesis of de Broglie waves was advanced in 1924. The technological prerequisites were created by H. Busch, the German physicist who investigated the focusing properties of axisymmetric fields and who devised a magnetic electron lens in 1926.
In 1928, the German scientists M. Knoll and E. Ruska initiated the development of the first magnetic transmission electron microscope (TEM); three years later, they obtained an image produced by an electron beam. The first scanning electron microscopes (SEM’s) were built by M. von Ardenne (Germany) in 1938 and by V. K. Zworykin (USA) in 1942. In a SEM, a finely focused electron beam, or electron probe, moves sequentially from point to point over an object. By the mid-1960’s, a high level of technical perfection had been achieved in SEM’s, and such electron microscopes began to be widely used in scientific research.
Transmission electron microscopes. TEM’s have the highest resolving power, surpassing light microscopes by a factor of several thousand with respect to resolution. In TEM’s, the limit of resolution, which characterizes the ability of an instrument to yield individual images of closely spaced fine details of an object, amounts to 2–3 angstroms (Å). Under favorable conditions, individual heavy atoms may be photographed. When periodic structures—such as the lattice planes of a crystal—are photographed, a resolution of better than 1 Å may be obtained. Such high resolutions are achieved owing to the extremely short de Broglie wavelength of electrons (seeDIFFRACTION OF PARTICLES). For a sufficiently small diffraction error, the spherical aberration of the objective, an aberration that affects the resolving power, can be reduced by means of optimum focusing. No effective methods have been found for correcting the aberrations in electron microscopes (see). Therefore, in TEM’s, magnetic lenses, which have the smallest aberrations, have completely supplanted electrostatic lenses. TEM’s are manufactured for various purposes. They may be divided into the following three groups: high-resolution TEM’s, simplified TEM’s, and high-voltage TEM’s.
HIGH RESOLUTION TRANSMISSION ELECTRON MICROSCOPES. High-resolution TEM’s have a resolving power of 2–3 Å. As a rule, they are general-purpose instruments. Additional equipment and attachments may be used for various purposes, including the following: to tilt a specimen in various planes at large angles to the optical axis; to heat, cool, or deform the specimen; to carry out X-ray diffraction analysis; and to perform research using electronographic methods. The accelerating voltage, which may be as high as 100–125 kilovolts (kV), is adjusted in steps and is characterized by high stability. Over a period of 1–3 min, the accelerating voltage varies by not more than one or two millionths of the initial value.
A diagram of a typical high-resolution TEM is presented in Figure 1. A high vacuum—that is, a pressure of down to 10–6 mm Hg—is created in the optical system, or the microscope column, by means of a special vacuum system. A schematic diagram of a
Figure 1. A transmission electron microscope: (1) electron gun, (2) condenser lenses, (3) objective, (4) projector lenses, (5) light microscope for additional magnification of the image observed on the screen, (6) barrel with viewing ports through which the image may be observed, (7) high-voltage cable, (8) vacuum system, (9) control panel, (10) stand, (11) high-voltage power unit, (12) lens power supply
TEM column is shown in Figure 2. The source of the electrons is a hot cathode. The electron beam is formed in an electron gun and is then focused twice, first by an upper condenser and then by a lower condenser. As a result, a small electron spot is produced on the specimen; the diameter of the spot may be adjusted over the range from 1 to 20 micrometers (μm). After passing through the specimen, some of the electrons are scattered and are removed from the beam by the objective aperture. The unscattered electrons pass through the objective aperture and are focused by the objective in the object plane of an intermediate lens, where the first magnified image is formed. Projector lenses produce a second image, a third image, and so forth. The last projector lens forms an image on a fluorescent screen, which emits light when bombarded by the electrons. The final magnification is equal to the product of the magnifications of all the lenses.
The extent and the nature of the electron scattering are not the same at various points of a specimen because the thickness, density, and chemical composition of the specimen vary from point to point. Consequently, the number of electrons removed from the beam by the objective aperture after passing through different points of the specimen also varies. As a result, the current density in the image varies. The variation of the current density is converted into contrast on the screen.
A cassette holder and photographic plates are located below the screen. When photographs are taken, the screen is removed and the photographic emulsion is exposed by the electrons.
In a high-resolution TEM, the image is focused by smoothly varying the current that generates the magnetic field of the objective. The currents of the other lenses are adjusted in order to vary the magnification of the microscope.
