Aspects of the present disclosure are related to advanced processing systems and more particularly, to quantum processing systems and to methods and systems for initializing and measuring processing elements in such quantum processing systems.
The developments described in this section are known to the inventors. However, unless otherwise indicated, it should not be assumed that any of the developments described in this section qualify as prior art merely by virtue of their inclusion in this section, or that those developments are known to a person of ordinary skill in the art.
Large-scale quantum processing systems hold the promise of a technological revolution, with the prospect of solving problems which are out of reach with classical machines.
To date, a number of different structures, materials, and architectures have been proposed to implement quantum bits (or qubits) and corresponding quantum control and processing systems.
Before such large-scale quantum computers can be manufactured commercially, a number of hurdles need to be overcome. Precise measurement of qubit states at any given time in a quantum processing device is one such essential requirement. Different types of sensors and qubit measurement and initialization techniques have been proposed in the art. Some of these techniques however may be prone to errors.
Accordingly, improved techniques for measuring qubit states and/or initializing qubits in quantum processing systems are desirable.
Reference to any prior art in the specification is not an acknowledgment or suggestion that this prior art forms part of the common general knowledge in any jurisdiction or that this prior art could reasonably be expected to be understood, regarded as relevant, and/or combined with other pieces of prior art by a skilled person in the art.
According to a first aspect of the present disclosure there is a provided a method for measuring the spin of an electron in a quantum dot that is tunnel coupled to a reservoir. The method includes measuring a spin state of the injected electron while applying a ramped detuning for a time period.
In some embodiments, the method further includes emptying the quantum dot.
The method may further include injecting an electron in the quantum dot by applying a constant voltage pulse for a first time period.
In some embodiments, applying the ramped detuning includes continuously ramping electrochemical potentials of the spin states of the injected electron from below the Fermi energy of the reservoir to above the Fermi energy of the reservoir.
In some embodiments, measuring the spin state of the injected electron includes monitoring a charge signal of the quantum dot using a charge sensor.
Further, a blip in the charge signal of the quantum dot before a threshold time may indicate a spin-up state of the injected electron. Alternatively, an increase in the charge signal of the quantum dot after a threshold time may indicate a spin-down state of the injected electron.
In some embodiments, a rate for the injected electron to tunnel out of the quantum dot and into the reservoir increases during the time period and a rate for an electron to tunnel from the reservoir to the quantum dot decreases during the time period.
According to a second aspect of the present disclosure, there is provided a method for initializing a spin qubit in a quantum dot that is tunnel coupled to a reservoir. The method includes injecting an electron in the quantum dot by applying a ramped detuning for a first time period.
In some embodiments, the method further includes emptying the quantum dot.
In some embodiments, the method further includes measuring a spin state of the injected electron while applying a ramped detuning for a second time period
Applying the ramped detuning for the first time period may include continuously ramping electrochemical potentials of the spin states of the injected electron from above the Fermi energy of the reservoir to below the Fermi energy of the reservoir.
In some embodiments, the spin state of the injected electron comprises monitoring a charge signal of the quantum dot using a charge sensor.
In some embodiments, a spin state of the injected electron depends on a rate of the ramped detuning.
In some embodiments, the electron has a spin-down state in response to a slow rate of the ramped detuning.
The method further includes the step of selecting the first time period to control the relative population of the two spin states.
According to a third aspect of the present disclosure, there is provided a method for measuring the spin of an electron in a first quantum dot that is tunnel coupled to a second quantum dot. The measure includes measuring a quantum state of the first and second quantum dots while applying a ramped detuning for a time period.
In some embodiments, the method further includes injecting an electron in the second quantum dot to set the state of the quantum dots to a singlet state such that the first quantum dot includes zero unpaired electrons and the second quantum dot includes two electrons by applying a constant voltage pulse for a first time period.
In some embodiments, the method further includes applying the ramped detuning comprises continuously ramping electrochemical potentials of the spin states of the injected electron from below Fermi energy to above the Fermi energy.
In some embodiments, the method further includes measuring the quantum state of the first and second quantum dots comprises monitoring a charge difference between the first and second quantum dots using a charge sensor.
