04.09.2024, Wednesday, 9:0010:40



Coherent Control of a Triangular Exchangeonly Spin Qubit
Joseph Broz^{1}, Edwin Acuna^{1}, Jason Petta^{2}
^{1}HRL Laboratories, ^{2}Department of Physics and Astronomy, University of California  Los Angeles
Abstract: We investigate a triple quantum dot (TQD) with the dots arranged in a closepacked triangular geometry. This includes measurements of the charge stability in the fewelectron regime as well as the dynamical characterization of a threeelectron exchange only (EO) qubit encoded into TQD, which serves as a proxy for the coherent performance of the joint electronic spin state of the dots. In all cases, we find that performance is comparable to stateoftheart demonstrations of EO qubits in linear dot arrays. Thus, our results represent a step towards scaling up quantum dot arrays to larger and more complex structures.






Quantum operations and statistics of a dense 10 spin qubit array
Valentin John^{1}, Cécile Yu^{1}, Stefan Oosterhout^{2}, Lucas Stehouwer^{1}, Floor van RiggelenDoelman^{1}, Maximilian RimbachRuss^{1}, Stefano Bosco^{1}, Giordano Scappucci^{1}, Francesco Borsoi^{1}, Menno Veldhorst^{1}
^{1}QuTech, Delft University of Technology, ^{2}QuTech, TNO
Abstract: The study of qubits encoded in single spins has so far been limited to small quantum systems, predominantly arranged in 1D lattice configurations and small 2D arrays. While scaling further in two dimensions would be critical for the implementation of error correction schemes, these efforts have been hindered by challenges in material uniformity and noise, in the design and fabrication of a dense gate layout, and in the implementation of efficient tuning strategies.
Here, we investigate an extended 10qubit system defined on a twodimensional array of 10 quantum dots in the fewhole regime. Quantum dots are defined on a Ge/SiGe heterostructure grown on a Ge substrate and subjected to an inplane magnetic field of a few tens of MHz.
We perform an indepth characterisation of the system obtaining statistics in various properties of each of the 10 qubits in the same condition. Our study first encompasses gfactors and coherence times (T2*), finding distributions with variabilities of 6% and 9%, respectively. We then obtain the EDSR driving strength of each qubit when driven by 22 different gates. To understand our results, we model our system by including capacitive and spinorbit couplings, and obtain insights on the locality, directionality and potential crosstalk of EDSR in such a dense qubit array. By driving each qubit with the most efficient gate at fixed Rabi frequencies, we benchmark the singlequbit gate performance obtaining fidelities all above 99.4%. In this talk, we will also present to the community some of the challenges encountered in tuning and calibration of our qubit system, and discuss steps forward.






Junction element for twodimensional connectivity of spin qubits in Ge/SiGe
Inga Seidler, Konstantinos Tsoukalas, Leonardo Massai, Felix Schupp, Matthias Mergenthaler, Gian Salis, Andreas Fuhrer, Patrick HarveyCollard
IBM Research Europe  Zurich
Abstract: Current spin qubit devices are mostly based on linear qubit arrays, allowing only for couplings to two neighboring qubits. Holes in Ge/SiGe heterostructures offer the perfect platform for twodimensional (2D) quantum dot (QD) arrays, thanks to the high confinement quality, high tunability, and larger gate structure dimensions. We implement a junction element in a Y geometry, which couples three double quantum dots via one mediator quantum dot at the intersection (Fig. a). The strong latching effects arising from the absence of reservoirs are mitigated by tuning techniques based on constant charge occupation of the QD array. We show single hole occupation for all seven quantum dots simultaneously and tunnel rate tunability.
Towards the goal of operating this junction as a spin coupling mechanism, we show manipulation of two qubits in one of the arms of the junction (Fig. b). We measure the angular dependence of the qubit frequencies on the applied magnetic field and find that the principal axes of the gtensor align with the axis of the junction arm at a 150 degree angle, as opposed to one of the main crystal directions (Fig. c). This reinforces the interpretation that the gtensor properties are determined by electrostatic confinement or strain commensurate with the QD layout.
With the goal of implementing threeway and 2D qubit connectivity, we will show the progress towards linking qubits in different arms of the junction.






FourSpin Qubit Chain in SiMOS with Independent Parity Readout
Cameron Jones^{1}, Santiago Serrano Ramirez^{2}, Jonathan Yue Huang^{3}, MengKe Feng^{4}, Nard Dumoulin Stuyck^{2}, Tuomo Tanttu^{2}, Wee Han Lim^{2}, Andre Saraiva^{2}, Arne Laucht^{3}, Andrew Dzurak^{2}, ChihHwan Yang^{3}
^{1}University of New South Wales, ^{2}UNSW & Diraq, ^{3}UNSW Sydney, ^{4}University of New South Wales / Diraq
Abstract: To date, high fidelity single and two qubit gates have been routinely achieved in SiliconMOS (SiMOS) devices. Demonstration of qubit entanglement with more than two qubits in such devices has however remained elusive. In this work we present a fourqubit device formed by a linear array of SiMOS quantum dots, each populated by an electron spin qubit, as seen in figure 1.a. An onchip antenna allows for electron spin resonance for single qubit operations, and electrode gates between dots provide tuneable exchange coupling required for twoqubit operations. With this device, we demonstrate coherent single qubit operation of all four qubits, as well as pairwise exchange coupling between each nearest neighbour qubit pair. This allows for universal qubit control of the fourqubit processor. Additionally, we are able to achieve high fidelity initialisation of all qubits, and independent parity readout of the Q1Q2 and Q3Q4 pairs in a single measurement shot. Finally, we demonstrate the generation of a 3qubit Greenberger–Horne–Zeilinger (GHZ) state. By performing quantum state tomography on the entangled state, we reconstruct the density matrix shown in figure 1.b. From these measurements, we calculate a Mermin witness value for the state that is greater than the classical limit of 2. This shows successful 3 qubit entanglement in a SiMOS device, using two pairs of parity readout.






Grover's algorithm in a fourqubit silicon processor above the faulttolerant threshold
Ian Thorvaldson^{1}, Dean Poulos^{2}, Christian Moehle^{2}, Saiful Misha^{2}, Hermann Edlbauer^{2}, Jonathan Reiner^{2}, Helen Geng^{2}, Benoit Voisin^{2}, Michael Jones^{2}, Matthew Donnelly^{1}, Luis Pena^{2}, Charles Hill^{1}, Casey Myers^{1}, Joris Keizer^{1}, Yousun Chung^{2}, Samuel Gorman^{1}, Ludwik Kranz^{1}, Michelle Simmons^{1}
^{1}Silicon Quantum Computing, CQC2T, ^{2}Silicon Quantum Computing
Abstract: Spin qubits in silicon are strong contenders for realizing a practical quantum computer. This technology has made remarkable progress with the demonstration of single and twoqubit gates above the faulttolerant threshold and entanglement of up to three qubits. However, maintaining high fidelity operations while executing multiqubit algorithms has remained challenging due to challenges of crosstalk. In this talk, we present results of a fourqubit silicon processor with every operation above the fault tolerant limit and a demonstration of Grover's algorithm with a ~95% probability of finding the marked state (see Fig. a), one of the most successful implementations to date [1]. Our fourqubit processor is made of three phosphorus atoms and one electron spin precisionpatterned into 1.5 nm^{2} isotopically pure silicon (see Fig. b). The strong resulting confinement potential, without the need for confinement gates reduces crosstalk and leverages the benefits of alltoall connectivity of the nuclear spins provided by the hyperfine interaction. This not only allows for efficient multiqubit operations, but also provides individual qubit addressability. Together with the long coherence times of the nuclear and electron spins, this results in all four single qubit fidelities above 99.9% and controlledZ gates between all pairs of nuclear spins above 99% fidelity. The high control fidelities, combined with >99% fidelity readout of all nuclear spins, allows for the creation of a threequbit GreenbergerHorneZeilinger (GHZ) state with 96.2% fidelity (see Fig. c), the highest reported for semiconductor spin qubits so far. Such nuclear spin registers can initialised with high fidelity [2] and coupled via electron exchange [3], establishing a path for larger scale faulttolerant quantum processors.
[1] I. Thorvaldson et. al., arXiv:2404.08741 (2024)
[2] J. Reiner et. al., Nat. Nanotechnol. (2024)
[3] L. Kranz et. al., Phys. Rev. Appl. 19, 024068 (2023)





04.09.2024, Wednesday, 11:1012:50




Lightinduced offset charge in Si/SiGe quantum dots as a proxy for radiation impacts
Brighton Coe^{1}, Michael Wolfe^{2}, Jared Benson^{2}, Tyler Kovach^{2}, Alysa Rogers^{2}, Deanna Campbell^{3}, Spencer Weeden^{2}, Robert Mcdermott^{2}, Gabriel Bernhardt^{2}, Donald Savage^{4}, Max Lagally^{4}, Shimon Kolkowitz^{5}, Mark Eriksson^{2}
^{1}University of Wisconsin Madison, ^{2}University of Wisconsin Madison Department of Physics, ^{3}Sandia National Laboratories, ^{4}University of Wisconsin Madison Department of Material Science and Engineering, ^{5}University of California Berkeley Department of Physics
Abstract: Stateoftheart quantum error correction codes cannot correct for correlated errors. Recently, it has been shown that radiation impacts induce correlated errors in superconducting qubits. An important open question for semiconductor qubits is whether electronhole pairs induced by radiation impacts in the bulk of the chip can propagate through the SitoSiGe interface and induce offset charge shifts in gate defined Si/SiGe quantum dot qubits.
Here, we imitate such radiation impacts in Si/SiGe quantum dot devices using a fiber optic connection in a 3K cryogenic refrigerator to deposit energy from multiple 1.6 eV photonimpacts on the back side of the host silicon substrate. We show that such photon impacts shift the Coulomb blockade behavior of gatedefined quantum dots at the surface of the chip. We track multiple offset charge shifts from a series of photon bursts by using active feedback to sit on the side of a Coulomb blockade peak. Using this technique, we uncover a strong correlation between photon bursts and observed offset charge shifts.
Using G4CMP we simulate the propagation of large numbers of electronhole pairs through the substrate. Surprisingly, but consistent with the experimental data, we find that electronhole pairs induced by illuminating on the backside of the wafer can propagate long distances and impact operation of quantum dot qubits on the front side of the wafer. This finding is important because radiation impacts would likely occur in the bulk of the wafer, and electronhole pairs will affect the qubits if they propagate upwards to surface or nearsurface regions of the Si/SiGe heterostructure.
We observe such offset charge shifts in experiments on devices from different growth systems and with different substrate thicknesses: one grown at UWMadison with fabrication done in part at Sandia National Laboratories, and the other an Intel Tunnel Falls device.






