Quantum transistor

Abstract

A quantum computing device includes an optical resonator having a resonant wavelength band. A crystalline material including a crystal defect is contained within the optical resonator. The crystal defect has a ground state and an excited state, which has an emission wavelength in the resonant wavelength band. A source electrode and a drain electrode are disposed on opposing sides of the crystal defect and configured to apply a first electric field in the crystalline material along a longitudinal axis. A gate electrode is disposed in proximity to the crystal defect and configured to apply to the crystalline material a second electric field transverse to the longitudinal axis. Control circuitry is configured to apply a first voltage between the source and drain electrodes to control a charge state of the crystal defect and to apply a second voltage to the gate electrode to tune the emission wavelength.

Claims

1. A quantum computing device, comprising: an optical resonator having a resonant wavelength band; a crystalline material comprising a crystal defect contained within the optical resonator, the crystal defect having a ground state and an excited state, which has an emission wavelength in the resonant wavelength band; a source electrode and a drain electrode disposed on opposing sides of the crystal defect and configured to apply a first electric field in the crystalline material along a longitudinal axis; a gate electrode disposed in proximity to the crystal defect and configured to apply to the crystalline material a second electric field transverse to the longitudinal axis; semiconductor donor and acceptor layers overlying the crystalline material between at least one of the electrodes and the crystal defect; and control circuitry, which is configured to apply a first voltage between the source and drain electrodes to control a charge state of the crystal defect and to apply a second voltage to the gate electrode to tune the emission wavelength, wherein application of the first electric field releases carriers from the semiconductor donor layer to the semiconductor acceptor layer in a vicinity of the crystal defect to create a two-dimensional PIN structure over the crystal defect, and switching a polarity of the first voltage switches the charge state of the crystal defect by shifting positive and negative carrier bands in the crystalline material via the PIN structure over the crystal defect.

2. The device according to claim 1, wherein the crystalline material comprises diamond.

3. The device according to claim 2, wherein the crystal defect comprises a nitrogen vacancy (NV) defect.

4. The device according to claim 3, wherein application of the first voltage between the source and drain electrodes switches the NV defect between an NV.sup.0 state and an NV.sup. state.

5. The device according to claim 1, wherein application of the first electric field with a first polarity switches the crystal defect to a first charge state, and application of the first electric field with a second polarity, opposite to the first polarity, switches the crystal defect to a second charge state having the ground state and the excited state, which has the emission wavelength in the resonant wavelength band.

6. The device according to claim 5, wherein the control circuitry is configured to apply the first voltage so as to maintain the crystal defect in the second charge state during a period of quantum computation.

7. The device according to claim 5, and comprising an optical waveguide, which is configured to convey one or more excitation beams to the crystal defect at wavelengths selected to convert one or more electronic spin states of the crystal defect to respective charge states, and wherein the controller is configured to detect the charge states of the crystal defect by measuring a current between the source electrode and the drain electrode.

8. The device according to claim 5, and comprising an optical waveguide, which is configured to convey one or more excitation beams to the crystal defect, including a first excitation beam at a transition wavelength of the second charge state and a second excitation beam at an infrared wavelength chosen to initialize the crystal defect in the second charge state while preventing conversion of the second charge state to the first charge state.

9. The device according to claim 1, wherein the control circuitry is configured to adjust the second voltage so as to tune the emission wavelength of the crystal defect.

10. The device according to claim 1, wherein the semiconductor donor layer comprises one or more of a transition metal oxide (TMO) and a transition metal dichalcogenide (TMD).

11. The device according to claim 1, and comprising a piezoelectric element coupled to modify a length of the optical resonator, wherein the control circuitry is configured to drive the piezoelectric element to adjust the resonant wavelength band of the optical resonator.

12. The device according to claim 1, wherein the optical resonator comprises periodic structures formed in the crystalline material on opposing sides of the crystal defect.

13. The device according to claim 12, wherein the periodic structures comprise holes extending through a layer of the crystalline material that contains the crystal defect.

14. The device according to claim 12, wherein the periodic structures comprise indentations in a surface of the crystalline material.

