Microscopic origin and predictive modeling of AC dissipation in epitaxial-graphene quantum Hall effect devices

Precision measurements on epitaxial-graphene quantum Hall devices established that their alternating-current (ac) dissipation factor and magnetocapacitance are tightly correlated, that in-plane electric fields cause roughly three orders of magnitude more ac dissipation than perpendicular fields, and that compressible puddles in the quantized bulk contribute to the loss — yet the microscopic mechanism behind the surprisingly large local dissipation factors (of order unity at 1 kHz), the magnetic-field dependence of the screening-edge width, and the residual stray in-plane fields of double-shielded device geometries remain without a quantitative predictive model. Later work reproduces and works around the phenomenology but does not resolve it: the most recent review still names the understanding and engineering of screening and dissipation an active international focus. The open core is a microscopically grounded, quantitative model of ac dissipation in quantized epitaxial graphene that predicts the measured magnetocapacitance and dissipation factor across voltage, magnetic field, and frequency, including the stray-field budget of real double shields.

Background

The quantum Hall effect (QHE) — the quantization of the Hall resistance of a two-dimensional electron gas (2DEG) at integer Landau-level filling factors — provides the primary representation of the ohm, and at alternating current the quantized Hall resistance (QHR) can serve as a quantum impedance standard from which the farad and the henry are directly realized. In ac operation, however, parasitic capacitive currents flow between the 2DEG and nearby metal electrodes, and any loss associated with them makes the measured Hall impedance frequency-dependent, corrupting its use as a standard. The loss is quantified by the dissipation factor (loss tangent) $\tan\delta$, the ratio of the real (dissipative) to the imaginary (reactive) part of the capacitive current; the reactive part is conventionally expressed as a magnetocapacitance $C(U,B)$, measured between the 2DEG and a metal electrode as a function of the applied rms voltage $U$ and magnetic field $B$. Because the loss can be suppressed by device design — in particular by the double-shield technique, in which the Hall bar is surrounded by top and back shield electrodes so that stray capacitive currents are intercepted and returned to the measurement loop, leaving predominantly perpendicular electric fields at the 2DEG — understanding what physical process sets the residual loss is central to impedance metrology. In the quantum Hall regime the edge of a 2DEG is described by electrostatic edge-channel theory (Chklovskii, Shklovskii and Glazman, 1992): the carrier density does not fall abruptly to zero at the device boundary but over a region containing compressible strips (which screen electric fields like a metal) and incompressible strips (which have no free states at the Fermi level and do not screen); in the quantized bulk, incompressible regions embed metallic compressible puddles. Schurr, Kalmbach, Ahlers, Hohls, Kruskopf, Müller, Pierz, Bergsten and Haug (Phys. Rev. B 96, 155443 (2017)) measured $C(U,B)$ and $\tan\delta(U,B)$ at cryogenic temperature on epitaxial graphene Hall bars with two contrasting electrode configurations — a side electrode producing strong in-plane fields, and a double-shield arrangement producing predominantly perpendicular fields — and compared them with a GaAs double-shield device. They found: (i) $C$ and $\tan\delta$ show the same functional dependence on $U$ and $B$ (up to sign and an offset), consistent with the Kramers–Kronig relations between the real and imaginary parts of a causal dielectric response; (ii) both the anticorrelated inverse-root voltage dependence of the side-electrode device and the positively correlated linear dependence of the double-shield device are limiting cases ($U\gg U_0$ and $U\ll U_0$) of one model, in which the width of the nonscreening edge region follows $d(U,B)=f(B)\,(a_{\mathrm{B}}l)^{1/2}\,[U_0/(U+U_0)]^{1/2}$ (with $a_{\mathrm{B}}$ an effective Bohr radius, $l$ a density-profile width, and $U_0$ a threshold voltage), and the dissipation factor follows $\tan\delta=(U/C)(dC/dU)\,\chi+\tan\delta_0$ with $\chi=\tan\varphi/(1+p)$, where $\tan\varphi$ is a local dissipation factor of the cyclically charged nonscreening 2DEG area and $p$ the exponent of the capacitance power law; (iii) in-plane field components cause about three orders of magnitude more ac dissipation than perpendicular ones ($\tan\delta\sim 10^{-1}$ versus below $10^{-4}$), qualitatively attributed to electrons being mobile within the 2DEG plane but confined in the third dimension; (iv) in the quantized state, compressible bulk puddles contribute measurably to the loss, with a fitted charging time constant of 59 $\mu$s; and (v) the extracted local dissipation factors are surprisingly large — $\tan\varphi=0.38\pm0.05$ and $0.55\pm0.05$ for the side-electrode graphene device at 14 T and 0 T, $1.30\pm0.12$ for the double-shielded graphene device, and $1.64\pm0.16$ for the GaAs device, all at 1 kHz. What the paper does not supply is any microscopic scattering or relaxation mechanism that would make $\tan\varphi$ of order unity, any explicit form of the function $f(B)$, any quantitative model of the stray in-plane field components that remain in a real (finite-size, imperfectly aligned) double shield, and any experimental test of the predicted weak-to-strong-field crossovers at voltages far outside the measured window. This matters because these are exactly the ingredients needed to predict, rather than empirically fit, the residual frequency dependence of a graphene-based quantum impedance standard — the goal the follow-up metrology literature has continued to pursue.

