Advances in Organic Multiferroic Junctions

Abstract

Typically, organic multiferroic junctions (OMFJs) are formed of an organic ferroelectric layer sandwiched between two ferromagnetic electrodes. The main scientific interest in OMFJs focuses on the magnetoresistive properties of the magnetic spin valve combined with the electroresistive properties associated with the ferroelectric junction. In consequence, memristive properties that couple magnetoelectric functionalities, which are one of the most active fields of research in material sciences, are opening a large spectrum of technological applications from nonvolatile memory to elements in logic circuits, sensing devices, energy harvesting and biological synapsis models in the emerging area of neuromorphic computing. The realization of these multifunctional electronic elements using organic materials is presenting various advantages related to their low-cost, versatile synthesis and low power consumption functioning for sustainable electronics; green disintegration for transient electronics; and flexibility, light weight and/or biocompatibility for flexible electronics. The purpose of this review is to address the advancement of all OMFJs including not only the achievements in the charge and spin transport through OMFJs together with the effects of electroresistance and magnetoresistance but also the challenges and ways to overcome them for the most used materials for OMFJs.

1. Introduction

Multiferroic tunnel junctions consisting of a ferroelectric barrier in magnetic spin valves were first proposed about 15 years ago [1,2,3] and are capturing considerable scientific interest, especially in the field of electric-field-controlled spintronics and information technologies. As magnetic tunnel junctions, these systems allow for the tunneling spin-polarized current to be efficiently filtered according to the orientation of the magnetization of the ferromagnetic electrodes, which generate high resistive and low resistive states for their antiparallel and parallel relative orientation, respectively, based on the tunnel magnetoresistance (MR) effect [4,5]. The activity of the ferroelectric barrier in a tunnel junction allows for switching between different resistive states defined by the orientation of its electrical polarization that alters the induced charge densities and the electrostatic potential at the barrier–electrode interface and thus the charge transport characteristics of the tunneling current, based on the electroresistance (ER) effect [6,7,8]. In addition, the piezoelectric effect of the ferroelectric that affects the strain at the interface may couple with the magnetostrictive effects of the ferromagnetic layers (Figure 1). This affects the magnetic anisotropy and the magnetization, leading to drastic changes in the magnetoresistive properties, such as the reversal of the magnetoresistance hysteresis induced, for instance, by the electrical polarization [3].

Furthermore, advanced spin and charge transport properties may be added to the multilevel-resistive or memristive states of the multiferroic junctions when the separating layer between the ferromagnetic electrodes of the junctions is an organic material. Organic electronics and spintronics constitute developing fields for the next generation of applications. For organic spintronics, the light molecular elements present weak spin–orbit and hyperfine interactions that lead to a long spin relaxation time and thus long spin transport distances even at room temperature [9,10] that open routes for their inclusion in electronic devices, such as memory and computing elements to store and process information, neuromorphic computing and models of synaptic and neuronal operations, etc. Moreover, important assets for technological applications of organic systems are the low-cost and versatile fabrication, sustainability, light weight, mechanical flexibility and the possibility of a controlled degradation for electronic waste management [11,12].

The purpose of this article is to review the major advances in OMFJs through the analysis of mechanisms, functioning and applications. It is envisaged not only to merge current scientific knowledge but also to identify new paths for research and innovation in this field. This report contains a detailed description of the main characteristics of OMFJs in Section 2 that comprise the properties of organic spin valves and organic ferroelectric junctions, respectively. This section includes the origins of the functional properties of OMFJs, i.e., the magnetoresistance and the electroresistance effects and the control of the multilevel-resistive properties with external electric and magnetic fields. Section 2.1 describes the mechanism of the spin and charge transport through organic spin valves and summarizes magnetic and organic materials that are mostly used in spin valves and can be potentially used in OMFJ junctions. Furthermore, Section 2.2 summarizes the organic ferroelectric junctions and their properties, together with a description of organic ferroelectrics. Section 3 describes the reported OMFJs, their properties that are based on the coupling of magnetoresistive and electroresistive effects and their optimal configuration for adjusted functional devices. This review also includes challenges regarding the operational processes of OMFJs, together with proposed solutions to overcome these difficulties. Finally, concluding remarks on the progress strategies in OMFJs and a general perspective are presented in Section 4.

2. Characteristics of Organic Spin Valves and Organic Ferroelectric Junctions

2.1. Magnetic Junctions and Organic Spin Valves

Magnetic junctions are originally formed of two ferromagnetic films, preferably with different magnetic coercive fields, separated by a non-magnetic spacer that can be a tunneling barrier or a spin transport medium. In magnetic tunnel junctions, the tunneling current between metallic ferromagnetic films is spin-dependent and generates a tunneling MR that is based on the relative orientation of the magnetization of the two ferromagnetic electrodes [4]. The MR ratio is defined by the Julliere expression as follows:

$$\mathrm{M}\mathrm{R}=\frac{R_{ap}-R_p}{R_p}100 \left(\%\right)=\frac{2P_1{P}_2}{1+P_1{P}_2}100(\%)$$

(1)

where R_ap and R_p are resistances for the antiparallel and parallel orientations of the magnetizations, respectively, and P₁, P₂ are the spin polarizability of the two ferromagnetic electrodes. Equivalently, the MR is defined for spin valves where spin-polarized currents are transported through the separating medium. Usually, the MR ratio is the maximum for an antiparallel magnetic configuration and the minimum for the parallel configuration (Figure 2a). However, these values can appear reversed, leading to negative MR values, observed for different junctions (Figure 2b).

Table 1. Organic spin valves. Abbreviations of molecule names are as follows: PC71BM—phenyl-C71-butyric-acid-metyl-ester; TCNE—tetracyanoethylene; BCP—buthocuproine; TPD—N,N-bis(3-methylphenyl)-N,N-diphenylbenzidine; BTQBT—bis(l,2,5-thiadiazolo)-P-quinobis(l,3-dithiole); TIPS-pentacene—6,13-bis(triisopropylsilylethynyl)-pentacene; DOO-PPV—poly(dioctyloxy)phenylenevinylene; Alq₃—tris(8-hydroxyquinolinato)aluminum; Gaq₃—tris-(8-hydroxyquinoline)gallium; IC₁₂H₂(PO₃Et₂)₂—diethyl(11-iodoundecyl)phosphonate; Tb[Pc(C₁₂OPO₃Et₂)]₂—terbium bis-phthalocyaninato-diethyl-phosphonate; NitPO—4-methylbenzylphosphonate)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide; P3HT—poly(3-hexylthiophene); F₄TCNQ—2,3,5,6-Tetrafluoro-7,7,8,8-tetracyanoquinodimethane; P3MT—poly(3-methylthiophene); T₆—sexithienyl; CVB—4,4′-bis99-(ethyl-3-carbazovinylene)-1,1′-bipheny; C8-BTBT—2,7-dioctyl [1]benzothieno[3,2-b][1]benzothiophene; P(NDI2OD-T2)—poly{[N,N′-bis(2-octyldodecyl)-1,4,5,8-naphthalenedicarboximide)-2,6-diyl]-alt-5,5′-(2,2′-bithio-phene)}; F8BT—poly(9,9-dioctylfluorene-co-benzothidiazole); Py-Np—4,4′,4″,4′′′-(pyrene-1,3,6,8-tetrayl)tetraaniline-naphthalene-2,6-dicarbaldehyde; TPP—tetraphenyl porphyrin; H₂Pc—hydrogen-phthalocyanine; CuPc—copper-phthalocyanine; FePc—iron-phthalocyanine; F₁₆CuPc—fluorinated copper-phthalocyanine; and BDMT—1,4 benzenedimethanethiol.

