Reliable Spin Valves of Conjugated PolymerBased on Mechanically Transferrable TopElectrodesShuaishuai Ding,†,∥ Yuan Tian,*,† Hanlin Wang,† Zhang Zhou,§,∥ Wenbo Mi,⊥ Zhenjie Ni,† Ye Zou,†

Huanli Dong,† Hongjun Gao,§ Daoben Zhu,† and Wenping Hu*,†,‡

†Beijing National Laboratory for Molecular Sciences, Key Laboratory of Organic Solids, Institute of Chemistry, Chinese Academy ofSciences, Beijing 100190, China‡Tianjin Key Laboratory of Molecular Optoelectronic Sciences, Department of Chemistry, School of Sciences, Tianjin University,Tianjin 300072, China§Beijing National Laboratory for Condensed Matter Physics and Institute of Physics, Chinese Academy of Sciences, Beijing 100190,China⊥Tianjin Key Laboratory of Low Dimensional Materials Physics and Preparation Technology, School of Science, Tianjin University,Tianjin 300354, China∥University of Chinese Academy of Sciences, Beijing 100049, China

*S Supporting Information

ABSTRACT: Organic spintronic devices present one ofthe most appealing technologies for future spintronicdevices by taking advantage of the spin degree of freedom.Conjugated polymers are attractive for the exemplifieddevice of organic spin valves (OSVs) due to their weakspin−orbit coupling, solution-processability, low produc-tion cost, and mechanical flexibility. However, theperformance of polymer SVs is a matter of debate, as theevaporated top ferromagnetic (FM) electrode will pene-trate into the organic layer during a typical fabrication process, especially in the device with an organic layer thickness ofnanometers. It will cause a severe problem in controllable and reproducible spin manipulations, not to mention theclarification of the spin-dependent transport mechanism. Here, a universal, simple, and low-cost method based on atransferred electrode is developed for a polymer spin valve with stable and reliable state operation. It is demonstrated inan OSV device with a vertical structure of La2/3Sr1/3MnO3 (LSMO)/P3HT/AlOx/Co/Au that this approach not onlybuilds a damage-free interface between magnetic electrodes and an organic spacer layer but also can be generalized forother devices with delicate active layers. Furthermore, a multistate writing and reading prototype is achieved on thepremise of robust and quick magnetic response. The results reveal the importance of a spinterface and effective thicknessof the organic layer in fundamental spintronic research and may lead to a strong potential in future flexible, large-area,and robust organic multifunctional circuits.KEYWORDS: organic spin valves, polymer spintronics, organic spintronics integration, metal penetration, spin-interface

Spintronics is a principle that exploits the magnetic natureof electrons and manipulates electronics with the spindegree of freedom, which shows great potential in

nonvolatile data storage and quantum computing.1−3 As aprincipal spintronics prototype of hard-disk-drive read heads,spin-valves (SVs) have been highly attractive and hold promisein the next-generation flexible and optoelectronics.4 It is asandwiched structure composed of two soft ferromagnetic(FM) electrodes with different coercivity decoupled by anonmagnetic interlayer.5,6 Defined as MR = (Rap − Rp)/Rp,

7,8

magnetoresistance (MR) is the relative change in electrical

resistance of an SV caused by different spin-dependentscatterings with respect to the relative alignment of magnet-ization (parallel or antiparallel) in FM electrodes when theexternal magnetic field is swept. A considerable variation canbe detected only if the spin information can be reserved duringthe spin transport process in the nonmagnetic layer. Theintroduction of organic materials into the field of spintronics is

Received: September 30, 2018Accepted: November 9, 2018Published: November 9, 2018

Artic

lewww.acsnano.orgCite This: ACS Nano 2018, 12, 12657−12664

© 2018 American Chemical Society 12657 DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

Dow

nloa

ded

via

INST

OF

PHY

SIC

S on

Jan

uary

9, 2

019

at 0

1:39

:59

(UT

C).

Se

e ht

tps:

//pub

s.ac

s.or

g/sh

arin

ggui

delin

es f

or o

ptio

ns o

n ho

w to

legi

timat

ely

shar

e pu

blis

hed

artic

les.

highly preferable due to their weak spin-scatterings induced bylow spin−orbit coupling.8,9 Especially, the detected spinrelaxation time of organics (in the range of μs)10 is muchlonger than that of their inorganic counterparts by orders ofmagnitude.2,11 Moreover, π-conjugated polymers benefit froman abundance of supply, solution-processability, low cost, lightweight, and mechanical flexibility,12 serving as good candidatesin large-scale optoelectronic applications.13 Thus, the compre-hensive investigations of π-conjugated polymer based organicspin valves (OSVs) will not only provide some insights into thefundamental spintronics research but also boost the develop-ment for the multifunctional integration of organic circuitsmanipulated by light and electric and magnetic fields.However, delicate organic materials often suffer from the

unexpected performance of MR inversion since the first verticalOSV device achieved by Xiong et al.14 in 2004. Even in thesame device configuration with identical fabrication proce-dures, both positive and negative MR could be observed.15,16

