TY - JOUR
AU - Rose, Harald, H.
AB - Abstract A brief history of the development of direct aberration correction in electron microscopy is outlined starting from the famous Scherzer theorem established in 1936. Aberration correction is the long story of many seemingly fruitless efforts to improve the resolution of electron microscopes by compensating for the unavoidable resolution-limiting aberrations of round electron lenses over a period of 50 years. The successful breakthrough, in 1997, can be considered as a quantum step in electron microscopy because it provides genuine atomic resolution approaching the size of the radius of the hydrogen atom. The additional realization of monochromators, aberration-free imaging energy filters and spectrometers has been leading to a new generation of analytical electron microscopes providing elemental and electronic information about the object on an atomic scale. Scherzer theorem, aberration correctors, atomic resolution Introduction Aberration correction in electron microscopy dates back to Otto Scherzer who proved in 1936 that chromatic and spherical aberrations of rotationally symmetric electron lenses are unavoidable [1]. His finding was so important that it was named ‘Scherzer theorem’. The validity of this theorem requires that several conditions be satisfied. Object and image must both be real, the electromagnetic field must be static and rotationally symmetric, the electric potential and its derivatives must be continuous (no space charges within the region of the electron beam) and the mirror mode has to be excluded. Contrary to the aberration-free compound lenses of high-performance light microscopes, the quality of electron lenses is very poor. As a result, the resolution limit of standard electron microscopes equals about one hundred times the wavelength, whereas modern light microscopes have reached a resolution limit somewhat smaller than the wavelength. In 1947, Scherzer found an ingenious way for enabling aberration correction. He demonstrated in a famous article that it is in theory possible to eliminate chromatic and spherical aberrations by lifting any one of the constraints of his theorem, either by abandoning rotational symmetry or by introducing time-varying fields, or space charges [2]. Moreover, he proposed a multipole corrector compensating for the spherical aberration of the objective lens. Seeing the arrangements of atoms has been an old dream in science. In 1959, Richard Feynman articulated this wish by stating: ‘It would be very easy to make an analysis of any complicated chemical substance; all one would have to do would be to look at it and see where the atoms are. The trouble is that the electron microscope is one hundred times to poor. Is there no way to make the electron microscope more powerful?’ At that time the resolution of the standard electron microscopes had not yet reached the so-called Scherzer limit ⁠, where and are the electron wavelength and the coefficient of the spherical aberration, respectively. However, even after this limit had been achieved ∼20 years later, it did not suffice to image and study the atomic structure of matter because atom displacement by knock-on collisions prevents one from reducing sufficiently the wavelength. As a result, modern standard electron microscopes operating at 200 or 300 kV do not yield a resolution limit below 2.5 or 2.0 pm, respectively. The resolution is limited by the axial geometrical aberrations resulting from static lens defects and chromatic aberration and instabilities. The latter incoherent defects determine the information limit because they suppress the transfer of the high spatial frequencies. Fortunately, chromatic aberration does not prevent atomic resolution for voltages above ∼200 kV. However, the suppression of the instabilities is a very demanding ongoing task. Early correction attempts Scherzer always believed that departure from rotational symmetry offered the most promising approach for correcting the axial aberrations of electron lenses. Therefore, he allocated to his student R. Seeliger the task to built and test the electrostatic corrector shown in Fig. 1 within the frame of his doctoral thesis. Fig. 1 Open in new tabDownload slide Scheme of the Scherzer corrector consisting of a stigmator St, two electrostatic cylinder lenses and ⁠, a round lens R and three octopoles and compensating for the spherical aberration of the objective lens O. Fig. 1 Open in new tabDownload slide Scheme of the Scherzer corrector consisting of a stigmator St, two electrostatic cylinder lenses and ⁠, a round lens R and three octopoles and compensating for the spherical aberration of the objective lens O. Starting in 1948, Seeliger built and tested the Scherzer corrector for a period of ∼5 years. He aligned the constituent elements mechanically by means of adjustment screws. However, the experiments showed that this approach was a major obstacle due to insufficient stability of the mechanical alignment. As a result, the correction did not improve the resolution of the microscope because it was limited by mechanical and electromagnetic instabilities rather than by the static defects of the objective lens. Although Seeliger could not improve the resolution limit of the basic electron microscope, he could demonstrate that the corrector provided a