EP0127735B1 - Photo cathode and method of manufacturing such a cathode - Google Patents

Abstract

An infrared detector includes a vacuum tube containing a photo sensitive layer comprised of densely packed needles arranged vertically on a substrate and having been grown as metal whiskers in a porous portion of the substrate. The substrate includes a metallic layer either in contact with or insulated from the needles depending upon the mode of the detecting system. The needles may face the incoming radiation, or may face away therefrom, in which case at least part of the substrate has to be transparent to infrared radiation. The radiation is either acquired directly or through an infrared optic or through a raster or line-scanning system. Photo emission from the needles can be used either directly for the production of an image or indirectly through a scanning process. The diameter and distance of the needles is significantly smaller than the radiation band to be detected.

Description

The invention relates to a photocathode according to the preamble of claim 1 and a method for its production (claims 8 and 9). It works on the principle of photofield emission and offers advantageous and new applications for photocells, photomultipliers, image converters and electron beam tubes (Vidicons).

The currently available night vision systems can be divided into two categories:

The first category is the residual light amplifier. They are based on the use of photoemissive surface layers (photocathodes) with high sensitivity to radiation in the visible and in the near infrared up to wavelengths of around 900 nm. These systems are relatively cheap, compact and powerful, but their effects are greatly impaired by haze and fog.

The second category is thermal imaging devices that work in the 3-13 µm range and capture the natural radiation (temperature radiation) emanating from all objects. These systems offer increased range, even in bad weather, but they are relatively expensive due to their complex image conversion and imaging technology - examples include vidicon picture tubes, optomechanical scanning and self-scanning integrated circuits. The sensor field must be cooled when using semiconductor detectors.

• - Höhere Leistungsfähigkeit der Nachtsichtgeräte der ersten Kategorie (Restlichtverstärker). Sowohl die Einstrahlung (Photonenfluss) des Nachthimmels, als auch der Reflexionsgrad des Chlorophylls in den Pflanzen nehmen oberhalb von λ=900 nm stark zu, so dass wesentlich kontrastreichere Bilder entstehen.
• - Bei etwa λ,=1,6 µm ist die Strahlungsstärke des Restlichtes des bedeckten Nachthimmels von gleicher Intensität wie die thermische Eigenstrahlung von Objekten mit normaler Umgebungstemperatur. Das heisst, dass ein im 1-2 µm-Band empfindlicher Detektor als Restlichtverstärker und zugleich als Wärmestrahlungsempfänger eingesetzt werden könnte.

The aim has always been to shift the upper limit of use XK of the photocathodes towards longer waves. An electron-emitting photodetector that can be used in the 1-2 pm range and above would offer the following advantages:

• - Higher performance of the night vision devices of the first category (residual light amplifier). Both the irradiation (photon flux) of the night sky and the degree of reflection of chlorophyll in the plants increase sharply above λ = 900 nm, so that images with much higher contrast are created.
• - At about λ, = 1.6 µm, the radiance of the residual light from the overcast night sky is of the same intensity as the thermal radiation of objects with a normal ambient temperature. This means that a detector sensitive in the 1-2 µm band could be used as a residual light amplifier and at the same time as a heat radiation receiver.

• - Die Erfassung des Temperaturstrahlungsbereiches mit Hilfe von Photokathoden - insbesondere oberhalb von λ=2 µm hätte eine drastische Vereinfachung und Kostensenkung der Kameratechnik zur Folge, da man die Vorteile der ersten und zweiten Kategorie von Nachtsichtgeräten praktisch kombinieren könnte. Die Vereinfachung entsteht u.a. dadurch, dass die ausgelösten Photoelektronen unmittelbar zur Bilderzeugung auf einem Leuchtschirm verwendet werden können, wie das beispielsweise bei den sogenannten Kaskadenröhren oder den Multi-Mikro-Kanal Röhren geschieht.

Since both methods work complementarily with regard to the object properties, the coupling results in an increased information content, higher sensitivities and completely new possibilities and applications of night vision technology that cannot be overlooked today.

• - The detection of the temperature radiation area with the help of photocathodes - especially above λ = 2 µm would result in a drastic simplification and cost reduction of the camera technology, since one could practically combine the advantages of the first and second category of night vision devices. The simplification arises, among other things, from the fact that the triggered photoelectrons can be used directly for image generation on a fluorescent screen, as is the case, for example, with the so-called cascade tubes or the multi-micro-channel tubes.

In the past, numerous solution proposals have been made to extend the detection range of photocathodes over λ = 1 µm.