Noise Correlations in a Silicon FiveQubit Array
Leon Camenzind^{1}, YiHsien Wu^{1}, Juan RojasArias^{1}, Akito Noiri^{1}, Kenta Takeda^{1}, Takashi Nakajima^{1}, Takashi Kobayashi^{2}, Ik Kyeong Jin^{1}, Amir Sammak^{3}, Peter Stano^{1}, Giordano Scappucci^{4}, Daniel Loss^{5}, Seigo Tarucha^{6}
^{1}RIKEN Center for Emergent Matter Science (CEMS), Wako, Japan, ^{2}RIKEN Center for Quantum Computing (RQC), Wako, Japan, ^{3}QuTech, Delft University of Technology, Delft, The Netherlands & Netherlands Organization for Applied Scientific Research (TNO), Delft, The Netherlands, ^{4}QuTech, Delft University of Technology, Delft, The Netherlands & Kavli Institute of Nanoscience, Delft University of Technology, Delft, The Netherlands, ^{5}RIKEN Center for Emergent Matter Science (CEMS), Wako, Japan & RIKEN Center for Quantum Computing (RQC), Wako, Japan & Department of Physics, University of Basel, Basel, Switzerland, ^{6}RIKEN Center for Emergent Matter Science (CEMS), Wako, Japan & RIKEN Center for Quantum Computing (RQC), Wako, Japan
Abstract: Understanding the noise environment of semiconductor qubits is crucial for advancing spinbased quantum computing. While quantum error correction techniques can effectively address spontaneous, uncorrelated errors, the presence of correlated noise poses significant challenges to these protocols.
In our Silicon28/SiliconGermanium spin qubits, the surrounding environment of semiconductors and oxides introduces sources of twolevel systems (TLS), which, through the micromagnets used for spin manipulation, convert electrical noise from local TLSs into qubit energy noise (Fig. a). Importantly, these TLS can be shared among qubits, resulting in correlated noise and, thus, correlated quantum gate errors. Here, we report on qubitqubit correlated noise in a fivequbit array.
We measure individual qubit energy fluctuations through Bayesian estimations on Ramsey sequences and additionally measure energy fluctuations of two flanking charge sensors. Our methodology facilitates a comprehensive exploration of the noise environment by enabling the characterization of individual qubit noise power spectral densities and crosscorrelation power spectral densities. This reveals distinct noise profiles and correlation spectra influenced by local TLS environments in the array.
For an integration time of 2.7h, our five qubits show T_{2}^{*} around 45 μs matching expectations from the measured noise spectral densities. We observe relatively strong correlations between neighboring qubits, with a notable reduction for nextneighbors and only minimal correlations to thirdneighbor qubits (Fig. b). This trend is confirmed in the qubitsensor correlations. Further, we can alter the position of our qubits by gate voltages and demonstrate a polynomial decay of noise correlations with qubitqubit spatial separation.
Our findings indicate an associated correlation length between 100 and 200nm, thus underscoring the feasibility of building large arrays of spin qubits for Siliconbased quantum computing architectures. The observed short correlation distance and the potential to mitigate correlations through local gating hold significant promise for scalable implementation of quantum error correction codes.






Extracting noise crossspectra from singleshot measurements
Juan RojasArias^{1}, Peter Stano^{1}, Daniel Loss^{2}
^{1}RIKEN, ^{2}University of Basel, RIKEN
Abstract: As qubit arrays continue to scale in size, there is a growing need to include the analysis of crosscorrelations in the toolbox for studying noise in quantum processors. In this work, we introduce a novel approach for extracting the noise cross power spectral density (crossPSD) of a qubit pair experiencing dephasing. Our method leverages singleshot readouts from a Ramseytype experiment (Fig. 1a), offering resilience against errors in state preparation and measurement. Notably, this approach allows to perform spectroscopy across a wide frequency range, extending the spectral range when compared to current methods. In Fig. 1b we present the successful implementation of the method to simulated data, where both the magnitude (upper panel) and phase (lower panel) of a predefined nonmonotonic crossPSD are properly extracted.






Using valley relaxation hotspots to boost spinshuttling fidelity in Si quantum wells
Merritt Losert^{1}, Rajib Rahman^{2}, Lars Schreiber^{3}, Susan Coppersmith^{2}, Mark Friesen^{1}
^{1}University of WisconsinMadison, ^{2}University of New South Wales, ^{3}JARA Institute for Quantum Information
Abstract: Highly variable valley splittings in Si/SiGe heterostructures pose a challenge for highfidelity shuttling of Si spin qubits. Regions of low valley splitting lead to valley excitations, which in turn cause dephasing of the spin qubit. In this work, we propose a scheme to strongly enhance the spinshuttling fidelity by making use of valleyrelaxation hotspots. In contrast with conventional devices, where valley relaxation rates are typically in the range of 0.1 – 10 kHz, we show here that much higher relaxation rates of 50 MHz or more can be achieved in structures where the electron overlaps significantly with Ge, such as narrow quantum wells or wells containing Ge. Such hotspots are prevalent in regions with large valley splittings and large intervalley dipolar matrix elements, arising from SiGe randomalloy disorder. Here, we derive analytical models for the statistical distributions of valley splittings, dipolar matrix elements, and relaxation rates due to alloy disorder, and we verify these models with tightbinding simulations. Then, by performing simulations of spin shuttling in the presence of alloy disorder, we show how these hot spots can reduce the shuttling infidelities, with average shuttling infidelities near 0.1 – 0.01%, for shuttling velocities between 1 – 10 m/s and shuttling trajectories of 5 microns. This is in contrast with shuttling infidelities near unity for conventional heterostructures, over the same parameter regime. Our work therefore provides an effective and simple solution to the a key obstacle for spin shuttling in Si.






Impact of growth front segregation and postgrowth annealing on the valley energy splitting of spin qubits in silicon heterostructures
Jan Klos^{1}, Jan Tröger^{2}, Jens Keutgen^{3}, Merritt Losert^{4}, Helge Riemann^{5}, Nikolay Abrosimov^{5}, Joachim Knoch^{6}, Hartmut Bracht^{7}, Susan Coppersmith^{8}, Mark Friesen^{4}, Oana CojocaruMirédin^{9}, Lars Schreiber^{10}, Dominique Bougeard^{11}
^{1}JARAFIT Institute for Quantum Information, Forschungszentrum Jülich GmbH & RWTH Aachen University, Aachen, Germany, ^{2}Institute of Materials Physics, University of Münster, Münster, Germany; Tascon GmbH, Münster, Germany, ^{3}I. Physikalisches Institut IA, RWTH Aachen University, Aachen, Germany, ^{4}University of WisconsinMadison, Madison, Wisconsin, USA, ^{5}LeibnizInstitut für Kristallzüchtung (IKZ), Berlin, Germany, ^{6}Institute of Semiconductor Electronics, RWTH Aachen University, Aachen, Germany, ^{7}Institute of Materials Physics, University of Münster, Münster, Germany, ^{8}University of New South Wales, Sydney, Australia, ^{9}I. Physikalisches Institut IA, RWTH Aachen University, Aachen, Germany; INATECH, AlbertLudwigs Universität Freiburg, Freiburg im Breisgau, Germany, ^{10}JARAFIT Institute for Quantum Information, Forschungszentrum Jülich GmbH & RWTH Aachen University, Aachen, Germany; ARQUE Systems GmbH, Aachen, Germany, ^{11}Institut für Experimentelle und Angewandte Physik, Fakultät für Physik, Universität Regensburg
Abstract: We present isotope concentration depth profiles of a ^{28}Si/SiGe quantum well (QW) heterostructure analyzed with atom probe tomography (APT) and timeofflight secondaryion mass spectrometry. The profiles are then used as an input for a tightbinding model to predict realistic valley energy splittings. We have experimentally observed spinecho dephasing times T_{2} =128 μs and valley energy splittings E_{VS} around 200 μeV for single spin qubits in this molecular beam epitaxy (MBE) QW heterostructure previously. With APT, we find the concentration of nuclear spincarrying ^{29}Si to be 50 ppm in the ^{28}Si QW.
The resolution limits of APT allow to uncover that both the top SiGe/^{28}Si and the bottom ^{28}Si/SiGe interfaces of the asgrown QW are shaped by epitaxial growth front segregation signatures on a few monolayer scale.
A post growth thermal treatment of the heterostructure  representative of the thermal budget experienced during qubit device processing – additionally indicates minimal thermallydriven, isotropic bulk diffusion, inducing a widening of the top SiGe/^{28}Si QW interface by about two monolayers, while the width of the bottom ^{28}Si/SiGe interface remains unchanged.
The tightbinding model including SiGe alloy disorder and the experimental APT concentration, suggests that the subtle combination of the slight thermally driven post growth diffusion and of a minimal Ge concentration around 0.3 % in the QW, as a result of a bottom ^{28}Si/SiGe QW interface segregation trailing edge, is instrumental for the observed large valley splitting of E_{VS}=200 μeV and the predicted probability of 62% to find E_{VS}>100 μeV in the annealed heterostructure.
Minimal Ge additions < 1 % hence seem to support high EVS without compromising coherence times. Note, that the probability to induce very small Ge additions into the QW during epitaxy gets more likely in thin QWs with diffused interfaces, which are used more and more for spin qubit devices.





