Electric field tuning of phase separation in manganite thin films

In this paper, we investigate the electric field effect on epitaxial Pr${}_{0.65}$(Ca${}_{0.75}$Sr${}_{0.25}$)${}_{0.35}$MnO${}_3$ thin films in electric double-layer transistors. Different from the conventional transistors with semiconducting channels, the sub(micrometer)-scale phase separation in the manganite channels is expected to result in inhomogeneous distribution of mobile carriers and local enhancement of electric field. The field effect is much larger in the low-temperature phase separation region compared to that in the high-temperature polaron transport region. Further enhancement of electroresistance is achieved by applying a magnetic field, and a 250% modulation of resistance is observed at 80 K, equivalent to an increase of the ferromagnetic metallic phase fraction by 0.51%, as estimated by the general effective medium model. Our results illustrate the complementary nature of electric and magnetic field effects in phase-separated manganites, providing insights on such novel electronic devices based on complex oxides.

Figure 1

(a) X-ray diffraction θ-2θ scan of the PCSMO film grown on LSAT substrate. (b) Schematic side view of the EDLT. In this particular gate polarity, the PCSMO is in the accumulation state. Also shown are the cationic and ionic molecular components of the electrolyte DEME-TFSI. (c) Schematic of the device, along with the measurement setup. Typical AFM image of the PCSMO film taken (d) right after the PLD growth and (e) after the gating experiments and rinsing off the electrolyte. The scan size of both images is 1 × 1 μm${}^2$. The RMS roughness measured for (d) and (e) is 0.91 and 1.04 nm, respectively.

Figure 2

(a) Temperature-dependent resistance of the PCSMO channel (device PCS-A) measured during both warming and cooling (as marked by the arrows) under a series of applied gate voltages (−3, −2, 0, +2, and +3 V). (b) Corresponding ER vs temperature curves, where ER or Δ$R$/$R$(0) is defined as $\left\{\right[R$($V_g$) − $R$(0)]/$R\left(0\right)\}$. The open symbols represent the warming processes, while the filled ones represent the cooling processes; this is also indicated by the arrow directions for +3 V gate bias.

Figure 3

(a) Small polaron fitting to the high-temperature transport data for both the warming and the cooling processes. For clarity, only the zero gate data are shown. (b) Plot of the activation energy $E_{\mathrm{hop}}$ as calculated from the fitting to the small polaron model versus the gate voltage.

Figure 4

A two-dimensional toy model illustrating the local electric field distribution inside an inhomogeneous media when a voltage is applied across the electrodes. Four metallic spheres with a radius of 5 nm are embedded in an insulating host, and the total thickness of the media between the electrodes is 25 nm, which is close to the actual thickness of the PCSMO film used in our experiments.

Figure 5

(a) Temperature dependence of the normalized magnetization $\frac{M}{M_S}$, where $M$ was the magnetization of the PCSMO film measured under a magnetic field of 0.1 kOe in a field-cooled process and $M_S$ is the saturation magnetization determined in the 10 K hysteresis loop shown in the inset. (b) Comparison of the measured conductivity of the PCSMO film (green solid circles) with the calculated conductivity (green open squares) based on the GEM theory with the percolation threshold $f_c$ = 0.24. Also shown are the conductivity data of PCMO (blue circles) and PSMO (red circles) taken from the paper by Blake et al. [64]. (c) Variation of the FM fraction at 80 K calculated from the GEM theory as a function of the applied gate voltage.

Figure 6

(a) Magnetic field-dependent MR measured on the device PCS-B under the gate bias of −3 V (open circles) and +3 V (filled circles). For clarity, only the data taken on increasing magnetic fields are shown. The small abrupt steps in the data are measurement artifacts. (b) Variation of the ER with the applied magnetic field. Here, ER is defined as Δ$R$/$R$(−3 V) =  $\left\{\right[R$(+3 V) − $R$(−3 V)]/$R(-3\mathrm{V})\}$. The filled (open) circles represent the ER data measured on the increasing (decreasing) magnetic field. (c) Plot of the temperature-dependent values of the magnetic field at the peak ER as deduced from Fig. 6. (d) Plot of ER vs temperature for different magnetic fields.

Figure 7

Fraction of the FM phase vs the magnetic field at 80 K for gate voltages of +3 and −3 V as estimated from the GEM theory.

Figure 8

Plot of equivalent magnetic field $H_{\mathrm{equi}}$ as a function of temperature. $H_{\mathrm{equi}}$ represents the value of a magnetic field, which needs to be introduced on the +3 V gate biased channel in order to produce the same resistance modulation as reversing the gate bias to −3 V.