|Publication number||USRE29578 E|
|Application number||US 05/795,189|
|Publication date||Mar 14, 1978|
|Filing date||May 9, 1977|
|Priority date||Jun 11, 1974|
|Publication number||05795189, 795189, US RE29578 E, US RE29578E, US-E-RE29578, USRE29578 E, USRE29578E|
|Original Assignee||Massachusetts Institute Of Technology|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (4), Non-Patent Citations (3), Referenced by (8), Classifications (14)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The invention herein described was made in the course of work performed under Contact No. F19628- 7-C-0150 with the Electronics Systems Division of the Department of the Air Force and under Contract No. N00014- 67-A-0204- 0014 with the Office of Naval Research, Department of the Navy.
This application is a continuation in part of my copending applications entitled Generating and Using Coherent Optical Radiation, Ser. No. 389,970 filed Aug. 20, 1973 and Solid State Optical Junction Devices and Arrays and Systems Incorporating Same, Ser. No. 389,783, filed Aug. 20, 1973, now U.S. Pat. No. 3,898,453, which in turn are continuations in part of my prior application Ser. No. 62,380 filed Aug. 10, 1970, now U.S. Pat. No. 3,755,678. The disclosures thereof are all thereby incorporated by reference.
This invention relates to devices capable of oscillation, radiation, amplification, frequency subdivision and other functions at the far infrared and adjacent regions of the frequency spectrum. In this invention quantum mechanical electron tunneling is employed in a metal-dielectric-metal deposit upon a solid substrate. The deposit comprises ultra thin dielectric films and microscopic contact areas, and is integrated with other components of the device, including inductance and feedback antenna arrangements.
As noted by L. F. Saki in "Long Journey into Tunnelling," Science, Mar. 22, 1974, p. 1149, et seq, the tunneling phenomenon in metal-dielectric-metal contracts per se was demonstrated in 1930, and since that time has been the subject of extensive research. Also, as noted in that article and articles by I Giaever in "Electron Tunneling and Superconductivity," Science, Mar. 29, 1974, p. 1253, et seq, and Davis and Hasack in "Double Barrier in Thin Film Triodes," Journal of Applied Physics, Apr. 1963 p. 864, et seq, such contacts have been known to produce a negative resistance characteristic, meaning amplifiers, oscillators, and other devices, and have been known also to be useful to form triodes, etc.
In prior work the dielectric barrier has been of 30 Angstrom or usually considerably greater thickness and typical contact areas have been of the order of one square millimeter. The relatively large capacitance of such contacts, the rapid increase in impedance when one decreases contact area for reducing capacitance, and the problems of non-uniformity in deposited layers have seemingly limited the practical speed of response of such tunneling contacts to frequencies much lower than those of concern here.
According to the present invention it has been discovered to be possible to obtain a true response over a broad range extending into the far infrared and infrared regions employing solid deposits having contact areas of the order of 1 square micron (1 million times smaller than the above example), with dielectric thicknesses of less than about 10 Angstrom (less than 5 atoms thickness). Significantly it is found that the ultra-thin dielectric layer between the metal deposits successfully resists breakdown under the applied voltage and current densities and has contact impedance at the most of a few hundred ohms (impedance so low that in a larger contact it might be considered a metallic short). (For actual tests see the paper authored by my colleagues and myself: Small, Elchinger, Javan, Sanchez, Bachner and Symthe, Applied Physics Letters Vol. 24, No. 6, Mar. 15, 1974, P. 275 et seq., the disclosure of which is hereby incorporated by reference.)
Within the parameters of construction described above, tunneling is obtained, with response at freqencies of the order of 1012 Hz and higher, with contact capacitance of the order of 10-12 farad or indeed 10-14 or lower. While a full explanation of these results is not yet available, it is probable that a number of compensating factors have, fortuitously, and in combination, allowed these results to be obtained. Thus, although prediction of non-uniformities in deposit thicknesses of about 10A has discouraged experiments, it has been now realized that the drastically reduced contact area of the present device reduces the probability of non-uniformity effects; and while reduction in contact area per se would greatly increase impedance, the reduction in barrier thickness more than compensates, while the enlarged current densities do not lead to failure perhaps because of increased cooling effects attributable to the increase of surface to volume ratio attendant with the reduction in size.