The method further includes a step in the charge difference between the quantum dots before a threshold time indicates a singlet quantum state.
The method further includes an increase in the charge difference between the quantum dots after a threshold time indicates a triplet zero quantum state.
The method further includes a rate for the injected electron to tunnel out of the second quantum dot and into the first quantum dot increases during the time period.
While the invention is amenable to various modifications and alternative forms, specific embodiments are shown by way of example in the drawings and are described in detail. It should be understood, however, that the drawings and detailed description are not intended to limit the invention to the particular form disclosed. The intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present invention as defined by the appended claims.
The spin states of electrons or nuclei in a semiconductor material are good candidates to carry quantum information and act as quantum bits (or qubits) for a quantum computing system. In order to perform quantum computation however three important steps are required—initialization of qubits, control of qubits, and readout of individual qubits. This present disclosure provides new and improved techniques for two of these steps—qubit initialization and measurement (also referred to as readout in this disclosure).
Single-shot spin readout is an important requirement for error-corrected quantum computing in semiconductor spin qubits.
In some cases, single-shot spin readout is performed by mapping the spin state of a qubit onto a charge state (i.e., spin-to-charge conversion), which can then be detected using a nearby charge sensor such as a single-electron transistor (SET), a single-lead quantum dot (SLQD) charge sensor, a quantum point contact (QPC), a tunnel junction, or a single lead sensor.
One such single-shot readout technique is referred to as energy-selective measurement (ESM). ) state and in
) state. In ESM, a magnetic field (B0) is applied to split the spin-up and spin-down states by the Zeeman energy. The quantum dot potential is then tuned such that if the electron has spin-down (as seen in
The above description of the ESM techniques is described with reference to a system where the spin-up electron has lower energy than the spin-down electron. For example, the measurement technique may be performed in a GaAs (Gallium arsenide) material. In this case, the lower-energy spin-up electron cannot tunnel to the reservoir. In other examples, the spin-up state has higher energy compared to the spin-down state. For example, in a silicon material. In these examples, the spin-down state cannot tunnel to the reservoir.
In some cases, to measure the spin of a qubit using ESM a three stage process (as shown in , its electrochemical potential, μ0↔↓ is below EF, so the electron remains on the dot. If the spin is |↑
, its electrochemical potential, μ0↔↑ is above EF, so the electron tunnels to the reservoir 106. Now Coulomb blockade is lifted and an electron with spin-up can tunnel onto the dot 104. After tread, the pulse ends and the dot is emptied again.
In the read phase, the voltage pulse quickly moves μ0↔↓ and μ0↔↑ such that EF is situated between them. In this position a spin-up electron is more likely to tunnel out of the quantum dot 104 and to the reservoir 106 compared to a spin-down electron. If a spin-up electron tunnels to the reservoir 106 then a spin-down electron can tunnel onto the quantum dot 104 from the reservoir 106 at this stage.
This two-step tunnelling process creates a ‘blip’ in a nearby charge sensor response which is a signature that state-to-charge conversion of the electron spin state has been performed.
Importantly, the density of states in the reservoir 106 are broadened by an effective electron temperature Te with an energy, kBTe where kB is Boltzmann's constant. Therefore, there is some finite probability that a detected blip 122 is actually due to a spin-down electron tunnelling to the reservoir 106, which causes the spin-down electron to be counted as a spin-up. These unintentional spin-down tunnel events are the main source of error when performing ESM at small B0 or at high temperatures.
The ESM described above has allowed for high-fidelity electron spin readout. However, the protocol relies on the precise alignment of both the spin-up and spin-down energy levels μ0↔↓ and μ0↔↑ about EF. The alignment procedure is usually performed by measuring a so-called ‘spin-tail’ by varying the read level from below EF to above EF and then determining the optimal read level by analysing the measurement fidelity. In a large-scale quantum computing architecture such precise alignment may not be feasible for every qubit on a timescale that is allowed by charge noise.