In the case of photoelectron emission, the incident photons must have at least that

Have quantum energy h _VK = work function 0 to trigger the _{photo effect} , ie the work function is decisive for the cutoff frequency v _K or the cutoff wavelength λ _K = c _{/ νK} (c = speed of light). The situation is somewhat modified for semiconductors because the Fermi level is not occupied. In the case of a p-type semiconductor in which only the valence band is filled with electrons, the photoelectron must have the minimum energy E _G + E _{A in} order to be able to be emitted (EG = band gap, E _A = electron affinity).

The work function of most metals is about 4.5 eV, which corresponds to a cutoff wavelength of XK = 0.28 pm, i.e. ultraviolet light is required to trigger photoelectrons.

In known photocathodes, the work function by surface treatment is e.g. Coating with cesium or cesium compounds reduced to values by about 1 eV, so that there is also sensitivity to visible light and partly in the adjacent near infrared. For a further reduction of the work function, two methods have become known: the NEA method and the field-assisted photoemission.

In the NEA photocathodes (Negative electron affinity, J. Electron. Mater. 3, 9 (1974)), a strong bending of the band edges means that the vacuum level E _{vac is} lowered below the conductivity band edge E _L. The electron affinity as the difference between these two quantities (E _A = E _vac -E _L ) thus becomes negative, so that electrons that have been excited into the conductivity band can leave the solid. Although NEA cathodes have been known for 20 years and good quantum yields have been measured in the laboratory, their practical implementation has so far been less than encouraging, mainly due to manufacturing difficulties. In addition, NEA cathodes have a fundamental disadvantage, since the semiconductors used - for example gallium arsenide and silicon - also lead to a limitation of the spectral range due to the bandgap; the absorption edge is λ _K = 1.12 µm (Si) or XK = 0.92 µm (GaAs).

The well-known electrical peak effect is used in field-assisted photoemission. Due to the high field strengths occurring at the peaks, the potential barrier on the hard body surface reduced in height and width. The step-wise dependence of the potential energy of an electron on the distance from the metal surface without an electric field is deformed to a lower wall in the presence of a strong electric field. Due to the tunnel effect, electrons can also leave the solid whose energy is less than the work function. In metals, electrons are emitted from states just below the Fermi level. The field electron microscope is a known practical application of this effect.

• - Die Eindringtiefe des Lichtes muss gross gegen die Spitzenhöhe h sein.
• - Die Reichweite der Photoelektronen im Festkörper (Diffusionslänge) muss etwa gleich der Eindringtiefe sein, die im Material erzeugten Elektronen sollen ja die Oberfläche erreichen können.

To date, photofield cathodes (PFE cathodes) have been experimentally realized with semiconductor materials (IEEE Trans. Electr. Dev. 21, 785, 1974). For this purpose, a large number of tips are produced on a semiconductor crystal, for example made of silicon, by selective etching. The typical radius of curvature of the tips is r = 10 nm, while the distance between the tips is about 2 h = 20 pm. A schematic representation is shown in FIG. 1. The ratio of the emitting area to the total area r ² / h ² in this case is 2.5 · 10-8. In order to achieve high sensitivity, it is desirable that not only photoelectrons which are generated in the immediate vicinity of the tips can leave the surface, but that as large an area of the crystal as possible contributes to the photocurrent. This in turn can only happen if two elementary requirements are met:

• - The depth of penetration of the light must be large compared to the peak height h.
• - The range of the photoelectrons in the solid (diffusion length) must be approximately equal to the penetration depth, the electrons generated in the material should be able to reach the surface.

It has been found that on the basis of weakly doped p-type silicon material, favorable values for penetration depth and diffusion length are in the order of 100 pm. From this example it is also understandable that photofield emitters are currently only possible on the basis of semipermeable semiconductor material. In the case of metals, the penetration depth of the light is limited to approximately one tenth of the light wavelength due to the high electron density, so that metallic structures of the type shown in FIG. 1 are only photosensitive in the vicinity of the tips, and the detector sensitivity, based on the total area, thus becomes negligibly small.

• - Der endliche Bandabstand des verwendeten Halbleiters führt wiederum zu einer Grenzwellenlänge, ähnlich wie bei NEA-Kathoden und herkömmlichen Photokathoden.
• - Die Aktivierung von Photoelektronen im gesamten Oberflächenbereich zwischen den Spitzen, die einerseits für eine hinreichende Quantenausbeute unerlässlich ist, hat einen unerwünschten und unvermeidbaren Nebeneffekt: Es entsteht ein sehr hoher Dunkelstrom, aufgrund thermischer Anregung aus stets vorhandenen Oberflächenzuständen. Aufgrund der grossen Diffusionslänge und des Feldeinflusses gelangen die thermisch erzeugten Elektronen, ähnlich wie die Photoelektronen mit nahezu 100% Wahrscheinlichkeit in den Spitzenbereich und werden emittiert. Ein Betrieb ist daher nur bei starker Kühlung der Halbleiterkathode möglich.