Further, according to the invention, devices including oscillators, radiation sources, amplifiers, multivibrators and frequency subdividers are constructed incorporating metal-dielectric-metal deposited contact regions within the parameters just noted. Preferably there are realized in the form of multibarrier structures, i.e. structures having two or more layers of dielectric separated by appropriate metal layers in a series arrangement or by operating with the metals of the junction in superconductive state.
In certain embodiments the contact exhibits a negative impedance characteristic, means are provided to bias the contact into the negative impedance region, and the contact is connected in parallel with an inductance to define an integrated tank circuit with resonant frequency in the desired region. According to one preferred embodiment the RC time constant of the circuit is as fast as the speed of response of the contact and oscillations take place in the form of CW. In another embodiment where the speed of response of the contact is not matched to the RC time constant of the circuit, a relaxation oscillator is achieved, its oscillation appearing in the form of pulses. In either embodiment avantageously a means to alter the bias over the negative impedance curve is provided to enable the frequency of oscillation to be pulled, thus to achieve a tunable oscillator at far infrared or adjacent frequencies.
In preferred embodiments an antenna structure is deposited on the substrate matched to the frequency of oscillation, for emitting radiation at that frequency. According to one preferred embodiment, a contact having a negative impedance characteristic is connected across a pair of parallel conductors deposited upon the substrate, of length related to the oscillation wave length, to provide the inductance of the tank circuit, the outer ends of these lines being connected to an antenna structure, preferably each line connected to one branch of a dipole antenna. According to another preferred embodiment, a loop is formed by multiple deposits which also, in a region of juncture, form the effective contact, extensions of the deposits forming an antenna.
In a device according to another preferred embodiment, a multiple barrier structure is provided in which an intermediate metal layer forms in effect a control grid or base, its voltage controlling the current flowing from one side to the other of the structure.
In another embodiment two or more such devices are provided with the collector of the first connected to the control or base of the second. By adding a deposited antenna optically coupling the collector of the second device with the control or base of the second, a feedback oscillator operable in the far infrared region or adjacent frequencies is achieved. In similar arrangements monostable and bistable multivibrators are achievable.
In still another embodiment a series of oscillator structures is provided of progressively lower resonating frequency, employed to subdivide the frequency e.g. of a laser, to a lower frequency, at which it may be radiated or conducted through a transmission line. In other embodiments selection of input frequencies relative to local quantized modes of the dielectric barrier are employed to enhance nonlinearities or to introduce nonlinear reactive components.
These and numerous other objects and features of the invention will be understood from the following detailed description of certain embodiments taken in conjunction with the drawings wherein:
FIG. 1 is a plan view illustrating initial steps in fabricating a device according to the invention;
FIG. 2 shows a diode device according to the invention;
FIG. 3 shows a multibarrier diode according to the invention, which with addition of dotted portions becomes a triode device;
FIG. 4 is a cross-sectional view taken on line 4--4 of FIG. 3; and FIG. 4a is a diagram of the resulting multibarrier junction;
FIG. 5 is a current-voltage curve (I vs V) illustrating a negative impedance region of active devices employed in certain embodiments;
FIG. 6 is a plan view of a negative impedance oscillator incorporating an active device according to the invention; and FIG. 6a illustrates details of fabrication thereof;
FIGS. 7 and 8 are plan views of alternative embodiments to that of FIG. 6;
FIG. 9 is a plan view of an oscillator employing triodes and antennas providing feedback in accordance with the invention.
FIG. 10 is a diagrammatic view of a frequency divider network employing a series of negative impedance oscillators, responsive to input from a laser;
FIG. 11 is a diagram illustrating the local quantized modes of the dielectric barrier of devices according to the invention in relation to a far infrared or infrared input; and
FIGS. 12a, b, c and d illustrate equivalent circuits achievable from appropriate selection of the variables of the device of FIG. 11.
FIG. 13 is a diagram of a parametric frequency down conversion unit.
FIG. 13a is an illustration of the invention employing a fabre-Perot resonator.
