The Progress on Magnetic Material Thin Films Prepared Using Polymer-Assisted Deposition

Polymer-assisted deposition (PAD) has been widely used in the preparation of high-quality oxides and sulfides for basic research and applications. Specifically, diverse PAD-prepared magnetic material thin films such as ZnO, Ga2O3, SrRuO3, LaCoO3, LaMnO3, Y3Fe5O12, MoS2, MoSe2, and ReS2 thin films have been grown, in which thickness-dependent, strain-modulated, doping-mediated, and/or morphology-dependent room-temperature ferromagnetism (RTFM) have been explored. Inspired by the discovery of intrinsic low-temperature FM in two-dimensional (2D) systems prepared using mechanical exfoliation, the search for more convenient methods to prepare 2D ferromagnetic materials with high-temperature FM has seen explosive growth, but with little success. Fortunately, the very recent synthesis of 2D NiO by PAD has shed light on this challenge. Based on these abovementioned developments, the difficulties of PAD when preparing a-few-nanometer single-crystalline materials and the opportunities in PAD for novel materials such as chiral magnetic soliton material Cr1/3NbS2 are discussed.


Polymer-Assisted Deposition (PAD) 2.1. Main Processing Steps
Similar to other traditional chemical solution deposition (CSD) methods, PAD also involves preparing precursor solutions using coatings and heat treatment of precursor films. As shown in Figure 2, the main difference lies in the preparation of polymer precursor solutions ( Figure 2a) and ultrafiltration (Figure 2b). In PAD, metal ions coordinate with polymers in the solution to form precursors. This polymer mainly refers to polyethyleneimine (PEI), which can complex with the vast majority of metals.
Molecules 2023, 28, x FOR PEER REVIEW 2 of 23 RTFM in these materials, are presented and discussed. Then the challenges and opportunities for PAD in the preparation of 2D ferromagnetic materials are also manifested. In the end, we compare the advantages of PAD, mechanical exfoliation (ME) [72], and chemical vapor transport [73][74][75] (CVT) in the preparation of novel materials such as chiral magnetic soliton material CrNbS1/3 (CNS).

Main Processing Steps
Similar to other traditional chemical solution deposition (CSD) methods, PAD also involves preparing precursor solutions using coatings and heat treatment of precursor films. As shown in Figure 2, the main difference lies in the preparation of polymer precursor solutions ( Figure 2a) and ultrafiltration (Figure 2b). In PAD, metal ions coordinate with polymers in the solution to form precursors. This polymer mainly refers to polyethyleneimine (PEI), which can complex with the vast majority of metals. (a-d) Schematic illustration of the main processing steps used to grow thin films using PAD [20].
In 2006, Lin and Xie et al. [93] prepared ZnO TFs on sapphire substrates by means of PAD. The films were composed of nanometric particles and had a high c-axis orientation. The crystallinity, c-axis orientation, and surface morphology of ZnO TFs could be modified with rapid heat treatment. It was found that more oxygen vacancies (Vo) could enhance the green photoluminescence properties.
In 2012, we developed ZnO TFs using PAD [71] where all the films exhibited RTFM. Results from the superconducting quantum interference device (SQUID) and X-ray photoelectron spectroscopy (XPS) show that RTFM was not directly related to Vo. In order to clarify the origin of RTFM, we tested the samples using positron annihilation spectroscopy (PAS). Unexpectedly, zinc vacancies (V Zn ) were found to be responsible for the RTFM. Actually, this phenomenon of RTFM caused by V Zn was also found in samples prepared using PLD [87,94] and ILE [64]. In contrast, it was reported that Vo was the origin of RTFM in the undoped ZnO samples' prepared sol-gel [84] and pulsed electron beam deposition (PED) [95].

Ga2O3
High-quality β-Ga2O3 TFs [96] with a single preferential growth orientation of (−201) and low defect density were successfully grown on sapphire substrates using PAD. Interestingly, the Mn concentration was closely related with the saturation magnetization (Ms) and coercive field (Hc) values in amorphous Mn-doped gallium oxide [97] TFs, which are different from crystalline Mn-doped Ga2O3 TFs. Notably, the samples annealed in air exhibited a Ms [98] as strong as 170% times that of the samples annealed in pure O2 gas, which can be quantitatively explained by oxygen-vacancy-controlled ferromagnetism due to bound magnetic polarons established between delocalized hydrogenic electrons of Vos and local magnetic moments of Mn 2+ , Mn 3+ , and Mn 4+ ions in the samples.

SrRuO3 (SRO)
SrRuO3, as a ferromagnetic conductive material, is usually used as a metal layer at magnetic tunnel junctions [99]. Its unique perovskite structure facilitates integration with other functional metal oxides. Epitaxial ferromagnetic SRO TFs were obtained experimentally through several experimental methods, including PLD [99,100,101], solgel [102,103], sputtering [104,105,106], molecular beam epitaxy (MBE) [107], and metalorganic chemical vapor deposition (MOCVD) [108,109]. PAD has its unique advantages as a low-cost method for preparing large-area metal oxides coatings. In 2007, Jia et al. [33] successfully prepared epitaxial SRO TFs with high crystallinity using PAD, as shown in Figure 4A. Their curie temperature (Tc) is about 160 K in Figure 4B-D, which is equivalent to that in the other literature [104]. The thickness of PAD-prepared SRO TFs is usually around 100 nm.

