Automated Extraction of Capacitive Coupling for Quantum Dot Systems

Editors' Suggestion

Automated Extraction of Capacitive Coupling for Quantum Dot Systems

Joshua Ziegler, Florian Luthi, Mick Ramsey, Felix Borjans, Guoji Zheng, and Justyna P. Zwolak

Phys. Rev. Applied 19, 054077 – Published 24 May 2023

Abstract

Gate-defined quantum dots (QDs) have appealing attributes as a quantum computing platform. However, near-term devices possess a range of possible imperfections that need to be accounted for during the tuning and operation of QD devices. One such problem is the capacitive crosstalk between the metallic gates that define and control QD qubits. A way to compensate for the capacitive crosstalk and enable targeted control of specific QDs independent of coupling is by the use of virtual gates. Here, we demonstrate a reliable automated capacitive coupling identification method that combines machine learning with traditional fitting to take advantage of the desirable properties of each. We also show how the cross-capacitance measurement may be used for the identification of spurious QDs sometimes formed during tuning experimental devices. Our systems can autonomously flag devices with spurious dots near the operating regime, which is crucial information for reliable tuning to a regime suitable for qubit operations.

Received 30 January 2023
Accepted 12 April 2023

DOI:https://doi.org/10.1103/PhysRevApplied.19.054077

Physics Subject Headings (PhySH)

Capacitance Quantum control Quantum gates Quantum information architectures & platforms Quantum information with solid state qubits

Quantum dots Semiconductor compounds

Machine learning

Condensed Matter, Materials & Applied PhysicsQuantum Information, Science & Technology

Authors & Affiliations

Joshua Ziegler ^1,†, Florian Luthi², Mick Ramsey², Felix Borjans ², Guoji Zheng ², and Justyna P. Zwolak ^1,3,*

¹National Institute of Standards and Technology, Gaithersburg, Maryland 20899, USA
²Intel Components Research, Intel Corporation, 2501 NW 229th Avenue, Hillsboro, Oregon 97124, USA
³Joint Center for Quantum Information and Computer Science, University of Maryland, College Park, Maryland 20742, USA

^*jpzwolak@nist.gov
^†Current address: Intel Components Research, Intel Corporation, 2501 NW 229th Avenue, Hillsboro, Oregon 97124, USA.

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 19, Iss. 5 — May 2023

Subject Areas

Reuse & Permissions

Access Options

Article Available via CHORUS

Download Accepted Manuscript

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
An example 2D scan and corresponding pixel classification, class clusters, and linear fits. (a) A simulated voltage scan showing left and right transitions as well as a polarization line. (b) Pixel classification for the scan shown in (a). (c) Regions of pixels and linear fits from the pixel classification. The large dark points indicate the centers of pixel regions.
Reuse & Permissions
Figure 2
(a) Box plot of root-mean-square error (RMSE) for all transition classes (left, central, and right [LT, CT, RT]) as a function of the synthetic noise level. The notch indicates the 95% confidence level. (b) RMSE as a function of noise level for the LT class. (c) RMSE as a function of noise level for the RT class.
Reuse & Permissions
Figure 3
(a) Large experimentally measured charge stability diagram with a scatter plot of centers of pixel class regions overlaid. The colors of the points indicate the virtual gate off-diagonal values identified by fits to the region. The sizes of the points indicate the weights used when averaging. Only points with relative error less than 20% are plotted. (b),(c) Charge stability diagram after applying virtual gates acquired near the $(0, 0)$ – $(1, 1)$ charge transition in (b) and near the $(1, 3)$ – $(2, 4)$ charge transition in (c). In both (b) and (c) the virtualization is performed off-line, via an affine transform to the original scan shown in (a), and the points are plotted using the same parameters as in (a).
Reuse & Permissions
Figure 4
Histograms of the off-diagonal elements of the virtualization matrix for an experimentally measured scan shown in Fig. 3 as a function of plunger gates (a) $V_{P_{1}}$ and (b) $V_{P_{2}}$ . Off-diagonal values are shifted and scaled by the mean of the off-diagonal elements from virtual gates in the $(1, 1)$ charge state for ease of comparison. Virtual gate values are extracted from a strip of small scans shifted by 3 mV (two pixels) in each opposing direction to better visualize variation at each plunger gate value.
Reuse & Permissions
Figure 5
Spurious dot detection. The top row shows two charge stability diagrams capturing properly formed QD [panels (a),(b)] and three charge stability capturing spurious QD [panels (c)–(e)]. The black boxes in (b) and (e) highlight small 2D scans, denoted by $V_{R_{1}}$ and $V_{R_{2}}$ , typical of the autotuning approaches proposed in Refs. [13, 14]. Panels (f)–(j) in the middle row show pixel classification results for the charge stability diagrams shown in the top row. Plots of fitting results used to determine whether a spurious QD is present are shown in the bottom row [panels (k)–(o)]. The different groups of transition are shown with different shades of green. The monotonicity within each group of transitions is clearly visible in panels (k) and (l). In contrast, in the three plots shown in panels (m)–(o), there is a clear divergence from the expected trend for the spurious QD, as indicated by the nonmonotonic change between the first (leftmost point) and second (middle point) transition. Error bars indicate one standard deviation.
Reuse & Permissions

Physical Review Applied