DE19629583A1 - Emitter and / or detector component for submillimeter-wave radiation and method for its production - Google Patents

Abstract

The invention concerns a superconductive electronic component which displays specific properties of an emitter and/or detector for electromagnetic radiation in the submillimetre wave range. The component comprises a planar network of microbridges (webs) formed in a thin layer of a high-temperature superconductor. The latter is grown epitaxially on the substrate with the CuO2 planes either perpendicular or inclined at an angle THETA (1 DEG < THETA <89 DEG ) to the substrate surface. In this way, each microbridge receives a sequence of stacks of superimposed (intrinsic) Josephson contacts. The invention also concerns superconductive connections (series and parallel) between individual microbridges, whereby switching circuit parameters, such as adaptation of the impedance to the radiation space and maximizing of the radiated output, can be optimized. The frequency and intensity, for example, of the radiation field can be influenced (e.g. modulated) by external means of an electronic control. In particular, this component can continuously cover the frequency range between the far infrared and the microwave range. The invention also concerns some applications of the proposed component, covering both the emission and detection of electromagnetic radiation.

Description

Field of application of the invention

The invention relates to a superconducting component, production technologies and applications fields and in particular a novel emitter and / or detector for the submillime terwelle area, as well as its diverse applications.

Characteristic of the known technical solutions 1. Arrangements of natural Josephson contacts

The first findings on the usability of Josephson contacts as tunable Microwave emitters and detectors go back to the early work of B. Josephson and S. Shapiro back. However, it was recognized very early on that a single Josephson- Contact emits too little power and also has too wide a spectral distribution points to find practical application as a microwave radiator. These shortcomings are known to know about the use of stacking sequences of Josephson contacts become sore [Jain u. a. 1984; Bindslev Hansen and Lindelof 1984; Lukens 1990].

Once the coupling between the Josephson contacts is strong enough, it can go to itself synchronization with the effect of a coherent radiation emission from all contacts [Lukens 1990; Konopka 1994]. Possible coupling mechanisms and coupling strengths are have already been examined in detail [Jain u. a .; Lukens 1990]. The line width of the elektroma The genetic radiation emitted by the crystal stack decreases to the extent that the number of Josephson contacts in the array increases. It can become very small when there is a large number of Josephson contacts [Lukens 1990; Wiesenfeld u. a. 1994; Konopka 1994].

The power of the emitted radiation also increases with the number of Josephson Contacts and can also be remarkable in large arrangements (P 1 mW), which for many practical applications are sufficient [Bindeslev Hansen and Lindelof 1984; Jain u. a. 1984; Knockout  nopka u. a. 1994; Wiesenfeld u. a. 1994]. It is important that a good impedance match to the Load resistance (e.g. the free space) is reached because otherwise most of the radiation does not emerge from the component, but is used up inside by multiple reflection.

Another point of view are different loss mechanisms in the contacts. For example, the capacities of the contacts are responsible for a loss of power at higher frequencies [Lukens 1990; Wiesenfeld et al. 1994], which indicates the use of the smallest possible Josephson contacts. Small contact areas are also favorable for reducing the noise and the spectral line width [Kunkel and Siegel 1994; Konopka 1994 ]. The current flow through the contact becomes inhomogeneous as soon as the relationship exists between the width w of the transition and the Josephson penetration depth λ _j , w 4λ _j ., [Kunkel and Siegel 1994 ]. It is also clear that for optimal phase synchronization the critical currents (I _c ) through all Josephson contacts have to be very similar. In general, a homogeneity of at least ± 5% in linear stacking sequences is required [Konopka 1994].

The high requirements mentioned so far can be somewhat weakened if one distributed arrangement of equidistant Josephson contacts along a propagation line is chosen. However, the distance between the Josephson contacts and the selected radiation wavelength match [Kukens 1990; Han et al. a. 1994]. Apparently it will the tunability of the radiation frequency was severely curtailed, but in favor a significantly increased emission performance.

There have been numerous experimental studies on the arrangement of Josephson contacts, and some remarkable results have been achieved. Most of them are based on conventional superconductors with low transition temperatures, such as. B. Nb / Al-AlO _x / Nb three-layer structures. Complete phase synchronization has been demonstrated in linear arrangements of 100 such Josephson contacts [Han et al. 1993]. In some cases, a broadband antenna (e.g. logarithmic spiral antenna, etc.) was integrated on the chip, and the radiation emitted to the outside was measured. In other cases, a further Josephson contact was coupled as a detector to the arrangement of the radiation emitters via a propagation line. Some of the best results achieved are as follows: An emission of P = 50 µW at a frequency of ν = 400-500 GHz has been demonstrated in a distributed arrangement of 500 Josephson contacts [Han et al. 1994]. In another circuit (arrangement of 10 × 10 contacts) radiation with a line width of Δν = 10 kHz and a tunability of ν = 53 - 230 GHz [Booi and Benz 1994] was measured.

With the discovery of high-temperature superconductors in La-Ba-Cu-O by G. Bednorz and KA Müller in 1986, and the associated increase in critical temperature in similar copper oxide (cuprate) compounds up to T _c <160K, has created an enormous expectation for superconducting electronics at the temperature of liquid nitrogen and above. In fact, dozens of research laboratories around the world have made Jo sephson contacts in these new materials using several different techniques. The radiation emission as a result of the Josephson alternating current in artificial Josephson contacts in high-temperature superconductors was measured and analyzed [Kunkel and Siegel 1994]. In the same work, phase synchronization in step contacts over a broad frequency band of ν = 80-500 GHz was demonstrated. In larger arrangements (arrays), only a partial (up to 4 Josephson contacts) and also unstable phase synchronization could be achieved [Konopka 1994]. The reason has been found to be that this is a result of the inhomogeneity (dissimilarity) of such step contacts in high-temperature superconductor-Josephson contacts, where critical current variations of up to ± 50% occur [Konopka 1990]. In another work, arrays of 5 and 10 Josephson junctions from adjacent step contacts are investigated, in turn with only partial phase synchronization and very low radiation power [Kunkel and Siegel 1994].

