Picosecond Optoelectronic Chronography Discoverer
After his wonderful constructive period of life at the Kazan State University (1931-1947) Yevgeny Konstantinovich worked as the trainee of the Academician P.L. Kapitsa (in 1945) at the Institute for Physical Problems in Moscow for a short period of time. There he successfully confirmed the results of his experiments on EPR and defended doctoral dissertation at the Physical Institute of the Academy of Sciences on the 30th of January, 1945. Soon after these events, I.V. Kurchatov invited him to the Design Engineering Bureau-11 (Arzamas-16) (1947-1951). Yevgeny Knostantinovich was awarded the Stalin Prize and the Order of Lenin “For the Development of Electromagnetic Methods of Registration of Fast Processes on the Study of Core Part of Nuclear Bomb Charge” (1949) for the works accomplished during these years. In 1951 Y.K. Zavoisky was appointed to the position of Head of the Sector No. 74 in the Laboratory of Measuring Instruments of the Academy of Sciences of the USSR (LIPAN) at the Kurchatov Institute. There, in 1950s together with his colleagues he constructed the first luminescence camera in the country for the registration of ionizing particle tracks and created the spark counters of ionizing particles using the multi-stage IITs which were designed by M.M. Butslov and his colleagues. IITs which could register individual light quanta and ensure picosecond time resolution during the study of fast processes (FPs) were designed for the first time in the world.
Birth of Picosecond Electron-Optical Chronography
When the first image intensifier tubes were invented in the early 1930s  nobody thought that they could be used for the high-speed photography in order to register FPs. In order to improve the object observation during the night time the author suggested the prior illumination of that invention with invisible IR radiation. First IITs contained photocathode and luminescent screen located parallel to each other and accelerating potential was applied between them. In homogeneous electric field photoelectrons spread along the parabolic trajectories giving the blurred object image on the screen. In the early 1940s in the papers of L.A. Artsimovich  in our country and studies of German and US scientists [6-8] the results connected with the obtainment of spatial resolution in IITs with the value of tens of line pairs per millimeter were published. It was achieved at the expense of the image focusing in IIT by electrostatic and magnetic lens. The first efforts for the application of pulse voltage on accelerating electrodes of night vision IIT were made in the late 1940s – early 1950s with the further use of such IITs for the study of time characteristics of non-stationary light sources. At the State Optical Institute (SOI) M.P. Vaniukov with his colleagues implemented the exposures with the duration of fractions of microsecond (4·10-7s)  using the night vision IIT of AEG type (with electrostatic focusing of photoelectron images) by their activation with short voltage pulses. J. Courtney-Pratt (England) applying alternating magnetic field to the similar IITs achieved the time resolution up to the frictions of nanosecond (2.8·10-10 s) . However, the real revolution in the development of methods and devices for electron-optical photography with picosecond time resolution was accomplished by Yevgeny Konstantinovich Zavoisky with his colleagues in 1950s . Works were performed at the Institute of Nuclear Power Engineering and great scientist Sergei Dmitrievich Fanchenko left bright memories about these works .
The idea of image deflection in IIT by the system which is similar to oscillographic system completely belongs to the Academician Y.K. Zavoisky too. This decisive suggestion consisted in the radical modernization of IIT which had just been developed and designed by M.M. Butslov at that time in 1949. This device known as PIO-1 (single-frame pulse converter) was intended for the frame-by-frame photography of FPs with electrostatic focusing. The main point of the suggestion consisted in the introduction of two pairs of oscillographic-type deflection plates oriented towards each other at 90° to the area of crossover which dimensions were: diameter – fractions of millimeter, length – several centimeters. Upgraded device (PIM-3 – multiframe pulse converter) represented the first time-analyzing IIT in the world . Mikhail Mikhailovich demonstrated the high-speed frame-by-frame photography with the cutoff frequency up to 107-108 frame/s and quite satisfactory dynamic spatial resolution reaching 15-20 line pair/mm experimentally. Result showed the resolution just by two-three times worse than the resolution obtained under the static conditions. But in these first experiments the rate of photoelectron image deflection on output screen (3 · 107 cm/s) exceeded the result achieved by Courtney-Pratt in his experiments with the magnetic deflection of photoelectron images. So this is how the principle of registration of FP individual phases at the expense of the deflection of photoelectron images in fast-changing electric fields was implemented experimentally for the first time in the world.
