Degradation Effect and Magnetoelectric Transport Properties in CrBr3 Devices

Two-dimensional (2D) magnetic materials exhibiting unique 2D-limit magnetism have attracted great attention due to their potential applications in ultrathin spintronic devices. These 2D magnetic materials and their heterostructures provide a unique platform for exploring physical effect and exotic phenomena. However, the degradation of most 2D magnetic materials at ambient conditions has so far hindered their characterization and integration into ultrathin devices. Furthermore, the effect of degradation on magnetoelectric transport properties, which is measured for the demonstration of exotic phenomena and device performance, has remained unexplored. Here, the first experimental investigation of the degradation of CrBr3 flakes and its effect on magnetoelectric transport behavior in devices is reported. The extra magnetic compounds derived from oxidation-related degradation play a significant role in the magnetoelectric transport in CrBr3 devices, greatly affecting the magnetoresistance and conductivity. This work has important implications for studies concerning 2D magnetic materials measured, stored, and integrated into devices at ambient conditions.

However, most 2D magnetic materials suffer from degradation in the ambient environment, which originates from the sensitivity to O 2 and H 2 O [13][14][15]. There are also photocatalytic degradation and photochemical or photothermal oxidation-related degradation in the presence of oxygen and moisture, closely depending on various factors such as light exposure and temperature [13,16]. For chromium trihalides (CrX 3 , X = I, Br, Cl), chromium bromide (CrBr 3 ) and chromium chloride (CrCl 3 ) are more stable than chromium Materials 2022, 15, 3007 2 of 9 triiodide (CrI 3 ), and this condition is related to the stability of lone-pair electrons of the halogen element (X) atoms that terminate the surfaces of layers. CrI 3 shows the first visible signs of degradation within seconds under air and light conditions under an optical microscope [13]. It slowly degrades in ambient environment with the coverage of Al 2 O 3 , PMMA, or h-BN. This result also depends on light exposure and temperature, of which the former is related to the photocatalytic substitution of iodine by water. The fragility of 2D magnetic materials to ambient conditions hinders their characterization and integration into ultrathin devices that are functionalized by interfacial effect, spin-orbit torque-driven magnetization switching, exotic phenomena, and so on [3,10,11,17]. Some studies have focused on avoiding the degradation or improving the air stability of 2D magnetic materials in ambient environment [18][19][20][21][22][23]. Generally, to avoid or reduce degradation, the exfoliation of these 2D magnetic materials and their heterostructures stacking process are carried out in a glove box under inert atmosphere. Whole 2D magnetic material-based devices can be encapsulated by hexagonal boron nitride (h-BN) before being taken out. As a typical example, graphene/CrI 3 /graphene spin-filter magnetic tunnel junctions were successfully fabricated by encapsulating h-BN, and they showed a tunneling magnetoresistance ratio of 19,000% [23]. Encapsulation of 2D materials with exfoliated h-BN single crystal flakes has been widely used in 2D material research [8,12,18,23]. However, 2D magnetic materials that are deposited or exfoliated on SiO 2 /Si substrates still degrade when covered by Al 2 O 3 , PMMA, or h-BN capping films [13], due to moisture and oxygen leakage through the capping film/SiO 2 and the 2D magnetic material/SiO 2 interfaces. In addition, great efforts have been devoted to improving the air stability of 2D magnetic materials from synthesis and modification strategies [14,21,22,24], such as atomic layer deposition passivation [21], passivation from air oxidation by thiol molecules [24], and growing air-stable CrSe 2 nanosheets on a dangling-bond-free WSe 2 substrate [22].
Understanding the stability of 2D magnetic materials, the degradation effect on device performance, and the derived physical phenomena at ambient conditions is crucial for their characterization and applications in ambient conditions. Thus far, a limited number of studies have investigated the degradation effect of 2D magnetic materials on device performance and the derived physical phenomena [25]. Furthermore, the effect of degradation on magnetoelectric transport behavior (such as magnetoresistance), which is usually measured in a cryogenic transport measurement system for demonstrating exotic phenomena and device performance, remains unexplored.