negative spherical aberration, which could be adjusted to compensate for the spherical aberration of the objective lens [3]. G. Moellenstedt first demonstrated experimentally the effective correction of spherical aberration by means of this corrector [4]. By employing critical illumination with a large cone angle of rad, he enlarged the spherical aberration to such an extent that it became by far the dominant aberration, which strongly blurred the image. After compensating for the spherical aberration by means of the corrector octopoles, the resolution improved by a factor of about 7, accompanied by a striking increase in the contrast. Although Seeliger failed to improve the experimental resolution by eliminating the spherical aberration, Scherzer was convinced that aberration correction was the most appropriate procedure for improving the performance of electron microscopes. In order to find the reasons for the failure of the experiment, he confronted W. E. Meyer in 1957 with the task to investigate for his thesis the obstacles limiting the actual resolution of the spherically corrected microscope and to find means for eliminating these perturbations. The results of Meyer's intensive and painstaking investigations revealed that one has to consider two groups of parasitic perturbations [5]. He defined the first group as alignment aberrations. They are produced by static imperfections and misalignment of the objective lens and the constituent elements of the corrector. Nowadays, we characterize the axial misalignment aberrations of second order as axial coma and 3-fold astigmatism, and those of third order as star aberration and 4-fold astigmatism. It follows from Meyer's representation of the corresponding aberration coefficients that one can eliminate all axial astigmatisms by means of multipole stigmators. These results remained largely unnoticed because only the first-order astigmatism affected the resolution at that time. The second group of aberrations originates from time-dependent aberrations caused by charging, alternating external electromagnetic fields and mechanical instabilities. Meyer defined the deleterious effect of these incoherent parasitic aberrations as image blur. At that time, contrast transfer theory was not yet established in electron microscopy. Within the frame of notation of this theory, the resolution limit resulting from the incoherent aberrations is the so-called information limit. In 1964, Deltrap built a telescopic quadrupole–octopole corrector to eliminate the spherical aberration of a probe-forming lens. Although he nullified this aberration, he failed, like his predecessors, to improve the actual resolution of the uncorrected system because at that time the resolution was not limited by the spherical aberration of the objective lens. About 7 years later, H. Rose proved that all known correctors introduce large off-axial coma and are not suitable for transmission electron microscopes (TEMs) [6]. In order to compensate for spherical aberration, chromatic aberration and off-axial coma, he proposed a novel aplanatic corrector utilizing symmetry properties. This corrector was built and tested successfully in a test microscope within the frame of the so-called Darmstadt project and it demonstrated for the first time the simultaneous correction of chromatic and spherical aberration [7]. Unfortunately, the project was abandoned after the death of O. Scherzer in 1982, although it was successful as far as it went. Scherzer's ingenious idea to compensate for the unavoidable spherical aberration of round lenses by means of a corrector consisting of non-rotationally symmetric elements and the experiments of Seeliger and Moellenstedt initiated correction efforts at other places, predominantly in England. In order to simplify the arrangement of the Scherzer corrector, Archard replaced the two cylinder lenses and the round lens of the Scherzer corrector shown in Fig. 1 by four quadrupoles forming an anti-symmetric quadrupole quadruplet [8]. The quadrupole fields were excited within octopole elements, which also produced octopole fields compensating for the combined spherical aberration of the round lens and the corrector together with the 2- and 4-fold aperture aberrations introduced by the quadrupole fields of the corrector. In 1963, Dymnikov and Yavor proposed a symmetric quadrupole quadruplet with anti-symmetric excitation of the quadrupole fields [9]. Figure 2 illustrates the arrangements of the quadrupoles and the course of the fundamental rays in the x–z and y–z sections of the telescopic version. Fig. 2 Open in new tabDownload slide Course of the fundamental paraxial rays in the x–z and the y–z sections of the telescopic quadrupole quadruplet; astigmatic line images of the infinitely distant plane are formed at the centers of the inner quadrupoles. Fig. 2 Open in new tabDownload slide Course of the fundamental paraxial rays in the x–z and the y–z sections of the telescopic quadrupole quadruplet; astigmatic line images of the infinitely distant plane are formed at the centers of the inner quadrupoles. Kelman and Yavor demonstrated theoretically in 1963 that the chromatic aberration coefficient of a combined electrostatic–magnetic quadrupole can have negative sign [10]. By choosing the strengths of the electric and magnetic components such that