15. The device according to claim 1, wherein the crystalline material is configured as an optical waveguide, which is coupled to convey optical radiation emitted from the crystal defect at the emission wavelength from the optical resonator to a detector.

16. The device according to claim 15, wherein the optical waveguide is further coupled to convey one or more excitation beams at one or more excitation wavelengths from one or more beam sources to the crystal defect.

17. The device according to claim 16, and comprising: a substrate, on which the optical waveguide is disposed; an input waveguide, which is disposed on the substrate and is coupled to inject the one or more excitation beams into the optical waveguide; and an output waveguide, which is disposed on the substrate and is coupled to receive the emitted optical radiation from the optical waveguide.

18. The device according to claim 17, wherein the optical waveguide and the input and output waveguides have tapered ends, which are overlaid on the substrate so as to inject the one or more excitation beams into the optical waveguide and inject the emitted optical radiation into the output waveguide by adiabatic coupling.

19. The device according to claim 17, wherein the output waveguide comprises a filter configured to block the excitation wavelength.

20. The device according to claim 15, wherein the optical waveguide comprises a ridge disposed on a slab of the crystalline material.

21. The device according to claim 15, wherein the optical waveguide has a cross-sectional profile selected from a group of profiles consisting of a rectangular profile, a triangular profile, and a pentagonal profile.

22. The device according to claim 1, and comprising: a photonic integrated circuit (PIC) substrate, on which the crystalline material is disposed; and one or more optical waveguides, which are disposed on the substrate and are optically coupled to the crystalline material.

23. The device according to claim 22, wherein the one or more optical waveguides comprise at least one inverse taper configured to couple optical radiation between the crystalline material and the one or more optical waveguides.

24. The device according to claim 22, wherein the PIC comprises a reflective surface, which is configured to couple optical radiation between the crystalline material and the one or more optical waveguides.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a block diagram that schematically illustrates a quantum computing system, in accordance with an embodiment of the invention;

(2) FIGS. 2A, 2B, and 2C are schematic top, side, and sectional views, respectively, of a quantum transistor, in accordance with an embodiment of the invention;

(3) FIGS. 3A, 3B, and 3C are schematic sectional views of semiconductor layers on a diamond crystal in a quantum transistor, in accordance with an embodiment of the invention;

(4) FIG. 4 is a plot that schematically illustrates emission wavelengths of an NV color center in a quantum transistor as a function of voltages applied to the electrodes of the quantum transistor, in accordance with an embodiment of the invention;

(5) FIG. 5A is a schematic side view of waveguides in a quantum transistor showing details of an optical resonator and filter implemented in the waveguides, in accordance with an embodiment of the invention;

(6) FIG. 5B is a schematic detail view of a periodic structure within the resonator of FIG. 5A, in accordance with an embodiment of the invention;

(7) FIG. 6 is a schematic detail view of a periodic structure used in the resonator of a quantum transistor, in accordance with another embodiment of the invention;

(8) FIG. 7 is a schematic sectional view of an optical resonator used in a quantum transistor, in accordance with an alternative embodiment of the invention; and

(9) FIGS. 8 and 9 are schematic side views of quantum transistors, in accordance with alternative embodiments of the invention.

DETAILED DESCRIPTION

Overview

(10) For the sake of concreteness and clarity, the description that follows will focus on qubits based on NV color centers in diamond. These embodiments, however, are described solely by way of example. The principles of these embodiments may similarly be applied, mutatis mutandis, in qubits based on other types of color centers in diamond, as well as based on defects in other types of crystals.