04.09.2024, Wednesday, 14:2016:00




Onchip Artificial Atomic Parametric Amplifier based on Semiconductor Quantum Dots
Yongqiang Xu, Rui Wu, Gang Cao, Guoping Guo
University of Science and Technology of China
Abstract: In circuit quantum electrodynamics (circuit QED), superconducting microwave resonator is an important means for qubit readout. As one of the pivotal functional component, the parametric amplifier significantly enhances the performance of circuit QEDbased readout. Quantum dot as an artificial atom, can serve as an ideal platform for the realization of an ultimate miniaturized parametric amplifier. It facilitates high fidelity readout, with advantages in convenient integration, strong tunability and resilience against magnetic fields.
Here, based on the interaction between a microwave cavity and a GaAs double quantum dot (DQD), we develop an artificial twolevel atomic parametric amplifier. Harnessing the intrinsic nonlinearity of the DQD, a parametric gain of transmission exceeding 11~dB is achieved. Specifically, we exploit this onchip amplifier to read out the other integrated DQD, demonstrating a significant enhancement in readout performance with a threefold increase in signaltonoise ratio (SNR). Our results open a new avenue to develop onchip quantum technologies for weak signal measurement.






Operating a coherent spin qubit allelectrically with highQ gatereflectometry for fast readout
Rafael Eggli^{1}, Taras Patlatiuk^{1}, Toni Berger^{1}, Eoin Kelly^{2}, Alexei Orekhov^{2}, Gian Salis^{2}, Richard Warburton^{1}, Dominik Zumbühl^{1}, Andreas Kuhlmann^{1}
^{1}University of Basel, ^{2}IBM Research Zurich
Abstract: Combining all electrical spin manipulation and insitu gatedispersive spin readout promises to be the road to high density, large scale spin qubit processors. The rapidly progressing hole spin qubit platforms in silicon and germanium provide purely electrical spin control at recordbreaking speed. Superconducting offchip inductors have been introduced to enhance the dispersive response by boosting the internal quality factor (Q) of the resonator and have enabled fast, high fidelity spin readout. However, demonstrating electrical spin manipulation in the presence of a highQ resonator has been a longstanding challenge. A recent report (Kelly et al., APL, 2023) on the inadvertent ringup of a superconducting highQ dispersive sensor circuit caused by capacitive cross talk suggest resonator ringing to be a major obstacle when implementing electrical driving with such resonators, potentially limiting qubit coherence if qubit drive sequences spectrally overlap with the tank resonance frequency. This especially affects systems with strong spinorbit interaction like silicon holes.
Here, we report on a silicon fin fieldeffect transistor hole spin qubit integrated with a niobium nitride nanowire inductor and coherently controlled allelectrically at 1.5 K. We investigate the mechanism by which resonator ringup impacts qubit coherence and initialisation, by measuring the qubit in transport and contrasting data taken with first a lowQ wire wound surfacemount inductor and finally a highQ inductor connected to the identical device and gate. We find a large parameter space for which resonator ring up causes a significant reduction in initialisation/readout efficacy, suggesting that primarily state preparation and measurement (SPAM) errors are introduced. Importantly, we find that the ring up does not limit our coherence time, indicating that efficient highQ resonators in gate sensing are compatible with allelectrical spin control. These findings support the vision for largescale, dense and hot qubit arrays based on allelectrical control with cointegrated gatedispersive highQ readout capabilities.






Spin Qubits with Scalable millikelvin CMOS Control
Sam Bartee^{1}, Will Gilbert^{2}, Kun Zuo^{1}, Kushal Das^{3}, Tuomo Tanttu^{2}, Chih Huan Yang^{2}, Nard Stuyck^{2}, Sebastian Pauka^{3}, Rocky Su^{4}, Wee Han Lim^{2}, Santiago Serrano^{2}, Christopher Escott^{2}, Fay Hudson^{2}, Kohei Itoh^{5}, Arne Laucht^{2}, Andrew Dzurak^{2}, David Reilly^{3}
^{1}The University of Sydney, ^{2}Diraq, ^{3}Microsoft Quantum Sydney, ^{4}UNSW, ^{5}Keio University
Abstract: A key virtue of spin qubits is their submicron footprint, enabling a single silicon chip to host the millions of qubits required to execute useful quantum algorithms with error correction. With each physical qubit needing multiple control lines however, a fundamental barrier to scale is the extreme density of connections that bridge quantum devices to their external control and readout hardware. A promising solution is to colocate the control system proximal to the qubit platform at millikelvin temperatures, connected via miniaturized interconnects. Even so, heat and crosstalk from closely integrated control has potential to degrade qubit performance, particularly for twoqubit entangling gates based on exchange coupling that are sensitive to electrical noise. Here, we benchmark silicon MOSstyle electron spin qubits controlled via heterogeneouslyintegrated cryoCMOS circuits with a power envelope sufficiently low to enable scaleup. Demonstrating that cryoCMOS can efficiently enable universal logic operations for spin qubits, we go on to show that millkelvin control has little impact on the performance of single and twoqubit gates. Given the complexity of our milli kelvin CMOS platform, with some 100thousand transistors, these results open the prospect of scalable control based on the tight packaging of semiconductor qubits with a ‘chiplet style' control architecture.






Fast readout of planar SiMOS quantum devices using gatebased sensing
Frederic Schlattner^{1}, Giovanni Oakes^{2}, David Ibberson^{2}, John Morton^{1}, Ross Leon^{2}, M. Fernando GonzalezZalba^{2}
^{1}UCL/Quantum Motion, ^{2}Quantum Motion
Abstract: Spins in planar metaloxidesemiconductor (MOS) quantum dots are a promising platform to scale semiconductor quantum computing architectures. Particularly, they offer compatibility with semiconductor manufacturing lines and extensibility in two dimensions. As the technology scales up, techniques to ameliorate the impact of readout sensors on qubit connectivity need to be brought into place. Gatebased readout offers the advantage that the very same gates that define the qubit array can be used for sensing, resulting on a negligeable impact on qubit connectivity. However, gatebased readout has been hindered in planar MOS devices due to a low lever arm, lower quality factors and detrimental accumulation in the fanout regions.
In this work, we develop a methodology for gatebased readout of planar MOS quantum dots reaching a stateoftheart minimum integration time of 300 ns for a SNR=1, corresponding to an electrical fidelity of >99.9% in just 2 µs and on a par with some of the best charge sensing demonstrations. We then show how the methodology can be used to read multiple quantum dots with just one gate/resonator enabling efficient and highly scalable readout.






Frequencymultiplexed readout of quantum dots with integrated cryoCMOS current amplifiers
Baptiste Jadot^{1}, Quentin Schmidt^{1}, Brian Martinez^{1}, Thomas Houriez^{1}, JeanBaptiste Casanova^{2}, Adrien Morel^{3}, Tristan Meunier^{4}, Gaël Pillonnet^{1}, Gérard Billiot^{1}, Aloysius Jansen^{5}, Xavier Jehl^{5}, Yvain Thonnart^{2}, Franck Badets^{1}
^{1}Univ. Grenoble Alpes, CEA, Leti, F38000 Grenoble, France, ^{2}Univ. Grenoble Alpes, CEA, List, F38000 Grenoble, France, ^{3}Univ. Savoie Mont Blanc, SYMME, F74000 Annecy, France, ^{4}Univ. Grenoble Alpes, CNRS, Institut Néel and Quobly, F38000 Grenoble, France, ^{5}Univ. Grenoble Alpes, CEA, IRIG, Pheliqs, F38000 Grenoble, France
Abstract: As spinqubit based quantum cores increase in size, signal multiplexing techniques appear as a musthave to maintain a scalable approach. In particular, spin qubit readout techniques usually require the use of large inductors, circulators and amplifiers. Another approach is the use of current amplifiers with a high enough bandwidth (Fig. a) to probe the state of several frequencymultiplexed qubits. In this talk, we present a cryogenic capacitive transimpedance amplifier (CTIA) operating at 4K, able to measure up to 70 different channels with a readout fidelity of 99.99% in 8.5μs. Varying the frequency separation between channels, we study the threshold above which channel overlap is suppressed (Fig. b) and compare our approach to state of the art reflectometry techniques.
In a second part of the talk, we exploit this cryogenic amplifier to demonstrate the charge readout of two multiplexed singleelectron transistors (SETs). These SETs are made from dualgates FDSOI 28nm transistors with adapted dimensions. Under a strong backbiasing voltage, a quantum dot is formed between the two gates and Coulomb diamonds are observed. Using this smallscale demonstrator, we verify the metrics obtained by the circuit alone and probe the electrostatic environment of each SET independently at 4K.





05.09.2024, Thursday, 9:0010:40




Optimal pulse control for Pauli spin blockade initialization and readout
Christian Ventura Meinersen, Stefano Bosco, Maximilian RimbachRuss
QuTech, TU Delft
Abstract: Semiconductor spin qubits in electrostatically defined quantum dots have recently raised attention by demonstrating highfidelity operations and proofofprinciple error correction algorithms [1, 2]. An essential component of quantum error correction is a reliable, yet fast, initialization and readout process. Fast readout is typically performed using Pauli spin blockade (PSB), where the spin information is converted into a charge degree of freedom using conditional hopping of an electron or hole inside a double quantum dot. The charge information can then be detected, for example, by a charge sensor. For initialization, the process is reversed.
A bottleneck for fast and highfidelity PSB readout is a fast yet adiabatic population transfer of the spin states to the desired charge states by pulsing the electrostatic potential of the double quantum dot. On the one hand, a fully adiabatic transfer is slow and can further lead to significant errors through state leakage induced by noise, e.g. 1/f charge noise at the anticrossing [3]. On the other hand, diabatic protocols are fast and potentially less prone to such errors but induce undesired transitions.
We theoretically study multiple shortcutstoadiabaticity strategies for population transfer and discuss transfer fidelities, speed, and feasibility. We particularly focus on a multilevel fastquasiadiabatic pulse [4], which provides an optimized pulse for the detuning of the double quantum dot potential. Furthermore, we analyze the impact of charge noise by combining the Lindblad and filter function formalisms and compute the corresponding transfer fidelity. Our results show a drastic improvement in fidelity compared to the commonly implemented linear ramp (See Figure).
[1] van Riggelen et al., npj Quantum Inf 8, 1–7 (2022).
[2] Takeda et al., Nature 608, 682–686 (2022).
[3] Krzywda et al., Phys. Rev. B 104, 075439 (2021).
[4] Fehse et al., Phys. Rev. B 107, 245303 (2023).