Additionally, ESM requires that the ratio of qubit tunnel rates to be ˜800 corresponding to EZ>13kBTe when μ0↔↓=−μ0↔↑ for high-fidelity readout. However, recent proposals and experiments are investigating the possibility of operating spin qubits at temperatures above 1K to increase cooling power for cryogenic electronics. Therefore, it would be beneficial to have a spin readout technique that can still be used when EZ˜kBTe.
To overcome some of these issues, aspects of the present disclosure provide new qubit readout and initialization methods. These methods are more resilient to charge noise, high-temperatures, and low magnetic field operation. Further, these methods can be easier to scale to a larger number of qubits compared to ESM.
The qubit initialization and measurement methods of the present disclosure are referred to as ramped spin initialization (RSI) and ramped spin measurement (RSM) methods. In the RSM/RSI methods, the tunnel rates of the electron spin states are continuously varied in time rather than being fixed. Further details about the RSI and RSM method will be described in the following sections.
The RSI and RSM methods will be described with respect to donor based quantum processing systems. One such quantum processing device 200 is displayed in
The whole device 200 may be epitaxial—i.e., the donor dot 202 may be fabricated within a substrate (such as a p-type Si substrate (1-10 Ωcm). Positioning the donor dot 202 epitaxially can significantly reduce impact of noise on the qubit. In some examples, the quantum dot 202 is formed about 20-50 nm from the surface.
The quantum dot 202 may be tunnel coupled to a charge sensor 206 (such as a single-electron transistor SET) and electron reservoir 208 to load the electrons 204 onto the donor dot 202. Further, the qubit 212 thus formed may be controlled by one or more gates 210.
Further, the load stage and empty stage are equivalent to the load and empty stages of the ESM method. However, the read phase now involves determining when a blip occurs during the read phase rather than if a blip is detected.
The ramping 401 of the detuning fundamentally changes the way in which the spin-to-charge conversion process operates in RSM. First, the tunnel rates of the |↓ and |↑
states, Γin/out↓ and Γin/out↑ respectively, become time-dependent. Specifically, the tunnel out rates are slow at the beginning of the ramp (generally of a few Hz) 401 and then increase over the duration of the ramp while the tunnel in rates are fast at the start of the ramp (generally less than a MHz) 401 and then decrease over time. These time-dependent tunnel rates are the reason that RSM can outperform ESM.
Secondly, in the RSM technique, the time threshold is redefined. For ESM, the time threshold is equivalent to the readout time and is chosen to determine if a blip 122 is detected. If a blip is detected the qubit state is assigned to be |↑, whereas if no blip is detected before the time threshold then the spin state is assigned to be |↓
. However, for RSM the time threshold is used to threshold when a blip occurs. If a blip 404 is detected before the time threshold then the spin state is assigned to be |↑
and if the blip 404 occurs after then the spin state is assigned as Ii).
Since the electrochemical potentials of both spin states are above EF at the end of the ramp 401, a blip 404 or step is always detected in the charge sensor signal during the read phase; that is, the electron spin has to be in either |↑ or |↓
.
Such a ramp pulse 401 in the read phase is advantageous for a number of reasons. For high temperature or low magnetic field conditions the optimal energy position occurs where both α0↔↓ and μ0↔↑ are below EF and the tunnel rates to the electron reservoir 208 are slow. The slow tunnel rates mean that the readout time becomes impractical. Ramped spin measurement mitigates this long readout time that would otherwise be required by attempting readout at every detuning position for a short amount of time. Therefore, in such situations, the spin measurements can achieve high-fidelity while keeping the readout time short.
Additionally, the alignment of the RSM method 300 is comparatively simple and fast over the traditional ‘spin-tail’ measurement and involves performing the same pulse sequence as for the actual readout protocol.
Finally, the use of ramp pulses 401 significantly reduces the bandwidth requirements for fast pulsing which could possibly lead to lower electron temperatures since additional filtering can be implemented.
electrons tunnel to the SET and a |↓
electron tunnels back onto the quantum dot 202, respectively.
