JP2016111541A - Resonance tunnel diode oscillator and method of manufacturing the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a highly reliable resonance tunnel diode oscillator, without increasing the number of process steps, by suppressing parasitic resonance of a resonance tunnel diode (RTD) and an external circuit with no metal wiring, and to provide a method of manufacturing the same.SOLUTION: A resonance tunnel diode oscillator 30 includes a first resonance tunnel diode RTD1 including a first anode A1 and a first cathode K1, and having asymmetric current voltage characteristics, and a second resonance tunnel diode RTD2 including a second anode A2 and a second cathode K2, and having asymmetric current voltage characteristics. The first RTD1 and second RTD2 are connected in reverse parallel, and when any one of the first RTD1 or second RTD2 is biased to negative resistance oscillation state, the other is biased to the resistance state.SELECTED DRAWING: Figure 1

Description

  The present embodiment relates to a resonant tunneling diode type oscillator and a manufacturing method thereof.

In recent years, electronic devices such as transistors have been miniaturized, and the size has become nano-sized. Therefore, a new phenomenon called a quantum effect has been observed. And development aiming at realization of ultra-high-speed devices and new functional devices using this quantum effect is in progress. In such an environment, an attempt to perform large-capacity communication, information processing, imaging, measurement, or the like using a frequency range of 0.1 THz (10 ¹¹ Hz) to 10 THz, particularly called a terahertz band. Has been done. This frequency region is an undeveloped region between light and radio waves. If a device that operates in this frequency band is realized, in addition to the above-mentioned imaging, large-capacity communication / information processing, physical properties, astronomy, living things, etc. It is expected to be used for many purposes such as measurement in various fields.

  As an element that oscillates a high-frequency electromagnetic wave having a frequency in the terahertz band, an element having a structure in which a resonant tunneling diode (RTD) and a fine slot antenna are integrated is known. Further, an element having an MIM (Metal Insulator Metal) structure in which a metal and an insulator are stacked at both ends of the slot antenna, the insulator is sandwiched between upper and lower electrode metals, and short-circuited in a high frequency is disclosed.

  On the other hand, a Schottky barrier diode (SBD) is well known as a terahertz detection element, but this cannot be used as a terahertz oscillation element.

  On the other hand, the RTD can be used as both an oscillation element and a detection element, but when used as a terahertz oscillation element, a resistor such as Bi is connected between the anode and the cathode in order to suppress parasitic oscillation.

  A low-noise terahertz oscillation detecting element in which a resonant tunneling diode and SBD are integrated is also disclosed.

JP 2007-124250 A JP 2012-217107 A

T. Wei and S. Stapleton, “Equivalent circuit and capacitance of double barrier resonant tunneling diode,” J. Appl. Phys. 73 (2), 15 January 1993, pp. 829-834. C. Bayram, Z. Vashaei, and M. Razeghi, “Reliability in room-temperature negative differential resistansce characteristics of low-aluminum content AlGaN / GaN double-barrier resonant tunneling diodes,” Appl. Phys. Lett. 97, 181109 (2010 ). Lin ’an Yang, Hanbing He, Wei Mao, and Yue Hao,“ Quantitative analysis of the trapping effect on terahertz AlGaN / GaN resonant tunneling diode, ”Appl. Phys. Lett. 99, 153501 (2011). Z. Suet, D.J.Paul, J. Zhang and S.G.Turner, “Si / SiGe n-type resonant tunneling diodes fabricated using in situ hydrogen cleaning,” Appl. Phys. Lett. 90, 203501 (2007).

  A terahertz wave oscillator having an RTD has a problem that the main oscillation in the terahertz band is restricted by parasitic oscillation with an external circuit caused by the negative resistance of the RTD. As a method for suppressing the parasitic oscillation, a method is used in which a resistor is arranged in parallel with the diode so that the negative resistance is not visible to the external circuit. Until now, wiring has been performed with bismuth (Bi), which has a relatively high resistance, and metals such as nickel (Ni), titanium (Ti), and platinum (Pt), which are commonly used in semiconductor processes, to suppress parasitic oscillation. This oscillation was obtained. However, structurally, a metal layer for metal wiring different from the RTD is required, and the number of processes increases. In addition, the reliability of metal wiring varies and is unstable. Furthermore, when looking at applications such as high-speed wireless communication exceeding several tens of Gbps in the future, it is necessary to realize such high-speed modulation, but there is a limit in response speed with metal wiring.

  The present embodiment provides a resonant tunneling diode type oscillator that suppresses the parasitic oscillation of the RTD and the external circuit without metal wiring, and has high reliability without increasing the number of process steps, and a manufacturing method thereof.

  According to one aspect of the present embodiment, the first resonant tunneling diode having an asymmetric current-voltage characteristic, a second anode, and a second cathode are provided, the first anode and the first cathode, and an asymmetric current. A second resonant tunneling diode having voltage characteristics, wherein the first resonant tunneling diode and the second resonant tunneling diode are connected in antiparallel, and the first resonant tunneling diode and the second resonant tunneling diode are A resonant tunneling diode oscillator is provided in which one is biased into a negative resistance oscillation state and the other is biased into a resistance state.

  According to another aspect of the present embodiment, a semiconductor substrate, a first semiconductor layer disposed on the semiconductor substrate, a first cathode region formed by patterning the first semiconductor layer, and a first A first cathode connected to the first cathode region, a first resonant tunneling diode having a first anode connected to the second cathode region, and a second cathode connected to the second cathode region. A second resonant tunneling diode having a second anode connected to the first cathode region, wherein either one of the first resonant tunneling diode and the second resonant tunneling diode is biased to a negative resistance oscillation state. When done, a resonant tunneling diode type oscillator is provided in which the other is biased into a resistive state.

  According to another aspect of the present embodiment, a step of forming a first semiconductor layer on a semiconductor substrate, and a step of patterning the first semiconductor layer to form a first cathode region and a second cathode region Forming a first resonant tunneling diode in which a first cathode is connected to the first cathode region and a first anode is connected to the second cathode region; and a second cathode is connected to the second cathode region. Forming a second resonant tunneling diode having a second anode connected to the first cathode region, and forming a first cathode electrode commonly connected to the second anode electrode on the first cathode region. And a step of forming a second cathode electrode commonly connected to the first anode electrode on the second cathode region. There is provided.