Formation of spin-hybridization-induced polarized states17

between the first molecular layer and the top electrodeinterface11,18 is considered one of the most probable reasons.Both the metal−organics contact and metal penetration inorganic materials could lead to a spin-dependent hybrid-interface state, while the latter is dominant and uncontrollablein most experiments,15,16 which impairs spin injection andtransport in organic layers.19,20 In some cases, a properformation of an interface state is beneficial for spin injection.However, it is quite different for individual devices since the

damage to an organic layer caused by metal penetration duringdirect evaporation is a random process, let alone for theirreversibility and severe decay in performance for fragileorganic materials.21,22 The magnetic atoms induced by directevaporation can even cause pinholes and inclusion in theorganic layer over a distance of ∼100 nm,23 in which theeffective thickness of the organic layer cannot be clearlydefined and the contributions of such a dead layer and pureorganic materials in spintronics are hard to distinguish. Thisuncontrollable interface state and irreversible damages to theorganic layer caused by metal penetration will not only makethe spin mechanism research more complex and debatable butalso limit the practical applications of organic spin valves dueto poor quality and performance control. Therefore, the keypoint toward a reliable and stable OSV device is to obtain ametal-penetration-free interface with little damage to theorganics.To date, it is still difficult to suppress the top metallic

electrode penetration, the damage involving “hot” atoms to thespinterface and organic layer during the traditional evaporationprocess16 in OSV fabrication. Insertion of a thin tunnelingbarrier (e.g., AlOx) between the organic layer and theelectrodes seems to be an efficient way to reduce such metalpenetration;23−25 however, the thickness of AlOx is hard tocontrol. Thick AlOx might hinder effective spin transport,7

while thin AlOx cannot fully prevent metal penetration into thefragile organic layers. Another strategy is using a low-temperature fabrication method by cooling the substrate with

Figure 1. Organic spin valve fabrication and structure. (a, b) Fabrication process of the organic spin valve. Top ferromagnetic electrodes aretransferred onto the polymer surface by a sacrificial water-soluble layer. (c) Cross-sectional schematic diagram of the device structure andmolecular structure used in experiments. (d) Evaporated top electrodes on spin-coated PSS. (e) Self-encapsulated electrodes by PSsupporting film. (f) Bottom electrodes LSMO and spin-coated P3HT. (g) Top view of the spin valve device fabricated by a transferred topferromagnetic electrode. All of the fabricated devices have a fixed junction area of 500 × 500 μm2. Scale bar: 100 μm.

ACS Nano Article

DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

12658

liquid N2 during the deposition, for which special experienceand high-cost high-vacuum equipment reformation areneeded.9 An indirect deposition method with the assistanceof an inert gas is not an option for the same reasons.26 Asuccessful approach named buffer layer assisted growth(BLAG) with a top Co electrode exhibits an extraordinaryMR ratio of up to ∼300%. Nevertheless, even such a skillfulmethod is incapable of hindering the ill-defined layercompletely when the organic layer thickness is less than 23nm.27 All these existing methods are either expensive andexclusive of a high threshold of technique or cannot completelyprevent the metal penetration, limiting their practicalapplication for most laboratories and further industrialintegration. Hence, for the majority of researchers, a universal,low-cost, and straightforward method that eliminates interfacialdiffusion is desired for the achievement of a reliable and stableOSV device.Herein we report a general approach for the formation of

top FM electrodes in polymer SV devices for a stable andreliable state operation. In our method, the mechanicallytransferrable top electrode is thermally deposited on a water-solvable polymer poly(sodium 4-styrenesulfonate) (PSS),peeled off, and transferred onto the target polymer/bottomelectrode, forming a mechanical contact at the top interface.This damage-free strategy for metal penetration, justified inboth organic28 and inorganic electronics29 before, hasadvantages of generality, practicability, low manufacturingcost, and simple technology in organic spintronics devicefabrication compared to the methods mentioned above. Theoptical microscope image, X-ray photoelectron spectroscopy(XPS), and M−H characterization show a well-preservedquality of the as-transferred FM electrode, which is applicableto the OSV device. As a noted classical material with better

compatibility in the solution process, poly(3-hexylthiophene-2,5-diyl) (P3HT) is taken in a traditional vertical spin valvestructure as an example to show that our method is capable ofobtaining real MR signal output stably and reliably with a clearevaluation of the effective thickness in the organic layer. Amultistate writing and reading prototype is further achieved onthe premise of such a robust and quick magnetic response. Ourresearch provides an advanced strategy to improve OSV devicereliability hindered by the sensitive spinterface and thermaldamage to organics due to metal penetration. It also offers apotential advance to solution-accessible OSV units in massproduction, benefiting future applications in large-scale organicintegrated circuits as well.