The second fundamental experiment initiated by Y.K. Zavoisky was carried out in 1952 by M.M. Butslov. The fact is that in the world practice IIT use was restricted due to the low luminance of obtained images. In the thirties German scientist G. Holst patented the scheme of image luminance intensification at the expense of series connection of several single-stage IITs in biplanar configuration. However, nobody could implement it practically. M.M. Butslov, great technologist and vacuum expert, managed to collect several two-electrode IITs with magnetic image focusing in one glass bulb. Uniqueness of this multi-stage IIT consisted in the fact that every subsequent stage was separated from the previous one by the vacuum-tight 10-micron mica; luminescent screen of the previous stage was applied on one side of mica, and cesium-antimonide (Cs3Sb) photocathode of the following stage was formed on the other side of mica. Coefficient of luminance intensification of one such stage was 20-30. Connecting five-six stages sequentially it was possible to implement the total intensification of image luminance by 105-108 times. And depending on the required coefficient of luminance intensification it was possible to choose either single-stage IIT with the magnetic focusing of photoelectron images (of M9 type) or IIT containing up to six single intensifying chambers connected to each other in one glass encapsulation .
On one significant day in 1953 Y.K. Savoisky and his colleagues tested the first six-stage luminance intensifier in the world – UM-95 type device designed by M.M. Butslov and his colleagues . Since then, similar devices have never been reproduced in any other place of the world. New intensifier allowed observing the image of individual electrons on the screen. Remembering this experiment, S.D. Fanchenko emphasized its scientific significance: fundamental limit of luminance intensification was achieved – further increase of intensification coefficient did not result in the increase of electron number in the registered image and therefore did not give any additional information except for the increase of electron noise level.
Afterwards, Y.K. Zavoisky, who has been the only possessor of image intensifiers with the marginal coefficient of intensification in the world for many years, managed to implement multi-stage IITs in science, engineering and medicine. Professor V.V. Prokofieva, worker of the Crimean Astronomical Observatory, was great enthusiast with the respect to the use of multi-stage IITs in astronomical research.
Open-mindedness rapidly led Yevgeny Konstantinovich to the explanation of the physical principles of pico-femtosecond electron-optical chronography. Y.K. Zavoisky finished his report on the seminar at the Kurchatov Institute in 1953 with the words: “I am not sure yet but it seems like IITs of PIM-3 type are capable to accomplish the time resolution shorter than 10 ns obtained under the conditions of frame-by-frame photography. Image focusing in it is very sharp – size of the image from perfect point source is 3-5 µm. As far as I know, there is no good theory on IIT. However, in accordance with the Fermat’s principle upon the focusing of point source radiation into the point image every elementary wave arrives to the image point without phase distortions. Therefore, basically obtainment of the perfect stigmatic image enables to transmit the infinitely-short signal. I invite all of you, my listeners, to solve this problem in my sector”.
In order to organize the next decisive experiment Yevgeny Konstantinovich instructed M.M. Butslov to attach the time-analysing IIT PIM-3 to the luminance intensifier UM-94. This is how the legendary time-analyzing IITs of UMI series, representing the combination of PIM-3 and intensifying stages M-9, were designed for the first time in the world and then started to be produced in our country [16-17].
Shortly, Sergei Dmitrievich Fanchenko received from M.M. Butslov six-chamber IIT UMI-95 and attached 300 MHz 100 W continuous generator to its PIM part. Two outputs of the generator were connected to the deflection plates of PIM-3 through 75 Ohm coaxial cables. Resonant circuit was assembled on the output terminals of each of two pairs. With its help it was possible to introduce the phase shift between two sinusoidal signals fed to every pair of plates. As result of this shift, there was opportunity to observe either circular scan (phase shift π/2) or Lissajous figures on IIT screen. When the scan speed was (2–3)·109 cm/s and spot had the size of (1–2)·10-2 cm the technical time resolution was limited to the value of 2-5 ps.
In order to carry out the dynamic tests with UMI-95 it was necessary to find the proper light source generating the radiation pulses with the duration which was less than the technical time resolution or shorter than 2-5 ps. It should be taken into account that then in 1953 lasers have not existed yet. Researchers had to use the glow of small spark dischargers in nitrogen occurring at high pressure. Duration of such spark discharge was intentionally shorter than the period of elliptical scan (3 ns). Densitogram of the spark glow scan showed that the time of spark discharge build up was less than ten picoseconds. The shape of time-spark intensity profile differed greatly from the shape of Gaussian curve. In order to measure the instrument function of UMI-95 there were no other independent light sources.
At the same time, photometer curves allowed to differentiate the changes of intensity time profile with the accuracy to several picoseconds. Further studies were carried out using PIM-3 attached to the five-chamber luminance intensifier through mica films. It ensured the registration of every single photoelectron leaving the input photocathode of time-analyzing IIT. Results led to the conclusions that IIT own time resolution was not worse than 10 ps.
