Room Temperature Planar Hall Transistor

Bao Zhang1, Kangkang Meng

1, Mei-Yin Yang

1, K. W. Edmonds

2, Hao Zhang

1,

Kai-Ming Cai1, Yu Sheng

1,3, Nan Zhang

1, Yang Ji

1, Jian-Hua Zhao

1, Kai-You Wang

1*

1 SKLSM, Institute of Semiconductors, CAS, P. O. Box 912, Beijing 100083, People’s

Republic of China

2 School of Physics and Astronomy, University of Nottingham, Nottingham NG7 2RD, United

Kingdom

3 Department of Physics, School of Sciences, University of Science & Technology

Beijing, Beijing 100048, China

Controlling the spin transport in solids electrically is central to the application of

spintronics for the development of information technology1. Since the

spin-transistor was proposed by Datta and Das in 1990s, due to the difficulties of

spin injection and detection in semiconductors, until now there is still no efficient

spin field effect transistor developed2-4

. 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 channel5. Here

we report the tunability of the planar Hall resistance in ferromagnetic half metal

Co2FeAl devices solely by piezo voltages from positive to negative and from

negative to positive, which can be analogously used as the n-type and p-type field

effect transistor, respectively. The magnetic NOT and NOR gates are

demonstrated based on the Co2FeAl planar Hall transistors without external

magnetic field at room temperature. Our demonstration can pave a way for the

application of future spintronics, realizing both the information storage and

processing using ferromagnetic materials.

Apart from the proposals of spin based logic in semiconductors6,7

,approaches to

control the domain mall motion, spin waves, spin Hall effect electrically using spin

transfer torque, spin-orbit torque, spin Hall effect, magnetoelectrical coupling and

* Corresponding author’s E-mail: kywang@semi.ac.cn

piezo voltages etc. have been proposed for spin logic based on ferromagnetic

metals8-17

. Among them, the piezo voltage is one of the most effective methods to

control the magnetization, in which a deformation of the crystal structure of the

magnetic materials, induces a change in the magnetocrystalline anisotropy which is

directly related with the spin orbit interaction in the crystal18-24

. 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) technology25

. Here we propose and achieve planar

Hall transistors controlled by piezo voltages which can be used for both the storage

and logic devices.

The active layer in our devices is the half metallic Co2FeAl24,26

. The

magnetization of the Co2FeAl is controlled by the piezo voltages and the detection is

provided by the planar Hall voltage across the Hall bar devices. The planar Hall

voltage in magnetic materials is sensitive to the relative direction between the

electrical current and magnetization vector in the plane27

. Here we demonstrate the

planar Hall resistance (RH) of the half metallic 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. Without the

semiconductor channel, the functionality of n-type field effect transistors (n-FET) and

p-type field effect transistors (p-FET) can be realized in our planar Hall transistors

using half magnetic metals by piezo control of the magnetization switching. Utilizing

the planar Hall transistors, we demonstrate the NOT and NOR logic operations. The

simple device structure allows us to build large scale building blocks for future logics.

The half metallic Co2FeAl (CFA) thin film was grown on semi-insulating GaAs

(001) by using molecular beam epitaxial (MBE) technique at 280℃28. The Hall bar

devices along different in-plane major crystalline orientations were fabricated using

standard photolithography (See Methods). After polishing the GaAs substrate down to

100 m, the devices were bonded to the piezoelectric ceramic transducer (PZT). The

positive/negative voltage produces a uniaxial tensile/compressive strain perpendicular

to the stacks (the direction of strain is marked in Fig.1a). In order to ensure the

deterministic switching of the Co2FeAl moment, all the devices were bonded ~2 from

the [110] to [010] direction. The deformation of the Co2FeAl devices under the piezo

voltage was measured using a strain gauge, and 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 Information Fig.

S1). 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 (LMOKM). At different deformation states by the

piezo voltages, the longitudinal and transverse resistances were measured

simultaneously with a fixed electrical current injected through the channel. All

measurements were performed at room temperature.

The 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 a rather large

magnetic field of 500 Oe (which is much larger than the coercive field), then swept

the external magnetic field to zero. The RH measured with a fixed current I = 50 μA

along the channel was recorded with periodic piezo voltage pulses applied to both

devices, as is shown in Fig. 1b. Strikingly, the RH of the [100] orientation device

periodically switched from -0.12 to 0.12 Ω with the voltage pulse changing from 0 to

-30 V. In contrast, the RH of the [010] orientation device was periodic switched

simultaneously from 0.12 to -0.12 Ω with switching the voltage pulse from 0 to -30 V.