Tomography of silicon spin qubits dressed in a global field
Kevin Guo^{1}, Ensar Vahapoglu^{2}, MengKe Feng^{2}, Amanda Seedhouse^{2}, Wee Han Lim^{2}, Fay Hudson^{2}, James SlackSmith^{1}, Nicola Meggiato^{3}, Andre Saraiva^{2}, Chih Yang^{2}, Arne Laucht^{2}, Andrew Dzurak^{2}, Jarryd Pla^{1}
^{1}UNSW, ^{2}UNSW, Diraq, ^{3}ETH Zürich
Abstract: A spinbased quantum computer capable of running practical quantum algorithms will likely require an architecture which can be scaled to millions of qubits. Individually addressing and controlling many qubits presents a challenge in scalability due to qubit crosstalk. One possible solution utilizes a dielectric resonator to provide a uniform driving field used for global qubit control. By dressing the qubits in an onresonance global field, all qubits are driven simultaneously with a single microwave source.
In this study, we demonstrate control of electron spin qubits in silicon driven by a dielectric resonator. We use gate set tomography (GST) to benchmark bare qubits and qubits dressed in both continuouswave and sinusoidally modulating global fields (Fig. 1a), obtaining dressed single qubit gate fidelities exceeding 99% (Fig.1b). Crucially, the dressed identity gate error is four times smaller than the bare error, which is particularly relevant for common quantum algorithms where qubit idle time is high.
Finally, GST experiments are performed with introduced offsets in the Rabi and Larmor frequencies, finding that the dressed gates are more robust to variability in the Larmor frequency. Using a combination of GST and noise spectroscopy we posit potential sources of noise in the experimental setup and suggest improvements to further increase qubit fidelities. These results demonstrate the viability of dielectric resonators for global control in scalable silicon quantum computing architectures.






Dressed singlettriplet qubit in germanium driven via resonant exchange interaction
Kostas Tsoukalas, Alexei Orekhov, Uwe Lüpke, Felix Schupp, Matthias Mergenthaler, Gian Salis, Patrick HarveyCollard, Andreas Fuhrer
IBM Research
Abstract: In a typical spin qubit device, the exchange interaction of two spins is tuned via a barrier gate that controls the wavefunction overlap between neighboring charges. This interaction can be changed with a baseband pulse on the barrier gate voltage to perform a SWAPlike gate operation. However, in most cases the splitting of the two antiparallel spins competes with the exchange and results in a rotation around a tilted axis. Driving the exchange at the frequency of the antiparallel spin splitting can recover a clean, orthogonal SWAPlike gate. In this work, we study this resonant exchange interaction of two neighboring hole spins in a germanium quantum dot array (Figure, left inset).
We first demonstrate coherent Rabi and full Bloch sphere control in the subspace spanned by the two antiparallel spin states (Figure, right inset). Through Clifford randomized benchmarking, we estimate an average gate fidelity above 99%. We then use the resonant interaction to create a continuously driven qubit in this antiparallel spin subspace, realizing a dressed singlettriplet (ST0) qubit. On this dressed qubit we demonstrate the ability to initialize, perform single qubit operations and readout while achieving substantially increased coherence times compared with the bare qubit.






Geometry of the dephasing sweet spots of spinorbit qubits
Esteban RodriguezMena^{1}, Marion Bassi^{1}, Boris BrunBarriere^{1}, Simon Zilhmann^{1}, Lorenzo Mauro^{1}, Jose Carlos AbadilloUriel^{2}, Benoit Bertrand^{1}, Heimanu Niebojewski^{1}, Romain Maurand^{1}, YannMichel Niquet^{1}, Xavier Xehl^{1}, Vivien Schmitt^{1}, Silvano de Franceschi^{1}
^{1}CEA Grenoble, ^{2}Instituto de Ciencia de Materiales de Madrid
Abstract: Hole spin qubits in semiconductor quantum dots have attracted much attention as possible building blocks for quantum computers and simulators. They can be manipulated electrically thanks to strong intrinsic spinorbit coupling without the need for extrinsic elements such as micromagnets. This electrical addressability comes at the cost of a higher sensitivity to electrical and charge noise, limiting the dephasing time T^{*}_{2}. However, there may exist "dephasing sweet spots" where the qubit decouples (to first order) from the noise so that T^{*}_{2} reaches a maximum.
Here we discuss the geometrical nature of the dephasing sweet spots of a spinorbit qubit electrically coupled to one or more fluctuator(s). We show theoretically that the sweet spots usually draw lines (rather than points) on the unit sphere describing magnetic field orientation, providing more opportunities for optimal operation (see Fig a.). For that purpose, we characterize the qubit and its response to the fluctuator(s) by a Zeeman tensor G and its derivative(s) G' with respect to the fluctuating field(s). We discuss Si & Ge hole spin qubits as an illustration. Moreover, we experimentally probe the sweet lines in SiMOS devices (see Fig b.), and demonstrate that their position can be tuned by gate voltages. We achieve efficient electricdipole resonance on the sweet lines with a quality factor as high as 690, side by side with the values reported for electrons in silicon.
Our results provide guidelines for highly coherent operation of hole spin qubits regardless of the architecture, and insights for the design of devices more resilient to electrical noise.






Classification and magic magnetic field directions for spinorbitcoupled double quantum dots
Aritra Sen^{1}, Gyorgy Frank^{2}, Baksa Kolok^{2}, Jeroen Danon^{3}, András Pályi^{2}
^{1}Budapest University of Technology and Economics (BME), ^{2}Budapest University of Technology and Economics, ^{3}Norwegian University of Science and Technology
Abstract: The spin of a single electron confined in a semiconductor quantum dot is a natural qubit candidate. Fundamental building blocks of spinbased quantum computing have been demonstrated in double quantum dots with significant spinorbit coupling (SOC), for example with holes in Silicon and Germanium. Here we show that spinorbitcoupled double quantum dots can be categorised in six classes (A to F in the figure), according to a partitioning of the multidimensional space of their g tensors (g_{L}, g_{R}) which are in general real and nonsymmetric. In particular, for charge carriers in the valence bands of Silicon and Germanium subject to significant SOC, the g tensors are highly anisotropic. We predict that the spin physics is highly simplified due to pseudospin conservation, whenever the external magnetic field is pointing to special directions ("magic directions"), where the number of special directions is determined by the class. The magic directions yield spinrelaxation, electronshuttling sweetspots and simpler single spin readout using Pauli spin blockade. We also analyze the existence and relevance of "magic loops" in the space of magneticfield directions, corresponding to equal local Zeeman splittings. Magic loops provide dephasing sweet spots and stopping points in Pauli spin blockade readout. All together each class determines physical characteristics of the double dot, i.e., features in transport, spectroscopy, and coherence measurements, as well as qubit control, shuttling, and readout experiments. These results present an important step toward precise interpretation and efficient design of spinbased quantum computing experiments in materials with strong spinorbit coupling.





05.09.2024, Thursday, 11:1012:30




Large Scale Characterization of Qarpet at mK temperatures
Asser Elsayed^{1}, Federico Poggiali^{1}, Alberto Tosato^{1}, Davide Degli Esposti^{2}, Lucas Stehouwer^{1}, Menno Veldhorst^{1}, Giordano Scappucci^{3}
^{1}QuTech / TU Delft, ^{2}QuTech /TU Delft, ^{3}TU Delft, QuTech
Abstract: Germanium spin qubits are emerging as a promising technology for spinbased quantum processors based on electrostatically defined quantum dots. Advancements in optimizing Gebased materials and a deeper understanding of their fundamental physics are essential to drive progress. Traditional hero devicestyle measurements are no longer sufficient to drive the required progress. To meet these demands, largescale cryogenic characterization and statistical analysis, integrated into the design cycle of Gebased quantum devices, are imperative.
At the SiQEW 2023 we introduced Qarpet (in name of the tightly knit fabric of electrostatic gates defining a highly dense array of qubits). The Qarpet architecture is tailored for largescale characterization of quantum dot qubits and features a crossbar array of repeating unit cells, each comprising a sensing dot and qubit dots. The device preparation and bonding is engineered for individual control yet maximizing shared lines to minimize room temperature resource requirements. Following the proof of principle qubit demonstration, our efforts are now focusing on the comprehensive characterization of quantum dots. We showcase the operation of singlehole transistors for sensing dots and lasthole occupation for qubit dots, allowing to address uniformity over extensive length scales. We then provide a detailed statistical investigation of charge noise, further discuss the potential for highvolume qubit measurements, and present proofofconcept results paving the way for future characterization endeavors.






An Industrial Triple Metal Gate Process for a 2D Shuttling Architecture
Wolfram Langheinrich^{1}, Pascal Muster^{1}, Sebastian Pregl^{1}, Felix Reichmann^{2}, Yuji Yamamoto^{2}, Varvara Brackmann^{3}, Michael Friedrich^{3}, Nikola Komerički^{4}, Laura Diebel^{5}, Dominique Bougeard^{5}, Till Huckemann^{6}, Lars Schreiber^{6}, Hendrik Bluhm^{6}
^{1}Infineon Technologies Dresden GmbH & Co. KG, ^{2}IHP LeibnizInstitut für innovative Mikroelektronik, Frankfurt/Oder, Germany, ^{3}Fraunhofer IPMS, Dresden, Germany, ^{4}Fraunhofer IAF, Freiburg, Germany, ^{5}Institut für Experimentelle und Angewandte Physik, Fakultät für Physik, Universität Regensburg, ^{6}JARA Institute for Quantum Information
Abstract: Scaling qubit numbers while improving gate fidelities is a general challenge for any quantum computing platform. In case of siliconbased spin qubits an obvious approach are crossbar arrays, but they additionally suffer issues like wiring fanout and crosstalk. A shuttlingbased 2Darchitecture is a promising alternative at the cost of increased fabrication complexity [1]. Therefore, a triple metal gate process on Si/SiGe heterostructures was developed within an industrial production line, enabling high yield and reliability for conveyormode shuttling devices with micrometer length. In order to achieve low defect density, pattern transfer is done using reactive ion etching, compared to liftoff techniques, typically used in research lab processes. Only one gate layer (Gate1) requires electron beam lithography, whereas the screening gate (Gate0) for the shuttling channel and the top gate (Gate2) use optical lithography. Furthermore, optimisation of the gate oxides is crucial, since charge noise and disorder will affect the shuttling fidelity. Room temperature characterisation of our CVD SiO_{2} at MOS devices shows low interface trap densities as low as 10^{10} cm^{2}. Lowfrequency charge noise was measured at double quantum dots in the 50mK region, where one dot is operated as a SET. Values below 1 µeV/√Hz were achieved. The valley splitting energy was obtained via magnetospectroscopy at the same device and values around 50 µeV were found. First results also show that singleelectron shuttling is possible in devices fabricated with this process.
Figure: Short conveyormode shuttling device with two SETs left and right. The darkfield TEM shows four claviature gates for adiabatically moving a single electron.






