Charge-Induced Artifacts in Nonlocal Spin-Transport Measurements: How to Prevent Spurious Voltage Signals

Editors' Suggestion

Charge-Induced Artifacts in Nonlocal Spin-Transport Measurements: How to Prevent Spurious Voltage Signals

Frank Volmer, Timo Bisswanger, Anne Schmidt, Christoph Stampfer, and Bernd Beschoten

Phys. Rev. Applied 18, 014028 – Published 13 July 2022

Abstract

To conduct spin-sensitive transport measurements, a nonlocal device geometry is often used to avoid spurious voltages that are caused by the flow of charges. However, in the vast majority of reported nonlocal spin-valve, Hanle spin precession or spin Hall measurements, background signals have been observed that are not related to spins. We discuss seven different types of these charge-induced signals and explain how these artifacts can result in erroneous or misleading conclusions when falsely attributed to spin transport. The charge-driven signals can be divided into two groups: signals that are inherent to the device structure and/or the measurement setup and signals that depend on a common-mode voltage. We designed and built a voltage-controlled current source that significantly diminishes all spurious voltage signals of the latter group in both dc and ac measurements by creating a virtual ground within the nonlocal detection circuit. This is especially important for lock-in-based measurement techniques, where a common-mode voltage can create a phase-shifted, frequency-dependent signal with an amplitude several orders of magnitude larger than the actual spin signal. Measurements performed on graphene-based nonlocal spin-valve devices demonstrate how all spurious voltage signals that are caused by a common-mode voltage can be completely suppressed by such a current source.

Received 6 December 2021
Revised 17 May 2022
Accepted 20 May 2022

DOI:https://doi.org/10.1103/PhysRevApplied.18.014028

Physics Subject Headings (PhySH)

Charge Metrology Spintronics Transport phenomena

Graphene Spin valves

Spin

Transport techniques

Condensed Matter, Materials & Applied Physics

Authors & Affiliations

Frank Volmer ^1,2, Timo Bisswanger ¹, Anne Schmidt¹, Christoph Stampfer ^1,3, and Bernd Beschoten ^1,*

¹2nd Institute of Physics and JARA-FIT, RWTH Aachen University, Aachen 52074, Germany
²AMO GmbH, Advanced Microelectronic Center Aachen (AMICA), Aachen 52074, Germany
³Peter Grünberg Institute (PGI-9), Forschungszentrum Jülich, Jülich 52425, Germany

^*bernd.beschoten@physik.rwth-aachen.de

Article Text (Subscription Required)

Click to Expand

Supplemental Material (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 18, Iss. 1 — July 2022

