Introduction

Since the spin field effect transistor was proposed by Datta and Das in 1990s, due to the difficulties of spin injection and detection, until now there is still no efficient spin field effect transistor developed1,2,3. Recently, the spin Hall transistor based on two-dimensional electron gas has been demonstrated, in which the spins were generated optically rather than electrically in the semiconductor channel4. However, all-electrical manipulation of the magnetization is more desirable for spintronic applications. Approaches to control the spin of magnet bits electrically, such as spin-orbit torque5,6,7,8,9,10,11,12,13,14,15, magneto-electrical coupling16,17, voltage18,19, polarized light20 and piezo voltages21,22 have been proposed based on ferromagnetic metals. Among them, the piezo voltage is one of the most effective methods to control the magnetization switching, in which a deformation of the crystal structure of the magnetic materials induces a change of the magnetocrystalline anisotropy which is directly related with the spin-orbit interaction in the crystal23,24,25,26,27,28. The control of the charge transport in semiconductors by piezo voltages has also been demonstrated for high speed piezotronics, which has been proposed for post Complementary Metal Oxide Semiconductor (CMOS) technology29.

Here we demonstrate planar Hall effect devices in which the planar Hall resistances/voltages of ferromagnetic Co2FeAl devices can be tuned by piezo voltages from positive (negative) to negative (positive) effectively, which is associated with magnetization switching in the plane by 90°. Room temperature magnetic NOT and NOR gates are demonstrated based on the Co2FeAl piezo voltage controlled planar Hall effect devices without the external magnetic field. The simple device structure allows us to build large scale building blocks for future logics.

Results

The design of piezo voltage controlled planar Hall effect devices

The Heusler alloy Co2FeAl films used in our planar Hall effect devices were grown by molecular beam epitaxy (MBE)28,30. The magnetization of the Co2FeAl is controlled by the uniaxial deformation induced by a piezo voltage applied to a PZT transducer and the detection is provided by the planar Hall voltage across the Hall bar devices (See Methods and Supplementary S1). Schematic diagrams of the two Co2FeAl Hall bars with the respect to the GaAs crystal orientation along [100] and [010] axes are shown in Fig. 1a. We first fully magnetized the Hall bar devices along the [110] direction with magnetic field of 500 Oe (which is much larger than the coercive field), then swept the external magnetic field to zero. The measured RH with periodic piezo voltage (UP) pulses applied to both devices is shown in Fig. 1b. A constant Hall voltage offset of the Hall device has been removed. Strikingly, the RH of the [100] orientation device periodically switched from −0.12 to 0.12 Ω with UP changing from 0 to −30 V. In contrast, the RH of the [010] orientation device was periodically switched simultaneously from 0.12 to −0.12 Ω. The RH transitioned from negative low value to positive high value for sample [100], whereas the [010] orientated device changed from positive high to negative low under the same range of UP, which is analogous to the n-FET and p-FET in CMOS technology. The advantage of this planar Hall effect device is that the RH (induced by planar Hall effect) changes sign during the tuning, whereas the conventional FET is switched on and off by accumulating and depleting the electrons (holes), while the resistance of the FET is always positive. In the piezo voltage control planar Hall effect devices, the changing of RH sign originates from the rotation of the magnetization vector respective to the electrical current under piezo voltages.

Figure 1
figure 1

The planar Hall resistance of Co2FeAl device structure controlled by the piezo voltages.

(a) The schematic diagrams for measurements of planar Hall effect and application of piezo voltages for both the [010] and [100] orientated devices. (b) The periodic changes of the RH for both the [010] and [100] orientated devices with the periodic change of the piezo voltage between 0 V and −30 V without external magnetic field. (c) The change of the RH induced by a change of the piezo voltage from 0 V to certain values for both the [010] and [100] orientated devices. (d) The dependence of the RH for both the [010] and [100] orientated devices on the angle of magnetization under a fixed magnetic field at 2,000 Oe rotated in the plane, where the dots are the experimental results and the lines are the fitted results.

