Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

# Achieving large and nonvolatile tunable magnetoresistance in organic spin valves using electronic phase separated manganites

## Abstract

Tailoring molecular spinterface between novel magnetic materials and organic semiconductors offers promise to achieve high spin injection efficiency. Yet it has been challenging to achieve simultaneously a high and nonvolatile control of magnetoresistance effect in organic spintronic devices. To date, the largest magnetoresistance (~300% at T = 10 K) has been reached in tris-(8-hydroxyquinoline) aluminum (Alq3)-based organic spin valves (OSVs) using La0.67Sr0.33MnO3 as a magnetic electrode. Here we demonstrate that one type of perovskite manganites, i.e., a (La2/3Pr1/3)5/8Ca3/8MnO3 thin film with pronounced electronic phase separation (EPS), can be used in Alq3-based OSVs to achieve a large magnetoresistance (MR) up to 440% at T = 10 K and a typical electrical Hanle effect as the Hallmark of the spin injection. The contactless magnetic field-controlled EPS enables us to achieve a nonvolatile tunable MR response persisting up to 120 K. Our study suggests a new route to design high performance multifunctional OSV devices using electronic phase separated manganites.

## Introduction

Efficient spin injection from a ferromagnetic (FM) material into a nonmagnetic semiconductor is highly desirable for spin-based computing concepts such as spin transistors and spin storage/memory devices1,2,3,4. It is challenging due to conductivity mismatch between the FM contact and the semiconductor5,6,7 that greatly suppresses the spin injection efficiency. Inserting a high-quality tunneling barrier/spin filterer at the FM/semiconductor interface8 will improve the spin injection efficiency whereas a sophisticated epitaxial layer growth is required making it difficult to be compatible with other heterostructure systems9,10,11. The discovery of MR effect through an organic semiconductor (OSEC) medium between two FM contacts, namely organic spin valves (OSVs)12,13,14, suggests a new route to enhance the spin injection efficiency into semiconductors: The formation of OSEC/FM interface (so-called: spinterface15,16,17,18,19,20,21) offers a promising spin-filtering effect to enhance the spin polarization of FM metals leading to a large MR effect (up to 300%) at low temperature17,22,23. While most recent OSV studies focus on optimizing the MR performance by tailoring the OSEC/FM spinterface and exploring the multi-functionalities of OSVs by utilizing photovoltaic response in OSECs9,19,24,25,26,27,28, the influence of the FM contacts on the MR effect has not been emphasized.

In general, FM materials with high spin polarization are favorable for the FM contacts in the OSVs. In this regard, the half metallic perovskite manganites with near 100% spin polarization would be the most popular candidates. In spite of a large number of members in manganites family, La0.67Sr0.33MnO3 (LSMO) has been almost the only material used in OSVs10,14,22,29,30. This is probably because most other manganite films have pronounced electronic phase separation (EPS) featured by coexistence of ferromagnetic metallic (FMM) and non-ferromagnetic insulating phases, which would presumably lower the spin polarization and are thus considered as non-favorable materials for spin injection.

Here we demonstrate that by substituting the LSMO with a (La2/3Pr1/3)5/8Ca3/8MnO3 (LPCMO) film, a surprisingly large MR effect up to 440% is achieved in a conventional OSV geometry. Benefited from such a large magnetoresistance, a signature of electrical Hanle effect is successfully observed in the LPCMO-based OSV device, unambitiously showing the electrical spin injection and detection in the Alq3 layer. The LPCMO system has been well known for its pronounced EPS between the ferromagnetic metallic (FMM) and antiferromagnetic charge ordered insulating (COI) phases31. We show that magnetic field induced modulation of the FMM and COI phase in the LPCMO thin film leads to a tunable MR effect up to 120 K above which the phase separation disappears. Our results open a new pathway of designing high performance organic spintronic devices by utilizing the electronic phase separation of the complex manganites, thus bridging the gap between organic spintronics and complex manganites materials.

## Results

### Device structure and characterizations of LPCMO thin film

We demonstrate EPS-modulated large MR response in a vertical-junction molecular spin valve device consisting of a 60-nm-thick Alq3 (tris-(8-hydroxyquinoline) aluminum, Aldrich) molecule thin film sandwiched between a 60-nm-thick LPCMO film epitaxially grown on the SrTiO3 substrate and a 10-nm-thick cobalt layer with Au capping. A schematic diagram of the fabricated OSV structure is shown in Fig. 1a, b (see Methods). The Alq3 films were used because it is the well-tested molecular material for constituting the organic/molecular spintronic devices. It provides a large spin filtering effect via the Alq3/ferromagnet spinterface as evidenced by two-photon photoemission spectroscopies19. The LPCMO thin film was selected since the high-quality LPCMO film grown by pulsed laser deposition exhibits a large-scale electronic phase separation that can be controlled by applying a pre-set field (Bpre) or varying the temperature31,32. The LPCMO thin film is also suitable for an efficient spin injection owing to its half-metallic nature that would significantly suppress the conductivity mismatch between organic molecules and oxide surface. In support of this assertion, the spin polarization of LPCMO is calculated from the tunneling magnetoresistance (TMR) response in a LPCMO/Al2O3/Co devices using the Jullière model (i.e., $${\mathrm\it {TMR}} = \frac{{2P_{{\mathrm{LPCMO}}}P_{{\mathrm{Co}}}}}{{1 - P_{{\mathrm{LPCMO}}}P_{{\mathrm{Co}}}}}$$)33. The obtained largest value of TMR is 93% at 10 K. Considering PCo = 34% from the literature report34, we estimate that the PLPCMO is ~93% at 10 K (see Supplementary Fig. 1 and Supplementary Note 1). Note that the application of the Jullière model here is questionable due to the large down-spin channel resistance of the LPCMO electrode, which is neglected by the Jullière model and tends to lower the estimated spin polarization. The actual value of PLPCMO would be even higher.