Problem Statement

Provide a predictive, microscopically grounded account of alternating-current dissipation in quantized epitaxial-graphene Hall devices, replacing the phenomenological parameters of Schurr et al. (2017) with derived physics. Concretely, a model of the cyclic charging of the nonscreening 2DEG regions must address all three of the following facets of this single open modeling question: (1) the microscopic scattering or relaxation mechanism responsible for the extracted local dissipation factors $\tan\varphi$ of order unity at kHz frequencies, and the observable consequences (magnitude, frequency dependence, weak magnetic-field dependence) that identify it; (2) the magnetic-field dependence $f(B)$ of the screening-edge width $d(U,B)=f(B)(a_{\mathrm{B}}l)^{1/2}[U_0/(U+U_0)]^{1/2}$ — derived from electrostatics rather than absorbed into an unspecified prefactor, together with a physical determination of the threshold voltage $U_0$ instead of a fit value; and (3) a quantitative model of the stray in-plane electric-field components of nominally double-shielded device geometries, including their dependence on shield alignment and finite-size effects, establishing how closely real double shields can approach the perpendicular-field dissipation bound ($\tan\delta\lesssim 10^{-4}$) in production devices. The model must be valid across both the weak-field ($U\ll U_0$) and strong-field ($U\gg U_0$) limiting regimes of the same general voltage dependence, must be consistent with the measured magnetocapacitance and dissipation-factor behavior of the side-electrode and double-shield graphene devices (the correlated voltage dependences, the about-three-orders-of-magnitude in-plane/perpendicular directionality, and the compressible-bulk-puddle contribution to the loss in the quantized state), and should predict the residual frequency dependence of the Hall resistance that a real shielded device will show. The published datasets — including the fitted 59 $\mu$s bulk-puddle charging time and the GaAs comparison-device measurements — remain available as experimental constraints on any such model.