A 300mm SiliconBased Platform for Quantum Computing Device Technologies
Bart Raes, Clement Godfrin, Ruoyu Li, George Simion, Stefan Kubicek, Sofie Beyne, Shana Massar, Yann Canvel, Julien Jussot, Roger Loo, Yosuke Shimura, Massimo Mongillo, Danny Wan, Kristiaan De Greve
Imec
Abstract: Silicon spin qubits have emerged as leading contenders for largescale quantum computing, owing to their extended coherence times and compatibility with CMOS technology. At IMEC, we have devised a comprehensive strategy to systematically optimize all process parameters, leveraging industrial 300mm fabrication processes, to enhance qubit performance. The scalability and high reproducibility inherent in 300mm processes enable a deterministic exploration of qubit metrics and their sensitivity to process parameters, crucial for advancing qubit quality.
We showcase the efficacy of our approach by presenting the latest findings, focusing on fully integrated qubit structures, on our electron SiMOS, hole SiMOS and Si/SiGe platforms. We will discuss, among others, high fidelity (9999.9%), high speed operation and readout, and the overall approach into scaling up our qubit platforms.






Fidelity of Si/SiGe spin qubits fabricated by an industrial 300 mm process
Viktor Adam^{1}, Thomas Koch^{1}, Daniel Schroller^{1}, Bart Raes^{2}, Julian Ferrero^{1}, Stefan Kubicek^{2}, Shana Massar^{2}, Ruoyu Li^{2}, Clement Godfrin^{2}, Danny Wan^{2}, Kristiaan De Greve^{2}, Wolfgang Wernsdorfer^{1}
^{1}Karlsruhe Institute of Technology (KIT), Germany, ^{2}Interuniversity Microelectronics Centre (imec), Belgium
Abstract: Silicon spin qubits are one of the most promising candidates for largescale quantum computing due to their long coherence times and compatibility with existing industrial complementary metaloxidesemiconductor (CMOS) fabrication processes. The Si/SiGe heterostructure aims to decouple the qubits from the semiconductoroxide interface, known as the main source of charge noise, but often features low valleysplitting values and is more challenging in terms of heterostructure growth.
In this work, we demonstrate the operation of natural Si/SiGe spin qubit devices, fabricated in a 300 mm semiconductor manufacturing facility using a combination of optical and ebeam lithography with an industry compatible process. A Coµ magnet generates a magnetic field gradient that enables us to drive the qubits at MHz rate via the electric dipole spin resonance (EDSR) drive line and to address both qubits individually via their respective resonance frequency. We measure valleysplittings of 85 µeV and charge noise values of 1 µeV/√Hz. Furthermore, we observe spin relaxation times above 1 s and qubit coherence times around 1 µs, which are common for natural silicon quantum wells. Finally, we present qubit drives of several MHz and a single gate fidelity above 99%.





06.09.2024, Friday, 9:0010:40




The Qube: a lattice of vertically and laterally coupled quantum dots
Hanifa Tidjani^{1}, Michael Chan^{2}, Dario Denora^{2}, Jann HinnerkUngerer^{3}, Alexander Ivlev^{2}, Alberto Tosato^{2}, Corentin Déprez^{2}, Lucas Stehouwer^{2}, Amir Sammack^{4}, Stefan Oosterhout^{4}, Giordano Scappucci^{2}, Menno Veldhorst^{2}
^{1}QuTech, TU Delft, ^{2}TU Delft, ^{3}Harvard University, ^{4}TNO
Abstract: Quantum dot based spin qubits have made rapid developments in the complexity of devices, with scaling of qubits being a priority. Approaches towards the scale up of quantum dot qubits are primarily based on one or twodimensional planar arrays. Here, introduce the third spatial dimension by increasing the number of quantum wells, forming a threedimensional quantum dot lattice in a Ge/SiGe heterostructure. To achieve this, we show that on a bilayer heterostructure a single top plunger gate can be used to form a verticallycoupled double quantum dot, both in transport and in charge sensing[1]. We demonstrate independent control of the occupation of the vertically aligned quantum dots by virtualizing to the surrounding gates. This virtualization is facilitated by differing gatetodot capacitances due to the separation of the quantum wells in the vertical zdirection, resulting in differences in electrostatic confinement. By accumulating under two plunger gates, we form a 2x2 quantum dot array, aligned on the xz plane[2]. Despite a small interlayer separation of 4nm, we control the occupation and tune to the (1,1,1,1) regime. Expanding this to a device with four plunger gates (a) and a larger interlayer separation of 10nm (b), we form a 3D lattice of quantum dots (c). Confinement plays an important role in operating the lower layer. When strongly confined, only the upper quantum well is occupied, and in this regime we form lateral singlettriplet qubits. The expansion of the quantum dot devices to the third spatial dimension presents an exciting opportunity to extend the framework of gatedefined semiconductor quantum dots beyond planar implementations, for the development of scalable quantum computation and simulation.






A hole spin qubit on a planar silicon 300 mm CMOS platform
Isaac Vorreiter^{1}, Joseph Hillier^{2}, Scott Liles^{3}, Jonathan Huang^{1}, Aaquib Shamim^{3}, Stefan Kubicek^{4}, Clement Godfrin^{4}, Danny Wan^{4}, Ruoyu Li^{4}, Bart Raes^{4}, Kristiaan De Greve^{4}, Alex Hamilton^{1}
^{1}UNSW, ^{2}QED group, School of Physics, University of New South Wales, Sydney NSW, 2052, Australia, ^{3}University of New South Wales, ^{4}imec, KU Leuven
Abstract: Spin qubits implemented within GroupIV semiconductor hole quantum dots are demonstrating favourable properties as building blocks for scalable quantum processors [15]. Holes possess a strong intrinsic spinorbit interaction which enables fast allelectrical spin control via electricdipole spin resonance (EDSR), allowing for the implementation of ‘spinorbit' qubits. Implementing spinorbit qubits within 2D planar MOS structures provides the benefits of flexible integration, enabling the creation of densely packed 2D tunnelcoupled arrays or sparse arrays [6]. However, spinorbit qubits realised in 2Dlike planar MOS silicon have not been extensively studied, and as such their single and two qubit properties are not wellcharacterised in these systems.
Here we demonstrate a holespin spinorbit qubit in a planar silicon double quantum dot. Furthermore, we demonstrate Rabi frequencies reaching 15 MHz and controllable twoqubit exchange of order 40 MHz. We investigate the qubit control and fidelity as a function of magnetic field, demonstrating strategies for operating these qubits in the presence of strong spinorbit interaction. Additionally, the device was fabricated using a 300mm integration flow [7] compatible with foundrybased fabrication processes. Our results affirm industrially fabricated 2D MOS silicon quantum dots as feasible platforms for implementing spinorbit qubits.
Figure: a) Falsecolour SEM image of the device, where the quantum dots are formed under the plunger gates P1 and P2, with confinement provided by the barrier gates B1, B2 and the Cgate. An adjacent singlehole transistor is use for charge sensing and readout. b) Charge stability diagram in the fewhole, weaklycoupled regime where spin readout is performed. c) Rabi oscillations as a function of frequency detuning showing a typical chevron pattern.
[1] Nat Commun 7,13575,(2016).
[2] 17,1072–1077,(2022).
[3] Nat. Nanotechnol. 16,308–312,(2021).
[4] Nat Electron 5,178–183,(2022).
[5] Nature 577,487–491,(2020).
[6] npj Quantum Inf 3,34,(2017).
[7] 2021 Symposium on VLSI Circuits, Kyoto, Japan, 2021,pp.12






Fully autonomous tuning of a spin qubit
Jonas Schuff^{1}, Miguel Carballido^{2}, Madeleine Kotzagiannidis^{3}, Juan Carlos Calvo^{3}, Marco Caselli^{3}, Jacob Rawling^{3}, David Craig^{1}, Barnaby van Straaten^{1}, Brandon Severin^{1}, Federico Fedele^{1}, Simon Svab^{2}, Pierre Chevalier Kwon^{2}, Rafael Eggli^{2}, Taras Patlatiuk^{2}, Nathan Korda^{3}, Dominik Zumbühl^{2}, Natalia Ares^{1}
^{1}University of Oxford, ^{2}University of Basel, ^{3}Mind Foundry
Abstract: Spanning over two decades, the study of qubits in semiconductors for quantum computing has yielded significant breakthroughs. However, the development of largescale semiconductor quantum circuits is still limited by challenges in efficiently tuning and operating these circuits. Identifying optimal operating conditions for these qubits is complex, involving the exploration of vast parameter spaces. This presents a real 'needle in the haystack' problem, which, until now, has resisted complete automation due to device variability and fabrication imperfections. In this study, we present the first fully autonomous tuning of a semiconductor qubit, from a grounded device to Rabi oscillations, a clear indication of successful qubit operation. We demonstrate this automation, achieved without human intervention, in a Ge/Si core/shell nanowire device. Our approach integrates deep learning, Bayesian optimization, and computer vision techniques. We expect this automation algorithm to apply to a wide range of semiconductor qubit devices, allowing for statistical studies of qubit quality metrics. As a demonstration of the potential of full automation, we characterise how the Rabi frequency and gfactor depend on barrier gate voltages for one of the qubits found by the algorithm. Twenty years after the initial demonstrations of spin qubit operation, this significant advancement is poised to finally catalyze the operation of large, previously unexplored quantum circuits.