RESULTS AND DISCUSSION

The device fabrication steps are sketched in Figure 1a,b.Briefly, 1 nm Al followed by immediate oxidation, 10 nm ofCo, and 60 nm of Au were subsequently evaporated on thespin-coated water-soluble poly(4-styrenesulfonate) (PSS)layer. Then a polystyrene (PS) membrane was spin-coatedonto the top electrodes (60 nm Au/10 nm Co/∼1 nm AlOx)to facilitate the transfer process and encapsulate the device. Bya poly(dimethylsiloxane) (PDMS)-assisted transfer technique,it is easy to completely peel off the top electrode in water andtransfer it onto the target P3HT/LSMO substrate. Detailedfabrication methods are shown in the Experimental Section.Once batch fabrication of evaporated top electrodes is done,such a simple transfer procedure could be finished within 15min (including baking time) even in the laboratory with onlythe most basic and fundamental equipment, i.e., a spin coaterand a hot-stage. As schematically illustrated in Figure 1c, theas-prepared polymer SV has a typical vertical cross-barstructure with the following layers: PS capping layer/Au (60

Figure 2. Characterization of top Co electrodes. Magnetic hysteresis loops for top ferromagnetic electrodes (a) before and (b) after transfer,fabricated by e-beam evaporation with AlOx and measured at T = 2 K, respectively. (c) Na 1s XPS spectra for different layers during thetransfer process. (d) Comparison of Al 2p XPS spectra between inverted top electrodes after transfer and fully oxidized AlOx on a Sisubstrate. (e) EDS mapping of Co element. The electrodes were transferred as described in the Experimental Section, then some of theelectrodes were removed by peeling off the PDMS stamps.

ACS Nano Article

DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

12659

nm)/Co (10 nm)/AlOx (∼1 nm)/P3HT (∼20 nm)/LSMO(100 nm). LSMO, with high spin polarization and stability tooxygen/water,14 serves as a source, while Co as a detector ofthe spin-polarized transport in the P3HT spacer (Figure S1,Supporting Information). A semioxidized AlOx is inserted tophysically prevent direct contact with the chemically active Co.Optical microscope images in Figure 1d−g declare thecompleteness and flatness of the top electrodes before andafter the transfer process, partly confirming the validity of themethod since the magnetic properties will be substantiallyreduced in wrinkled electrodes.To study the influence of the fabrication method and

transfer procedure on top FM electrode quality, a detailedinvestigation on Co electrodes is shown in Figure 2. Themagnetization of all three kinds of electrodes, including AlOx/thermal-evaporated (TE) Co, AlOx/e-beam-evaporated (EB)Co, and e-beam-evaporated Co without AlOx, has declined tosome extent after the peel-off and transfer procedure (FigureS2, Supporting Information). The quality of thermal-evaporated Co is worse than those of e-beam-evaporatedones both before and after transfer, indicating a looserstructure in the TE electrode, which could be attributed tothe lower amount of energy and heat accumulation for atomreconstruction during the thermal evaporation. In the presenceof AlOx the magnetic properties of e-beam-evaporated Co ispreserved as much as possible during the transfer procedure(Figure 2a,b), which is crucial for the spin detection. XPSspectra for different layer surfaces during the transfer processare depicted in Figure S3, Supporting Information. No peakswere detected in the inverted AlOx/Co/Au/PS membrane inNa 1s spectra (Figure 2c) compared to the standard PSSsample, indicating minimal PSS residual during the peel-off

process. The removal of the PSS residual is necessary so thatwe can detect the nature of the organic layer itself instead of aspin transport in the organic layer/PSS buffer layerheterostructure. In Al 2p spectra, the peak position of theinverted AlOx/Co sample shifts toward lower binding energythan fully oxidized Al evaporated on a Si substrate (Figure 2d).This result points out that the transferred thin AlOx layerprobably deviates from the standard Al2O3,

30 indicating anactually electrically leaky state with transport properties closeto those of a bad metal. Such a leaky AlOx barrier is crucial forachieving effective spin transport in vertical OSVs.7 Todemonstrate the gentle, “low-energy” material integration, wemechanically peeled off the transferred metal electrodes fromthe P3HT interface after the device fabrication. Elementalmapping based on energy-dispersive X-ray spectroscopy (EDS)(Figure S4, Supporting Information) confirmed that themechanical contact without direct chemical bonding of Coand the P3HT molecule can highly suppress the spin-dependent hybrid-interface state induced by metal penetration,since most of the Co could be removed after peeling off theelectrodes (Figure 2e). Compared to the directly evaporatedapproach, we provide a gentle method to form a mechanicallyformed electric contact since there is no extra energy to impelthe metal atoms to defuse into the organic layer like theevaporation one. Theoretically, it is compatible with anythermally fragile organic materials that would be easilydamaged by aggressive fabrication in a tradition procedure,dramatically expanding the range of organic spintronic researchin materials. The optical microscopy images (Figure 1d−g),M−H characterizations (Figure 2a,b), XPS spectrum (Figure2c,d), and EDS mapping (Figure 2e) together testify theprotection of AlOx and the quality of transferred magnetic

Figure 3. Electrical and magnetotransport measurement in an LSMO/P3HT/AlOx/Co/Au spin valve. (a) Resistance characteristics withtemperatures ranging from 300 to 2 K at a constant current of I = 0.01 μA. (b) dI−dV curves for the device at different temperatures. Itshows parabolic curves near zero-bias. (c) Typical square MR response curve measured at T = 2 K, I = 0.15 μA. The black line and red linerepresent the MR value while the magnetic field is swept from positive to negative and back from negative to positive, respectively. (d)Temperature dependence of MR at I = 0.01 μA and current dependence of MR at T = 2 K.