In the presently preferred form a multibarrier device is manufactured in accordance with generally known microelectronic methods, and presented with other structure in a predetermined array, such as described below. For devices operable in the far infrared and infrared region, the techniques of Photolithography e.g. as described in Smith, Bachner and Efremow, Journal of Electrochemical Society; Electrochemical Technology, May 1971, Vol. 118, No. 5 are applicable, in which deposited lines to an accuracy of less than 1 micron are realizable. Alternatively, and of particular utility for shorter wave lengths, one may employ the methods of electron beam microfabrication such as are described in "Microcircuits by Electron Beam," Broers and Hatzaki; Scientific American, November, 1972.
The ultrathin dielectric layer employed in the device according to the invention may be obtained by choosing metal which has an inherently thin oxide layer (e.g. nickel) or by employing metals usually having thicker oxide layers, limiting the amount of oxygen to which the metal is exposed and then by applying additional layers of metal and dielectric to seal the contact area from further exposure to oxygen. Other dielectric substances may be employed too.
Referring to FIGS. 1 and 4 the first step in forming a device according to the invention is to deposit upon a non-conductive substrate 10 (preferably transparent to the radiation of interest, e.g. silicon, aluminum oxide or sapphire for infrared radiation) a conductive line structure 12 including leg 14 of an appropriate first metal, e.g. copper, nickel or aluminum. The leg 14 has a width d of the order of 1 micron and a controlled thickness of less than 1 micron. The metal of this line structure or at least the tip of leg 14 for the contact region is then allowed to oxidize (or is coated) to form what will become a potential barrier, a dielectric layer 16 of a few Angstrom (less than about 1 nm) thickness. Referring to FIG. 2 in the next step a second conductive line structure 18 with leg 20 similar to leg 14 but at right angles to it, of a second metal which may be the same or different from the first is deposited on the substrate, with width d and a controlled thickness less than 1 micron and slightly overlapping in region E the top of leg 14, a distance kept small, preferably less than about 1 micron. The metal of this second leg 20 is then provided with a dielectric layer 22 as above. Then, referring to FIG. 3, a third conductive line structure 24 of a third metal is deposited with leg 26 similar to leg 14, aligned with it and pointed in the opposite direction. It slightly overlaps the top of leg 20 in region E, a distance t, also kept small, preferably less than 1 micron.
The particular metals and oxides or other dielectric are chosen with their work functions and similar characteristics in mind, as well as the mode and conditions of proposed operation. Thus where it is desired to have a nickel and chromium sandwich, with oxide dielectric, it is preferable to deposit the nickel first and allow it to oxidize, as it oxidizes more gradually than chromium, offering better opportunity to control the oxide to obtain the ultra thin oxide layer. Similarly, where the device is intended to provide a negative impedance phenomenon in accordance with super lattice theory, not only should the various thicknesses, including the intermediate metal thickness, be properly designed and controlled to ensure conservation of energy and momentum, but also the substances should be selected with a view to proper reflectivities, as explained in the literature. Where it is desired to have both or only one side of the device superconductive appropriate choice of metals is again dictated, to employ metals which are superconductive at the chosen temperature of operation.
Under certain circumstances even a diode formed in accordance with FIG. 2 will produce an I-V characteristic having a region of negative slope, i.e. negative resistance, which can lead to negative resistance oscillators. One mechanism through which this can occur is by operating at low temperature with both metals superconducting, see the tunneling articles cited above. Another mechanism for achieving negative impedance is by utilization of multi-barrier structures, i.e. structures in which there are two or more dielectric layers separated by a metal layer of predetermined thickness. FIG. 4 can be taken as a representation of such a multi-barrier construction, see also FIG. 4a.
The multi-barrier structure, constructed in accordance with appropriate parameters, can produce a negative impedance region in accordance with multi-layer tunneling barrier phenomena of the super lattice, as explained by Tsu and Esaki, Applied Physics Letters, Vol. 22, No. 11, June 1, 1973, p. 562. In FIG. 3 of that article the I-V curve for the double and triple barriers is given, in which regions of negative impedance are clearly seen.
Multi-layer devices within the general paramaters given above, i.e. contact areas of less than about 1 micron2 and barrier thicknesses less than about 10 Angstrom, can achieve the desired high-speed response while demonstrating the negative impedance characteristic just mentioned. According to the invention, such a structure is integrated with deposits forming a tank circuit, thus to provide a negative impedance oscillator.