In several publications, spectacular statements about promising to be future applications of grouped Josephson contacts made. At for example, applications as generators and detectors of radiation in the GHz and THz area predicted [Jain u. a. 1984], and even more specific in radio astronomy and radio spectroscopy of heavy molecules [Konopka 1994], in voltage normals [Ono u. a. 1995], etc. So far, all of these applications have been difficult because of the above have not been put into practice. It is generally believed that for spectroscopic applications a radiated power of 0.1-1 mW with continuous tunability is required. And so far this has not succeeded [Konopka u. a. 1994].  

US Pat. No. 3,725,213 describes a superconducting barrier component and its production technology, which is also used as a generator and detector for Millimeter wave and sub millimeter wave radiation is suitable and is based on a granular Structure of the superconducting material supports. While a higher radiation power or Sensitivity due to the summation of many Josephsonian contacts between crystal bodies there are few opportunities for reproducible production, electronic control, phase synchronization and impedance matching to the vacuum. The This component can use a current-pulse-induced magnetic field between the two contacts be switched normally and superconducting, however, a control of the emitted radiation frequency over the magnetic field-dependent energy gap not tried.

A superconducting device is described in US Pat. No. 4,837,604, the vertical Sta Sequences of Josephson contacts and serial interconnection of stack sequences applies. It is designed as a switch with 3 electrical contacts and is intended to be simple Jo sephson contacts or lateral arrangements of such contacts in analog and digital Replace switching arrangements. Radiation emission is not a goal for the device, yet his technical interpretation would allow such a goal.

In US Pat. No. 5,114,912, a high-frequency oscillator is described which relates to one in the Flat lay arrangement of Josephson contacts establishes. It is powered by a direct current source excited, and the frequency can be tuned using this direct current. In Pedance adjustment can be done by choosing the number of Josephson contacts in the area Arrangement, or by connecting appropriate shunt resistors. As A tunable DC-AC converter is used in the GHz range and even specified up to THz frequencies.

A disadvantage of this component is its explicit limitation to a two-dimensional one two-dimensional arrangement of adjacent Josephson contacts on one chip. A such geometry implies a strict limitation of the maximum possible number of Jo sephson contacts. The minimum area of a single contact is due to the limited performance of photolithography at about 1 µm², the requirements the uniformity is high, a considerable critical current is required (not a little less than 1 mA for optimal performance), which is also resistive via cables with low contact stand must be fed. On the other hand, the requirement for phase synchronization limits the  Longitudinal expansion of the component to about λ / 4, a value that for a frequency of 1 THz is about 75 mm. If you allow small distances between the contacts, then you can one can expect a maximum of one to two thousand such individual elements. In practice Arrangement of 10 × 10 = 100 contacts realized. As already explained, such provides Component design not the minimum performance required for interesting applications from 0.1-1 mW. It should be pointed out that the alternative proposed according to the invention Designs allow a much higher packing density of Josephson contacts with the The prospect of a significantly higher output and thus a large number of new products Turnings that are not possible with flat (planar) arrangements, as shown there.

EP 446146 describes a three-layer Josephson contact which has superconducting electrodes on both sides which consist of L _y Ba₂Cu₃O₄ (L _y is Y or a rare earth metal, 6 <y7). The non-superconducting barrier here consists of Bi₂Y _x Sr _y Cu _z O _w , 0 x 2, 1 y 3, 1 z 3 and 6 w 13. However, information on the properties of the component (such as I _c , R _n , IV characteristic, microwave modulation) or for the representation of lateral or vertical stack structures and their possible applications.

A magnetic control mode for the emitted or detected radiation frequency is used described in EP 513,557, where the component according to the invention consists of vertical stacks follow from Josephson contacts, there is also the galvanically connected to the side have conclusions. Between each pair of adjacent superconductors / barriers / superconductors Structures (SIS structures) a further superconducting layer is provided, which is characterized by a Insulator layer is separated on both sides from the adjacent Josephson contact. This layer should serve to control the individual component: By running a current over the side Conducting contacts through this control layer (see also US Pat. No. 3,725,213) is said to be a magnetic field can be generated, which affects the energy gap of the Josephson contact.

This component has several disadvantages, which its practical implementation under the currently known material properties and available microfabrication Make technologies impossible. In particular, (i) its manufacture requires deposition superconducting contacts of about 0.01 µm width over insulating layers on two against overlying side surfaces of stack structures grown perpendicular to it with Josephson Contacts. No technology is currently available for this. Furthermore, they are (ii) none Superconductors are known to have such large currents to generate the high magnet required  field strengths that can be used to reduce the energy gas in high-temperature Superconductors within the limits of the required geometric component dimensions would do. Even assuming that this was somehow possible, (iii) the Ma gnetfeld affect a certain control layer more than just a Josephson contact undesirable "cross-talk" between multiple Josephson contacts and the sponsors thwart tab view. After all, (iv) the radiated power would be too low for many users due to a circuit optimization of the Josephson contacts for impedance matching the outside space is not provided.

2. Natural stacking sequences of Josephson contacts in copper oxide superconductors

In the very first publication on copper oxide superconductors, J.G. Bednorz and K.A. Müller 1986 the opinion that La-Ba-Cu-O due to its distinctive layer structure is a quasi two-dimensional superconductor. Subsequently, this hypothesis was confirmed by a Large number of experimental results confirmed in various cuprate superconductors [e.g. B. Bozovic 1991], being direct evidence in the observation of superconductivity in Bi₂Sr₂CaCu₂O₈ layers the thickness of a unit cell is [Bozovic u. a. 1994]. Of further is the critical current along the c-axis (i.e. perpendicular to the CuO₂ layers) much smaller than in the direction parallel to the CuO₂ levels. It is not zero, i. that is, it can by flow from a supra current in the direction of the c-axis. However, this implies the property that the planar 2-dimensional superconducting rods over Josephson tunnels are weak are coupled.

In other words, cuprate superconductors can be considered a natural (intrinsic) stacking film of Josephsen contacts with a distance of 6. . . Take 25 Å. A theoretical one A model for a stacking sequence of Josephson-coupled superconducting layers was developed already published 25 years ago [by Lawrence and Doniach 1971], and since then extensive richly examined. The model's predictions relate to a nonlinear current-voltage (I-V) characteristic curve, microwave radiation-induced I-V stages (Shapiro stages), and the micro wave emission by applying DC voltage.

