Superconductive switch circuit and quantum arithmetic circuit having the same

Abstract

To provide a superconductive switch circuit which is small and has a variable frequency.SOLUTION: The superconductive switch circuit includes: a first transmission line between a first input port and a first output port, the first transmission line having N (N is an integer of at least 1) serially connected first unit cells including a first inductor and a first capacitor; a second transmission line between a second input port and a second output port, the second transmission line having N serially connected second unit cells including a second inductor and a second capacitor; a variable coupling between the first transmission line and the second transmission line, the variable coupling adjusting the coupling state between the first transmission line and the second transmission line; and a control line for controlling the coupling strength of the variable coupling. The first unit cell, the second unit cell, and the variable coupling are formed of lumped constant elements.SELECTED DRAWING: Figure 2B

Description

Technical Field

[0001] The present invention relates to a superconducting switch circuit and a quantum computing circuit using the same, and particularly to an on-chip type superconducting switch circuit and a quantum computing circuit.

Background Art

[0002] Superconducting switch circuits are applied to quantum computing and devices operating at extremely low temperatures. Semiconductor-based mechanical pulse latch switches and thermoelectric switches release a large amount of heat and thus cannot correspond to temperatures on the order of millikelvins. PIN diode switches and field effect transistors also cannot be used for low-temperature operation at the single-photon level of power. On the other hand, a superconducting quantum interference device (SQUID) having a Josephson junction is used as a low-power consumption and low-loss inductance element in a switch circuit for low-temperature operation. A single-pole single-throw (SPST) switch with one circuit and one contact is limited to use in a single channel, but with a single-pole multi-throw (SPMT) or multi-pole multi-throw (MPMT) switch, the propagation of an electromagnetic field can be controlled more flexibly.

[0003] Pairs of fixed-frequency hybrid beam splitters and double-pole double-throw (DPDT) superconducting switches using a SQUID chain have been proposed (see, for example, Non-Patent Document 1). A filter-based quantum signal router using a SQUID as a variable inductor (see, for example, Patent Document 1) and a superconducting switch system (see, for example, Patent Documents 1 and 2) are also known.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Non-Patent Literature

[0005]

Non-Patent Literature 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] The superconducting switch disclosed in Non-Patent Literature 1 has, in addition to a pair of hybrid beam splitters and a SQUID chain, a transmission line type resonator and two direct current (DC) control lines on the chip, and requires many on-chip components. The size of the switch circuit is 5 mm × 4 mm, which makes it difficult to widely use in an on-chip multi-quantum bit processor. Also, the operating frequency of the switch circuit is fixed to the frequency determined by the beam splitter. Although the superconducting routers and superconducting switches disclosed in Patent Literatures 1 and 2 can adjust the filter frequency to some extent by changing the magnetic flux passing through the SQUID to change the inductance, the operating frequency of the switch circuit itself is fixed.

[0007] One object of the present invention is to provide a small-sized and frequency-variable superconducting switch circuit.

Means for Solving the Problems

[0008] In an embodiment, the superconducting switch circuit includes a first transmission line in which N (N is an integer of 1 or more) first unit cells each including a first inductor and a first capacitor are connected in series between a first input port and a first output port, A second transmission line formed by connecting N second unit cells including a second inductor and a second capacitor in series between a second input port and a second output port, a variable coupling provided between the first transmission line and the second transmission line for adjusting a coupling state between the first transmission line and the second transmission line, a control line for controlling a coupling strength of the variable coupling, and the first unit cell, the second unit cell, and the variable coupling are formed of lumped elements.

Advantages of the Invention

[0009] With the above configuration, a small-sized and frequency-variable superconducting switch circuit is realized.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2A

Figure 2B

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12A

Figure 12B

Figure 12C

Figure 12D

Figure 12E

Figure 12F

Figure 13

Figure 14

Figure 15

[0011] A superconducting switch circuit according to an embodiment will be described below with reference to the drawings. The following description is intended to embody the technical concept of the present invention, and unless otherwise specified, does not limit the present invention to the following configuration or method. In each drawing, components having the same function are given the same reference numerals, and redundant explanations may be avoided. The size and positional relationship of components shown in each drawing may be exaggerated for clarity of explanation.

[0012] <Compact and frequency-tunable superconducting switch circuit> FIG. 1 is a diagram comparing a superconducting switch circuit 10 of the embodiment with a known superconducting switch 1000. FIG. 1A is a micrograph of the superconducting switch circuit 10 of the embodiment formed on a chip as a lumped-constant circuit 100, and FIG. 1B is a micrograph of the superconducting switch 1000 disclosed in Non-Patent Document 1. The known superconducting switch 1000 has two fixed-frequency hybrid beam splitters H, two resonators R, and two SQUID arrays S connected to two DC control lines, and is 4 mm × 5 mm in size. When a signal input to port (1) matches the resonant frequency of the circuit, a signal is output to port (3). When the frequency of the input signal deviates from the resonant frequency, resonator R acts as a reflector, and the signal is output to port (2). This superconducting switch 1000 operates at a fixed frequency determined by the hybrid beam splitter H, and its operating bandwidth is approximately 150 MHz at most.