ACS Nano Article

DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

12660

electrodes, which is the foundation for obtaining reliable andstable MR signal.Figure 3a displays device resistance as the temperature

dropped from 300 to 2 K for the 20-nm-thick P3HT device(Figure S1, Supporting Information) at a current of I = 0.01μA. The stability and reliability are manifested by comparingthe R−T curves with the same experimental settings fabricatedin different batches (Figure S5). Such a uniform resistance ishard to achieve in traditional evaporated samples even in thesame batch.15,16 In order to explore the current transport modein P3HT-based SVs, the differential conductance (dI−dV)characteristics of the OSV under different temperatures wererecorded (Figure 3b). The dI−dV curves are almost temper-ature independent and nearly parabolic, indicating themultistep-tunneling31-dominated conducting mode in thelow-temperature zone, in which the spin-polarized carrierstransport via the intermediate states inside the HOMO−LUMO gap. This tendency is consistent with the resistancechanges shown in Figure 3a, considering the P3HT layer isthin enough to form a conducting mode by multisteptunneling. Note that the absence of a zero-bias dip in thedI−dV curve is proof of a well-defined interface withoutmagnetic impurities or pinholes between the transferredelectrodes and the P3HT layer.32 Figure 3c represents thetypical traces of resistance versus magnetic field of the OSVwith 20 nm P3HT measured at T = 2 K. Both the square shapeand the coercive force are consistent with the M−H loopshown in Figure 2b. The rapid response and steep change atboth parallel and antiparallel state suggest that the device isreliable and well stabilized, which were preserved at differentmeasurement conditions (Figure S6, Supporting Information).It should be pointed out that the junction resistance of thefabricated OSV device is about 2 or 3 orders of magnitudehigher than that of LSMO/AlOx/Co/Au

32 without a P3HTlayer, indicating the MR was mainly contributed from theorganic semiconductor material. The influence of theanisotropic magnetoresistance (AMR) effect of LSMO andCo FM electrodes can be ruled out considering their relativelow resistance value.12 The small step at low coercive fieldsaround the antiparallel state might possibly be induced bymagnetic coupling between FM electrodes4 at such a thin layerof 20 nm P3HT. The transferred TE Co could also beapplicable for such an OSV device (Figure S7, SupportingInformation) with less stability and quality control. Figure 3dsummarizes the MR ratios as functions of applied current andtemperature, which is consistent with previously reporteddata.33 Such a steep decrease of MR is partly a consequence ofweakened surface spin polarization of the LSMO electrode34

with increased temperature. Our primary goal here is tointroduce a low-cost, simple, and universal method to getreliable and stable MR signals so that many kinds of organicmaterials could be involved, while room-temperature MR canalso be expected with carefully optimized LSMO electrodes.35

The reliable OSV device with a rapid magnetic response isintrinsically related to the superior fabrication method. Inorder to gain more insight on the importance of the clearinterface and the effective thickness, three typical MR curveswith different interfaces between FM electrodes and the P3HTlayer and the effective thickness of P3HT are shown in Figure4. A thick polymer device with direct-evaporated Co usuallyexhibits an MR response with triangular background curves(Figure 4a). Especially in the interval from antiparallel toparallel state there exist several intermittent spiny noises

(Figure S8, Supporting Information), indicating an unstablemagnetic response. Furthermore, the curved shape of the MRsuggests that either the antiparallel or parallel state is not wellstabilized, most likely due to the penetrated Co conductingfilament16 in the P3HT layer generated during direct thermalevaporation (Figure 4b). It is deduced that the device exhibitsincoherent transport behavior dominated by variable rangehopping (VRH)36 or multiple trapping and release (MTR)37

with sacrificing the effective thickness of the P3HT. In someareas where the layer thickness is strongly shortened, forinstance, through pinholes/filament caused by thermalirradiation, the device might display a parallel current pathprone to show tunneling magnetoresistance (TMR).38 In thissituation both the spin diffusion length and magnetic responseperformance might be overestimated in the presence of a “deadlayer”. Especially, the direct thermal evaporation method is notapplicable to deal with a thin film whose thickness iscomparable to that of the sacrificed “dead layer” thickness(Figure 4d). Such kind of device usually shows a peculiarnegative MR response (Figure 4c, Figure S9, SupportingInformation) ascribed to the severely damaged interfaceinduced by metal penetration.11,15,16 The greatest advantageof transfer technology is that it avoids the issues of pinholescaused by artificial and external-induced metal filaments, whichmeans reliable and stable MR signals can be expected (Figure4e) on any substrate and material only if the organicsemiconductor layer is continuous and flat (Figure 4f).Moreover, the spin diffusion length, another key parameterin OSV performance evaluation, can be evaluated moreprecisely with this method since the effective transportthickness is nearly equal to the actual thickness of the organiclayer. External traps and defects induced by thermal damage tothe organic layer can be limited with less spin and chargecarrier scattering, in which MR performance will be morecontrollable with precise evaluation.By introducing transferred top electrodes, no pseudosto-

chastic behavior is observed in the interval from the parallel toantiparallel state due to the inhibited magnetic pinning sites,which usually occurs in the traditional evaporation method byatom migration from the top electrodes. The forward biascurrent and resistance output were monitored by continuously