  According to the present embodiment, it is possible to provide a resonant tunneling diode type oscillator having a high reliability without suppressing the parasitic oscillation of the RTD and the external circuit without using metal wiring, and without increasing the number of process steps, and a method for manufacturing the same. .

The typical bird's-eye view block diagram of RTD based on a basic technique. Schematic circuit representation of RTD according to basic technology. Current-voltage characteristics of RTD according to basic technology. In RTD which concerns on basic technology when there is no parallel resistance, (a) Simple equivalent circuit diagram, (b) Explanatory drawing of parasitic oscillation with an external circuit. In RTD which concerns on basic technology in case there exists parallel resistance, (a) Simple equivalent circuit diagram, (b) Explanatory drawing of RF basic oscillation by which the parasitic oscillation with the external circuit was suppressed. An example of a device surface photograph of an RTD according to the basic technology when there is a parallel resistance. It is another device surface photograph example of the RTD according to the basic technology when there is a parallel resistance, (a) an example in which the resistance wiring width is 3 μm, (b) an example in which the resistance wiring width is 5 μm, (c) a resistance wiring width (B) An example in which the resistance wiring width is 20 μm. (A) Typical cross-sectional structure diagram of RTD1 applicable to RTD type oscillator concerning embodiment, (b) Typical cross-section figure of RTD2 applicable to RTD type oscillator concerning embodiment. 6 shows an example of an asymmetric current-voltage characteristic of RTD 1 that can be applied to the RTD type oscillator according to the embodiment. The equivalent circuit block diagram of the RTD type oscillator which concerns on embodiment. 6 shows an example of an asymmetric current-voltage characteristic of RTD 1 that can be applied to the RTD type oscillator according to the embodiment. 6 shows an example of an asymmetric current-voltage characteristic of RTD 2 that can be applied to the RTD type oscillator according to the embodiment. 5 is a current-voltage characteristic example estimated from a combined resistance of RTD1 and RTD2 in the RTD oscillator according to the embodiment. The typical plane block diagram of the wiring structure of the RTD type oscillator which concerns on embodiment. FIG. 15 is a schematic sectional view taken along the line II of FIG. FIG. 15 is a schematic sectional view taken along line II-II in FIG. 14. FIG. 15 is a schematic sectional view taken along line III-III in FIG. 14. The typical cross-section figure of the modification of RTD1 applicable to the RTD type | mold oscillator which concerns on embodiment. FIG. 4 is a schematic cross-sectional structure diagram that can be formed in the simultaneous process of RTD1 and RTD2 applicable to the RTD type oscillator according to the embodiment. The typical plane pattern block diagram of the RTD type oscillator which concerns on embodiment provided with a dipole antenna. The typical plane pattern block diagram of the RTD type oscillator which concerns on embodiment provided with a taper slot antenna.

  Next, the present embodiment will be described with reference to the drawings. In the following, the same reference numerals are assigned to the same blocks or elements to avoid duplication of explanation and simplify the explanation. It should be noted that the drawings are schematic and different from the actual ones. Moreover, it is a matter of course that portions having different dimensional relationships and ratios are included between the drawings.

The embodiment described below exemplifies an apparatus and a method for embodying the technical idea, and does not specify the arrangement of each component as described below. This embodiment can be modified in various ways within the scope of the claims.
[Basic technology]
A schematic bird's-eye view configuration of the RTD according to the basic technology is expressed as shown in FIG. Also, a schematic circuit representation of the RTD according to the basic technology is expressed as shown in FIG. 2, and the current-voltage characteristics of the RTD according to the basic technology are expressed as shown in FIG.

  As shown in FIG. 1, the RTD according to the basic technology is disposed on a semiconductor substrate 1, a second electrode 2 disposed on the semiconductor substrate 1, and a second electrode 2 on the semiconductor substrate 1. The first electrode 4, the MIM reflector 50 formed between the first electrode 4 and the second electrode 2 with the insulating layer interposed therebetween, and the first electrode facing the semiconductor substrate 1 adjacent to the MIM reflector 50. An active element 90 disposed substantially at the center of the resonator disposed between the first electrode 4 and the second electrode 2; and the first electrode 4 adjacent to the active element 90 and facing the semiconductor substrate 1; A horn opening disposed between the first electrode 4 and the second electrode 2 facing on the semiconductor substrate 1 adjacent to the waveguide disposed between the second electrodes 2, and the first electrode 4 And a resistance element 114 made of bismuth connected between the second electrodes 2. The active element 90 is typically an RTD.

  In the terahertz wave oscillator having the RTD, the main oscillation in the terahertz band is regulated by parasitic oscillation with an external circuit due to the negative resistance of the RTD. As a method for suppressing the parasitic oscillation, as shown in FIG. 1, a resistive element 114 made of bismuth is arranged in parallel with the RTD so that the negative resistance is not visible to the external circuit.

The resistance R _Bi of the resistance element 114 made of bismuth is connected in parallel between the anode and cathode of the _{RTD with} respect to the resistance R _RTD of the _RTD , as shown in FIG. As a result, the combined resistance R _t between the anode A · cathode K of the RTD, RTD resistor parallel-connected resistor R _RTD · R of the resistance R _Bi of R _RTD and resistance element 114 consisting of bismuth _Bi / (R _RTD + R _Bi ).

By arranging the resistance R _Bi of the resistance element 114 made of bismuth in parallel with the resistance R _RTD of the _RTD , as shown in FIG. 3, the negative resistance ( The occurrence of −ΔV / ΔI) is shifted to a relatively large voltage and relatively large current side, and it is confirmed that there is an effect of suppressing parasitic oscillation.

  Since parasitic oscillation occurs between the negative resistance region and the external circuit, the resistance is arranged in parallel with the RTD to make it difficult to see the negative resistance from the external circuit. In this way, parasitic oscillations other than the main oscillation can be suppressed.