their combined action nullifies for electrons with nominal energy, the element acts as a first-order Wien filter, which can be utilized to correct chromatic aberration in one section. The first experimental study with electric–magnetic quadrupoles was performed by Hardy in 1967 [11]. He showed in a proof-of-principle experiment that these elements enable chromatic correction. About 10 years later, Koops et al. achieved chromatic correction in the image of a real object within the frame of the Darmstadt project [12]. In order to increase the chromatic aberration, they added an alternating voltage of 130 V to the cathode potential, thereby strongly blurring the image. The image became sharp again after he corrected the chromatic aberration by properly adjusting the electric and magnetic quadrupoles, as shown in Fig. 3. Fig. 3 Open in new tabDownload slide Image of a holey carbonaceous foil taken by an electric–magnetic quadrupole triplet with unit magnification and two projector lenses. The images on the left are obtained without chromatic correction and those on the right-hand side with chromatic correction. The images on the bottom are formed by adding an alternating voltage of 130 V to the 40 kV of the cathode potential. Fig. 3 Open in new tabDownload slide Image of a holey carbonaceous foil taken by an electric–magnetic quadrupole triplet with unit magnification and two projector lenses. The images on the left are obtained without chromatic correction and those on the right-hand side with chromatic correction. The images on the bottom are formed by adding an alternating voltage of 130 V to the 40 kV of the cathode potential. In 1972, A. Crewe and V. Beck started at the University of Chicago project another attempt to correct the spherical aberration of a scanning transmission electron microscope (STEM). They built and tested over a period of 6 years a magnetic quadrupole–octopole corrector consisting of a symmetric quadruplet with combined quadrupoles and octopoles. Because the corrector is aimed for the STEM, the difficulties encountered by the Darmstadt project of correcting a sufficiently large field of view do not arise. Although Beck and Crewe did introduce many stigmator coils for producing weak dipole and hexapole fields, they were unable to find a suitable setting [13]. Although all fruitless experimental attempts demonstrated that correction works in principle, none of them achieved an improvement in resolution. The main reasons for theses failures were as follows: (i) the basic microscope was not stable enough, (ii) the deleterious interference with the environment had been underestimated, (iii) it was not possible to determine the state of alignment with the required precision and (iv) the resolution-limiting residual aberrations could not be measured with the required accuracy and eliminated within a period of time which must be shorter than the duration of the overall stability of the entire system. These obstacles were so severe that at the end of the 1980s, a panel of experts advised the American National Science Foundation to no longer fund the fruitless correction projects. As a result, experimental work on aberration correction was abandoned worldwide. At about the same time during the 1989 Electron Microscopy Meeting at Salzburg, Austria, Haider, Rose and Urban had intensive discussions on the prospects of aberration correction with the help of a new semi-aplanatic corrector involving two sextupoles and two round-lens transfer doublets [14]. This system is corrected for spherical aberration and isotropic coma. Since the corrected system is rather simple and because hexapoles do not affect the paraxial path of rays, they were convinced that the successful correction of spherical aberration and radial off-axial coma in a modern 200 kV TEM equipped with a field emission gun should be feasible contrary to the belief of all experts. Evolution of the hexapole corrector Sextupoles were for a long time not considered to be suitable candidates for correcting the aberrations of round lenses because they introduce in first approximation a 3-fold second-order path deviation. A hexapole field produces a force whose strength is proportional to the square of the lateral distance of the electron from the optic axis. The direction of the force depends on the azimuthal angle and changes its sign every 60 angular degrees, as illustrated schematically in Fig. 4. Therefore, an assemble of electrons propagating on the mantle of a cylinder parallel to the optic axis experiences a primary 3-fold deformation of second order after passing the hexapole field. However, due to the non-linear force, an additional rotationally symmetric divergent third-order deformation arises resulting in a negative spherical aberration at the image plane. Accordingly, hexapoles can compensate for the spherical aberration of round lenses provided that one can eliminate the large primary 3-fold astigmatism. Unfortunately, it is not possible to eliminate this aberration in a system consisting solely of hexapoles. Fig. 4 Open in new tabDownload slide Action of an electrostatic sextupole on electrons propagating parallel to the optic axis; the arrows indicate the direction of the force. Fig. 4 Open in new tabDownload slide Action of an electrostatic sextupole on electrons propagating parallel to