[0013] In contrast, the superconducting switch circuit 10 of the embodiment is formed as a lumped constant circuit 100 using elements that are sufficiently smaller than the wavelength of the frequency being handled, and is 0.45 mm × 0.1 mm in size. The superconducting switch circuit 10 is a DPDT switch having two input ports Pin1 and Pin2 and two output ports Pout1 and Pout2. The superconducting switch circuit 10 of the embodiment realizes a DPDT switch with a size that is only 1 / 450 of the size of the known superconducting switch 1000, and is widely applicable to on-chip multi-qubit processors.

[0014] As will be described later, the superconducting switch circuit 10 of this embodiment has a variable operating frequency ranging from 4.5 GHz to 7.5 GHz, and an expanded bandwidth up to 800 MHz. This is achieved by routing signals by controlling the coupling state between two transmission lines formed by lumped elements. As will be described later, this is because the coupling state between the two transmission lines is determined by the phase of the signal envelope, regardless of the frequency of the input signal. The specific configuration and operation theory of the superconducting switch circuit 10 will be described with reference to FIGS. 2A and 2B.

[0015] <Configuration and Operating Theory of Superconducting Switch Circuit> FIG. 2A is a diagram showing a theoretical model of a superconducting switch circuit according to an embodiment. The superconducting switch circuit 10M of the theoretical model includes a first transmission line 11, a second transmission line 12, and a variable coupling Lcoup that adjusts the coupling state between the first transmission line 11 and the second transmission line 12. The coupling strength or coupling phase of the variable coupling Lcoup is controlled by a control line (not shown).

[0016] The first transmission line 11 includes N (N is an integer of 1 or more) first unit cells 101-1 to 101-N connected in series between a first input port Pin1 and a first output port Pout1. Each of the first unit cells 101-1 to 101-N (which may be simply referred to as "first unit cell 101") has an inductor represented by the symbol L and a capacitor represented by the symbol C.

[0017] The second transmission line 12 includes N (N is an integer of 1 or more) second unit cells 102-1 to 102-N connected in series between a second input port Pin2 and a second output port Pout2. Each of the second unit cells 102-1 to 102-N (which may be simply referred to as "second unit cell 102") has an inductor represented by the symbol L and a capacitor represented by the symbol C. The inductance of the first unit cell 101 and the second unit cell 102 may be a geometric inductance such as wiring, a kinetic inductance realized by a thin film, or a Josephson inductance by a Josephson junction. The inductances L of the first unit cell 101 and the second unit cell 102 provided at corresponding positions along the transmission line may be the same or different.

[0018] The capacitors of the first unit cell 101 and the second unit cell 102 are interdigital or parallel-plate thin-film capacitors. From the viewpoint of forming the capacitors as lumped constant elements, parallel-plate thin-film capacitors are preferable. The capacitances C of the first unit cell 101 and the second unit cell 102, which are provided at corresponding positions along the transmission line, may be the same or different. The variable coupling Lcoup is realized by utilizing variable capacitance, kinetic inductance, or a SQUID chain. The variable coupling Lcoup determines the mutual inductance between the first transmission line 11 and the second transmission line 12.

[0019] 2B is a schematic diagram of a superconducting switch circuit 10 in which the model of FIG. 2A is realized by a lumped constant circuit 100. The superconducting switch circuit 10 includes a first transmission line 11, a second transmission line 12, a variable coupling provided between the first transmission line 11 and the second transmission line 12, and a control line CL that controls the coupling strength of the variable coupling. The variable coupling is realized by N SQUIDs 15-1 to 15-N (which may be collectively referred to as "SQUIDs 15" as appropriate) connected in series.

[0020] The first transmission line 11 includes N (N is an integer greater than or equal to 1) first unit cells 101-1 to 101-N connected in series between a first input port Pin1 and a first output port Pout1. The second transmission line 12 includes N (N is an integer greater than or equal to 1) second unit cells 102-1 to 102-N connected in series between a second input port Pin2 and a second output port Pout2. The first unit cells 101-1 to 101-N, the second unit cells 102-1 to 102-N, and the SQUID 15 forming the variable coupling are formed of lumped elements.

[0021] In the example of FIG. 2B, each of the first unit cells 101-1 to 101-N includes a first Josephson junction 111 as a first inductor and a first thin-film capacitor 112 as a first capacitor. Each of the second unit cells 102-1 to 102N includes a second Josephson junction 121 as a second inductor and a second thin-film capacitor 122 as a second capacitor. A Josephson junction has a configuration in which a thin insulating film is sandwiched between two superconductors, and is easily fabricated as a lumped element on a chip by film formation such as sputtering and vacuum evaporation, and etching.