The phase competition between FMM and COI domains in the LPCMO thin film35,36,37 is characterized by both temperature-dependent resistivity (R vs. T) and magnetization (M vs. T) measurements, as shown in Fig. 1c (please also see Supplementary Figs. 2 and 3). These two plots present a typical FMM-to-COI interconversion in the LPCMO thin film as a function of temperature38,39: When the temperature decreases, the paramagnetic (PM) phase is replaced by the COI phase at TCO ~ 200 K, as revealed by the sharp increase of the thin film resistance. This transition is consistent with the onset of ferromagnetism as presented in the M vs. T and the derivative peak of $$|dM/dT|$$ vs. T curve (inset of Fig. 1c) at the same temperature. The short-range FMM phase domains grow with decreasing temperature and percolate to form conducting channels at the insulator-to-metal transition temperature (~100 K). The long-range ferromagnetic order forms at Curie temperature (Tc ~ 59 K) when most of COI phase domains are melt into the FMM phase.

Figure. 1d shows the typical nonlinear current-voltage (I–V) curves of the LPCMO-based OSV device at 10 K, which exhibit similar characteristics to those of a conventional LSMO/Alq3/Co OSV14. All I–V curves were measured at 1 T after the following magnetic field history: (1) a pre-set magnetic field, Bpre, was applied along the in-plane direction; (2) The field was reduced to zero and then increased to 1 T. This procedure was repeated for higher Bpre without any demagnetization process in between. The conductance (resistance) increases (decreases) with increasing Bpre, suggesting that the percolation of FMM phase in the LPCMO bottom electrode reduces the overall device resistance. For comparison, the LPCMO bottom electrode is replaced by La0.67Sr0.33MnO3 to form a LSMO-OSV. I–V curves of the LSMO-OSV do not exhibit any measurable dependence on Bpre (see Supplementary Fig. 4).

### Large MR effect in the LPCMO-OSV device

MR responses of the LPCMO-OSV as a function of pre-set magnetic field (Bpre) are presented in Fig. 2a. The MR value was measured after the following procedure: (1) A chosen Bpre was applied along the in-plane direction and then reduced to zero; (2) A sweeping in-plane field was applied from 1 T to −1 T and back to 1 T. Upon sweeping field from 1 T to −1 T, the relative magnetization orientation of the two FM electrodes changes from parallel (P) to antiparallel (AP) and to parallel configurations, leading to a change of the device resistance (R) from RP to RAP and back to RP. A typical positive MR peak appears (RP < RAP) corresponding well with the coercive fields of Co and LPCMO electrodes determined by SQUID measurements (see Supplementary Fig. 5). When the top Co electrode is replaced by a Au layer, there is no MR signal (see Supplementary Fig. 6). This implies that the observed MR response indeed originates from the switching of magnetization of the Co and LPCMO electrode, excluding the possibilities of the tunneling anisotropic magnetoresistance35,40,41 and organic magnetoresistance42,43. The abrupt change of device resistance between the parallel and antiparallel configuration indicates a high-quality device fabrication and well-defined interfaces between Alq3 molecules and ferromagnetic electrodes. We note that the sign of MR response is opposite to that in a LSMO OSV14,17,22,29,30 (see Supplementary Fig. 4). This indicates that the LPCMO electrode injects a spin current into the Alq3 layer with an opposite sign comparing to that from LSMO electrode. The sign of spin current can vary depending on the relative energy level alignment and interface between Alq3 and FM layer as manifested in previous reports9,11,14,17,22.

By applying Bpre = 1 T, a large MR signal close to +130% is obtained in the LPCMO-OSV device. Surprisingly, the MR amplitude increases with increasing Bpre. We confirmed that this cannot be understood by simply counting the Bpre induced increase of FMM volume ratio, which would only reduce the device overall resistance but not MR of the device: Under a large preset magnetic field (e.g., Bpre = 7 T), both RAP and RP decrease but the decrease of RP is much larger than RAP, resulting in a substantial enhancement of MR amplitude (up to ~440%) at high Bpre. Such a large MR amplitude has not been observed in similar OSV devices using LSMO electrode10,14,17,22,29,30, suggesting a higher spin polarization injection from LPCMO electrode that is superior to the well-known LSMO electrode. The change of MR amplitude shows a non-linear increase as a function of Bpre and saturates at high preset magnetic fields (see Fig. 2b).

Three control experiments were performed to cross-check our results as summarized below (see Supplementary Fig. 4, 6, 7 and Fig. 2b): (1) A LSMO/Alq3/Co device exhibits a MR value of −37% at 10 K using the same growth condition and measurement procedures. But Bpre dependence of MR response was not observed; (2) A LPCMO/Alq3/Au device has no hysteretic MR response except for a weak anisotropic MR response from the LPCMO electrode; (3) A LPCMO/Al2O3/Co device exhibits a MR value of +30%. The device shows a negative Bpre-dependent MR response, namely the MR ratio decreases as Bpre increases which is opposite to that in the LPCMO/Alq3/Co device. Hence we conclude that the formation of spinterface between Alq3 and LPCMO is critical to achieve a large and tunable MR response by Bpre.

Bias dependence of MR response in the LPCMO-OSV device under different Bpre is plotted in Fig. 2c. The MR value increases with decreasing bias voltage, reaching a maximum at the lowest applied V = 0.1 V. A similar trend of bias dependence of the MR signal has been routinely observed in organic-based OSVs11,14,22,29 and in inorganic semiconductor/FM structures44,45, which might be attributed to the electric field dependence of the spin relaxation time/spin diffusion length in the molecule film46,47 or magnon excitation inside the FM electrodes11,25,29,48. Because the normalized $$|MR|$$ response shows no difference as a function of bias under different Bpre (inset of Fig. 2c), we can conclude that Bpre does not affect the electric field dependent spin relaxation time/spin diffusion length in Alq3 molecule film.