In this work, the degradation of CrBr 3 , a ferromagnetic semiconductor 2D material that has been extensively studied recently, was investigated in an ambient environment by optical microscopy. The X-ray photoelectron spectroscopy (XPS) spectra demonstrated extra compounds derived from oxidation-related degradation on the CrBr 3 surface. Furthermore, the degradation effect on magnetoelectric transport behavior in CrBr 3 devices was investigated by the analysis of temperature-dependent Hall magnetoresistance and conductivity in CrBr 3 devices.

Characterization and Measurements
X-ray photoelectron spectroscopy (XPS) data were acquired on PHI Quantera II XPS spectrometer employing Al Kα radiation. The XPS binding energies were calibrated using the C 1s peak (284.5 eV). Magnetic measurements of CrBr 3 bulk crystals and temperaturedependent magnetoelectric transport measurements of CrBr 3 devices were performed in a physical property measurement system (PPMS DynaCool, Quantum Design). The Hall magnetoresistance was obtained using DC measurement and AC measurement, in order to confirm the repeatability of data. For AC measurements by the lock-in technique, the AC current was provided by the Keithley 6221 current source, and the voltage signal was measured by Model SR830 DSP lock-in amplifiers. For DC measurements, the Keithley 2400 SourceMeter and Keithley 2182A Nanovoltmeter were used to provide the current and to detect the voltage, respectively.

Results and Discussion
CrBr 3 is a layered ferromagnetic semiconductor with out-of-plane magnetic order below the Curie temperature (T C ). CrBr 3 crystallizes in a lamellar structure with trigonal space group R3 (148), in which Cr 3+ ions arrange in the honeycomb magnetic lattice and are surrounded by edge-sharing octahedra formed by Br − ions [28,29]. As shown in Figure 1a, the Br-Cr-Br layers are stacked along the c-axis and stacked together via van der Waals force. The magnetic moments defined by Cr layers are oriented perpendicular to the ab plane and aligned ferromagnetically within each layer, and the layers are also ferromagnetically coupled. The magnetic properties of a CrBr 3 platelet were characterized by a vibrating sample magnetometer in PPMS. As shown in Figure 1b, zero-field-cooled (ZFC) and field-cooled (FC) magnetization and their first derivatives exhibited a paramagnetic-ferromagnetic phase transition at T C = 33 K, nearly close to the T C obtained by temperature-dependent M-H curves ( Figure S1), and it is also in good accordance with the T C = 32-40 K of CrBr 3 observed in previous studies [28][29][30].
nm thick SiO /Si substrates. The pre-patterned Pt (6 nm) electrodes on SiO /Si substrates were fabricated through micro-nano fabrication processes. Finally, two types of CrBr devices were prepared by transferring CrBr flake onto pre-patterned Pt electrodes using a poly(bisphenol A carbonate) (PC)/polydimethylsiloxane (PDMS) polymerbased transfer method [27].

Characterization and Measurements
X-ray photoelectron spectroscopy (XPS) data were acquired on PHI Quantera II XPS spectrometer employing Al Kα radiation. The XPS binding energies were calibrated using the C 1s peak (284.5 eV). Magnetic measurements of CrBr bulk crystals and temperaturedependent magnetoelectric transport measurements of CrBr devices were performed in a physical property measurement system (PPMS DynaCool, Quantum Design). The Hall magnetoresistance was obtained using DC measurement and AC measurement, in order to confirm the repeatability of data. For AC measurements by the lock-in technique, the AC current was provided by the Keithley 6221 current source, and the voltage signal was measured by Model SR830 DSP lock-in amplifiers. For DC measurements, the Keithley 2400 SourceMeter and Keithley 2182A Nanovoltmeter were used to provide the current and to detect the voltage, respectively.