The requirement for this is that the combined resistance R _t > = 0,

R _Bi ≦ ΔV / ΔI (= R _RTD ) (1)

It is represented by

  Until now, wiring was made of Bi, which has a relatively high resistance value, or a metal such as Ni, Ti, Pt or the like generally used in semiconductor processes, and this oscillation was obtained by suppressing parasitic oscillation. However, a layer different from the RTD is required, and the number of processes increases. In addition, the reliability of metal wiring varies and is unstable. Furthermore, when looking at applications such as high-speed wireless communication exceeding several tens of Gbps in the future, it is necessary to realize such high-speed modulation, but there is a limit in response speed with metal wiring. Certainly, it is possible to cope with a high-speed response by reducing the thickness of the wiring and optimizing the shape, but there is a concern that the manufacturing error and the reliability decrease accordingly.

In the RTD according to the basic technique when there is no parallel resistance, a simple equivalent circuit configuration is expressed as shown in FIG. 4A, and an explanatory diagram of parasitic oscillation with an external circuit is shown in FIG. Represented as shown. In addition, in the RTD according to the basic technique when there is a parallel resistance R _p , a simple equivalent circuit configuration is expressed as shown in FIG. 5A, and the RF basic oscillation in which parasitic oscillation with an external circuit is suppressed. The explanatory diagram is expressed as shown in FIG.

In FIGS. 4 and 5, L1 and L2 correspond to inductances of wirings such as bonding wires and lead wires that are externally viewed from the RTD portion. C _p indicates the parasitic capacitance of the RTD portion. In addition, the RTD of the RTD portion is indicated by a diode display. The antenna portion is represented by a parallel circuit of an antenna inductance LA and an antenna resistance RA. Further, an external circuit such as a connector or a drive circuit is connected between the anode terminal A and the cathode terminal K.

(Parasitic oscillation with external circuit)
When an attempt is made to produce a terahertz wave oscillator using an RTD, as shown by the solid line arrow in FIG. 4B, a parasitic oscillation occurs between a portion that is externally viewed from the RTD. In FIG. 4B, the broken line shows how the main oscillation is suppressed, and the solid line shows how the parasitic oscillation occurs. If the negative resistance of the RTD is seen with respect to the external circuit, the Q value is higher when the external circuit oscillates at a low frequency than the resonant circuit that performs the RF basic oscillation, and the oscillation is easier. As a result, parasitic oscillation with the external circuit occurs remarkably.

(Parallel resistance)
In general, as a method of suppressing parasitic oscillation, as shown in FIGS. 5A and 5B, a parallel resistance R _p is arranged with respect to the RTD, so that the Q value of the parasitic oscillation is lowered and oscillation is performed. Is devised such that power is efficiently transferred to the oscillation side. In FIG. 5B, the broken line indicates a state where the Q value of the parasitic oscillation is reduced to suppress the oscillation, and the solid line indicates a state where the main oscillation is efficiently performed.

If the parallel resistance R _p is arranged and the negative resistance becomes invisible from the outside, the Q of the parasitic oscillation that oscillates at a low frequency with the outside becomes low, and the oscillation becomes harder than the RF basic oscillation. Therefore, the original RF fundamental oscillation occurs.

(Q value)
Briefly explaining the Q value, it is an index indicating how stably the oscillation state exists. An expression showing the Q value when a parallel resonant circuit is considered is described below.

Q = 1 / R _t · √ (LA / C _p ) (2)

Arranging the parallel resistance R _p corresponds to increasing the combined resistance R _t .

If the Q value of the parasitic oscillation as shown above is considered from this equation (2), disposing the parallel resistor R _p corresponds to increasing R _t, and as a result, the Q value is reduced. It is done. This shows that the parallel resistance R _p qualitatively suppresses parasitic oscillation.

(RTD device surface micrograph)
An example of a device surface photomicrograph of RTD according to the basic technique in the case where there is a parallel resistance is expressed as shown in FIG. FIG. 6 corresponds to the top view of FIG.

  FIG. 7A shows another example of the device surface micrograph of RTD according to the basic technology in the case where there is a parallel resistance, and the resistance wiring width is 3 μm, and the resistance wiring width is 5 μm. 7B is represented as shown in FIG. 7B, an example where the resistance wiring width is 10 μm is represented as shown in FIG. 7C, and an example where the resistance wiring width is 20 μm is illustrated in FIG. 7D. It is expressed as follows.

  In the RTD according to the basic technology, a parallel resistor has been produced by using Bi having a relatively high resistance value and metals such as Ti, Pt, and Ni generally used in a semiconductor process. Due to the structure of the device, we have successfully improved the film-covering property of the stepped part, such as oblique deposition, and achieved device operation verification.

  However, in order to perform metal wiring, it is necessary to increase the number of process steps. Further, when the characteristics of the Bi wiring and the like were analyzed by an energization test or the like, a change in resistance value was observed with respect to the operating environment temperature. Although this is a general characteristic of metal materials, it is designed with consideration given to matching with the RTD, not just the wiring, so resistance value changes with temperature may affect the stable operation of the device. . There is also a problem that the metal wiring is disconnected during use as a failure mode of the element.

  In addition, the metal wiring has a certain degree of impedance, and becomes a cause of loss in high-frequency driving. Therefore, there is a possibility that it becomes a handicap for high-speed operation of the device.

[Embodiment]
A schematic cross-sectional structure of the first RTD 1 applicable to the RTD type oscillator 30 according to the embodiment is represented as shown in FIG. 8A, and a schematic cross-sectional structure of the second RTD 2 is shown in FIG. 8B. It is expressed as follows. 8A and 8B, the direction indicated by the arrow is the direction in which the forward current flows. Details of the device structure will be described later.

  Also, an example of the asymmetric current-voltage characteristic of the first RTD 1 applicable to the RTD oscillator 30 according to the embodiment is expressed as shown in FIG.

  In addition, an equivalent circuit configuration of the RTD oscillator 30 according to the embodiment is expressed as indicated by a portion indicated by reference numeral 30 in FIG.

  As shown in FIG. 10, the RTD oscillator 30 according to the embodiment includes a first RTD1 including a first anode A1 and a first cathode K1, and a second RTD2 including a second anode A2 and a second cathode K2. Prepare. Here, the first RTD 1 and the second RTD 2 are connected in antiparallel, and when one of the first RTD 1 and the second RTD 2 is biased to the negative resistance oscillation state, the other is biased to the resistance state.