### Signature of electrical Hanle effect in the LPCMO-OSV device

To verify the spin-aligned carrier injection into the Alq3 layer, we conducted the electrical Hanle effect measurements49 in the LPCMO/Alq3/Co device at T = 10 K. An out-of-plane magnetic field, Bz was applied to the device (Fig. 3a) when the magnetization of the LPCMO and Co electrode are maintained in the antiparallel (Fig. 3b) or parallel configuration (Fig. 3c), respectively (see Methods). A typical Hanle effect is shown in the LPCMO-OSV device in both antiparallel and parallel configurations, which confirms the success of electrical spin injection and detection in the Alq3 interlayer. While using the same LPCMO electrode, the magnetic tunnel junction (i.e., LPCMO/Al2O3/Co) shows no Hanle effect (Fig. 3d, e), excluding the influence of the titled magnetization of FM electrodes on the device resistance. The applied perpendicular field (< ± 0.2 T) during the Hanle measurement does not title the magnetization of the LPCMO thin film (see Supplementary Fig. 8). The maintained magnetization states of two FM electrodes are also confirmed by repeating the in-plane MR loop after the Hanle measurement.

Here the spin lifetime, τs may be extracted by fitting measured Hanle effect with the equation50,51:

$$\Delta R_{{\mathrm{AP}}({\mathrm{P}})}\left( {B_{\mathrm{z}}} \right) = \Delta R_{{\mathrm{AP}}({\mathrm{P}})}(B_{\mathrm{z}} = 0)/(1 + (\omega _{\mathrm{L}}\tau _{\mathrm{s}})^2)$$
(1)

where ωL is the Larmor frequency defined by $$\omega _{\mathrm{L}} = g\mu _{\mathrm{B}}B_{\mathrm{z}}/\hbar$$, the g-factor of holes in the Alq3 interlayer14 is equal to 2, μB is the Bohr magneton, is the reduced Planck’s constant, and $$R_{{\mathrm{AP}}({\mathrm{P}})}(B_{\mathrm{z}} = 0)$$ is the device resistance in the antiparallel and parallel configuration at Bz = 0T, respectively. By taking account of both magnetic configurations, we found the spin lifetime, τs ~ 45 ps for the injected spin ½ holes in the Alq3 OSV.

In contrast to the diffusion-dominated non-local Hanle measurement52, we found the change of device resistance in the antiparallel configuration is subtle (changing from ~500 MΩ to ~490 MΩ). It doesn’t decrease to the same level of the resistance as that in the parallel configuration (~90 MΩ), or vice versa. This indicates that the measured Hanle effect would be attributed to the partial spin precession under Bz when the spin carrier promptly transmits through the Alq3 interlayer53 due to the exchange-mediated spin transport54,55. Thus the obtained spin lifetime from the Hanle effect should be considered as the transit time, τtransit or the lower bound of spin lifetime in the Alq3 interlayer.

It is the first time that the electrical Hanle effect is demonstrated in the Alq3-based OSVs. Due to the large in-plane magnetic anisotropy of the LPCMO electrode and the efficient spinterface, the large MR response up to 440% allows us to probe such a subtle Hanle effect in the LPCMO-OSV device, namely less than 4% of change in the device resistance at Bz = 200 mT that is two orders of magnitude smaller than its MR value. As a comparison, it is expected that the similar Hanle effect in the LSMO-based OSV device may be much smaller and beyond the instrument sensitivity (<0.4% if the MR value is ~40%14), probably accounting for the absence of Hanle effect in most Alq3-based OSVs. In addition, the short transit time in the Alq3 layer (τtransit ~ 45 ps) would result in a very broad Hanle effect in the need of the perpendicular magnetic field at least up to 100–200 mT which is inaccessible in most OSVs using the soft LSMO electrode.

### Electronic phase separation correlated MR effect

The observed results suggest that the electronic phase separation (EPS) inside the LPCMO film plays a key role in the unusually large and tunable MR effect. To unravel the correlation between EPS and the observed MR response, a variable-temperature, high field magnetic force microscope (MFM) was used to visualize the FMM-to-COI phase interconversion as a function of Bpre and temperature (see Methods).

Figure 4a, c show the MFM images of the LPCMO bottom electrode and Fig. 4d, f show the MR response of the same LPCMO-based OSV device measured at 10 K after application of Bpre = 1 T, 3 T, and 7 T, respectively. MFM images were recorded following the same procedure used in the MR measurements. All the MFM images were taken at a constant magnetic field (Bext = 1 T) after withdrawing Bpre. It is clear that increasing Bpre reduces the area fraction of the COI phase (blue area) and eventually all the COI phase domains transit into the FMM phase at a high Bpre. This is consistent with the increase of magnetization as Bpre increases (see Supplementary Fig. 9). The area fraction of the FMM phase is calculated to be 50.2%, 86.7%, and 100% at Bpre = 1 T, 3 T, and 7 T, respectively. In addition, the comparison of MFM images, the data analysis and magnetic hysteresis loops of the LPCMO film, the LPCMO/Alq3(1 nm) and LPCMO/Al2O3(1 nm) indicates that the phase boundary and FMM volume fraction remain virtually unchanged with the Alq3 or Al2O3 capping layer (see Supplementary Figs. 10, 11, 12). The MR response increases accordingly with increasing FMM area fraction, as shown in Fig. 4d, f.

Figure 4g to i show the MFM image of the LPCMO bottom electrode and Fig. 4j, l show the MR response of the same LPCMO-based OSV device measured at 75 K after application of Bpre = 1 T, 3 T, and 7 T, respectively. The MR response at 75 K again increases with Bpre induced increasing area fraction of the FMM phase. The maximum MR value obtained at 75 K (Bpre = 7T) is about two times smaller than that at 10 K, which is not surprising for OSV devices because the spin diffusion length in Alq3 thin film decreases with increasing temperature11,15. The high field MR response (near linear slope) becomes more pronounced at 75 K than at 10 K due to the decreased overall hysteric MR response in the device. The relations between the obtained MR values and the area fraction of the FMM phase at different temperature are summarized in Fig. 5a. The MR value increases nonlinearly as a function of the area fraction of the FMM phase.