For comparison,
The resulting histogram is shown in or |↓
(aligned with the dashed lines in
and the longer events are associated with a spin |↓
. Additionally, by monitoring the position of the spin down |↓
distribution using a moving average, the threshold time can be adjusted either in real-time or in post-processing as shown by line 508 in
The bimodal distribution of the measured signal in
During RSM, the detuning is ramped across the Fermi broadened electron transition during readout from an initial detuning point ϵ0 relative to the centre of the transition with a ramp rate of r=ϵramp/tramp where ϵramp is the change in qubit detuning relative to the reservoir 208, and tramp is the duration of the ramp 401. The rate function is given by:
where Γ is the maximum tunnel rate of the electron transition and kB is the Boltzmann constant.
The tunnelling event distributions during RSM are equivalent to the probability density function (PDF) f(t) which for a NHP process can be calculated as:
f(t)=λ(t)exp(−∫0tλ(y)dy (2)
Substituting the rate equation from Equation 1 into Equation 2 results in the general normalised PDF for RSM as
Using this equation it is possible to analytically model the electron system during RSM.
The fidelity of RSM can be quantified by the spin up and spin down fidelities which are the proportion of electrons correctly identified for each spin state respectively.
Given that for RSM a time threshold is used to distinguish between the two spin states, and hence their respective PDFs, the spin up fidelity can be described in terms of the normalised cumulative density function (CDF) F↑ and the spin down fidelity can be described in terms of the CDF F↓. In general, for a given time T, the CDF is defined as
F(T)=∫0Tf(t)dt (4)
Hence, by substituting in Equation 3, this results in the specific CDF F(t) for RSM as
For a given value of the time threshold t, the spin up fidelity is given as F↑(t), and the spin down fidelity is given as 1−F↓(t). The fidelities can then be combined into a single metric, the visibility V as defined by:
which can be used for optimising the fidelity.
To optimise and calculate the fidelity of RSM the respective PDFs of the spin up and down electron states are first determined. Firstly, the PDF for the spin-up (spin-down) electrons is scaled by the proportion of spin up (down) electrons N↑(↓). Additionally, when in a global magnetic field, the two electron spin states are separated in energy by the Zeeman energy splitting EZ. The energy separation results in the tunnelling transition off the quantum dot 202 to be centred around a different detuning position resulting in an offset (equal to the Zeeman energy) in ϵ0 for each electron spin state. An example for the PDFs for the spin up and spin down electrons tunnelling off the quantum dot 202 are calculated and presented in
To determine the measurement fidelity of the RSM using time thresholding, the optimal signal value and time value need to be determined for determining when a |↑ or |↓
is detected. First, with respect to the optimal time threshold topt, the maximal visibility 602 is found when the following condition is met
Solving for topt it is found that
Together, these equations allow the analytic determination of the optimal time threshold and measurement fidelity for a given set of RSM parameters.
In
Next,
To validate the initialization process, a subsequent RSM method 300 may be performed to measure the spin-up population of the quantum dot 202. In particular, . From previous calibration of the measurement fidelity and MT, the initialisation fidelity can be estimated using Eq. 13 below. Interestingly, the relative population of the two spin states can be controlled continuously from ˜50% to ˜0% by changing the ramp time (r/Γ0˜10−3 to r/Γ0˜10−5), which could be used for initialising different spin states.
Initialisation fidelity, Finit can be defined as the accuracy at which an electron can be deterministically initialised in a spin-down state |↓. To accurately determine Finit the measurement fidelity is taken into account since that contributes to the number of spin-up states that can be correctly identified. The experimentally measured spin-up count, M↑ is given by,
M
↑=(1−Finit)F↑+Finit(1−F↓) (12)
which can be rearranged to obtain,
Using the data from
The expected initialisation fidelity Finit can also be theoretically calculated for a range of tramp values using the previously derived PDF (Equation 3) and the experimental fit parameters. The relevant electron tunnelling rates between the reservoir 208 and quantum dot 202 for initialisation are the complement of the relevant tunnelling rates for readout. This symmetry means that both initialisation and readout can be described by the same NHP process as long as a polarity change is made to the detuning axis. Therefore, f↑/(↓)(t) describes the PDF of a single spin up (down) electron tunnelling from the reservoir 208 to the quantum dot 202.
