Ferromagnetism in two-dimensional CrTe 2 epitaxial films down to a few atomic layers

Two-dimensional (2D) van der Waals ferromagnetic materials have attracted intense attention due to their potential impact on both fundamental and applied research studies. Recently, a new 2D ferromagnet CrTe 2 , prepared by mechanical exfoliation or chemical vapor deposition, has gained interest due to its novel magnetic properties. In this work, high quality CrTe 2 epitaxial thin ﬁlms were prepared on GaAs (111)B substrates using solid source molecular beam epitaxy, with the thickness varying from 35 to 4 monolayers (MLs). The magnetic easy axis of all the ﬁlms is oriented along the c-axis. A Curie temperature of 205 K is found in the 35 ML CrTe 2 ﬁlm, measured by the temperature-dependent anomalous Hall resistance ( R AHE ). Importantly,

materials have attracted extensive attention 1 owing to their remarkable properties in diverse experimental fields, such as electronics, 2,3 optoelectronics, 4-6 magnetism, 7 thermoelectrics, 8 and spintronic devices. 9 Among these, intrinsic 2D magnetic materials, such as CrI 3 , 10 Fe 3 GeTe 2 , 11,12 and Cr 2 Ge 2 Te 6 , 13 have aroused great interest not only due to their fascinating magnetic properties in the 2D limit but also for opening new avenues for applications in spintronics. [13][14][15] Recently, a new 2D ferromagnetic material CrTe 2 has been found. 16,17 Freitas et al. reported that they had synthesized the metastable compound 1T-CrTe 2 for the first time. 16 Purbawati et al. reported thin CrTe 2 flakes with a Tc above room temperature. 18 Sun et al. found that the ferromagnetism could hold above room temperature in a metallic phase of 1T-CrTe 2 down to the ultra-thin limit. 19 However, all of the reported CrTe 2 are prepared in bulk forms, and the devices are made by mechanical exfoliation and stacking techniques. 12,[19][20][21] The size and thickness of 2D materials obtained by exfoliation are uncontrolled, which prevents them being applied in actual production. Thus, the preparation of the large-scale samples with high quality is of great significance.
In this work, a series of CrTe 2 thin films with various thicknesses from 35 to 4 monolayers (MLs) were epitaxially grown on high resistivity (≈10 MΩ cm) GaAs (111)B substrates using molecular beam epitaxy (MBE). The base pressure of the growth chamber is below 2 × 10 −10 Torr. The GaAs (111)B substrates were cleaned under a standard procedure before being loaded into the chamber. 22 Then, the substrates were annealed at 580 ○ C in the growth chamber until the streak patterns appeared as monitored by the real-time RHEED, 23 as shown in Fig. 1(b). High-purity Cr (99.999%) and Te (99.9999%) were evaporated by conventional effusion cells. During the growth, the Cr effusion cell was kept at 1180 ○ C and the Te effusion cell was kept at 255 ○ C to keep the flux ratio of Cr/Te at 1/20 to maintain a Te-rich environment. The GaAs (111)B substrates were set at 265 ○ C during the growth for all the samples.   18 while between the layers, the interactions are mainly of the van der Waals type. 24 Figures 1(b) and 1(c) show representative RHEED patterns of the GaAs (111)B substrate and the CrTe 2 , respectively. Sharp streaky patterns were observed after growth, which indicates the flat surface morphology of the films. The doubleheaded arrows between the two first order stripes represent the d-spacing, 25 which is inversely proportional to the in-plane lattice constant. For GaAs (111)B, the atoms are close-packed in the (111) plane, which makes its in-plane lattice constant to be aGaAs/ √ 2 of ∼4.0 Å, as shown by the black dashed line in Fig. 1(d). From this, the lattice constant of the films can be extracted to be 3.81 Å, which is very close to the theoretical value of 3.79 Å, 18,26 as shown by the red dashed line in Fig. 1(d). Figure 1(d) illustrates the evolution of the lattice constant of CrTe 2 film during growth. Soon after the beginning of the growth, the lattice constant of the film decreases from the GaAs substrates and reaches CrTe 2 after the first layer, as indicated by the blue dashed line. This suggests that the films have completely released the strain from the substrate after the growth of the first layer, which is consistent with the van der Waals epitaxial growth mode. 27 Figure 2(a) shows a typical atomic force microscopy (AFM) (NT-MDT, INTEGRA SPECTRA II, Probe aperture size ∼100 nm) image of an as-grown CrTe 2 film. Hexagonal terraces can be observed with a typical size of ∼200 nm, reflecting the hexagonal crystal structure inside the (0001) plane. 28 Figure 2(b) displays the height profile of the dashed line in Fig. 2(a). The height is equal to ∼6.1 Å, consistent with a CrTe 2 ML thickness. 16,26 The phase purity and crystal structure of the CrTe 2 thin film have been identified by x-ray diffraction (Bruker D8 Discover single crystal diffractometer, λ ≙ 1.5406 Å) and the spectra are shown in Fig. 2(c). Compared with the Joint Committee on Powder Diffraction Standards (JCPDS) data of CrTe 2 , the film has been found to exhibit rhombohedral crystal geometry with no other detectable phases. 24 The lattice parameter c was calculated as 6.144 Å according to the position of the XRD peaks, which is consistent with the value of 6.166 Å by the theoretical calculation. 19,24 After growth, the samples are patterned into standard Hall bar devices using Ion Beam Etching (IBE) with Ar gas. The magnetic properties of the samples were examined by the magnetotransport measurements inside an Oxford instruments' cryogenic system (TeslatronPT), as shown in Fig. 3(a). A Keithley 6221 AC/DC current source was used to apply an AC of 1uA with a frequency of 13 Hz. Lock-in amplifier SR830 was employed to obtain the longitudinal and transverse voltage signals, while the magnetic field up to 2 T was scanned back and forth perpendicular to the sample surface.