[0022] The first thin-film capacitor 112 and the second thin-film capacitor 122 are easily fabricated on the chip by film formation such as sputtering and vacuum evaporation, and etching, for example, as parallel-plate capacitors. A parallel-plate capacitor has a stacked structure in which a dielectric layer is sandwiched between a lower electrode and an upper electrode. The dielectric layer is formed of, for example, a metal oxide. When the lower electrode is formed of a metal that can be naturally oxidized, a natural oxide film may be used as part of the dielectric layer.

[0023] SQUID 15 has a configuration in which two Josephson junctions are coupled by a superconducting loop, and is fabricated as a lumped element on a chip by film formation such as sputtering and vacuum evaporation, and etching. A direct current is applied as a control current to the control line CL. The current-voltage characteristics of SQUID 15 are determined by the magnetic flux generated by the control current passing through the SQUID loop.

[0024] By controlling the direct current flowing through the control line CL, the magnetic flux passing through the SQUID 15 is controlled, and the mutual inductance between the first transmission line 11 and the second transmission line 12 is controlled. By adjusting the mutual inductance between the first transmission line 11 and the second transmission line 12, the coupling strength between the first transmission line 11 and the second transmission line is controlled. This coupling strength is determined by the phase of the envelope of the input signal, regardless of the frequency of the input signal. Therefore, a signal input from the input port Pin1 or the second input port Pin2 can be output from one or both of the first output port Pout1 and the second output port Pout2 over a wide frequency range.

[0025] FIG. 3 is a diagram showing the flow of potential φ and current I at a specific node of the superconducting switch circuit 10. The n-th first unit cell 101 including the first Josephson junction 111 and the first thin-film capacitor 112, the n-th second unit cell 102 including the second Josephson junction 121 and the second thin-film capacitor 122, and the n-th SQUID 15 between the first unit cell 101 and the second unit cell 102 constitute the unit circuit 110. The superconducting switch circuit 10 is formed by connecting N of these unit circuits 110 in series.

[0026] Let the current flowing into the n-th first unit cell 101 be Ia. The current Ia enters the node 104 through the first Josephson junction 111. The first thin-film capacitor 112 is connected to the node 104 on the output side of the first Josephson junction 111. Let the voltage at the node 104 be φa. Let the current flowing through the second unit cell 102 be Ic. The current Ic enters the node 105 through the second Josephson junction 121. The second thin-film capacitor 122 is connected to the node 105 on the output side of the second Josephson junction 121. Let the voltage at the node 105 be φc.

[0027] A SQUID 15 is connected to a node 103 between a first thin-film capacitor 112 and a second thin-film capacitor 122. Let the voltage at the node 103 be φb. The capacitance between the nodes 104 and 103 is determined by the voltage difference (φa - φb) between the nodes 104 and 103 and the charge of the current Ia. The capacitance between the nodes 105 and 103 is determined by the voltage difference (φc - φb) between the nodes 105 and 103 and the charge of the current Ic.

[0028] The SQUID 15 connected to the node 103 has two Josephson junctions 151 and 152 coupled by a superconducting loop. A control line CL for controlling the magnetic flux penetrating the loop of the SQUID 15 is commonly used for N series-connected SQUID 15s, and the coupling state between the first transmission line 11 and the second transmission line 12 can be controlled by only one control line CL.

[0029] <Signal Routing by Variable Coupling> FIG. 4 is a schematic diagram showing signal routing by variable coupling of the superconducting switch circuit 10 of the embodiment. When the coupling between the first transmission line 11 and the second transmission line 12 is weak, no interaction occurs between the first transmission line 11 and the second transmission line 12. By increasing the coupling, a part of the electromagnetic field of the signal input to the first input port Pin1 is coupled to the second transmission line 12 due to the mutual inductance between the first transmission line 11 and the second transmission line 12. By controlling the magnetic flux penetrating the SQUID by the control current flowing through the control line CL, the coupling strength can be controlled, and the coupling state between the first transmission line 11 and the second transmission line 12 can be controlled.

[0030] FIG. 5 is a diagram showing different switch states according to the coupling state. (A) of FIG. 5 shows the routing when the coupling between the first transmission line extending between the first input port Pin1 and the first output port Pout1 and the second transmission line extending between the second input port Pin2 and the second output port Pout2 is maximum. Let the first input port Pin1 be port number 1, the first output port Pout1 be port number 2, and the second output port Pout2 be port number 3.

[0031] A signal input to the first input port Pin1 is routed to the second output port Pout2 with the maximum coupling strength. The S parameter representing the propagation from the first input port Pin1 to the first output port Pout1 is S 21 OFF , the S parameter indicating the propagation from the first input port Pin1 to the second output port Pout2 is S 31 ON At this time, the coupling phase χN in the Nth unit circuit 110 is π / 2, where χ is the envelope of the input high-frequency signal.