Results and Discussion
CrBr is a layered ferromagnetic semiconductor with out-of-plane magnetic order below the Curie temperature ( ). CrBr crystallizes in a lamellar structure with trigonal space group 3 (148), in which Cr ions arrange in the honeycomb magnetic lattice and are surrounded by edge-sharing octahedra formed by Br ions [28,29]. As shown in Figure 1a, the Br-Cr-Br layers are stacked along the c-axis and stacked together via van der Waals force. The magnetic moments defined by Cr layers are oriented perpendicular to the ab plane and aligned ferromagnetically within each layer, and the layers are also ferromagnetically coupled. The magnetic properties of a CrBr platelet were characterized by a vibrating sample magnetometer in PPMS. As shown in Figure 1b, zero-fieldcooled (ZFC) and field-cooled (FC) magnetization and their first derivatives exhibited a paramagnetic-ferromagnetic phase transition at = 33 K, in good accordance with the = 32-40 K of CrBr observed in previous studies [28][29][30].  The chemical composition of CrBr 3 flakes on SiO 2 /Si substrates was analyzed by XPS. The CrBr 3 flakes were mechanically exfoliated on SiO 2 /Si substrates in the air. Figure 2a shows the XPS spectra of two CrBr 3 flakes on SiO 2 /Si substrates, and the inset presents the optical images of the corresponding CrBr 3 flakes. The expected elemental peaks of Cr 2p, Br 3d, Si 2p, C 1s, and O 1s were observed quantitatively. The SiO 2 /Si substrate contributed to Si 2p and O 1s peaks. For CrBr 3 flakes, the binding energies of the Br 3d 5/2 and Br 3d 3/2 core level were located at 68.9 and 69.9 eV, respectively ( Figure 2b). In addition, the two peaks located at 576.1 and 585.7 eV corresponded to Cr 2p 3/2 and Cr 2p 1/2 (Figure 2c), respectively. However, an additional peak at 579.1 eV could be seen, most likely related to chromium oxides (such as CrO 2 and Cr 2 O 3 ) [31]. This additional Cr 2p 3/2 peak centered at 579.1 eV could be fitted to obtain a large FWHW of 6.6 eV, associated with the mixture of chromium oxides and other compounds. In addition, the main features of the XPS spectra were identical with other CrBr 3 flakes on SiO 2 /Si substrates but differed in the content of additional compounds. Our XPS results are in good agreement with the oxidation-related degradation resulting in the formation of chromium oxide [32]. ure 2a shows the XPS spectra of two CrBr flakes on SiO /Si substrates, and the inset presents the optical images of the corresponding CrBr flakes. The expected elemental peaks of Cr 2p, Br 3d, Si 2p, C 1s, and O 1s were observed quantitatively. The SiO /Si substrate contributed to Si 2p and O 1s peaks. For CrBr flakes, the binding energies of the Br 3d5/2 and Br 3d3/2 core level were located at 68.9 and 69.9 eV, respectively ( Figure 2b). In addition, the two peaks located at 576.1 and 585.7 eV corresponded to Cr 2p3/2 and Cr 2p1/2 (Figure 2c), respectively. However, an additional peak at 579.1 eV could be seen, most likely related to chromium oxides (such as CrO and Cr O ) [31]. This additional Cr 2p3/2 peak centered at 579.1 eV could be fitted to obtain a large FWHW of 6.6 eV, associated with the mixture of chromium oxides and other compounds. In addition, the main features of the XPS spectra were identical with other CrBr flakes on SiO /Si substrates but differed in the content of additional compounds. Our XPS results are in good agreement with the oxidation-related degradation resulting in the formation of chromium oxide [32]. The XPS results indicate the extra compounds derived from degradation on the CrBr surface, although no obvious degradation could be observed under optical microscopy after storage in ambient atmosphere for 2 days. Furthermore, the degradation process of CrBr flakes on the SiO /Si substrate was monitored by optical microscopy in ambient atmosphere (Figure 3). The degradation of CrBr flakes was slow in ambient atmosphere, and they showed no significant change after storage in ambient atmosphere for 15.6 h (Figure 3a-c). Subsequently, dark-gray bubbles or droplets cluttered on the edges, and the surfaces of flakes gradually increased under light irradiation (Figure 3df). This irradiation was directly provided by the light emitted from the 50× objective of an The XPS results indicate the extra compounds derived from degradation on the CrBr 3 surface, although no obvious degradation could be observed under optical microscopy after storage in ambient atmosphere