The biggest difference between the two processes is that initialisation involves many possible electrons in the reservoir of which only one (or none) is initialised onto the quantum dot. To account for this difference when calculating the probability of initialising a spin-up (down) electron P↑(↓)init(t), the probability that an electron of the opposite spin had not already tunnelled onto the quantum dot must be considered. Hence,
where the initialisation fidelity is equal to the total probability that a spin down electron was initialised within the ramp time, i.e.,
F
init=∫0t
Using Equation 16 it is possible to predict the required ramp time to initialise a certain level of initialisation fidelity given a set of experiment parameters.
Accordingly, aspects of the present disclosure propose and demonstrate a readout technique for semiconductor spin qubits that can be used to measure the spin of qubits in currently implemented and future implementations of quantum processing/computing systems. The readout protocol offers a number of advantages over known readout techniques including simple alignment (with feedback), short measurement/initialisation times, and reduced gate bandwidths.
Using the measurement technique, the present disclosure shows that high-fidelity readout can be readily performed at low magnetic field to temperature ratios, which is crucial for high temperature qubit operation. Finally, the reverse pulse sequence can be used to deterministically load the |↓ with greater than 99% fidelity or controllably load a particular mixture of |↓
and |↑
states. The RSM advantages over ESM will allow for simple single-shot, low magnetic field measurements that can be readily scaled to large qubit systems.
Ramped Readout with Two Quantum Dots
In the previous sections, the ramped initialization and readout methods were described with respect to a quantum dot and a reservoir—i.e., based on tunnel coupling between a quantum dot and a reservoir. These ramped methods can be extended to include spin-state selective readout between pairs of quantum dots.
In both and |↑
respectively. Combining the two spin states of each electron yields four new states. There is one singlet state with total spin angular momentum 0. This state is:
Where the first arrow and second arrow in each combination indicate the spin direction of the first and second electron 1003, 1004 respectively. There are three triplet states, each with total spin angular momentum 1:
Here, the T+ triplet state corresponds to both electrons 1003, 1004 being in the spin-up state and the T− triplet state corresponds to both electrons 1003, 1004 being in the spin-down state.
In =|↑↑
state. In this case, no tunnelling is permitted. That is, electron 1003 cannot move to quantum dot 1002 and thus will remain on quantum dot 1001. There is also no tunnelling allowed if the electrons occupy the other triplet states: |T−
=|↓↓
or |T0
=|↑↓
+|↓↑
.
In =|↑↓
−|↓↑). In this state, when the potential of quantum dot 1002 is detuned below the potential of quantum dot 1001, the electron 1003 can tunnel to quantum dot 1002 as the spin states of the electrons on the two quantum dots are opposite.
In some cases, to measure the state of a qubit using PSB, a three stage process (as shown in
The different stages are typically controlled by a three-level voltage pulse on nearby gate(s). The voltage pulse shifts the quantum dots' energy levels relative to each other. The shape of the voltage pulse in this case is the same as that shown in
Before a pulse is applied, both electrons 1003, 1004 are in quantum dot 1001. The electrons can either be in the Singlet, |S (0,2), or Triplet-0, |T0 (0,2)
, states. After sufficient time, the Triplet-0 state relaxes to the Singlet ground state.
Then at the second stage a voltage pulse (called the load phase) is applied for time twait. This pulse pulls the energy levels of both spin states below the fermi energy EF. An electron 1003 can now energetically tunnel into the dot 1002 from the reservoir 106. The quantum state of the electrons 1003 and 1004 after the tunnelling can either be a Singlet or a Triplet-0 state.
During this stage 1010, the electron 1003 (1004) is trapped on the dot 1001 (1002) and the Coulomb blockade prevents a second electron from being added to the dots 1001 (1002).
After twait, the voltage pulse is increased and maintained for a duration tread, in order to position the energy levels of the Singlet and Triplet-0 states in a read-out configuration.