[0032] FIG. 5B shows routing when the coupling between the first transmission line and the second transmission line is intermediate. Half of the signal input to the first input port Pin1 propagates directly to the first output port Pout1 without being coupled to the second transmission line. The remaining half of the input signal is coupled to the second transmission line and routed to the second output port Pout2. The S-parameter that indicates the propagation from the first input port Pin1 to the first output port Pout1 is S 21 HALF , the S parameter showing the propagation from the first input port Pin1 to the second output port Pout2 is S 31 HALF At this time, the coupling phase χN in the Nth unit circuit 110 is 3π / 4.

[0033] FIG. 5C shows routing when the coupling between the first transmission line and the second transmission line is minimum. A signal input to the first input port Pin1 propagates through the first transmission line and is routed to the first output port Pout1 without coupling to the second transmission line. The S parameter indicating the propagation from the first input port Pin1 to the first output port Pout1 is S 21 ON , the S parameter indicating the propagation from the first input port Pin1 to the second output port Pout2 is S 31 OFF At this time, the coupling phase χN in the Nth unit circuit 110 is π.

[0034] The basis for the routing in Figure 5 will be explained. For a given operating angular frequency ω = 2πf, the coupling phase χN that achieves a desired branching ratio is determined. As shown in the model diagram of Figure 2A and the unit circuit 110 of Figure 3, in the nth unit circuit 110, the current flowing through the first transmission line 11 is Ia, the current flowing through the second transmission line 12 is Ic, and the variable coupling between the first transmission line 11 and the second transmission line 12 is Lcoup. For simplicity, the inductance L of each first unit cell 101 on the first transmission line 11 is assumed to be the same as the inductance L of each second unit cell 102 on the second transmission line 12. Furthermore, the capacitance C of each first unit cell 101 is assumed to be the same as the capacitance C of each second unit cell 102. The wave equation describing the nth unit circuit 110 is expressed by Equation (1).

[0035]

number

[0036] The Fourier transform of the current is expressed by equation (2).

[0037]

number

[0038]

number

[0039] If we set the determinant of equation (3) to 0, [(1+Lcoup / L)ω 2 -k 2 / LC] 2 -[(Lcoup / L)ω 2] 2 =0 From this, we obtain two modes with the dispersion relation of equation (4) for the wave vector.

[0040]

number

[0041] When only the first input port Pin1 is excited, the solution to equation (1) is given by equation (5).

[0042]

number

[0043] S parameter S of the first output port Pout1 21 and the S-parameter S of the second output port Pout2 31 When the high frequency component K is time-averaged with respect to the power of the oscillator, only the envelope χ of the oscillatory current remains.

[0044]

number

[0045] The phase of the envelope signal determines how much of the signal is transferred to other channels. As shown in Figure 5(A), for a signal completely coupled to the second transmission line 12, Ia = 0, Ic = Imax in the Nth unit circuit 110, and χN, which satisfies cos(χN) = 0 and sin(χN) = ±1, is the solution. Therefore, the coupling phase χN is π / 2 as shown in equation (7).

[0046]

number

[0047] As shown in Figure 5(B), when the signal is branched equally between the first output port Pout1 and the second output port Pout2, the coupling phase χN is 3π / 4. The signal coupled half to the first output port Pout1 and half to the second output port Pout2 has Ia = Imax / 2 and Ic = Imax / 2 in the Nth unit circuit 110, and the solution is χN that satisfies cos(χN) = ±1 / √2 and sin(χN) = 1 / √2. As shown in Figure 5(C), when the signal is routed only to the first output port Pout1, the coupling phase χN is π.

[0048] Thus, for switching or routing, the phase (χN) of the envelope signal at the output end is important, and the frequency (oscillation frequency) of the signal passing through the transmission line is almost irrelevant. Regardless of the frequency of the input signal, the superconducting switch circuit 10 can route the input signal at the desired rate and in the desired direction by controlling the magnetic flux passing through the SQUID 15 to change the coupling state.

[0049] Equation (7) is satisfied and the impedance Z is Z=(L / C) 1 / 2 The device parameters are selected taking into consideration manufacturing conditions so as to satisfy the condition of Ω = 50Ω. The inductance L on the transmission line realized by the first Josephson junction 111 and the second Josephson junction 121 and the coupling inductance Lcoup realized by the SQUID 15 include the minute internal capacitance of the Josephson junction. The Josephson junction capacitance C JJ and the capacitance of the SQUID, C SQUID Considering this, the inductance L of the transmission line * and coupling inductance Lcoup * is expressed by equation (8). 1 / L * =1 / L-ω 2 C JJ 1 / Lcoup * =1 / Lcoup-ω 2 C SQUID (8)

[0050] The inductance L of each Josephson junction on the transmission line is Φ0 / (2πI cri ) The coupling inductance Lcoup of each SQUID 15 is approximated as Φ0 / [4πI cri (cos(πΦ / Φ0)) 1 / 2 ], where I cri is the critical current, and Φ is the quantum. When the magnetic flux passing through the superconducting loop is confined inside the loop, the magnetic flux in the loop takes on discrete values. The magnetic flux quantum Φ is the smallest unit of quantized magnetic flux. The signal routing in FIG. 5 based on the control of the coupling state is performed by controlling the current Ia flowing through the first Josephson junction 111, the current Ic flowing through the second Josephson junction 121, and the current flowing through the SQUID 15 to generate a critical current I cri This constraint is effective as long as the operating power of the superconducting switch circuit 10 is smaller than the critical current I cri Although this may be limited by , the upper limit of operating power can be increased by successively increasing the critical current of each Josephson junction and increasing the number N of unit circuits 110 while maintaining the total inductance of the SQUID chain.