Organic Material	System (Thickness)	MR/Temp. (K)	Ref.
Fullerene	LSMO(50 nm)/C60(120 nm)/Co(15 nm)	−13.3%/20 K	[13]
	LSMO(50 nm)/C60(180 nm)/Co(15 nm)	~−7%/20 K	[13]
	LSMO(50 nm)/C60(180 nm)/Co(15 nm)	~−4.5%/120 K	[13]
	LSMO(50 nm)/C60(180 nm)/Co(15 nm)	~−1%/290 K	[13]
	Co(15 nm)/AlOx(1 nm)/C60(5 nm)/Py(20 nm)	9%/300 K	[14]
	Co(15 nm)/AlOx(1 nm)/C60(28 nm)/Py(20 nm)	5.5%/300 K	[14]
	Co(15 nm)/AlOx(1 nm)/C60(18 nm)/Py(20 nm)	~14%/80 K	[14]
	Fe3O4/AlOx/C60(80 nm)/Co(10 nm)	6.9%/150 K	[15]
	Fe3O4/AlOx/C60(80 nm)/Co(10 nm)	5.3%/300 K	[15]
	Fe3O4/AlOx/C60(110 nm)/Co(10 nm)	~2.2%/150 K	[15]
	Fe3O4/AlOx/C60(110 nm)/Co(10 nm)	~0.5%/300 K	[15]
	LSMO(50 nm)/C70(120 nm)/Co(15 nm)	−8.9%/20 K	[13]
	LSMO(50 nm)/C70(180 nm)/Co(15 nm)	~−6%/20 K	[13]
	LSMO(50 nm)/C70(180 nm)/Co(15 nm)	~−4%/120 K	[13]
	LSMO(50 nm)/C70(180 nm)/Co(15 nm)	~−0.8%/290 K	[13]
PC71BM	Co(20 nm)/LiF(1.4 nm)/PC71BM(45 nm)/Py(12 nm)	~6.2%/300 K	[16]
	Co(20 nm)/LiF(1.5 nm)/PC71BM(45 nm)/Py(12 nm)	~1.3%/300 K	[16]
	Co(20 nm)/LiF(1.6 nm)/PC71BM(45 nm)/Py(12 nm)	~−0.6%/300 K	[16]
	Co(20 nm)/LiF(1.8 nm)/PC71BM(45 nm)/Py(12 nm)	~−2.4%/300 K	[16]
	Co(20 nm)/LiF(2 nm)/PC71BM(45 nm)/Py(12 nm)	~−1.2%/300 K	[16]
Rubrene	Fe3O4(100 nm)/AlOx(2 nm)/rubrene(2 nm)/Co(10 nm)	6%/300 K	[17]
	Fe3O4(100 nm)/AlOx(2 nm)/rubrene(2 nm)/Co(10 nm)	11%/150 K	[17]
	Fe3O4(100 nm)/AlOx(2 nm)/rubrene(6 nm)/Co(10 nm)	3.3%/300 K	[17]
	Fe3O4(100 nm)/AlOx(2 nm)/rubrene(6 nm)/Co(10 nm)	9%/150 K	[17]
	Co(8 nm)/Al2O3(0.5 nm)/rubrene(4.6 nm)/Fe(10 nm)/CoO(1.5 nm)	6%/295 K	[18]
	Co(8 nm)/Al2O3(0.5 nm)/rubrene(4.6 nm)/Fe(10 nm)/CoO(1.5 nm)	13%/80 K	[18]
	Co(8 nm)/Al2O3(0.5 nm)/rubrene(4.6 nm)/Fe(10 nm)/CoO(1.5 nm)	16%/4.2 K	[18]
	Co(15 nm)/AlOx(2.5 nm)/rubrene(5 nm)/Fe(15 nm)	~−0.058%/100 K	[19]
	Co(15 nm)/AlOx(2.5 nm)/rubrene(10 nm)/Fe(15 nm)	~−0.058%/100 K	[19]
	Co(15 nm)/AlOx(2.5 nm)/rubrene(5 nm)/Fe(15 nm)	~−0.025%/300 K	[19]
	Co(15 nm)/AlOx(2.5 nm)/rubrene(10 nm)/Fe(15 nm)	~−0.035%/300 K	[19]
	LSMO(50 nm)/LAO(1.2 nm)/rubrene(5 nm)/Fe(30 nm)	~12.5%/10 K	[20]
	LSMO(50 nm)/LAO(1.2 nm)/rubrene(5 nm)/Fe(30 nm)	~3.5%/150 K	[20]
	LSMO(50 nm)/LAO(1.2 nm)/rubrene(5 nm)/Fe(30 nm)	~0.1%/250 K	[20]
	LSMO(50 nm)/LAO(1.2 nm)/rubrene(20 nm)/Fe(30 nm)	~0.2%/10 K	[20]
Rubrene/TCNE	Fe(50 nm)/rubrene(10 nm)/V[TCNE]x(300 nm)	−0.06%/100 K	[21]
	Fe(50 nm)/rubrene(10 nm)/V[TCNE]x(300 nm)	−0.02%/200 K	[21]
	Fe(50 nm)/rubrene(10 nm)/V[TCNE]x(300 nm)	0.01%/300 K	[21]
	Al(50 nm)/V[TCNE]x(50 nm)/rubrene(10 nm)/V[TCNE]x(300 nm)/Al(30 nm)	−0.035%/100 K	[22]
	Al(50 nm)/V[TCNE]x(50 nm)/rubrene(10 nm)/V[TCNE]x(300 nm)/Al(30 nm)	−0.01%/200 K	[22]
BCP	Co(11 nm)/AlOx(1.5 nm)/BCP(5 nm)/Py(11 nm)	~4.5%/300 K	[23]
	Co(11 nm)/AlOx(1.5 nm)/BCP(10 nm)/Py(11 nm)	~4%/300 K	[23]
	Co(11 nm)/AlOx(1.5 nm)/BCP(30 nm)/Py(11 nm)	~4%/300 K	[23]
	Co(11 nm)/AlOx(1.5 nm)/BCP(10 nm)/Py(11 nm) bending/in air	~3%/300 K	[24]
TPD	Co2MnSi(20 nm)/TPD(200 nm)/Co(7 nm)	7.8%/300 K	[25]
	Co2MnSi(20 nm)/TPD(200 nm)/Co(7 nm)	~11%/150 K	[25]
	Co2MnSi(20 nm)/TPD(200 nm)/Co(7 nm)	10.7%/5 K	[25]
	LSMO/TPD(200 nm)/Co(7 nm)	~1.5%/250 K	[25]
	LSMO/TPD(200 nm)/Co(7 nm)	~5%/150 K	[25]
	LSMO/TPD(200 nm)/Co(7 nm)	19%/5 K	[25]
pentacene	LSMO/pentacene(200 nm)/LSMO	~5.5%/5.3 K	[26]
	LSMO/pentacene(30 nm)/LSMO	~2%/9 K	[27]
BTQBT	LSMO/BTQBT(200 nm)/LSMO	~8%/10 K	[26]
	LSMO/BTQBT(50 nm)/LSMO	~28%/9.1 K	[26]
TIPS-Pentacene	CoPt(10 nm)/TIPS-pentacene(75 nm)/AlOx(2 nm)/Co(10 nm)	−0.08%/300 K	[28]
	CoPt(10 nm)/TIPS-pentacene(40 nm)/AlOx(2 nm)/Co(10 nm)	~−0.45%/300 K	[28]
	CoPt(10 nm)/TIPS-pentacene(40 nm)/AlOx(2 nm)/Co(10 nm)	~−0.05%/175 K	[28]
	CoPt(10 nm)/TIPS-pentacene(100 nm)/AlOx(2 nm)/Co(10 nm)	~−0.05%/300 K	[28]
	LSMO/TIPS-pentacene single crystal(269 nm)/Co(10 nm)	~−17%/30 K	[28]