FIG. 5 denotes the current-voltage characteristic curve of such a device, and FIG. 6 represents the incorporation of such a device into an oscillator.
In FIG. 6, a multi-barrier diode 30 of the metal-dielectric-metal type (as illustrated in FIG. 3) is constructed in accordance with the foregoing parameters, having a region of negative slope in its current-voltage characteristic and provided with leads 32 and 34 for connection to a voltage source to bias the junction into the negative conductance region. The junction typically has between its legs 14, 26 a capacitance of the order of 10-12 to 10-14 farad, the smaller values being advantageous for higher frequency operation. A pair of parallel electrically conductive line structues 36 and 38 is deposited upon the substrate to form an inductive element, the ends 36a and 38a being connected to respective sides of the junction. The length of these line structures is approximately a quarter wavelength of the frequency of interest taking into consideration the fact that the wavelength for the integrated structure upon the substrate will be less than the free space wavelength of the same frequency. In the embodiment illustrated in FIG. 6, these line structures are each less than about 2 microns in width and thickness, and the selected spacing S between them is about 5 microns. The inductance of the tank circuit is determined in accordance with well known theory by the length of the structures 36, 38 which is selected with regard to the frequency that will be used in the device.
The right-hand ends of the parallel line structures 36 and 38 are connected to the inner ends of dipole antenna 40, each branch 40a, 40b of this antenna extending at a right angle to its respective line structure and the overall length of the dipole being approximately a half wavelength of the operating frequency. For an example the length of the parallel lines 36 and 38 may be 10 or 20 or 50 microns (and each branch of the dipole antenna the same length) depending upon the frequncy at which resonance is desired. The capacitance for the thus formed tank circuit is provided dominantly by the junction 30, but also in part by the parallel line structures 36, 38; the inductance of the circuit is dominated by the properly spaced parallel lines; the resistance is dominated by the radiation resistance of the antenna, and the more detailed design considerations readily follow from known work in negative resistance oscillators, tailoring the oscillator expressions to the very small size and the very small capacitance and inductance present, and being careful to take into consideration effects that the leads 32, 34 may have at very high frequency.
Alternative structures will be evident to those skilled in the art as illustrated in FIG. 7 wherein elements identified with primed numerals have functions corresponding to those with unprimed numerals in FIG. 6.
In operation, a DC voltage is applied across pads 32 and 34, sending a current through the tunneling contact 30 and driving the tunneling contact with the negative impedance region. Then this system will oscillate at a frequency determined by the circuit as a whole. The antenna will radiate at the resonant frequency, thus providing a source, for instance, of far infrared of a frequency determined by the particular dimensions chosen in depositing the structure. Also, by changing the bias between the contact pads 32 and 34, it is possible to change from one portion to another of the negative impedance region of the current-voltage curve of the active device, thus to pull the frequency. Using this technique it is possible to achieve a tunable range of about 10 percent of the frequency.
FIG. 6a illustrates how the oscillator of FIG. 6 may be advantageously fabricated by first depositing integral metallic structure I defining the first metal layer of the active device, first lead 32, inductance line 36 and the upper branch 40a of the dipole antenna, then after oxidizing, depositing a controlled thickness metal layer 42, and after again oxidizing, depositing a second line structure II defining the third metal layer of the active device, the second lead 34, line 38 and the lower branch 40b of the antenna. It may be noted that only three deposits and two oxidation steps are required for the construction of the complete oscillator.
Typically a contact area of 1 micron2 may be assumed with a metal-to-metal separation at each dielectric layer on the order of the atomic dimension, that is, less than about 10 Angstrom. The inductance L provided dominantly by the parallel line structure is of the order of 3 × 10-12 Henry and the capacitance C is equal to 10-14 or 10-15 farad, contributed dominantly by the junction. Accordingly the frequency f is given by:
f = 1/(2 π√LC) = 1012 Hz
The dipole antenna 40,/is then constructed with overall length λ/2 of 100 microns suitably corrected to take into account the change in wavelength due to the presence of the substrate upon which the antenna is deposited.