Ga 2 O 3
High-quality β-Ga 2 O 3 TFs [96] with a single preferential growth orientation of (−201) and low defect density were successfully grown on sapphire substrates using PAD. Interestingly, the Mn concentration was closely related with the saturation magnetization (M s ) and coercive field (H c ) values in amorphous Mn-doped gallium oxide [97] TFs, which are different from crystalline Mn-doped Ga 2 O 3 TFs. Notably, the samples annealed in air exhibited a M s [98] as strong as 170% times that of the samples annealed in pure O 2 gas, which can be quantitatively explained by oxygen-vacancy-controlled ferromagnetism due to bound magnetic polarons established between delocalized hydrogenic electrons of Vos and local magnetic moments of Mn 2+ , Mn 3+ , and Mn 4+ ions in the samples.

SrRuO 3 (SRO)
SrRuO 3 , as a ferromagnetic conductive material, is usually used as a metal layer at magnetic tunnel junctions [99]. Its unique perovskite structure facilitates integration with other functional metal oxides. Epitaxial ferromagnetic SRO TFs were obtained experimentally through several experimental methods, including PLD [99][100][101], sol-gel [102,103], sputtering [104][105][106], molecular beam epitaxy (MBE) [107], and metalorganic chemical vapor deposition (MOCVD) [108,109]. PAD has its unique advantages as a low-cost method for preparing large-area metal oxides coatings. In 2007, Jia et al. [33] successfully prepared epitaxial SRO TFs with high crystallinity using PAD, as shown in Figure 4A. Their curie temperature (T c ) is about 160 K in Figure 4B-D, which is equivalent to that in the other literature [104]. The thickness of PAD-prepared SRO TFs is usually around 100 nm.

LaCoO3 (LCO)
Goodenough et al. [110] discovered the phase transition of Co from the low-spin state to the high-spin state in LCO samples, which were grown using high-temperature ceramic techniques. However, it was very difficult to control the sample uniformity and oxygen dose. Later, they obtained uniform single-phase LCO samples using precursor coprecipitation [111] and studied the magnetic and transport properties. The temperaturedependent electronic structure model can explain this phase transition from low-spin Co (III) to high-spin Co 3+ well. In 2013, Rivadulla et al. [67] successfully prepared high-quality epitaxial LCO TFs on SrTiO3 (STO) using PAD. The film was uniform and smooth throughout the region, as shown in Figure 5A. As shown in Figure 5B, LCO films were grown epitaxially on STO substrates. Its Tc is around 85 K in Figure 5C. The thicknessdependent magnetism [27,63] in Figure 5D has also been found in LCO film systems, which is closely related to biaxial tensile strain. When the film thickness is reduced, the FM is enhanced.
LCO TFs induced an additional thermal strain through thermal expansion coefficient mismatch [65], which would lead to larger lattice parameters. Interestingly, Tc and magnetism can be significantly increased by adjusting the in-plane tensile strain. However, for PAD-grown epitaxial LCO TFs, in-plane lattice tensile strain and CoO6 octahedral rotation occur. In the epitaxial growth process, in-plane biaxial tensile