Full size image

The change of the planar Hall resistance (ΔRH) on switching the piezo voltage from zero to certain values for both the [010] and [100] orientated devices is shown in Fig. 1c. The ΔRH remains almost zero with switching the UP from 0 to values above −27 V, but a sharp jump and then a flat plateau were observed for both devices with switching the piezo voltages to a further negative value. Opposite values of ΔRH (0.25 Ω and −0.25 Ω) were observed for devices along [100] and [010] orientations at the plateau. The dependence of the RH on the angle of magnetization (with respect to the [110] direction) is shown in Fig. 1d for both devices at zero UP. The magnetization is rotated using an applied magnetic field of 2,000 Oe, which is large enough to force the magnetization vector to follow the field direction. The angular dependence of the RH can be fitted well using the single domain model31, , where is the sheet resistance with the current parallel to the magnetization, is the sheet resistance with the current perpendicular to the magnetization, θ + π/4 and θ−π/4 represent the angle between the electrical current and the magnetization vector for [010] and [100] orientated devices respectively and γ is the deviation angle between the applied strain direction and [110] direction which is about 2°. The periodic RH for both devices has the same magnitude and frequency but with π/2 phase shift. The maximum magnitude of the Hall resistance occurs when θ = nπ/2, where n is an integer. The angular dependence of RH shows that switching magnetization by 90° induces a change ΔRH of 0.27 Ω, which coincides with the large ΔRH value in Fig. 1c. Thus, the piezo voltage can fully switch the magnetization of the Co2FeAl devices by 90° in the plane.

The magnetic properties controlled by the piezo voltages

To obtain further insight into the switching behavior of the Co2FeAl planar Hall effect devices, the magnetic hysteresis loops various UP were investigated using longitudinal magneto-optical Kerr microscopy (LMOKM). Figure 2a shows the Kerr rotation angle during the magnetization reversal at zero UP (without deformation) with magnetic field applied close to the in-plane major crystalline [110], and [010] orientations, respectively. [110] orientation is the easy axis and the hard-axis-like behavior is seen for the [010] direction with saturation occurring around 220 Oe. However, the orientation hysteresis loop shows a two-step switching with a continuous reversible rotation between the two steps. The observed magnetic hysteresis loops are the consequence of the superposition of a uniaxial and a fourfold anisotropy, where the uniaxial easy axis is along [110] orientation under zero strain (UP = 0) and the fourfold easy axes are along [110] and orientations32,33. To demonstrate the evolution of the magnetic anisotropy under piezo voltages, the hysteresis loops along of Co2FeAl with UP at zero and ±30 V are plotted in Fig. 2b. With UP = 30 V, although the two-step jumping was also observed during the magnetization reversal, the field range of the two sharp jumps increased dramatically by more than a factor of two. The continuous reversible region between the two-step jumps increases with increasing the applied the piezo voltage (details in Supplementary Fig. S2), indicating the axis becomes harder. With UP = −30 V as shown in Fig. 2b, one step magnetization reversal along orientation was observed, suggesting the magnetic easy axis is switched by 90° from [110] to in the plane with UP at −30 V.

Figure 2
figure 2

The magnetic states of Co2FeAl device controlled by the piezo voltages.

(a) The magnetic hysteresis loops of Co2FeAl device measured by LMOKM with magnetic field applied in the [110], and [010] directions with piezo voltage at zero. (b) The magnetic hysteresis loops with magnetic field in orientation with piezo voltages at −30, 0 and 30 V. (c) The magnetic domain images (a–e) of the Co2FeAl device without deformation during the magnetization reversal by applied magnetic field along orientation. The magnetic domain images (f–j) of the Co2FeAl device controlled by piezo voltages without external magnetic field. (d) The angular dependence of the magnetic energy density at zero magnetic field for the Co2FeAl device with piezo voltages at −30, 0 and 30 V, where the minimum energy is in [110]/ direction for both 0 and 30 V (stretched) and it is switched to / orientation with piezo voltage at −30 V (compressed).