Automated realtime gate virtualization of a 10 quantum dot array
Anantha Rao^{1}, Donovan Buterakos^{2}, Valentin John^{3}, Cécile Yu^{3}, Stefan Oosterhout^{4}, Lucas Stehouwer^{3}, Giordano Scappucci^{3}, Menno Veldhorst^{3}, Francesco Borsoi^{3}, Justyna Zwolak^{2}
^{1}University of Maryland, ^{2}NIST, ^{3}QuTech and Kavli Institute of Nanoscience, Delft University of Technology, Delft, The Netherlands, ^{4}Netherlands Organisation for Applied Scientific Research (TNO), Delft, The Netherlands
Abstract: Arrays of gatedefined quantum dots are potential candidates for realizing scalable multiqubit devices and efficiently performing quantum computation. As the capacitive coupling between different voltage gates leads to crosstalk, implementing virtual gates is crucial for achieving orthogonal control of all the qubit's parameters, such as energies and couplings. However, determining efficiently and accurately the compensations on each layer of virtual gates frameworks is challenging and requires iterative procedures. In our work, we propose and test in realtime an automated method for defining multiple layers of virtual gates using machine learning techniques. Earlier methods relied on automatic detection of the slope of transition lines from individual twodimensional charge stability diagrams (2DCSD). However, we show that such strategies are prone to errors, particularly due to effects such as slow loading via decoupled reservoirs that lead to "latchy" transitions. In our simple and robust approach, we combine machine learning and traditional curvefitting to define virtual barrier gates. Our algorithm identifies the locations and tracks the trajectories of charge interdots in 2DCSDs while stepping a barrier gate voltage [Fig. 1(B), see Fig. 1(D) for trajectory of charge interdot 6 shown in the first plot in Fig. 1(B)]. The slope of the linear fit in Fig.1(D) is used to extract the required compensation for the virtual plunger gates preventing the movement of the interdot as barrier voltage is varied [Fig. 1(C)]. With our approach, we autonomously virtualize a stateoftheart holebased 10 quantum dot system [Fig. 1(A)], with 12 barrier gates and 10 plunger gates defined on planar germanium, producing a 10x12 virtual matrix in under an hour, leading to a factor of 10x to 1000x reduction in time with respect to manual operations.






Investigating phononinduced frequency shifts in semiconductor spin qubits
Irina Heinz^{1}, Jeroen Danon^{2}, Guido Burkard^{1}
^{1}University of Konstanz, ^{2}Norwegian University of Science and Technology
Abstract: Spin qubits have proven to be a feasible candidate for quantum computation and benefit from the advanced device manufacturing in semiconductor industry making them a good candidate for scalable quantum computation [1]. Compared to superconducting platforms spin qubits can operate at higher temperatures up to hundreds of mK. However, recent experiments [2,3] show a nontrivial dependence of the spin qubit frequency on the temperature. In this work we aim to gain insight into the underlying physics causing frequency shifts in the lowtemperature limit and to understand the interaction between qubit and phonons.
[1] G. Burkard, T. D. Ladd, A. Pan, J. M. Nichol, and J. R. Petta, Semiconductor spin qubits, Rev. Mod. Phys. 95, 025003 (2023)
[2] B. Undseth, X. Xue, M. Mehmandoost, M. RimbachRuss, P. T. Eendebak, N. Samkharadze, A. Sammak, V. V. Dobrovitski, G. Scappucci, and L. M. K. Vandersypen, Nonlinear Response and Crosstalk of Electrically Driven Silicon Spin Qubits, Phys. Rev. Appl. 19, 044078 (2023)
[3] B. Undseth, O. PietxCasas, E. Raymenants, M. Mehmandoost, M. T. Mądzik, S. G. J. Philips, S. L. de Snoo, D. J. Michalak, S. V. Amitonov, L. Tryputen, B. P. Wuetz, V. Fezzi, D. D. Esposti, A. Sammak, G. Scappucci, and L. M. K. Vandersypen, Hotter is Easier: Unexpected Temperature Dependence of Spin Qubit Frequencies, Phys. Rev. X 13, 041015 (2023)





06.09.2024, Friday, 11:1012:30




Remote quantum state transfer using spinphoton vacuum Rabi oscillations
Xiao Xue^{1}, Jurgen Dijkema^{1}, Patrick HarveyCollard^{2}, Maximilian RimbachRuss^{1}, Tobias Bonsen^{1}, Sander de Snoo^{1}, Guoji Zheng^{1}, Amir Sammak^{1}, Giordano Scappucci^{1}, Lieven Vandersypen^{1}
^{1}QuTech, TU Delft, ^{2}IBM Research Zurich
Abstract: With the demonstrations of strong spinphoton couplings in semiconductor and distant twoqubit logic mediated by virtual microwave photons, circuit quantum electrodynamics becomes a promising approach to link spin qubits in spatially separated quantum modules. Here, we use a real microwave photon which is on resonance with two spin qubits that are 250 µm apart to coherently transfer a quantum state between them.
First, we report the first observation of vacuum Rabi oscillations between spin qubits and microwave photons. The oscillations can be rapidly switched on and off by a simple voltage pulse on the double quantum dot detuning. The oscillation frequencies match the observed avoided crossing of the energy levels in spectroscopy measurements.
Second, we utilize consecutive vacuum Rabi oscillations of the two spin qubits to transfer quantum states from one to the other. The quantum state is first transferred coherently from one qubit to the photonic state in the resonator with a halfperiod vacuum Rabi oscillation, and then similarly transferred to the second qubit.
These experiments provide new avenues for scaling quantum networks on a chip.






Strong ChargePhoton Coupling in Planar Ge with Granular Aluminium Superconducting Resonators
Marian Janik^{1}, Kevin Roux^{1}, Carla Borja Espinosa^{1}, Oliver Sagi^{1}, Abdulhamid Baghdadi^{1}, Thomas Adletzberger^{1}, Andrea Ballabio^{2}, Marc Botifoll^{3}, Alba Garzon Manjon^{3}, Jordi Arbiol^{4}, Daniel Chrastina^{2}, Giovanni Isella^{2}, Ioan Pop^{5}, Georgios Katsaros^{1}
^{1}ISTA, Institute of Science and Technology Austria, Am Campus 1, 3400 Klosterneuburg, Austria, ^{2}LNESS, Physics Department, Politecnico di Milano, via Anzani 42, 22100, Como, Italy, ^{3}ICN2, Catalan Institute of Nanoscience and Nanotechnology CSIC and BIST, Campus UAB, Bellaterra, 08193 Barcelona, Catalonia, Spain, ^{4}ICN2, Catalan Institute of Nanoscience and Nanotechnology CSIC and BIST, Campus UAB, Bellaterra, 08193 Barcelona, Catalonia, Spain; ICREA, Catalan Institution for Research and Advanced Studies, Passeig de Lluís Companys 23, 08010 Barcelona, Catalonia, Spain, ^{5}IQMT, Institute for Quantum Materials and Technology, Karlsruhe Institute of Technology, 76344 EggensteinLeopoldshafen, Germany; PHI, Physikalisches Institut, Karlsruhe Institute of Technology, 76131 Karlsruhe, Germany
Abstract: Charges and spins confined in quantum dots (QDs) coherently coupled to a microwave photon in a superconducting resonator are interesting for quantum computation, quantum optics, or analog quantum simulations. Reaching a sizeable coupling strength exceeding loss rates is intrinsically challenging in these systems due to the small dipole moment of charges confined in QDs. Since it can be compensated with highimpedance resonators, they have received notable attention in the past decade. Nevertheless, previous QD circuit quantum electrodynamics implementations have not exceeded the impedance of ∼ 3.8 kΩ, leaving opportunities for significant improvement. The large kinetic inductance of granular aluminium (grAl) could provide an orderofmagnitude enhancement.
Here, we report an in situ resistance control of grAl evaporation, which allows us to reproducibly fabricate grAl coplanar waveguide resonators with characteristic impedances reaching Z = 22.3 ± 0.3 kΩ due to the large sheet kinetic inductance up to L_{k} ≅ 3 nH/☐. Magnetic field resilience of B_{⊥}^{max} = 281 ± 1 mT and B_{∥}^{max} = 3.5 ± 0.05 T is achieved for 100 nmnarrow superinductors. We implement these resonators with QDs in planar germanium and reach the strong holephoton coupling regime with a rate of g_{c} = 566 ± 2 MHz and a cooperativity of C = 251 ± 8.
The demonstrated properties make grAl resonators suitable for boosting the spinphoton coupling strength, a crucial requirement for fast, highfidelity, longdistance twoqubit gates.






Lifetime and coherence of a cQED hole spin qubit
Leo Noirot^{1}, Simon Zihlmann^{2}, Cecile Yu^{3}, Etienne Dumur^{2}, Romain Maurand^{4}
^{1}CEA Grenoble Lateqs, ^{2}CEA Grenoble, ^{3}TU Delft, ^{4}CEA
Abstract: Spins in semiconductor quantum dots constitute a promising platform for scalable quantum information processing. Coupling them strongly to the photonic modes of superconducting microwave resonators would enable fast nondemolition readout and longrange, onchip connectivity, well beyond nearestneighbor quantum interactions. As the field of spin circuit quantum electrodynamic (cqed) is growing, new experiments showed spinphoton coupling rates as high as 330 MHz and a 2qubit gate mediated by a photonic interaction. However, up to now, all of the semiconductor spincqed devices have showed fast decoherence and relaxation, hindering the high fidelity control and readout usually achieved for spin qubits.
We present here an experimental study of a hole spin qubit embedded in a Si double quantum dot, strongly coupled to a microwave cavity thanks to the intrinsic spinorbit interaction (SOI) of holes in Si. We measure relaxation (Fig. a) and dephasing (Fig. b) as a function of magnetic field and gate voltage (not provided here), therefore controlling the spinphoton coupling as well as the spin's energy through the magneticfield dependence of the SOI and gatedependence of the electric dipole. We span its energy over a range of 10GHz, crossing several cavity modes and identify photon emission (multimode Purcell) as the limiting mechanism for lifetime. The dephasing shows signs of chargenoise induced dephasing as its magneticfield dependence follows the secondorder electrical susceptibility of the qubit at the charge sweetspot. Accordingly, the Ramsey dephasing time is maximal at the sweetspot while it degrades strongly away from it. Surprisingly however, the Echo dephasing time is minimal at the sweetspot and increases away from it. This trend is qualitatively in agreement with a dephasing induced by thermal photons or it could be the consequence of the interplay between first and second order coupling to charge noise. Further studies are needed to fully clarify the situation.