Figure 2. Schematic diagram of a transmission electron microscope column: (1) tungsten-wire hairpin cathode, which is heated to 2800°K by a current passing through it; (2) Wehnelt cylinder; (3) anode; (4) upper (short-focal-length) condenser, which produces a reduced image of the electron source; (5) lower (long-focal-length) condenser, which focuses the reduced image of the electron source on the specimen; (6) specimen; (7) objective aperture; (6) objective; (9), (10), and (11) projectorlens system; (12) cathodoluminescent screen, on which the final image is formed
SIMPLIFIED TRANSMISSION ELECTRON MICROSCOPES. In research that does not require a high resolving power, simplified TEM’s are used. Such TEM’s contain one condenser and two or three lenses for image magnification. They are simpler in design than high-resolution TEM’s and are characterized by a lower and less stable accelerating voltage, usually 60–80 kV. The resolving power ranges from 6 to 15 Å. Simplified TEM’s are also used for instructional purposes and to carry out preliminary examinations of specimens and routine investigations. The specimen thickness that can be penetrated by the electron beam depends on the accelerating voltage. Specimens with a thickness of 10 to several thousand angstroms are studied in 100-kV electron microscopes.
HIGH-VOLTAGE TRANSMISSION ELECTRON MICROSCOPES. Specimens that are two to three times thicker than those examined in standard TEM’s are studied in high-voltage TEM’s, which have an accelerating voltage of up to 200 kV. The resolving power of a high-voltage TEM may be as high as 3–5 Å. To provide for stability and to ensure that the instrument can withstand the high voltages, the electron gun has two anodes, one of which is supplied with an intermediate potential that amounts to half the accelerating voltage. The magnetomotive force of the lenses is greater than that in a 100-kV TEM, and the lenses are larger and heavier.
Figure 3. A very-high-voltage electron microscope: (1) housing in which an insulating gas (sulfur hexafluoride) is pumped to a pressure of 3–5 atmospheres; (2) electron gun; (3) accelerating tube; (4) high-voltage power-supply capacitors; (5) condenser lens unit; (6) objective; (7), (8), and (9) projector lenses; (10) light microscope; (11) control panel
Very-high-voltage electron microscopes. Very-high-voltage electron microscopes (Figure 3) are large instruments with a height of 5 to 15 m and an accelerating voltage of 0.5–0.65,1–1.5, or 3 megavolts. Contained in special enclosures, such electron microscopes are used to study specimens with a thickness of 1–10 μm, or 104–105 Å. The electrons are accelerated in an electrostatic accelerator, which is located in a housing filled with a pressurized
Figure 4. A scanning electron microscope: (1) electron-gun heat shield, (2) hairpin hot cathode, (3) Wehnelt electrode, (4) anode, (5) system of two condenser lenses, (6) aperture, (7) two-stage deflecting system, (8) objective, (9) aperture, (10) specimen, (11) secondary-electron detector, (12) crystal spectrometer, (13) proportional counter, (14) preamplifier, (15) amplifier unit, (16) and (17) equipment for X-ray detection, (18) amplifier unit, (19) magnification adjustment unit, (20) horizontal scanning unit, (21) vertical scanning unit, (22) and (23) cathode-ray tubes
insulating gas. A stabilized high-voltage power supply is located in the same housing or in an additional housing. Work is in progress on the development of a very-high-voltage electron microscope with a linear accelerator, in which the electrons are accelerated to energies of 5–10 megaelectron volts. In the study of thin specimens, the resolving power of a very-high-voltage electron microscope is lower than that of a TEM. In the case of thick specimens, the resolving power is ten to 20 times higher than that of a 100-kV TEM.
Scanning electron microscopes, SCANNING ELECTRON MICROSCOPES WITH A HOT CATHODE. Thick specimens are studied at a resolution of 70 to 200 Å by means of SEM’s with a hot cathode. The accelerating voltage may be adjusted over the range from 1 to 30–50 kV.
The design of a SEM is shown in Figure 4. Two or three electron lenses are used to focus a narrow electron probe on the surface of a specimen. Magnetic deflection coils scan the probe over a given area on the specimen. When the probe electrons interact with the specimen, several types of radiation are produced (Figure 5), including secondary electrons, backscattered electrons, electrons that pass through the specimen (if it is thin), X-ray bremsstrahlung, characteristic X rays, and light. Each type of radiation may be detected by an appropriate collector containing a sensor that converts the radiation into electric signals. After being amplified, the electric signals are supplied to a cathode-ray tube (CRT) and modulate the CRT display. The CRT and the specimen are scanned in synchronism, and a magnified image of the specimen is observed on the CRT display. The magnification is equal to the ratio of the scan amplitude in the display to that at the specimen. The image is photographed directly from the CRT display. The primary advantage of the SEM is the instrument’s high information output, which is due to the possibility of observing the image by using the signals from the various sensors.
SEM’s may be used for such purposes as X-ray diffraction analysis, the study of microrelief, the investigation of the distribution of chemical composition over a specimen, and the examination of p-n junctions. A specimen is usually studied without being subjected to preliminary preparation. SEM’s are also used in industrial processes, for example, in the inspection of defects in micro-circuits.