(11) Despite the promise and potential advantages of qubits based on NV defects in diamond (and other types of crystal defects), practical quantum computing devices based on NV defects have not yet been demonstrated. Experimental systems based on NV defects are still far from meeting the minimal levels of stability, efficiency, and fidelity that are needed for multi-qubit devices. A number of problems must be overcome in order to create a quantum computer based on NV defects: Quantum efficiencyThe luminescence spectrum of the NV.sup. center comprises a zero-phonon line (ZPL) and a broad phonon sideband, which contains most of the spontaneous luminescent energy emitted from the excited triplet state. Only the ZPL has the well-defined frequency and polarization needed for qubit operations. Nonradiative losses reduce the quantum efficiency of ZPL emission still further. Spectral diffusionThe NV.sup. center has electric dipole moments in both the ground and the excited states. Therefore, the NV.sup. transition frequency, on which the qubit is based, is sensitive to local electric fields, which tend to fluctuate with location, thus causing frequency variations, known as spectral diffusion, among the qubits in a multi-qubit device. For multi-qubit entanglement and other operations, the spectral diffusion must be held within tight limits. Charge stabilityThe NV center has two metastable charge states: NV.sup. and NV.sup.0. Optical excitation of the NV charge state using a laser frequently causes ionization, with a resulting transition to the NV.sup.0 state and loss of quantum information.

(12) Embodiments of the present invention that are described herein address these problems using a novel quantum computing device that is referred to herein as a quantum transistor. Strictly speaking, the device is not a transistor in the sense in which this term is conventionally used in electronic circuits. It is referred to herein as a quantum transistor, however, because of its topological similarity to an electronic transistor, with three electrodes coupled to a solid-state core.

(13) In the quantum transistors that are described herein, a crystal defect is contained in an optical resonator having a resonant wavelength band that contains the emission wavelength of the crystal defect (i.e., the ZPL). The resonator may comprise, for example, periodic structures formed on opposing sides of the defect in the crystalline material (such as diamond). The resonator is designed to have a high Q factor and low mode volume, to give strong Purcell enhancement of the spontaneous emission in the ZPL and thus increase the quantum efficiency of the device.

(14) Electrodes are connected to the crystalline material, including source, drain, and gate electrodes. The source and drain electrodes are disposed on opposing sides of the crystal defect and define a longitudinal axis running through the defect. Application of a voltage between the source and drain electrodes gives rise to an electric field in the crystalline material along the longitudinal axis. Application of a voltage to the gate electrode gives rise to an additional electric field transverse to the longitudinal axis. By applying the proper voltages between the source and drain electrodes, while also applying a voltage to the gate electrode, control circuitry of the quantum transistor is able to switch the charge state of the crystal defect. Thus, an NV center in diamond, for example, can be switched reliably and stably from the NV.sup.0 state to the NV.sup. state required for qubit operations, as well as maintaining the NV center in the NV.sup. state during a period of quantum computation.

(15) Furthermore, by adjusting the voltage applied to the gate electrode, the control circuitry is able to tune the emission wavelength of the ZPL and thus offset the effect of spectral diffusion. In this manner, multiple qubits can be spectrally calibrated so that their ZPL frequencies are substantially identical, thus facilitating entanglement among the qubits and enabling multi-qubit operations.

(16) In some embodiments, the optical resonator is coupled efficiently to one or more optical waveguides, which convey the optical radiation emitted from the crystal defect at the emission wavelength of the defect (for example, the red ZPL emission of the NV.sup. defect) from the optical resonator to a detector. The optical waveguides may also convey one or more excitation n beams at appropriate excitation wavelengths of the crystal defect (for example, a green beam for exciting the NV.sup. defect) from one or more beam sources to the crystal defect. In some embodiments, the waveguides are formed on a photonic integrate circuit (PIC) substrate, on which the crystalline material containing the defect is mounted. Multiple diamond chiplets containing respective NV defects may be mounted together on such a PIC, thus creating a multi-qubit quantum computing device, while the waveguides form a photonic network that interconnects the qubits for purposes of computing operations and output detection.

System Description

(17) FIG. 1 is a block diagram that schematically illustrates a quantum computing system 20, in accordance with an embodiment of the invention. System 20 comprises an array of quantum transistors 22 mounted on a PIC 24. Quantum transistors 22 comprise diamond chiplets, each including an NV defect contained in an optical resonator, with electrodes connected to the chiplet, as shown in detail in the figures that follow. In some embodiments, the diamond chiplet comprises a waveguide, and the optical resonator is formed in this diamond waveguide, which couples to the PIC.