the optic axis; the arrows indicate the direction of the force. It was first pointed out by P. W. Hawkes in 1965 that in second approximation, the sextupoles introduce additional third-order aberrations equivalent to those of round lenses [15]. In the beginning of the 1970s, Rose and Plies investigated in detail the aberrations of electron optical systems with a curved axis consisting of magnetic round lenses, dipoles quadrupoles, sextupoles and octopoles [16]. In particular, they derived formulae for the coefficients of the third-order aberrations and showed that the spherical aberration can have negative sign even if the index of refraction is rotationally symmetric in the paraxial domain [17]. However, they did not explore further the properties of sextupoles in combination with round lenses because the dominant second-order aberrations of the sextupoles seemed to rule out their usefulness for correcting the much smaller third-order spherical aberration of a good objective lens. In 1979, V. Beck showed that a combination of a round lens and two sextupoles has a negative spherical aberration and can be made free of the second-order aperture aberration [18]. Unfortunately, his system and that suggested later by Crewe introduce a large 3-fold fourth-order aperture aberration [19]. This aberration results from the combination of the large non-corrected second-order field aberrations with the third-order axial aberrations and prevents an appreciable improvement in the resolution. The problem was solved by Rose in 1981 [20], who proposed a sextupole or hexapole corrector consisting of a telescopic round-lens doublet and two identical sextupoles, one of them centered at the front focal point of the first round lens and the other at the back focal point of the second lens, as shown in Fig. 5. The transfer doublet images the first hexapole inversely onto the second hexapole, whereupon all second-order path deviations cancel out. However, the rotationally symmetric third-order path deviations add up. The resulting negative spherical aberration is proportional to the square of the hexapole strength, which can be adjusted to eliminate the combined spherical aberration of the corrector and the objective lens. Since the corrector introduces primarily a negative spherical aberration, we can conceive it as ‘spectacles’ for the imperfect objective lens of the electron microscope. Fig. 5 Open in new tabDownload slide Arrangement of the elements and path of the paraxial fundamental rays within the telescopic sextupole corrector; f is the focal length of the round lenses. Fig. 5 Open in new tabDownload slide Arrangement of the elements and path of the paraxial fundamental rays within the telescopic sextupole corrector; f is the focal length of the round lenses. The front focal point of the first lens and the back focal plane of the second lens of the transfer doublet represent the nodal planes of the corrector, which coincide with its coma-free planes. In order to form an aplanat, we must match the front coma-free point with the corresponding point of the objective lens. Unfortunately, this is not directly possible because the coma-free point of the round lens lies within the lens field and is located at the center of the first sextupole. However, we can match these points optically by means of another telescopic transfer doublet if we place its front nodal point in the coma-free point of the objective lens and its back nodal point at the point of the corrector, as illustrated in Fig. 6. This procedure eliminates the radial or isotropic component of the off-axial coma yet not the azimuthal component, which results from the Larmor rotation of the objective lens. Fortunately, this component is small and does not appreciably limit the number of image points as long as the resolution is not significantly higher than ∼1 Å. Fig. 6 Open in new tabDownload slide Coma-free arrangement of the sextupole corrector and the objective lens by means of a telescopic transfer doublet resulting in an electron optical semi-aplanat. Fig. 6 Open in new tabDownload slide Coma-free arrangement of the sextupole corrector and the objective lens by means of a telescopic transfer doublet resulting in an electron optical semi-aplanat. In order to get funding, Haider, Rose and Urban had to convince the electron microscopists that the new correction approach was realistic and promising because the required technology was available in 1990. At the end of this year, they submitted a joint grant application to the Volkswagen Foundation to obtain the necessary funds. All other granting agencies refused to fund a ‘non-feasible’ project. Fortunately, the Volkswagen foundation was willing to take the risk and approved funding in 1991. In January 1992, Haider started the experimental work at the European Molecular Biology Laboratory (EMBL) at Heidelberg. Already in 1994, he demonstrated the correction of spherical aberration on an electron-optical test bench. Based on this proof of principle, the Volkswagen Foundation approved the final funding for realizing the aberration-corrected TEM to be employed in Juelich for atomic characterization of materials. In July 1997, B. Kabius, a member of Urban's group, obtained the first atomic-resolution images of gallium arsenide after the resolution limit of the microscope was reduced