Figure 4. Schematic illustration of metal penetration influence inthe MR response. Typical MR response and related interfacebetween polymer and top ferromagnetic electrodes of (a, b) thickpolymer with direct-evaporated Co, (c, d) thin polymer withdirect-evaporated Co, and (e, f) thin polymer with transferred Co,respectively.

ACS Nano Article

DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

12661

alternating positive and negative magnetic fields throughoutthe stability test, as plotted in Figure S10, SupportingInformation. On the premise of a stably and reliably operateddevice with a sharp and quick MR response, the resistancevalue will maintain a steady state at constant applied biascurrent and external magnetic field. We note that the stabilityis hard to reproduce in those devices with noisy and triangularMR curves (Figure 4a, Figure S8, Supporting Information)since the metal filaments might migrate irreversibly during themeasurement. Based on the cycle stability results, a multistatewriting (W) and reading (R) operation was realized on theP3HT-based OSV device (Figure 5). When the device is

initialized by a negative magnetic field (−1000 Oe), therelative change of resistance value (R1) is consistent with thechange of applied magnetic field (W), as illustrated in the redarea of the left component, while a positive initialized magneticfield (+1000 Oe) will cause an opposite resistance response(R4). The resistance signals are significantly distinguishableunder different applied bias current (Wi), and taking thisdevice for example, six distinct sets of resistance states (Ri)were formed. Besides, the relative change of resistance washighly symmetric at the same bias current; such success iscredited with the perfect MR response curve. The test resultsare mostly stable under a low operating bias current andmagnetic field, demonstrating its potential application in thefield of information processing with low consumption.

CONCLUSION

In conclusion, a universal method of transferred top electrodeis proposed for the fabrication of a metal-penetration-free

vertical OSV device. Instead of MR overestimation due to theexistence of a dead layer, our device fabricated by a transferredelectrode not only shows a reliable and pure spin-related MRsignal but also can evaluate the effective thickness of theorganic layer more precisely without the dead layercontribution from metal filaments and thermal irradiation.Based on its reliability, a multistate writing and readingprototype was established. Compared to a traditionalmemristor, it is operated by spin-polarized electrodes so thatthe magnetic field can act as a second input line, integratingboth storage and processing operations within the same device.Furthermore, various potential applications in large-areaoperational organic spintronics can be expected due to thesolution processability and simple fabrication process demon-strated in this work.

EXPERIMENTAL SECTIONMaterials. The following materials and solvents were used as

received without further purification: polystyrene from AldrichChemical Co.; poly(styrene sulfonic acid) sodium salt (MW70 000) from Alfa Aesar; regioregular P3HT (rr-P3HT) from TCIChemicals, 1,2-dichlorobenzene (99%, pure) and N-butyl acetate(99+%, extra pure) from Acros Organics, and poly(dimethylsiloxane)(Sylgand 184) from Dow Corning. Other normal chemicals werepurchased from Beijing Chemical Works.

Fabrication of Top Self-Encapsulated Ferromagnetic Elec-trodes. An n-type Si wafer containing 300 nm thick SiO2 was cleanedsuccessively with deionized water, boiled piranha solution (sulfuricacid (98%)/hydrogen peroxide ≥ 3:1), deionized water, and 2-propanol. Then the substrate was dried under a stream of N2 gas. ThePSS (30 mg mL−1 in deionized water) was spin coated at 2500 rpmfor 60 s to form a water-soluble layer. A 1 nm Al layer was depositedon a PSS-coated substrate by thermal evaporation at a deposition rateof about 0.1 Å s−1. An electrically leaky AlOx barrier was obtained bysubsequent oxidation in a stream of oxygen (400 sccm) for 5 min atroom temperature. Co FM layers (10 nm thickness) and a top Au film(60 nm thickness) were grown by e-beam evaporation (pbase = 1 ×10−7 Torr) through a shadow mask at a rate of 0.3 Å s−1. For thefabrication of a support membrane during the transfer process, thePS/n-butyl acetate solution (80 mg mL−1) was spin-coated onto theprepared substrate at a spinning speed of 2000 rpm for 30 s, forming asoft self-encapsulated protection layer on the electrodes.