  For example, in FIG. 9, when the forward voltage value is +200 mV, the first RTD 1 can be regarded as a pure resistance, but when the forward voltage value is −200 mV, the first resistance value is Can be considered. Here, as shown in FIG. 9, the first RTD 1 has asymmetric current-voltage characteristics. Similarly, the second RTD 2 also has asymmetric current-voltage characteristics.

  Further, as shown in FIG. 10, in the RTD oscillator 30 according to the embodiment, the first anode A1 and the second cathode K2 are connected, and the first cathode K1 and the second anode A2 are connected.

(Asymmetric current-voltage characteristics)
FIG. 11 is an example of an asymmetric current-voltage characteristic of the first RTD 1 that can be applied to the RTD oscillator 30 according to the embodiment, and an enlarged view of a forward voltage value in a range of −400 mV to +400 mV is shown in FIG. Is done. FIG. 12 shows an example of the asymmetric current-voltage characteristics of the second RTD 2 that can be applied to the RTD type oscillator 30 according to the embodiment, and an enlarged view of the forward voltage value in the range of −400 mV to +400 mV. It is expressed in

  Furthermore, in the RTD oscillator 30 according to the embodiment, an example of current-voltage characteristics estimated from the combined resistance of the first RTD 1 and the second RTD 2 based on the characteristics of FIGS. 11 and 12 is expressed as shown in FIG. . According to this calculation result, it can be seen that connecting two RTDs in parallel with their polarities reversed has a sufficient effect of suppressing parasitic oscillation.

  For example, in the example of the asymmetric current-voltage characteristic of the first RTD 1 (FIG. 11), the first RTD 1 is a pure resistance in a bias state with a forward voltage value of +200 mV, and a negative resistance in a bias state with a forward voltage value of −200 mV. In the example of the asymmetric current-voltage characteristics of the second RTD 2 (FIG. 12), in the bias state where the forward voltage value is +200 mV, the second RTD 2 is in the negative resistance state and the forward voltage value is −200 mV bias. In a state, it becomes a pure resistance state.

  Therefore, in the RTD type oscillator 30 according to the embodiment, the first RTD 1 and the second RTD 2 are connected in anti-parallel, so when either one of the first RTD 1 and the second RTD 2 is biased to the negative resistance oscillation state, the other Can be biased into a resistive state.

  When two RTDs having the current-voltage characteristics shown in FIGS. 11 and 12 are used and the polarities are reversed, an equivalent circuit as shown in FIG. 10 is obtained. At this time, it is assumed that RF (THZ band) oscillation is attempted using the negative resistance region on the negative side of one RTD. If +200 mV is applied, one RTD takes -200 mV, so that the negative resistance region is biased and becomes a drive voltage. At this time, a voltage of +200 mV is applied to the other RTD and it behaves as a pure resistor.

  As is clear from comparison between the current-voltage characteristics of the first RTD 1 shown in FIGS. 9 and 11 and the current-voltage characteristics estimated from the combined resistance shown in FIG. 13, in the RTD oscillator 30 according to the embodiment, the first RTD 1 Since the second RTD 2 is connected in antiparallel, when the first RTD 1 is biased to the negative resistance oscillation state, the second RTD 2 can be biased to the resistance state. Thereby, according to the RTD type oscillator according to the embodiment, parasitic oscillation of the RTD and the external circuit can be suppressed without metal wiring.

  The RTD oscillator 30 according to the embodiment can provide a terahertz oscillator using a second RTD 2 that suppresses parasitic oscillation with an external circuit, as a device structure that suppresses parasitic oscillation of the first RTD 1.

  This technique is not limited to the terahertz wave band, and is a basic technology that can be generalized as a technique for suppressing parasitic oscillation in the RTD and its external circuit. Specifically, it is a method in which two RTDs having asymmetric current-voltage characteristics on the plus side and the minus side are realized by inverting the polarity and arranging them in parallel.

  The RTD behaves as if it is a pure resistor in a voltage range lower than the voltage range showing a negative resistance. Reversing the polarity means that when one RTD is approaching a negative resistance, the other RTD behaves as a resistor.

  As a result, as will be described later, it is possible to produce a device structure that suppresses parasitic oscillation in the same epitaxial substrate and in the same process. Since there is no extra wiring, reliability and high-frequency characteristics can be improved.

  In the RTD type oscillator 30 according to the embodiment, the number of steps for metal wiring can be reduced, and the device reliability can be improved. Further, since the same epitaxial substrate is used, it is not necessary to add a new epitaxial film or process. This can be realized by changing the mask from an existing device, and can contribute to the reduction of the final manufacturing cost. In addition, since the inductance of the metal wiring portion can be removed, the characteristics of high-speed modulation can be improved.

Although the SBD may be applied to suppress parasitic oscillation, the RTD oscillator 30 according to the embodiment in which the RTD is applied to suppress parasitic oscillation has the following effects. That is,
(A) If the SBD uses the same epi substrate, an epi layer constituting the SBD must be formed separately from the RTD. Development of epitaxial film growth technology is necessary, raising the technical hurdle of film formation. In contrast, in the RTD oscillator 30 according to the embodiment, the current epi substrate can be used as it is. The technical hurdle is relatively low because the same epi substrate can be used.

  (B) It can be integrated in the same process. Since the current process can be handled by changing the mask, it can be realized without increasing the number of process steps.

  (C) There is a cost reduction effect. By eliminating the metal wiring, the number of processes is reduced, which can contribute to a reduction in manufacturing cost.

(Device structure)
A schematic plan configuration of the wiring structure of the RTD type oscillator 30 according to the embodiment is expressed as shown in FIG. 14, and a schematic cross-sectional structure taken along line II of FIG. 14 is shown as shown in FIG. 14 is represented as shown in FIG. 16, and the schematic sectional structure along the line III-III in FIG. 14 is represented as shown in FIG. .