This behavior cannot be attributed to the simple enlargement of the device cross-section due to the expansion of FMM phase, which would only cause the MR response to be unchanged or reduced as observed in the LPCMO/Al2O3/Co device (i.e., equivalent to the case by applying a larger current flow at a higher bias voltage, see Supplementary Fig. 5). Instead, we plot MR as a function of L/A, where L is the total length of the phase boundary and A is the area fraction of the FMM phase (inset of Fig. 5a). We find that the result can be approximately described by the relationship,

$$\frac{{{\mathrm\it{MR}} + 1}}{{{\mathrm\it{MR}}_{{\mathrm{max}}} + 1}} \approx \frac{1}{{1 + \alpha \frac{L}{A}}}$$
(2)

The fitting parameter, α = 35.6 ± 3.5 μm is obtained throughout the entire temperature range, indicating the key role of boundary formation between FMM and COI phase on the observed tunable MR response (see Fig. 5b).

### Two-current model with spin-flip scattering

The two-current model is the basis of all theoretical models of giant and tunneling magnetoresistance56. The celebrated Jullière model is in fact a special limit of the two-current model in which the electrode resistance is set to zero and the tunnel barrier is assumed to be spin-independent (see Supplementary Note 2 and Supplementary Fig. 13). In order to explicitly include the effect of the electrode resistance, whose change under the magnetic field is essential for understanding the nonlinearly enhanced MR effect under Bpre in the LPCMO-OSV device, and the strong spin-dependence of the spinterface resistance, we cannot use the simple Jullière model (see Fig. 6a, b). Here we start from the original two-current model that can capture all the key features of the experimental results including the LPCMO electrode resistance and interfacial spin filtering effect at the metal/molecule interface: The spin up (↑) and spin down (↓) carrier undergoes two different paths through LPCMO electrode, LPCMO/Alq3 spinterface21, Alq3 bulk layer, Alq3/Co spinterface19, and Co electrode from which a spin-dependent resistance is generated (Ri↑(↓), i denotes the different layer), respectively (see Fig. 6c, d). The interfacial spin filtering effect17 induced spin-dependent resistance at LPCMO/Alq3 and Alq3/Co spinterface are represented by $$R_{{\mathrm{spinterL}}({\mathrm{R}}) \downarrow }$$, with the corresponding up-spin channel spinterface resistances set to zero. In our model, we propose that the formation of EPS in the LPCMO film produces a significantly large spin-orbit coupling at the FMM/COI boundary because of the inversion symmetry breaking57. It leads to an additional spin-flip scattering between the spin-up and spin-down channel that can be equivalently represented as Rsf in the circuit (which effectively provides a short circuit due to the large difference in resistance between the two spin channels). The spin-flip scattering rate is approximately proportional to the total area of the phase boundary. Thus Rsf is an effective resistance due to spin-flip scattering and should be inversely proportional to the area of the FMM/COI phase boundary. The modified two-current model yields the following (see Supplementary Note 3):

For the device resistance in the parallel configuration (Rp):

$$R_{\mathrm{P}} \approx R_{{\mathrm{LPCMO}} \uparrow } + R_{{\mathrm{Alq}}3} + R_{{\mathrm{Co}} \uparrow } \approx R_{{\mathrm{Alq}}3};$$
(3)

For the device resistance in the antiparallel configuration (RAP):

$${R_{{\mathrm{AP}}} \approx} \\ {\frac{{R_{\mathrm{LPCMO} \downarrow} }\left( {R_{{\mathrm{Alq}}3} + R_{{\mathrm{spinterL}} \downarrow }} \right)\left( {R_{{\mathrm{Alq}}3} + R_{{\mathrm{spinterR}} \downarrow }} \right) + R_{{\mathrm{sf}}}\left( {R_{{\mathrm{LPCMO}} \downarrow } + R_{{\mathrm{Alq}}3} + R_{{\mathrm{spinterL}} \downarrow }} \right)\left( {R_{{\mathrm{Co}} \downarrow } + R_{{\mathrm{Alq}}3} + R_{{\mathrm{spinterR}} \downarrow }} \right)}{{R_{{\mathrm{LPCMO}} \downarrow }\left( {2R_{{\mathrm{Alq}}3} + R_{{\mathrm{spinterL}} \downarrow } + R_{{\mathrm{spinterR}} \downarrow }} \right) + R_{{\mathrm{sf}}}\left( {R_{{\mathrm{LPCMO}} \downarrow } + 2R_{{\mathrm{Alq}}3} + R_{{\mathrm{spinterL}} \downarrow } + R_{{\mathrm{spinterR}} \downarrow }} \right)}}}$$
(4)

where $$R_{{\mathrm{Alq}}3}$$ is the spin independent junction resistance of the Alq3 bulk, $$R_{{\mathrm{LPCMO}} \uparrow }$$ and $$R_{{\mathrm{LPCMO}} \downarrow }$$ is the resistance in the majority (spin-up) and minority (spin-down) channel of LPCMO film, respectively. $$R_{{\mathrm{spinterL}} \downarrow }$$ and $$R_{{\mathrm{spinterR}} \downarrow }$$ are the resistance arising from the spin-filtering effect in the down-spin channel on the LPCMO (L) side and the Co (R) side, respectively.

With zero or a small Bpre, the area of FMM/COI boundaries inside LPCMO thin film is large because of strong EPS. A strong spin-flip scattering at the boundary causes Rsf to be close to zero and randomize the spin-up and spin-down channel, resulting in a significantly decreased spin polarization degree via LPCMO film. Hence there will be no net spin polarization current transmitting through the LPCMO film and thus MR value is small (see Fig. 6a).