As shown in FIG. 8 it is also possible to deposit an integrated structure with an active device in a loop where the inductance would appear essentially from the loop. Referring to FIG. 8, a first metal deposit 41 of e.g., 20 micron length and curved to form half of an inductance loop is vacuum deposited on the substrate and then allowed to oxidize to provide a dielectric layer thereupon. Then intermediate layer 42 is deposited at A and oxidized. Subsequently, a second half loop 43 of metal of line width and thickness of about 2μ is deposited with ends overlapping the ends of the half loop 41 at A and at B. The overlap in the area A is sized to produce a multi-barrier junction with a contact area of 1 micron2 or less in accordance with the principles outlined above. The area of overlap B at the opposite ends of the loop however is sized much larger and therefore is effectively shorted by its own capacitance at the frequencies of interest. (Or area B may be protected from oxidation effects and an ohmic contact formed.) Thus there is provided a tank circuit with the inductances of the loop appearing in parallel with the capacitance of the junction. Dipole antenna 44 and biasing leads 46, 48, may be provided by the same operations employed to form the diode as described in connection with FIG. 6a. In this structure the single loop defines the inductance of the tank circuit and again the tiny contact establishes the capacitance, it being remembered that the very small separation of the metal layers of the active device, on the order of the atomic dimension, 5 or 10 Angstrom, makes the capacitance of the contact dominant. Typical dimensions and parameters for this construction are readily calculated from well known theory for any desired operating frequency in accordance with the principles discussed above.
One application of devices according to the invention is as a source of far infrared or infrared radiation, tunable by varying the bias voltage on the contact and thus shifting the point of operation along the I-V curve in the negative impedance region. Thus it can be seen that in a way a new laser is formed for generating radiation; and it is possible to incorporate the device in a Fabre Perot resonator to achieve a more intense radiation level or to employ a phased array of the devices to achieve a similar effect.
The oscillator is particularly useful in spectroscopy and in atomic clocks for the purpose of up-conversion or in other instances for the purpose of down-conversion, discussed below.
Multi-barrier active devices such as those described above are also useful to provide amplification and feedback controlled circuits. In such cases the intermediate metal layer is extended as shown in dotted lines in FIG. 3. The intermediate metal acts as a control element along the line of the theoretical consideration given by Davis and Hosack in "Double Barrier in Thin-Film Triodes", Journal of Applied Physics, Apr. 1963, p. 864, et seq. and references there cited, to which reference is made. Feedback from one portion of a device to another may be accomplished by radiative coupling through antennas deposited as integrated structures with the active devices as illustrated in FIG. 9 where two multi-layer active devices are deposited upon a substrate. The first device 50 has its intermediate metal layer 52 connected to an λ/2 antenna leg 54 and the second device 56 has its intermediate metal layer 58 connected to the emitter 60 of the first device. Also the collector of the second device is connected to an λ/2 antenna leg 62. The two antenna legs are parallel and spaced in accordance with known antenna theory to enable positive feedback which thus supports resonance in the selected frequency region. As with other such oscillators the system has a band through which it can be tuned by varying the bias across the devices by means of bias pads 61, 64, 66, and 68.
Additional elements may be added to obtain more complex circuits as indicated in FIG. 9 where in dotted lines a third ultra-thin dielectric layer triode is indicated, controlled by the second, and with antenna coupled to the second for positive feedback. Thus a phased array of three or more active devices can be achieved, all coupled for positive feedback, and their antennas also radiating to the outside world at the resonant frequency.
It is further apparent that by appropriately adjusting antenna separation negative feedback can be achieved and by appropriate changes in the printed circuitry and with reference to the extensive literature on circuit theory, both monostable and bistable multi-vibrators can be achieved, with obvious application as signal sources and storage components in computers and other logic networks operable at speeds greatly increased over those heretofore known.
Negative impedance diode oscillator or a triode configuration used as a multivibrator or as a flip-flop circuit element can be used to subdivide frequencies. This can be done in accordance with principles well known at radio frequencies. See Principles and Applications of Electron Devices, Paul D. Ankrum, international Text Book Company. In such devices the frequency of oscillation can become synchronized by an external signal. An applied voltage of a frequency higher than the free-running oscillation frequency of the system can cause it to lock at the frequency of the synchronizing signal or it can become locked at a subharmonic frequency of it. Accordingly the invention described here enables subdividing frequencies starting from far infrared or infrared frequencies. Possible applications of this principle include high frequency counter applications.