  The RTD oscillator 30 according to the embodiment may include the semiconductor substrate 1, and the first RTD 1 and the second RTD 2 may be disposed on the semiconductor substrate 1. Here, as the semiconductor substrate 1, for example, a semi-insulating InP substrate can be applied.

  Further, the first RTD 1 and the second RTD 2 may be integrated on the semiconductor substrate 1 as shown in FIGS.

  As shown in FIGS. 14 to 17, the RTD oscillator 30 according to the embodiment patterns the semiconductor substrate 1, the first semiconductor layer 91a disposed on the semiconductor substrate 1, and the first semiconductor layer 91a. A first cathode region, a second cathode region, a first cathode K1 connected to the first cathode region, a first RTD1 having a first anode A1 connected to the second cathode region, and a second cathode; A second cathode K2 is connected to the region, and a second RTD2 is connected to the second anode A2 and the first cathode region. Here, when one of the first RTD1 and the second RTD2 is biased to the negative resistance oscillation state, the other is biased to the resistance state.

  The first RTD 1 and the second RTD 2 have asymmetric current-voltage characteristics.

Furthermore, RTD oscillator 30 according to the embodiment, as shown in FIGS. 14 to 17, a first cathode electrode 2 ₁ disposed on the first cathode region on, disposed on the second cathode region on the first a second cathode electrode 2 _2, first the anode electrode 4 ₁ connected to the first anode A1, a second anode electrode 4 ₂ connected to the second anode A2, the first cathode electrode 2 ₁ a 2 and anode electrode 4 ₂ is commonly connected, a second cathode electrode 2 ₂ are commonly connected 4 ₁ and the first anode electrode.

Note that an interlayer insulating film 9 is formed in a place where electrical insulation is required, as shown in FIG. The interlayer insulating film 9 can be formed of, for example, a SiO ₂ film. The interlayer insulating film 9 can be formed by a CVD method or a sputtering method.

-RTD-
A configuration example of the first RTD 1 and the second RTD 2 applicable to the RTD oscillator 30 according to the embodiment is a semiconductor substrate 1 made of a semi-insulating InP substrate, as shown in FIGS. A GaInAs layer 91a, which is disposed on the GaInAs layer 91a, and is doped on the GaInAs layer 91a, and is doped on the GaInAs layer 92a. The undoped layer is disposed on the GaInAs layer 92a. A GaInAs layer 93a, an RTD portion composed of an AlAs layer 94a / InGaAs layer 95 / AlAs layer 94b disposed on the GaInAs layer 93a, an undoped GaInAs layer 93b disposed on the AlAs layer 94b, and a GaInAs layer 93b On the GaInAs layer 92b disposed on and doped with an n-type impurity, and on the GaInAs layer 92b The GaInAs layer 91b is disposed on the GaInAs layer 91b, and is doped on the GaInAs layer 91b. The GaInAs layer 91c is disposed on the GaInAs layer 91c. 1 electrode 4 and the second electrode 2 disposed on the GaInAs layer 91a.

As shown in FIG. 8A, the quantum well structure QW1 of the first RTD 1 is formed by sandwiching a GaInAs layer 95 between AlAs layers 94a and 94b. The quantum well structure QW1 stacked in this way has n-type GaInAs layers 92a and 92b and n ⁺ -type GaInAs layers 91a and 91b or undoped GaInAs layers 93a and 93b used as spacer layers SP11 and SP12. The second electrode 2 ₁ and the first electrode 4 ₁ are ohmically connected via 91c.

Similarly, as shown in FIG. 8B, the quantum well structure QW2 of the second RTD 2 is formed by sandwiching a GaInAs layer 95 between AlAs layers 94a and 94b. The quantum well structure QW2 stacked in this manner has n-type GaInAs layers 92a and 92b and n ⁺ -type GaInAs layers 91a and 91b or undoped GaInAs layers 93a and 93b used as spacer layers SP21 and SP22, or The second electrode 2 ₁ and the first electrode 4 ₁ are ohmically connected via 91c.

  In the RTD oscillator 30 according to the embodiment, the first RTD 1 and the second RTD 2 are, as shown in FIGS. 8A and 8B, a quantum well layer 95 and a tunnel barrier layer 94a sandwiching the quantum well layer 95. First spacer layers SP11 (93a) and SP21 (93a) disposed on the cathode K side through 94b and second spacer layers SP12 (93b) and SP22 (93b) disposed on the anode A side are provided. The first spacer layer SP11 and the second spacer layer SP12 may have different thicknesses.

  In the RTD oscillator 30 according to the embodiment, the first RTD 1 and the second RTD 2 may be configured such that the thicknesses of the first spacer layers SP11 and SP21 are different from each other.

  In the RTD oscillator 30 according to the embodiment, the first RTD1 and the second RTD2 may be configured such that the thicknesses of the second spacer layers SP12 and SP22 are different from each other.

  In the first RTD1 and the second RTD2 applicable to the RTD oscillator 30 according to the embodiment, the quantum well structures QW1 and QW2 that exhibit the resonant tunneling effect, and the spacer layers SP11 and By making the thicknesses of SP12 and SP21 / SP22 different, the current and voltage characteristics are asymmetric on the plus side and the minus side. This is because the number of carriers injected into the quantum well layers 95 of the quantum well structures QW1 and QW2 varies depending on the bias direction depending on the thicknesses of the spacer layers SP21 and SP22 and SP21 and SP22. The introduction of asymmetry by making the thicknesses of the spacer layers SP21 and SP22 and SP21 and SP22 different makes it possible to realize the RTD oscillator 30 according to the embodiment.

  In the first RTD 1 applicable to the RTD oscillator 30 according to the embodiment, the thickness of the spacer layers SP11 and SP12 is asymmetrical, for example, 20 nm and 2 nm, so that the asymmetrical characteristic as shown in FIG. 9 or FIG. Current-voltage characteristics can be obtained.

  Similarly, in the second RTD 2 applicable to the RTD type oscillator 30 according to the embodiment, the thickness of the spacer layers SP21 and SP22 is asymmetrical, for example, 2 nm and 20 nm, so that the asymmetry as shown in FIG. Current-voltage characteristics can be obtained.

  Here, the thickness of each layer is as follows, for example.