In fact, all of these properties were observed in cuprate superconductors, first in Bi₂Sr₂CaCu₂O₈ [Kleiner et al. 19921 and then also in (Pb _y Bi _1-y ) ₂Sr₂CaCu₂O₈, Tl₂Ba₂Ca₂Cu₃O₁₀, etc. [Kleiner and Müller 1994; Müller 1994, 1996, as well as through other groups, R´gi and others 1994, Irie and others 1994, Schmidl and others 1995, Yasuda and others 1996, Tanabe and others 1996, Yurgens and others 1996, Seidel and others 1996, Xiao and others 1996]. Most of these results were found in small single crystals, but mesa structures were also etched from single crystals or thin layers [Schlenga et al. 1995]. In some cases, phase synchronization of more than 1000 Josephson contacts was achieved [Schlenga et al. 1995], but in general only partially achieved, as shown by the large number of branches in the IV characteristics (in all previous work). In particular, if the Jo sephson contacts belonging to a stack are not all identical, ie if their critical current differs from contact to contact, they do not return to the normal conducting state at the same point on the characteristic curve as soon as the current flow is increased . As a result, the radiation emission must result from the superposition of coherent (narrow-band) radiation, which has also been observed [Schlenga et al. 1995, Müller 1996]. These differences in the properties of the Josephson contacts result from imperfect crystal growth and the lithography process for defining the mesa structures, which also leads to differences in the cross-sectional area of the mesa structures. The highest frequency of emitted radiation measured so far is ν = 95 GHz, which is, however, the result of inadequate laboratory detection technology [Müller 1996].

No practically usable components or applications have been reported so far, although there is evidence of the use of possible future emitters of submillimeter radiation are sufficient [Schlenga 1995].

Although such mesa structures in thin cuprate layers are nonlinear I-U characteristics with numerous branches and switching between branches, hysteresis, as well as through Shapiro steps caused by microwave radiation, and even show microwave emission ten, the I-V characteristics and other properties of these mesa structures remain far behind ideal phase-synchronized batch sequences of Josephson contacts. Ins In particular, the emitted microwave line is only in the pW area, which is synonymous there with is that such stacking sequences are of no interest for practical applications.

There are three main reasons for this: (a) The Josephson contacts were not uniform in their values of I _c and R _N ; (b) The areas of the Josephson contacts were generally too large, typically about 30 × 3 µm² ≈ 10 ^-5 ; and (c). The components were not optimized with regard to their impedance matching to the free space. As a result of (a) and (b), phase synchronization is not perfect, unstable and random. Natural Josephson contacts in cuprate superconductors have relatively large critical current densities in the c-axis direction (10⁴-10⁶ A / cm²), so that such contacts have by far too large critical currents, 100 mA and above. Large Josephson contacts show many complex excitation modes, super currents flowing in both directions, Fluxon movement, etc. Although these problems are known, the scientists in the groups mentioned were unable to significantly increase the contact area due to excessive electrical resistances to reduce. Excessive electrical contact resistances cause the stack sequences to heat up significantly, which makes phase synchronization more difficult and the critical currents become inhomogeneous. In extreme cases, the Josephson contacts burn out.

3. Thin layers of high-temperature superconductors with a-axis orientation

A current development that is suitable for practical implementation in a new component is a technology for growing thin high-temperature cuprate superconductors, in which the CuO₂ levels are not arranged parallel to the substrate. These can actually consistently perpendicular to the substrate, and we want such a thin layer as "a- Axes "are oriented layers. This name is derived from the specific case that the single crystal layer is grown in such an epitaxial relationship with the substrate is that the a-axis is perpendicular to the surface of the substrate. We want this case do not differ from that that the b axis (in orthorhombic crystals with a ≠ b) is perpendicular to the substrate and also if the a-axis is arbitrarily inclined, so as long as only the c-axis runs parallel to the substrate surface, d. H. so as long as the CuO₂- Layers are perpendicular to the substrate surface.

Using a similar commonly accepted (though inaccurate) termino We will use the entire group of epitaxial orientations in which the CuO₂ planes are identical assume an angle to the substrate other than 90 ° or 0 °, in the term "inclined summarize a-axis "layer. A finer distinction and a more precise description Exercise is quite possible, for example using the concept of Miller's In indices of substrate surface (on which the thin layer is to be deposited) and thin layer Surface. However, this is not necessary for the discussion here.

By far the largest group of high quality thin films produced by High-temperature superconductors have the c-axis orientation (i.e. they are with them CuO₂ levels parallel to the substrate surface). The reason for this is the strong anisotropy phy  physical properties of layered cuprates, including growth rates in different crystallographic directions. Nevertheless, there is considerable effort the ge on the cultivation of "a-axis" oriented and "inclined a-axis" layers are directed, whereby the main motive stems from the much greater coherence length in Use the direction of the a-axis for three-layer Josephson contacts. The following will be some of the most successful breeding experiments of this kind are reported.

There have already been thin Bi₂Sr₂CaCu₂O₈ layers with 45 ° inclined CuO₂ levels Magnetron sputter technology deposited on SrTiO₃ substrates with an angle of 5 ° deviate from the (110) direction. MgO substrates also came with the same characteristics stik and additional buffer layers for use [Tanimura u. a. 1993]. The desired inclined a-axis orientation was determined using RHEED (diffraction of high-energy electro reflection) and TEM (transmission electron microscopy) in cross-section education as well as by measuring transport properties.

Essentially completely a-axis-oriented YBa₂Cu₃O Schichten layers were obtained on (100) -oriented LaSrCaO₄ substrates with a PrBa₂Cu₃O₇ buffer layer by means of single target sputtering [Suzuki et al. 1993]. Similar oriented films made of Nd _{1 + x} Ba _2-x Cu₃O _7-d were applied directly to (100) -oriented SrTiO₃ substrates using laser vapor deposition, whereby special care was taken to set the growth parameters correctly [Badaye et al. 1995 ].

These examples demonstrate that it is technically possible to "a-axis" oriented and "inclined to produce a-axis "layers of high quality from cuprate superconductors. An embodiment the invention is based on the above facts.