for 2 days. Furthermore, the degradation process of CrBr 3 flakes on the SiO 2 /Si substrate was monitored by optical microscopy in ambient atmosphere (Figure 3). The degradation of CrBr 3 flakes was slow in ambient atmosphere, and they showed no significant change after storage in ambient atmosphere for 15.6 h (Figure 3a-c). Subsequently, dark-gray bubbles or droplets cluttered on the edges, and the surfaces of flakes gradually increased under light irradiation (Figure 3d-f). This irradiation was directly provided by the light emitted from the 50× objective of an optical microscope during the whole measurement. The above phenomena demonstrate that the degradation of CrBr 3 flakes was moderate but became faster under light exposure. Furthermore, the degradation was accelerated by a reduction in flake thickness (Figure 3 and Figure S2 in Supplementary Materials). The degradation process for the CrBr 3 flakes with different thicknesses were also monitored using another optical microscope that could provide a stronger light irradiation from a 100× objective. As shown in Figure 3g-i, some tiny and gray droplets mainly formed at the flake edges and grain boundaries, growing rapidly under stronger light irradiation. Then, the droplets merged into larger ones, which spread from the flake edges and finally covered the majority of the flake surface until the flakes were completely decomposed. A short decomposition time of several minutes was observed due to the stronger light irradiation and higher temperature. Furthermore, the degradation was accelerated by a reduction in flake thickness (Figures 3  and S2 in Supplementary Materials). The degradation process for the CrBr flakes with different thicknesses were also monitored using another optical microscope that could provide a stronger light irradiation from a 100× objective. As shown in Figure 3g-i, some tiny and gray droplets mainly formed at the flake edges and grain boundaries, growing rapidly under stronger light irradiation. Then, the droplets merged into larger ones, which spread from the flake edges and finally covered the majority of the flake surface until the flakes were completely decomposed. A short decomposition time of several minutes was observed due to the stronger light irradiation and higher temperature. Oxygen, humidity, and light exposure are crucial factors that affect the air stability of 2D magnetic materials [13][14][15]. Moreover, the size, thickness, starting condition, and quality of each flake, even the substrate, play a particular role in the degradation, which are difficult to be quantitatively evaluated in ambient atmosphere. The observation of scanning electron microscope images further confirmed that defects, such as grain boundaries, present a dominant contribution to material degradation, acting as the sources of inducement and propagation [25]. This condition can be used to explain where the bubbles or droplets appear firstly. Oxygen, humidity, and light exposure are crucial factors that affect the air stability of 2D magnetic materials [13][14][15]. Moreover, the size, thickness, starting condition, and quality of each flake, even the substrate, play a particular role in the degradation, which are difficult to be quantitatively evaluated in ambient atmosphere. The observation of scanning electron microscope images further confirmed that defects, such as grain boundaries, present a dominant contribution to material degradation, acting as the sources of inducement and propagation [25]. This condition can be used to explain where the bubbles or droplets appear firstly.
A tentative explanation for degradation is the reaction with ambient O 2 and H 2 O, as well as the subsequent formation of intermediate chromium oxide bromides, chromium oxides or aqua chromium halide [13,32]. According to previous studies of CrI 3 [13], water reacts with CrBr 3 to possibly form partially hydrated CrBr 3−x (H 2 O) x+ x under ambient conditions, and the reactivity can be accelerated by light irradiation. Thus, the gray droplets on the CrBr 3 flakes shown in Figure 3 and Figure S2 are related to photocatalytic degradation, in which the main degradation pathway is the photocatalytic substitution of bromine by water. In addition, O 2 can react with CrBr 3 to form chromium oxides at high temperatures, similar to the case of CrI 3 where degradation causes the formation of Cr 2 O 3 at 180 • C, resulting in teal-colored CrI 3 thin flakes [13].