The thicknesses of the n ⁺ -type GaInAs layers 91a, 91b, and 91c are, for example, about 400 nm, 15 nm, and 8 nm, respectively. The thicknesses of the n-type GaInAs layers 92a and 92b are substantially equal, for example, about 25 nm. The thicknesses of the undoped GaInAs layers 93a and 93b are, for example, thicknesses that enable the above-described asymmetry to be about 2 nm · 20 nm. The thicknesses of the AlAs layers 94a and 94b are equal, for example, about 1.1 nm. The thickness of the GaInAs layer 95 is about 4.5 nm, for example.

8A and 8B, a SiO ₂ film, a Si ₃ N ₄ film, a SiON film, a HfO ₂ film, an Al ₂ O ₃ film, or the like is formed on the side wall portion of the stacked structure shown in FIGS. An insulating film made of a multilayer film can also be deposited. The insulating layer can be formed by a chemical vapor deposition (CVD) method, a sputtering method, or the like.

(Modification)
A schematic cross-sectional structure of a modified example of the first RTD 1 applicable to the RTD type oscillator 30 according to the embodiment is expressed as shown in FIG. In FIG. 18, the formation of the GaInAs layer 91c doped with a high concentration of n-type impurities disposed on the GaInAs layer 91b is omitted. When the contact between the first electrode 4 and the GaInAs layer 91b does not matter, the formation of the GaInAs layer 91c may be omitted as shown in FIG. Note that the structure of this modified example is also applicable to the second RTD 2.

(Antenna structure)
As the resonant tunnel diode type oscillator according to the present embodiment, any antenna that can be integrated in a plane, such as a slot antenna, a bow tie antenna, a patch antenna, and a Yagi-Uda antenna, can be applied.

  A schematic planar pattern configuration of the RTD oscillator according to the embodiment including a dipole antenna is expressed as shown in FIG.

  As shown in FIG. 20, the RTD oscillator 30 according to the embodiment may include dipole antennas 20D and 40D connected to the first cathode K1 and the second cathode K2.

  Further, as shown in FIG. 20, the RTD oscillator 30 according to the embodiment includes feed lines 20F and 40F connected to the dipole antennas 20D and 40D, and pad electrodes 20P and 40P connected to the feed lines 20F and 40F. May be provided.

  The RTD oscillator 30 according to the embodiment may include an MIM reflector (not shown) connected between the first cathode K1 and the second cathode K2.

  A schematic planar pattern configuration of the RTD oscillator 30 according to the embodiment including the tapered slot antennas 20T and 40T is expressed as shown in FIG.

  As shown in FIG. 21, the RTD oscillator 30 according to the embodiment may include tapered slot antennas 20T and 40T connected to the first cathode K1 and the second cathode K2.

  Further, in the RTD type oscillator 30 according to the embodiment, the tapered slot antennas 20T and 40T are conductive conductors disposed between the first cathode K1 and the second cathode K2 facing each other on the semiconductor substrate 1, as shown in FIG. A waveguide 70 and a horn opening 80 disposed between the first cathode K1 and the second cathode K2 facing the semiconductor substrate 1 adjacent to the waveguide 70 may be provided. Here, in the RTD oscillator 30 according to the embodiment, the first cathode K1 and the second cathode K2 are connected to the second anode A2 and the first anode A1.

  The RTD oscillator 30 according to the embodiment including the tapered slot antennas 20T and 40T is connected between the first cathode K1 and the second cathode K2, and the resonator 60 of the resonant tunnel diode oscillator 30 is disposed in the arrangement. On the other hand, the MIM reflector 50 disposed on the opposite side of the waveguide 70 may be provided.

  As shown in FIG. 21, the RTD type oscillator according to the embodiment includes a semiconductor substrate 1, a second antenna electrode 20T disposed on the semiconductor substrate 1, and an insulating layer for the second antenna electrode 20T. And the first antenna electrode 40T disposed on the semiconductor substrate 1 so as to face the second antenna electrode 20T, and between the first antenna electrode 40T and the second antenna electrode 20T with an insulating layer interposed therebetween. The MIM reflector 50 formed on the semiconductor substrate 1, the resonator 60 disposed between the first antenna electrode 40 T and the second antenna electrode 20 T facing the semiconductor substrate 1 adjacent to the MIM reflector 50, and the resonator 60 The first RTD 1 and the second RTD 2 arranged in the substantially central portion of the first antenna electrode 40 T and the second antenna electrode 20 adjacent to the resonator 60 and facing the semiconductor substrate 1. A waveguide 70 disposed in between, and a horn opening 80 disposed between the first antenna electrode 40T and the second antenna electrode 20T facing the semiconductor substrate 1 adjacent to the waveguide 70. .

The length L1 of the resonator 60 is, for example, about 30 μm or less. Further, the width of the recesses 5 and 6 (the distance between the first antenna electrode 40T and the second antenna electrode 20T) is, for example, about 4 μm. The dimensions of the first RTD 1 and the second RTD ² are, for example, about 1.4 μm ² . However, the sizes of the first RTD 1 and the second RTD ² are not limited to this value, and may be, for example, about 5.3 μm ² or less. The size of each part of the resonator 60 is not limited to the above dimensions, and is appropriately set in design according to the frequency of the terahertz electromagnetic wave.

  Further, as shown in FIG. 21, the first antenna electrode 40 </ b> T in the portion where the resonator 60 is formed compared to the distance between the first antenna electrode 40 </ b> T and the second antenna electrode 20 </ b> T in the waveguide 70. The interval between the second antenna electrodes 20T is narrow.

  The MIM reflector 50 is disposed at the closing portion on the opposite side of the opening of the resonator 60. Due to the laminated structure of the metal / insulator / metal MIM reflector 50, the first antenna electrode 40T and the second antenna electrode 20T are short-circuited in high frequency. Further, the MIM reflector 50 has an effect that it can reflect a high frequency while being open in terms of direct current.