On the basis of this data we can certainly conclude that picosecond electron-optical chronography came into being in 1954 in Moscow at the Institute of Nuclear Power Engineering by the efforts of Y.K. Zavoisky and his colleagues. This event was documented in the article of Y.K. Zavoisky (at that time he was corresponding member of the Academy of Sciences of the USSR) and S.D. Fanchenko called “On the Study of Ultra-Fast Processes” sent to the editorial body on the 1st of September, 1954, and published in 1955 in the Proceedings of the Academy of Sciences of the USSR (Proceedings of the Academy of Sciences, v. 100, No. 4).
Physics of Pico-Femtosecond Electron-Optical Chronography
Results of the unique experiments of Y.K. Zavoisky with IIT literally in several months found their brilliant interpretation in his new article sent on the 1st of December 1955 to the Proceedings of the Academy of Sciences and called “Physics of Electron-Optical Chronography” (Proceedings of the Academy of Sciences, 1956, v. 108, No. 2).
Yevgeny Konstantinovich understood very well that Fermat’s postulate is not applicable for the zero time of image placing in the focusing area. Since German scientist O. Scherzer  showed the mandatory availability of chromatic aberrations for the electron lens used in IIT in 1930s. Academician Lev Andreevich Artsimovich was actively involved into the study of the theory of photoelectron images formation in electrostatic IITs intended for the night viewing in 1940s. Relying on the general considerations he derived the formula for longitudinal chromatic aberrations. According to his research  longitudinal chromatic aberrations DZ in the focusing electrostatic lens of IIT can be expressed as follows:
ΔZ = m ∙ ΔVoz ∙ V/eE,
where Z is the axial coordinate; m and e are the electron mass and charge respectively; E is the intensity of electric field close to photocathode; ΔVoz is the spread of photoelectron initial speed along the axial component; V is the longitudinal component of axial speed which photoelectrons obtain at thr combined accelerating potential applied to IIT. The theory of spatial chromatic aberrations of L.A. Artsimovich became conventional theory for the description of quality of static images formed in IIT.
After ten years, on the basis of L.A. Artsimovich formula Y.K. Zavoisky and S.D. Fanchenko easily obtained the expression for the temporal aberrations of electron optics. The time of image formation Δσ in the focusing area was estimated taking into account the following:
Δσ = ΔZ/V.
Authors meant that if the accelerating potential on IIT is constant the electron speed close to the focusing area is constant too. Thus, in 1955 the famous correlation accepted by the world science was derived:
Δτ = m ∙ ΔVoz /e∙E.
Nowadays, it is known as the formula of Zavoisky-Fanchenko for the determination of contribution of longitudinal chromatic aberrations to the maximum achievable time resolution of IIT.
For the qualitative evaluations which are usually used by the developers authors introduced the simplified expression:
Δτ = a · 10–11/Е ,
where Е is the intensity of electric field which is in close proximity to photocathode expressed in the units of CGSE charge (1 CGSЕ = 30 V/mm), a is the dimensionless coefficient equal to 1-2 for silver-oxygen-cesium photocathodes. In accordance with the calculations of these authors for the device PIM-3 the maximum time resolution turned out to be equal to 5-10 ps. And the values Е = 2 СGSE (60 V/mm) and a = 2 are used here and it agrees well with the experimental results.
Zavoisky and Fanchenko did not limit their attention with theoretical consideration of unavoidable longitudinal chromatic aberrations of focusing lens. Laying the groundwork for the pico-femtosecond electron-optical chronography they considered the whole path of the formation of photoelectron images from photocathode to IIT output screen. Having analyzed the resulting images they came to the conclusion on the possibility of achievement of time resolution up to 10 femtoseconds (10-14 s). And indeed, during the next two decades the leap was made in the development of electron-optical photography. In the late 70s many papers of national and foreign researchers devoted to the improvement of IIT time resolution by more than one order were written; the value of time resolution reached and exceeded one picosecond.
In the article “Pico-Femtosecond Electron-Optical Chronography” (Proceedings of the Academy of Sciences of the USSR, 1976, v. 226, No. 5) Y.K. Zavoisky draws the following fundamental conclusion: “In order to master the range 10-12–10-14 s there is only one way: minimization of the chromatic aberration effect through the increase of electric field intensity close to IIT photocathode up to 103-104 CGSE units (30-100 kV/mm) particularly on the basis of pulse supply”.
Thus, the main physical principles underlying in pico-femtosecond electron-optical registration are as follows .
Due to the dispersion of group velocities in glass lens of deflective-type objectives minimum time of the image formation cannot be shorter than 50-100 fs. Therefore, in order to reach the range of short times of 10-100 fs it is necessary to use the projection lens of reflective type projecting the image plane (for example, entrance micrometer slit) on IIT photocathode.