Device Fabrication of the Organic Spin Valve. The 500-μm-wide LSMO strip electrode with a thickness of 100 nm was fabricatedby a DC facing-target magnetron sputtering method as reportedbefore.15 It is stable in an oxygen atmosphere and water and can berepeatedly utilized after cleaning without apparent degradation.14 A10 mg amount of P3HT powders was dissolved in 1 mL of 1,2-dichlorobenzene and stirred at 80 °C for 30 min to form dispersivesolutions. After filtration through a 0.22 μm syringe filter, 50 μL ofP3HT solution was spin coated onto the SrTiO3/LSMO substrate at5000 rpm for 60 s and subsequently baked at 120 °C for 5 min.Finally, the top FM electrodes were transferred onto the P3HT layerby a PDMS-assisted transfer technique.

PDMS-Assisted Transfer Technique. The self-encapsulated topCo electrodes were mechanically scratched at the border of the PSmembrane by a knife to facilitate the PSS dissolution. For transferholder fabrication, hollow PDMS molds were prepared by mixing baseand curing agent in a 10:1 weight ratio and baking at 70 °C overnight.The PDMS transfer holder was attached onto the PS membrane, andthen the entire substrate was put in water. Once the PSS wasdissolved and the membrane was floated on the water surface, thetransfer process was done by moving the target electrodes onto theorganic semiconductor layer with the assistance of a PDMS holderfollowed by a baking step at 70 °C for 10 min. Next the PDMS moldwas removed by cutting the edge of the PS membrane in the hollowarea and then peeling it off carefully from the substrate with tweezers.

Figure 5. Write (W)−read (R) cycles performed on an LSMO/P3HT/AlOx/Co/Au device with different currents at 2 K. Themultistate nonvolatility was proved by the retained magneticresponse sequence after the application of an initial writingmagnetic field. The gray area represents the initialization of themagnetic field. The corresponding resistance response results (R1to R6) recorded at different writing bias currents (W1 to W6) I =0.01, 0.02, and 0.03 μA are highlighted in red, green, and blueareas, respectively.

ACS Nano Article

DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

12662

Characterizations. Optical microscope images were taken with anOlympus BX51 optical microscope at 10×. Surface topographies wereimaged using a Veeco atomic force microscope operating in tappingmode. Elemental distribution was conducted using a scanningelectron microscope-energy dispersive spectrometer (SEM-EDS,SU8010, Hitachi, Japan) combination. The hysteresis loop measure-ments of FM electrodes were performed using a vibrating samplemagnetometer (Quantum Design, PPMS). The XPS spectra wereacquired using an Axis Ultra spectrometer (Kratos Analytical, UK)with a monochromatic Al Kα source. All the electrical characteristicswere carried out with standard four-probe method in QuantumDesign PPMS with a closed-cycle helium cryostat. The wire-bondingand measurement procedure were reported elsewhere in detail.15

ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsnano.8b07468.

Figures S1−S10 (PDF)

AUTHOR INFORMATIONCorresponding Authors*E-mail: tianyuan@iccas.ac.cn.*E-mail: huwp@iccas.ac.cn.ORCIDWenbo Mi: 0000-0002-9108-9930Hongjun Gao: 0000-0002-6766-0623Daoben Zhu: 0000-0002-6354-940XWenping Hu: 0000-0001-5686-2740NotesThe authors declare no competing financial interest.

ACKNOWLEDGMENTSThe authors are grateful for the financial support from theMinistry of Science and Technology of China(2016YFB0401100, 2017YFA0204503), the National NaturalScience Foundation of China (51725304, 51633006,51703159, 51733004, 91433115), National Program forSupport of Top-notch Young Professionals, and the StrategicPriority Research Program of Chinese Academy of Sciences(Grant No. XDB12000000). The authors acknowledge theLaboratory of Microfabrication, Institute of Physics, CAS, fortheir assistance in electrode fabrication.

REFERENCES(1) Hueso, L. E.; Pruneda, J. M.; Ferrari, V.; Burnell, G.; Valdes-Herrera, J. P.; Simons, B. D.; Littlewood, P. B.; Artacho, E.; Fert, A.;Mathur, N. D. Transformation of Spin Information into LargeElectrical Signals Using Carbon Nanotubes. Nature 2007, 445, 410−413.(2) Dediu, V. A.; Hueso, L. E.; Bergenti, I.; Taliani, C. Spin Routesin Organic Semiconductors. Nat. Mater. 2009, 8, 707−716.(3) Drew, A. J.; Hoppler, J.; Schulz, L.; Pratt, F. L.; Desai, P.; Shakya,P.; Kreouzis, T.; Gillin, W. P.; Suter, A.; Morley, N. A.; Malik, V. K.;Dubroka, A.; Kim, K. W.; Bouyanfif, H.; Bourqui, F.; Bernhard, C.;Scheuermann, R.; Nieuwenhuys, G. J.; Prokscha, T.; Morenzoni, E.Direct Measurement of the Electronic Spin Diffusion Length in aFully Functional Organic Spin Valve by Low-energy Muon SpinRotation. Nat. Mater. 2009, 8, 109−14.(4) Gobbi, M.; Golmar, F.; Llopis, R.; Casanova, F.; Hueso, L. E.Room-temperature Spin Transport in C60-based Spin Valves. Adv.Mater. 2011, 23, 1609−1613.(5) Dieny, B.; Speriosu, V. S.; Gurney, B. A.; Parkin, S. S. P.; Wilhoit,D. R.; Roche, K. P.; Metin, S.; Peterson, D. T.; Nadimi, S. Spin-valve