By applying a large Bpre, the FMM/COI boundaries diminish because of the percolation of FMM phases (Fig. 6b). The effective cross-section of the device increases, causing RAlq3, RspinterL↓,RspinterR↓, and RLPCMOL↓ to decrease inversely proportionally to the effective cross section area of the FMM phase. Presumably, the MR effect should remain unchanged if all the resistances decrease with the same proportion. However, the boundary induced spin flip scattering is greatly suppressed due to diminished FMM/COI phase mixing, i.e., Rsf becomes significantly larger ( → ∞). The increase of Rsf compensates the overall reduction of RLPCMO↓ in Eq. (4). It leads to a slightly reduced RAP in the antiparallel configuration, compared to the greatly reduced RP as shown in Fig. 2a. Thus ΔR(Bpre) increases as a function of Bpre, making the MR response increase dramatically. We note that the absence of the enhanced MR response in the LPCMO/Al2O3/Co device suggests that a highly polarized spin filtering effect (i.e., a large RspinterL(R)↓) at the LPCMO/Alq3 spinterface is required to realize the observed results (see Supplementary Fig. 7). A nonlinear increase of spin filtering resistance as a function of spin polarization degree from the LPCMO/Alq3 spinterface may also contribute to the overall MR response. Therefore, the observed large MR response can be understood as synergistic contributions from the tunable spin polarization degree at the FMM/COI boundary and spin polarization dependent spin filtering effect at the LPCMO/Alq3 spinterface. To connect Eq. (4) with the MR change as a function of FMM ratio and phase boundary length, we assume that Rsf, RLPCMO↓, and RspinterL(R)↓ are larger than all other resistances, and $$R_{{\mathrm{spinterL}}({\mathrm{R}}) \downarrow } \ll R_{{\mathrm{LPCMO}} \downarrow }$$. The AP resistance is simplified to,

$$R_{{\mathrm{AP}}} \approx \frac{{R_{{\mathrm{sf}}}R_{{\mathrm{spinterR}} \downarrow }}}{{R_{{\mathrm{spinterL}} \downarrow } + R_{{\mathrm{spinterR}} \downarrow } + R_{{\mathrm{sf}}}}};$$
(5)

We further assume that $$R_{{\mathrm{spinterL}}({\mathrm{R}}) \downarrow }\left( {B_{{\mathrm{pre}}}} \right) = R_{{\mathrm{spinterL}}({\mathrm{R}}) \downarrow }(0)/A$$, $$R_{{\mathrm{Alq}}3}\left( {B_{{\mathrm{pre}}}} \right) = R_{{\mathrm{Alq}}3}(0)/A$$ where A is the FMM ratio, and Rsf = C/L where L is the total length of the phase boundary and C is a constant. The MR becomes,

$$\begin{array}{l}{\mathrm{MR}} + 1 = \frac{{R_{{\mathrm{AP}}}}}{{R_{\mathrm{P}}}} \approx \\ \frac{{CR_{{\mathrm{spinterL}} \downarrow }\left( 0 \right)}}{{R_{{\mathrm{Alq}}3}\left( 0 \right)\left[ {R_{{\mathrm{spinterL}} \downarrow }\left( 0 \right) + R_{{\mathrm{spinterR}} \downarrow }(0)} \right]\frac{L}{A} + R_{{\mathrm{Alq}}3}\left( 0 \right)C}} = \\ \frac{{MR_{{\mathrm{max}}} + 1}}{{1 + \frac{{R_{{\mathrm{spinterL}} \downarrow }\left( 0 \right) + R_{{\mathrm{spinterR}} \downarrow }(0)}}{C}\frac{L}{A}}};\end{array}$$
(6)

Comparing to Eq. (2), we find $$\alpha = \frac{{R_{{\mathrm{spinterL}} \downarrow }\left( 0 \right) + R_{{\mathrm{spinterR}} \downarrow }(0)}}{C}$$.

Our proposed model is further validated by the temperature dependence of MR response in the LPCMO-OSV device. Figure 6e shows the obtained MR values as a function of temperature under different Bpre at V = 0.2 V. In the presence of Bpre, the MR response decreases monotonically with increasing temperature, and disappears above T ~ 150 K due to diminished magnetization of the LPCMO film (Fig. 1c). The ratio between MR(Bpre = 7 T) and MR(Bpre = 1 T) is plotted as a function of temperature in Fig. 6f. The MR(Bpre = 7 T)/MR(Bpre = 1 T) ratio shows a maximum around T = 50 K, which is consistent with the trend of $$|dM/dT|$$ vs. T plot in the inset of Fig. 1c. The experimentally obtained values of MR, RP, and RAP at each temperature and Bpre are put into Eqs. (5) and (6), from which Rsf can be numerically calculated and is plotted in Fig. 6g. We found that the temperature-dependence of the ratio between Rsf at Bpre = 7 T and Bpre = 1 T (Fig. 6h) agrees well with that of the MR(Bpre = 7 T)/MR(Bpre = 1 T) ratio shown in Fig. 6f. This provides a compelling evidence that the boundary-induced spin-flip scattering (Rsf) plays an essential role in determining the MR response in the LPCMO-OSV device.

## Discussion

We demonstrate that a large MR effect of ~440% can be created in Alq3 molecule-based spin valve device using a LPCMO thin film with pronounced EPS as the bottom electrode. The EPS state can be tuned by applying a Bpre, allowing the contactless control of area fraction of the FMM phase in the bottom electrode. As a result, the MR response can be tuned by Bpre within six times difference in magnitude. Such an unusually large yet tunable MR effect is likely governed by a tunable spin-flip scattering at the boundaries between FMM and COI phases inside the LPCMO film and a large spin filtering effect at the molecule/oxide spinterface. The large MR effect enables us to demonstrate the first electrical Hanle effect in the Alq3-based OSV device as the success of spin injection and detection in the organic layer. The obtained spin lifetime (or the transit time) of electrically injected spin carriers (~45 ps) is surprisingly shorter than that of the bulk Alq3 layer despite its weak spin-orbit coupling, indicating the dominant role of the exchange-mediated spin transport. Our results demonstrate a novel approach for tuning MR response in organic spintronic devices via controlling EPS state in manganite electrodes. The proposed phase boundary-induced spin flip mechanism sheds light on understanding the origin of spin injection efficiency in the manganite films. Our study opens a new route towards logic switching for future high-performance molecular memory/storage devices.