  The first antenna electrode 40T and the second antenna electrode 20T each have, for example, a metal laminated structure of Au / Pd / Ti or Au / Ti, and the Ti layer is a semiconductor substrate 1 made of a semi-insulating InP substrate. It is a buffer layer for making a contact state favorable. The thickness of each part of the first antenna electrode 40T and the second antenna electrode 20T is, for example, about several hundred nm, and a flattened laminated structure as shown in FIG. 21 is obtained as a whole. Note that both the first antenna electrode 40T and the second antenna electrode 20T can be formed by a vacuum evaporation method, a sputtering method, or the like.

  Note that the Ti layer forming the surface layer of the first antenna electrode 40T is desirably removed in order to reduce contact resistance when the extraction electrode is formed by a bonding wire (not shown). Similarly, the Ti layer forming the surface layer of the second antenna electrode 20T is desirably removed in order to reduce the contact resistance when the extraction electrode is formed by a bonding wire (not shown).

The insulating layer of the MIM reflector 50 can be formed of, for example, a SiO ₂ film. In addition, a Si ₃ N ₄ film, a SiON film, a HfO ₂ film, an Al ₂ O ₃ film, or the like can be applied. The thickness of the insulating layer can be determined in consideration of the geometric plane size of the MIM reflector 50 and the required capacitor value in terms of circuit characteristics, and is, for example, about several tens nm to several hundreds nm. The insulating layer can be formed by a CVD method or a sputtering method.

(Production method)
The manufacturing method of the RTD type oscillator according to the embodiment includes a step of forming the first semiconductor layer 91a on the semiconductor substrate 1, and patterning the first semiconductor layer 91a to form the first cathode region and the second cathode region. Forming the first RTD1 in which the first cathode K1 is connected to the first cathode region and the first anode A1 is connected to the second cathode region, and the second cathode K2 is connected to the second cathode region. is to form a step of forming a first 2RTD2 the second anode A2 is connected to the first cathode region, a first cathode electrode 2 ₁ are commonly connected to the second anode electrode 4 ₂ to the first cathode region on and a step, and forming a second cathode electrode 2 ₂ which are commonly connected to the first anode electrode 4 ₁ on the second cathode region.

  The steps of forming the first RTD1 and the second RTD2 include the steps of forming the first spacer layers SP11 (93a) and SP21 (93a) on the first semiconductor layer 91a, and the first spacer layers SP11 (93a) and SP21. (93a) a step of forming the first tunnel barrier layers 94a, 94a, a step of forming the quantum well layers 95, 95 on the first tunnel barrier layers 94a, 94a, and a step of forming the first tunnel barrier layers 94a, 94a on the quantum well layers 95, 95 A step of forming two tunnel barrier layers 94b and 94b, and a second spacer layer SP12 (93b) of a thickness different from that of the first spacer layers SP11 (93a) and SP21 (93a) on the second tunnel barrier layers 94b and 94b. Forming SP22 (93b).

  The structures of the first RTD1 and the second RTD2 applicable to the RTD type oscillator 30 according to the embodiment can be formed in the same manufacturing process.

  19 is a schematic cross-sectional structure of the first RTD 1 and the second RTD 2 manufactured in the same process, and a schematic cross-sectional structure of the first RTD 1 and the second RTD 2 part of the RTD type oscillator 30 according to the integrated embodiment is shown in FIG. Represented as shown. Since the configuration of each layer is the same as that in FIGS. 8A and 8B, a duplicate description is omitted.

  In the RTD oscillator 30 according to the embodiment, the thickness of the spacer layers SP11 and SP12 of the first RTD1 is, for example, 20 nm · 2 nm, and the thickness of the spacer layers SP21 and SP22 of the second RTD2 is, for example, 2 nm · 20 nm For this, it is necessary to separately use a process of epitaxially growing in common, a process of epitaxially growing only the first RTD1 side by masking the second RTD2 side, and a process of epitaxially growing only the second RTD2 side by masking the first RTD1 side. There is. The manufacturing processes of the other layers can be formed by a common process of the first RTD1 and the second RTD2.

  For example, in the first RTD 1 and the second RTD 2 manufactured in the same process, the oscillation frequency observed at room temperature is about 300 GHz.

  According to the manufacturing method of the RTD type oscillator according to the embodiment, parasitic oscillation of the RTD and the external circuit can be suppressed without the metal wiring, and the manufacturing process of the metal wiring becomes unnecessary, thereby reducing the number of process steps. be able to.

  In the RTD oscillator according to the present embodiment, an example in which the first tunnel barrier layer / quantum well layer / second tunnel barrier layer has a configuration of AlAs / InAlAs / AlAs is shown. It is not limited to. For example, the first tunnel barrier layer / quantum well layer / second tunnel barrier layer may have an AlGaAs / GaAs / AlGaAs configuration. In addition, the first tunnel barrier layer / quantum well layer / second tunnel barrier layer may have an AlGaN / GaN / AlGaN configuration. Further, the first tunnel barrier layer / quantum well layer / second tunnel barrier layer may have an SiGe / Si / SiGe configuration.

  In any configuration, the first spacer layer is disposed in contact with the first tunnel barrier layer, the second spacer layer is disposed in contact with the second tunnel barrier layer, and the thicknesses of the first spacer layer and the second spacer layer are set. The first RTD1 and the second RTD2 having such asymmetric current-voltage characteristics are connected in antiparallel, and one of the first RTD1 and the second RTD2 is negative. Similarly, when biased to a resistive oscillation state, the other is biased to a resistive state.

  According to the present embodiment, it is possible to provide a resonant tunneling diode type oscillator having a high reliability without suppressing the parasitic oscillation of the RTD and the external circuit without using metal wiring, and without increasing the number of process steps, and a method for manufacturing the same. .

[Other embodiments]
As described above, the RTD type oscillator according to the embodiment and the modification example thereof has been described. However, the description and drawings constituting a part of this disclosure are exemplary, and the embodiment is limited. Should not be understood. From this disclosure, various alternative embodiments, examples and operational techniques will be apparent to those skilled in the art.

  As described above, this embodiment includes various embodiments not described here.

  The RTD oscillator of this embodiment can be applied to a terahertz oscillator, a high-frequency resonance circuit, a signal amplifier, etc. on a device basis, and has a large capacity such as a terahertz wave imaging device, a sensing device, a high-speed wireless communication device, etc. In addition to communication and information processing, it can be applied to a wide range of fields such as measurement and security in various fields such as physical properties, astronomy, and living things.