	LSMO/TIPS-pentacene single crystal(457 nm)/Co(10 nm)	~−1.5%/30 K	[29]
	LSMO/TIPS-pentacene single crystal(269 nm)/Co(10 nm)	~−2.8%/100 K	[29]
	LSMO/TIPS-pentacene polycryst. film(97 nm)/Co(10 nm)	~−6%/30 K	[29]
	LSMO/TIPS-pentacene polycryst. film(97 nm)/Co(10 nm)	~−3.5%/100 K	[29]
DOO-PPV	LSMO(200 nm)/Hydrogen-DOO-PPV(25 nm)/Co(15 nm)	2%/10 K	[30]
	LSMO(200 nm)/Deuterium-DOO-PPV(25 nm)/Co(15 nm)	~40%/10 K	[30]
	LSMO(200 nm)/H/D-DOO-PPV(25 nm)/Co(15 nm)	~5%/10 K	[30]
	LSMO(200 nm)/H/D-DOO-PPV(25 nm)/Co(15 nm)	~2.5%/150 K	[30]
	LSMO(200 nm)/H/D-DOO-PPV(25 nm)/Co(15 nm)	~0.2%/300 K	[30]
	LSMO(200 nm)/H-DOO-PPV(55 nm)/Co(15 nm)	~0.2%/10 K	[30]
	LSMO(200 nm)/D-DOO-PPV(55 nm)/Co(15 nm)	~2.5%/10 K	[30]
Alq3	LSMO(100 nm)/Alq3(130 nm)/Co(3.5 nm)	−40%/11 K	[31]
	LSMO(100 nm)/Alq3(160 nm)/Co(3.5 nm)	~−14%/11 K	[31]
	LSMO(100 nm)/Alq3(200 nm)/Co(3.5 nm)	~−7%/11 K	[31]
	LSMO(100 nm)/Alq3(250 nm)/Co(3.5 nm)	~−2%/11 K	[31]
	LSMO(100 nm)/Alq3(160 nm)/Co(3.5 nm)	~−2%/130 K	[31]
	LSMO(100 nm)/Alq3(160 nm)/Co(3.5 nm)	~−0.3%/230 K	[31]
	Co(8 nm)/Al2O3(0.6 nm)/Alq3(1.6 nm)/Py(10 nm)	4.6%/300 K	[32]
	Co(8 nm)/Al2O3(0.6 nm)/Alq3(1.6 nm)/Py(10 nm)	6.8%/77 K	[32]
	Co(8 nm)/Al2O3(0.6 nm)/Alq3(1.6 nm)/Py(10 nm)	7.5%/4.2 K	[32]
	LSMO(15–20 nm)/Alq3 (300 nm)/Al2O3 (2 nm)/Co(35 nm)	~−0.1%/20 K	[33]
	LSMO(15–20 nm)/Alq3 (200 nm)/Al2O3 (2 nm)/Co(35 nm)	~−5%/20 K	[33]
	LSMO(15–20 nm)/Alq3 (100 nm)/Al2O3 (2 nm)/Co(35 nm)	−11%/20 K	[33]
	LSMO(15–20 nm)/Alq3 (100 nm)/Al2O3 (2 nm)/Co(35 nm)	−6%/100 K	[33]
	LSMO(15–20 nm)/Alq3 (100 nm)/Al2O3 (2 nm)/Co(35 nm)	−0.15%/300 K	[33]
	LSMO/Alq3 (23 nm)/Co-nanodots/Co(7 nm)	−1%/10 K	[34]
	LSMO/Alq3 (67 nm)/Co-nanodots/Co(7 nm)	−7%/10 K	[34]
	LSMO/Alq3 (93 nm)/Co-nanodots/Co(7 nm)	−300%/10 K	[34]
	LSMO/Alq3 (135 nm)/Co-nanodots/Co(7 nm)	−13%/10 K	[34]
	LSMO/Alq3 (93 nm)/Co(7 nm)	−35%/10 K	[34]
	LSMO/Alq3 (135 nm)/Co(7 nm)	−4%10 K	[34]
	LSMO/Alq3 (40 nm)/Co(20 nm)	~−18%/100 K	[35]
	LSMO/Alq3 (40 nm)/Co(20 nm)	~−10%/150 K	[35]
	LSMO/Alq3 (40 nm)/Co(20 nm)	−0.07%/300 K	[35]
	LPCMO(60 nm)/Alq3(60 nm)/Co(10 nm)	100–440%(1−7T)/10 K	[36]
	LPCMO(60 nm)/Alq3(60 nm)/Co(10 nm)	~200%(7T)/25 K	[36]
	LPCMO(60 nm)/Alq3(60 nm)/Co(10 nm)	~100%(7T)/75 K	[36]
	LPCMO(60 nm)/Alq3(60 nm)/Co(10 nm)	~1%(7T)/150 K	[36]
	LSMO/Alq3(60 nm)/Co(10 nm)	−37%/10 K	[36]
	SFMO(150 nm)/Alq3(45 nm)/Co(16 nm)	30%/10 K	[37]
	SFMO(150 nm)/Alq3(45 nm)/Co(16 nm)	25%/100 K	[37]
	SFMO(150 nm)/Alq3(45 nm)/Co(16 nm)	10%/200 K	[37]
	SFMO(150 nm)/Alq3(45 nm)/Co(16 nm)	2%/300 K	[37]
	Fe3O4(110 nm)/AlOx(2 nm)/Alq3(2 nm)/Co(10 nm)	~12.5%/150 K	[38]
	Fe3O4(110 nm)/AlOx(2 nm)/Alq3(2 nm)/Co(10 nm)	~6%/300 K	[38]
	Fe3O4(110 nm)/AlOx(2 nm)/Alq3(10 nm)/Co(10 nm)	~4%/150 K	[38]
	Fe3O4(110 nm)/AlOx(2 nm)/Alq3(10 nm)/Co(10 nm)	~2%/300 K	[38]
	Fe3O4(110 nm)/AlOx(2 nm)/Alq3(20 nm)/Co(10 nm)	~1%/150 K	[38]
	Fe3O4(110 nm)/AlOx(2 nm)/Alq3(20 nm)/Co(10 nm)	~0.5%/300 K	[38]
TPD/Alq3	FeCo(17 nm)/TPD(50 nm)/Alq3(200 nm)/LiF(1.9 nm)/NiFe(17 nm)	~0.3%/40 K	[39]
Gaq3	LSMO(20 nm)/Gaq3 (10–15 nm)/AlOx(2 nm)/Co(7 nm) electrically controlled-high-resistive state (oxygen migration)	−2%/100 K	[40]
	LSMO(20 nm)/Gaq3 (10–15 nm)/AlOx(2 nm)/Co(7 nm) electrically controlled-high-resistive state (oxygen migration)	−8%/100 K	[40]
IC12H2(PO3Et2)2/ Gaq3	LSMO/IC12H2(PO3Et2)2(ML)/Gaq3(40 nm)/AlOx(2 nm)/Co	~−1.9%/200 K	[41]
	LSMO/IC12H2(PO3Et2)2(ML)/Gaq3(40 nm)/ AlOx(2 nm)/Co	~−9.8%/100 K	[41]
	LSMO/IC12H2(PO3Et2)2(ML)/Gaq3(40 nm)/ AlOx(2 nm)/Co	~−18%/10 K	[41]
Tb[Pc(C12OPO3Et2)]2 /Gaq3	LSMO/Tb[Pc(C12OPO3Et2)]2(ML)/Gaq3(40 nm)/AlOx(2 nm)/Co	~−2%/200 K	[41]
	LSMO/Tb[Pc(C12OPO3Et2)]2(ML)/Gaq3(40 nm)/AlOx(2 nm)/Co	~−9.3%/100 K	[41]
	LSMO/Tb[Pc(C12OPO3Et2)]2(ML)/Gaq3(40 nm)/AlOx(2 nm)/Co	~−13%/100 K	[41]