LaCoO 3 (LCO)
Goodenough et al. [110] discovered the phase transition of Co from the low-spin state to the high-spin state in LCO samples, which were grown using high-temperature ceramic techniques. However, it was very difficult to control the sample uniformity and oxygen dose. Later, they obtained uniform single-phase LCO samples using precursor coprecipitation [111] and studied the magnetic and transport properties. The temperaturedependent electronic structure model can explain this phase transition from low-spin Co (III) to high-spin Co 3+ well. In 2013, Rivadulla et al. [67] successfully prepared highquality epitaxial LCO TFs on SrTiO 3 (STO) using PAD. The film was uniform and smooth throughout the region, as shown in Figure 5A. As shown in Figure 5B, LCO films were grown epitaxially on STO substrates. Its T c is around 85 K in Figure 5C. The thicknessdependent magnetism [27,63] in Figure 5D has also been found in LCO film systems, which is closely related to biaxial tensile strain. When the film thickness is reduced, the FM is enhanced.
Molecules 2023, 28, 5004 6 of 23 film magnetism, as in Figure 5E. The LCO films with three different orientations show significantly different magnetic behaviors from the bulk, which is related to the biaxial compressive strain and the tetragonal distortion of CoO6 octahedra ( Figure 5H-K). It was also found that PEI molecular weight, heat treatment conditions, and spin-coating rate had significant effects on the crystallinity and epitaxial quality of LCO TFs [28]. In addition, epitaxial LCO TFs were obtained with PLD [69]. Similarly, the FM to PM phase transition was also observed at about 80 K.  (111) oriented LCO films grown on (100), (110), and (111) LAO substrates, respectively (reprinted with permission from [65]; reprinted with permission from [27]).
Modulating material properties through strain [22,[112][113][114] has always been a mysterious topic, especially the regulation of RTFM [4,[115][116][117]. However, it has always been a challenge to introduce strain directly into materials in experiments. Although strain has been introduced into materials with many methods, such as using flexible substrates [118], using lattice mismatch [118,119] or thermal [120,121] strain, introducing wrinkles, LCO TFs induced an additional thermal strain through thermal expansion coefficient mismatch [65], which would lead to larger lattice parameters. Interestingly, T c and magnetism can be significantly increased by adjusting the in-plane tensile strain. However, for PAD-grown epitaxial LCO TFs, in-plane lattice tensile strain and CoO 6 octahedral rotation occur. In the epitaxial growth process, in-plane biaxial tensile deformation is induced, causing the Co-O distance (r Co−O ) to stretch in the in-plane direction and compress in the out-of-plane direction, as shown in Figure 5F-G. The inhibition of the CoO 6 octahedral rotation is more evident in thinner films. In addition, the rectangular distortion of the CoO 6 octahedron caused by the deformation can also reduce the e g -t 2g gap [63]. In general, the strain-induced FM originates from the reduction in the e g -t 2g gap and the inhibition of the CoO 6 octahedral rotation.
Notably, the orientation of LAO substrates has a significant effect on the epitaxial film magnetism, as in Figure 5E. The LCO films with three different orientations show significantly different magnetic behaviors from the bulk, which is related to the biaxial compressive strain and the tetragonal distortion of CoO 6 octahedra ( Figure 5H-K). It was also found that PEI molecular weight, heat treatment conditions, and spin-coating rate had significant effects on the crystallinity and epitaxial quality of LCO TFs [28]. In addition, epitaxial LCO TFs were obtained with PLD [69]. Similarly, the FM to PM phase transition was also observed at about 80 K.

LaMnO3 (LMO)
In 1965, Solovyev et al. [126] explained the AFM of LMO perovskite by using the itinerant-electron model based on the local spin density approximation. Lattice distortion strongly affects magnetism. When La 3+ is partially placed, the FM is introduced into the LMO system. As a typical example, Rivadulla et al. [127] reported an oriented LMO with RTFM grown using PAD, which is up to a number of centimeters in size and is greatly compatible with standard microfabrication techniques. Furthermore, they found that the type of substrate affects the orientation of LMO films [128], and some non-stoichiometric diffraction peaks may even appear. Even after annealing at 900 °C in air, the epitaxial characteristics of LMO TFs were still preserved.
Jain et al. [30] attempted to prepare high-quality Sr-and Ca-doped LMO epitaxial films using PAD. The annealing temperature determined the crystallinity, microstructure, resistance, magnetism, and magnetoresistance of TFs. In addition, dense polycrystalline films of La1−xCaxMnO3 [129] were also deposited on Si substrates. Moreover, it was found that the annealing atmosphere could affect the morphology of TFs, and the film was composed of sintered grains after annealing under oxygen. Similarly, the Tc and Ms could also be modulated through annealing conditions. Generally, the introduction of the second phase will also greatly modulate the properties of the materials. SrTiO3 [50], as the second phase, was introduced to regulate the magnetic transport properties of LCMO epitaxial TFs using PAD. This phase increases the resistance of LCMO TFs by increasing the height of the spin-dependent tunneling barrier between magneto-crystals and significantly changes the magnetoresistance (MR) by increasing the MR value and reducing the metal-insulator transition temperature. Similarly, NiO or Co3O4 are also added, which can have a similar effect [37]. Antiferromagnetic Co3O4 has a more obvious inhibitory effect on the magnetism of LCMO TFs.