Generally, there is a close correlation between magnetism and transport in magnetic materials, and the anomalous Hall effect (AHE) is utilized to demonstrate the prevailing ferromagnetism. The Hall resistance (Rxy) follows an empirical relation 29 Rxy ≙ R 0 H + RAM(H), where the first term is the ordinary Hall effect and the second is the AHE with respective coefficients R 0 and RA. 30 Here, R 0 is the ordinary Hall coefficient and is solely connected to the carrier density, and RA is proportional to the longitudinal resistance Rxx or Rxx 2 depending on the dominant extrinsic scattering mechanisms. 31 By subtracting the linear ordinary Hall component R 0 H, the anomalous Hall data RAHE ≙ Rxy-R 0 H of 35, 24, and 4 ML CrTe 2 films were plotted at different temperatures in Figs. 3 respectively. For the 35 ML CrTe 2 film, a square hysteresis loop can be observed at low temperatures, suggesting its ferromagnetism and the out-of-plane easy axis. As the temperature rises, the saturated RAHE first increases and reaches the maximum at about 150 K, then decreases, and finally disappears around 200 K, as shown in Fig. 4(a). The coercive field (Hc) decreases and eventually vanished around 200 K. Similar behaviors are also observed in the 24 ML sample, as shown in Fig. 3(c).
For the 4 ML CrTe 2 film, as the temperature increases, the saturated RAHE first decreases from positive to negative at 80 K, reaches its minimum at 150 K, then increases again, and finally disappears at around 190 K. This strange temperature dependent saturated RAHE, only found at very thin samples, may be associated with the change of their intrinsic anomalous Hall coefficient, which is related to their Berry curvatures. 32 On the other hand, its coercive field (Hc) The evolution of the Curie temperature T C (red) and the coercive field H C (blue) at 5 K vs CrTe 2 thicknesses. The Curie temperature raises with the increasing film thicknesses, while the coercivity correspondingly decreases.
decreases monotonically with the increased temperature and reaches zero at 191 K. The temperature dependent saturated RAHE of the 35 ML CrTe 2 is plotted in Fig. 4(a), and the Curie temperature of 202 K can be determined as it reaches zero. 33 Figure 4(b) shows the HC as a function of temperature. As the temperature increases, HC drops and reaches zero at 203 K. In addition, the method of the Arrott-plot has been adopted to further confirm the ferromagnetic transition. 34 Here, we take the assumption of side jump mechanism at high fields and use the ratio of (R xy/R 2 xx ) to estimate the magnetization M. 35,36 Thus, the Arrott-plot can be plotted as where B is magnetic induction, 37 as shown in Fig. 4(c). It is well known that the intercept on the y-axis of the extrapolated line at high field is positive for the ferromagnetic state and negative for the paramagnetic state. 38 When the intercept goes to zero, the Curie temperature TC of 204.6 ± 0.8 K can be determined, 39 as shown in the inset of Fig. 4(c). All three methods presented here give a consistent TC of ∼205 K for the 35 ML CrTe 2 films. The TC for all other prepared samples has been determined similarly, as shown in Fig. 5(b). It decreases from 205 to 191 K, as the film thickness decreases from 35 to 4 MLs. Figure 5(a) presents the RAHE of various CrTe 2 thin films at 5 K. All the samples exhibit similar square hysteresis loops, demonstrating strong ferromagnetism with an out-of-plane easy axis even for the 4 ML sample. 40 As shown in Fig. 5(b), the HC increases as the sample thickness decreases and reaches 8200 Oe at 4 ML. This increase in coercivity is probably due to the increase in shape anisotropy originating from magnetostatic or dipole interactions. 41 In summary, a series of CrTe 2 thin films with various thicknesses have been synthesized by MBE. All the films show strong ferromagnetism with an out-of-plane easy axis. More importantly, as the film thickness decreases from 35 to 4 ML, the Curie temperature decreases only slightly from 205 to 191 K, while the coercive field (HC) increases from 5120 to 8200 Oe. With distinct magnetic properties, large area CrTe 2 thin films grown by MBE are promising for stacking with other 2D materials and suitable for all van der Waals material electronic devices.
See the supplementary material for RHEED patterns of the CrTe 2 thin films with different thicknesses, RHEED intensity oscillations during growth, x-ray diffraction methods to calculate the lattice parameter c, and RAHE of all the CrTe 2 films vs magnetic field at various temperatures.

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