(18) PIC 24 comprises a substrate, such as a silicon-on-insulator (SOI) substrate, on which a network of waveguides with low propagation loss is formed, for example SiN waveguides, for conveying optical radiation to and from quantum transistors 22. Alternatively, other types of substrates may be used, such as a silicon or glass substrate. In the present example, these waveguides include input waveguides 28, which inject one or more optical excitation beams, for example a green beam output by a laser 26 at around 532 nm, into quantum transistors 22 via a suitable coupler. In addition, output waveguides 36 on PIC 24 receive the radiation emitted from the quantum transistors, for example red radiation in the ZPL of the NV.sup. defects, at approximately 637 nm via another coupler. Methods for optically coupling waveguides 28 and 36 to the diamond chiplet containing the NV defect are described further hereinbelow.

(19) In some embodiments, multiple excitations beams are applied at different wavelengths to excite the NV defects. In one such embodiment, one excitation beam, such as a green beam at 532 nm, is applied at a transition wavelength of the NV.sup. state, and a second excitation beam is applied at an infrared wavelength chosen to initialize the NV defect to its NV.sup. charge state while avoiding ionizations of the charge state (i.e., preventing transition to the NV.sup.0 charge state). The infrared wavelength is typically chosen in the range of 780-1900 nm, for example at 1550 nm, and is linearly polarized along a direction perpendicular to the axis of the NV defect structure (i.e., perpendicular to the axis between the nitrogen atom and the crystal vacancy).

(20) In addition, a microwave source 30 generates a microwave input 32 to quantum transistors 22, to drive transitions between the electron spin energy levels of the NV.sup. ground state. These transitions can be exploited both in reading the state of a qubit, for example using conversion of spin to charge state as described further hereinbelow, and in shifting the NV defects between ground and superposition states, which can be used both in single-qubit operations and in n creating entanglement between qubits.

(21) Electronic control circuitry 34 applies electrical signals to the electrodes of quantum transistors 22 (as shown in the figures that follow) in order to switch the defect charge states and tune the emission wavelengths of the crystal defects. These functions are described further hereinbelow with reference to FIGS. 3A-C and 4.

(22) In the pictured embodiment, output waveguides 36 couple quantum transistors 22 to a photonic network 38, which in turn connects the output signals from the quantum transistors to optical detectors 44. In the pictured example, photonic network 38 comprises optical switches 40, such as phase-controlled Mach-Zehnder interferometers. Detectors 44 may comprise, for example, high-speed avalanche photodiodes (APDs) or single-photon avalanche diodes (SPADs), which are fabricated on a separate detector chip 42 coupled optically to PIC 24. Alternatively, the optical detectors may be fabricated on PIC 24, as well. High-speed SPADs with fast quenching can be useful in overcoming spectral among diffusion the quantum transistors. The details of network 38 and detectors 44, however, are beyond the scope of the present disclosure.

(23) Optical and Electrical Structure of a Quantum Transistor

(24) FIGS. 2A, 2B, and 2C are schematic top, side, and sectional views, respectively, of quantum transistor 22, in accordance with an embodiment of the invention. In this embodiment, an NV defect 54 is formed in a diamond waveguide 52, such as a ridge waveguide in the pictured embodiment. NV defect 54 can be formed, for example, by electron irradiation, ion nano-implantation, or pulsed laser irradiation of the diamond crystal. Waveguide 52 in this example comprises a narrow, rectangular ridge disposed on a broader slab 50 of diamond. The waveguide and slab can be produced, for example, by photolithographic etching of a diamond membrane. In a typical implementation, waveguide 52 has a length in the range of 10-50 m, and a height and width in the range of 100-300 nm. The thickness of slab 50 beneath waveguide 52 is typically in the range of 50-150 nm. The dimensions of waveguide 52 may advantageously be chosen to support single-mode operation at both the excitation wavelength and the emission wavelength of defect 54. Alternatively, other dimensions and other types of diamond waveguides may be used, such as a strip with a rectangular (fishbone), triangular, or pentagonal cross-sectional profile.