LaMnO 3 (LMO)
In 1965, Solovyev et al. [126] explained the AFM of LMO perovskite by using the itinerant-electron model based on the local spin density approximation. Lattice distortion strongly affects magnetism. When La 3+ is partially placed, the FM is introduced into the LMO system. As a typical example, Rivadulla et al. [127] reported an oriented LMO with RTFM grown using PAD, which is up to a number of centimeters in size and is greatly compatible with standard microfabrication techniques. Furthermore, they found that the type of substrate affects the orientation of LMO films [128], and some non-stoichiometric diffraction peaks may even appear. Even after annealing at 900 • C in air, the epitaxial characteristics of LMO TFs were still preserved.
Jain et al. [30] attempted to prepare high-quality Sr-and Ca-doped LMO epitaxial films using PAD. The annealing temperature determined the crystallinity, microstructure, resistance, magnetism, and magnetoresistance of TFs. In addition, dense polycrystalline films of La 1−x Ca x MnO 3 [129] were also deposited on Si substrates. Moreover, it was found that the annealing atmosphere could affect the morphology of TFs, and the film was composed of sintered grains after annealing under oxygen. Similarly, the T c and M s could also be modulated through annealing conditions. Generally, the introduction of the second phase will also greatly modulate the properties of the materials. SrTiO 3 [50], as the second phase, was introduced to regulate the magnetic transport properties of LCMO epitaxial TFs using PAD. This phase increases the resistance of LCMO TFs by increasing the height of the spin-dependent tunneling barrier between magneto-crystals and significantly changes the magnetoresistance (MR) by increasing the MR value and reducing the metal-insulator transition temperature. Similarly, NiO or Co 3 O 4 are also added, which can have a similar effect [37]. Anti-ferromagnetic Co 3 O 4 has a more obvious inhibitory effect on the magnetism of LCMO TFs.
PAD-grown epitaxial multilayer TFs were also systematically studied. The behavior of the interfacial magnetic coupling of LCO/LMO [41] systems was studied by Rivadulla et al. The LCO precursor is spin-coated after the preparation of LMO TFs, and the LCO/LMO epitaxial bilayers can be prepared after annealing. The coercivity of LCO/LMO is 30 times higher than that of single-film LMO or LCO due to superexchange related to a redox reaction at the interface. However, the coercivity of La 0.92 MnO 3 /LCO [42] is 5 times higher than that of LMO film, which is caused by FM superexchange at the interface. Furthermore, the tunneling conduction phenomenon was also found in ferromagnetic epitaxial bilayers of LCO/LSMO [40]. The negative temperature coefficient of tunneling resistance shows that the quality of the heterojunction meets the basic research and application. Interestingly, substrate orientation also has a significant effect on the magnetic and transport properties of the LCMO/BSTO bilayer films [52]. In addition, metal-insulator phase transition highly related to substrate orientation was also found. Moreover, a large magnetoresistance [31] (−71% at 5T) was observed in multilayer-coated LSMO/LCMO TFs. Therefore, spincoating multilayer ferromagnetic materials is an effective way to modulate phase transition temperature and magnetoresistance.
Easy doping [1,[29][30][31][32]34,36,51,130] is an important feature of PAD. Sr-doped LaMnO 3 [30,31,40,50,52] TFs and Co-doped ZnO [29] TFs are representatives of magnetic materials prepared using PAD. Doping can not only mediate the magnetic strength [29], but also drive the magnetic phase transition [30,31]. Rare earth metal element doping can introduce the biaxial compressive strain in RCMO [68] and RNMO [77] TFs, as seen in Figure 7A,B. Due to the ionic radius difference of rare elements, the in-plane lattice parameter is changed, and the biaxial compressive strain is introduced in the film due to the lattice relaxation effect. Moreover, there is a linear relationship between T c and biaxial compressive strain. Furthermore, the preparation of a monolayer-doped system using PAD is a challenging research direction. LCO/LMO epitaxial bilayers can be prepared after annealing. The coercivity of LCO/LMO is 30 times higher than that of single-film LMO or LCO due to superexchange related to a redox reaction at the interface. However, the coercivity of La0.92MnO3/LCO [42] is 5 times higher than that of LMO film, which is caused by FM superexchange at the interface. Furthermore, the tunneling conduction phenomenon was also found in ferromagnetic epitaxial bilayers of LCO/LSMO [40]. The negative temperature coefficient of tunneling resistance shows that the quality of the heterojunction meets the basic research and application. Interestingly, substrate orientation also has a significant effect on the magnetic and transport properties of the LCMO/BSTO bilayer films [52]. In addition, metal-insulator phase transition highly related to substrate orientation was also found. Moreover, a large magnetoresistance [31] (−71% at 5T) was observed in multilayer-coated LSMO/LCMO TFs. Therefore, spin-coating multilayer ferromagnetic materials is an effective way to modulate phase transition temperature and magnetoresistance. Easy doping [1,[29][30][31][32]34,36,51,130] is an important feature of PAD. Sr-doped LaMnO3 [30,31,40,50,52] TFs and Co-doped ZnO [29] TFs are representatives of magnetic materials prepared using PAD. Doping can not only mediate the magnetic strength [29], but also drive the magnetic phase transition [30,31]. Rare earth metal element doping can introduce the biaxial compressive strain in RCMO [68] and RNMO [77] TFs, as seen in Figure 7A,B. Due to the ionic radius difference of rare elements, the in-plane lattice parameter is changed, and the biaxial compressive strain is introduced in the film due to the lattice relaxation effect. Moreover, there is a linear relationship between Tc and biaxial compressive strain. Furthermore, the preparation of a monolayer-doped system using PAD is a challenging research direction.

Layered Material Thin Films
Since long-range intrinsic ferromagnetic order was observed in exfoliated Cr2Ge2Te6 [153] and CrI3 [62] monolayer systems using polar magneto-optical Kerr effect (MOKE) measurements in 2017, the research on 2D ferromagnetic materials has become a fastgrowing field. However, the size and thickness of 2D materials obtained with most of the methods, including mechanical exfoliation [62,153,154], chemical vapor transport [63,155,156,157], microwave irradiation [158], direct vapor transport technique [159], and so on, are difficult to control.