Full size image

The magnetic domain images of the Co2FeAl [100] orientated device without deformation were taken by LMOKM with magnetic field applied in orientation, which are shown in Fig. 2c(a–e). The magnetization corresponding to, [110] and oriented domains are marked with arrows. Domain wall propagation could be detected during the sweeping of magnetic field around −12 Oe and 33 Oe, which was associated with the two-step jumps of the hysteresis loop along in Fig. 2a. Figure 2c(f–j) represents the domain images of the device controlled by piezo voltages without magnetic field, which clearly shows the magnetization switched between and [110] axes.

The 90° manipulation of magnetization in the piezo voltage controlled Co2FeAl planar Hall effect device is due to an extra uniaxial anisotropy introduced by piezo voltages. The magnetic energy density of the system without deformation can be written as34,35:

where θ is the angle between magnetization and easy axis [110] direction, α is the angle between the external magnetic field and [110] direction, KC is cubic anisotropy, KU is the uniaxial anisotropy, MS is the saturated magnetization. The magnetic anisotropy constants can be obtained by analyzing the magnetic hysteresis loops along the uniaxial hard orientation, in the reversible region between the two-step jumps (details in Supplementary S3). The KC and KU were obtained to be (108 ± 5)MS and (41 ± 2)MS, respectively. Then the angular dependence of the magnetic energy can be plotted, as shown in Fig. 2d. An additional strain-induced uniaxial magnetic anisotropy term KPsin2(θ − γ) is added to the magnetic energy density equation (1) when UP≠ 0, where KP has the same/opposite sign of KU at positive/negative piezo voltages. The obtained KP/MS is (26 ± 1) and (−46 ± 4) Oe for UP at 30 and −30 V, respectively. The angular dependence of the magnetic energy at UP = ±30 V is also plotted in Fig. 2d. The lowest energy state ε/MS is along [110] orientation for UP at 0 and 30 V. However, the lowest energy is along orientation at UP = −30 V because the piezo voltage induced KP is larger than that of the KU but with opposite sign. The magnetic energy landscape in Fig. 2d suggests piezo voltages can switch the magnetic easy axis by 90°, which has also been confirmed using ferromagnetic resonance28.

The logic operation architectures of NOT and NOR logic gates

In the present work, we propose and demonstrate NOT and NOR logic gates using the piezo voltage controlled planar Hall effect devices. Firstly, a high Hall voltage state (ON, digital ‘1’) can be defined as an output voltage of +5 μV or larger. A low Hall voltage state (OFF, digital ‘0’) is defined as an output voltage of + 2 μV or smaller. For the piezo voltages, the 0 V and −30 V correspond to the logic state ‘0’ and ‘1’. Based on the single piezo voltage controlled planar Hall effect device shown in Fig. 3a, the piezo voltages can effectively switch the magnetization between the [110] and magnetic states, which function as the NOT gate and produce the planar Hall voltage output as shown in Fig. 1b. The truth table in Fig. 3b represents the NOT gate operation . The NOR gate was built based on one [100] and one [010] planar Hall effect device connected as shown in Fig. 3c. The two piezo voltages (UP1 and UP2) control the [010] and [100] orientated devices separately. The magnetizations of two devices were preset to [110] orientation by external magnetic field. Then all the operations were executed without magnetic fields. Inputting the [0, 0] to the logic with both piezo voltages at zero sets the magnetization of both devices along [110], so that the Vo is 11.8 μV and the output equal to 1. When the magnetizations of both devices is rotated to direction by the piezo voltages, corresponding to the magnetic state [1, 1], the Vo is −11.6 μV < 2 μV and output = 0. If only switching the magnetization of [100] or [010] device to orientation, corresponding to the [1, 0] or [0, 1] magnetic states, the output is 0 since both Vo are less than 2 μV (as shown in Fig. 3d). The non-zero value of Hall voltage at [1, 0] and [0, 1] states is because the two devices are slightly different. Figure 3e illustrates the NOR gate logic operations depending on the four separately measured values as defined in the Fig. 3d. The experimental in-situ measured output voltages were shown in the Supplementary Fig. S4, indicating that the logic devices have very good reproducibility.