NitPO/Gaq3	LSMO(15 nm)/NitPO(ML)/Gaq3(150 nm)/AlOx(2 nm)/Co	−4.34%/100 K	[42]
	LSMO(15 nm)/NitPO(ML)/Gaq3(150 nm)/AlOx(2 nm)/Co	−7.1%/3 K	[42]
P3HT	Fe50Co50(20 nm)/P3HT(75 nm)/Ni81Fe19(20 nm)	0.1%/300 K	[43]
	Fe50Co50(40 nm)/P3HT(150 nm)/Ni81Fe19(20 nm)	−0.04%/300 K	[43]
	Ni81Fe19(12 nm)/P3HT(30 nm)/AlOx(1 nm)/Co(10 nm)	~1%/10 K	[44]
	Fe3O4(80 nm)/P3HT(12 nm)/AlOx(1 nm)/Co(12 nm)	~2.3%/100 K	[45]
	Fe3O4(80 nm)/P3HT(12 nm)/AlOx(1 nm)/Co(12 nm)	~0.75%/200 K	[45]
	Fe3O4(80 nm)/P3HT(25 nm)/AlOx(1 nm)/Co(12 nm)	~1.75%/100 K	[45]
	LSMO/P3HT(100 nm)/Co(10 nm)	80%/5 K	[46]
	LSMO/P3HT(100 nm)/Co(10 nm)	1.5%/300 K	[46]
	LSMO(100 nm)/P3HT(80 nm)/AlOx(1 nm)/Co(10 nm)	−0.2%/2 K 15.6%/2 K	[47]
	LSMO/P3HTannealed(45 nm)/Co	~15%/20 K	[48]
	LSMO/P3HTannealed(72 nm)/Co	~11%/20 K	[48]
	LSMO/P3HTannealed(103 nm)/Co	~8%/20 K	[48]
	LSMO/P3HTannealed(175 nm)/Co	~4%/20 K	[48]
	LSMO/P3HT(45 nm)/Co	~12%/20 K	[48]
	LSMO/P3HT(72 nm)/Co	~6%/20 K	[48]
	LSMO/P3HT(103 nm)/Co	~3%/20 K	[48]
	LSMO/P3HT(175 nm)/Co	~2%/20 K	[48]
	LPCMO(60 nm)/P3HT(30 nm)/AlOx/Co(10 nm)	93%/30 K	[49]
	LPCMO(60 nm)/P3HT(30 nm)/AlOx/Co(10 nm)	~53%/50 K	[49]
	LPCMO(60 nm)/P3HT(30 nm)/AlOx/Co(10 nm)	~30%/75 K	[49]
	LSMO(100 nm)/P3HT(20 nm)/AlOx(1 nm)/Co(10 nm)	~7.5%/2 K	[50]
	LSMO(100 nm)/P3HT(20 nm)/AlOx(1 nm)/Co(10 nm)	~4%/5 K	[50]
	LSMO(100 nm)/P3HT(20 nm)/AlOx(1 nm)/Co(10 nm)	~3%/10 K	[50]
	LSMO(100 nm)/P3HT(20 nm)/AlOx(1 nm)/Co(10 nm)	~0.1%/30 K	[50]
P3HT/F4-TCNQ	LSMO(100 nm)/P3HT(40 nm)/F4TCNQ dopant/Co(15 nm)	~19%/2 K	[51]
	LSMO(100 nm)/P3HT(40 nm)/F4TCNQ dopant/Co(15 nm)	~16%/5 K	[51]
	LSMO(100 nm)/P3HT(40 nm)/F4TCNQ dopant/Co(15 nm)	~14%/10 K	[51]
	LSMO(100 nm)/P3HT(40 nm)/F4TCNQ dopant/Co(15 nm)	~13%/15 K	[51]
	LSMO(100 nm)/P3HT(40 nm)/F4TCNQ dopant/Co(15 nm)	~11%/20 K	[51]
P3MT	LSMO/P3MT(15 nm)/Co(15 nm)	~−43%/20 K	[52]
	LSMO/P3MT(15 nm)/Co(15 nm)	~−0.3%/280 K	[52]
T6	LSMO/T6(70–140 nm)/LSMO	15–30%/300 K	[9]
	LSMO/T6(200 nm)/LSMO	7–10%/300 K	[9]
CVB	LSMO/CVB(100 nm)/Co	~−15%/14 K	[53]
	LSMO/CVB(100 nm)/Co	~−3.5%/140 K	[53]
C8-BTBT	Ni78Fe22/C8-BTBT(2 nm)/Ni78Fe22	~−0.3%/300 K	[54]
	Ni78Fe22/C8-BTBT(4 nm)/Ni78Fe22	~0.5%/300 K	[54]
	Py/C8-BTBT(1ML)/Co	~−1%/10 K	[55]
	Py/C8-BTBT(2ML)/Co	~−12%/10 K	[55]
P(NDI2OD-T2)	LSMO(100 nm)/P(NDI2OD-T2)(35 nm)/AlOx(1.5 nm)/Co(10 nm)	~30%/4.2 K	[56]
	LSMO(100 nm)/P(NDI2OD-T2)(35 nm)/AlOx(1.5 nm)/Co(10 nm)	~18%/50 K	[56]
	LSMO(100 nm)/P(NDI2OD-T2)(35 nm)/AlOx(1.5 nm)/Co(10 nm)	~5%/150 K	[56]
	LSMO(100 nm)/P(NDI2OD-T2)(35 nm)/AlOx(1.5 nm)/Co(10 nm)	~1%/250 K	[56]
	LSMO(100 nm)/P(NDI2OD-T2)(100 nm)/AlOx(1.5 nm)/Co(10 nm)	~10%/4.2 K	[56]
	LSMO(100 nm)/P(NDI2OD-T2)(100 nm)/AlOx(1.5 nm)/Co(10 nm)	~5%/50 K	[56]
	LSMO(100 nm)/P(NDI2OD-T2)(100 nm)/AlOx(1.5 nm)/Co(10 nm)	~2%/150 K	[56]
PIID-CNTVT-C1: X=C PAIID-CNTVT-C1: X=N	LSMO(50 nm)/PIID-CNTVT-C1(40 nm)/Py(10 nm)	~19%/50 K	[57]
	LSMO(50 nm)/PIID-CNTVT-C1(40 nm)/Py(10 nm)	~14%/100 K	[57]
	LSMO(50 nm)/PIID-CNTVT-C1(40 nm)/Py(10 nm)	~8%/150 K	[57]
	LSMO(50 nm)/PIID-CNTVT-C1(40 nm)/Py(10 nm)	~3%/200 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C1(40 nm)/Py(10 nm)	~25%/50 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C1(40 nm)/Py(10 nm)	~17%/100 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C1(40 nm)/Py(10 nm)	~10%/150 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C1(40 nm)/Py(10 nm)	~5%/200 K	[57]
PIID-CNTVT-C3: X=C-H PAIID-CNTVT-C3: X=N	LSMO(50 nm)/PIID-CNTVT-C3(40 nm)/Py(10 nm)	~16%/50 K	[57]
	LSMO(50 nm)/PIID-CNTVT-C3(40 nm)/Py(10 nm)	~12%/100 K	[57]
	LSMO(50 nm)/PIID-CNTVT-C3(40 nm)/Py(10 nm)	~6%/150 K	[57]