## Methods

### Sample fabrication

LPCMO films (60 nm) were grown on SrTiO3 (100) substrates under the oxygen pressure of 8 × 10−4 Torr using pulsed laser deposition58. During the deposition, the film thickness was monitored by in-situ Reflection High Energy Electron Diffraction (RHEED). The LPCMO bottom electrodes were fabricated from the LPCMO thin films by wet photo-lithography59. After ultrasonic cleaning in alcohol and acetone, the LPCMO electrodes were immediately transferred into a vacuum chamber with a base pressure of 7.5 × 10−11 Torr. The Alq3 (99.995%, Aldrich) spacer layer (thickness: ~60 nm) was thermally evaporated at room temperature, followed by e-beam deposition of Co (10 nm)/Au (10 nm) top electrodes in a crossbar configuration using a shadow mask. The substrate was kept at 280 K during the top electrode deposition process to suppress the interdiffusion between the top electrode and the Alq3 layer10. The thickness of each layer was monitored by a quartz thickness monitor which had been calibrated by a profilometer. The active device area was about 20 × 120 μm.

### Magnetization and magnetoresistance measurements

Magnetization measurements were carried out using a Quantum Design superconducting quantum interference device system (SQUID). Magnetic fields were applied parallel to the device plane, which is the easy magnetization direction for both bottom and top magnetic electrodes.

Magnetoresistance measurements were carried out using a Quantum Design Physical Property Measurement System (PPMS) combined with a Keithley 2400 source meter. The magnetic fields were applied along the in-plane direction of the device. The MR is defined as MR = (RantiparallelRparallel)/Rparallel, where Rantiparallel (Rparallel) is the device resistance in the antiparallel (parallel) magnetic configuration of the two ferromagnetic electrodes. The device resistance was measured by two-point probe method.

The MR response is controlled by three factors: temperature (T), pre-set magnetic field (Bpre), and applied bias (V). Prior to measurements at each chosen temperature (e.g., 10 K, 25 K, 50 K, 75 K, 100 K, 125 K, and 150 K), the device was always set to room temperature following by cooling under zero magnetic field. After the chosen temperature was reached, Bpre (1 T, 2 T, 3 T, 4 T, 5 T, 6 T, and 7 T) was first applied to the device and then removed allowing us to control the area fraction of the FMM phase in the LPCMO bottom electrode. A voltage bias (e.g., V = 0.1 V, 0.2 V, 0.3 V, 0.4 V, 0.5 V, and 0.6 V) was applied to the device from which the resistance of device was recorded using a LabVIEW program. The MR response was measured by sweeping the magnetic field continuously from 1 T to −1 T and back to 1 T.

### Hanle measurements

The Hanle effect measurements were taken at T = 10 K by applying a voltage bias V = 0.1 V in the PPMS system. For the Hanle measurement in the parallel configuration, the MR loop is firstly measured along the in-plane magnetic field direction under the Bpre = 7 T to achieve the largest MR value in this device (~440%). After the MR loop measurement, the in-plane magnetic field is directly sweeping back to zero in order to maintain a ‘parallel’ configuration of two magnetic layers of the device (i.e., the low resistance state). Second, the device is rotated by 90 degrees to apply a perpendicular magnetic field (i.e., Bz is sweeping from +0.2 T to −0.2 T) to measure the Hanle effect in this parallel configuration of the device.

For the Hanle effect in the antiparallel configuration, the MR loop is firstly measured along the in-plane field direction to achieve the largest MR. From the Bin-plane = +1 T, the magnetic field is sweeping down to Bin-plane = −30 mT and then reducing to zero to achieve an antiparallel configuration of two ferromagnetic layers (i.e., the high resistance state). Then the device is rotated by 90 degrees to apply the perpendicular magnetic field (i.e., Bz is sweeping from +0.2 T to −0.2 T) to perform the Hanle measurement.

After finishing the Hanle measurement, the device is rotated back to the in-plane field geometry to make sure that the same MR loop is repeated after applying the perpendicular magnetic field.

### Magnetic force microscope measurements

MFM images were acquired from the same LPCMO thin film which was later used as the bottom electrode for LPCMO-OSV device. The MFM measurements were carried out in the PPMS System using Attocube scanning probe microscope following the same procedure for magnetoresistance measurements. For measurements at different temperatures, the temperature was first raised to room temperature at which the LPCMO becomes paramagnetic, and then cooled under zero field to a chosen temperature for MFM measurements on the LPCMO thin film. Note that the temperature cycle does not change the global resistivity and magnetization of the LPCMO film at a chosen temperature, although the EPS pattern varies even at the same chosen temperature after different temperature cycle. All the MFM images were taken at a constant magnetic field (Bext = 1 T) after withdrawing Bpre. The calculated area fraction of the FMM phase based on the MFM images should thus be similar to that in the LPCMO bottom electrode of the device.

## Data availability

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

## References

1. 1.

Muccini, M. A bright future for organic field-effect transistors. Nat. Mater. 5, 605 (2006).

2. 2.

Capelli, R. et al. Organic light-emitting transistors with an efficiency that outperforms the equivalent light-emitting diodes. Nat. Mater. 9, 496 (2010).

3. 3.

Kent, A. D. & Worledge, D. C. A new spin on magnetic memories. Nat. Nanotechnol. 10, 187 (2015).

4. 4.

Chappert, C., Fert, A. & Van Dau, F. N. The emergence of spin electronics in data storage. Nat. Mater. 6, 813 (2007).

5. 5.