Effect in Soft Ferromagnetic Sandwiches. J. Magn. Magn. Mater. 1991,93, 101−104.(6) Zutic, I.; Fabian, J.; Sarma, S. D. Spintronics: Fundamentals andApplications. Rev. Mod. Phys. 2004, 76, 323.(7) Sun, X.; Gobbi, M.; Bedoya-Pinto, A.; Txoperena, O.; Golmar,F.; Llopis, R.; Chuvilin, A.; Casanova, F.; Hueso, L. E. Room-temperature Air-stable Spin Transport in Bathocuproine-Based SpinValves. Nat. Commun. 2013, 4, 2794.(8) Jang, H. J.; Richter, C. A. Organic Spin-Valves and Beyond: SpinInjection and Transport in Organic Semiconductors and the Effect ofInterfacial Engineering. Adv. Mater. 2017, 29, No. 1607239.(9) Sun, X.; Bedoya-Pinto, A.; Mao, Z.; Gobbi, M.; Yan, W.; Guo, Y.;Atxabal, A.; Llopis, R.; Yu, G.; Liu, Y.; Chuvilin, A.; Casanova, F.;Hueso, L. E. Active Morphology Control for Concomitant LongDistance Spin Transport and Photoresponse in a Single OrganicDevice. Adv. Mater. 2016, 28, 2609−2615.(10) Krinichnyi, V. I.; Chemerisov, S. D.; Lebedev, Y. S. EPR andCharge-transport Studies of Polyaniline. Phys. Rev. B: Condens. MatterMater. Phys. 1997, 55, 16233−16244.(11) Barraud, C.; Seneor, P.; Mattana, R.; Fusil, S.; Bouzehouane,K.; Deranlot, C.; Graziosi, P.; Hueso, L.; Bergenti, I.; Dediu, V.;Petroff, F.; Fert, A. Unravelling the Role of the Interface for SpinInjection into Organic Semiconductors. Nat. Phys. 2010, 6, 615−620.(12) Li, F.; Li, T.; Chen, F.; Zhang, F. Excellent Spin Transport inSpin Valves Based on the Conjugated Polymer with High CarrierMobility. Sci. Rep. 2015, 5, 9355.(13) Dong, H.; Hu, W. Multilevel Investigation of Charge Transportin Conjugated Polymers. Acc. Chem. Res. 2016, 49, 2435−2443.(14) Xiong, Z.; Wu, D.; Vardeny, Z. V.; Shi, J. Giant Magneto-resistance in Organic Spin-Valves. Nature 2004, 427, 821−824.(15) Ding, S.; Tian, Y.; Li, Y.; Mi, W.; Dong, H.; Zhang, X.; Hu, W.;Zhu, D. Inverse Magnetoresistance in Polymer Spin Valves. ACS Appl.Mater. Interfaces 2017, 9, 15644−15651.(16) Vinzelberg, H.; Schumann, J.; Elefant, D.; Gangineni, R. B.;Thomas, J.; Buchner, B. Low Temperature Tunneling Magneto-resistance on (La,Sr)MnO3/Co Junctions with Organic Spacer Layers.J. Appl. Phys. 2008, 103, No. 093720.(17) Cinchetti, M.; Dediu, V. A.; Hueso, L. E. Activating theMolecular Spinterface. Nat. Mater. 2017, 16, 507−515.(18) Shi, S.; Sun, Z.; Bedoya-Pinto, A.; Graziosi, P.; Li, X.; Liu, X.;Hueso, L.; Dediu, V. A.; Luo, Y.; Fahlman, M. Hybrid Interface Statesand Spin Polarization at Ferromagnetic Metal-Organic Heterojunc-tions: Interface Engineering for Efficient Spin Injection in OrganicSpintronics. Adv. Funct. Mater. 2014, 24, 4812−4821.(19) Sun, D.; Ehrenfreund, E.; Vardeny, Z. V. The First Decade ofOrganic Spintronics Research. Chem. Commun. 2014, 50, 1781−1793.(20) Drew, A. J.; Szulczewski, G.; Nuccio, L.; Gillin, W. P. The Roleof Interfaces in Organic Spin Valves Revealed through Spectroscopicand Transport Measurements. Phys. Status Solidi B 2012, 249, 9−17.(21) Tang, Q.; Jiang, L.; Tong, Y.; Li, H.; Liu, Y.; Wang, Z.; Hu, W.;Liu, Y.; Zhu, D. Micrometer- and Nanometer-Sized Organic Single-Cruystalline Transistors. Adv. Mater. 2008, 20, 2947−2951.(22) Tang, Q.; Tong, Y.; Li, H.; Ji, Z.; Li, L.; Hu, W.; Liu, Y.; Zhu,D. High-Performance Air-Stable Bipolar Field-Effect Transistors ofOrganic Single-Crystalline Ribbons with an Air-Gap Dielectric. Adv.Mater. 2008, 20, 1511−1515.(23) Sidorenko, A. A.; Pernechele, C.; Lupo, P.; Ghidini, M.; Solzi,M.; De Renzi, R.; Bergenti, I.; Graziosi, P.; Dediu, V.; Hueso, L.;Hindmarch, A. T. Interface Effects on an Ultrathin Co Film inMultilayers Based on the Organic Semiconductor Alq3. Appl. Phys.Lett. 2010, 97, 162509.(24) Dediu, V.; Hueso, L. E.; Bergenti, I.; Riminucci, A.; Borgatti, F.;Graziosi, P.; Newby, C.; Casoli, F.; De Jong, M. P.; Taliani, C.; Zhan,Y. Room-temperature Spintronic Effects in Alq3-Based HybridDevices. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 78, 115203.(25) Mooser, S.; Cooper, J. F. K.; Banger, K. K.; Wunderlich, J.;Sirringhaus, H. Spin Injection and Transport in a Solution-processedOrganic Semiconductor at Room Temperature. Phys. Rev. B: Condens.Matter Mater. Phys. 2012, 85, 235202.