(25) An optical resonator 56, containing defect 54, is formed in the central part of waveguide 52. Resonator 56 has a high-Q resonant wavelength band that contains the ZPL emission wavelength of defect 54. The wavelength band of the resonator may be tunable, for example using a piezoelectric element in as shown FIG. 8. The piezoelectric element may be positioned alongside or beneath resonator 56, or it may be formed on PIC 24. Resonator 56 is defined in the present embodiment by periodic structures formed in waveguide 52, for example as shown in FIGS. 5A/B, 6 and 7. This type of resonator is advantageous in achieving high Q and a small mode volume. Alternatively, other types of resonant structures may be used.

(26) Input waveguide 28 and output waveguide 36 terminate in respective adiabatic tapers 60 and 62 on a PIC substrate 58 within quantum transistor 22. Tapers 60 and 62 are produced, for example, by suitable etching of the SiN layer on PIC 24. Tapers 60 and 62 are overlaid by corresponding inverse tapers of the ends of diamond waveguide 52 in respective overlap regions 61 and 63. Waveguide 52 may be separated in the Z-direction from the underlying tapers 60 and 62 by a thin dielectric layer, for example comprising SiO.sub.2. The green light input through waveguide 28 is transferred adiabatically, with low loss, by evanescent wave coupling from taper 60 into the overlying part of diamond waveguide 52 in region 61; and the red light output from defect 54 is similarly transferred from diamond waveguide 52 into taper 62 in region 63. For efficient adiabatic coupling, the lengths of overlap regions 61 and 63 in the X-direction are typically in the range of 10-20 m.

(27) Conductive electrodes are deposited on diamond slab 50, including a source electrode 64, a drain electrode 66, and a gate electrode 68, in proximity to defect 54. Source electrode 64 and drain electrode 66 are deposited on opposing sides of defect 54, defining a longitudinal axis through their respective locations. (In the present example, this longitudinal axis runs parallel to the X-axis, but other orientations of the longitudinal axis may alternatively be used.) When a voltage is applied through an active conductive channel between source electrode 64 and drain electrode 66, it will give rise to an electric field within diamond waveguide 52 along the X-axis. Gate electrode 68 is offset from the longitudinal axis, so that a voltage applied between the gate electrode and drain electrode 66 (or between the gate electrode and source electrode 64), for example, will give rise to a transverse electric field component along the Y-direction within diamond waveguide 52.

(28) Control circuitry 34 (FIG. 1) drives electrodes 64, 66 and 68 to control the charge state of crystal defect 54 and to tune the ZPL wavelength emitted by the crystal defect. To enable this functionality, a semiconductor hole donor layer 70 (which may be a monolayer or multilayer structure) is formed over the surface of diamond waveguide 52, between the electrodes (particularly gate electrode 68) and NV defect 54. This donor layer 70 may comprise, for example, a transition metal oxide (TMO) and/or transition metal dichalcogenide (TMD), such as MoO.sub.3, MoS.sub.2, V.sub.2O.sub.5, WO.sub.3, ReO.sub.3, CrO.sub.3, WS.sub.2, or MoSe.sub.2, for example. Donor layer 70 may be formed by atomic layer deposition or other deposition techniques, such as thermal evaporation, sputtering, or e-beam. The TMO or TMD may be combined with two-dimensional (2D) materials such as graphene, hexagonal boron nitride, or other Vander Waals heterostructures. Alternatively, other hole donor materials (electron acceptors) may be used, or the donor layer may be formed by suitable doping of the diamond itself. A number of possible donor layer configurations are shown, by way of example, in FIGS. 3A-C.