Aim of the invention

It is an object of the present invention to provide means which achieve the effects discussed above allow parts and difficulties to bypass known solutions and a component develop that avoids the disadvantages discussed.

It is another object of the present invention to provide a novel radiation emitter and detector for electromagnetic radiation in the submillimeter wave range make a large number of stacking sequences of Josephson contacts in one area chigen (2-dimensional) arrangement applies, so that the generation of microwave  or the sensitivity to detect microwave radiation is considerably increased becomes.

It is another object of the present invention to provide a novel radiation emitter in the submillimeter wave region with such an impedance that Impe danz adaption to the radiation receiving space is achieved, whereby the ab radiated power takes maximum values.

It is another object of the present invention to provide a novel radiation emitter to make available in the submillimeter wave area, which has very low emission lines width (less than a millionth of the radiation frequency) within its up to a few THz-reaching submillimeter wave band.

It is another object of the present invention to provide a novel radiation emitter and detector in the submillimeter wave area with an electronic controller for continuous ier change in the emission frequency or the detection frequency over a wide to create spectral range.

It is another object of the present invention to provide a novel radiation emitter and detector in the submillimeter wave area to create its emitted microwaves radiation can be electronically modulated or switched on and off, which is also true corresponds to fast electronic switches for superconducting electronics.

It is another object of the present invention to provide a novel radiation emitter and detector in the submillimeter wave area, which have the functions emission and Detection independently, in channels with different frequencies or electrically variable and controlled frequencies.

It is another object of the present invention to provide a novel radiation emitter and detector in the submillimeter wave area, which inverses the emission and detection mode allowed with external electronic means.

It is another object of the present invention to provide a novel radiation emitter and detector in the submillimeter wave area, which is intended for inclusion in Supra conductor / semiconductor hybrid circuits is suitable.  

State the nature of the invention

These objectives are achieved with the present invention of a two-dimensional network of superconducting micro bridges with stacking sequences of (native) Josephson contacts achieved, parts of the network grouped in a predetermined manner and (serial, parallel) interconnected, as well as via contacts with an external electronic control unit can be operated. In particular, the task is performed in the characterizing part of claim 1 and claim 13 specified features. Advantageous design gene and applications of the invention, claims 2 to 12 and 14 to 19.

Embodiment

The invention will be explained in more detail with the aid of exemplary embodiments. In the belonging The drawing shows:

Fig. 1 A unit cell of the crystal of a high-temperature superconductor which contains a CuO₂- plane,

Fig. 2 An "a-axis" oriented layer of a high temperature superconductor. The CuO₂ levels are perpendicular to the substrate.

Fig. 3 A thin layer of a high-temperature superconductor with "inclined a-axis". The CuO₂ levels are inclined at an angle Θ to the substrate.

Fig. 4.A A microbridge made from a high temperature superconductor with "a-axis" oriented growth direction.

Fig. 4.B An equivalent circuit for the device shown in Fig. 4.A: A simple, linear (serial) arrangement of Josephson junctions.

Fig. 5 A microbridge made from a high temperature superconductor with "inclined a-axis" growth direction. The equivalent circuit corresponds to that from Fig. 4.B.

Figure 6.A A parallel arrangement of micro-bridges, each including stacking sequences natural Jo sephson contacts made of an epitaxial "a-axis." - or "a- inclined axes" layer by chemical or ion etching of the trenches, which electrically isolate the micro bridges from one another.

Fig. 6.B An equivalent circuit for the component of Fig. 6.A Parallel connection of two linear stacking sequences of Josephson contacts.

Fig. 7.A A parallel connection of three identical groups (clusters), each with 10 Mi krobrückeit

Fig. 7.B An equivalent circuit for the device of Fig. 7.A

Figure 8 shows an arrangement included. More groups of micro-bridges, each stacking sequences of Josephson junctions, along a strip line. The distance between the individual segments of combined emitters corresponds to the wavelength λ of the electromagnetic radiation within the structure.

Fig. 9 shows a basic structure of the device according to the invention, which contains in a thin superconducting epitaxial layer a two-dimensional lateral arrangement of parallel-connected micro bridges (webs) (a) which are separated from one another by trenches (b) extending to the substrate and thus the Direct current flow through the micro bridges or groups of micro bridges. Each microbridge includes a stacking sequence of Josephson junctions, the arrangement in Fig. 2 and will be explained FIG. 3. The electrical connections (c) and (d) allow a connection to external control electronics (e).

Based on an embodiment of the component according to the invention, the invention is described below dung described in more detail, the basic structural unit, d. H. the linear stacking order of Josephson contacts in a high quality thin film of a high temperature superconductor is formed, the thin layer in a special epitaxial relationship with the Substrate is located.

1. Micro bridges in "a-axis oriented" and "inclined a-axis oriented" thin layers

For explanation, it must be noted with reference to FIG. 1 that the crystallographic unit cell of all known cuprate superconductors represents an extended parallelepiped with the side lengths a≈b≈3.8Å and c <a, b. For example, c≈6.5 Å in La _1.85 Sr _0.15 CuO₄, c≈11.7 Å in YBa₂Cu₃O₇, c≈15.4 Å in Bi₂Sr₂CaCu₂O₈, etc. [More specifically, the c-axis periodicity in La _1.85 Sr _0.15 CuO₄ and Bi₂Sr₂CaCu₂O₈ is twice as much large because of the presence of a sliding plane, although this fact is of little importance for the present discussion, just like that of a slight orthorhombic distortion with a ≠ b in YBa₂Cu₃O₇].

All of the high-temperature superconductors that have been considered so far were down with the c-axis have been grown right to the substrate (they are commonly referred to as "c-axes  oriented HTS layers "). This is the most common epitaxial relationship usually sets when high temperature superconductors on lattice matched substrates, such as e.g. B. (100) cut SrTiO₃, LaAlO₃, etc. are grown. In this case reasonable fen all CuO₂ levels parallel to the large area of the substrate.

However, if specially cut substrates are used, it is possible to achieve other epitaxial arrangements in which the CuO₂ planes are perpendicular to the large area of the substrate. In such a case, the a-axis of the cuprate superconductor (or completely equivalent to the b-axis) is arranged perpendicular to the large area of the substrate, as shown in FIG. 2. Such high-temperature superconductor thin layers are referred to for short as "a-axis oriented HTS layers".