	LSMO(50 nm)/PIID-CNTVT-C3(40 nm)/Py(10 nm)	~2%/200 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C3(40 nm)/Py(10 nm)	~23%/50 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C3(40 nm)/Py(10 nm)	~15%/100 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C3(40 nm)/Py(10 nm)	~10%/150 K	[57]
	LSMO(50 nm)/PAIID-CNTVT-C3(40 nm)/Py(10 nm)	~5%/200 K	[57]
PIID-TVTCN: X=C-H PAIID-TVTCN: X=N PFIID-TVTCN: X=C-F	LSMO(30 nm)/PIID-TVTCN(25 nm)/Co(10 nm)	~−23%/50 K	[58]
	LSMO(30 nm)/PIID-TVTCN(25 nm)/Co(10 nm)	~−8%/200 K	[58]
	LSMO(30 nm)/PAIID-TVTCN(25 nm)/Co(10 nm)	~−10%/50 K	[58]
	LSMO(30 nm)/PAIID-TVTCN(25 nm)/Co(10 nm)	~−3%/200 K	[58]
	LSMO(30 nm)/PFIID-TVTCN(25 nm)/Co(10 nm)	~−16%/50 K	[58]
	LSMO(30 nm)/PFIID-TVTCN(25 nm)/Co(10 nm)	~−5%/200 K	[58]
	LSMO(30 nm)/PIID-TVTCN(25 nm)/AlOx(2 nm)/Co(10 nm)	~10%/50 K	[58]
	LSMO(30 nm)/PIID-TVTCN(25 nm)/AlOx(2 nm)/Co(10 nm)	~4%/200 K	[58]
	LSMO(30 nm)/PAIID-TVTCN(25 nm)/AlOx(2 nm)/Co(10 nm)	~17%/50 K	[58]
	LSMO(30 nm)/PAIID-TVTCN(25 nm)/AlOx(2 nm)/Co(10 nm)	~5%/200 K	[58]
	LSMO(30 nm)/PFIID-TVTCN(25 nm)/AlOx(2 nm)/Co(10 nm)	~10%/50 K	[58]
	LSMO(30 nm)/PFIID-TVTCN(25 nm)/AlOx(2 nm)/Co(10 nm)	~3%/200 K	[58]
PNVT-8: X=H PNVT-CN-8:X=CN	NiFe(12 nm)/PNVT-8(15 nm)/Co(12 nm)	~0.35%/10 K	[59]
	NiFe(12 nm)/PNVT-8(15 nm)/Co(12 nm)	~0.18/150 K	[59]
	NiFe(12 nm)/PNVT-8(15 nm)/Co(12 nm)	~0.12/300 K	[59]
	NiFe(12 nm)/PNVT-CN-8(15 nm)/Co(12 nm)	~0.22/10 K	[59]
	NiFe(12 nm)/PNVT-CN-8(15 nm)/Co(12 nm)	~0.1%/150 K	[59]
	NiFe(12 nm)/PNVT-CN-8(15 nm)/Co(12 nm)	~0.08%/300 K	[59]
	NiFe(12 nm)/Au(3 nm)/PNVT-CN-8(25 nm)/Co(12 nm)	~−0.1%/100 K	[60]
	NiFe(12 nm)/Au(3 nm)/PNVT-CN-8(25 nm)/Co(12 nm)	~−0.08%/300 K	[60]
PNVT-10: X=H PNVT-CN-10:X=CN	NiFe(12 nm)/PNVT-10(15 nm)/Co(12 nm)	~0.19%/10 K	[59]
	NiFe(12 nm)/PNVT-10(15 nm)/Co(12 nm)	~0.07/150 K	[59]
	NiFe(12 nm)/PNVT-10(15 nm)/Co(12 nm)	~0.02/300 K	[59]
	NiFe(12 nm)/PNVT-CN-10(15 nm)/Co(12 nm)	~0.13%/10 K	[59]
	NiFe(12 nm)/PNVT-CN-10(15 nm)/Co(12 nm)	~0.05/150 K	[59]
	NiFe(12 nm)/PNVT-CN-10(15 nm)/Co(12 nm)	~0.02/200 K	[59]
	NiFe(12 nm)/Au(3 nm)/PNVT-CN-10(25 nm)/Co(12 nm)	~0.05%/10 K	[60]
	NiFe(12 nm)/Au(3 nm)/PNVT-CN-10(25 nm)/Co(12 nm)	~−0.03%/100 K	[60]
F8BT	LSMO(20 nm)/F8BT(45 nm)/MoOx(3 nm)/Co(20 nm) magneto-electroluminescence	~9%/20 K ~2.4%/20 K	[61]
	LSMO(20 nm)/F8BT(45 nm)/MoOx(3 nm)/Co(20 nm)	~7%/30 K	[61]
	LSMO(20 nm)/F8BT(45 nm)/MoOx(3 nm)/Co(20 nm) magneto-electroluminescence	~3%/100 K ~0.4%/100 K	[61]
	LSMO(20 nm)/F8BT(45 nm)/MoOx(3 nm)/Co(20 nm)	~1%/150 K	[61]
Py-Np	LSMO/Py-Np(90 nm)/Co(20 nm)	~−27%/30 K	[62]
	LSMO/Py-Np(90 nm)/Co(20 nm)	~−19%/100 K	[62]
	LSMO/Py-Np(90 nm)/Co(20 nm)	~−9%/150 K	[62]
	LSMO/Py-Np(90 nm)/Co(20 nm)	~−2%/200 K	[62]
TPP	LSMO(50 nm)/TPP(20 nm)/Co(5 nm)	~−15%/80 K	[63]
	LSMO(50 nm)/TPP(20 nm)/Co(5 nm)	~−6%/80 K	[63]
	LSMO(50 nm)/TPP(20 nm)/Co(5 nm)	~−3%/80 K	[63]
H2Pc	Co(12 nm)/AlOx(1.5 nm)/H2Pc(90 nm)/Py spin-photovoltaic-resistive levels control with light intensity	~7%/300 K	[64]
CuPc	Fe(25 nm)/CuPc(100 nm)/Co(5 nm)	6.4%/40 K	[65]
	Fe(25 nm)/CuPc(100 nm)/Co(5 nm)	3.2%/80 K	[65]
	Fe(25 nm)/CuPc(100 nm)/Co(5 nm)	1.8%/120 K	[65]
	LSMO/CuPc(110 nm)/Co	~−6%/10 K	[66]
	LSMO/CuPc(110 nm)/Co	~−4.5%/110 K	[66]
	LSMO/CuPc(110 nm)/Co	~−1%/300 K	[66]
	LSMO/CuPc(50 nm)/Co	~−20%/10 K	[66]
	LSMO/CuPc(160 nm)/Co	~−2.5%/10 K	[66]
	LSMO/CuPc(200 nm)/Co	~−0.5%/10 K	[66]
FePc	Co/FePc(50 nm)/Co	4%/10 K	[67]
	Co/FePc(50 nm)/Co	1.2%/50 K	[67]
	Co/FePc(50 nm)/Co	0.5%/100 K	[67]
F16CuPc	Co/AlOx/F16CuPc(15 nm)/Py	~4%/300 K	[68]
	Co/AlOx/F16CuPc(90 nm)/Py Spin-photovoltaic (vacuum or air)	~4%/295 K −30/−25%/295 K	[68]
BDMT	NiFe(15 nm)/FeCo(10 nm)/BDMT(1 nm)/AlOx(1 nm)FeCo(35 nm)	2.4%/20 K	[69]
	NiFe(15 nm)/FeCo(10 nm)/BDMT(1 nm)/AlOx(1 nm)/FeCo(35 nm)	1.9%/100 K	[69]
	NiFe(15 nm)/FeCo(10 nm)/BDMT(1 nm)/AlOx(1 nm)/FeCo(35 nm)	0.25%/300 K	[69]