(29) When a suitable voltage is applied to gate electrode 68, it releases carriers from donor layer 70 in the vicinity of defect 54, thus creating a two-dimensional PIN layer structure, comprising an undoped intrinsic semiconductor region between p-type semiconductor and n-type semiconductor regions, over the defect. Depending on the electrode configuration, defect 54 can then be switched between the NV.sup.0 and NV.sup. charge states by switching the polarity of the voltage applied between source electrode 64 and drain electrode 66. In other words, applying a sufficient voltage of the proper polarity will switch defect 54 to the NV.sup. charge state by shifting the positive and negative carrier bands via the PIN layer structure in the diamond crystal and/or in the diamond surface or interface, irrespective of the application of the excitation beam from laser 26. Applying a voltage of the opposite polarity will switch defect 54 to the NV.sup.0 charge state. The polarities and magnitudes of the voltages to be applied to electrodes 64, 66 and 68 for the purposes of charge state switching depend on the geometrical and electrical characteristics of quantum transistor 22.

(30) The electric field due to the voltage applied to gate electrode 68 also modifies the ZPL wavelength of defect 54. Changing the gate voltage over a small range can thus be used to tune the ZPL wavelength, as shown in FIG. 4, for example over a range of 1-2 nm. Different voltages may be applied to different quantum transistors 22 in system 20 to compensate for ZPL variations due to other, uncontrolled local electric fields, and thus cancel the spectral diffusion among the quantum transistors.

(31) In an alternative embodiment, the charge state and ZPL wavelength of a crystal defect, such as NV defect 54, may be controlled by application of appropriate voltages between electrodes 64, 66 and 68 without incorporating a donor layer and PIN structure in the quantum transistor. In this case, acoustoelectric and/or Stark interactions may be applied in controlling the charge state and ZPL wavelength.

(32) FIG. 3A is a schematic sectional view of semiconductor layers overlying the diamond crystal in quantum transistor 22, in accordance with an embodiment of the invention. Donor layer 70 is deposited over a P-type hydrogen surface termination 71 of the diamond in resonator 56. Application of a voltage to gate electrode 68 across donor layer 70 gives rise to Fermi energy level pinning along the heterostructure between the carriers in a two-dimension hole gas (2DHG) 72 in the subsurface diamond, between the donor layer and defect 54 in diamond resonator 56, and the carriers formed in a two-dimensional electron gas (2DEG) 74 on the opposite side of the donor layer. The P-type and N-type layers play the role of acceptor (giving rise to the 2DHG in the diamond subsurface) and donor (giving rise to the 2DEG). These three layers: the P-type and N-type electron gases with the intervening undoped layer, create a two-dimensional PIN structure. Application of a voltage between source electrode 64 and drain electrode 66 will transfer a positive or negative bulk charge from the PIN structure to the diamond bulk, thus switching defect 54 between NV.sup.0 and NV.sup. states, or vice versa.

(33) FIG. 3B is a schematic sectional view of semiconductor layers overlying the diamond crystal in quantum transistor 22, in accordance with another embodiment of the invention. In this embodiment, one part of the diamond in resonator 56, with P-type hydrogen surface termination 71, is overlaid by hole donor layer 70. An adjacent part of the diamond, with an N-type nitrogen surface termination 79, is overlaid by an electron donor layer 80. Application of a voltage to gate electrode 68 will give rise to Fermi energy level pinning between a 2DHG 76 and a 2DEG 78 side by side. In other words, the PIN structure in this case is arranged laterally rather than vertically. This structure can be controlled in a similar manner as the structure of FIG. 3A to change the charge state of defect 54 by switching the source-drain voltage.

(34) FIG. 3C is a schematic sectional view of doped semiconductor layers in diamond overlying the undoped diamond crystal in quantum transistor 22, in accordance with an alternative embodiment of the invention. In this case, the PIN structure over defect 54 is created by bulk doping of the diamond subsurface, creating a diamond P-type layer 82 and a diamond N-type layer 84. In other words, the PIN structure is based on doped multilayers in the diamond itself, serving as the acceptor and donor layers. Layers 82 and 84 can be formed, for example, by heavily doping thin layers near the surface of the diamond crystal with boron and phosphorus dopants. The thickness of layers 82 and 84 is typically in the range of 1-40 nm. Application of a voltage to gate electrode 68 will give rise to bulk charge in the diamond crystal in a manner similar to the embodiment of FIG. 3A.