1 ... Semiconductor substrate 2, 2 ₁ , 2 ₂ ... Second electrode (cathode electrode)
4, 4 ₁ , 4 ₂ ...... First electrode (anode electrode)
5, 6 ... Recess 9 ... Interlayer insulating film 20D, 40D ... Dipole antenna (antenna electrode)
20T, 40T ... Tapered slot antenna (antenna electrode)
20F, 40F ... feed lines 20P, 40P ... pad electrode 30 ... RTD oscillator 50 ... MIM reflector 60 ... resonator 70 ... waveguide 80 ... horn opening 90 ... active element 91a ... first semiconductor layer (GaInAs layer)
94a, 94b ... tunnel barrier layer 95 ... quantum well layer 114 ... resistance elements A, A1, A2 ... anodes K, K1, K2 ... cathodes SP11, SP12. SP21, SP22 ... spacer layers QW1, QW2 ... quantum well structure

Claims (19)

A first resonant tunneling diode comprising a first anode and a first cathode;
A second resonant tunneling diode comprising a second anode and a second cathode;
The first resonant tunneling diode and the second resonant tunneling diode are connected in antiparallel, and one of the first resonant tunneling diode and the second resonant tunneling diode is biased to a negative resistance oscillation state. A resonant tunneling diode type oscillator, wherein the other is biased in a resistance state.   2. The resonant tunneling diode oscillator according to claim 1, wherein the first resonant tunneling diode and the second resonant tunneling diode have asymmetric current-voltage characteristics.   The resonant tunneling diode oscillator according to claim 1 or 2, wherein the first anode and the second cathode are connected, and the first cathode and the second anode are connected. Comprising a semiconductor substrate,
4. The resonant tunneling diode oscillator according to claim 1, wherein the first resonant tunneling diode and the second resonant tunneling diode are disposed on the semiconductor substrate. 5.   5. The resonant tunneling diode oscillator according to claim 4, wherein the first resonant tunneling diode and the second resonant tunneling diode are integrated on the semiconductor substrate. A semiconductor substrate;
A first semiconductor layer disposed on the semiconductor substrate;
A first cathode region and a second cathode region formed by patterning the first semiconductor layer;
A first resonant tunneling diode having a first cathode connected to the first cathode region and a first anode connected to the second cathode region;
A second resonant tunneling diode having a second cathode connected to the second cathode region and a second anode connected to the first cathode region, wherein the first resonant tunneling diode and the second resonant tunneling diode are A resonant tunneling diode type oscillator characterized in that when one of them is biased to a negative resistance oscillation state, the other is biased to a resistance state.   The resonant tunneling diode oscillator according to claim 6, wherein the first resonant tunneling diode and the second resonant tunneling diode have asymmetric current-voltage characteristics. A first cathode electrode disposed on the first cathode region;
A second cathode electrode disposed on the second cathode region;
A first anode electrode connected to the first anode;
A second anode electrode connected to the second anode,
8. The resonant tunneling diode type oscillator according to claim 7, wherein the first cathode electrode is commonly connected to the second anode electrode, and the second cathode electrode is commonly connected to the first anode electrode. . The first resonant tunneling diode and the second resonant tunneling diode are:
A quantum well layer;
A second spacer layer disposed on the anode side through a tunnel barrier layer sandwiching the quantum well layer, and a first spacer layer disposed on the cathode side,
The resonant tunnel diode oscillator according to claim 8, wherein the first spacer layer and the second spacer layer have different thicknesses.   9. The resonant tunneling diode oscillator according to claim 8, wherein the first resonant tunneling diode and the second resonant tunneling diode have different thicknesses of the first spacer layers.   9. The resonant tunneling diode oscillator according to claim 8, wherein the first resonant tunneling diode and the second resonant tunneling diode have different thicknesses of the second spacer layers.   The resonant tunneling diode oscillator according to any one of claims 8 to 11, further comprising a dipole antenna connected to the first cathode and the second cathode.   The resonant tunnel diode type oscillator according to claim 12, further comprising: a feed line connected to the dipole antenna; and a pad electrode connected to the feed line.   12. The resonant tunneling diode oscillator according to claim 8, further comprising a tapered slot antenna connected to the first cathode and the second cathode.   The resonant tunneling diode oscillator according to any one of claims 8 to 14, further comprising an MIM reflector connected between the first cathode and the second cathode. The tapered slot antenna includes a waveguide disposed between the first cathode electrode and the second cathode electrode facing each other on the semiconductor substrate;
The resonant tunnel according to claim 14, further comprising: a horn opening disposed between the first cathode electrode and the second cathode electrode facing the semiconductor substrate adjacent to the waveguide. Diode type oscillator.   The MIM reflector connected between the said 1st cathode and the said 2nd cathode, and is arrange | positioned on the opposite side to the said waveguide with respect to arrangement | positioning of the said resonant tunnel diode type | mold oscillator, It is characterized by the above-mentioned. The described resonant tunneling diode type oscillator. Forming a first semiconductor layer on a semiconductor substrate;
Patterning the first semiconductor layer to form a first cathode region and a second cathode region;
Forming a first resonant tunneling diode having a first cathode connected to the first cathode region and a first anode connected to the second cathode region;
Forming a second resonant tunneling diode having a second cathode connected to the second cathode region and a second anode connected to the first cathode region;
Forming a first cathode electrode commonly connected to a second anode electrode on the first cathode region;
Forming a second cathode electrode commonly connected to the first anode electrode on the second cathode region. A method for manufacturing a resonant tunneling diode type oscillator, comprising: Forming the first resonant tunneling diode and the second resonant tunneling diode;
Forming a first spacer layer on the first semiconductor layer;
Forming a first tunnel barrier layer on the first spacer layer;
Forming a quantum well layer on the first tunnel barrier layer;
Forming a second tunnel barrier layer on the quantum well layer;
19. The method of manufacturing a resonant tunneling diode oscillator according to claim 18, further comprising: forming a second spacer layer having a thickness different from that of the first spacer layer on the second tunnel barrier layer.