Microsecondlived quantum states in a carbonbased circuit in cQED
Benoit Neukelmance^{1}, Hue Benjamin^{2}, Quentin Schaeverbeke^{3}, Lucas Jarjat^{4}, Arnaud Théry^{4}, Jules Craquelin^{4}, William Legrand^{5}, Cubaynes Tino^{4}, Gulibusitan Abulizi^{3}, Jeanne Becdelievre^{3}, Mariah El Abassi^{3}, Joseph Sulpizio^{3}, Davide Stefani^{3}, Audrey Cottet^{6}, Matthieu Desjardins^{3}, Kontos Takis^{7}, Matthieu Delbecq^{8}
^{1}1) Laboratoire de Physique de l'École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, Paris, France; 2) C12 Quantum Electronics, Paris, France, ^{2}1) aboratoire de Physique de l'École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, Paris, France; 2) C12 Quantum Electronics, Paris, France, ^{3}C12 Quantum Electronics, Paris, France, ^{4}Laboratoire de Physique de l'École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, Paris, France, ^{5}1)Laboratoire de Physique de l'École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, Paris, France; 2) C12 Quantum Electronics, Paris, France, ^{6}1) Laboratoire de Physique de l'École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, Paris, France; 3) Laboratoire de Physique et d'Etude des Matériaux, ESPCI Paris, Université PSL, CNRS, Sorbonne Université, Paris, France, ^{7}1) aboratoire de Physique de l'École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, Paris, France; 3) Laboratoire de Physique et d'Etude des Matériaux, ESPCI Paris, Université PSL, CNRS, Sorbonne Université, Paris, France, ^{8}1) Laboratoire de Physique de l'École normale supérieure, ENS, Université PSL, CNRS, Sorbonne Université, Université Paris Cité, Paris, France; 3) Laboratoire de Physique et d'Etude des Matériaux, ESPCI Paris, Université PSL, CNRS, Sorbonne Université, Paris, France; 4) Institut universitaire de France (IUF)
Abstract: Electron spins in quantum dots represent an attractive path towards the realization of quantum processors due to the high spin resilience to environmental noise and the large electric dipole allowed by the DQD. While multiqubit gates are commonly mediated through nearestneighbor exchange interaction, achieving coherent longrange coupling between spins remains a major challenge for such architectures. Enabling spinphoton interaction is thus appealing.
Here, we manipulate the quantum state of an ultraclean suspended carbon nanotube double quantum dot with ferromagnetic contacts embedded in a microwave cavity. By performing quantum manipulations in the time domain via the cavity photons, we demonstrate coherence times T_{2}* of the order of 1 μs as shown by the Ramsey fringes experiment shown in panel 1. This is two orders of magnitude larger than what was ever measured in any carbon quantum circuit and one order of magnitude larger than silicon based quantum dots in comparable environment. Thanks to a peculiar Rabi chevrons pattern, we are able to infer precisely the spectrum of our double quantum dot. In particular it allows us to understand why our transition exhibits a very weak dispersion with detuning ε_{δ} as shown in panel 2 (dashed line is theory). This regime is favorable to drastically reduce charge noise dephasing. Combining the regime we achieved with a high impedance resonator to boost the bare electronphoton coupling should enable high fidelity twoqubit gates in future works. Overall this holds promise for carbon as a contender host material for spin qubits in circuit quantum electrodynamics.





06.09.2024, Friday, 14:0015:40




Fabrication and characterization of multirail Si/SiGe exchangeonly spin qubits in the SLEDGE architecture
Jacob Blumoff
HRL Laboratories LLC
Abstract: Si/SiGe exchangeonly spin qubits encoded in a decoherencefree subsystem (DFS) are a compelling platform for quantum computing because of their compatibility with advanced fabrication techniques and their exclusive use of baseband pulses for control. Using the SingleLayer EtchDefined GateElectrode (SLEDGE) architecture, which implements a CMOSlike separation between active frontend gates and electrical routing layers, we recently demonstrated highfidelity twoqubit DFSencoded gates in a singlerail device. Scaling to multirail geometries, where qubits are connected to more than two neighbors, is an essential step towards quantum fault tolerance because it improves robustness and connectivity. We report on the fabrication of a tworail, sixdot device with three distinct backend routing layers. We also discuss the electrostatic tuneup, the initial parametric characterization, and the singlequbit randomized benchmarking performance of this device.






Highfidelity coupling between nuclear spin registers via electron exchange
Junliang Wang Wang, Hermann Edlbauer, Ian Thorvaldson, Christian Moehle, Billy Pappas, A F M Saiful Haque Misha, Michael Jones, Fabian Pena, Yousun Chung, Joris G. Keizer, Ludwik Kranz, Michelle Y. Simmons
Silicon Quantum Computing
Abstract: Nuclear spin qubits in silicon exhibit extremely long coherence times exceeding seconds, making them an excellent candidate for quantum computation [1]. Scanning tunnelling microscopy (STM) enables the fabrication of atomically engineered nuclearspin registers with subnm precision, which are readily operable with single and twoqubit gate fidelity above the faulttolerant threshold [2]. A key challenge for scalingup is the ability to couple nuclear spin registers whilst maintaining high fidelity. Here, we demonstrate such a highly efficient quantum link using exchangecoupled electron spins that are located on neighbouring nuclear spin registers. Applying interleaved randomized benchmarking to these electron spins, we report a CNOT gate fidelity exceeding 99%. By controlling the nuclear spins in both registers [3], we engineer the energy difference between the electron spin qubits and explore its impact on the CNOT gate fidelity. Establishing a highly efficient link between nuclear spin registers, our results continue to pave the way towards a largescale silicon quantum processor.
References:
[1] J. T. Muhonen et al., Nature Nanotechnology 9 (2014)
[1] I. Thorvaldson et al., arXiv:2404.08741 (2024)
[2] J. Reiner et al., Nature Nanotechnology (2024)






Engineering the excited state spectra of Si/SiGe quantum dots.
Tom Watson, Daniel Keith, Sam Neyens, Otto Zietz, Andrew Wagner, Ekmel Ercan, Joelle Corrigan, John Rooney, Peter Bavdaz, Rambert Nahm, David Kohen, Nathan Bishop, Stephanie Bojarski, Jeanette Roberts, James Clarke
Intel Corporation
Abstract: A key challenge for building larger spinbased quantum computers is to maintain high uniformity across the qubit array. For Si/SiGe spin qubits, one of the main sources of nonuniformity are lowlying excited states in quantum dots that result in poor readout and initialization fidelity. In addition, these states can lead to issues with spin shuttling and exchangebased gate operations. For one electron in a quantum dot, the lowest excited state is due to the valley degree of freedom and is determined by the atomistic details of the Si quantum well. Adding a second electron results in singlet/triplet states where the splitting is determined by both the valley and orbital degree of freedom and how strongly they couple.
In this talk, we discuss the techniques we employ for the fast millikelvin characterization of quantum dot excited states to build statistically relevant data sets to compare different wafers. With these techniques, we show that we can increase the mean valley splitting of our quantum dots from Ev = 60ueV (σEv = 40ueV) to Ev = 230ueV ( σEv =130ueV) by introducing a small percentage of Ge (~2%) into the silicon quantum well. Furthermore, we show that despite high valley splittings, the singlet/triplet splitting can still be strongly suppressed due to confinement effects and valleyorbit coupling.






Efficient 3D TCAD simulations of twoqubit gates in semiconductor quantum dots
Pericles Philippopoulos^{1}, Mohammad Mostaan^{2}, Raphaël Prentki^{1}, Thomas Baker^{3}, Marek Korkusinski^{4}, Felix Beaudoin^{5}
^{1}Nanoacademic Technologies, ^{2}Nanoacademic Technologies, Simon Fraser University, ^{3}University of Victoria, ^{4}National Research Council of Canada, ^{5}Nanoacademic Technologies Inc.
Abstract: The design and engineering of classical semiconductor chips often relies on a mature set of computational tools. Among these tools are technology computeraided design (TCAD) software, used to predict device performance and trends before fabrication. As we move toward using semiconductors for quantum technologies like spin qubits, it seems plausible that we will need to adopt best practices from classical electronics. However, due to fundamental differences in operating principles between classical and quantum hardware, specialized quantum TCAD tools must be developed.
In recent years, spinqubit systems in quantumdot arrays have been scaled up to the 10particle regime. In this regime, using 3D TCAD simulations to accurately model quantum operations on each qubit (e.g., singlequbit gates and readout) and pairs of qubits (e.g., twoqubit gates) becomes imperative to mitigate the significant costs of experimental rounds of design, fabrication, and characterization. Additionally, realistic 3D finiteelement simulations of these systems are challenging due to mesh complexity and poor scaling of accurate manybody solvers with respect to particle number and size of the singleparticle basis.
In this presentation, we will present computationally efficient approaches to simulate twoqubit gates in electron or hole quantumdot systems with realistic 3D geometries. As illustrated in the figure, our approach encompasses the following key simulation steps: arbitrary 3D geometry definition (a), simulation of device electrostatics, computation of tunnel splittings (b) and envelope functions [(c) and (d)] from a singleparticle effective Schrödinger equation, manybody simulations yielding charge stability diagrams and the exchange interaction strength (e), and timedependent simulations of twoqubit gates under realistic noise sources (f). These results will be presented for practically relevant device geometries such as Fully Depleted Silicon On Insulator (FDSOI) transistors. Leveraging recent advances in Coulomb interaction calculations and manybody physics, this work paves the way for highthroughput twoqubit gate simulation workflows for quantumdot arrays.