One way to achieve this orientation is as follows (here La _1.85 Sr _0.15 CuO₄ is chosen as an example, where a ≈ 3.8 Å and c ≈ 6.5 Å):
The substrate and its "cut" (ie the direction of polishing or the orientation of the surface plane of the substrate) are selected so that the periodicity of the surface plane is 3.8 Å × 6.5 Å. Then there is the possibility that the layer grows with its a-axis perpendicular to the substrate. Another aspect is the number of "favorable contacts" between cations and anions of the layer and the substrate, which influence the energetic relationships in the interface. Of course, different substrates or at least substrate sections are required for each different high-temperature superconductor.

For the present purpose, it is sufficient to state that such substrates and substrate sections for several cuprate high-temperature superconductors have already been discovered, and complete epitaxial a-axis growth has actually been demonstrated, namely for La _1.85 Sr _0.15 CuO₄ and YBa₂Cu₃O₇ [Suzuki et al. 1993, Badaye et al. 1995].

Another embodiment of the present invention is provided by a possible other epi taxial relationship shown in FIG. 3. For this all CuO₂ levels of the Cu prat films are grown at an angle to the substrate which is different from 0 ° and 90 °. Such an angle depends on the substrate cut and can typically be 1 ° to 10 °, although larger or smaller angles are also possible. For the purpose of the present discussion, we assume that the angle is approximately 5 °. This requires the existence of steps in the substrate that are 3.8 Å high and have a distance of approximately d = 3.8 × tan 85 ° Å. Epitaxial growth of such inclined a-axis oriented HTS layers has also been carried out [Tanimura et al. 1993, Kataoka et al. 1993].

As already stated, high-temperature cuprate superconductors, such as e.g. B. Bi₂Sr₂CaCu₂O₈, Tl₂Ba₂Ca₂Cu₃O₁₀, HgBa₂CaCu₂O _6.2 , La _1.85 , Sr _0.15 CuO₄, etc. natural superconducting supergit from SIS. . . Or SINIS type is. If one uses one of said orientations of the thin film, then a simple micro-bridges can be produced in Fig. 4.A and Fig. 5 type shown, and thus that is, a stacking order of Josephson junctions. Such stacking sequences in micro bridges represent the basic structure for the component according to the invention. However, to set an optimal performance of the component, several such micro bridges must be interconnected in a superconducting network, as described in the text below.

2. Interconnection of micro bridges: circuit optimization

A single stacking sequence of Josephson contacts is considered first. The maximum allowed critical current through a single contact is about I _c = 1 mA. Larger values lead to very disturbed current-voltage characteristics due to the special flow conditions that arise. The Josephson contact behaves like a "long contact" with currents flowing back and forth in different places. Using the example of I _c R = 10 mV and I _c = 1 mA, the typical lower limit for the electrical resistance is R = 10 Ω. Impedance matching of the component to the vacuum requires a total resistance R _ToT = NR≅300 Ω. With R = 10 Ω one obtains N = 30 as the value for the number of Josephson contacts to be inserted in the stacking sequence of a mesa structure. If I _{c were} less than 1 mA, an even larger value for R and thus a smaller value for N would have occurred. Accordingly, this value N would be a lower limit for a single mesa structure with impedance matching to the vacuum.

The geometric dimensions of the contact are also already defined with the values mentioned. If j _c ^⟂ = 10⁴A / cm², I _c = 1 mA is the area of the Josephson contact

A = I _c / j _c = 10 ^-3 A / 10⁴A / m² = 10 ^-7 cm² = 10 µm².

If you were to use square Josephson contacts, they should have an edge length of 3 µm.  

This value of j _c ^⟂ = 10⁴A / cm² relates to natural Josephson contacts in very anisotropic high-temperature superconductor materials, such as. B. Bi₂Sr₂CaCu₂O₈. With less anisotropic high-temperature superconductor compounds, j _c ^⟂ can ^assume higher values (up to 10⁶A / cm² in YBa₂Cu₃O₇), which could result in even smaller contact cross-sectional areas.

However, the output power of such a simple stacking sequence is quite modest:

P _MAX ^out = 1 / 8Ni ²R _c = (1/8) × 30 × 10⁶ × 10A²Ω = 5 × 10 ^-4 mW.

Higher radiated power can only be expected if more complex systems with series and parallel groups of mesa structures are used. Essentially, the idea is to increase the total current through the component without exceeding I _c ^MAX = 1 mA for a single contact by connecting several stack sequences in parallel ( FIG. 6). A reduction in overall resistance can be easily compensated for by increasing the number of Josephson contacts in each stacking sequence.

For a parallel connection of M stack sequences, in which each contains N Josephson contacts in series connection, the total resistance is R _TOT = (N / M) R; For R _TOT = 300 Ω and R = 10 Ω you get N / M = 30 or N = 30M.

The total current becomes 1 = MI _c , so the total power

P _TOT ^out = (1/8) × 10 ^-6 × 300 × M²A²Ω≈0.4 × M² mW arises.

It can be seen that the total radiated power is proportional to M² and it could seem as if the performance could be increased arbitrarily. However, restrictions are imposed by the maximum possible size of the network and of course the still acceptable dimensions of the imposed on individual batch sequences.

In particular for a grouped arrangement of Josephson contacts, phase synchronism can be used tion can only be achieved over a distance of approximately λ / 4, where λ is the waves length of the emitted electromagnetic radiation. Would be the desired one emitted Frequency ν = 3 THz, then λ = 100 µm; for ν = 1 THz, one obtains λ = 300 µm.

To optimize component parameters (e.g. to match impedance to the vacuum and a reduction in the internal reflection of electromagnetic radiation at the kri  stall / vacuum interface), it is desirable to have micro bridges in every will ten way to combine electrically. However, the electrical connections should be off Superconductors exist and do not lead to the formation of "weak links" gen), which then act as Josephson contacts in series with the stacking sequence of Jo sephson contacts work in the mesa structure. In other words, they are High-temperature superconductor electrodes of the same type of superconducting material conducive.