It is worth noting, with negative current applied to the [100] or the [010] orientated

device, same RH but opposite planar Hall voltage switching was observed in both

devices during the piezo voltages changing from 0 to -30 V. The Hall resistance

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 piezo voltage, which is analogous to the n-FET and p-FET in CMOS

technology. The advantage of this Planar Hall transistor is that the RH (induced by

planar Hall effect) changes sign during the tuning, whereas the conventional field

effect transistors are switched on and off by accumulating and depleting the electrons

(holes) in the channel through electrical gating, while the resistance of the FET is

always positive. However, in our piezo voltage control planar Hall devices, the RH

sign change originates from the rotation of the magnetization vector respective to the

electrical current under piezo voltages. Thus the two magnetic states of planar Hall

transistor tuned by piezo voltages can not only be used for information storage, but

also used as a building block for new functional logic devices.

The change of the planar Hall resistance (ΔRH) of [010] and [100] orientated

devices as a function of the piezo voltage is shown in Fig. 1c. The ΔRH remains

almost zero when switching the piezo voltages from 0 to -27 V for both devices. But a

sharp jump was observed for both devices with switching the piezo voltages from

zero to a further lower value, and then raised slightly before it reaches to a flat plateau.

Opposite value of ΔRH (0.25 Ω and -0.25 Ω) was observed for devices along [100]

and [010] orientations when the piezo voltage is changed from zero to a value lower

than -28 V. It is well established that the planar Hall resistance arises as a result of the

non-equivalence of components of the resistance tensor which are perpendicular and

parallel to the magnetization direction, leading to the appearance of off-diagonal

resistance components. Thus the planar Hall resistance is strongly dependent on the

relative direction between the electric current and the magnetization vector. The

angular dependence of the RH was measured for devices along both samples at piezo

voltage UP = 0, where a fixed 2000 Oe external magnetic field was rotated in the

plane anticlockwise starting at [110] orientation, which is shown in Fig. 1d. The

external magnetic field is much larger than the in-plane magnetic anisotropy fields, so

thatthe magnetization vector of the Co2FeAl follows the external magnetic field

direction. As shown in Fig. 1d, the angular dependence of the planar Hall resistance

can be fitted well using the single domain model29

,

𝑅𝐻 =1

2(𝑅𝑠ℎ𝑒𝑒𝑡

∥ − 𝑅𝑠ℎ𝑒𝑒𝑡⊥ ) sin [2( ± π/4) + 𝛾], where 𝑅𝑠ℎ𝑒𝑒𝑡

∥ is the sheet resistance

with the current parallel to the magnetization, 𝑅𝑠ℎ𝑒𝑒𝑡⊥ is the sheet resistance with the

current perpendicular to the magnetization, 𝜃 + π/4 represents the angle between

the electrical current and the magnetization vector for [010] orientated device while

𝜃 − π/4 represents the angle between the electrical current and magnetization vector

for [100] orientated device, and γ is the deviation angle between the PZT

strain/compress direction and [110] direction which is about 2. The periodic RH for

both devices has the same magnitude and frequency, with π/2 phase shift. The

maximum magnitude of the Hall resistance occurs when 𝜃 = π/4 + nπ/2, where n is

an integer. The angular dependence of RH gave us the information that switching

magnetization by 90 induces a change in RH of 0.27 Ω for both devices which

coincides with the value shown in figure 1b on switching the piezo voltage from 0 to

-30 V. Thus, the piezo voltage can fully switch the magnetization of the Co2FeAl

devices by 90 in the plane.

To have a more insight into the switching behavior of Co2FeAl planar Hall

transistors, the magnetic properties Co2FeAl devices with various UP were

investigated using LMOKM. The hysteresis loops and the corresponding magnetic

domain structures were recorded with magnetic field applied in the plane. Fig. 2a

shows the Kerr rotation angle during the magnetization reversal without the piezo

voltage with magnetic field applied close to the in-plane major crystalline [110],

[1̅10] and [010] orientations, respectively. [110] orientation is the easy axis with the

full magnetic moment at remanence. Hard-axis-like behavior is seen for the magnetic

field in [010] direction with saturation occurring around 220 Oe (the black loop in Fig.