The finite thickness of photocathode determines the spread of the moment of photoelectrons departure in photocathode plane turned into vacuum. Upon the photocathode thickness of 100~(10-6 cm) and initial electron velocity of ~108 cm/s the value of this spread should not exceed 10 fs. Evaluation is accomplished under the condition of the absence of multiple scattering of electrons in photocathode depth. (Unfortunately, correctness of this statement has not been subjected to the careful theoretical and particularly experimental verifications yet).
Photocathodes must have low surface resistance. Photocathode materials must withstand the impact of high-intense electric fields (at the surface intensity of 1-6 kV/mm under static conditions and up to 30-100 kV/mm under pulse conditions). (According to these recommendations the specialized studies on photocathodes for femtosecond IITs should be accomplished and it is necessary to learn how to measure the distribution of photoelectron initial energies in actual practice. The reason is that at the abovementioned intensities near-cathode electric fields can severely distort the energy levels in semiconductors and influence on the energy distribution of photoelectrons upon the external photoeffect).
It is evident that time resolution (up to 10 fs) can be achieved at the phase velocities of images deflection on the screen which exceed the light speed by many times. It is so-called technical time resolution, τт = δх/V or the time during which the scan travels with the velocity V equal to one resolution spatial element δх. However, the influence of edge effects in deflecting plates at such deflection velocities and finite size of the beam in deflection plates can considerably deteriorate the spatial quality of the images deflected at such velocity.
Interaction of the electrons carrying the FP image with fast-changing deflecting field has quantum-mechanical character. It can cause the peculiar “shot effect” of deflection. If the scan velocity is subjected to the random fluctuations then the limit of time measurement accuracy will be not more than several units of femtoseconds, as the evaluations showed.
The effect of signal tailing exists at the expense of Coulomb repulsion of the electrons which are the parts of one resolution element during the time of transit through IIT. Indeed, the electrons, which departure from photocathode during the time of 10 fs, form groups in the bundle with the thickness which is not more than 10-6 cm. In its thin layer during the whole transit through IIT Coulomb repulsion occurs and it unavoidably causes the bundle thickening. Estimated correlation for Coulomb component of time resolution indicates the fact that number of electrons N contained in the image spatial element does not exceed one electron at the time resolution of 10 fs (τк ~10-14 N). It means that the time analysis of photoelectron images in femtosecond IITs must be carried out at the level of counting of single photoelectrons.
Compulsory condition for the reliable registration of the images scanned on luminescent screen with pico-femtosecond time resolution is their considerable (up to 104–106) luminance intensification. Meeting this condition will ensure the mode of the registration of every single photoelectron departing from the photocathode of time-analyzing IIT and carrying the information of the registered FP. Therefore, high-quality luminance intensifiers for the images analyzed in time (possibly with the use of micro-channel plates (MCP)) as well as ultra-sensitive CCD-matrices (including the electron-sensitive matrices) for the further signal input and images processing on computer are required.
And in conclusion we want to remind that evaluating the contribution of individual components to the maximum time resolution of IIT our classics never forgot about well-known effect of electron spectral broadening determined on the basis of quantum-mechanical uncertainty relation: ΔεΔτ ≥ ħ. If we take into account that in ultra-short photoelectron pulse electrons unavoidably have velocity spread estimated on the basis of uncertainty relation, this spread can be compared with the velocity spread estimated on the basis of external photoeffect character. Taking into account this additional (quantum-mechanical) velocity spread the estimation formula for the maximum τmax time resolution was derived:
τпред = ħ / (еV0Е)1/2,
where ħ is the Planck’s constant; Е is the field close to photocathode in the units of CGSE charge; V0 is the initial velocity of photoelectrons; е is the electron charge. According to this formula, at the intensity of the field close to photocathode of 30-100 kV/mm using the silver-oxygen-cesium photocathode maximum time resolution will be maintained at the level of 10 fs.
* * *
Summarizing this section it should be noted that theoretical basics of pico-femtosecond electron-optical photography formulated in 1950s in the papers of the school of Y.K. Zavoisky and submitted to the scientific community at the IV International Congress for High-Speed Photography in Cologne (Federal Republic of Germany) in 1958  withstood the test of time during the half of century. Their validity was tested and successfully confirmed on the basis of the results of many IIT applications during the study of FPs in laser physics, laser plasma physics, semiconductor physics, fiber and nonlinear optics, photobiology and medicine. Time-analyzing IITs are actively applied in nuclear physics for the study of ionizing particle tracks in luminescent materials , for the photography of Cherenkov radiation of individual relativistic charged particle , synchronization of the moment of arrival of colliding beams of relativistic electrons in time in the experiments with synchrotron radiation , in the experiments with controlled thermonuclear fusion (CTF) in order to estimate the density, fractional composition and electron temperature of plasma by the luminance of spectral lines . Please, read the second part of the review in the next issue of the magazine on how applied physics devices impulse the rapid development of electron-optical diagnostic equipment.