MoS2
As early as 2001, Remskar et al. [160] used C60 as a growth promoter to prepare chiral MoS2 nanotubes using the catalytic transport method. Inspired by this experimental method, Jagličić et al. [161] and Mihailovic et al. [162] each studied the magnetism of Lidoped MoS2 nanotube structures in 2003. Li-doped MoS2 nanotubes seem to be the realization of a near ideal 1D system, showing a strong new qualitative correlation in behavior. Moreover, edge-oriented MoS2 TFs [163,164] exhibit weak magnetism (~1−2 emu•g −1 ). The magnetism is related to the edge spin on the edge of the prism of the nanosheet. Remarkably, the first preparation of MoS2 TFs [26] using PAD was achieved in 2016. The absorption spectrum showed that MoS2 TFs have thickness-dependent band gap regulation. When the film thickness was reduced to 2.5 nm, discontinuities appeared. Inspired by this work, we prepared MoS2 TFs [3,5] with adjustable thicknesses, shown in Figure 9.
Due to the mismatch of thermal expansion coefficients between the film and the substrate, residual compressive strain is introduced at the interface between film and substrate during high temperature growth. After buckling, MoS2 TFs change from very flat to rough, as seen in Figure 9A,B. As shown in Figure 9C, the roughness of the 400 nm thick film is about 1 nm. The crystallinity of the film in Figure 9D, in which there are polycrystalline and amorphous components, is not high. However, the distribution of Mo and S elements is relatively uniform (Figure 9E), and their valences are +4 and −2 ( Figure  9F), respectively. Raman measurements show that there is compressive strain at the bottom of the film, as seen in Figure 9G.

Layered Material Thin Films
Since long-range intrinsic ferromagnetic order was observed in exfoliated Cr 2 Ge 2 Te 6 [153] and CrI 3 [62] monolayer systems using polar magneto-optical Kerr effect (MOKE) measurements in 2017, the research on 2D ferromagnetic materials has become a fast-growing field. However, the size and thickness of 2D materials obtained with most of the methods, including mechanical exfoliation [62,153,154], chemical vapor transport [63,[155][156][157], microwave irradiation [158], direct vapor transport technique [159], and so on, are difficult to control.

MoS 2
As early as 2001, Remskar et al. [160] used C 60 as a growth promoter to prepare chiral MoS 2 nanotubes using the catalytic transport method. Inspired by this experimental method, Jagličić et al. [161] and Mihailovic et al. [162] each studied the magnetism of Li-doped MoS 2 nanotube structures in 2003. Li-doped MoS 2 nanotubes seem to be the realization of a near ideal 1D system, showing a strong new qualitative correlation in behavior. Moreover, edge-oriented MoS 2 TFs [163,164] exhibit weak magnetism (~1-2 emu·g −1 ). The magnetism is related to the edge spin on the edge of the prism of the nanosheet. Remarkably, the first preparation of MoS 2 TFs [26] using PAD was achieved in 2016. The absorption spectrum showed that MoS 2 TFs have thickness-dependent band gap regulation. When the film thickness was reduced to 2.5 nm, discontinuities appeared. Inspired by this work, we prepared MoS 2 TFs [3,5] with adjustable thicknesses, shown in Figure 9.
Due to the mismatch of thermal expansion coefficients between the film and the substrate, residual compressive strain is introduced at the interface between film and substrate during high temperature growth. After buckling, MoS 2 TFs change from very flat to rough, as seen in Figure 9A,B. As shown in Figure 9C, the roughness of the 400 nm thick film is about 1 nm. The crystallinity of the film in Figure 9D, in which there are polycrystalline and amorphous components, is not high. However, the distribution of Mo and S elements is relatively uniform (Figure 9E), and their valences are +4 and −2 ( Figure 9F), respectively. Raman measurements show that there is compressive strain at the bottom of the film, as seen in Figure 9G. Due to the disturbance of humidity or temperature, the residual compressive strain in the film is released, resulting in the formation of web buckles [3][4][5]22]. According to the change rate of peak frequency (1/ꞷ) × (dꞷ/dɛ), we can estimate the strain change on the surface of the film [3,165]. The film changes from compressive strain to tensile strain after buckling. Interestingly, the RTFM of web buckles is 7.5 times stronger than that of the flat TFs in Figure 10A. In detail, Ms is increased at different test temperatures ( Figure 10B). The fluorescence spectra in Figure 10C,D show that the enhanced magnetism is related to Vs. Biaxial strain can regulate the magnetic ordering temperature as shown by the dot line, which decreases from 367 K to 338 K after buckling ( Figure 10E,F). There exist small protuberances around 57 K in the green area corresponding to the Neel point (TN), which is the consequence of the antiferromagnetic forces. In general, the decrease in compressive strain and the increase in tensile strain produce more defects, thereby enhancing their RTFM. Due to the disturbance of humidity or temperature, the residual compressive strain in the film is released, resulting in the formation of web buckles [3][4][5]22]. According to the change rate of peak frequency (1/ω) × (dω/d ), we can estimate the strain change on the surface of the film [3,165]. The film changes from compressive strain to tensile strain after buckling. Interestingly, the RTFM of web buckles is 7.5 times stronger than that of the flat TFs in Figure 10A. In detail, M s is increased at different test temperatures ( Figure 10B). The fluorescence spectra in Figure 10C,D show that the enhanced magnetism is related to Vs. Biaxial strain can regulate the magnetic ordering temperature as shown by the dot line, which decreases from 367 K to 338 K after buckling ( Figure 10E,F). There exist small protuberances around 57 K in the green area corresponding to the Neel point (T N ), which is the consequence of the antiferromagnetic forces. In general, the decrease in compressive strain and the increase in tensile strain produce more defects, thereby enhancing their RTFM. Molecules 2023, 28, x FOR PEER REVIEW 11 of 23