[0051] <Fabrication of superconducting switch circuits> Fig. 6 is a micrograph of a superconducting switch circuit 10 fabricated on a chip. The superconducting switch circuit is formed as a lumped constant circuit 100. Fig. 7 is an enlarged image of the lumped constant circuit 100 in Fig. 6, and Fig. 8 is an enlarged image of region A1 in Fig. 7.

[0052] 6, the superconducting switch circuit 10 is formed as a lumped constant circuit 100 measuring 80 μm × 420 μm approximately in the center of a 5 mm × 2.5 mm silicon chip. The lumped constant circuit 100 is connected to a first input port Pin1, a second input port Pin2, a first output port Pout1, and a second output port Pout2 by coplanar waveguides with an impedance of 50 Ω. The coplanar waveguides are formed, for example, by sputtering niobium (Nb) into a thin film with a thickness of 50 nm, and then patterning it by reactive ion etching using CF4.

[0053] As shown in FIG. 7, in the lumped constant circuit 100, 24 unit circuits 110 are connected in series. As shown in FIG. 8, each unit circuit 110 includes a first Josephson junction 111 of the first transmission line 11 (see FIG. 3), a first thin film capacitor 112, and a SQUID 15 as a variable coupling. The SQUID 15 has two Josephson junctions 151 and 152. The first thin film capacitor 112 and the SQUID 15 are joined by a bandage 107 to prevent corrosion due to dissimilar metal contact. The second transmission line 12 (see FIG. 3) side of the unit circuit 110 is arranged symmetrically to the first transmission line 11 side with respect to the SQUID chain.

[0054] Each Josephson junction and thin-film capacitor is formed of a three-layer structure of Al / Al2O3 / Al, with each layer being 50 to 80 nm thick. This three-layer structure is formed by electron beam evaporation of aluminum followed by oxidation treatment. After deposition, the film is patterned into the shape of each element using standard photolithography techniques. The bandage 107 is formed by removing the Al and Nb oxides on the chip by argon milling, then depositing and etching a 100 nm thick aluminum film. The area of the first Josephson junction 111 on the transmission line is 1.8 μm 2 , the area of the first thin film capacitor 112 is 22 μm 2 , the area of SQUID15 is 26.4 μm 2 The area of each of the two Josephson junctions 151 and 152 is 12.6 μm 2 The size of the lumped constant circuit 100 in which 24 unit circuits 110 are connected in series is 80 μm×420 μm, as described above.

[0055] <Measurement and Evaluation> The switching characteristics are measured and evaluated using the fabricated superconducting switch circuit 10. The chip of FIG. 6 on which the superconducting switch circuit 10 is formed is attached to the base stage of a dilution refrigerator at a temperature of 12 mK. A high-frequency signal generated by a vector network analyzer (VNA) is attenuated and filtered and then input to the chip, and the output signal is amplified and acquired. A continuous wave generated by the VNA is input to the first input port Pin1, and the raw magnitudes of the complex transmission coefficients |S 21 | and |S 31 | are acquired from the first output port Pout1 and the second output port Pout2 and calibrated by another path within the experimental setup.

[0056] First, the magnetic flux dependence response of the S-parameters is acquired. |S 21 | and |S 31 | are measured as functions of frequency and external magnetic flux. For the purpose of the measurement, an external magnetic flux is generated by a coil instead of a DC control current. When observing the periodic patterns of the measured |S 21 | and |S 31 |, the transmission of the electromagnetic wave propagating through the transmission line behaves periodically with respect to the normalized magnetic flux Φ / Φ0 normalized by the magnetic flux quantum Φ0. The normalized magnetic flux Φ / Φ0 changes greatly near ±0.5 and hardly changes in the range from 0.0 to ±0.4. Therefore, |S 21 | and |S 31 | when Φ / Φ0 is near 0.5 are analyzed.

[0057] FIGS. 9, 10, and 11 are diagrams showing the magnetic flux dependence of the S-parameters as functions of frequency. In FIG. 9, Φ / Φ0 = 0.42; in FIG. 10, Φ / Φ0 = 0.46; and in FIG. 11, Φ / Φ0 = 0.48. The |S 21 | and |S 31 | shown in FIGS. 9 to 11 indicate the values after calibration, not the raw magnitudes.