by the corrector from 0.24 to ∼0.12 nm [21]. After this success, the aberration-corrected electron microscope was transferred to Juelich and installed there. During the following 2 years the electrical stability of the instrument was further improved and some modifications were implemented. In 2000, the microscope was in perfect shape, and it operates on a routine basis with extraordinary results up to this date. The impressive results obtained by K. Urban and co-workers with the aberration-corrected microscope have received great attention and recognition, especially among the members of the materials science community. Employing the negative phase contrast mode by over-compensating the spherical aberration of the objective lens, they imaged for the first time oxygen and other low-Z elements [22]. Revival of the quadrupole–octopole corrector Due to the continuous failure to achieve any real gain in resolution, the experimental work on aberration correction stopped until J. Zach tackled the problem experimentally anew in the early 1990s [23]. Because the chromatic aberration determines primarily the resolution of the LVSEM rather than the mechanical and electromagnetic instabilities, the chances to improve the instrumental resolution by correcting the aberrations seemed high enough to justify another correction attempt. Therefore, a new Cc/Cs quadrupole–octopole corrector was designed and built by Lanio and Haider at the EMBL. Unfortunately, the first tests of the corrector revealed that the hysteresis causes a cross-talk between magnetic multipole fields that are combined in a single element. The unavoidable cross-talk prevented a suitable adjustment of the corrector, as had been the case for the Chicago corrector. To eliminate this difficulty, J. Zach substituted electric multipoles for the magnetic multipoles apart from the two inner magnetic quadrupoles, which are mandatory for eliminating the chromatic aberration. With the rearranged corrector, Zach and Haider (1995) achieved for the first time a real improvement in the resolution of an actual electron microscope [24]. The chromatic correction is schematically illustrated in Fig. 7. Fig. 7 Open in new tabDownload slide Schematic arrangement of the elements of the corrected LVSEM and course of the axial trajectories for different energy deviations illustrating chromatic correction. Fig. 7 Open in new tabDownload slide Schematic arrangement of the elements of the corrected LVSEM and course of the axial trajectories for different energy deviations illustrating chromatic correction. Owing to the advancement of technology, also O. Krivanek started working on aberration correction around 1995 at the Cavendish Laboratory in Cambridge. He developed an improved version of the Deltrap quadrupole–octopole corrector aimed to compensate for the spherical aberration of a VG STEM operating at an accelerating voltage of 100 kV. Krivanek and co-workers constructed the corrector in such a way that apart from the basic magnetic quadrupole and octopole fields for the correction of the spherical aberration, additional multipole fields could be excited under computer control to compensate for deleterious parasitic geometrical aberrations caused by misalignment, magnetic inhomogeneities and mechanical inaccuracies. The experiments showed that the cross-talk between the quadrupole and octopole fields posed a major difficulty for the alignment, as had been the case with the Chicago corrector. To avoid this obstacle, Krivanek and his co-workers designed a new corrector such that the quadrupole and octopole fields could be excited in separate elements [25]. The corrector was incorporated in a 100 kV VG STEM. For adjusting the system and for determining the resolution-limiting aberrations, they developed a procedure based on the evaluation of Ronchigrams. By means of these improvements, they succeeded in performing rapidly the many adjustment steps in a systematic way. The results demonstrated for the first time a genuine improvement in the resolution of a STEM by correction of the spherical aberration [26]. Murray Gibson at the Argonne National Laboratory had in 1999 the visionary idea to realize a sub-Å and sub-eV in situ electron microscope. He intended to convert the standard electron microscope into a materials science laboratory with adequate space to carry out experiments under various environmental conditions. The dynamic micro-laboratory required an increased space within the pole piece of the objective lens. However, increasing the gap between the pole pieces enlarges the focal length and the chromatic aberration preventing sub-Å resolution. Hence, atomic resolution can only be achieved for the proposed microscope by correcting spherical and chromatic aberration. In order to establish a national project for the development of a Transmission Electron Achromatic Microscope (TEAM), Murray Gibson organized a workshop on this topic in July 2000. After extensive discussions, the numerous experts came to the conclusion that the project was feasible. Four national laboratories decided to develop a mutual proposal for the construction of the microscope with Murray Gibson as coordinator. A crucial part of the project was the design of a novel corrector compensating for spherical and chromatic aberrations and for off-axial coma to obtain