2a). However, the loop of the magnetic field applied along [1̅10] shows a two-step

switching with a continuous reversible rotation in between the two steps. The different

switching behaviors for the [1̅10] and [110] directions is not expected on the

grounds of cubic crystal symmetry. The observed magnetic hysteresis loops along the

major crystalline orientations are the consequence of the superposition of the uniaxial

and the fourfold anisotropy, where the uniaxial easy axis is along [110] orientation

and the cubic easy is along [110] and [1̅10] orientations30

.

The Kerr rotation angle during the magnetization reversal at ±30 V with

magnetic field applied along in-plane major crystalline [110], [1̅10] and [010]

orientations was also investigated. To demonstrate the evolution of the magnetic

anisotropy under piezo voltages, the magnetic hysteresis loops along [1̅10] of

Co2FeAl under piezo voltages 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 compared to the original state namely UP = 0 V. The continuous reversible

magnetic field range between the two step jumps increases with increasing the applied

the piezo voltages (details are shown in Supplementary information), indicating the

[1̅10] becomes harder with increasing the tensile strain along [110] orientation. With

the piezo voltage at UP = -30 V as shown in Fig. 2b, the magnetic hysteresis loop

along [1̅10] orientation has one step magnetization reversal, suggesting the magnetic

easy axis has been switched by 90 from [110] to [1̅10] in the plane under piezo

voltages of -30 V.

The magnetic anisotropy variation under piezo voltages is due to an extra

uniaxial anisotropy introduced by the strain under the piezo voltages. The magnetic

energy density of the system without deformation can be written as31

:

𝝐(𝜽) = −𝟏

𝟒𝑲𝑪 𝒔𝒊𝒏𝟐(𝟐𝜽) + 𝑲𝑼 𝒔𝒊𝒏𝟐(𝜽) − 𝑯𝑴𝒔 𝒄𝒐𝒔(𝜽 − 𝜶) (1),

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, and the last term is

the Zeeman energy32

. The magnetic anisotropy constants can be obtained by

analyzing the magnetic hysteresis loops along the uniaxial hard orientation, where the

two step jumping appears. The 𝐾𝐶 and 𝐾𝑈 were obtained to be −108𝑀𝑆 and

41𝑀𝑆 , respectively (details in Supplementary Information). Using the obtained

magnetic anisotropy constants, the angular dependence of the magnetic energy for the

state without deformation is plotted by the red line in Fig. 2c. The minimum values of

the 𝜖/𝑀𝑆 are at [110] and [1̅1̅0] orientation when the piezo voltage is 0 V, so the

[110] orientation is an easy axis. An additional stress-induced uniaxial magnetic

anisotropy term 𝐾𝑝 𝑠𝑖𝑛2(𝜃) is added to the magnetic energy density equation (1)

when Up 0, where KP has the same sign of KU at tensile strain (positive piezo

voltages) and opposite sign at compressive strain (negative piezo voltages),

respectively. The angular dependence of the magnetic energy at UP = ±30 V obtained

by analyzing the magnetic hysteresis loop using the modified magnetic energy density

formula are also plotted in Fig. 2c. The obtained KP/MS is 25.7 Oe for UP = 30 V and

-46.4 Oe for UP = -30 V. The lowest energy state is along [110] orientation for piezo

voltage at 0 and 30 V. However, because the piezo voltage at -30 V induced KP is

larger than that of the KU and with the opposite sign, the lowest energy state is along

the [1̅10] orientation at UP = -30 V. The magnetization switching will happen if the

gaining energy of magnetic domains is larger than the energy barrier of the two

neighbor local minimums. The magnetic energy landscape in figure 2c confirms piezo

voltages can switch the magnetic easy axis by 90. The piezo voltage control of the

magnetization switching by 90 in the plane has also been confirmed using

ferromagnetic resonance24

.

The magnetic domain images of the Co2FeAl [100] orientated device without

deformation were taken by LMOKM with magnetic field applied in [1̅10]

orientation, which are shown in Fig. 2d (a-e). At relatively large positive magnetic

field of 80 Oe, the magnetic images are homogenously dark, because the

magnetization vector of the device is fully aligned in [1̅10] orientation. Decreasing

the positive field to 12 Oe, part of the domain image turns from dark to grey as shown

in Fig. 2d (b), indicating only part of the device was switched 90 to the easy axis of

[110] orientation. Further decreasing external magnetic field to zero, the observed

homogenously grey domain images shown in Fig. 2d (c) suggest the magnetization

has been fully switched by 90 to [110] orientation due to the energy minimum is in

[110] orientation, which is confirmed by the very small Kerr signal shown in Fig. 2a.