Baseband control of singleelectron silicon spin qubits in twodimensions
Brennan Undseth^{1}, Florian Unseld^{1}, Yuta Matsumoto^{1}, Eline Raymenants^{1}, Oriol PietxCasas^{1}, Saurabh Karwal^{2}, Sergey Amitonov^{2}, Amir Sammak^{2}, Giordano Scappucci^{1}, Lieven Vandersypen^{1}
^{1}Delft University of Technology, ^{2}Netherlands Organization for Applied Scientific Research (TNO)
Abstract: Micromagnetenabled electricdipole spin resonance (EDSR) is an established means of highfidelity singlespin control in silicon. However, the resulting architectural limitations have restrained stateoftheart quantum processors to onedimensional arrays, and heating effects from the associated microwave dissipation exacerbates crosstalk for multiqubit operations. In contrast, spin control based on hopping spins has recently emerged as a compelling primitive for highfidelity baseband control in sparse hole arrays in germanium [1]. In this work, we commission a ^{28}Si/SiGe 2x2 quantum dot array both as a 4qubit quantum processor using established EDSR techniques and as a 2qubit device using hopping spins in a low magnetic field regime. This control method is previously unexplored in the silicon platform but benefits from engineerable micromagnetdominated stray fields that induce a measurable tip in quantization axis between adjacent quantum dots. The figure illustrates how the measured spin fraction after a particular shuttling sequence can be used to infer the tip when fit to the expected unitary evolution. We can directly compare the two modes of operation in terms of fidelity, coherence, and crosstalk. We find that the shuttling gate fidelity of 99.7% is on par with the benchmarked resonant gate while offering a shorter gate time. Lowering the external field to the shuttling regime nearly doubles the measured T_{2}^{Hahn} suggesting a reduced coupling to charge noise. Finally, the shuttling gate circumvents the transient pulseinduced resonance shift. These results establish new opportunities for engineering spin qubit arrays in silicon.
[1] Wang, C., John, V., Tidjani, H., et al. Operating semiconductor quantum processors with hopping spins. arXiv:2402.18382





06.09.2024, Friday, 16:1017:30




Er sites in Si for quantum information processing
Alexey Lyasota^{1}, Ian Berkman^{2}, Gabriele de Boo^{2}, John Bartholomew^{3}, Shao Lim^{4}, Brett Johnson^{5}, Jeffrey McCallum^{4}, BinBin Xu^{2}, Shouyi Xie^{2}, Rose Ahlefeldt^{6}, Matthew Sellars^{6}, Chunming Yin^{7}, Sven Rogge^{2}
^{1}Centre of Excellence for Quantum Computation and Communication Technology, School of Physics, University of New South Wales, ^{2}Centre of Excellence for Quantum Computation and Communication Technology, School of Physics, University of New South Wales, Sydney, New South Wales 2052, Australia, ^{3}Centre for Engineered Quantum Systems, School of Physics, The University of Sydney, Sydney, New South Wales 2006, Australia; The University of Sydney Nano Institute, The University of Sydney, Sydney, New South Wales 2006, Australia, ^{4}Centre of Excellence for Quantum Computation and Communication Technology, School of Physics, University of Melbourne, Victoria 3010, Australia, ^{5}School of Science, RMIT University, Victoria 3001, Australia, ^{6}Centre of Excellence for Quantum Computation and Communication Technology, Research School of Physics, Australian National University, Canberra, Australian Capital Territory 0200, Australia, ^{7}Centre of Excellence for Quantum Computation and Communication Technology, School of Physics, University of New South Wales, Sydney, New South Wales 2052, Australia; Hefei National Laboratory for Physical Sciences at the Microscale, CAS Key Laboratory of Microscale Magnetic Resonance and School of Physical Sciences, University of Science and Technology of China, Hefei 230026, China
Abstract: Rareearth ions in a solidstate host exhibit low homogeneous broadening and long spin coherence at cryogenic temperatures, making them promising for a range of quantum applications, such as optical quantum memories and opticalmicrowave transductions. Emitters with long electron spin and optical coherence in Si, a leading material platform for electronic and photonic technologies, are especially attractive for quantum applications.
Here, we report on the observation of eight Er sites in Si that have both long optical coherence and electron spin lifetime. We measured 1 ms spin coherence for two sites in a nuclear spinfree silicon crystal (<0.01% 29Si), which appeared to be instrumentally limited. Using AlternatingPhase CPMG sequence, we extended the spin coherence of one of the sites to 40 ms. Measurements with naturally abundant Si revealed that the Er electron spin coherence was limited by coupling to 29Si nuclear spins. The measured homogeneous linewidths of all 8 sites are below 100 kHz, and inhomogeneous broadening approaches 100 MHz [1]. These results were achieved for Er implanted from 200 and 700 nm from 28Si surface at 1016 cm3 level. The Er homogeneous linewidth and spin coherence were addressed using optical combbased spectral hole burning and optically detected magnetic resonance techniques. To enhance Er emission collection efficiency, samples were directly positioned atop specially fabricated superconducting single photon detectors and resonantly excited via fibre optics. The demonstration of a long spin coherence time and narrow optical linewidth in multiple sites show that Er in 28Si is an exceptional candidate for future quantum information and communication applications and can be used for single photon frequency multiplexing schemes.
[1] Ian R. Berkman et al, arXiv:2307.10021v2 (2023).






Creation and manipulation of Schrödinger cat states of a nuclear spin qudit in silicon
Xi Yu^{1}, Benjamin Wilhelm^{1}, Danielle Holmes^{1}, Arjen Vaartjes^{1}, Daniel Schwienbacher^{1}, Martin Nurizzo^{1}, Anders Kringhoj^{1}, Mark van Blankenstein^{1}, Alexander Jakob^{2}, Pragati Gupta^{3}, Fay Hudson^{4}, Kohei Itoh^{5}, Andrew Dzurak^{1}, Barry Sanders^{3}, David Jamieson^{2}, Andrea Morello^{1}
^{1}School of Electrical Engineering and Telecommunications, UNSW Sydney, Sydney, NSW 2052, Australia, ^{2}School of Physics, University of Melbourne, Melbourne, VIC 3010, Australia, ^{3}Institute for Quantum Science and Technology, University of Calgary, Alberta T3A 0E1, Canada, ^{4}Diraq Pty. Ltd., Sydney, NSW, Australia, ^{5}School of Fundamental Science and Technology, Keio University, Kohokuku, Yokohama, Japan
Abstract: Highdimensional quantum systems are a valuable resource for quantum information processing. They can be used to encode errorcorrectable logical qubits, for instance in continuousvariable states of oscillators such as microwave cavities or the motional modes of trapped ions. Powerful encodings include ‘Schrödinger cat' states, superpositions of widely displaced coherent states. Recent proposals suggest encoding logical qubits in highspin atomic nuclei, which can host hardwareefficient versions of continuousvariable codes on a finitedimensional system.
Here we demonstrate the creation and manipulation of Schrödinger cat states using the spin7/2 nucleus of a single antimony atom, embedded in a silicon nanoelectronic device. We use a coherent multifrequency control scheme to produce spin rotations that preserve the SU(2) symmetry of the 8level qudit. These SU(2)covariant rotations (CR) constitute logical Pauli operations for logical qubits encoded in the Schrödinger cat states. Together with the set of `virtualSNAP' gates, which impart an arbitrary phase on each qudit level, the CR are used to generate the cat state (see the pulse sequence in figure a). After the generation protocol, We measure an equal superposition of the two spincoherent states with opposite magnetization (figure b) and the Wigner function of the cat states (figure d) exhibits parity oscillations with a contrast up to 0.982(5) (figure c), and state fidelities up to 0.913(2).
Furthermore, a strong orientationdependent lifetime of the spin7/2 cat state is observed, with T2* values of 15.0(6) ms for cat states parallel to the spin quantization axis and 49(2) ms for perpendicular orientations. These findings hold promise for encoding logical qubits into perpendicular cat states, leveraging inherent biases in physical noise affecting nuclear spins.
These results demonstrate highfidelity preparation and logical control of nonclassical resource states and underscore the feasibility of quantum error correction on a single atomic site within a semiconductor platform.






Inferring the shape of a fewhole Ge quantum dot from magnetospectroscopy data
Mitchell Brickson, Andrew Miller, N Jacobson, TzuMing Lu, Dwight Luhman, Andrew Baczewski
Sandia National Laboratories
Abstract: The magnetic properties of hole quantum dots in Ge are sensitive to their shape due to
the interplay between strong spinorbit coupling and strong confinement. We show that
the inclusion of the splitoff band, surrounding SiGe layers, and holehole interactions
have a strong influence on calculations of the effective g factor of a lithographic
quantum dot in a Ge/SiGe heterostructure. Comparing predictions from such a detailed
model to raw magnetospectroscopy data, we apply maximumlikelihood estimation to
infer the shape of a quantum dot with up to four holes. We expect that methods like this
will be useful in assessing qubittoqubit variability critical to further scaling quantum
computing technologies based on spins in semiconductors.






Magnon propagation and localization in a 2x4 Ge quantum dot array
Daniel Jirovec^{1}, Pablo CovaFariña^{2}, Stephan Oosterhout^{3}, TzuKan Hsiao^{1}, Xin Zhang^{1}, Elizaveta Morozova^{1}, Amir Sammak^{4}, Giordano Scappucci^{1}, Menno Veldhorst^{1}, Lieven Vandersypen^{1}
^{1}TU Delft/ QuTech, ^{2}TU Delft/QuTech, ^{3}TNO / QuTech, ^{4}TNO/ QuTech
Abstract: Semiconductorbased quantum dot arrays are versatile platforms for analog quantum simulations, potentially offering insights into classically intractable manybody quantum phenomena with fewer resources compared to digital processors. The ability to engineer a variety of interesting regimes has led to the demonstration of exotic phases of matter, from Mott insulators and Nagaoka ferromagnetism to implementations of a Heisenberg spinchain and signatures of resonating valence bonds. However, for quantum advantage, large scale systems and new tuning strategies are required. Here, we present advancements in this direction in a 2x4 Gebased quantum dot array. We apply new tuning methods to the observation of magnon dynamics in a disordered system, where magnons represent spin excitations traveling through the array via nearestneighbor exchange interactions, amidst disorder provided by random effective gfactors in each dot, typical for holes in Ge.
With our improved gate design, we achieve exchange tunability up to 500 MHz, surpassing disorder by a factor of 50 at our operating magnetic field, while mitigating exchange crosstalk through a novel compensation method. By optimizing ramptimes and idling points, we leverage the Hamiltonian's features to initialize target spinstates and extract singlesite spinup probabilities across the array. This enables us to track magnon evolution in tailored configurations as depicted in the figure showing a quantum walk in weakly coupled double dots. By increasing exchange we observe a transition from localization to free propagation, a phase transition reminiscent of manybody localization, a phenomenon of significant recent interest.
Our experiment bridges single qubit properties with manybody physics concepts, indicating progress towards largescale analog simulators and realistic nearterm applications of semiconductor quantum dot systems. Furthermore, we anticipate that the techniques demonstrated here will be directly transferable to digital spin qubit processors, expanding their capabilities.