Figure 3
figure 3

Programmable logic operations demonstrated by a NOT and a NOR gate.

(a) The schematic diagram of a piezo voltage controlled [100] orientated Co2FeAl device built for NOT gate. (b) Truth table summary of the operation described in NOT gate. (c) The schematic diagram of piezo voltages controlled [010] and [100] Co2FeAl devices built for NOR gate, where the piezo voltages UP1 and UP2 are for the [010] and [100] devices, respectively. (d) Truth table summary of operation for the NOR gates with piezo voltage input. (e) Illustration of the NOR logic states operation with four separately measured output values as shown in the Fig. 3d.

Full size image

Discussion

In this experiment, we demonstrated that the piezo voltage is an effective way to control the magnetization and the resulting value of planar Hall voltage. When the piezo voltage UP > −27 V, the magnetization is varied slightly and the Hall voltage change is small. However, when the piezo voltage UP < −27 V, the Hall voltage has an obvious jump. It shows that the magnetization can be manipulated effectively by the piezo voltage.

The logic gates realized by the piezo voltage controlled planar Hall effect devices are desirable to be applied to the current computation technology. The time response of the piezo voltage controlled device has been investigated, where the rising and falling time are 220 μs and 70 μs, respectively (details in Supplementary S5). However, a smaller input voltage and a faster magnetization switching for the piezo voltage control should be achieved. Using the methodology of integrating the piezoelectric layers, the application of a smaller input voltage switching the magnetization is achievable by scaling down the devices to nanometer sizes36. Using the thin piezoelectric layer can also largely reduce the impedance thus largely increasing the device response speed. Our demonstration could pave a new way for the spin logic, realizing the information processing in only ferromagnetic metals.

Methods

Sample preparation

The Heusler alloy Co2FeAl (CFA) thin film was grown on semi-insulating GaAs (001) by using molecular beam epitaxial (MBE) technique at 280 °C. After deposition of 10 nm-thick CFA, the film was capped with an aluminum layer of 3 nm to avoid oxidation. The Hall bar devices along different in-plane major crystalline orientations were fabricated using standard photolithography and ion beam etching, where the device width is 20 μm and the distance between the neighbor arms is 80 μm. After polishing the GaAs substrate down to 100 μm thickness, the devices were bonded to the piezoelectric ceramic transducer (PZT) by glue. In order to ensure the deterministic switching of the Co2FeAl moment, all the devices were bonded ~2° from the [110] to [010] direction. The positive/negative voltage produces a uniaxial tensile/compressive strain in [110] orientation perpendicular to the stacks (the direction of strain is marked in Fig. 1a). The deformation of the Co2FeAl devices under the piezo voltage was found to be linearly changed with the applied piezo voltages, with tensile strain under positive piezo voltages and compressive strain under negative piezo voltages (See Supplementary Fig. S1).

The measurements of the sample

The magnetization vectors and the magnetic domains of the devices during magnetization reversal along in-plane orientations were measured by longitudinal magneto-optical Kerr microscopy (Nano MOKE3). At different deformation states controlled by the piezo voltages, a fixed current I = 50 μA passing through the channel was used to perform all the magnetotransport measurements, where the planar Hall voltages were detected using a Keithley nanovolt meter 2182. The time-dependent response of the device was measured by MOKE3 and the piezo voltages were applied by Agilent B1500A and the leading and trailing time are both 100 ns.

Additional Information

How to cite this article: Zhang, B. et al. Piezo Voltage Controlled Planar Hall Effect Devices. Sci. Rep. 6, 28458; doi: 10.1038/srep28458 (2016).