The known semiconductor photo field emitters are burdened with some serious disadvantages, which question the successful use for IR photocathodes:

• - The finite bandgap of the semiconductor used in turn leads to a cutoff wavelength, similar to NEA cathodes and conventional photocathodes.
• - The activation of photoelectrons in the entire surface area between the tips, which is essential on the one hand for a sufficient quantum yield, has an undesirable and unavoidable side effect: A very high dark current arises due to thermal excitation from surface conditions that are always present. Due to the large diffusion length and the influence of the field, the thermally generated electrons, like the photoelectrons, reach the tip area with almost 100% probability and are emitted. Operation is therefore only possible when the semiconductor cathode is strongly cooled.

• - hohe Quantenausbeute,
• - niedriger Dunkelstrom,
• - einfache, zuverlässige Herstellungstechnik auch auf gekrümmten Flächen.

The invention is based on the object, starting from the aforementioned PFE cathodes, to create a photocathode for a photodetector for the infrared range above λ = 1 pm, which has the following properties:

• - high quantum yield,
• - low dark current,
• - Simple, reliable manufacturing technology even on curved surfaces.

• - hohe zulässige Betriebstemperaturen,
• - keine Kühlung,
• - gute Umgebungsstabilität,
• - keine Caesium-Oberflächenbehandlung und
• - lange Lebensdauer.

Other desirable properties are:

• - high permissible operating temperatures,
• - no cooling,
• - good environmental stability,
• - no cesium surface treatment and
• - long lifetime.

According to the invention, this object is achieved by a photocathode with the characterizing features of claim 1.

Embodiments of the photocathode according to the invention, applications of the photocathode and manufacturing processes are the subject of subclaims.

The transport of the photoelectrons to the tip is fundamentally different in the whisker structure (needle structure) according to the invention than in macroscopic semiconductor crystals. Since there is no band gap in metals, the lifetime of the photoelectron is significantly shorter, i.e. it relaxes in a short time and releases all of its excitation energy to the grid until it is in equilibrium with the other electrons. The average free path length is also limited to typical values of 10 nm by the slim shape of the needles. The average energy loss per scatter of the electron on a lattice atom or on the surface is about 0.1 eV at room temperature. When excited with infrared radiation of e.g. λ = 2 µm, i.e. a quantum energy of 0.62 eV, the electron will experience around 60 impacts until it is thermalized and cover a distance of about 0.6 pm during this time. Since the absorption length for 2 pm radiation in the needle structure is also of the order of 1 pm, a considerable part of the “hot” electrons will reach the tip area, where it will be sucked off by the external field. Electron-electron collisions can be neglected in these circumstances.

According to the knowledge gained so far, the emission behavior of the metal structure according to the invention is shaped by another effect which can prove to be very important for image conversion purposes. The part of the excited electrons that thermalizes without the Reaching the top releases all of its energy to the metal grid, thereby heating it up. With increasing temperature, the number of electrons increases in an energy interval above the average electron energy (broadening of the Fermi energy distribution). These thermally excited electrons have an increased tunneling probability in the potential threshold, the total emission current thereby increases significantly.

The total emission current is made up of directly emitted photoelectrons (photoemission) and indirectly emitted electrons by heating (thermal emission), so that overall a very high sensitivity arises. The relative proportion of the two types of emissions can be specifically influenced by dimensioning the structure, selecting the material and operating temperature and adapted to the respective task. The optical emission is advantageous for fast responding detectors, while the thermal induced emission is particularly suitable for image acquisition with mechanical or electronic scanning due to its storage and accumulation effect.

• - Die Eindringtiefe des Lichtes wird wesentlich gesteigert. Während sie bei kompakten Metallen weniger als ein Zehntel der Wellenlänge beträgt, kann sie je nach Packungsdichte und Metallart bis zu einem Mehrfachen der Wellenlänge betragen.
• - Die Metallmikrostruktur ist ein hervorragender Lichtabsorber, d.h. Reflexionsverluste können hier vernachlässigt werden, während sie bei Halbleiter-PFE-Kathoden bei 30% liegen.
• - Die einfallende Strahlung wird flächenhaft absorbiert, obwohl die Querschnittsfläche der Nadeln nur einen Bruchteil (1/2 bis 1/10) der Grundfläche beträgt. Da die Strukturelemente viel kleiner als die Wellenlänge sind, verläuft die Absorption nicht nach den Regeln der geometrischen Optik, sondern nach einem Prozess, der als kohärente Anregung von Streuzentren beschrieben werden kann. Es entfällt somit die bei makroskopischen PFE-Kathoden trivialerweise vorhandene Bedingung, dass auch die Täler zwischen den Spitzen zum Photoeffekt beitragen müssen.
• - Metalle besitzen im Gegensatz zu Halbleitern keine Bandlücke, welche durch Photoanregung überwunden werden muss. Damit entfällt grundsätzlich die Abschneidebedingung; es gibt keine Grenzwellenlänge λ_«.
• - Die Verwendung von Metallkristalliten statt ultrareiner, perfekter Halbleiterkristalle eröffnet entscheidende fertigungstechnische Vorteile:

Thanks to the special microstructuring, a number of advantageous effects can be combined:

• - The penetration depth of the light is significantly increased. While it is less than a tenth of the wavelength for compact metals, it can be up to a multiple of the wavelength depending on the packing density and type of metal.
• - The metal microstructure is an excellent light absorber, ie reflection losses can be neglected here, while they are 30% for semiconductor PFE cathodes.
• - The incident radiation is absorbed areally, although the cross-sectional area of the needles is only a fraction (1/2 to 1/10) of the base area. Since the structural elements are much smaller than the wavelength, the absorption does not proceed according to the rules of geometric optics, but according to a process that can be described as a coherent excitation of scattering centers. The condition, which is trivial in macroscopic PFE cathodes, that the valleys between the tips also have to contribute to the photo effect is thus eliminated.
• - In contrast to semiconductors, metals do not have a band gap, which must be overcome by photoexcitation. The cut-off condition is therefore basically eliminated; there is no cutoff wavelength λ _« .
• - The use of metal crystallites instead of ultra-pure, perfect semiconductor crystals opens up decisive manufacturing advantages:

Simple, cost-effective method of production thanks to the electrochemical coating processes described below, the production of large and arbitrarily curved cathodes, the use of environment-stable and operationally stable materials, the fact that the operating characteristics can be influenced to a large extent by the band structure of the metals used and the geometry of the needle structure.

The dark current behavior in the metal whisker structure is considerably more favorable than in the case of PFE semiconductor cathodes, since metals do not show the phenomenon of the surface states and the high diffusion lengths due to the missing band gap. At constant operating temperature, the dark current of the photocathode according to the invention is determined solely by the external field, and can therefore be adjusted very sensitively to an optimal low level by means of the pulling voltage or an auxiliary voltage. Since there is neither a cut-off condition nor reflection losses, incident photons can be detected by optical and thermal excitation with maximum quantum efficiency. The sensitivity threshold is limited solely by noise effects.

The surface resolution of a metal structure cathode is incomparably higher than that of conventional detectors with specific elements such as photodiode arrays, PFE semiconductor cathodes or polycrystalline coatings. Since the needle distances are smaller than the light wavelength, the resolving power of the detector according to the invention is in principle even better than that of the optical imaging. In real systems, where the resolution is limited anyway by other components, the microscopic character of the needle structure has a positive effect in other respects.

This is to be discussed in the example of the Vidocon tube: the electron beam of today's Vidocon tubes has a diameter at the sensor layer (retina) of typically 35 gm, while the needle diameter of the structure according to the invention is only 0.01 Jlm. One pixel is thus created by the combined effect of around 10 ⁶ neighboring peak emitters. This fact largely compensates for the inevitable short-range material and manufacturing-related fluctuations in the emission behavior of the tips. In addition, the current load of the whiskers is reduced to a tolerable level in the picoamp area, at which self-heating or burning out of the needle tips cannot occur.

• - Extrem kurze Ansprechzeiten und höchste Empfindlichkeit aufgrund der sehr kleinen eingesetzten Masse der Nadeln (1 pm lang, 10 nm stark). Zum Vergleich: pyroelektrisch empfindliche Schichten sind durchgehend 30 Jlm dick.
• - Der Hauptvorteil des erfindungsgemässen thermischen Detektors dürfte jedoch daraus resultieren, dass er in «Emission» arbeitet, im Gegensatz zu allen herkömmlichen thermischen Sensoren, und dadurch Photomultiplier und Bildwandlerröhren für das mittlere und ferne Infrarot ermöglicht, also Nachweiselemente, welche heute noch nicht verfügbar sind.
• - Der Dualismus von optisch und thermisch erzeugter Elektronenemission bietet nicht nur den eingangs erwähnten Vorteil, die Grenzwellenlänge in den Übergangsbereich von Restlicht und Wärmestrahlung ( 1-2 Jlm-Band) zu verschieben, sondern rückt die Verwirklichung einer «multispektralen» Bildröhre, welche sowohl im Sichtbaren als auch im nahen Infrarot und im Wärmestrahlungsbereich arbeiten kann, in den Bereich des Möglichen.