Subject Areas

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
(a) Lock-in-based, nonlocal measurement scheme of a lateral spin-valve device. A charge current gets spin polarized by flowing through the ferromagnetic electrodes $I_{+}$ and $I_{-}$ and, therefore, creates the quasichemical potentials depicted in panel (b) for majority and minority spins. (c) As the spin signal is directly linked to the current in the injection circuit, both signals should be in phase with each other (the latter can be measured, for example, over a shunt resistor). As long as the voltage-to-current converter is driving the current in phase to the applied reference voltage from the lock-in, the spin signal should appear only in the $X$ channel of the lock-in. In real measurements charge-induced signals also appear that can be either in phase or out of phase with the applied current. (d) Without these spurious signals, both spin-valve (solid line) and Hanle spin precession (dashed lines) measurements should be symmetric around $R_{nl} = 0$ . Instead, in experiments a background signal is almost always measured that changes, e.g., by going from the electron into the hole regime (e) or by merely increasing the measurement frequency (f).
Reuse & Permissions
Figure 2
(a) Schematic top view of the transport channel illustrating the effect of current spreading. Because of a spatially asymmetric injection of charge carriers, a small part of the overall current flows on a curved path from $I_{+}$ into the nonlocal part of the device before it drains into $I_{-}$ . This leads to a measurable potential difference between the two voltage probes $V_{+}$ and $V_{-}$ (colored lines illustrate equipotential lines), which can be changed by an out-of-plane magnetic field (dashed lines). (b) Simplified equivalent circuit of both the device under test and the measurement setup. Considered are the contact resistances $R_{C i}$ , the resistances $R_{M i}$ of the material, the capacitances $C_{W i}$ of the wiring, and the input impedances $Z_{I i}$ of the measurement equipment. The bias voltage $V_{bias} (t)$ that is necessary to drive a current through the injection circuit can, in principle, be freely distributed ( $0 \leq α \leq 1$ ) between electrodes $I_{+}$ and $I_{-}$ . However, there is only one value of $α$ , which depends on the voltage divider consisting of $R_{C 1}$ , $R_{M 1}$ , and $R_{C 2}$ , that shifts the whole nonlocal part of the device into a virtual ground (green letters). For every other setting of $α$ , a nonzero common-mode voltage $V_{CM} (t) \neq 0$ will drive currents over every component that is referenced to ground in the nonlocal part of the device (red arrows).
Reuse & Permissions
Figure 3
Demonstration on how changing the $R C$ time constant of the $V_{+}$ electrode impacts the measured nonlocal voltage. (a),(b) Spin-valve measurements are repeated with additional resistors and BNC cables (100 pF/m) connected to the $V_{+}$ electrode. The impact on the recorded nonlocal signal is depicted in panel (a) for the $X$ channel and in panel (b) for the $Y$ channel of the lock-in. (c) Equivalent circuit that is used to simulate the effect of varying $R C$ time constants. (d) Corresponding amplitude $M = \sqrt{X^{2} + Y^{2}}$ and phase of the nonlocal voltage measured between $V_{+}$ and $V_{-}$ if the resistance $R_{C 3}$ in (c) is varied for a fixed capacitance $C_{W}$ of the external wiring. (e) Same as in panel (d) but for a fixed value $R_{C 3}$ and different values for $C_{W}$ .
Reuse & Permissions
Figure 4
(a) Equivalent circuit that is used to simulate the detrimental impact of additional $R C$ low-pass filters on the nonlocal spin measurement. The high capacitances to ground of these filters significantly increase the spurious nonlocal signals that are caused by the charging and discharging currents over these capacitors. (b) The simulated signals in the $X$ channel (solid lines) and $Y$ channel (dashed lines) as a function of the applied measurement frequency for different resistances $R_{M}$ of the material. (c),(e) To verify that the readout of the $X$ channel actually captures the whole spin signal, a rotation of the coordinate system by a reference phase $φ_{ref}$ can be utilized. As only the projection of each signal on the respective axis is measured, the amplitude of the spin signal should decrease as soon as the $X$ axis is rotated. (d) Corresponding experiment in which the amplitude of the spin-related switching in a spin-valve measurement is recorded as a function of $φ_{ref}$ . The maximum around $0^{\circ}$ demonstrates that the presented current source creates a current signal and, therefore, a spin signal that is in phase with the output of the lock-in [compare to Fig. 1].
Reuse & Permissions
Figure 5
Demonstration that charge-induced signals can be significantly reduced by adjusting the virtual ground with the help of the presented current source. (a) Hanle curves measured with a symmetric bias applied to electrodes $I_{+}$ and $I_{-}$ (red curves) show a distinct background signal. Instead, the green curves depict the same Hanle measurement after the nonlocal part of the device is put to a virtual ground. This prevents the flow of any charge current in the detection circuit over resistances or capacitances to ground and, therefore, diminishes the charge-induced background signals. (b),(c) Nonlocal signal as a function of measurement frequency for both the $X$ channel (b) and the $Y$ channel (c). For these graphs, the values in the parallel configuration of the Hanle curves at $B = 0 T$ are measured for both bias conditions. In the case of a symmetric bias voltage, the charge-induced signal gets much larger than the actual spin signal, especially in the $Y$ channel. Instead, in the case of the adjusted virtual ground, the signal in the $Y$ channel is negligible over the whole measurement range and the signal in the $X$ channel consists of only the actual spin signal.
Reuse & Permissions

Physical Review Applied