Optical microscopes had been around for centuries, and while you can still find them in classrooms across the country, their dependence on light had become a problem. Light's tendency to diffract, or bend around the edges of optical lenses, limits the magnification capability and resolution of optical microscopes.
As a result, scientists began to develop new ways to examine the microscopic world around them and, in , produced the world's first transmission electron microscope TEM. This instrument directs a beam of electrons through the sample under observation and then projects the resulting image on a fluorescent screen.
Since development of TEMs was well under way by the time SEMs came along, the latter were initially considered unnecessary. It took the unwavering resolution of C. Oatley, a professor of engineering at Cambridge University, to move the newer microscope forward.
Working closely with several of his colleagues and graduate students, Oatley was able to demonstrate both the SEM's magnification potential and the astonishing 3-D quality of images it produced. Today, SEMs are routinely used in tasks like inspecting semiconductors for defects or exploring how insects work.
Next, was the scanning transmission electron microscope STEM , which was developed by Manfred Ardenne in the late s, followed by the development of the prototype of the modern SEM by Vladimir Zworykin in the early s. Electron microscopes surpassed the limitations of optical microscopes and dramatically improved the resolution so that it is possible to observe objects as tiny as an atom. In addition to improvements in resolution, many enhancements are still being made to the electron microscope.
One is the development of the environmental-scanning electron microscope which maintains the sample chamber at low vacuum for observation of samples containing moisture. The analysis covered the limitations on probe diameter due to lens aberrations and the calculation of the current in the probe.
He also showed how detectors should be placed for bright-field and dark-field STEM and for imaging a solid sample in an SEM, and considered the effects of beam and amplifier noise on imaging.
In order to fulfil the Siemens contract he built the first STEM and demonstrated the formation of probes down to 4 nm diameter. But in the short time available he was limited to employing existing technology, and because there was no suitable low-noise electronic detector available, he used photographic film to record the image. Consequently there was no immediately visible image.
A schematic of the microscope column is shown in fig. Immediately below the sample was a drum around which was wrapped the photographic film. The image was recorded by rotating the drum and simultaneously moving it laterally by means of a screw while the currents in the deflection coils were controlled by potentiometers mechanically coupled to the drum mechanism.
Since the image was not visible until the film had been developed, focussing could only be accomplished indirectly by using the stationary probe to produce a shadow image of a small area of the sample on a single-crystal ZnS screen which was observed through an optical microscope and prism system.
The recordings were inferior to those from the TEM that was being constructed by Ruska and von Borries at Siemens, and the hoped-for advantages of STEM with thick samples were not fulfilled. He spent a short time trying to use the instrument in the SEM mode on bulk samples, but only low resolution images could be obtained because of the detector problem: the sample current was amplified by thermionic tubes and a large probe current was needed.
He did not publish any images. In total, von Ardenne worked for less than two years on scanning electron microscopy before concentrating on the development of his universal TEM von Ardenne ; and then, with the start of the war, on building a cyclotron and isotope separators for nuclear energy projects. If he had been able to continue, there is little doubt that he would have built an efficient SEM within a year or two: this is evidenced by a patent von Ardenne which included a proposal for double-deflection scanning, two papers von Ardenne a,b , and a book von Ardenne Two of the chapters in the book were on scanning microscopy and were based on the papers but included additional material relating to imaging the surfaces of solid samples.
Most importantly he proposed a detector using an electron multiplier with beryllium copper dynodes see fig. Measurements of the secondary emitting ratio of beryllium copper and its stability when exposed to the atmosphere were first reported only in by I. In his book he also discussed the interaction between the beam electrons and the sample and suggested that back-scattering would cause a loss of resolution, illustrating this with a diagram which has a quite modern look fig.
He argued that the incident beam electrons produce secondary electrons at or near the surface from an area approximately equal to the beam diameter and give a high resolution image "nutzbare Strahlung" ; the beam electrons then penetrate the sample and a proportion of them are back-scattered and reach the surface where they produce further secondaries.
The back-scattered electrons are emitted from an area of diameter comparable with the penetration depth, and the secondaries they produce "schadliche Strahlung" may impair the resolution however he did not consider the case of a sample with small inclusions below the surface. He concluded that good resolution might be obtained either with a very low energy beam 1 keV , or with one having a high energy 50 keV.
In the first case the back- scattered electrons would emerge from an area of the surface little larger than the incident beam and the resolution would be unaffected. On the other hand with a keV beam the secondary electrons would be produced by the back-scattered electrons over a very much larger area; moreover they would be evenly distributed so that their main effect would be to increase the background reduce the contrast rather than to affect the resolution.
Von Ardenne's scanning microscope was destroyed in an air raid on Berlin in , and after the war he did not resume his work in electron microscopy but researched in other fields: first in Russia and from in Dresden, which was then in the D. Additional information about von Ardenne's scientific work has been given in his autobiography von Ardenne and by McMullan This work was done in parallel with the development of a TEM and by the same staff, in particular Hillier, Ramberg, Vance and Snyder as well as Zworykin himself.
Although Zworykin had every microscope paper from Germany translated as soon as it was received Reisner he was apparently not influenced by von Ardenne's work on the SEM. Instead he started by, in effect, repeating Knoll's beam scanner experiments using a "Monoscope"; this was a pattern-generating cathode-ray tube which had been developed by RCA for television use Burnett and was very similar to Knoll's apparatus.