(35) FIG. 4 is a plot that schematically illustrates emission wavelengths of an NV color center in quantum transistor 22 as function of voltages applied to a electrodes 64, 66 and 68 of the quantum transistor, in accordance with an embodiment of the invention. The horizontal axis shows the voltage (in volts) applied between drain electrode 66 and source electrode 64. When the voltage is positive, in the region to the left of the dashed vertical line, defect 54 remains in the NV.sup.0 charge state and does not emit radiation in the red transition band. (NV.sup.0 emission is in a different range at around 575 nm.) When the drain-source voltage is negative, to the right of the dashed vertical line, defect 54 transitions to the NV.sup. charge state.

(36) In other words, positive drain-source voltage switches quantum transistor 22 off, and negative drain-source voltage switches quantum transistor 22 on, thus enabling qubit operations. Furthermore, continued application of the negative drain-source voltage maintains quantum transistor 22 in the NV.sup. state and prevents transitions to NV.sup.0, thus lengthening the coherence period during which quantum calculations can be carried out.

(37) Each curve 90a, 90b, 90c, . . . , 90d shows the relative emission wavelength as a function of the drain-source voltage for a different value of the voltage applied between gate electrode 68 and source electrode 64. The gate-source voltage increases in increments of 0.2 volts from curve to curve, up to a maximum value of 2.8 volts. In the pictured example, adjustment of the gate-source voltage tunes the ZPL over a range of about 420 m.

(38) At the conclusion of a quantum computation, control circuitry 34 reads out the spin state of NV defect 54 in each quantum transistor 22. As noted earlier, one way to read out the spin states is to ply one or more excitation beams via input waveguides 28 to the quantum transistors at wavelengths chosen so that NV defects 54 output optical signals to optical detectors 44 and to measure the optical signals.

(39) Alternatively, control circuitry 34 may read out the spin states electronically, by a process of spin-to-charge conversion. For this purpose, input waveguides 28 convey one or more excitation beams to NV defects 54 at wavelengths selected to convert one or more electronic spin states of the NV defect to respective charge states. For example, laser beams at green and near infrared wavelengths (such as 532 nm and 1064 nm) may be applied to preferentially excite and ionize the NV.sup. |0> spin state, without ionizing the |1> states. Control circuitry 34 detects the charge state of each NV defect by measuring the current between source electrode 64 and drain electrode 66. High-intensity optical excitation can be applied to read out the photocurrent from each NV defect with high speed and high signal/noise ratio.

(40) Optical Resonators and Filters

(41) Reference is now made to FIGS. 5A and 5B, which schematically illustrate optical structures in waveguides 52, 60 and 62 of quantum transistor 22, in accordance with an embodiment of the invention. FIG. 5A is a side view, showing features of optical resonator 56 and of a filter 98 implemented in waveguide 62. FIG. 5B is a detail view of a periodic structure 94 within resonator 52.

(42) Resonator 56 comprises periodic structures 94 and 96 formed in waveguide 52, serving as Bragg reflectors on opposing sides of defect 54. In the present embodiment, these periodic structures comprise holes 104 extending through the diamond waveguide, as illustrated in FIG. 5B. The holes have a period P, which is chosen to reflect light in a band containing the ZPL wavelength. The holes may be filled with air or with another material, such as SiO.sub.2. For example, for a narrow diamond waveguide with air holes and surrounded by air, P may be approximately /3212 nm for the NV.sup. emission line at 637 nm. To maximize the Purcell enhancement of the ZPL emission, structures 94 and 96 are designed for high Q in the ZPL band, for example Q >2000, and are positioned close together around defect 54 to minimize the optical mode volume V.sub.m. (The Purcell enhancement is proportional to Q/V.sub.m.) At the same time, for efficient collection of the radiation emitted from defect 54 into waveguide 62, as indicated by arrows 102, it is desirable that periodic structure 94 be fully reflective, while structure 96 is only partially reflective.

