† Corresponding author. E-mail:
Project supported by the National Key Research and Development Program of China (Grant No. 2016YFA0300702), Shanghai Municipal Natural Science Foundation, China (Grant Nos. 19ZR1402800, 18JC1411400, 18ZR1403200, and 17ZR1442600), the Program of Shanghai Academic Research Leader, China (Grant Nos. 18XD1400600 and 17XD1400400), and the China Postdoctoral Science Foundation (Grant Nos. 2016M601488 and 2017T100265).
We reported a study of tunnel magnetoresistance (TMR) effect in single manganite nanowire via the combination of magnetotransport and magnetic force microscopy imaging. TMR value up to 290% has been observed in single (La1 − yPry)1 − xCaxMnO3 nanowires with varying width. We find that the TMR effect can be explained in the scenario of opening and blockade of conducting channels from inherent magnetic domain evolutions. Our findings provide a new route to fabricate TMR junctions and point towards future improvements in complex oxide-based TMR spintronics.
Over the past decades, tunnel magnetoresistance (TMR) effect has attracted lots of research interests due to its conceptual importance in spintronics and potential of technological applications.[1–4] One of the most well-known geometries to fabricate TMR architectures is the magnetic tunnel junction (MTJ).[5–7] The basic structure of a MTJ usually consists of a thin insulating layer sandwiched between two ferromagnetic electrode layers. Depending on the relative spin orientation of the two ferromagnetic electrodes being parallel or antiparallel, the low/high resistance state (RP/RAP) can be achieved. Such structure has proven to be successful, but carries complications in materials synthesis because controlled growth of multiple materials is required. This unavoidably mandates extra care in the growth process, since different materials can have different optimal thermodynamic growth parameters and their interfaces can have dramatic impacts on the TMR properties.[8,9]
An alternative approach to achieve TMR effect is to utilize the intrinsic coexistence of different electronic phases in complex oxide systems. Perovskite manganites such as (La1 − yPry)1−xCaxMnO3 (LPCMO) provide a good platform. The interaction of electron, spin, orbital, and lattice produces the rich phases in the system, such as the emergence of antiferromagnetic charge ordered insulating (COI) phase and ferromagnetic metallic (FMM) phase. These phases can coexist at wide ranges of temperatures and magnetic fields, effectively having the opportunity to form intrinsic FMM/COI/FMM junction with TMR effect. Such intrinsically formed junctions have the merit of eliminating chemical interfaces, thus the fabrication of multiple layers is not required. Although TMR effects have been observed in LPCMO nanowires and anti-dots systems and several mechanisms and models have been proposed,[10–14] a clear TMR-based FMM/COI/FMM feature has not been observed, clouding the underlying mechanism on the TMR effect.
In this work, we study the TMR effect in LPCMO nanowires with a particular emphasis on the underlying mechanism. The magnetic force microscope (MFM) measurements directly visualized the magnetic domain evolutions under magnetic fields. Combining with the magnetotransport measurements, we conclude that the TMR effects can be understood in the scenario of opening and blockage of filamentary conducting channels in FMM/COI/FMM domain configurations. Such mechanism is completely different from that in the traditional TMR junctions and can be only realized under spatially confined nanowires.
A 30 nm (La2/3Pr1/3)5/8Ca3/8MnO3 (LPCMO) thin film sample is grown by pulse laser deposition (PLD) on a LaAlO3 (001) substrate. The substrate is heated to 800 °C in the atmosphere of 8 × 10−1 Pa oxygen containing 8% ozone. Reflection high-energy electron diffraction (RHEED) is used to ensure the high quality of layer-by-layer epitaxial growth via its oscillations of intensity during sample growth. After growth, the sample is post annealed at 850 °C for 3 h in 1 atmosphere oxygen to minimize the influence of oxygen vacancies. Nanofabrication techniques are employed to fabricate LPCMO nanowires. The LPCMO nanowires are fabricated by electron beam lithography. 200 nm AR-N 7520, a negative E-beam resist, is covered on the thin film by spin-coating. Electron beam lithography (Zeiss SIGMA SEM and Raith Elphy Plus pattern processor) is employed to pattern the wires on the negative resists. Developing liquid (RZX 3038:H2O = 4: 1) is used to remove the part without electron beam writing after patterning. Argon ion beam etching is applied to corrode the part uncovered by resist and retain the nanowires pattern. Reactive-ion etching with oxygen plasma is applied to remove the residual resist. At last, the LPCMO naonwire is patterned with gold electrodes by optical lithography.
The length of the wires is 20 μm. These straight wires have smooth edges, indicating high quality of the sample after nanofabrication. The electronic transport is measured by a physical property measurement system (PPMS) with 2-probe connection. The current density through the nanowires is controlled around 3 × 103 A/cm2, using the source Agilent 2900 series, in order to reduce the effect of current-induced heating.
Figure
Figure
When the magnetic field reaches −200 Oe, a sudden jump of resistance to a higher value occurs. The resistance remains stable at this high value until the magnetic field reaches −900 Oe, at which the resistance of the nanowire drops abruptly to a lower value and then decreases continuously. When sweeping the magnetic field from −3000 Oe to 3000 Oe, a similar behavior is observed with the corresponding upward and downward jumps of the resistance occurring at 400 Oe and 800 Oe, respectively. The MR ratio is calculated as MR = (R - R0.3T)/R0.3T, and it reaches 290%. Such TMR effects are observed in the temperature range between 96 K and 111 K as highlighted in Fig.
The domain evolution during the TMR process is directly visualized in a 500 nm nanowire using MFM with different applied magnetic fields along surface normal, as shown in Fig.
We now turn to discuss the correlation between the MFM and transport measurements during the TMR process. We first identify the nature of the observed blue domains which nucleate with decreasing magnetic field. Based on the imaging principle of MFM, these blue domains may correspond to either COI domains or the FMM domains with the magnetization direction antiparallel to the MFM tip. For manganites, it is well established that the magnetic field can melt the COI state into FMM state. The fact that the blue domains continue to grow with increasing field from 0 to −300 Oe (Figs.
Figure
The above results illustrate that the TMR effect observed in a single manganite wire is of distinctively different nature from that of the conventional TMR junctions which consist of three different layers of materials. From application point of view, two issues need to be resolved for the single nanowire TMR in manganites. First, since the domain distributions are random and different at each thermal cycle,[19,20] the TMR effects are not always repeatable under different temperature cycles, as observed in several publications[10,11,14] and our work here. Therefore, it is important to develop strategies to obtain TMR structures with well-defined domain positions and sizes in order to gain better control of the TMR effect. Local electric-field gating[21] or magnetic patterning[22] can be possible solutions. Second, from application point of view, it is essential to elevate the operation temperature by developing new materials or new heterostructures.[23]
In summary, via the combination of nanofabrication, magnetotransport, and MFM imaging, we achieve large TMR effect in single manganite nanowires and propose a plausible mechanism. Our research provides a new route to design TMR devices based on electronic phase separated systems without chemical interfaces.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] |