The possibility of using the thermal emission effect mentioned for infrared radiation detection has not been recognized or realized so far, at most this effect has been registered as a disturbance in the form of high dark currents. In addition to the known thermal detectors, the invention defines a new, technically relevant type of detector with remarkable properties, which are based on the four processes: bolometer effect, thermovoltaic effect, thermopneumatic effect and pyroelectric effect:

• - Extremely short response times and maximum sensitivity due to the very small mass of the needles used (1 pm long, 10 nm thick). For comparison: pyroelectrically sensitive layers are 30 Jlm thick throughout.
• - The main advantage of the thermal detector according to the invention should, however, result from the fact that it works in «emission», in contrast to all conventional thermal sensors, and thereby enables photomultipliers and image converter tubes for the middle and far infrared, that is to say detection elements which are not yet available today .
• - The dualism of optically and thermally generated electron emission not only offers the advantage mentioned at the beginning of shifting the cut-off wavelength into the transition range from residual light and heat radiation (1-2 Jlm band), but also moves towards the realization of a «multispectral» picture tube, which is both in the Visible as well as in the near infrared and in the heat radiation area can work in the area of the possible.

production method

The use of microstructures of the type described as a surface emitter or as an image converter requires that the geometry of the needles, that is to say the needle height, tip radius and tip spacing, can be formed to the greatest extent evenly. It has now been found that this difficult task can be solved in two steps with a relatively simple electrochemical process, which is described below.

In a first step, a thin, porous oxide layer is produced on a suitable conductive substrate by anodic oxidation. In the second step, metallic nuclei are generated in the oxide pores, which eventually grow in the form of whiskers beyond the oxide surface. Similar processes are known in the field of the production of solar absorber layers (e.g. DE-AS 26 16 662, DE-AS 27 05 337).

The first step:

For optimal formation of the porous oxide mask, a number of conditions with regard to the substrate, the electrolyte and the anodizing parameters must be met. Suitable substrates are metals which provide dense, firmly adhering, electrically insulating surface oxides by anodization, such as aluminum, magnesium, titanium, tin, tantalum, zircon.

In order to build up layers that are as defect-free as possible, contamination, surface defects and gross structural defects must be avoided. The use of single crystals is ideal, although sufficient homogeneous cover layers can also be achieved on polycrystalline material, since the grain structure does not necessarily continue in the oxide layer. Photocathodes for electron beam picture tubes are advantageously operated in «transmission», i.e. the radiation falls from the back onto the photosensitive layer, the substrate must be permeable and carry an electrically conductive electrode. In this case, single-crystalline silicon or germanium can be considered as a suitable substrate for the embodiment according to the invention.

The electrolyte must contain at least one oxygen-containing compound, preferably dilute acid such as sulfuric acid, phosphoric acid, tartaric acid or salt solutions, alcohols etc. included in order to form the oxide. On the other hand, the electrolyte must have a certain ability to redissolve the oxide under the influence of the anodic field. If these two mechanisms interact correctly, an oxide layer of approximately 0.5 gm thick or thicker is formed, which has cylindrical pores in an even distribution. The pores are to be regarded as current paths, which allow the oxide-metal interface to progress continuously into the substrate. A strong redissolution of the oxide takes place in the pores during the anodizing process. At the base of the pore, a thin oxide skin, the barrier layer, of a few nm thickness forms. This layer is comparable to the thin anodic oxide skin, which arises in non-redissolving electrolytes and whose thickness grows in proportion to the voltage.

This anodizing method, known per se, which is used in the art e.g. is used in the anodizing of aluminum materials, can be used very advantageously to form the oxide mask according to the invention, since extremely homogeneous and reproducible pore structures arise. Although the effect of an electrolyte cannot generally be predicted, the pore diameter, oxide thickness and pore spacing can be set in a systematic manner by means of temperature, concentration and current density for a given system of electrolyte and substrate.

The second step:

In the second process step, metal whiskers (needles) are stored from electrolyte containing metal salts, starting at the base of the pores. This process also has a certain similarity to the anodizing technique (electrolytic coloring of anodic Al ₂ 0 ₃ layers). The difference, however, is that here the metal structure protrudes from the thin oxide skin, while metal pigments are embedded in thick oxides. The metals nickel, cobalt, iron, manganese and chromium have proven to be a favorable material for the needles, which tend to columnar or whisker-like growth during electrochemical crystallization and have relatively high lattice stability against thermal and electrical loads. The thickness and height of the metal needles can be set within certain limits, which also depend on the oxide pore structure, using the deposition parameters.