MoSe2
Recently, MoSe2 has received more attention because it has a similar hexagonal atomic-layers structure to MoS2 and even better electronic and optical properties because of the more suitable bandgap (~1.1-1.6 eV) [166] compared to MoS2 (~1.3-1.9 eV) [167]. Notably, Zhang et al. [168] found a spin-splitting of ~180 meV at the top of the valence of MoSe2 film prepared using MBE, which was larger than that of ~100 meV for single-layer MoS2 prepared using mechanical exfoliation [169], suggesting MoSe2 has a greater application potential than MoS2 in spintronic devices. To investigate the magnetism of MoSe2, in 2015, Xia et al. [170] synthesized 2H-MoSe2 nanoflakes with zigzag edges using CVD, which exhibited Ms of 1.39 emu·g −1 that diminished when the flakes became bigger. Furthermore, they also synthesized 1T-MoSe2 incorporated nanosheets [171] and nanoflowers [172] using hydrothermal and solvothermal, respectively, but their Ms values (8.36 × 10 −3 emu·g −1 and 2.70 × 10 −2 emu·g −1 ) were even lower than those in 2H-MoSe2 nanoflakes with zigzag edges, despite predictions that 1T-MoSe2 provided a magnetic moment of 2 µB/Mo atom equal to that provided by Mo atoms at the zigzag edges [25].
The PAD was first used to synthesize MoSe2 TFs in 2020 [173]. A photodetector was fabricated based on the 22 nm thickness of MoSe2 TFs on Au substrate, which showed near-perfect absorption in visible wavelengths. Based on the above work, we prepared 4 cm × 4 cm of MoSe2 TFs with a smooth surface (roughness average ~0.22 nm) using PAD [25], as shown in Figure 11A,B.

MoSe 2
Recently, MoSe 2 has received more attention because it has a similar hexagonal atomiclayers structure to MoS 2 and even better electronic and optical properties because of the more suitable bandgap (~1.1-1.6 eV) [166] compared to MoS 2 (~1.3-1.9 eV) [167]. Notably, Zhang et al. [168] found a spin-splitting of~180 meV at the top of the valence of MoSe 2 film prepared using MBE, which was larger than that of~100 meV for single-layer MoS 2 prepared using mechanical exfoliation [169], suggesting MoSe 2 has a greater application potential than MoS 2 in spintronic devices. To investigate the magnetism of MoSe 2 , in 2015, Xia et al. [170] synthesized 2H-MoSe 2 nanoflakes with zigzag edges using CVD, which exhibited M s of 1.39 emu·g −1 that diminished when the flakes became bigger. Furthermore, they also synthesized 1T-MoSe 2 incorporated nanosheets [171] and nanoflowers [172] using hydrothermal and solvothermal, respectively, but their M s values (8.36 × 10 −3 emu·g −1 and 2.70 × 10 −2 emu·g −1 ) were even lower than those in 2H-MoSe 2 nanoflakes with zigzag edges, despite predictions that 1T-MoSe 2 provided a magnetic moment of 2 µ B /Mo atom equal to that provided by Mo atoms at the zigzag edges [25].
The PAD was first used to synthesize MoSe 2 TFs in 2020 [173]. A photodetector was fabricated based on the 22 nm thickness of MoSe 2 TFs on Au substrate, which showed near-perfect absorption in visible wavelengths. Based on the above work, we prepared 4 cm × 4 cm of MoSe 2 TFs with a smooth surface (roughness average~0.22 nm) using PAD [25], as shown in Figure 11A,B. The crystallinity of the films increases but the Ms decreases with the growth temperature, as shown in Figure 11C Figure 11F indicate that the Tc of MoSe2 TFs is higher than 400 K. Figure 11G shows the electron paramagnetic resonance (EPR) spectra of the samples, definitely indicating that The crystallinity of the films increases but the M s decreases with the growth temperature, as shown in Figure 11C Figure 11F indicate that the T c of MoSe 2 TFs is higher than 400 K. Figure 11G shows the electron paramagnetic resonance (EPR) spectra of the samples, definitely indicating that V Se s and V Mo s are both present in samples and suggesting that there is a relationship between the strong RTFM and the V Se s and V Mo s in the MoSe 2 TFs prepared using PAD. The first-principles calculations are used to further understand the relationship between the FM and the vacancy defects. The spin-resolved total density of states (TDOSs) and partial density of states (PDOSs) of Mo 4d electrons and Se 4p electrons, as shown in Figure 11H, suggest that only 2H-MoSe 2 with V Mo s can induce robust magnetism. Figure 11I shows that the relationship between J ij and V Mo -V Mo distances (3.32-13.27 Å) is an RKKY oscillation and the magnetic coupling is always FM for the two-V Mo s system. To sum up, our experimental and theoretical results both show that the strong RTFM of MoSe 2 TFs prepared using PAD mainly originates from the RKKY interactions between the magnetic moments of V Mo s.