He then built an SEM based on the Monoscope but with two magnetic lenses to produce a very small focussed probe, and a demountable vacuum system so that the sample could be changed Zworykin et al. The scan rate was the U.
He next tried to obtain a high current in a smaller probe by the use of a field emission gun with a single-crystal tungsten point Zworykin et al.
To obtain a sufficiently high vacuum he had to return to having the gun and the sample in a glass envelope which was baked and sealed-off. A single magnetic lens was used, and fleeting images were obtained at X magnification with scanning at TV rate and a thermionic tube amplifier.
Stable images could no doubt have been achieved but at that time a practical microscope would not have resulted because demountable UHV techniques did not exist. To overcome the noise problem Zworykin therefore decided to build a SEM with an efficient electron detector and a slow scan. The detector was the combination of phosphor and photomultiplier that Everhart and Thornley used nearly twenty years later in an improved form.
In order to bring the secondary electrons to it he designed an electrostatic immersion lens which retarded the beam electrons and accelerated the secondaries. The electron beam leaving the gun was accelerated to 10 keV in the intervening electron optics.
The accelerated secondary electrons diverged as they passed through the 4th electrostatic lens and hit the phosphor screen with an energy of 9. In the first instrument the scanning was done by moving the sample relative to the beam electro-mechanically using loudspeaker voice coils and later hydraulic actuators; it was only in the final version that magnetic scanning of the beam was employed.
The scan time was fixed at 10 min by the facsimile recorder which was used for image recording and which also controlled the microscope scans. There was no provision for a faster scan and the production of a visible image on a TV monitor; this seems strange remembering Zworykin's TV background, but it may have been because the signal bandwidth was limited by the decay time of the phosphor as was found in later work at Cambridge McMullan The focus setting was found by maximising the high frequency components in the video waveform observed on an oscilloscope, a method which was originally proposed by von Ardenne b.
Although the intention was to produce contrast by differences in the secondary emission ratio of the surface constituents, and the incident beam energy of eV was chosen with this in mind, contamination of the surface in the rather poor vacuum prevented meaningful compositional contrast being obtained. Actually, all of Zworykin's published micrographs were of etched or abraded samples, and contrast was topographic Zworykin et al.
The quality of the recorded images was rather disappointing, and together with the lack of a visible image must have been a factor in RCA deciding to discontinue the project.
But another reason was undoubtedly the excellent results that were, as mentioned in the Introduction, being obtained with replicas in TEMs. The Cambridge Scanning Electron Microscopes. Apart from a theoretical analysis of resolving power by a French author Brachet , no other work on the SEM had been reported by The feeling among electron microscopists appeared to be that it was not worth further consideration in view of the apparent failure of the RCA SEM - if such an experienced team were unsuccessful it was very unlikely that anyone else could produce an effective instrument; a notable exception to this general opinion was Gabor It was then that Charles Oatley at the Engineering Laboratories of the University of Cambridge decided that another look at the SEM might be worthwhile, although, as he has related, "several experts expressed the view that this [the construction of an SEM] would be a complete waste of time" Oatley et al He has explained at some length the reasons that brought him to this decision but the main technological ones were that "Zworykin and his collaborators had shown that the scanning principle was basically sound and could give useful resolution in the examination of solid surfaces" and "improvements in electronic techniques and components had resulted from work during the war" Oatley He was also of the opinion that the RCA detector had a low efficiency and only a small proportion of the secondaries were reaching it, with the result that the images were noisy in spite of the long recording time.
Independently of von Ardenne he proposed an electron multiplier with beryllium-copper electrodes Allen having been promised one by Baxter of the Cavendish Laboratory who was building multipliers of this type Baxter Sander; he had abandoned it at an early stage and had changed the subject of his research project to electron trajectory plotting Sander I converted it an STEM, and then to an SEM, by the addition of scan coils, the electron multiplier detector and a long persistence cathode-ray tube monitor McMullan It was still far from clear how Zworykin's results might be improved on.
A higher incident beam energy could be expected to be beneficial but it was not clear how image contrast would be formed. With Knoll and Ruska leading the way, other researchers quickly joined in the development effort.
In Brussels Ladislaus L. Marton made a primitive electron microscope to study the photoelectric effect, and went on to produce the first micrograph of a biological specimen. Manfred Von Ardenne in Berlin produced the earliest scanning-transmission electron microscope in Ruska at Siemens in Germany produced the first commercial electron microscope in the world in The initial projections that a handful of electron microscopes would saturate the worldwide market proved to be grossly pessimistic, and many companies entered the fray.
RCA was by far the leader in North America, with its electromagnetic lens technology, technological expertise, and corporate support for electron microscope development. General Electric tried to compete for a while with its line of electrostatic electron microscopes. Martin at Imperial University as early as Philips, Siemens, and Carl Zeiss each tried to grab a share of the European market. Starting in , scientists in Japan gathered to decide on the best way to build an electron microscope.
This group evolved into the Japan Electron Optics Laboratory JEOL that would eventually produce more models and varieties of electron microscopes than any other company.
0コメント