Referring now to FIG. 10 a laser 80 applies a radiation input 82 of infrared wave length λ through lense 84 to a first negative impedance oscillator structure 86 which has an antenna input structure 87 resonant to λ. The same oscillator structure 86 also has an output structure 89 resonating at 2λ. The radiation input through lens 84 at frequency V = c/λ will trigger the oscillator 86 to oscillate at frequency V1 = C/(2λ) which in turn, through its output, is coupled to a second stage 90 of similar construction but designed to oscillate at frequency V2 = c/(4λ). As indicated by the dots, further stages subdivide the frequencies in additional steps. The output can be coupled out for instance by a parallel pair of lines 92 forming a transmission line. More elaborate oscillator designs utilizing diodes as well as triode configurations can be utilized in accordance to the principles well known in the radio frequency regions, see for instance the Ankrum reference cited above, and thus multivibrators or flip-flop elements can be achieved.
Reference is made now to the aspect of the invention concerning enhancement of nonlinearities in the active devices and introduction of nonlinear reactive components as desired by appropriate relation of the operating frequencies to the quantized modes. In a diode structure, it is known that tunneling electrons can be scattered from local quantized modes of the dielectric barrier. This effect has been frequently observed in large area tunneling junctions which are not capable of responding at high frequencies. In these cases, the effect appears as changes in the DC I-V characteristic curve whose onsets occur at fairly well-defined values of bias voltages. These voltages correspond to the quantized energies of the vibrational modes associated with specific molecular bonds in the dielectric barrier (see J. Lambe and R. C. Jakleivic, Phys. Rev. 165,821 (1968). Employing the present invention, tunneling junctions can be made with a capability of operating at very high alternating frequencies and according to a further feature of the invention the input frequencies are then chosen to be in the regions where the local quantized modes of the dielectric barrier lie. In these regions resonant scattering of alternating currents from the local modes will result in dispersive effects causing a phase-shift to appear between alternating current and the applied voltage. This will in turn introduce inductive or capacitive components in the tunneling impedance. These reactive effects will be nonlinear functions of the applied voltages. According to this effect, a diode junction when used at frequencies close to the resonances of its local quantized modes, will show enhanced non-linearities. Beyond this according to the invention, in these regions, the nonlinear capacitive or inductive effects can be used to match the junctions to the antenna structure integrated to it.
Similar dispersive effects are also expected in multibarrier structures (super lattice); i.e. the non-linearities of tunneling impedance, the resistive as well as the reactive components of impedance, are enhanced in the regions where tunneling electrons resonate with the multibarrier structure.
Referring now to FIG. 11, a printed diode metal-to-metal junction is shown where the letters M denote the two metal components and 0 the intervening dielectric layer. The levels E1, E2, E3, denote quantized energies of the local modes. The voltage V applied across the dielectric layer 0 is V = V0 + v Cos ω t (where V0 is the biasing voltage and v the applied radiation field) and a current flows I = I0 + i Cos (ωt + φ) where i is is the amplitude of the alternating current and φ is the phase shift.
Depending upon the relationship of the frequency of the voltage applied across the metal-to-metal diode structure a phase shift, φ, occurs between the applied voltage V and the resultant current I. This results in the reactive components, as mentioned before and, depending upon the particular relationship of the frequency with respect to the exact resonances any of the equivalent circuits of FIGS. 12a, 12b, 12c and 12d can be realized. The variability of the resistive and inductive components indicated in FIG. 12a can be controlled by varying the biasing field V0 as well as the applied radiation yield v.
The effects here described can be introduced or enhanced by utilizing impurities in the dielectric layer as it is being deposited. The impurities can be for instance H2 O, or D2 O or organic molecules such as methyl aclohol, etc. Different impurities will give different energy spacings.