If the quantum state is Singlet, then the electrochemical potential of the Singlet states with one electron in each dot 1001 and 1002, |S (1,1), and with two electrons in 1002, |S (0,2)
, are equal, so the electron 1003 can tunnel between dots 1001 and 1002. If the quantum state is one of the triplet states, e.g., |T0
, the electrochemical potential of the Triplet-0 state with one electron in each quantum dot 1001 and 1002, |T0 (1,1)
, is lower than of the Triplet-0 state with two electrons in quantum dot 1002, |T0 (0,2)
, so the electron 1001 remains on quantum dot 1001—no tunnelling occurs.
After tread, the pulse ends and the Triplet-0 state relaxes to the Singlet state again. In the read phase, the voltage pulse quickly moves S11 and S02 such that they are equal. At this voltage level if the electrons 1003 and 1004 are in a Singlet state then the position of electron 1003 becomes a superposition of being in quantum dot 1001 and 1002.
This superposition creates a ‘step’ in a nearby charge sensor response which is a signature that state-to-charge conversion of the quantum state has been performed.
Aspects of the present disclosure provide a new method to perform readouts between two quantum dots. This method is more resilient to charge noise, high-temperatures, and low magnetic field operation. Further, this method can be easier to scale to a larger number of qubits compared to ESM.
Further, the load stage and empty stage are equivalent to the load and empty stages of the PSB method. However, the read phase now involves determining when a step occurs during the read phase rather than if a step is detected.
Further still, in the RSM technique, the time threshold is redefined. For PSB, the time threshold is equivalent to the readout time and is chosen to determine if a step 1016 is detected. If a step is detected, the qubit state is assigned to be in the |S state, whereas if no step is detected before the time threshold then the qubit state is assigned to be |T0
. However, for RSM, the time threshold is used to determine when a step occurs. If step 1104 is detected before the time threshold then the qubit state is assigned to be |S
and if the step 1106 occurs after the threshold, then the qubit state is assigned as |T0
.
The optimal time threshold topt scales proportionately with Γ0. For a given device and parameter values one can find the non-homogeneous Poisson distributions for each spin state (up and down). As such, topt is then the time threshold that distinguishes between these 2 distributions with the optimal fidelity.
In order to measure the state of a qubit using this method, a three stage process (as shown in
Since the electrochemical potentials of |S (1,1) and |T0 (1,1)
spin states are above those of |S (0,2)
and |T0 (1,1)
at the end of the ramp 401, a step 1104 or 1106 is always detected in the charge sensor signal during the read phase; that is, the quantum state has to be in either |T0
or |S
. That is, the spin state may be measured at step 1130.
Finally, the use of ramp pulses 1202 significantly reduces the bandwidth requirements for fast pulsing which could possibly lead to lower electron temperatures since additional filtering can be implemented.
At step 1204, the detuning between the dots is changed across the (0,2)→(1,1) transition, allowing one electron 1003 to tunnel from the right quantum dot to the left quantum dot resulting in the electrons having an unknown quantum state |S (1,1) or |T0 (1,1)
.
Next, at step 1206, the quantum state of the electrons 1003 and 1004 is measured by applying a ramped voltage. If a step is detected in the charge difference between the two dots before a threshold period of time, it is determined that the quantum dots were in the |S (1, 1) state and if the step 1106 is detected in the charge difference between the two dots after a threshold period of time, then it is determined that the quantum dots were in the |T0 (1,1)
state. These steps of the RSM method 1200 are controlled by applying voltage on a nearby gate 210, which shifts the energy levels of the quantum dots 1001 and 1002 around each other. The time threshold at which any events that occur before it during readout are assigned to the singlet state, and any time after the time threshold are assigned to a triplet.
The term “comprising” (and its grammatical variations) as used herein are used in the inclusive sense of “having” or “including” and not in the sense of “consisting only of”.
It will be appreciated by persons skilled in the art that numerous variations and/or modifications may be made to the invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.
Number | Date | Country | Kind |
---|---|---|---|
2021904192 | Dec 2021 | AU | national |
2022228109 | Sep 2022 | AU | national |