presents most of the results encountered for magnetic junctions with an organic semiconductor spacer, i.e., organic spin valve junctions, which are composed of different materials, including small molecules that contain acenes, thiophenes, fullerenes, metal complexes and their derivatives, but also larger molecules and polymers. This table includes the chemical representation of each organic layer, the junction sequence mentioning the thickness of each system and the MR at various temperatures extracted from the corresponding references. The organic films of the junctions were deposited by physical vapor deposition techniques or wet chemical routes. The MR strongly depends on the nature of the used materials, the thickness of the organic spacer and the temperature.

The variation in the MR values of organic spin valves is related to several parameters, such as the following:

• The spin polarization of the ferromagnetic electrodes;
• Tunneling/Spin injection from the ferromagnetic electrode into the organic layer;
• Spin transport through the organic layer;
• Spin detection at the other electrode.

Therefore, the MR of the organic spin valves can also be expressed by adapting the Julliere formula as follows:

$$\mathrm{M}\mathrm{R}=\frac{2P_1{P}_2{e}^{-d/{\lambda }_s}}{1+P_1{P}_2{e}^{-d/{\lambda }_s}}100(\%)$$

(2)

where d is usually the thickness of the spin transport medium and λ_s the spin diffusion length.

For efficient spin valves with high MR, the junctions preferably contain ferromagnetic electrodes with a high spin polarization. In particular, from oxides with the perovskite-based crystal structure, lanthanum strontium manganite with the general formula La_1−xSr_xMnO₃, where x describes the doping level, known by its abbreviation LSMO, has a 100% spin polarization at low temperatures [70] and is one of the most used ferromagnetic electrodes (especially as the bottom electrode) in organic spin valves (see Table 1). As another La-comprising perovskite oxide material, (La_2/3Pr_1/3)_5/8Ca_3/8MnO₃ (LPCMO) has a nearly maximum spin polarization (~97%) [71,72] and is starting to be employed in organic spin valves [36,49], leading to an incredibly large MR effect of up to 440% [36]. Moreover, it is worth to mention here the fact that LPCMO presents domains of electronic phase separation characterized by the coexistence of ferromagnetic metallic and antiferromagnetic insulating phases that can be magnetically controlled with preset magnetic fields, permitting the achievement of a nonvolatile tunable MR response in organic spin valves [36]. Starting from this result, one can imagine complex structured hybrid systems, like lateral arrays of tranches and nanowires [73,74,75,76], as templates for molecular assemblies that can provide an adjustable MR. The double perovskite Sr₂FeMoO₆ (SFMO) is also a potential material for organic spintronics with a high spin polarization (of 100% theoretically predicted [77] and ~85% experimentally detected [78]), used recently in organic-based junctions [37]. Additionally, magnetic oxides, such as Fe₃O₄ which has a high spin polarization of about -80% [79], are often used in organic spin valves [15,17,38,45]. Other materials with full spin polarization and a great potential in spintronics and multifunctional materials are Heusler compounds [80,81,82], from which Co₂MnSi has a 100% spin polarization [83] and can be employed in organic spin valves [25]. Magnetic transition metals Fe (44%), Co (34%) [84] and/or alloys of these metals, including FeCo (50%) [85], NiFe (45%) [86], Ni₈₀Fe₂₀ (~30%) [86,87] also called permalloy (Py) and close composition alloys such as Ni₈₁Fe₁₉ and Ni₇₈Fe₂₂ have an elevated Curie temperature (>700 °C) and are often used (mostly as a top electrode) in organic junctions (see Table 1).

Depending on the thickness of the separating layer, this can serve as a spin-conserved tunneling barrier or a spin transport medium. In the case of magnetic junctions with an organic spacer, the concept of tunneling is controversial, due to hybridization effects at interfaces and to the fact that the MR response is very similar either for tunneling or spin injection. However, for very thin thicknesses of only a few nm of the organic spacer (see Table 1), the tunneling mechanism is considered [63,88,89]. For organic spin valves, there are different mechanisms regarding the spin injection. A prime factor to be considered not only for the spin injection but also for the transport and the spin collection or detection is the energy level alignment at the metal–organic interface. Thus, the position of the Fermi level (E_F) of the metal with the work function (φ) in rapport with the positions of the molecular orbitals of the organic material with ionization energy (I_p) and the electron affinity energy (E_a) aligned at the vacuum level (Ev) determines the type of charge carrier. An example is shown in Figure 3. If the E_F of the metal is closer to the lowest unoccupied molecular orbital (LUMO), the main charge carriers are electrons (Figure 3a). If the E_F of the metal is closer to the highest occupied molecular orbital (HOMO), the main charge carriers are holes (Figure 3b). Moreover, one has to consider that because the spin polarization of the electronic states at different energy levels of the ferromagnetic electrode may be different, the spin injection and detection are likely different for different types of carriers and different energy alignments at the metal–organic interface. Additionally, alignment mismatches at interfaces may induce additional spin injection or collection problems (see, for example, Figure 3a).

Regarding the spin injection, every so often, a tunnel spin injection barrier consisting of a thin insulating film, such as AlO_x which is the most used (see Table 1), MoO_x [61], LAO (LaAlO₃) [20] or other films, like, for example, LiF [16] or organic layers [39,41,42,51], is inserted in between the ferromagnetic electrodes and the organic spin transport medium. These layers are used to optimize and tune the spin injection through processes such as solving the energy level mismatch at interfaces or tailoring the bend bending. Moreover, the inserted interfacial layer can improve the interface qualities, like, for example, in vertical junctions, it can reduce the penetration of the top ferromagnetic metallic electrode, or an organic spinterface [90,91,92,93] can ensure a less rough interface or better anchoring and tune the interface resistance [94,95,96,97].

Recently, several innovative spin injection methods based on studies on inorganic materials were proposed. From these, ferromagnetic resonance spin pumping injection is a pure spin injection process conducted by applying an external microwave field at its resonance frequency that generates a spin precession within the ferromagnetic layer, leading to an excess of spin angular momentum. When the precession frequency fits the microwave field, pure spin currents are injected into the organic layer through exchange interactions at the metal–organic interface [98,99,100], which lead to an induced electric field based on the inverse spin Hall effect, excluding the need for a bias voltage and resolving, thus, energy level alignment mismatches at interfaces.

Hot-electron spin injection or ballistic spin injection can be implemented in a three-terminal junction that contains three metal electrodes: an emitter, a base and a collector. From these layers, either the emitter or, in most cases, because a higher spin-polarized current is generated, the base should be ferromagnetic. The emitter and the base are separated by a tunnel barrier, while between the base and the collector is the spin transport medium, i.e., the organic layer. By applying a bias voltage between the emitter and the base, hot carrier electrons (or holes) are generated into the base that exhibit a ballistic spin filter effect based on different mean free paths of the majority and minority spin in the ferromagnet leading to a very high spin-polarized current that can be injected above the LUMO level for electrons (or below the HOMO level for holes) and measured at the collector if the applied voltage exceeds the injection barrier, i.e., the Schottky barrier, at the metal–organic interface [101,102,103]. In addition, hot electrons can be generated optically, using a photon photoemission process by applying an ultrashort laser pulse that generates hot electrons into the ferromagnet, from which a fraction can be ballistically injected at the metal–organic interface into the LUMO of the organic layer. Afterward, by applying a second laser pulse, the spin-polarized electrons remaining in the organic material can be excited into the vacuum and then photoemitted and detected [104]. Regarding spin-related optical effects, it is worth mentioning the coupling with luminescent and photovoltaic process that can lead to an enhancement in the electroluminescence intensity by the spin-polarized carriers [61] and to the manipulation of the spin-related magnetic response of the junction by the intensity of an irradiating light [64,68].