(43) One or more excitation beams, for example green laser radiation at 532 nm, as well as orange, red, or infrared radiation, are coupled from waveguide 60 into waveguide 52, as indicated by an arrow 100. This green radiation passes freely through periodic structures 94 and 96. To prevent the green radiation from propagating through output waveguide 36 to detector 44 (FIG. 1), an optical filter 98 is formed in waveguide 62. Filter 98 likewise comprises a periodic structure, but with a period selected to reflect the green radiation back toward overlap region 63.

(44) FIG. 6 is a schematic detail view of a periodic structure 106 used in the resonator of a quantum transistor, in accordance with another embodiment of the invention. In this case, periodic structure 106 comprises indentations 108 (creating a fishbone structure), which may not pass all the way through the diamond waveguide as in the preceding embodiment. This sort of periodic structure may be easier to fabricate than structures based on through-holes. Structure 106 may be made longer than structure 94 (FIGS. 5A/B) to compensate for the lower reflectivity of indentations 108 relative to holes 104. (This lower reflectivity is the result of lower overlap between the edge indentations and the center focused mode of the waveguide.)

(45) Although holes 104 and indentations 108 have rectangular profiles in FIGS. 5B and 6, in alternative implementations the holes and indentations may have other shapes, such as round or oval shapes (not shown in the figures).

(46) FIG. 7 is a schematic sectional view of an optical resonator 110 used in a quantum transistor, in accordance with an alternative embodiment of the invention. To form resonator 110, a two-dimensional pattern of holes 114 is etched through a diamond crystal 112. One or more of the holes are omitted or modified to define an optical cavity containing defect 54. Emission from defect 54 at the ZPL is guided in the horizontal directions by reflection from the periodic structure of holes 114 and in the vertical direction by total internal reflection within diamond crystal 112.

Alternative Embodiments

(47) FIG. 8 is a schematic side view of a quantum transistor 120, in accordance with an alternative embodiment of the invention. Components of quantum transistor 120 with similar functions to the components of quantum transistor 22, as described above, are labeled with the same indicator numbers.

(48) Quantum transistor 120 comprises a diamond chiplet 122, which is mounted on a PIC 126. Chiplet 122 contains defect 54 within an optical resonator 124, for example a Bragg resonator as described above, which is oriented vertically relative to PIC 126. Electrodes 64, 66 and 68 are formed on opposing sides of chiplet 122, with an intervening donor layer (not shown in this figure) between one or more of electrodes 64, 66 and 68 and the bulk of chiplet 122. The arrangement of the donor and semiconductor layers creates a PIN structure in chiplet 122 as detailed above in reference to FIGS. 3A-C. One or more waveguides 128 on PIC 126 convey light to and from input waveguide 28 and output waveguide 36. A microwave electrode 130 connects to microwave source 30 (FIG. 1) to create a signal that manipulates the ground spin states of defect 54 in the NV.sup. charge state.

(49) A piezoelectric element 132, for example a layer of a suitable piezoelectric crystal, can be actuated by control circuitry 34 (FIG. 1) to modify the effective length of optical resonator 124. Control circuitry 34 is thus able to adjust the resonant wavelength of resonator 124 by applying an appropriate drive voltage to piezoelectric element 132. This sort of piezoelectric resonator adjustment enables control circuitry 34 to tune the ZPL over a range of roughly 1-50 nm, thus compensating for large spectral diffusion if needed.

(50) FIG. 9 is a schematic side view of a quantum transistor 140, in accordance with yet another embodiment of the invention. Quantum transistor 140 is similar in design to quantum transistor 120 (FIG. 8), with the addition of a reflective surface 144 on a PIC 142, for coupling optical radiation between diamond chiplet 122 and one or more optical waveguides 146 on PIC 142. Reflective surface 144 may be etched into the PIC substrate at a 45 angle and coated for high reflectivity. This sort of coupling may be more efficient than the adiabatic taper-based couplers described earlier.

(51) The embodiments described above are cited by way of example, and the present invention is not limited to what has been particularly shown and described hereinabove. Rather, the scope of the present invention includes both combinations and subcombinations of the various features described hereinabove, as well as variations and modifications thereof which would occur to persons skilled in the art upon reading the foregoing description and which are not disclosed in the prior art.