Then with increasing negative magnetic field, part of the domain image of the device

firstly turns into white from grey at around H = -33 Oe, which is shown in Fig. 2d (d).

Then, the domain image of the whole device turns to white with increasing the

negative magnetic field further to -80 Oe as shown in Fig. 2d (e), where the

magnetization of the device is fully switched by the external magnetic field to [11̅0]

orientation. We then investigated the magnetic domain states controlled by piezo

voltages at zero magnetic fields using the LMOKM configuration, which is shown in

Fig. 2d (f-j). On removing the external magnetic field after firstly magnetizing the

device along [110] orientation with an external magnetic field of 100 Oe, the domain

image is homogenously grey at zero piezo voltage as shown in Fig. 2d (f), indicating

the magnetization vector stays in [110] orientation. When the piezo voltage of +30 V

is applied, the color of the domain image does not change. However, when the piezo

voltage of -30 V is applied as shown in Fig. 2d (h), the domain image of the device

turns into dark, indicating the magnetization was fully switched by 90 to [1̅10]

orientation. After decreasing the piezo voltage back to 0, the domain image of the

device returns back to grey in Fig. 2d (i). As shown in Fig. 2d (h-j), the homogeneous

reversible switch from grey to dark with switching the piezo voltages between 0 and

-30 V confirm that the magnetic states between [110] and [1̅10] orientations are

switchable by piezo voltages without external magnetic field.

The two magnetic states in the planar Hall transistors tuned by piezo voltages

cannot only be used to the information storage, but also can be used as logic 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 less. Based on the single planar Hall transistor shown in Fig. 3a,

the piezo voltages can effectively switch the magnetization between the [110] and

[1̅10] magnetic states, which function as the NOT gate and produce the planar Hall

voltage output as shown in Fig. 1b. When the piezo voltage is 0 V, the output Hall

voltage (Vo) is 6.5μV, which is larger than the 5 μV so that output =1. For piezo

voltage of -30 V, the Vo is changed to -5.7 μV < 2 μV by switching the magnetization

and the output = 0. The truth table in Fig. 3b represents the NOT gate operation Y = A ̅.

The p-type and n-type functionalities of the planar Hall transistors based on [100] and

[010] orientation Co2FeAl devices can be realized. It is also worth noting that piezo

voltages based on only [100] or [010] orientation devices can also fulfill the function

of the p and n type transistor by applying opposite current. The NOR gate was built

based on one p-type ([100] orientation) and one n-type ([010] orientation) planar Hall

transistor, which is shown in Fig. 3c. The two devices are connected as shown in Fig.

a

3c, where the two piezo voltages (UP1 and UP2) control the magnetization of 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 external magnetic fields with the fixed current of 50 μA. Inputting

the [0,0] to the logic with both UP1 and UP2 equal to zero sets the magnetization of

both devices along [110], so that the Vo is 11.8 μV (the sum of the Hall voltages from

these two devices). This is greater than the threshold voltage so that output = 1. On

switching magnetization of both devices to [1̅10] direction by 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 [1̅10] orientation,

corresponding to the [1,0] or [0,1] magnetic states, the Vo are -1.9 and 1.9 μV

respectively. Thus the output is 0 since both Vo are less than 2 μV (as shown in Fig.

3d). The non-zero value of Hall resistance at [1,0] and [0,1] states is because the

magnitude of the Hall resistance between these two devices is different, which is

either from the photolithography or the slightly different misalignment to [100] or

[010] orientation of these two devices. If the two devices have very different planar

Hall voltages, which can be realized by fabricating the electrical current channel

along different in-plane orientations, very clear four states gates controlled by piezo

voltages could be achieved through the above NOR logic design. The results are

summarized in the truth table representing the NOR gate Y = A + B as shown in Fig.

3e. Using the similar methodology to integrating the piezoelectric layers, the

application of a smaller input voltage switching the magnetization will be achieved by

scaling down the devices to nanometer sizes33

. Thus the magnetic NOT and NOR gate

functionalities can be realized by the logic gates designing.

In summary, we have presented that the planar Hall transistor in half metal

Co2FeAl devices can be controlled from positive (negative) to negative (positive) by

piezo voltages without external magnetic field, which is associated with piezo

voltages controlling the magnetization switching by 90 in the plane. The two

magnetic states of the planar Hall transistor controlled by the piezo voltages can not

only be used as the information storage, but also can be used in logic devices. Our

demonstration could pave a way for the application of future spintronics, realizing

both the information storage and processing in only ferromagnetic metals.