A resolution that is high for a SEM is obtained when an image is formed by means of secondary electrons. The resolution is determined by the diameter of the zone from which the secondary electrons are emitted. In turn, the size of the zone depends on such factors as the diameter of the electron probe, the properties of the specimen, and the velocity of the primary beam electrons. When the primary electrons penetrate to a large depth, the secondary processes, which develop in all directions, increase the diameter
Figure 5. Diagram showing the recording of information about a specimen in a scanning electron microscope: (1) primary electron beam, (2) secondary-electron detector, (3) X-ray detector, (4) back-scattered-electron detector, (5) visible-radiation detector, (6) detector for electrons that pass through the specimen, (7) instrument for measuring the electric potential induced in the specimen, (8) instrument for measuring the current due to electrons that pass through the specimen, (9) instrument for measuring the current due to electrons that are absorbed by the specimen
of the zone; as a result, the resolving power decreases.
A secondary-electron detector consists of a multiplier phototube and an image tube, the main component of which is a scintillator with two electrodes. One of the electrodes is a grid that is held at a positive potential of up to several hundred volts. The other electrode is an accelerating electrode; it furnishes the collected secondary electrons with the energy needed to excite the scintillator. A voltage of about 10 kV is applied to the accelerating electrode, which is usually an aluminum coating on the surface of the scintillator. The number of scintillations is proportional to the number of secondary electrons emitted from a given point of the specimen. After being amplified in the multiplier phototube and in a video amplifier, the signal modulates a CRT display. The signal level depends on the topography of the specimen, the presence of local electric and magnetic microfields, and the value of the secondary emission coefficient. In turn, the secondary emission coefficient is a function of the chemical composition of the specimen at a given point.
Backscattered electrons are detected by a semiconductor (silicon) detector. The image contrast is due to the dependence of the backscattering coefficient on the angle of incidence of the primary beam and on the atomic number of the substance of which the specimen is composed. The resolution of a backscattered-electron image is lower than that of a secondary-electron image, sometimes an order of magnitude lower. Because the electrons travel in straight lines to the collector, information about individual regions with no straight paths to the collector is lost; that is, a shadow is formed.
Characteristic X rays are detected either by a crystal spectrometer or by an energy-dispersive detector, which is a semiconductor X-ray detector that usually consists of pure silicon doped with lithium. When a crystal spectrometer is used, X-ray photons are first reflected by the crystal and then detected by a proportional counter. When an energy-dispersive detector is used, the signal from the detector is amplified by a low-noise preamplifier and is further amplified by an amplifier system; the preamplifier is cooled by liquid nitrogen to reduce the noise. The signal from a crystal spectrometer modulates a CRT display, and an image showing the distribution of some chemical element over the surface of the specimen is displayed.
SEM’s are also used to carry out local X-ray quantitative analysis. An energy-dispersive detector detects all elements from Na to U with a high sensitivity. By means of a set of crystals with various distances between planes (seeBRAGG-VULF CONDITION), a crystal spectrometer covers the range from Be to U.
A considerable disadvantage of the SEM is the long scan period when investigating a specimen. A relatively high resolving power may be obtained by using an electron probe with a sufficiently small diameter. In this case, however, the probe current is reduced. As a result, the influence of the shot effect, which reduces the signal-to-noise ratio, increases markedly. To prevent the signal-to-noise ratio from falling below a given level, the scanning frequency should be reduced so that a sufficiently large number of primary electrons and, consequently, secondary electrons is collected at each point of the specimen. Therefore, a high resolving power is obtained only at low scanning frequencies. Sometimes, a single frame is formed in 10–15 min.
SCANNING ELECTRON MICROSCOPES WITH A FIELD-EMISSION GUN. SEM’s with a field-emission electron gun have a resolving power of down to 30 A, which is high for a SEM. As does a field-emission microscope, a field-emission electron gun uses a cathode in the form of a pointed tip. At the end of the tip, a strong electric field is generated that draws electrons from the cathode (seeFIELD EMISSION). The brightness of an electron gun with a field-emitting cathode is 103–104 times higher than that of an electron gun with a hot cathode, and the probe current is greater. Therefore, a SEM with a field-emission gun is capable of rapid scanning; to increase the resolving power, the probe diameter is reduced. However, a field-emitting cathode operates stably only in an ultrahigh vacuum, that is, at a pressure of 10–9–10–11 mm Hg. As a result, the design and operation of SEM’s with a field-emission gun are more complicated than the design and operation of SEM’s with a hot cathode.
Scanning transmission electron microscopes. The scanning transmission electron microscope (STEM) has a resolving power as high as that of the TEM. In the STEM, a field-emission gun is used to provide a sufficiently high current in a probe with a diameter of 2–3 Å. A schematic diagram of a STEM is presented in Figure 6.