Further advantages, features and applications result from the designs shown in the figures.

• Figur 1 einen Photofeldemitter nach dem Stand der Technik,
• Figur 2 eine photoempfindliche Schicht eines erfindungsgemässen Photodetektors,
• Figuren 3 und 4 Raster-Elektronen-Mikroskopaufnahmen von photoempfindlichen Schichten,
• Figuren 5 und 6 zwei erfindungsgemässe Photodetektoren als Bildwandlerelemente,
• Figuren 7 und 8 zwei Aufnahmesysteme mit erfindungsgemässen Photodetektoren.

Show it:

• 1 shows a photofield emitter according to the prior art,
• FIG. 2 shows a photosensitive layer of a photodetector according to the invention,
• FIGS. 3 and 4 scanning electron microscope images of photosensitive layers,
• FIGS. 5 and 6 two photodetectors according to the invention as image converter elements,
• FIGS. 7 and 8 show two recording systems with photodetectors according to the invention.

FIG. 1 schematically shows a cross section through the photosensitive layer of a semiconductor photo-field emitter 2 according to the prior art. A large number of peaks 6 of height h = 10 μm were produced on a p-doped semiconductor crystal 4 by selective etching. The typical radius of curvature of the tips is r = 10 nm, the distance between the tips 6 is 2 - h = 20 _J lm. The absorption length W, which corresponds to the penetration depth of the light, is about 100 µm as large as the diffusion length 1 (range of the photoelectrons). Further properties of the photo field emitter are described in the assessment of the prior art.

FIG. 2 shows a schematic cross section through a photocathode 8 according to the invention. On a metallic base 10 there is a porous oxide layer 12, in the pore channels 13 of which metal needles 14 (rods or whiskers) are deposited, which protrude beyond the oxide surface 16.

Typical and favorable values are:

The absorption length W is then approximately 1 to 2 μm. The condition W> is useful when using a base 10 made of metal. In the embodiment shown, the metal needles 14 touch the base 10. The metal structure shown can be used as a photo-emitting cathode. The barrier layer between the base 10 and the needles 14, which is usually present after the oxidation, has been removed so that the needles 14 are in direct galvanic contact with the conductive base 10. The selective dissolution of the barrier layer is achieved by anodic etching in a non-oxidizing acid which does not or only weakly attacks the base 10.

On the other hand, if the metal structure according to the invention is to be used as a photosensitive retina in a vidicon tube, it is preferable not to remove the barrier layer, but rather to strengthen it. Compared to the base 10, the needles 14 then form a large number of specific capacitors, which charge positively on the needle side when exposed to photons due to the electron emission.

FIG. 3 shows a scanning electron microscope (SEM) image of the oxide layer 12 in cross section. The pore channels 13 are greatly expanded by etching for better visibility. The magnification is 24,000. The large number of pore channels 13 perpendicular to the oxide surface 16 can be seen.

FIG. 4 shows a SEM image of a finished metal structure photocathode 8 in a magnification of 20,000 times: to make the needles 14 easier to see, the surface was coated with a thin gold layer. This results in the matchstick-like rounding of the needles 14 (free ends 18), which taper to a point. The carpet-like structure can be seen, which is formed by the large number of needles 14 located next to one another.

FIG. 5 shows the application of the photocathode according to the invention to the photodetector 20, which is designed as an image converter element with secondary electron amplification and a fluorescent screen. An IR window 24, an IR-transparent base 26, the photosensitive layer 28 consisting of oxide layer 12 and the needles 14, a multi-microchannel amplifier 30 and a fluorescent screen 32 are located in a vacuum-tight socket 22 one behind the other in the direction of light incidence.

An object field is imaged on the photosensitive layer 28 with a transparent base 26 by means of infrared optics (not shown). As in conventional residual light amplifiers, the multi-microchannel amplifier 30 operating as a secondary electron multiplier is at anode potential. The accelerated and amplified electron stream then strikes the fluorescent screen 32. The resulting image is observed directly or processed further by means of fiber optics, light amplifier tubes or electronically.

FIG. 6 shows a photodetector 33 for carrying out a previously unknown image converter method in which the effect of the photocathode according to the invention is directly coupled to a plasma display element.

In the vacuum-tight socket 34, an IR window 36, an anode space 37 with a grid anode 38, a photocathode 40, consisting of photosensitive layer 28 on a metallic base and an insulating layer 41, a plasma gas space 42 and a viewing window 44 are located one behind the other in the direction of light incidence electrically conductive coating 48.