[0058] In Figure 9, when the normalized magnetic flux Φ / Φ0 is 0.42, signal separation, i.e., routing, is possible between the first output port Pout1 and the second output port Pout2 for input signals from 7 GHz to 8 GHz. If the operating bandwidth is defined as the range where an intensity difference of 20 dB or more is obtained between the output ports, then a bandwidth of 880 MHz is obtained.

[0059] In FIG. 10, when the normalized magnetic flux Φ / Φ0 is 0.46, signal separation (routing) is possible between the first output port Pout1 and the second output port Pout2 for input signals of 5.5 GHz to 6 GHz, and the bandwidth is 300 MHz.

[0060] In Figure 11, when the normalized magnetic flux Φ / Φ0 is 0.482, signal separation (routing) is possible for input signals between 4 GHz and 5 GHz, between 6.5 GHz and 7 GHz, and near 7.5 GHz. The operating bandwidth in the 4 GHz to 5 GHz range is 460 MHz, and the signal is coupled to the second output port Pout2. In the 6.5 GHz to 7 GHz range, signal separation is possible with a difference of nearly 40 dB, and the operating bandwidth is 180 MHz. Signal separation is also possible for 7.5 GHz signals with an intensity difference of more than 20 dB.

[0061] 9 to 11, it is confirmed that the superconducting switch circuit 10 of the embodiment can tune the frequency in the range of 4.5 GHz or more and 7.5 GHz or less, and can operate in a wide bandwidth of 100 MHz to 800 MHz, by controlling the magnetic flux passing through the SQUID 15. It is also deduced that the magnetic flux passing through the SQUID 15 can be controlled by controlling the DC control current flowing through the control line CL provided along the SQUID chain.

[0062] <Switching performance> 12A to 12F are graphs showing the switching performance at different frequencies. The input signal frequency is changed to 4.8 GHz, 5.5 GHz, 6.0 GHz, 6.5 GHz, 7.0 GHz, and 7.3 GHz within the operating frequency range of 4.5 GHz to 7.5 GHz, and the calibrated |S 21 | and |S31 Plot | as a function of the normalized magnetic flux Φ / Φ0. As the frequency changes, the curves of the S-parameters shift continuously. Therefore, within the variable frequency range, it is possible to predict how the S-parameters will be at what frequencies of the input signal.

[0063] In FIGS. 12A to 12F, the operating points are indicated by thick arrows together with the bandwidth and the S-parameter values. At any frequency, |S 21 | and |S 31 | are found to be separated by a difference of 20 dB or more. When the normalized magnetic flux is small, the minimum value of the transmission parameter becomes shallower compared to the case where the magnetic field is stronger, but it becomes less sensitive to magnetic field changes caused by external factors. External factors include noise caused by jumps in magnetic flux and instability of current sources. The high isolation and the insensitivity to magnetic noise are in a trade-off relationship, and the magnetic flux, that is, the control current, can be appropriately selected according to the required conditions.

[0064] Note that in the high-frequency region, when the normalized magnetic flux Φ / Φ0 is from 0.49 to 0.50, fluctuations in the transmission characteristics due to non-uniformity of the critical current of the SQUID 15 may occur. However, by controlling the magnetic flux to 0.49 or less, signals can be reliably separated with low loss.

[0065] FIG. 13 is a diagram showing the coupling phase between two transmission lines with different magnetic fluxes as a function of frequency. The horizontal axis of FIG. 13 is the frequency (GHz), and the vertical axis is the coupling phase χN / π. Using Equation (6), χ is extracted as a function of frequency from the measurement data for each magnetic flux. The frequency range is 4.8 GHz or more and 7.3 GHz, the same as in FIGS. 12A to 12F. The coupling phase χN / π for each value of the normalized magnetic flux is indicated by a solid line. Also, fitting the measurement data using FIGS. (7) and (8) to obtain the effective inductance Lcoup of each unit circuit 110 * is obtained. The coupling phase χN / π at each value of the obtained effective inductance Lcoup * is indicated by a dashed line. The frequency dependence of the coupling phase for each normalized magnetic flux is the effective inductance Lcoup* It shows a tendency that well matches the frequency dependence of each coupling phase.

[0066] Thus, the superconducting switch circuit 10 composed of the first transmission line 11 and the second transmission line 12 coupled by the variable inductor realized by the SQUID chain can operate with a variable frequency in the bandwidth from 100 MHz to 880 MHz and in the range from 4.5 GHz to 7.5 GHz. 21 and S 31 are separated with a difference of 20 dB or more, and in a specific bandwidth, they can be separated with a difference of 40 dB. Furthermore, by controlling the coupling phase χN, signal routing can be performed with a variable branching ratio.