METHODS SUMMARY

The Co2FeAl film was grown on a GaAs (001) substrate using molecular beam

epitaxy (MBE) technology at 280℃. 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 20μm and the

distance between the neighbor arms is 80μm. Ti/Au contacts were deposited by

thermal evaporation. The GaAs substrate was polished down to 100 m, and then the

heterostructure was bonded to the piezoelectric ceramic transducer (PZT). The

magnetic hysteresis loops and the magnetic domain images were taken using the

longitudinal magneto-optical Kerr microscopy Nano MOKE3. DC measurements

were used to perform all the magnetotransport measurements, where the Hall voltages

were detected using a Keithley nanovoltage meter 2182.

Figure 1丨 The Planar Hall resistance of Co2FeAl device structure controlled by the

Piezo voltages. a. The schematic diagrams for planar Hall effect controlled by piezo voltages

measurements for both the [010] and [100] orientated devices. b. The periodic changes of the

planar Hall resistances for both the [010] and [100] orientated devices with the periodic

change of the piezo voltage pulses between 0V and -30 V without external magnetic field. c.

The change of the planar Hall resistance dependent on the change of the piezo voltages from

0 to certain values for both the [010] and [100] orientated devices. d. The angular dependence

of the planar Hall resistance for both the [010] and [100] orientated devices with a fixed

magnetic field at 2000 Oe rotated in the plane, where the dots are the experimental results and

the lines are the fitted results.

Figure 2丨 The magnetic states of Co2FeAl device controlled by the piezo voltages. a.

The magnetic hysteresis loops of Co2FeAl device measured by longitudinal magneto-optical

Kerr system with magnetic field applied in the [110], [1̅10] and [010] directions with piezo

voltage at zero. b. The magnetic hysteresis loops measured using longitudinal

magneto-optical Kerr system with magnetic field in [-110] orientation with piezo voltages at

-30, 0 and 30 V. c. The angular dependence of the magnetic energy density at zero magnetic

field for the Co2FeAl device with piezo voltages at -30 (red), 0(blue) and 30 V (green), where

the minimum energy is switched from [110] to [1̅10] orientation when the piezo voltage is

-30 V. d. The magnetic domain images (a-e) of the Co2FeAl device during the magnetization

reversal along [1̅10] orientation without deformation. The magnetic domain images (f-j) of

the Co2FeAl device were controlled by piezo voltages without external magnetic field.

Figure 3丨 Programmable logic operation 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 for the [010] and [100] devices, respectively.

d. The output voltages of the NOR gates with varying the piezo voltages for both logic gates.

e. Truth table summary of the operation described in NOR gate.

References

1. Behin-Aein, B., Datta, D., Salahuddin, S. & Datta, S. Proposal for an all-spin logic

device with built-in memory. Nat. Nanotech. 5, 266–269 (2010).

2. Datta, S. & Das, B. Electronic analog of the electro-optic modulator. Appl. Phys. Lett. 56,

665–667 (1990).

3. Koo, H. C. et al. Control of spin precession in a spin-injection field effect transistor.

Science 325, 1515-1518 (2009).

4. Sahoo, S. et al. Electric field control of spin transport. Nature Phys. 1, 99-102 (2005).

5. Wunderlich, J. et al. Spin Hall effect transistor. Science 330, 1801–1804 (2010).

6. Dery, H. et al. Spin-based logic in semiconductors for reconfigurable large-scale circuits.

Nature 447, 573-576 (2007).

7. Kane, B. E. A silicon-based nuclear spin quantum computer. Nature 393, 119–120

(1998).

8. Hayashi, M., Thomas, L., Moriya, R., Rettner, C. & Parkin, S. S. P. Current-controlled

magnetic domain-wall nanowire shift register. Science 320, 209-211 (2008).

9. Xu, P. et al. An all-metallic logic gate based on current-driven domain wall motion.

Nature Nanotech. 3, 97–100, (2008).

10. Schneider, T. et al. Realization of spin-wave logic gates. Appl. Phys. Lett. 92, 022505

(2008).

11. Liu, L. et al. Spin-torque switching with the giant spin Hall effect of tantalum. Science

336, 555-558 (2012).

12. Bhowmik, D., You, L. & Salahuddin, S. Spin Hall effect clocking of nanomagnetic logic

without a magnetic field. Nature Nanotech. 9, 59-63 (2013).

13. Sinova, J. & Žutić, I. New moves of the spintronics tango. Nat. Mater. 11, 368–371

(2012).