Schmidt, G. et al. Fundamental obstacle for electrical spin injection from a ferromagnetic metal into a diffusive semiconductor. Phys. Rev. B 62, R4790 (2000).

6. 6.

Ando, K. et al. Electrically tunable spin injector free from the impedance mismatch problem. Nat. Mater. 10, 655 (2011).

7. 7.

Fert, A. & Jaffres, H. Conditions for efficient spin injection from a ferromagnetic metal into a semiconductor. Phys. Rev. B 64, 184420 (2001).

8. 8.

Min, B. C., Motohashi, K., Lodder, C. & Jansen, R. Tunable spin-tunnel contacts to silicon using low-work-function ferromagnets. Nat. Mater. 5, 817 (2006).

9. 9.

Schulz, L. et al. Engineering spin propagation across a hybrid organic/inorganic interface using a polar layer. Nat. Mater. 10, 39 (2011).

10. 10.

Sun, D. et al. Active control of magnetoresistance of organic spin valves using ferroelectricity. Nat. Commun. 5, 4396 (2014).

11. 11.

Santos, T. S. et al. Room-temperature tunnel magnetoresistance and spin-polarized tunneling through an organic semiconductor barrier. Phys. Rev. Lett. 98, 16601 (2007).

12. 12.

Tsukagoshi, K., Alphenaar, B. W. & Ago, H. Coherent transport of electron spin in a ferromagnetically contacted carbon nanotube. Nature 401, 572 (1999).

13. 13.

Dediu, V. et al. Room temperature spin polarized injection in organic semiconductor. Solid State Commun. 122, 181–184 (2002).

14. 14.

Xiong, Z. H., Wu, D., Vardeny, Z. V. & Shi, J. Giant magnetoresistance in organic spin-valves. Nature 427, 821–824 (2004).

15. 15.

Dediu, V. A., Hueso, L. E., Bergenti, I. & Taliani, C. Spin routes in organic semiconductors. Nat. Mater. 8, 707 (2009).

16. 16.

Sanvito, S. Molecular spintronics: the rise of spinterface science. Nat. Phys. 6, 562–564 (2010).

17. 17.

Barraud, Clément et al. Unravelling the role of the interface for spin injection into organic semiconductors. Nat. Phys. 6, 615–620 (2010).

18. 18.

Galbiati, M. et al. Spinterface: crafting spintronics at the molecular scale. Mrs Bull. 39, 602–607 (2014).

19. 19.

Steil, S. et al. Spin-dependent trapping of electrons at spinterfaces. Nat. Phys. 9, 242–247 (2013).

20. 20.

Raman, K. V. et al. Interface-engineered templates for molecular spin memory devices. Nature 493, 509 (2013).

21. 21.

Cinchetti, M., Dediu, V. A. & Hueso, L. E. Activating the molecular spinterface. Nat. Mater. 16, 507 (2017).

22. 22.

Sun, D. et al. Giant magnetoresistance in organic spin valves. Phys. Rev. Lett. 104, 236602 (2010).

23. 23.

Nguyen, T. D. et al. Isotope effect in spin response of pi-conjugated polymer films and devices. Nat. Mater. 9, 345–352 (2010).

24. 24.

Ando, K. et al. Solution-processed organic spin–charge converter. Nat. Mater. 12, 622 (2013).

25. 25.

Watanabe, S. et al. Polaron spin current transport in organic semiconductors. Nat. Phys. 10, 308 (2014).

26. 26.

Drew, A. J. et al. Direct measurement of the electronic spin diffusion length in a fully functional organic spin valve by low-energy muon spin rotation. Nat. Mater. 8, 109 (2009).

27. 27.

Cinchetti, M. et al. Determination of spin injection and transport in a ferromagnet/organic semiconductor heterojunction by two-photon photoemission. Nat. Mater. 8, 115 (2009).

28. 28.

Sun, X. et al. A molecular spin-photovoltaic device. Science 357, 677–680 (2017).

29. 29.

Wang, F. J., Yang, C. G., Vardeny, Z. V. & Li, X. G. Spin response in organic spin valves based on La2/3Sr1/3MnO3 electrodes. Phys. Rev. B 75, 245324 (2007).

30. 30.

Dediu, V. et al. Room-temperature spintronic effects in Alq3-based hybrid devices. Phys. Rev. B 78, 115203 (2008).

31. 31.

Uehara, M., Mori, S., Chen, C. H. & Cheong, S. W. Percolative phase separation underlies colossal magnetoresistance in mixed-valent manganites. Nature 399, 560 (1999).

32. 32.

Ward, T. Z. et al. Elastically driven anisotropic percolation in electronic phase-separated manganites. Nat. Phys. 5, 885 (2009).

33. 33.

Julliere, M. Tunneling between ferromagnetic films. Phys. Lett. A 54, 225–226 (1975).

34. 34.

Tedrow, P. M. & Meservey, R. Spin polarization of electrons tunneling from films of Fe, Co, Ni, and Gd. Phys. Rev. B 7, 318 (1973).

35. 35.

Singh-Bhalla, G. et al. Intrinsic tunneling in phase separated manganites. Phys. Rev. Lett. 102, 077205 (2009).

36. 36.

Sen, C., Alvarez, G. & Dagotto, E. Competing ferromagnetic and charge-ordered states in models for manganites: The origin of the colossal magnetoresistance effect. Phys. Rev. Lett. 98, 127202 (2007).

37. 37.

Du, K. et al. Visualization of a ferromagnetic metallic edge state in manganite strips. Nat. Commun. 6, 6179 (2015).

38. 38.

Zhai, H. Y. et al. Giant discrete steps in metal-insulator transition in perovskite manganite wires. Phys. Rev. Lett. 97, 167201 (2006).

39. 39.

Ward, T. Z. et al. Reemergent metal-insulator transitions in manganites exposed with spatial confinement. Phys. Rev. Lett. 100, 247204 (2008).

40. 40.