Both series and parallel connection of mesa structures is of practical relevance Cases required. The following section explains a method with which the Seri Switching of simple micro bridges or of groups (clusters) of micro bridges achieved that are already in parallel. Furthermore, realistic estimates to the maximum possible number of phase-locked Josephson contacts in such stacking sequences and networks.

3rd example

An analysis of the realistic limits of linear stack sequences and their parallel and serial interconnection takes place in the following way: Assuming I _c R ≈ 25 mV and I _c ^MAX = 1 mA as given, then R = 25 Ω is the resistance of a Josephson contact . N is therefore limited to values of 10-15 if impedance matching to the vacuum is sought with R _TOT = NR ≅ 300 Ω. This is obviously not very favorable for a pure a-axis oriented HTS layer ( Fig. 2, Fig. 4.A), since in such a case the length of the microbridges (in a-axis layers the term " Microbridge "for the previously designated" mesa structure "in c-axis layers). Would be less than 15 × 20 Å = 300 Å, which is too small for common lithographic processes.

The situation improves for inclined a-axis oriented HTS layers ( FIG. 3, FIG. 5), where the angle of inclination is relatively small (approximately 5 °). Here one gets a total length L of the microbridge of d / tanθ for N = 15 and c = 15 Å, where d is the layer thickness. For d = 1000 Å and θ = 5 °, L ≈ 10,000 Å = 1 µm is obtained, which is a reasonable value for currently conventional lithography processes. And this length can be increased further with increasing layer thicknesses.

However, we also have to examine the required lateral dimensions. Restrictions come from the current density j _c . A typical current density for c-axis oriented layers is j _c = 10⁴ A / cm². With a desired critical current of I _c = 1 mA, the aforementioned contact area A = I _c / j _c = 10 ^-3 / 10⁴ cm² = 10 µm² is obtained. However, it should be noted that the effective width is now equal to the layer thickness divided by the tangent of the angle of inclination, ie d / tanθ. With d = 1,000 Å and θ = 5 °, d / tanθ ≈ 1 µm is obtained. This means that the width of the microbridges should be 1 µm, which is certainly sufficient.

However, such a microbridge, which contains a stacking sequence of Josephson contacts, will only produce a relatively small power of emitted radiation, namely P _out = (1/8) I _c ²R _TOT = (1/8) (10 ^-3 ) ²300 W ≈ 0.4 mW.

In order to increase this power, a parallel connection of the M of such micro bridges is again required, since in this case P _out = (1/8) (MI _c ) 2 R _TOT = M 2 × 0.4 mW, provided that the number N is the natural Josephson contacts in each microbridge is set so that the impedance matching to the vacuum remains fulfilled. For R _TOT = RN / M ≅ 300 and R = 15-30 Ω we get N / M ≈ 10-20, or typically N = 15 M.

This provides the starting point to create a very simple component, which is shown in Fig. 8 ge and consists of a parallel arrangement of M identical micro bridges. These are formed by simply creating a series of equidistant trenches down to the substrate by chemical or ion etching.

For N = 15 M one obtains micro-bridges which are elongated by M times. A pure a-axis oriented HTS film leads to a microbridge length of 15 × 300 Å = 4,500 Å or 0.45 µm, which is still a very small value and technologically difficult to manufacture. On the other hand the length of the microbridges in inclined a-axis oriented HTS layers for d = 1,000 Å, N = 300 and θ = 5 ° 0 are approximately M µm.

We had already concluded above that for j _c = 10⁴A / cm² and I _c = 1 mA the contact area A = 10 µm² and the microbridge width was about 10 µm (our standard choice d = 1,000 Å and θ = 5 °). M is thus severely restricted, since the entire width of the component can be 25 µm for an emission frequency of ν = 3 THz (λ = 100 µm) and 75 µm for ν = 1 THz (λ = 300 µm). Consequently, for M = 3-7 one obtains, which would correspond to a quite considerable output power of P _out = 1-20 mW.

Further improvements in emission performance could be obtained by choosing less anisotropic high temperature superconductors with a higher critical current density for the c-axis direction. For example, for YBa₂Cu₃O₇, j _c ≈ 10⁶A / cm² in the c-axis direction. In such a case, much smaller contact areas would have to be provided, say 0.1-1 µm², which results in much narrower trenches, a larger number of micro bridges and thus a higher emission power. A practical limit for the width of micro bridges could be 1 µm with very narrow trenches in between. At 3 THz this would lead to values for M of around 25 and the considerable output power of around P ≈ 240 mW.

Another possibility for improvement is the very short length of micro bridges (1 µm) for a-axis oriented HTS layers. One could design a component accordingly in FIG. 7. Using the available component length of 25 μm (for 3 THz), three parallel segments (strips) of 10 parallel micro bridges each were interconnected in this FIG. 7. The superconducting electrical connection between the 3 segments is parallel. One therefore has a total value M of micro bridges, which is given by the number of micro bridges per segment multiplied by the number of segments. Realistic values can go up to 25 micro bridges per segment and up to 10 segments, which would correspond to an M = 250. Under ideal conditions, this would result in an immense emission output of

P = (250) ² 0.4 mW = 25 W!

Although the previous value for the output power may be due to a number of others limiting factors represents an overestimation, the potential that could be stuck in such a component. The simplicity of the manufacture is also remarkable position that a single lithographic process cut without any special alignment required.

Because of the greater width of the material strips perpendicular to the micro bridges, a higher current can flow here than in a single segment of micro bridges. The current flows in these strips along the CuO₂ levels, which corresponds to the "simple" (ie more conductive) direction; j _c can be 10-1000 times higher in this direction (parallel to the a-axis) than in the "hard" (ie c-axis) direction along which the current flows through the microbridges. In this way, weak links between these contact tracks and the micro bridges are avoided.

Finally, it should be noted that the electrical connection to the external electronics via dic ke (0.5-1 µm) and large-area (1 min²) normally conductive metal contacts (gold, silver, etc.) takes place, which were deposited on the superconducting layer. A heating of the component Resistive behavior of these contacts is considerably minimized. Anbon the external connections is easy.

The embodiment of the component according to the invention discussed here has the additional advantage that the Josephson contacts have a very small cross-sectional area and thus a very small electrical capacitance C. A pure a-axis-oriented YBa₂Cu₃O₇ film with a critical current in the c-direction of j _c = 10⁶A / cm² has a cross-sectional area of A = 10 ^-9 cm² = 0.1 µm² at an optimal current. This corresponds exactly to a layer thickness of 1,000 Å and the microbridge width of 1 µm, as already described above.