ReS 2
Bulk rhenium disulfide (ReS 2 ) is a direct band gap semiconductor [174][175][176] which is pristine and non-magnetic [177,178]. Early research mainly focused on theoretical research, trying to introduce magnetism through defect engineering [177], doping engineering [178,179], and strain engineering. Similar to MoS 2 web buckles with RTFM, we also prepared large-area ReS 2 [4] web buckles in Figure 10. The partial release of compressive strain causes the magnetism to increase to 1.69 times, as shown in Figure 12A. After buckling, M s and M r are increased in Figure 12B,C. However, H c is increased only at 5 K and is decreased at other temperatures in Figure 12D. In addition, the in-plane FM of ReS 2 is weaker than the out-of-plane FM, which is consistent with the ferromagnetic characteristics of 2D materials. The FC and ZFC curves in Figure 12E show that T c is higher than 400 K. The introduction of biaxial tensile strain is accompanied by more defects, which enhances RTFM in Figure 12F. VSes and VMos are both present in samples and suggesting that there is a relationship between the strong RTFM and the VSes and VMos in the MoSe2 TFs prepared using PAD. The first-principles calculations are used to further understand the relationship between the FM and the vacancy defects. The spin-resolved total density of states (TDOSs) and partial density of states (PDOSs) of Mo 4d electrons and Se 4p electrons, as shown in Figure 11H, suggest that only 2H-MoSe2 with VMos can induce robust magnetism. Figure  11I shows that the relationship between Jij and VMo-VMo distances (3.32-13.27 Å) is an RKKY oscillation and the magnetic coupling is always FM for the two-VMos system. To sum up, our experimental and theoretical results both show that the strong RTFM of MoSe2 TFs prepared using PAD mainly originates from the RKKY interactions between the magnetic moments of VMos.

ReS2
Bulk rhenium disulfide (ReS2) is a direct band gap semiconductor [174][175][176] which is pristine and non-magnetic [177,178]. Early research mainly focused on theoretical research, trying to introduce magnetism through defect engineering [177], doping engineering [178,179], and strain engineering. Similar to MoS2 web buckles with RTFM, we also prepared large-area ReS2 [4] web buckles in Figure 10. The partial release of compressive strain causes the magnetism to increase to 1.69 times, as shown in Figure  12A. After buckling, Ms and Mr are increased in Figure 12B,C. However, Hc is increased only at 5 K and is decreased at other temperatures in Figure 12D. In addition, the in-plane FM of ReS2 is weaker than the out-of-plane FM, which is consistent with the ferromagnetic characteristics of 2D materials. The FC and ZFC curves in Figure 12E show that Tc is higher than 400 K. The introduction of biaxial tensile strain is accompanied by more defects, which enhances RTFM in Figure 12F.

NiO
Although 2D zinc oxide [21,29,43,44,71,93] was not successfully grown using PAD, through unremitting efforts, Zou et al. [2] obtained wafer-scale two-dimensional NiO TFs with RTFM in 2021. NiO is an intrinsic p-type semiconductor [180,181], and its bulk is an antiferromagnetism (AFM) [181,182] material. However, the nickel oxide with cation deficiency (Ni 1−x O) [173,183] exhibits FM. Moreover, local deficient Ni at dislocations [184] could also cause FM in antiferromagnetic NiO single-crystal TFs fabricated using PLD. Similarly, NiO crystalline nanoclusters may be associated with RTFM in sputtered TFs [185]. Actually, previous studies have shown that when the size of NiO nanoparticles (NPs) is reduced to nanoscale, the surface of NPs will induce weak FM [184,186,187]. Interestingly, it was also found that M s is proportional to the reciprocal of the size of the nanoparticles [188].
However, the previously prepared NiO NPs [173,182,183,[185][186][187][188][189][190] or TFs [184,185] usually exhibit weak FM, which limits their application in the field of spintronics. Waferscale NiO TFs with RTFM are essential for the construction of spintronic devices. The successful preparation of wafer-scale TFs [3][4][5]16,17,[19][20][21][22]191,192] is the traditional advantage of PAD. By controlling the concentration of Ni-precursors and the spin-coating rate, sub-nanometer NiO TFs [2] were successfully obtained. When the precursor concentration of Ni was 4.396 mg·mL −1 and the spin-coating rate was 8000 rpm, the thickness of the film was about 0.92 nm. The strong FM hysteresis loops could be observed using SQUID. Moreover, M s (23.5 emu·g −1 at RT) was almost proportional to the reciprocal of the thickness of TFs, similar to NPs [188]. The lattice defects on the surface of TFs may cause this abnormal magnetic behavior. As a matter of fact, such films with sub-nanometer thickness are polycrystalline.