To investigate the degradation effect on the magnetoelectric transport properties of CrBr 3 and its heterostructure devices, two kinds of CrBr 3 devices were fabricated, namely, Materials 2022, 15, 3007 6 of 9 devices A and B. Device A was fabricated by transferring a CrBr 3 flake on pre-patterned Pt Hall bar electrodes (inset in Figure 4a). Device B was fabricated by transferring a CrBr 3 flake on Pt Hall bar electrodes without a crossing area (inset in Figure 4d). Device A could be used to study the magnetic proximity effect, spin Hall magnetoresistance, and current-induced magnetization switching in the CrBr 3 /Pt heterostructure. As shown in Figure 4a,b, device A showed a remarkable Hall magnetoresistance (R H ) curve under out-of-plane magnetic fields (H ⊥ ) and in-plane magnetic fields (H // ), similar to the ferromagnetic behavior, which was notably larger than that for the reference sample (only Pt Hall bar device). Figure 4c shows the temperature dependence of R H in device A. The ferromagnetic-like behavior persisted at 110 K, which is greatly higher than the T C of 32-40 K for CrBr 3 . Therefore, the origin of the anomalous Hall effect (AHE) in semiconductor CrBr 3 film or magnetized Pt film can be excluded because the production of a remarkable R H above Tc is impossible. The R H in device A was most likely produced by the AHE of extra magnetic compounds derived from the degradation of CrBr 3 . To further clarify the origin of R H in device A, the R H of the CrBr 3 flake on device B was measured. As shown in Figure 4d,e, device B featured R H curves under H ⊥ and H // . As shown in Figure 4d, the R H change of~1 Ω was obtained when H ⊥ changed from +0.04 T to −0.04 T, and the R H curve contained nearly linear magnetoresistance under positive and negative H ⊥ . It is worth pointing out that linear I-V curves of CrBr 3 flakes were observed in device B (inset in Figure 4d), indicating the metallic characteristic of the "CrBr 3 flake". In addition, the same R H curve was observed for device A under H ⊥ and device B under H // , (Figure 4a,e). Figure 4f shows the R H curve maintained at 100 K, which is greatly higher than the T C of CrBr 3 . These results on device B further confirmed the extra magnetic compounds derived from the degradation of the CrBr 3 flake. However, determining specific magnetic compounds from these magnetoelectric measurements is difficult due to multiple compounds included in the degraded CrBr 3 . Furthermore, degradation can be judged by the conductivity or resistance of the CrBr 3 flake. In device A, the resistance of the CrBr 3 flake was about 4662 Ω according to the I-V curves and parallel resistance formula R A +R Pt R A ·R Pt . In device B, the resistance of the CrBr 3 flake was in the range of 1000-2000 Ω according to the I-V curves (inset in Figure 4d). Thus, there was a resistance of~10 3 Ω for extra magnetic compounds derived from degradation.
The CrBr 3 flakes used for devices A and B were stored for 6 and 2 months in a low-vacuum environment (~10 −1 Pa), respectively, before transferring to pre-patterned electrodes. The above magnetoelectric measurement results confirmed that teal-colored CrBr 3 thin flakes without droplets (insets in Figure 4a,d) underwent oxidation-related degradation in which the main degradation pathway was thermocatalytic conversion to chromium oxides in an atmosphere with O 2 . The degradation occurred mainly during the transfer process, in which the polymer-based transfer medium was melted on the CrBr 3 flake at 155 • C. The magnetoelectric measurement results agree well with the above XPS results, indicating the formation of chromium oxides on CrBr 3 thin flakes without droplets. Therefore, in addition to the absence of oxygen and moisture, high-temperature treatment should be avoided while fabricating devices based on air-sensitive 2D magnetic materials, to minimize the influence of degradation on performance and the derived physical phenomena in these devices.
