MILLIOHM RESISTOR FOR RQL CIRCUITS
1. A resistor comprising:
- a first terminal made of a superconducting metal;
a second terminal made of a superconducting metal; and
a Josephson junction device electrically connected between the first and second terminals, the Josephson junction device comprising ferromagnetic or antiferromagnetic material,wherein the resistor has a resistance value of between about one milliohm and about fifty milliohms.
A milliohm resistor is fabricated as a Josephson junction device that contains ferromagnetic or antiferromagnetic material of sufficient thickness to render the device entirely resistive between terminals. The device can have a resistance on the order of milliohms and can be consume a much smaller chip footprint than resistors of the same resistance fabricated using conventional resistive materials. Because the device can be fabricated without modification to processes used to fabricate reciprocal quantum logic (RQL) circuitry, it can easily be incorporated in RQL circuits to mitigate flux trapping or to perform other functions where very small resistances are needed. In particular, the device can burn off circulating currents induced by trapped flux without affecting the transmission of SFQ pulses through RQL circuitry.
- 1. A resistor comprising:
a first terminal made of a superconducting metal; a second terminal made of a superconducting metal; and a Josephson junction device electrically connected between the first and second terminals, the Josephson junction device comprising ferromagnetic or antiferromagnetic material, wherein the resistor has a resistance value of between about one milliohm and about fifty milliohms.
- View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
- 17. A method of mitigating flux trapping in a reciprocal quantum logic (RQL) circuit, the method comprising:
fabricating a resistor in an inductive loop in the RQL circuit, wherein the resistor has a resistance of between about one milliohm and about fifty milliohms, wherein the resistor has a footprint of less than twenty-five square micrometers, and wherein the resistor reduces to zero, in between about one millisecond and about ten microseconds, current circulating in the inductive loop as the result of trapped flux.
- 18. A chip-fabricated device comprising:
a first superconducting metal layer forming a first galvanic terminal; a second superconducting metal layer forming a second galvanic terminal; and between the superconducting metal layers, a barrier comprising one of; a synthetic antiferromagnet (SAF) comprising two ferromagnetic material layers separated by a spacer layer, the ferromagnetic material layers arranged to have antiparallel magnetizations with respect to each other, or an antiferromagnet layer; wherein the device has a resistance of between about one milliohm and about fifty milliohms, and wherein the device has a fabrication footprint of less than twenty-five square micrometers.
- View Dependent Claims (19, 20)
The present invention relates generally to superconducting circuit devices, and specifically to a milliohm resistor for reciprocal quantum logic (RQL) circuits.
Superconducting electronics can be implemented to provide high-performance computing with low energy consumption. Many superconducting computing circuits, including those classed as reciprocal quantum logic (RQL) circuits, include loops that have (passive) inductors and Josephson junctions (as the active circuit elements) arranged so as to carry out logical functions.
Antiferromagnetism is a property of certain materials wherein, below the Neel temperature, the magnetic moments of atoms or molecules, typically related to the spins of electrons, can align in a regular pattern with neighboring spins (on different sublattices) pointing in opposite directions. Synthetic antiferromagnets (SAFs) are artificial antiferromagnets made of multiple thin ferromagnetic layers separated by a nonmagnetic layer. Dipole coupling of the ferromagnetic layers can result in antiparallel alignment of the magnetization of the ferromagnets.
Giant magnetoresistance (GMR), in which antiferromagnetism plays a role, is a quantum-mechanical magnetoresistance effect observed in multilayers composed of alternating ferromagnetic and non-magnetic conductive layers. The GMR effect is can be observed as a relatively high electrical resistance when magnetizations of adjacent ferromagnetic layers are in an antiparallel alignment state.
One example includes a resistor having a first terminal made of a superconducting metal, a second terminal made of a superconducting metal, and a Josephson junction device electrically connected between the first and second terminals. The Josephson junction device includes ferromagnetic or antiferromagnetic material. The resistor has a resistance value of between about one milliohm and about fifty milliohms.
Another example includes a method of mitigating flux trapping in a reciprocal quantum logic (RQL) circuit. A resistor is fabricated in an inductive loop in the RQL circuit. The resistor has a resistance of between about one milliohm and about fifty milliohms. The resistor has a footprint of less than twenty-five square micrometers. The resistor reduces to zero, in between about one millisecond and about ten microseconds, current circulating in the inductive loop as the result of trapped flux.