Figure 6. Schematic diagram of a scanning transmission electron microscope: (1) field-emission cathode; (2) intermediate anode; (3) anode; (4) deflecting system for beam adjustment; (5) illuminator aperture; (6) and (8) deflecting system for electron probe scanning; (7) long-focal-length magnetic lens; (9) objective aperture; (10) magnetic objective; (11) specimen; (12) and (14) deflecting system; (13) ring collector for scattered electrons; (15) collector for unscattered electrons (removed when a spectrometer is used); (16) magnetic spectrometer, in which the electron beams are bent through 90° by a magnetic field; (17) deflecting system for separating electrons with different energy losses; (18) spectrometer slit; (19) collector; (SE) secondary-electron flux; (hv) X rays
In a STEM, two magnetic lenses reduce the diameter of the probe. A central detector and a ring-shaped detector are located below the specimen. Unscattered electrons enter the central detector; after the corresponding signals are converted and amplified, a bright-field image appears on a CRT display. Scattered electrons, which produce a dark-field image, are collected by the ring detector. Specimens thicker than those studied in TEM’s can be investigated in STEM’s because an increase in the number of inelastically scattered electrons with increasing thickness does not affect resolution; in STEM’s, the optical system does not extend beyond the specimen. An energy analyzer is used to separate the electrons that pass through the specimen into an elastically scattered beam and an inelastically scattered beam. Each beam enters a separate detector; the corresponding image, which contains additional information about the scattering properties of the specimen, is observed on a CRT.
In STEM’s, a high resolution is obtained with slow scans because the current in a probe with a diameter of only 2–3 Å is too low for rapid scans.
Hybrid electron microscopes. By combining the principle of image formation with a fixed beam, as in a TEM, and the principle of scanning a fine probe over a specimen, the advantages of the TEM, the SEM, and the STEM may be obtained in a single instrument. At the present time, all TEM’s allow for the observation of specimens in the scanning mode by means of condenser-objective lenses, which produce a reduced image of the electron source; the reduced image is scanned over a specimen by deflecting systems. In addition to an image formed by a fixed beam, images are produced on CRT displays by scanning with the use of, for example, electrons that pass through the specimen, secondary electrons, or characteristic X rays. The optical system of such a TEM extends beyond the specimen and makes it possible to operate in modes that are not feasible in other instruments. For example, at the same time, an electron diffraction pattern may be observed on a CRT display and an image of the specimen may be viewed on a screen.
Emission electron microscopes. An emission electron microscope produces an electron image of a specimen by means of electrons that are emitted by the specimen when it is heated, bombarded by a primary electron beam, illuminated, or immersed in a strong electric field that draws electrons from it. Such electron microscopes are usually employed for highly specialized purposes.
Electron mirror microscopes. An electron mirror microscope is used mainly to visualize the charge pattern and magnetic micro-fields on the surface of a specimen. The main optical element of such an instrument is an electron mirror; the specimen itself serves as one of the electrodes. The specimen is biased slightly negative with respect to the cathode of the electron gun. The electron beam is aimed at the mirror and is reflected by the field in the immediate vicinity of the surface of the specimen. The mirror forms a reflected-beam image on a screen. Microfields near the specimen’s surface redistribute the reflected-beam electrons, producing image contrast that makes the microfields visible.
Future prospects. Increasing the resolution in images of aperiodic objects to 1 Å or better will make it possible to resolve not only heavy atoms but also light atoms and to visualize the organic world on the atomic level. The construction of electron microscopes with a comparable resolution requires, for example, higher accelerating voltages, the development of electron lenses with small aberrations, and the development of methods for correcting the aberrations of such lenses. Of special interest are cryogenic lenses, in which the effect of superconductivity at low temperatures is used.
The study of the formation mechanism for the frequency-contrast response curves of images in electron microscopes has resulted in the development of methods of image reconstruction that are implemented in the same way as in conventional optics. Comparable methods in conventional optics are based on Fourier transforms, and the pertinent calculations are performed by means of computers.
REFERENCES
Eighth International Congress on Electron Microscopy. Canberra, 1974.
Stoianov, P. A., V. V. Moseev, K. M. Rozorenova, and I. S. Renskii. “Elektronnyi mikropskop predefnogo razresheniia EMV-100L.” Izv. AN SSSR: Seriia fizicheskaia, 1970, vol. 34.
Hawkes, P. Elektronnaia optika i elektronnaia mikroskopiia. Moscow, 1974. (Translated from English.)
Derkach, V. P., G. F. Kiiashko, and M. S. Kukharchuk. Elektronnozondovye ustroistva. Kiev, 1974.
Stoianova, I. G., and I. F. Anaskin. Fizicheskie osnovy metodov prosvechivaiushchei elektronnoi mikroskopii. Moscow, 1972.
Oatley, C. W. The Scanning Electron Microscope. Cambridge, 1972.
Grivet, P. Electron Optics, 2nd ed. Oxford, 1972.P. A. STOIANOV