References

1. J. Schurr, C.-C. Kalmbach, F. J. Ahlers, F. Hohls, M. Kruskopf, A. Müller, K. Pierz, T. Bergsten, and R. J. Haug, "Magnetocapacitance and dissipation factor of epitaxial graphene-based quantum Hall effect devices," Physical Review B 96, 155443 (2017). DOI: 10.1103/PhysRevB.96.155443, https://doi.org/10.1103/PhysRevB.96.155443
2. D. B. Chklovskii, B. I. Shklovskii, and L. I. Glazman, "Electrostatics of edge channels," Physical Review B 46, 4026–4034 (1992). DOI: 10.1103/PhysRevB.46.4026, https://doi.org/10.1103/PhysRevB.46.4026
3. F. Delahaye, B. P. Kibble, and A. Zarka, "Controlling ac losses in quantum Hall effect devices," Metrologia 37, 659–670 (2000). DOI: 10.1088/0026-1394-37-6/3, https://doi.org/10.1088/0026-1394-37-6/3
4. B. P. Kibble and J. Schurr, "A novel double-shielding technique for ac quantum Hall measurement," Metrologia 45, L25–L27 (2008). DOI: 10.1088/0026-1394-45-5/N01, https://doi.org/10.1088/0026-1394-45-5/N01
5. J. Schurr, B. P. Kibble, G. Hein, and K. Pierz, "Controlling losses with gates and shields to perfect a quantum Hall impedance standard," IEEE Transactions on Instrumentation and Measurement 58, 973–979 (2009). DOI: 10.1109/TIM.2008.2006961, https://doi.org/10.1109/TIM.2008.2006961
6. J. Schurr, F. Ahlers, and K. Pierz, "Magnetocapacitance and loss factor of GaAs quantum Hall effect devices," Metrologia 51, 235–242 (2014). DOI: 10.1088/0026-1394-51-3/235, https://doi.org/10.1088/0026-1394-51-3/235
7. A. F. Rigosi and R. E. Elmquist, "The quantum Hall effect in the era of the new SI," Semiconductor Science and Technology 34, 093004 (2019). DOI: 10.1088/1361-6641/ab37d3, https://doi.org/10.1088/1361-6641/ab37d3
8. M. Kruskopf, S. Bauer, Y. Pimsut, A. Chatterjee, D. K. Patel, A. F. Rigosi, R. E. Elmquist, K. Pierz, E. Pesel, M. Götz, and J. Schurr, "Graphene quantum Hall effect devices for AC and DC electrical metrology," IEEE Transactions on Electron Devices 68, 3672–3677 (2021). DOI: 10.1109/TED.2021.3082809, https://doi.org/10.1109/TED.2021.3082809
9. D. Chae, M. Kruskopf, J. Kučera, J. Park, N. T. M. Tran, D. B. Kim, K. Pierz, M. Götz, Y. Yin, P. Svoboda, P. Chrobok, F. Couëdo, and F. Schopfer, "Investigation of the stability of graphene devices for quantum resistance metrology at direct and alternating current," Measurement Science and Technology 33, 065012 (2022). DOI: 10.1088/1361-6501/ac4a1a, https://doi.org/10.1088/1361-6501/ac4a1a
10. Y. Pimsut, S. Bauer, M. Kraus, R. Behr, M. Kruskopf, O. Kieler, and L. Palafox, "Development and implementation of an automated four-terminal-pair Josephson impedance bridge," Metrologia 61, 025007 (2024). DOI: 10.1088/1681-7575/ad2539, https://doi.org/10.1088/1681-7575/ad2539
11. B. Chenaud, C. Hernández, A. Delgard, and C. Chaubet, "Influence of coaxial cable resistance on impedance measurements at low frequency and guidelines for phase determination of ac electrical signals," Review of Scientific Instruments 94, 115112 (2023). DOI: 10.1063/5.0172200, https://doi.org/10.1063/5.0172200
12. D.-H. Chae, H. He, and M. Kruskopf, "Recent advances and perspectives in quantum electrical metrology with epitaxial graphene," Measurement Science and Technology 37, 151001 (2026). DOI: 10.1088/1361-6501/ae57c8, https://doi.org/10.1088/1361-6501/ae57c8
13. J. Schurr and S. A. Awan, "Precision measurement of a resistance standard at frequencies up to 2 MHz," Measurement Science and Technology 37, 085801 (2026). DOI: 10.1088/1361-6501/ae46c4, https://doi.org/10.1088/1361-6501/ae46c4