For the CrBr 3 flakes without degradation, the R H and linear I-V curves were difficult to obtain due to the semiconductor characteristics. To minimize the degradation, we exfoliated the thicker CrBr 3 flakes and immediately transferred them onto the Pt Hall bar electrodes for measurements. Device C was prepared by transferring a fresh and thicker CrBr 3 flake on Pt Hall bar electrodes without a crossing area (Figure 5a). No linear I-V curve was obtained for device C (Figure 5b), indicating the poor electrical conductivity of semiconductor CrBr 3 without considerable degradation. lets. Therefore, in addition to the absence of oxygen and moisture, high-temperature treatment should be avoided while fabricating devices based on air-sensitive 2D magnetic materials, to minimize the influence of degradation on performance and the derived physical phenomena in these devices. For the CrBr flakes without degradation, the RH and linear I-V curves were difficult to obtain due to the semiconductor characteristics. To minimize the degradation, we exfoliated the thicker CrBr flakes and immediately transferred them onto the Pt Hall bar electrodes for measurements. Device C was prepared by transferring a fresh and thicker CrBr flake on Pt Hall bar electrodes without a crossing area (Figure 5a). No linear I-V curve was obtained for device C (Figure 5b), indicating the poor electrical conductivity of semiconductor CrBr without considerable degradation.
Our study demonstrates that extra magnetic compounds mainly derived from oxidation-related degradation significantly affected the magnetoresistance and conductivity of CrBr devices, revealing the non-negligible effect of degradation on the magnetoelectric transport behavior of CrBr thin flakes and their heterostructure devices. For bulk platelets, surface degradation showed a remarkable effect on the magnetic properties of bulk VI [25]. Therefore, understanding the degradation process and its outcomes on magnetoelectric transport behavior in our study is necessary for the interpretation of the experimental results for functional devices based on air-sensitive 2D magnetic materials.

Conclusions
In summary, CrBr devices were successfully fabricated to investigate the degradation effect of CrBr flakes on the magnetoelectric transport properties. The extra compounds derived from oxidation-related degradation in CrBr were confirmed by XPS spectra, significantly affecting the magnetoresistance and conductivity in CrBr devices with remarkable Hall magnetoresistance and the conductive characteristic of the de- Our study demonstrates that extra magnetic compounds mainly derived from oxidationrelated degradation significantly affected the magnetoresistance and conductivity of CrBr 3 devices, revealing the non-negligible effect of degradation on the magnetoelectric transport behavior of CrBr 3 thin flakes and their heterostructure devices. For bulk platelets, surface degradation showed a remarkable effect on the magnetic properties of bulk VI 3 [25]. Therefore, understanding the degradation process and its outcomes on magnetoelectric transport behavior in our study is necessary for the interpretation of the experimental results for functional devices based on air-sensitive 2D magnetic materials.

Conclusions
In summary, CrBr 3 devices were successfully fabricated to investigate the degradation effect of CrBr 3 flakes on the magnetoelectric transport properties. The extra compounds derived from oxidation-related degradation in CrBr 3 were confirmed by XPS spectra, significantly affecting the magnetoresistance and conductivity in CrBr 3 devices with remarkable Hall magnetoresistance and the conductive characteristic of the degraded CrBr 3 flakes. Our study elucidates the effect of degradation on performance and the derived physical phenomena in devices based on air-sensitive 2D magnetic materials, offering practical guidance on the characterization and applications of 2D magnetic materials in ambient conditions. Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ma15093007/s1, Figure S1: The magnetization of the CrBr 3 bulk platelet as a function of in-plane magnetic field in the temperature regime from 3 to 100 K.; Figure S2: (a,b) Optical microscopy images of the CrBr 3 flakes exposed to a stronger light irradiation in air for 0 min and 4 min, respectively.