ACS Nano Article

DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

12663

(26) Wang, S.; Shi, Y. J.; Lin, L.; Chen, B. B.; Yue, F. J.; Du, J.; Ding,H. F.; Zhang, F. M.; Wu, D. Room-temperature Spin Valve Effects inLa0.67Sr0.33MnO3/Alq3/Co devices. Synth. Met. 2011, 161, 1738−1741.(27) Sun, D.; Yin, L.; Sun, C.; Guo, H.; Gai, Z.; Zhang, X. G.; Ward,T. Z.; Cheng, Z.; Shen, J. Giant Magnetoresistance in Organic SpinValves. Phys. Rev. Lett. 2010, 104, 236602.(28) Zhao, X.; Tong, Y.; Tang, Q.; Liu, Y. Wafer-Scale CoplanarElectrodes for 3D Conformal Organic Single-Crystal Circuits. Adv.Electron. Mater. 2015, 1, 1500239.(29) Liu, Y.; Guo, J.; Zhu, E.; Liao, L.; Lee, S.-J.; Ding, M.; Shakir, I.;Gambin, V.; Huang, Y.; Duan, X. Approaching the Schottky−MottLimit in Vander Waals Metal−semiconductor Junctions. Nature 2018,557, 696−700.(30) Mitsuzuka, T.; Matsuda, K.; Kamijo, A.; Tsuge, H. InterfaceStructures and Magnetoresistance in Magnetic Tunnel Junctions. J.Appl. Phys. 1999, 85, 5807−5809.(31) Tran, T. L. A.; Le, T. Q.; Sanderink, J. G. M.; van der Wiel, W.G.; de Jong, M. P. The Multistep Tunneling Analogue of ConductivityMismatch in Organic Spin Valves. Adv. Funct. Mater. 2012, 22, 1180−1189.(32) Li, F.; Li, T.; Chen, F.; Zhang, F. Spin Injection and Transportin Organic Spin-valves Based on Fullerene C60. Org. Electron. 2014,15, 1657−1663.(33) Devkota, J.; Geng, R.; Subedi, R. C.; Nguyen, T. D. OrganicSpin Valves: A Review. Adv. Funct. Mater. 2016, 26, 3881−3898.(34) Garcia, V.; Bibes, M.; Barthelemy, A.; Bowen, M.; Jacquet, E.;Contour, J. P.; Fert, A. Temperature Dependence of the InterfacialSpin Polarization of La2/3Sr1/3MnO3. Phys. Rev. B: Condens. MatterMater. Phys. 2004, 69, No. 052403.(35) Majumdar, S.; Laiho, R.; Laukkanen, P.; V ayrynen, I. J.;Majumdar, H. S.; Osterbacka, R. Application of RegioregularPolythiophene in Spintronic Devices: Effect of Interface. Appl. Phys.Lett. 2006, 89, 122114.(36) Lu, N.; Li, L.; Banerjee, W.; Sun, P.; Gao, N.; Liu, M. ChargeCarrier Hopping Transport Based on Marcus Theory and Variable-range Hopping Theory in Organic Semiconductors. J. Appl. Phys.2015, 118, No. 045701.(37) Horowitz, G. Organic Field-Effect Transistors. Adv. Mater.1998, 10, 365−377.(38) Grunewald, M.; Gockeritz, R.; Homonnay, N.; Wurthner, F.;Molenkamp, L. W.; Schmidt, G. Vertical Organic Spin Valves inPerpendicular Magnetic Fields. Phys. Rev. B: Condens. Matter Mater.Phys. 2013, 88, No. 085319.

ACS Nano Article

DOI: 10.1021/acsnano.8b07468ACS Nano 2018, 12, 12657−12664

12664