[0067] Next, using the Josephson junction for testing and the bantage 107 arranged on the same chip as the superconducting switch circuit 10, the room temperature resistance R'n is measured to confirm the electrical characteristics. Using the Ambegaokar - Baratoff relational expression I cri =Δπ / (2eRn), the inductance L = Φ0 / (2πI cri ) of each Josephson junction is estimated. Here, Rn is the tunnel resistance of the junction in the normal state, and Δ is the energy gap of the superconductor. Actually, Rn < R'n, but when calculated with Rn = R'n, the inductance L per unit circuit 110 is L = 0.28 nH. The capacitance C of the thin - film capacitor 112 per unit circuit is about 300 fF. The inductance L and the capacitance C are regarded as constants.

[0068] The propagation time of the constant part of the superconducting switch circuit 10 formed by connecting 24 unit circuits 110 in series is about 124 ps. Assuming the propagation time in the SQUID chain, which is the variable part of the superconducting switch circuit 10, is 12 ps, the switching time of the superconducting switch circuit 10 is estimated to be at most 136 ps, realizing high - speed switching.

[0069] To improve the performance of the superconducting switch circuit 10, for example, to suppress magnetic flux jumps and widen the variable range of the operating frequency, the number N of the unit circuits 110 may be increased while maintaining the other parameters. cri The same effect can be obtained by lowering the MOSFET and increasing the variable coupling Lcoup itself, but this may limit the operating power. Therefore, the critical current I cri The control line CL may have an air bridge structure formed, for example, using air bridge technology, on a chip on which the lumped constant circuit 100 is formed. In this case, multiple superconducting switch circuits 10 can be formed on the same chip to operate multiple circuits. This opens up the possibility of a variety of microwave multiplexing applications and configurable networks for quantum optics. By using the superconducting switch circuit 10 of the embodiment, it is possible to control and read out a superconducting quantum system with a small amount of control wiring. Furthermore, since the superconducting switch circuit 10 formed using lumped constant elements operates at the base temperature of a dilution refrigerator, the circuit is maintained at an extremely low temperature, making it possible to operate a device in the quantum regime using a single-photon source.

[0070] <Application to quantum computing> By using a plurality of superconducting switch circuits 10 according to the embodiment, which operate in a wide band, at high speed, and with high efficiency, it is possible to construct a superconducting quantum computer that operates in the cryogenic temperature range.

[0071] Fig. 14 is a diagram showing the coupling control of 2 × 2 quantum bits in a quantum operation circuit 200A using the superconducting switch circuit 10 of the embodiment in comparison with a conventional configuration. Fig. 15 is a diagram showing the coupling control of 4 × 4 quantum bits in a quantum operation circuit 200B using the superconducting switch circuits 10-1 to 10-5 of the embodiment in comparison with a conventional configuration.

[0072] Consider a circuit that uses couplers between qubits to reduce the error rate. When qubits are input to ports (1) and (4), the 2 x 2 qubit operations between ports (1) to (4) are 1-2, 1-3, 4-2, and 4-3. When using conventional variable couplers CPL1, CPL2, CPL3, and CPL4 in Figure 14(B), the same 2 x 2 coupling occurs, but control lines CL1, CL2, CL3, and CL4 are required to switch the coupling on and off for each variable coupler CPL1 to CPL4.

[0073] In contrast, in the superconducting switch circuit 10 of the embodiment shown in Figure 14(A), by using a single DC control line CL as described above, it is possible to control 2 x 2 coupling states among the four ports (1), (2), (3), and (4). For example, a quantum bit may be input to port (1) or (4) to manipulate the coupling to ports (2) and (3). Alternatively, quantum bit 1 may be input to port (1) and quantum bit 2 may be input to port (4) to manipulate the coupling between these quantum bits.

[0074] 15A, five switches, namely, superconducting switch circuits 10-1 to 10-5 (collectively referred to as "superconducting switch circuits 10" as appropriate), according to an embodiment, are arranged to realize the operation of 4×4 quantum bits. One control line CL is used to control the coupling state of each superconducting switch circuit 10, and a total of five control lines CL1 to CL5 are arranged. As described above, CL1 to CL5 can be realized by air bridge technology, and therefore, the coupling of 4×4 quantum bits can be realized with a compact configuration in which the entire quantum operation circuit 200B can be housed in a refrigerator.

[0075] In contrast, the conventional variable coupler shown in Figure 5(B) requires 16 variable couplers to control the on / off state of the coupling between input ports (1) to (4) and output ports (5) to (8). To control each of these variable couplers, 16 control lines CL1 to CL16 are required. This requires a very large area for the variable couplers, and the wiring becomes complicated.

[0076] The superconducting switch circuit 10 of the embodiment becomes more advantageous in terms of compact circuit scale and fewer control lines as the number of couplings between quantum bits used in quantum operations increases.