Gould, C. et al. Tunneling anisotropic magnetoresistance: a spin-valve-like tunnel magnetoresistance using a single magnetic layer. Phys. Rev. Lett. 93, 117203 (2004).

41. 41.

Singh-Bhalla, G., Biswas, A. & Hebard, A. F. Tunneling magnetoresistance in phase-separated manganite nanobridges. Phys. Rev. B 80, 144410 (2009).

42. 42.

Bobbert, P. A. et al. Bipolaron mechanism for organic magnetoresistance. Phys. Rev. Lett. 99, 216801 (2007).

43. 43.

Hu, B. & Wu, Y. Tuning magnetoresistance between positive and negative values in organic semiconductors. Nat. Mater. 6, 985 (2007).

44. 44.

Wang, F. & Vardeny, Z. V. Recent advances in organic spin-valve devices. Synth. Met. 160, 210–215 (2010).

45. 45.

Sun, J. Z., Abraham, D. W., Roche, K. & Parkin, S. S. P. Temperature and bias dependence of magnetoresistance in doped manganite thin film trilayer junctions. Appl. Phys. Lett. 73, 1008–1010 (1998).

46. 46.

Harmon, N. J. & Flatté, M. E. Distinguishing spin relaxation mechanisms in organic semiconductors. Phys. Rev. Lett. 110, 176602 (2013).

47. 47.

Ardavan, A. et al. Will spin-relaxation times in molecular magnets permit quantum information processing? Phys. Rev. Lett. 98, 057201 (2007).

48. 48.

Moodera, J. S., Nowak, J. & van de Veerdonk, R. J. Interface magnetism and spin wave scattering in ferromagnet-insulator-ferromagnet tunnel junctions. Phys. Rev. Lett. 80, 2941 (1998).

49. 49.

Wang, J. et al. Spin-optoelectronic devices based on hybrid organic-inorganic trihalide perovskites. Nat. Commun. 10, 129 (2019).

50. 50.

Han, W. et al. Spin injection and detection in lanthanum-and niobium-doped SrTiO3 using the Hanle technique. Nature. Communications 4, 2134 (2013).

51. 51.

Dash, S. P., Sharma, S., Patel, R. S., de Jong, M. P. & Jansen, R. Electrical creation of spin polarization in silicon at room temperature. Nature 462, 491 (2009).

52. 52.

Lou, X. et al. Electrical detection of spin transport in lateral ferromagnet–semiconductor devices. Nat. Phys. 3, 197 (2007).

53. 53.

Grünewald, M. et al. Vertical organic spin valves in perpendicular magnetic fields. Phys. Rev. B 88, 085319 (2013).

54. 54.

Yu, Z. G. Impurity-band transport in organic spin valves. Nat. Commun. 5, 4842 (2014).

55. 55.

Yu, Z. G. Suppression of the Hanle effect in organic spintronic devices. Phys. Rev. Lett. 111, 016601 (2013).

56. 56.

Fert, A. Two-current conduction in ferromagnetic metals and spin wave-electron collisions. J. Phys. C 2, 1784 (1969).

57. 57.

Yao, W., Xiao, D. & Niu, Q. Valley-dependent optoelectronics from inversion symmetry breaking. Phys. Rev. B 77, 235406 (2008).

58. 58.

Ma, J. X., Gillaspie, D. T., Plummer, E. W. & Shen, J. Visualization of localized holes in manganite thin films with atomic resolution. Phys. Rev. Lett. 95, 237210 (2005).

59. 59.

Minsky, M. S., White, M. & Hu, E. L. Room‐temperature photo enhanced wet etching of GaN. Appl. Phys. Lett. 68, 1531–1533 (1996).

## Acknowledgements

This work was supported by the National Key Research Program of China (2016YFA0300702), National Basic Research Program of China (973 Program) under Grants No. 2014CB921104, National Natural Science Foundation of China (11504053), the Program of Shanghai Academic Research Leader (18XD1400600) and the Shanghai Municipal Natural Science Foundation (18JC1411400, 18ZR1403200). D.S. was supported by the start-up funding at the North Carolina State University and NC State University-Nagoya Research Collaboration Grant. X.-G.Z. was supported by the US National Science Foundation DMREF-1534401.

## Author information

Authors

### Contributions

J.S. and D.S. conceived this study and the experiments. W.Y. fabricated the devices and performed the MR measurement. W.Y., Q.S., T.M., and Y.B. grew the LPCMO and LSMO films. Y.Z. and L.D. provided the LPCMO targets. Q.L. and H.L. was responsible for the MFM measurement. P.C. and H.-X.L. assisted in performing the MR measurement. W.Y. and Y.Y. did the wet photolithography. W.W. and L.Y. assisted in the data discussion. X.-G.Z. was responsible for the theoretical model. D.S. and W.Y. wrote the first draft of the paper. J.S. was responsible for the project planning, group managing, and paper final writing. All authors discussed the results, worked on data analysis and paper preparation.

### Corresponding authors

Correspondence to Dali Sun or Jian Shen.

## Ethics declarations

### Competing interests

The authors declare no competing interests.

Peer review information: Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work.

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## Rights and permissions

Reprints and Permissions

Yang, W., Shi, Q., Miao, T. et al. Achieving large and nonvolatile tunable magnetoresistance in organic spin valves using electronic phase separated manganites. Nat Commun 10, 3877 (2019). https://doi.org/10.1038/s41467-019-11827-0

• Accepted:

• Published:

• ### Cornerstone of molecular spintronics: Strategies for reliable organic spin valves

• Shuaishuai Ding
• , Yuan Tian
•  & Wenping Hu

Nano Research (2021)

• ### Unveiling the role of Fe3O4 in polymer spin valve near Verwey transition

• Shuaishuai Ding
• , Yuan Tian
• , Xiang Liu
• , Ye Zou
• , Huanli Dong
• , Wenbo Mi
•  & Wenping Hu

Nano Research (2021)