A capacitance of C ≈ 5 × 10 ^-15 F in such small contacts can be expected from the known material properties for cuprates and for insulating layers. [The specific capacitance of Nb / NbO _x / Nb-Josephson contacts is 50 ff / µm²]. This value can be made 5 times lower if only 200 Å were used instead of the layer thickness of 1 µm. A disadvantage would then be the reduction in the maximum critical current density and a reduction in the radiated power.

On the other hand, such contacts have the advantage of critical damping. The McCumber parameter is β = 2π I _c R²C / Φ₀ 1 where Φ₀ = h / 2e = 2 × 10 ^-15 Vs is the flux quantum, h is the Planck quantum of action and e is the electron charge. It is known that β should be small (1) when optimum radiated power and tunability of the frequency is required.

The above consideration is in the case of a single group (cluster) by Josephson- Limited contacts. In fact, it is possible to know the maximum radiated power ter by choosing a distributed arrangement of such groups. You can group  pen at an equidistant distance at distances of λ between them (say λ = 300 mm for arrange ν = 1 THz).

By covering the entire structure (layer) with an insulating material (e.g. SiO₂, MgO, CeO₂, etc.) and applying a metal layer (e.g. gold or silver), a transmission line can be created through which the electromagnetic radiation is transmitted (see Fig. 8).

In such structures, phase synchronization can be achieved over fairly long distances the, whereby the radiated power can be considerable. The disadvantage of such lei However, it is evident that the operating frequency is quite high due to the group spacing λ is set. Higher performance can only be tuned at the price of a reduced tuning achieve speed.

Other aspects of optimization include the integration of an antenna, being more convenient as the existing superconducting layer can be used for this.

4. Microwave radiation source

After a thin-film structure with an arrangement of micro bridges (in general in parallel and series connection) and with other standard components of the Microwave technology (antenna, propagation line) connected, you can connect this to a usual control electronics with adjustable current source (up to 100 mA), control devices, Connect displays for frequency and power, etc. This then represents a complete source for narrowband electromagnetic radiation that is in a wide frequency range and can be tuned up to high frequencies of 5-10 THz.

This component of the present invention is called the basic variant of the tunnel tron. It is shown in FIG. 9.

5. Areas of application

The present invention can be used in almost all areas of application in which Mil limeter and submillimeter-wave radiation is emitted or detected. Also open up completely new fields of application, some examples of which are given below:

Some lasers and reverse wave tubes (carcinotrons) work in the submillimeter wave Area, but they are voluminous radiation sources with high power consumption. Solid  per oscillators, such as B. GUNN or IMPATT diodes are on the millimeter waves richly limited.

Josephson contacts, which are adapted to the external resistance without reflection and to form a network are summarized, can be controlled by a voltage and a wide Cover the frequency range into the terahertz region.

Quantum detection of electromagnetic radiation - a widely used concept in visible and infrared spectral range - was previously in the microwave and millimeter wave Area only possible in the narrow range of the spectrum, which is about the resonance frequency grouped by MASER amplifiers. The standard methods of proof in this Fre quenz range use nonlinear electrical resistors, such as. B. Schottky diodes as classic rectifiers and superimposed receivers. Their working principle is based on the Conversion of received power between different frequency ranges, instead the conversion of photons into electrical charge carriers, the functional principle of Quan detectors.

The abrupt non-linearity in the IU characteristic of SIS tunnel barriers for single-particle tunnels represents a useful property for resistive mixing. Superimposition receivers with Josephson contacts as such mixing stages have a sensitivity that approaches the quantum limit at frequencies up to several GHz. The function of a heterodyne receiver is to mix a weak useful signal of frequency νs with the frequency of a local oscillator ν _Lo , with an intermediate frequency ν _if = | ν _s -ν _Lo | is created and processed electronically. A photon current with an arrival rate of one photon per nanosecond is a typical value of the detection sensitivity of such a receiver. Such light outputs are typical of radio astronomy, which examines interstellar matter in the millimeter and submillimeter wave range to explain the structure of the universe. Between 100 and 1000 µm wavelength, numerous lines of molecular rotation and vibration spectra appear, which can in principle be used to elucidate numerous physical properties in the universe. Research into the 115 GHz rotational vibration of interstellar carbon monoxide (CO) with a receiver for λ = 2.6 mm also indicates the performance of such microwave spectrometers in other practical applications.

Spectroscopy in its general meaning implies the study of absorption and emission of electromagnetic radiation from substances from an external source be excited or emit themselves. The Tunneltron component has a special characteristic shaft on that it is used as a source of excitation for the medium to be spectroscopic and as a beam tion receiver of the radiation emerging from the medium can be used. Around To give an example: The tunnel tron enables the investigation of organic and anorga compounds in vapors, liquids and solids with regard to their chemical Composition and geometric and energetic structure as well as change effects processes; All of this is possible in function of various external parameters and in time-resolved regime. There are numerous cases where spectroscopic measurements are of interest se are, e.g. B. in the detection of hidden substances of organic molecules and chemical Compounds (biological substances, drugs and plastic substances,...) After a suitable Excitation in the submillimeter wave area, or devices for quality control for specific chemical substances (e.g. water content or contamination in solid substances, fat layer dic on meat, etc.).

The tunnel tron as a coherent and tunable radiation source has properties of the Wave field on the z. B. come to bear in interferometry and holography. Ho In principle, photography is a method of creating a unique photographic image of a coherently exposed object in which an undisturbed (direct) beam and that from the Reflected beam originating in an object is brought into interference in a detection system the. The reconstruction of this interference image provides a three-dimensional image of the object.

The possibility of frequency tuning of the coherent tunnel tron, as well as the Possibility of spread through media that is opaque to the human eye the submillimeter radiation of the tunnel tron is transparent, however, opens up a wide range re applications.