[0077] The above description may include the following aspects. (Section 1) a first transmission line in which N (N is an integer equal to or greater than 1) first unit cells, each including a first inductor and a first capacitor, are connected in series between a first input port and a first output port; a second transmission line in which N second unit cells, each including a second inductor and a second capacitor, are connected in series between a second input port and a second output port; a variable coupling provided between the first transmission line and the second transmission line, the variable coupling adjusting a coupling state between the first transmission line and the second transmission line; a control line for controlling the coupling strength of the variable coupling; Equipped with the first unit cell, the second unit cell, and the variable coupling are formed of lumped elements; Superconducting switch circuit. (Section 2) the control line is a DC current line, and the coupling strength of the variable coupling is controlled by controlling a magnetic flux generated by a control current flowing through the control line, and a signal input from the first input port or the second input port is output from one or both of the first output port and the second output port. Item 1. The superconducting switch circuit according to item 1. (Section 3) a branching ratio of the signal between the first output port and the second output port is variable; Item 2. The superconducting switch circuit according to item 2. (Section 4) the first capacitor is connected to a first node on the output side of the first inductor; the second capacitor is connected to a second node on the output side of the second inductor; the variable coupling is connected to a third node between the first capacitor and the second capacitor. The superconducting switch circuit according to any one of items 1 to 3. (Item 5) The variable coupling is a superconducting quantum interference device connected to the third node. The superconducting switch circuit according to item 4. (Item 6) The variable coupling has N superconducting quantum interference devices connected in series. The superconducting switch circuit according to any one of items 1 to 3. (Item 7) The control line is a DC control line. The magnetic flux generated by the DC current flowing through the control line acts on the N superconducting quantum interference devices to control the mutual inductance between the first transmission line and the second transmission line. The superconducting switch circuit according to item 6. (Item 8) The control line has an air bridge configuration. The superconducting switch circuit according to any one of items 1 to 7. (Item 9) The first unit cell is a first lumped constant circuit formed by a first Josephson junction and a first thin film capacitor. The second unit cell is a second lumped constant circuit formed by a second Josephson junction and a second thin film capacitor. The superconducting switch circuit according to any one of items 1 to 8. (Item 10) A quantum computing circuit using the superconducting switch circuit according to any one of items 1 to 9, wherein a qubit is input to the first input port or the second input port, and the coupling state of the variable coupling is controlled to operate the coupling of the qubit.

Description of symbols

[0078] 10, 10-1 to 10-5 Superconducting switch circuit 11 First transmission line 12 Second transmission line 15, 15-1 to 15-N SQUID 100 Constant Concentration Circuit 101, 101-1 to 101-N First Unit Cell 102, 102-1 to 102-N Second Unit Cell 103, 103-1 to 103-N, 104, 105 Nodes 110 Unit Circuit 111, 111-1 to 111-N First Josephson Junction 112, 112-1 to 112-N First Thin Film Capacitor 121, 121-1 to 121-N Second Josephson Junction 122, 122-1 to 122-N Second Thin Film Capacitor Lcoup Variable Coupling Pin1 First Input Port Pin2 Second Input Port Pout1 First Output Port Pout2 Second Output Port

Claims

1. A first transmission line in which N (N is an integer of 1 or more) first unit cells each including a first inductor and a first capacitor are connected in series between a first input port and a first output port; A second transmission line in which N second unit cells each including a second inductor and a second capacitor are connected in series between a second input port and a second output port; A variable coupling provided between the first transmission line and the second transmission line for adjusting a coupling state between the first transmission line and the second transmission line; A control line for controlling a coupling strength of the variable coupling; Comprising; The first unit cell, the second unit cell, and the variable coupling are formed of lumped elements; A superconducting switch circuit.

2. The control line is a direct current line, and by controlling a magnetic flux generated by a control current flowing through the control line, the coupling strength of the variable coupling is controlled, and a signal input from the first input port or the second input port is output from one or both of the first output port and the second output port. The superconducting switch circuit according to Claim 1.

3. A branching ratio of signals between the first output port and the second output port is variable. The superconducting switch circuit according to Claim 2.

4. The first capacitor is connected to a first node on an output side of the first inductor; The second capacitor is connected to a second node on an output side of the second inductor; The variable coupling is connected to a third node between the first capacitor and the second capacitor. The superconducting switch circuit according to Claim 1.

5. The variable coupling is a superconducting quantum interference device connected to the third node. The superconducting switch circuit according to Claim 4.

6. The variable coupling has N superconducting quantum interference devices connected in series. The superconducting switch circuit according to Claim 1.

7. The control line is a direct current control line; A magnetic flux generated by a direct current flowing through the control line acts on the N superconducting quantum interference devices to control a mutual inductance between the first transmission line and the second transmission line. The superconducting switch circuit according to Claim 6.

8. The control line has an air bridge configuration. The superconducting switch circuit according to Claim 1.

9. The first unit cell is a first lumped constant circuit formed of a first Josephson junction and a first thin film capacitor. The second unit cell is a second lumped constant circuit formed by a second Josephson junction and a second thin film capacitor. The superconducting switch circuit according to claim 1.

10. A quantum computing circuit using the superconducting switch circuit according to any one of claims 1 to 9, A quantum computing circuit in which a qubit is input to the first input port or the second input port, and the coupling state of the variable coupling is controlled to operate the coupling of the qubit.

Images