Advantages of nonlinear reactive effects in parametric frequency conversion and frequency mixing are well-known from previous knowledge in the radio and microwave frequency work. All of these previous considerations are applicable to the invention described here. For instance in frequency down conversion, nonlinear reactive effects will not be limited in efficiency by the Manly Row relation and hence can be more efficient as compared to applicaton of resistive nonlinear effects. Referring to FIG. 13, an example of parametric frequency down conversion is to drive the nonlinear junction 90 (constructed according to the techniques described above) at a frequency ω. By introducing resonant tank circuits 92 and 94 by the thin film techniques already mentioned, these circuits being resonant at frequencies ω1 and ω2, such that ω1 + ω2 = ω, oscillation will then occur at ω1 and ω2 if the amplitude of the driving voltage at ω is sufficiently large. The outputs at ω1 and ω2 are then taken along the transmission line output coupling 96. The nonlinear reactive components employed in this embodiment can be obtained by, e.g., two local quantized modes, one of them having an energy spacing in the vicinity of h ω1 and the other in the vicinity of h ω2 where h is the Plank Constant. In addition to a tank-circuit, external Fabre Perot resonators can be used to enhance the Qs at the appropriate frequencies as suggested in FIG. 13a.
By limiting the junction contact area to about 1 micron2 and the dielectric thickness to less than 10 Angstrom, the devices according to the invention can be operated at frequencies in the submillimeter and far infrared region, and at even higher frequencies if precise manufacturing techniques such as X-ray or electron beam lithography are used to achieve very small dimensions.
Operation of the active devices in a negative impedance oscillator (whether in CW or as a relaxation oscillator) or in feedback arrangements, can in certain instances be enhanced by cryogenic cooling.
It will be understood that while a major aspect of the invention relates to the achievement of active devices, especially oscillators and vibrators operable in the far infrared and infrared regions, the discoveries, particularly with reference to the performance of ultra-thin dielectric layers and small areas of the printed structures, as well as the associated printed antenna, will lead in certain instances to application to adjoining frequency regions, both higher and lower.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3229106 *||Oct 16, 1961||Jan 11, 1966||Tektronix Inc||Method of modulating light with intrinsic semiconductor device and electrical signal modulator employing such device|
|US3462605 *||Sep 22, 1965||Aug 19, 1969||Gen Electric||Semiconductor light-emitter and combination light-emitter-photocell wherein the reflector of the light-emitter is comprised of a material different from that of the light-emitter|
|US3649838 *||Jul 25, 1968||Mar 14, 1972||Massachusetts Inst Technology||Semiconductor device for producing radiation in response to incident radiation|
|US3755678 *||Aug 10, 1970||Aug 28, 1973||Massachusetts Inst Technology||Apparatus and processes for detection, generation and frequency measurements of electro-magnetic radiation in the infrared and visible domain|
|1||*||Dees, The Microwave Journal, vol. 9, Sept. 1966, pp. 48-55.|
|2||*||Hocker et al., vol 12. No. 12, June 1968, pp. 401, 402.|
|3||*||Small et al., Applied Physics Letters, vol. 24, No. 6, 3/15/74, pp. 275-279.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US5093699 *||Mar 12, 1990||Mar 3, 1992||Texas A & M University System||Gate adjusted resonant tunnel diode device and method of manufacture|
|US5331162 *||Nov 22, 1991||Jul 19, 1994||Trw Inc.||Sensitive, low-noise superconducting infrared photodetector|
|US6208702 *||Jan 23, 1998||Mar 27, 2001||International Business Machines Corporation||High frequency clock signal distribution utilizing CMOS negative impedance terminations|
|US6979998 *||Apr 16, 2003||Dec 27, 2005||Hewlett-Packard Development Company, L.P.||Magnetic filter|
|US8115688 *||Oct 24, 2008||Feb 14, 2012||Intelleflex Corporation||RF conduit and systems implementing same|
|US20040207399 *||Apr 16, 2003||Oct 21, 2004||Manish Sharma||Magnetic filter|
|US20100103071 *||Oct 24, 2008||Apr 29, 2010||Farrell Edward M||Rf conduit and systems implementing same|
|US20110062329 *||Sep 13, 2010||Mar 17, 2011||David Ben-Bassat||Electromagnetic based thermal sensing and imaging|
|U.S. Classification||250/214.00R, 331/66, 257/104, 257/431, 505/849|
|International Classification||H01L49/00, H01L31/00, H01J29/45|
|Cooperative Classification||H01L49/00, H01L31/00, H01J29/458|
|European Classification||H01L31/00, H01L49/00, H01J29/45D|