Another strategy to generate spin polarization in organic materials includes the chirality-induced spin selectivity injection. This can be reached when electrons with a spin orientation like the chirality of their transport medium are selected in favor of others that do not match this criterium [105,106,107,108]. The approach can lead to very high spin polarizations, also more than 80% [109].

Moreover, charge transfer spin polarization induced by an asymmetric population of molecular orbitals can be generated at the molecular level [110,111,112].

The transport of spin-polarized carriers through the organic layer is realized mainly through a hopping mechanism; however, for molecular crystals and polymers, a band transport is possible. Exchange coupling modes may be considered as an additional contribution for a very high concentration of transport carriers in doped or impurity band organic films. The spin diffusion thorough the organic layer for the hopping mode is defined principally by two factors, the spin diffusion length (λ_s) and the spin relaxation time (τ), as follows:

$${\lambda }_s=\sqrt{\frac{k_{B}T}{q}\mu \tau }$$

(3)

where k_B is the Boltzmann constant, T is the temperature, q is the carrier charge and μ is the mobility. Therefore, a long spin diffusion length can be reached for a long spin relaxation time and high mobility. The spin relaxation time is mainly influenced by the spin orbit coupling and the hyperfine interactions. The spin orbit coupling characterizes the interaction between the spin and the orbital angular momentum of the charge carrier, which contributes to the spin–charge conversion and the spin relaxation. Usually, the spin–orbit coupling is proportional to the fourth power of the atomic number. Thus, in organic materials that contain light-weight elements, such as C, H, N and O, the strength of the spin–orbit coupling is weak leading to an extremely long spin relaxation time. The hyperfine interaction, reported to be strong only for localized carriers, characterizes the exchange interactions between the spin of the charge carriers and of the nuclei (which are the half-integer nuclear spins of the atoms) and causes spin relaxations, as demonstrated, for instance, for deuterium/hydrogen substitution is DOO-PPV [30] where λ_s was increased significantly by deuteration. The mobility of the charge carriers is influenced by the molecular structure, carrier type (the mobility of electrons is higher than the one of holes), morphology of the films, packing modes, aggregations, disorder, defects, impurities and temperature. Crystals, polymers and the highly ordered stacked molecular aggregation of π-conjugated systems have a high mobility and weak spin scattering effects leading to long spin transport distances. The impact of various parameters on the spin transport and the related value of the MR can be deduced by a comparison of various organic junctions mentioned in Table 1.

The mobility for organic materials is the factor that strongly decreases λ_s as the spin relaxation time is high compared to inorganic materials [113].

2.2. Ferroelectric Junctions and Organic Ferroelectrics

Ferroelectric materials are dielectrics that possess a spontaneous electrical polarization whose orientation can be switched by an external electric field, with a critical value denoted the coercive field. Ferroelectrics are also pyroelectrics, piezoelectrics and present second harmonic generation properties. In ferroelectric junctions that consist of a ferroelectric film sandwiched between metallic electrodes, the current across the thick ferroelectric film (ferroelectric diode [114,115]) or the thin ferroelectric barrier (ferroelectric tunnel junction [5,6,7,8]) can be modulated by the orientation of the electrical polarization of the ferroelectric film.

The electroresistance modulation associated with the electric field-induced polarization reversal is in principle related to electrostatic effects at ferroelectric–metal interfaces. At the interface, polarization-generated charges that have different polarities in function of the orientation of the polarization will have, accordingly, an attraction or repulsion effect on the transport charge carriers. This effect arises over a short distance δ in the metal within which the electrons screen the polarization charges depending on the density of states at the Fermi level, giving rise to different δ for different metals. The incomplete screening modifies the depolarization field [7] and the electrostatic potential at each ferroelectric–metal interface (Figure 4) seen by the transport electrons [8,116]. Thus, the electronic potential profile is asymmetric and is modulated by the orientation of the polarization (Figure 5), which is controlled with an external electric field. In Figure 5, for simplicity, it is considered that the metals have a similar work function (φ₁ ≈ φ₂) marked as φ, while δ₁ > δ₂. The interface electrostatic effects are stronger as the thickness of the ferroelectric is smaller; thus, the electroresistance modulation by the orientation of the polarization is larger for ferroelectric tunnel junctions than for ferroelectric diodes. Moreover, the electroresistance values can be modulated to several resistive or also called memristive levels by applying progressive voltage pulses that partially switch the polar ferroelectric domains, which can reach a very high of a few orders of magnitude on–off ratio for the opposite orientation on the electrical polarization [117,118]. The electroresistance modulation can also be implemented in metal –organic ferroelectric–metal junctions [119,120,121,122,123,124,125]. Moreover, an electroresistance percentage (ER) can be deduced in a similar way as for magnetic spin valves described by the following:

$$\mathrm{E}\mathrm{R}=\frac{R_{up}-R_{down}}{R_{down}}100 \left(\%\right)$$

(4)

where R_up and R_down denote the corresponding resistance for opposite orientations of the electrical polarizations of the ferroelectric field.

Ferroelectricity appears in films or crystals of organic materials by following a few conditions, such as the following:

• the presence of electrical dipoles;
• the arrangement of the dipoles in non-centrosymmetric structures to avoid canceling each other through dipole–dipole interactions; and
• the compliance of properties with electrical polarization switching transitions that should have a sufficiently low energy-switching barrier affordable by the organic system.

The switching transitions can be associated with disorder–order transitions (which consist of rearrangements of individual molecules or of polar components) or displacive transitions (which consist of molecular displacement or of charge or ion displacement) [126,127,128]. Nevertheless, complex organic materials and metal–organic structures may include combinations of these processes. Organic ferroelectrics, depending on the composition, can be classified in the following: 1. single molecular components; 2. polymers and oligomers; 3. charge transfer complexes, or metal–organic complexes and hydrogen-bonded co-crystals.

Table 2. Organic ferroelectrics. Abbreviations of molecule names are as follows: TCAA—trichloroacetamide; TEMPO or tanane—tetramethyl-2,2,6,6-piperidinyl-oxyle; CDA—cyclohexan-1,1′-diacetic acid; TCHM—tricyclohexylmethanol; DNP—1,6-bis(2,4-dinitrophenoxy)-2,4-hexadiyne; CBDC—cyclobutene-1,2-dicarboxylic acid; PhMDA—2-phenylmalondialdehyde; 3-HPLN—3-hydroxyphenalenone; MBI—2-methylbenzimidazole; DC-MBI—5,6-dichloro-2-methylbenzimidazole; 1P—2-(p-tolyl)-1H-phenanthro[9,10-d]imidazole; SF-PFA—N-salicylidene-2,3,4,5,6-pentafluoroaniline; FF—diphenylalanine; PVDF—poly(vinylidene fluoride); P(VDF-TrFE)—poly(vinylidene fluoride—trifluoroethylene); P(VDF-HFP)—poly(vinylidene fluoride-hexafluoropropylene); TTF/CA—tetrathiafulvalene/chloranil; TTF/BA—tetrathiafulvalene/bromanil; Phz/H₂ca—phenazine/2,5-dicloro-3,6-dihydroxy-p-benzoquinones; Phz/H₂ba—phenazine/2,5-dibromo-3,6-dihydroxy-p-benzoquinones; DMBP/H₂ia—5,5′-dimethyl-2,2′-bipyridine/2,5-diiodo-3,6-dihydroxy-p-benzoquinones; Hdabco/ReO₄—1,4-diazabicyclo[2.2.2]octane perrhenate; TGS—triglycine sulfate; TSCC—tris-sarcosine calcium chloride; and Rochelle salt—potassium sodium tartrate tetrahydrate.

thiourea [129]     TCAA [130]       TEMPO [131]       CDA [132]       TCHM [133]      Benzil [134]          DNP [135]
squaric acid [136]    CBDC [137]      croconic acid [138]  PhMDA [137]   3-HPLN [137]      MBI [139]         DC-MBI [139]
1P [140]           SF-PFA [141]		guanine [142]     glycine [143]        thymine [144]           FF [145]
PVDF [146,147]            P(VDF-TrFE) [148]          P(VDF-HFP) [149]      nylon 11 [150] 6,10,11,12-12 [151]     PANI [152]
TTF/CA [153,154]             TTF/BA [154]            Phz/H2ca [155]                 Phz/H2ba [155]               DMBP/H2ia [156]
Hdabco/ReO4 [157]                       TGS [158,159]                    TSCC [160]                                       Rochelle salt [161]

contains a few examples for each category. In order to be included in ferroelectric junctions, ferroelectric films can be deposited by thermal evaporation, e.g., organic molecular beam epitaxy, similarly to the organic spin valves for small single-component systems, or wet chemical methods that are mostly used [119,120,121,122,123,124,125], including spin coating, drop casting, Langmuir–Blodget, etc., for polymers and other macromolecular structures.