Communication and data transmission is another area of application of the here struck component, whereby a frequency band is accessible, which far above that gen, which is currently managed by the ITU (Intern. Telecom. Union). The new Fre quenzgebiet enables a considerable increase in the number of usable channels for both Sa tellite as well as earthbound communication. [Looking at a channel width of 20 MHz, the frequency range up to 5 THz provides about 250,000 channels. This can be compared  with for example 40 channels for satellite communication in the frequency band of 11.7-12.5 GHz, used by the ITU for Region 1, Africa, Europe and the former Soviet Union together, was made available.] Take a 4 kHz channel width for speech communication, you would generate 2 billion voice channels on such a carrier could be transported.

Transmission to a high quality standard requires digital systems with pulse modulation (PCM - pulse code modulation) because frequency modulation far exceeds the noise power Measure of 3 pW / km (about 52 dB) would drive that by the CCIR as the upper noise limit was recommended.

Digital systems allow signals to be regenerated in intermediate stations, which means that Accumulation of errors is avoided.

It is itself grateful that a completely new wireless terrestrial microwave communication ons network, a broadband wireless link to the Internet and a broadband link to Satellite communication with all the advantages of the high number of telephony channels mentioned and the added benefit of reduced radiation exposure due to the reduced Absorption depth of human skin for submillimeter radiation is created.

Imaging multifrequency microwave radiometer (MIMR) for the exploration of near-earth Territory from satellites or other flying objects already has a high potential for use. Future remote sensing satellites will carry hyperspectral equipment, which are suitable for imaging exploration in a large number of frequency channels place of the seven channels that are currently used with multispectral instruments men.

Imaging all-round observation at close range is important for safety and maneuverability of helicopters and the approach of aircraft.

The same applies to road transport and even for robot applications adverse environmental conditions. The basic requirement is always a performance powerful, tunable, possibly monochromatic and coherent submillimeter wave Emitter (and corresponding detector), such as that with the component of the present Er is made available.  

As an "on-chip" integrated or separate emitter and detector component, it is for radar suitable measurements in the most general sense, d. that is, for location, navigation and early libraries systems. The growing need for compact, low-power radar assemblies the component according to the invention is suitable for civil and military use. Alarming or collision warning systems can be improved.

The tunnel tron can be an active sensor for SAR (synthetic aperture radar) application. A SAR system sends microwave radiation to the object (the earth when research sa tellites are the starting point), and receives the returning radiation. The fine tuning of the emitter allows the frequency to be positioned within the absorption band Minima of the atmosphere too. The possibility of the direction of the wave field of the tunnel tron Changing electronically proves to be a further advantage for scanning the observer area.

Biological and medical applications (tomography, imaging thermography, etc.) are simply too large in number from the properties of the component according to the invention Conclude.

Claims (19)

1. Emitter and / or detector component for submillimeter-wave radiation, consisting of a substrate, electrical leads (lines) to the electrodes, which are located on both sides of the component on a superconductor and are connected to an external power source, thereby characterized in that the component is created in a thin, single-crystalline and a-axis-oriented layer of a high-temperature superconductor (HTS) by forming micro bridges (a) and insulating trenches (b) so that the micro bridges (a) on both sides of superconducting Contacts (c, d) of the high-temperature superconductor forming the layer go out and can be combined in any manner in series and / or parallel connection and can be controlled externally, individually or in groups, by appropriate routing of the trenches (b). 2. Component according to claim 1, characterized in that the HTS layer epitaxially with the c-axis has grown parallel to the substrate surface, and the a-axis at an angle of 1 ° to 89 ° relative to the normal direction on the substrate. 3. Component according to claim 1 or 2, characterized by an arrangement (array) of min at least two to a few hundred micro bridges connected in parallel, each with two to one around a hundred vertically stacked Josephson contacts per microbridge, with all Mi krobrücken have common superconducting electrical contacts. 4. The component according to claim 3, characterized in that the electrical contacts a Allow voltage measurement for detector operation. 5. The component according to claim 3, characterized in that a plurality of parallel assemblies through a suitable trench guide to form a segment, connected in series are. 6. The component according to one or more of the preceding claims, characterized by a distributed arrangement of complete groups of microbridges along a microwave Stripline to achieve a higher radiated power. 7. The component according to one or more of the preceding claims, characterized by a use for communication and data transmission. 8. The component according to one or more of the preceding claims, characterized by use in radar systems   9. The component according to one or more of the preceding claims, characterized for Applications as multi-channel emitter and detector in satellite-based SAR (Synthetic Aperture Radar). 10. The component according to one or more of the preceding claims, characterized by use in imaging multi-channel microwave radiometers. 11. The component according to one or more of the preceding claims, characterized by any kind of spectroscopic application. 12. The component according to one or more of the preceding claims, characterized by use as a local oscillator in heterodyne receivers, for example for astrono mixed applications used. 13. Method for producing an emitter and / or detector component for submillimeters Wave radiation according to one or more of the preceding claims, characterized records that the high temperature superconductor (HTS) layer by molecular beam epitaxy (MBE) and analog processes, such as. B. atomic layer epitaxy (ALE), fine focus Ion beam epitaxy (FIBE) is generated. 14. The method according to claim 13, characterized in that the molecular beam epitaxy (MBE) device spectroscopic means for the observation of the interface and for control of growth of the HTS layer. 15. The method according to claim 13 and / or 14, characterized by the following process steps:

• (a) Epitaxial growth of a HTS layer with a sloping a-axis, 100-500 nm thick, on a suitably cut wafer (substrate) with or without one or more buffer layers.
• (b) Lithography for making the trenches.
• (c) Ion etching down to the substrate surface.
• (d) removal of the photoresist.
• (e) attaching metal contacts to the HTS electrodes.

16. The method according to one or more of the preceding claims, characterized ge indicates that the HTS layer by chemical vapor deposition (CVD) and analog process is produced. 17. The method according to one or more of the preceding claims, characterized net that the HTS layer by laser pulse deposition (PLD) and analog methods ago  is provided, including the case that the PLD device via spectroscopic means for Interface monitoring and control of the growth process is equipped. 18. The method according to one or more of the preceding claims, characterized net that the HTS layer produced by sputter deposition (SD) and analog processes is included, the case that the SD device via spectroscopic means to the interface monitoring and control of the growth process. 19. The method according to one or more of the preceding claims, characterized in that that the device structure is created using micromasks.