Yet another example includes chip-fabricated device having a resistance of between about one milliohm and about fifty milliohms, and a fabrication footprint of less than twenty-five square micrometers. The device has a first superconducting metal layer forming a first galvanic terminal, a second superconducting metal layer forming a second galvanic terminal, and, between the superconducting metal layers, a barrier. The barrier is either a synthetic antiferromagnet (SAF) made up of two ferromagnetic material layers separated by a spacer layer, the ferromagnetic material layers arranged to have antiparallel magnetizations with respect to each other, or an antiferromagnet layer.
Although it is generally desirable in chip-fabricated superconducting circuits to eliminate sources of passive resistance in order to provide circuits of lower operating power, it may be desirable in certain circumstances to include passive resistances as a matter of design in order to alleviate undesirable side-effects of superconducting circuit operation or to permit specialized circuit operation. However, conventional resistive materials tend to have a resistance per minimum fabrication area square that is on the order of a few ohms, e.g., ten ohms or twenty ohms. Fabricating a very low resistance resistor, e.g., to provide a resistance of one ohm or less, e.g., to provide a resistance on the order of milliohms, out of conventional resistive materials can thus require impractically large plan-view device area that is expensive from a layout area standpoint, and which can also manifest parasitic capacitive or inductive side-effects in the circuit that may negatively impact circuit performance and may prevent a circuit from functioning as intended. A comparatively low-area milliohm resistor as described herein can be used in superconducting circuits and particularly in RQL circuits, for example, to absorb unwanted currents, for example, to eliminate flux trapping.
The resistance of device 102 can be tuned at fabrication time by adjusting the plan-view area of device 102 to provide a resistance value in the low milliohms to multiple tens of milliohms in value, e.g., between about one milliohm and about fifty milliohms, e.g., between about one milliohm and about twenty milliohms, e.g., between about one milliohm and about ten milliohms, e.g., between about one milliohm and about five milliohms. Accordingly, as the term “milliohm resistor” is used herein, such a resistor is not necessarily limited, except as otherwise specified, only to resistances of one milliohm or even to resistances of between one and ten milliohms. Milliohm resistor 100 can thus provide low resistance values not easily achievable in the RQL process. Device 102 can also include layer growth promoting outer spacer layers (not shown in
In each of
Upper superconducting galvanic contact 202 and lower superconducting galvanic contact 214 in milliohm resistor 200 can each be made, for example, of a superconducting metal, e.g., niobium, niobium nitride, or aluminum. These layers of device 200 galvanically connect the device to other parts of a circuit in which the milliohm resistor 200 is implemented, e.g., a structure such as the one shown in
Spacer layer 208 in milliohm resistor 200 can, for example, be made of any one of elemental copper (Cu), elemental ruthenium (Ru), elemental iridium (Ir), or elemental rhodium (Rh). Spacer layer 208 can have a thickness of, for example, between about 0.5 nanometers and about 1.0 nanometers.
Ferromagnetic layers 206, 210 can, for example, be made of any one of elemental cobalt, elemental iron, a cobalt-iron alloy (e.g., 1:1 CoFe), a cobalt-iron-boron alloy (CoFeB), a nickel-iron alloy (NiFe), a nickel-iron-niobium alloy (NiFeNb), a nickel-iron-chromium alloy (NiFeCr), or a nickel-iron-copper alloy (NiFeCu). In the case of each alloy mentioned, many different alloy concentrations can be used, provided that the alloy concentration is magnetic. The selection of materials for ferromagnetic layers 206, 210 and spacer layer 208 are interchangeable given appropriate selection of the layer thicknesses; thus, for example, the SAF barrier can be made with cobalt ferromagnetic layers 206, 210 and a ruthenium spacer layer 208, or, with appropriate selection of the layer thicknesses, can be made to work just as well with nickel-iron alloy ferromagnetic layers 206, 210 and a copper spacer layer 208.
Ferromagnetic layer thicknesses of about ten nanometers are sufficient to suppress all supercurrent in the junction 200. The ferromagnetic layers can therefore each be at least ten nanometers thick. The precise resistance of the junction can be adjusted by adapting the fabrication area of the junction. Spacer layer 208 should be fabricated to be of sufficient thickness to ensure antiferromagnetic coupling will prevent magnetic fields from causing failures in nearby digital circuits.
Antiferromagnetic barrier layers 416, 516 can be made, for example, of ferromanganese (FeMn), ferrous oxide (FeO), ferric oxide (Fe2O3), nickel (II) oxide (NiO), manganese (II) oxide (MnO), manganese (IV) oxide (MnO2), chromium (III) oxide (Cr2O3), or iridium manganese (IrMn). There are a wide variety of antiferromagnetic materials that can be used for antiferromagnetic layer 416, 516, and this list is not exclusive. Antiferromagnetic barrier layers 416, 516 can have thicknesses of, for example, between about ten nanometers and about fifty nanometers, e.g., about fifty nanometers. Antiferromagnetic barrier layers 416, 516 of lesser thicknesses can be effective, but thicker layers advantageously provide fabrication uniformity of the material layers (e.g., it is more practicable to achieve a one percent uniformity on five-angstrom variation than on one-angstrom variation).