Thickness-Dependent RTFM
The controllable thickness of thin films is an important feature of PAD. In detail, the thickness can be controlled by the concentration of precursors [2,26] and the spin-coating rate [2,5]. As shown in Figure 13, M s and bandgap are closely related to the thickness. Furthermore, the total magnetic moment of TFs with different thicknesses seems to be a constant. Interestingly, the thickness can even be adjusted to sub-nanometer level.

NiO
Although 2D zinc oxide [21,29,43,44,71,93] was not successfully grown using PAD, through unremitting efforts, Zou et al. [2] obtained wafer-scale two-dimensional NiO TFs with RTFM in 2021. NiO is an intrinsic p-type semiconductor [180,181], and its bulk is an antiferromagnetism (AFM) [181,182] material. However, the nickel oxide with cation deficiency (Ni1−xO) [173,183] exhibits FM. Moreover, local deficient Ni at dislocations [184] could also cause FM in antiferromagnetic NiO single-crystal TFs fabricated using PLD. Similarly, NiO crystalline nanoclusters may be associated with RTFM in sputtered TFs [185]. Actually, previous studies have shown that when the size of NiO nanoparticles (NPs) is reduced to nanoscale, the surface of NPs will induce weak FM [184,186,187]. Interestingly, it was also found that Ms is proportional to the reciprocal of the size of the nanoparticles [188].
However, the previously prepared NiO NPs [173,182,183,[185][186][187][188][189][190] or TFs [184,185] usually exhibit weak FM, which limits their application in the field of spintronics. Waferscale NiO TFs with RTFM are essential for the construction of spintronic devices. The successful preparation of wafer-scale TFs [3][4][5]16,17,[19][20][21][22]191,192] is the traditional advantage of PAD. By controlling the concentration of Ni-precursors and the spin-coating rate, sub-nanometer NiO TFs [2] were successfully obtained. When the precursor concentration of Ni was 4.396 mg·ml −1 and the spin-coating rate was 8000 rpm, the thickness of the film was about 0.92 nm. The strong FM hysteresis loops could be observed using SQUID. Moreover, Ms (23.5 emu·g −1 at RT) was almost proportional to the reciprocal of the thickness of TFs, similar to NPs [188]. The lattice defects on the surface of TFs may cause this abnormal magnetic behavior. As a matter of fact, such films with sub-nanometer thickness are polycrystalline.

Thickness-Dependent RTFM
The controllable thickness of thin films is an important feature of PAD. In detail, the thickness can be controlled by the concentration of precursors [2,26] and the spin-coating rate [2,5]. As shown in Figure 13, Ms and bandgap are closely related to the thickness. Furthermore, the total magnetic moment of TFs with different thicknesses seems to be a constant. Interestingly, the thickness can even be adjusted to sub-nanometer level.

Outlook
Take the chiral magnetic soliton CrNbS 1/3 (CNS) as an example, which is usually prepared using CVT [73][74][75]156] or ME [71]. Although single-crystal materials can be produced using CVT, its growth time is too long; it often takes several days or even a week. Furthermore, it is difficult to produce centimeter-scale materials. In addition, it is difficult to obtain large-size and controllable thickness samples using ME. So far, the thinnest CNS sample obtained was only 16 nm. Obtaining thinner samples is helpful to reveal the mystery of the disappearance of chiral magnetic solitons. However, PAD can prepare TFs with adjustable thickness, which provides the possibility of completing this task, as seen in Figure 14.
In this review, we summarized the recent developments of 2D ferromagnetic materials fabricated using PAD. Then, we introduced thickness-dependent, strain-modulation, doping, and morphology-dependent RTFM, respectively. Finally, we look forward to the advantages of PAD in the field of chiral magnetic soliton materials. This review provides a novel technique for preparing chiral magnetic soliton materials and may inspire new applications.  [2,3,4,5,25,26,72,73,74,75].
In this review, we summarized the recent developments of 2D ferromagnetic materials fabricated using PAD. Then, we introduced thickness-dependent, strainmodulation, doping, and morphology-dependent RTFM, respectively. Finally, we look forward to the advantages of PAD in the field of chiral magnetic soliton materials. This review provides a novel technique for preparing chiral magnetic soliton materials and may inspire new applications.