In this case, the photocathode 40 is designed as a multi-element detector field. The detector elements are electrically insulated from one another and from the plasma gas space 42 (layer 41) and have a size of approximately 1 mm ² . The photosensitive layer 28 of the detector field is directed towards the object and lies opposite a grid anode 38. The plasma gas space 42 on the back of the detector field is delimited by the viewing window 44, which is electrically conductive hige transparent coating 48 carries and acts as a counter electrode (plasma electrode). The anode compartment 37 is evacuated. In the plasma gas space 42 there is gas at low pressure, for example 0.1 mbar. A voltage is applied between the grid anode 38 and the coating 48 serving as counterelectrode, which is composed of a DC voltage component and a superimposed AC voltage; the photocathode 40 is not connected. As in the other embodiments, the DC voltage component is regulated in such a way that a low dark current is present. The AC voltage has the task of holding the potential of the photocathode 40 in the vicinity of the potential of the counterelectrode as long as there is no irradiation, which is easily achieved by the different distances and the dielectric properties of the detector base. If a detector element is irradiated, it emits electrons and charges positively, which means that the potential difference to the counterelectrode increases and the ignition voltage of the plasma is exceeded. The plasma, once ignited, reduces the internal resistance in the plasma gas space 42, so that the plasma is extinguished again if it is not continuously irradiated further.

In addition to this simple operation, a number of special operating modes are conceivable, as have been developed similarly with the known plasma panels, e.g. Storage mode with superimposed pulses, alternating operation with constant charging of the detector element by IR radiation and subsequent discharge by plasma excitation.

FIG. 7 shows a recording system (camera 50) with a photodetector 52 according to the invention with scanning on the detector side. An optical-mechanical scanning system 51 is connected upstream of the photodetector 52. The photodetector 52 consists of a vacuum-tight housing 54, an IR window 56, a grid anode 58 and the photosensitive layer 28 on a metallic base 62. The optomechanical scanning system 51 (raster system) conducts it Radiation from an object field onto the photosensitive layer 28, but not simultaneously as in an optical image, but, for example such that only a small section of the field of view is passed over the photosensitive layer 28. In this way, the signals of the object field are successively output by the photosensitive layer 28 and can then be displayed again on a display device by means of a signal processor. Many variations of this principle are known, e.g. object-side scanning, line-by-line scanning, etc. The detector-side scanning shown is suitable for the sensor layer according to the invention, since it can be produced particularly over a large area and homogeneously. An advantage of this arrangement is the almost unlimited geometric resolution, the simple production technique of the photosensitive layer 28 on a metallic base 62, the omission of integrated multiplex circuits and the low optical and structural complexity of the camera 50.

FIG. 8 shows a recording system (picture tube 64) with electron beam scanning (vidicon picture tube). The photocathode according to the invention serves as a retina layer 60.

The picture tube 64 has the following structure:

The light entry side of the vacuum-tight housing 66 is formed by an IR window 68. Behind the IR window 68 is the electrically conductive rear side 70 of a high-resistance, transparent base 72, which carries the light-sensitive layer (retina layer 60) according to the invention on its front side. Opposite the retina 60 is a grid anode 74. Further behind is a device 76 for generating an electron beam 78 for scanning. The picture tube 64 is surrounded by known deflection and focusing elements 80. The high-resistance, transparent base 72, which has a highly conductive surface on its rear side 70, can be formed, for example, from single-crystal silicon with n ⁺ doping on the rear side. The bias of the retina layer 60 is a few volts positive with respect to the hot cathode of the device 76, so that the electron beam 78 strikes the retina layer 60 with low energy. The anode voltage at the grid anode 74 is set to optimal dark current. A special device 76 for beam generation with an off-axis source and thermal shielding ensures that no undesired IR radiation from the system falls onto the retina layer 60.

The object field (left, not shown) is imaged onto the retina layer 60 from the rear by means of infrared optics (not shown). During the image change time of the system of, for example, 1/25 second, the retina layer 60 is positively charged by photoelectron emission in proportion to the radiation intensity at the individual pixels. The photoelectrons are sucked off at the grid anode 74. The electron beam 78 erases the charge in time with the frame rate and places the retina layer 60 at the potential of the hot cathode (in 76). This creates a displacement current proportional to the charge in the retina layer 60 (junction capacitance), which can be processed as usual as a video signal 82, stored and read out again on an image monitor. As with other vidicons, the incoming parameters retinal capacity, retinal conductivity, sampling frequency, beam intensity, potentials etc. must of course be carefully coordinated. In this case, to increase operational safety, it is advisable to strengthen the natural barrier layer of the oxide mask by special measures, e.g. through a space charge zone in the semiconductor base. Layer production example:

The photosensitive layer can be applied to a base made of pure aluminum, for example as follows: The substrate is first degreased in a conventional manner in organic or alkaline media, then pickled in 5% sodium hydroxide solution at 60 ° C. for 5 minutes, rinsed in water and briefly immersed in 10% nitric acid at room temperature and rinsed clean again. After this pretreatment, the oxide layer is built up in 10% phosphoric acid at a bath temperature of 18 ° C and an alternating voltage of 16 volts in 20 minutes. After the intermediate rinsing, the barrier layer is etched, in a solution of 60 g / l MgCl ₂ using 6 volt alternating voltage for a few minutes and then immediately thoroughly washed.

The metal structure is created in a bath of 70 g / l NJS0 ₄ - 6H ₂ 0 and 20 g / l boric acid at room temperature with an alternating voltage of 12 volts in 15 minutes. After careful rinsing in the cascade, lastly at least 10 minutes in running deionized water, the layer is dried in slightly warmed air and, if possible, kept under vacuum or further processed (sealed).

Claims (10)

1. Photocathode for a photodetector for conversion of infrared radiation into an electrical signal comprising an infrared sensitive layer being arranged in a vacuum tube and emitting electrons by photo-electric field emission on sharp tips, characterised in that the sensitive layer comprises a plurality of densely packaged needles (14) extending perpendicularly to a substrate (10), the needles (14) having an electric conductivity like a metal, the diameter and the distance of the needles (14) being smaller by about one order of magnitude or more than the wavelength of the radiation to be detected and the free ends (18) of the needles (14) being the electron-emitting tips. 2. Photocathode as in claim 1, the substrate (10) being electrically conductive, the needles (14) of the sensitive layer being in contact with said substrate (10).

3. Photocathode as in claim 1, the needles (14) being electrically insulated with respect to their substrate carrier (72) by an isolating layer.

4. Photodetector with a photocathode as in claim 1 or 2, characterised in that the substrate (26) of the sensitive layer is transparent to the radiation to be detected, that an optical system is used to form an image of the scene on the sensitive layer (28), and that the emitted electrons energize a fluorescent screen (32) and render the object visible.

5. Photodetector with a photocathode as in claims 1 and 2, characterised in that the sensitive layer (28) and the metallic substrate are separeted into sensor elements arranged on an isolating layer (41 ), that an optical system is used to form an image of the scene on the sensitive layer (28), and that the sensor elements charged by field emission produce a gas discharge on their back sides in a gas-filled cavity by an electrically conductive coating (48), which functions as a counter electrode, so rendering the object visible.

6. Photodetector with a photocathode as in claims 1 and 2, characterised in that the sensitive layer (28) is arranged on a metallic substrate (62), that an image of the scene is formed upon the layer (28) by means of a mechanical scanning system (51) and that a video signal is produced by field assisted photo emission at the photodetector (52), the video signal rendering the scene visible by means of a signal processor and a monitor.

7. Photodetector with a photocathode as in claims 1 and 3, characterised in that the substrate (72) of the sensitive layer is transparent to the radiation to be detected, that an optical system is used to form an image of the scene on the sensitive layer (60), that the sensitive layer (60) is charged by field assisted photoemission and that a video signal (82) is produced by scanning and extinguishing the charge by an electron beam (78).

8. Method of manufacturing a photocathode as in one of the claims 1 or 3, characterised in that

- in a first step a porous oxide layer (12) is produced by anodic oxidation on a substrate (10) and

- in a second step metallic needles (14) are provided in a galvanic process inside of the pores of the oxide, the needles (14) extending perpendicular to the substrate (10) and projecting beyond the surface of the oxid layer (12), the diameter and the distance of the needles (14) being smaller by about one order of magnitude or more than the wavelength of the radiation to be detected and the free ends (18) of the needles (14) being the electron-emitting tips.

9. Method of manufacturing a photocathode as in one of the claims 1 or 2, characterised in that

- in a first step a porous oxide layer (12) is produced by anodic oxidation on a substrate (10)

- in a second step the barrier layer dividing the pores from the substrate is etched anodically and

- in a third step metallic needles (14) are provided in a galvanic process inside of the pores of the oxide, the needles (14) extending perpendicular to the substrate (10) and projecting beyond the surface of the oxid layer (12), the diameter and the distance of the needles (14) being smaller by about one order of magnitude or more than the wavelength of the radiation to be detected and the free ends (18) of the needles (14) being the electron-emitting tips.

Images