The electrical polarization switching mechanism of organic ferroelectrics can be related to their structure and composition. The majority of single-component systems, which are mentioned in part 1 of Table 2, form crystallographic structures through hydrogen bonds such as N-H⋅⋅⋅O, O-H⋅⋅⋅N, O-H⋅⋅⋅O and N-H⋅⋅⋅N. These intermolecular interactions allow for both the disorder–order and the displacive transitions in the polarization orientation processes. Thus, electrical polarization switching involves molecular rotations (such as, for example, for thiourea [129], CDA [132], glycine [143], thymine [144]), the reorientation of functional groups (such as, for example, for TCAA [130], TEMPO [131], TCHM [133], benzil [134], DNP [135]) and the intermolecular displacement of H⁺ within hydrogen bonds (such as, for example, within O-H⋅⋅⋅O for squaric acid [136], CBDC [137], croconic acid [138], PhMDA [137] and 3-HPLN [137]; within N-H⋅⋅⋅N for MBI [139], DC-MBI [139], 1P [140] and guanine [142] and within N-H⋅⋅⋅O for FF [145]). The latest example can be photo-activated. Another photo-induced switching is produced by a H⁺ intramolecular displacement within the O-H⋅⋅⋅N bonds for SA-PFA molecules [141]. The polymeric systems that are included in part 2 of Table 2 all switch their electrical polarization by the reorientation of functional groups and rotations of polymeric chains [146,147,148,149,150,151,152]. The bicomponent systems from part 3 of Table 2 change their polarization state by electronic transfer for TTF/CA [153,154] and TTF/BA [154] and by H⁺ transfer within O-H⋅⋅⋅N bonds for Phz/H₂ca [155], Phz/H₂ba [155] and DMBP/H₂ia [156]. The polarization changes for the systems in the last row of Table 2 are all produced by molecular reorientations and order–disorder transitions [157,158,159,160,161].

3. Organic Multiferroic Junctions

OMFJs can be obtained by combining the previously described devices, for instance, introducing an organic ferroelectric film as the separating medium in an organic spin valve. The main characteristic of OMFJs is the possibility to electrically control the spin polarization at the ferromagnet–organic–ferroelectric interfaces by switching the electrical polarization of the ferroelectric and thus manipulating the MR and ER effects.

These coupled processes were first demonstrated for inorganic multiferroic tunnel junctions [2,3,162]. Experimentally, recent studies on OMFJs using a few organic ferroelectric materials, such as PVDF [163,164,165], P(VDF-TrFE) [166,167] and croconic acid [168], have confirmed their memristive behavior manipulated with electric and magnetic stimuli. The rich variety of possibilities (see Table 2) of including other ferroelectric organic and biomolecular materials in OMFJs opens further routes for tuning the magnetoelectric and electronic/spintronic properties of these systems. As for any other organic spin valves and ferroelectric junctions, the MR and ER effects depend on other parameters and conditions. Moreover, the ferroelectric control of the spin polarization can induce the reversal of the magnetoresistance hysteresis (see Figure 2), i.e., the spin rectification effect, by switching the electrical polarization.

The MR effect of OMFJs, as for any organic spin valve, strongly depends on temperature and the thickness of the organic film. Expected consequences of these factors can be observed on the examples represented in Figure 6a–d. Moreover, the ER effect is also affected by these parameters (Figure 6d). Furthermore, the interfaces play a crucial role in defining the MR, ER and the coupled effects. For LSMO/PVDF/Co systems presented in Figure 6c,e, the PVDF/Co interface is different as the Co layer was deposited at low and room temperature, respectively. The MR for a supposedly rougher interface, with diffusion channels that also affect the thickness of the spin transport medium, is negative and does not show spin rectification for any electrical polarization orientation of the ferroelectric film (Figure 6e). For a similar system, by intercalating a thin MgO layer, the MR becomes positive (Figure 6e,f).

A similar behavior of the MR, ER and spin rectification effect with the temperature and the thickness is encountered in the LSMO/P(VDF-TrFE)/Co system (Figure 7). As expected, the MR and ER decrease by increasing the temperature and the thickness (Figure 7b,d,f,g). Different tendencies at certain temperature ranges may be related to a different spin transport mechanism [164]. Moreover, the thickness is important for determining the sign of the MR and the spin rectification effect (Figure 7a–f). In addition, the memristive properties of the ferroelectric spacer and the ferromagnetic–ferroelectric interface coupling induce mainly an increase in the MR and ER with the magnitude of the external electric field applied successively to the OMFJ (Figure 7i).

The role of the electric field and measurement values of the bias voltage in the MR was carefully investigated for the LSMO/croconic acid/SiO₂/Co system [168]. Positive and negative electrical pulses are used to switch the polarization of the ferroelectric film to on and off states. But, additionally, a negative applied voltage with a magnitude above a certain threshold for measuring the MR induces the spin rectification effect (Figure 8). The ferroelectric–metallic interface effects of the OMFJs are used to fine-tune the MR.

4. Conclusions and Outlook

The recent realization of OMFJs by combining the knowledge on organic spin valves and organic ferroelectric junctions shows encouraging results on exploiting the magnetoelectric coupling, electrostatic and energy alignment conditions at the ferroelectric–ferromagnetic interfaces to tune the MR and ER. Depending on these factors and the related variable parameters like, for example, the materials, the thickness, an intercalated oxide layer at the organic–metal interface and the temperature, the memristive states of OMFJs can be manipulated by triggering the external magnetic and electric stimuli. The memristive properties of switching and retaining the electrical resistance using the spin degree of freedom of electrons can principally be used as nonvolatile memory elements to store and process information with less power consumption and can serve as models of synaptic and neuronal operations. The promising results of OMFJs [163,164,165,166,167,168] open routes for new research contributions exploring various materials and conditions to optimize their properties. For instance, a large variety of molecular and biomolecular ferroelectrics [129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,161] can be used as the spin transport medium in OMFJs. Extensively used magnetic films may be replaced by other magnetic compounds with high polarization in spin [25,80,81,82,83], new magnetic constituents such as 2D van der Waals materials [169,170,171,172,173,174,175], or 2D networks of organic and metal–organic magnets on surfaces [110,111,112,176,177,178,179,180,181,182,183], metal–organic frameworks [20,21], heterostructures [110,111,184,185,186] and organic spinterfaces [13,49,51,90,91,92,163,187] as electrodes to improve the spin injection and/or the functionality of OMFJs. A few recent successful examples already include some of these new alternative types of electrodes for photovoltaic [188], capacitive [189] and spin-crossover [190] devices. Further studies may be important to also explore the potential of OMFJs for flexible electronics, wearable electronics, transient electronics, bioelectronics and perchance coupling their properties with light-related phenomena in magneto-optics, optoelectronics or photovoltaics contributing to the advancement of nanostructures and thin films [191].