14. Locatelli, N., Cros, V. & Grollier, J. Spin-torque building blocks, Nature Mater. 13, 11

(2014).

15. Lei, N. et al. Strain-controlled magnetic domain wall propagation in hybrid

piezoelectric/ferromagnetic structures. Nature Commun. 4, 1378 (2013).

16. Wang, J., Meng, H. & Wang, J. Programmable spintronics logic device based on a

magnetic tunnel junction element. J. Appl. Phys. 97, 10D509 (2005).

17. Leem, L. & Harris, J. Magnetic coupled spin-torque devices for nonvolatile logic

applications. J. Appl. Phys. 105, 07D102 (2009).

18. Li, Y. et al. Voltage manipulation of the magnetization reversal in Fe/n-GaAs/

piezoelectric heterostructure. J. Magnetism and Magnetic Mater. 375, 148 (2015).

19. De Ranieri, E. et al. Piezoelectric control of the mobility of a domain wall driven by

adiabatic and non-adiabatic torques. Nature Mater. 12, 808–814 (2013).

20. Rushforth, A. W. et al. Voltage control of magnetocrystalline anisotropy in

Ferromagnetic-semiconductor/piezoelectric hybrid structures. Phys. Rev. B 78, 085314

(2008).

21. Lei, N.et al. Magnetization reversal assisted by the inverse piezoelectric effect in

Co-Fe-B/ferroelectric multilayers. Phys. Rev. B 84, 012404 (2011).

22. Parkes, D. P. et al. Non-volatile voltage control of magnetization and magnetic domain

walls in magnetostrictive epitaxial thin films. Appl. Phys. Lett. 101, 072402 (2012).

23. Li, P. et al. Electric Field Manipulation of Magnetization Rotation and Tunneling

Magnetoresistance of Magnetic Tunnel Junctions at Room Temperature. Adv. Mater. 26,

4320–4325 (2014).

24. Gueye, M. et al. Effective 90-degree magnetization rotation in Co2FeAl thin

film/piezoelectric system probed by microstripline ferromagnetic resonance. Appl. Phys.

Lett. 107, 032908 (2015).

25. Newns, D., Elmegreen, B., Liu, X. H. & Martyna, G. A low-voltage high-speed

electronic switch based on piezoelectric transduction. J. Appl. Phys. 111, 084509 (2012).

26. Galanakis, I. & Mavropoulos, P. Spin-polarization and electronic properties of

half-metallic Heusler alloys calculated from first principles, J. Phys.:Cond. Matter, 19,

315213 (2007).

27. Tang, H. X., Kawakami, R. K., Awschalom, D. D. & Roukes, M. L. Giant planar Hall

effect in epitaxial (Ga,Mn)As devices. Phys. Rev. Lett. 90,107201 (2003).

28. Meng, K. et al. Magnetic properties of full-Heusler alloy Co2Fe1−xMnxAl films grown by

molecular-beam epitaxy. App. Phys. Lett. 97, 232506 (2010).

29. Wang, K. et al. Anisotropic magnetoresistance and magnetic anisotropy in high-quality

(Ga,Mn)As films. Phys. Rev. B 72, 085201 (2005).

30. Dumm, M.et al. Magnetism of ultrathin FeCo (001) films on GaAs(001). J. Appl. Phys.

87, 5457-5459 (2000).

31. Cowburn,R. et al. Magnetic switching and inplane uniaxial anisotropy in ultrathin

Ag/Fe/Ag(100) epitaxial films. J.Appl. Phys. 78, 7210 (1995).

32. Wang, K. et al. Spin reorientation transition in single-domain (Ga, Mn)As. Phys. Rev.

Lett. 95, 217204 (2005).

33. Solomon, P. et al. Pathway to the Piezoelectronic Transduction Logic Device. Nano Lett.

15, 2391-2395 (2015).

Acknowledgements

This work was supported by “973 Program” No. 2014CB643903, and NSFC Grant

Nos. 61225021,11174272 and 11474272.

Author Contributions

K-Y. W. designed the whole experiments. B. Z., M-Y. Y., and Y-Y. L. fabricated the

devices and performed the measurements. M-Y. Y., K-M. C., H. Z., K-Y. W. and B. Z.

analyzed the data. K-K. M. and J-H. Z. provided the experimental materials. K. W., B.

Z., M-Y. Y., and K. W. E. wrote the paper. All authors discussed the results and

commented on the manuscript.

Additional information

See the supplementary information.

Competing financial interests

The authors declare no competing financial interests.