a sufficiently large field of view of at least 2000 equally resolved object points per image diameter. In order to find an appropriate corrector design, Murray Gibson invited Harald Rose by the end of 2000 to tackle this challenging task during a stay at the Argonne National Laboratory. The correction of chromatic aberration necessitates the incorporation of electric and magnetic quadrupoles, which affect the paraxial path of the electrons. By imposing symmetry conditions on the multipole fields and the course of the fundamental paraxial rays, a large number of aberrations cancel out. The largest reduction is achieved by imposing symmetry conditions on the system as a whole and on each half of it. The optimum system is obtained by replacing each sextupole element of the hexapole corrector by a telescopic quadrupole–octopole quintuplet, shown in Fig. 8. The resulting TEAM corrector is depicted schematically in Fig. 9. Detailed calculations proved the feasibility of the corrector concept. By optimizing the design, Mueller and Uhlemann (CEOS) obtained an achromatic aplanatic system, which satisfies theoretically all requirements. The most stringent constraints are the extremely high stabilities of ∼ for the electric and magnetic quadrupole fields correcting for the chromatic aberration. These fields are excited within the central element of each multipole quintuplet. Fortunately, CEOS has reached in 2006 the unprecedented relative stability of for the necessary currents and voltages. Fig. 8 Open in new tabDownload slide Schematic course of the x- and y-components of the principal ray and the nodal ray, respectively, within the telescopic quadrupole–octopole quintuplet [27]. Fig. 8 Open in new tabDownload slide Schematic course of the x- and y-components of the principal ray and the nodal ray, respectively, within the telescopic quadrupole–octopole quintuplet [27]. Fig. 9 Open in new tabDownload slide Schematic arrangement of the TEAM corrector and course of the fundamental rays; strongly anamorphotic images of the diffraction plane are located at the center planes and of the multipole quintuplets. Fig. 9 Open in new tabDownload slide Schematic arrangement of the TEAM corrector and course of the fundamental rays; strongly anamorphotic images of the diffraction plane are located at the center planes and of the multipole quintuplets. In order to place an anamorphoptic image of the diffraction plane at the symmetry plane of the quintuplet, we must match the incident principal ray with the field ray and the nodal ray with the axial ray ⁠, respectively. The axial ray starts from the center of the object plane and the field ray intersects the center of the diffraction plane. The central element of each multipole quintuplet is an electric magnetic dodecapole enabling independent excitation of electric and magnetic quadrupole fields and an octopole field. The mixed electric and magnetic quadrupole fields act as a quadrupole and a first-order Wien filter, which compensates for the axial chromatic aberration. The central octopole fields eliminate the spherical aberration of the round lenses without introducing any appreciable field aberrations. We nullify the remaining 4-fold axial astigmatism by an additional octopole element placed at a distortion-free image of the diffraction plane, either at the plane midway between the round lenses of the transfer doublet or at the conjugate plane located behind the corrector. The latter location is more favorable because it induces smaller fifth-order combination aberrations than the other location. The comparison of Figs. 6 with 9 demonstrates that the double symmetry of the fundamental rays with respect to the multipole fields of the hexapole corrector is also preserved in the TEAM corrector, although the x- and y-components of the fundamental rays differ from each other within the multipole quintuplets. The ray components exhibit exchange symmetry with respect to the midplane because the quadrupole fields are excited anti-symmetrically with respect to this plane. To obtain an achromatic aplanat, (i) the field ray must intersect the coma-free plane of the objective lens, (ii) an octopole must be centered at each of the two anamorphotic images of this plane and (iii) a third octopole must be centered at a distortion-free image of the coma-free plane, as depicted in Fig. 10. Fig. 10 Open in new tabDownload slide Coma-free arrangement of the TEAM corrector and location of the octopole O2 compensating for the 4-fold axial astigmatism; the front nodal point of the corrector is located at the first image of the object plane O. Fig. 10 Open in new tabDownload slide Coma-free arrangement of the TEAM corrector and location of the octopole O2 compensating for the 4-fold axial astigmatism; the front nodal point of the corrector is located at the first image of the object plane O. In 2002, Uli Dahmen from the Lawrence Berkeley National Laboratory became the coordinator of the TEAM project because M. Gibson became director of the Advanced Photon Source at Argonne. The proposal for funding of the project was submitted to the Department of Energy (DOE) and was approved in 2003. However, the original aim and its name were changed from the achromatic in situ electron microscope to the ‘Transmission Electron Aberration-Corrected Microscope’, although the