Because a milliohm resistor as shown in any of
A milliohm resistor as shown in any of
A milliohm resistor as shown in any of
A trapped static flux induces a supercurrent upon the inductive loop that will persist in the loop indefinitely absent a resistance in series with the loop, in which case the supercurrent will no longer be a supercurrent and will “burn off” (i.e., be completely converted from electrical current to another form, such as heat) and, within a time that is dependent upon the value of the resistance, be reduced to zero. A loop-series resistance that is too high (e.g., on the order of ohms) can impede the propagation of single flux quantum (SFQ) pulses representing data or control signals and for which the time spent in any one inductive loop is typically on the order of picoseconds or nanoseconds.
By contrast, a loop-series resistance that is not too high (e.g., on the order of milliohms) will only burn off currents over a time scale of milliseconds or microseconds (e.g., between about one millisecond and about ten microseconds, e.g., between about ten milliseconds and one microsecond) and thus will not affect the intended propagation of such SFQ pulses because such SFQ pulses do not remain in any one loop long enough to be burned away by the milliohm resistor. As such, it is desirable that the loop-series resistance is not so high that it breaks the logical function of the inductive loop in which it may reside. A milliohm resistor can have a small enough resistance value that it has no impact on the logical application of the circuit, but over a large enough time scale, will burn off any unwanted supercurrents as may result from flux trapping.
As another example application, a milliohm resistor as shown in any of
As another example of sub-ohm resistances used in circuit design, a milliohm resistor as described herein could be used to construct a circuit that takes advantage of timescale as a design variable. Thus, for example, a circuit can be arranged such that if two current pulses are input to the circuit within a first amount of time of each other, below a time threshold, i.e., such that the first current pulse has not been sufficiently burned off by the time of the receipt of the second current pulse, the circuit performs a first logical function (e.g., because a logical decision Josephson junction is sufficiently biased by the additive effect of currents from the two pulses, or because a logical decision Josephson junction is triggered by the additive effect of such currents), whereas if the two current pulses are input to the circuit within a second amount of time of each other, above the time threshold, i.e., such that the first current pulse has been sufficiently burned off by the time of the receipt of the second current pulse, the second performs a second logical function distinct from the first logical function (e.g., because the logical decision Josephson junction is insufficiently biased by the current from the second pulse alone, or because a logical decision Josephson junction is not triggered by the effect of such current).
A milliohm resistor as shown in any of
The milliohm resistor devices described herein provide a unique fabrication area advantage over resistors of equivalent resistance made using conventional processes. The following table compares reduced integrated circuit footprint areas of example milliohm resistor devices as set forth in the present disclosure, termed “present devices” and fabricated as being circular in a plan view, to the much larger footprint areas required for conventional one-ohm-per-square and ten-ohm-per-square superconducting fabrication processes. The width and area values in the one-ohm-per-square and ten-ohm-per-square superconducting fabrication processes assume a minimum resistor length of about 0.5 micrometers, where “width” and “length” are dimensions in the x-y plane. Each different fabrication process defines the thickness z for its own manufacturing. The one-ohm-per-square and ten-ohm-per-square resistors, to which the present-device milliohm resistors are compared in the below table, are essentially wires of some metal, the z dimension thereof being fixed.
As can be seen in the above table, the present devices, as described herein, provide an area advantage by roughly a factor of twenty-five over the one-ohm-per-square process, and by another factor of ten beyond that for the higher-resistance ten-ohm-per-square process. The milliohm resistors of the present disclosure are compatible with either of these processes and thus, by using the present devices, much smaller milliohm resistors can be fabricated on superconducting integrated circuits made using either of these conventional processes.
What have been described above are examples of the invention. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the invention, but one of ordinary skill in the art will recognize that many further combinations and permutations of the invention are possible. Accordingly, the invention is intended to embrace all such alterations, modifications, and variations that fall within the scope of this application, including the appended claims. Additionally, where the disclosure or claims recite “a,” “an,” “a first,” or “another” element, or the equivalent thereof, it should be interpreted to include one or more than one such element, neither requiring nor excluding two or more such elements. As used herein, the term “includes” means includes but not limited to, and the term “including” means including but not limited to. The term “based on” means based at least in part on.