acronym TEAM remained unchanged. In order to realize the required resolution limit of 0.5 Å, the information limit of the basic instrument must stay below 0.4 Å because the corrector also introduces incoherent aberrations, which enlarge this limit. Recently, an information limit of 0.048 nm has been reached at Berkeley with an improved version of the hexapole corrector. Moreover, a resolution of ∼1 Å has been achieved with energy-loss electrons by Kabius et al. with the first prototype of the TEAM corrector compensating for spherical and chromatic aberration [28]. Conclusion and future developments In the last decade, considerable theoretical and experimental progress has been made in the field of electron optics. Thanks to the development of efficient procedures for calculating and designing complex electron-optical systems, it has become possible to precisely design multi-element aberration correctors, imaging energy filters, monochromators and quadrupole systems operating as projector lenses with a very small focal length. The incorporation of these elements in the TEM will enable quantitative analytical electron microscopy with sub-Å resolution and energy resolution below 0.1 eV. Moreover, fast computer-aided alignment procedures have been established which show the state of alignment and determine the voltages and currents which are subsequently applied by microprocessors to the stigmators compensating for the residual aberrations. Within the frame of the TEAM Project, the electrical and mechanical stabilities of the electron microscopes have been improved to such an extent that the information limit has now reached 0.48 Å at a voltage of 300 kV. The progress in the resolution of the microscope is convincingly illustrated in Fig. 11. Fig. 11 Open in new tabDownload slide Increase in the resolution of the microscope as a function of time. Fig. 11 Open in new tabDownload slide Increase in the resolution of the microscope as a function of time. We cannot image organic materials, ceramics and nano-objects with atomic resolution at medium voltages because these objects will be destroyed by knock-on collisions and ionization. In order to obtain sub-Å resolution, we must operate at voltages below the threshold for atom displacement. The knock-on of surface atoms occurs at voltages, which are appreciably lower than those for atom displacement within the bulk. Thus, surface etching may occur for low-Z materials at electron energies well below 60 keV. For radiation-sensitive objects, the attainable specimen resolution depends primarily on the tolerable dose D of the incident electrons, the image contrast C and the signal-to-noise ratio required to discriminate an object detail from the noise. In the case ⁠, the specimen resolution limit equals the instrumental resolution limit ⁠. However, for organic materials, the tolerable dose is so small that we can neglect the influence of the instrumental resolution To avoid atom displacement by knock-on collisions, we must operate at voltages lower than 60 kV for most organic objects. We achieve highest specimen resolution by (i) correcting spherical aberration, chromatic aberration and off-axial coma, (ii) introducing an anamorphotic obstruction-free micro-phase plate and (iii) reducing the information limit as much as possible. Moreover, the higher order aberrations and the third-order chromatic aberration must be kept small in order that the high-angle scattered electrons do not miss the image spot. Otherwise, they increase the fluctuation of the background intensity, thus decreasing the SNR. We obtain highest contrast in the negative contrast mode because in this case the scattering contrast and the phase contrast add up with the same sign. In the ideal case, atoms, atom clusters and atom columns appear in the image as bright spots superposed on the almost uniform background intensity [29]. The additional use of the inelastic scattered electrons may greatly enhance the contrast of low-Z objects, if they are undergoing an additional elastic scattering process, as it happens for biological objects embedded in ice. The SALVE (Sub-Ångstrom Low-Voltage Electron microscope) Project initiated by Ute Kaiser at the University of Ulm aims for the realization of an aberration-corrected TEM operating in the range between 20 and 100 kV. In order to utilize as many scattered electrons as possible, usable aperture angles up to 50 mrad are necessary. To satisfy this condition, the state of correction must be very high and mechanical vibrations and parasitic electromagnetic stray fields must be sufficiently suppressed. Simultaneously, the highest available technical means must be employed to stabilize the lens currents and voltages with extreme accuracy. With further advancement of technology we can expect that future electron-optical instruments will reach such degree of perfection that the resolution is no longer limited by static aberrations. Then, the main task is no longer the design and construction of aberration-free electron optical imaging systems but the reduction of the information limit by improving the mechanical and electrical stabilities. References 1 Scherzer O . Ueber einige Fehler von Elektronenlinsen , Z. Physik , 1936 , vol. 101 (pg. 593 - 693 ) Google Scholar Crossref Search ADS WorldCat 2 Scherzer O . Sphaerische und chromatische Korrektur von Elektronenlinsen , Optik , 1947 , vol. 2 (pg. 114 - 132 ) OpenURL Placeholder Text WorldCat 3 Seeliger R . Die sphaerische Korrektur von Elektronenlinsen mittels nicht rotationssymmetrischer Abbildungselemente , Optik , 1951 , vol. 8 (pg. 311 - 317 ) OpenURL Placeholder Text WorldCat 4 Moellenstedt G . Elektronenmikroskopische Bilder mit einem nach O. Scherzer sphaerisch korrigierten Objektiv , Optik , 1956 , vol. 13 (pg. 209 - 215 ) OpenURL Placeholder Text WorldCat 5 Meyer W E . Das praktische Aufloesungsvermoegen von Elektronenmikroskopen , Optik , 1961 , vol. 18 (pg. 101 - 114 ) OpenURL Placeholder Text WorldCat 6 Rose H . Elektronenoptische Aplanate , Optik , 1971 , vol. 34 (pg. 285 - 311 ) OpenURL Placeholder Text WorldCat 7 Bernhard W . Erprobung eines sphaerisch korrigierten Elektronenmikroskops , Optik , 1980 , vol. 57 (pg. 73 - 94 ) OpenURL Placeholder Text WorldCat 8 Archard G D . Requirements contributing to the design of devices used in correcting electron lenses , Br. J. Phys. , 1954 , vol. 5 (pg. 294 - 299 ) Google Scholar Crossref Search ADS WorldCat 9 Dymnikov A D , Yavor S Y . Four quadrupole lenses as an analogue of an axially symmetric system , Sov. Phys. Tech. Phys. , 1963 , vol. 8 (pg. 639 - 643 ) OpenURL Placeholder Text WorldCat 10 Kelman V M , Yavor S Y . Achromatic quadrupole electron lenses , Zh. Tekh. Fiz. , 1961 , vol. 31 (pg. 1439 - 1442 ) OpenURL Placeholder Text WorldCat 11 Hardy D F . Combined magnetic and electrostatic quadrupole electron lenses , 1967 Dissertation, University of Cambridge OpenURL Placeholder Text WorldCat 12 Koops H , Kuck G , Scherzer O . Erprobung eines elektronenoptischen Achromators , Optik , 1977 , vol. 48 (pg. 225 - 236 ) OpenURL Placeholder Text WorldCat 13 Crewe A V . Some Chicago aberrations , Microsc. Microanal. , 2002 , vol. 8 Supplement 2 (pg. 4 - 5 ) OpenURL Placeholder Text WorldCat 14 Rose H . Outline of a spherically corrected semi-aplanatic medium-voltage TEM , Optik , 1990 , vol. 85 (pg. 19 - 24 ) OpenURL Placeholder Text WorldCat 15 Hawkes P W . The geometrical aberrations of general electron optical systems, I and II , Philos. Trans. R. Soc. A , 1965 , vol. 257 (pg. 479 - 552 ) Google Scholar Crossref Search ADS WorldCat 16 Rose H , Plies E . Hawkes P W . Correction of aberrations in electron optical systems with curved axis , Image Processing and Computer-Aided Design in Electron Optics , 1973 London Academic Press (pg. 344 - 370 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 17 Plies E . Korrektur der Oeffnungsfehler elektronenoptischer Systeme mit krummer Achse und durchgehend astigmatismusfreien Gausschen Bahnen , Optik , 1973 , vol. 38 (pg. 502 - 518 ) OpenURL Placeholder Text WorldCat 18 Beck V D . A hexapole spherical aberration corrector , Optik , 1979 , vol. 53 (pg. 241 - 255 ) OpenURL Placeholder Text WorldCat 19 Crewe A V . Studies on sextupole correctors , Optik , 1980 , vol. 57 (pg. 313 - 327 ) OpenURL Placeholder Text WorldCat 20 Rose H . Correction of aperture aberrations in magnetic systems with threefold symmetry , Nucl. Instrum. Methods , 1981 , vol. 187 (pg. 187 - 199 ) Google Scholar Crossref Search ADS WorldCat 21 Haider M , Rose H , Uhlemann S , Schwan E , Kabius B , Urban K . Towards 0.1 nm resolution with the first spherically corrected transmission electron microscope , J. Electron Microsc. , 1998 , vol. 47 (pg. 395 - 405 ) Google Scholar Crossref Search ADS WorldCat 22 Urban K . Studying atomic structures by aberration-corrected transmission electron microscopy , Science , 2008 , vol. 321 (pg. 506 - 510 ) Google Scholar Crossref Search ADS PubMed WorldCat 23 Zach J . Design of a high-resolution low-voltage scanning electron microscope , Optik , 1989 , vol. 83 (pg. 30 - 40 ) OpenURL Placeholder Text WorldCat 24 Zach J , Haider M . Aberration correction in a low-voltage SEM by a multipole corrector , Nucl. Instrum. Methods Phys. Res. A , 1995 , vol. 365 (pg. 316 - 325 ) Google Scholar Crossref Search ADS WorldCat 25 Krivanek O L , Dellby N , Lupini A R . Towards sub-Å electron beams , Ultramicroscopy , 1999 , vol. 78 (pg. 1 - 11 ) Google Scholar Crossref Search ADS WorldCat 26 Krivanek O L , Nellist P D , Dellby N , Murfitt N S , Silagyi Z . Towards 0.5 Å electron beams , Ultramicrocopy , 2003 , vol. 96 (pg. 229 - 237 ) Google Scholar Crossref Search ADS WorldCat 27 Rose H . , Geometrical Charged-Particle Optics , 2008 Heidelberg Springer Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 28 Kabius B , Hartl P , Haider M , Mueller H , Uhlemann S , Rose H . Aberration correction within the TEAM project , J. Electron Microsc. , 2009 forthcoming OpenURL Placeholder Text WorldCat 29 Jia C L , Lentzen M , Urban K . High-resolution transmission electron microscopy using negative spherical aberration , Microsc. Microan. , 2004 , vol. 10 (pg. 174 - 184 ) Google Scholar Crossref Search ADS WorldCat © The Author 2009. Published by Oxford University Press on behalf of Japanese Society of Microscopy. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org Oxford University Press
TI - Historical aspects of aberration correction
JF - Journal of Electron Microscopy
DO - 10.1093/jmicro/dfp012
DA - 2009-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/historical-aspects-of-aberration-correction-ZO0mF6JBH0
SP - 77
EP - 85
VL - 58
IS - 3
DP - DeepDyve
ER -