Magnetic and structural properties of antiferromagnetic Mn 2 VSi alloy films grown at elevated temperatures

Eighty nano meters thick polycrystalline Mn 2 VSi films have been deposited on silicon substrates with an 18 nm silver seed layer and a 3 nm aluminium capping layer using a sputtering system. The best quality film is obtained for 723 K growth. The Mn 2 VSi thin film is verified to be antiferromagnetic, where an exchange bias is found when a 3 nm ferromagnetic CoFe layer has been deposited on the top of the Mn 2 VSi layer. The exchange bias is measured to be 34 Oe at 100 K. The blocking and thermal activation temperature of Mn 2 VSi is estimated to be below 100 K and within a range between 100 K and 448 K, respectively. These properties can be improved by substituting the constituent atoms with the other elements (e.g. Co and Al), suggesting a potential of Mn 2 VSi to be used as an antiferromagnet in a spintronic device. 10.1088/1361-6463/aa80d5

(Some figures may appear in colour only in the online journal) Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. a seed layer, and was capped by an Al layer [8] as shown in figure 1. These layers and the Mn 2 VSi thin films (25-80 nm thick) with a Co 0.6 Fe 0.4 layer (0 or 3 nm thick) were deposited using a PlasmaQuest high target utilization sputtering (HiTUS) system. The plasma was generated by a radio frequency (RF) field of 13.56 MHz in an Ar atmosphere of 3 × 10 -1 Pa [7]. The plasma was focused onto a sputtering target using bias voltages between 250 V and 990 V. The deposition rate of the Mn 2 VSi films is typically 0.07 nm s −1 .
During the Mn 2 VSi deposition the substrates were heated to the temperatures between 523 K and 923 K. This was achieved by a heater lamp (Philips, Halogen 100 W) installed in the chamber directly above the back of the substrates with a separation of about 7 mm. The power supply (ALDA, Model 00160141) controls the temperature up to about 1000 K. A temperature sensor was placed on the heater housing which was connected to the thermometer. Ceramic rings were used to hold the substrate for thermal insulation to avoid the substrate holders from being melted. The overall performance of the heater was stable with a temperature fluctuation within ±12 K. A set of experiments were carried out to evaluate temperature effects during and after the alloy growth onto the crystallization using vibrating sample magnetometer (VSM, ADE, Model 10) and x-ray diffraction (XRD, Rigaku, SmartLab) with a Cu Kα as main x-ray source for structure analysis.
The θ−2θ XRD scan indicates the variation in the crystallization as shown in figure 2. Samples which were sputtered at the higher temperature (723 K) are found to form a significantly stronger Mn 2 VSi(2 2 0) peak at about 46.1° than those grown at lower temperature (<723 K) (see figure 2(a)). This confirms the formation of the A2 phase. In particular, the films grown at 723 K provided the strongest Mn 2 VSi(2 2 0) Heusler peak, whose full width half maximum (FWHM) is measured to be 0.396° for 80 nm thick case. This is almost a half in FWHM as compared with the other samples, such as those grown at 523 K (FWHM = 1.27°). Several satellite peaks near Si(4 0 0) are also seen in the figures due to multi x-ray sources (Kα, Kβ) with slightly different wavelengths.
Post-annealing was also used to compare with the heated growth method as described above. Here, the samples were grown at room temperature using HiTUS. The sputtered films  were then placed into an annealing furnace (Carbolite, MTF 10/25/130, maximum temperature: 1273 K) for post-annealing at elevating temperatures between 573 and 923 K for 3 h under 8 × 10 −5 mbar. The XRD results for the post-annealed samples are shown in figure 2(b). The Mn 2 VSi(2 2 0) peak at about 46.1° is much less significant (FWHM = 0.52° for 873 K, for example, as listed in table 1) than those for the films grown at high temperature. In figure 2(c), a structural comparison is shown between 25 and 80 nm thick Mn 2 VSi films. Here, 80 nm-thick Mn 2 VSi films show the Mn 2 VSi(2 2 0) peak, confirming the formation of the A2 phase.
The corresponding x-ray reflectivity (XRR) results are shown in figure 3. For the post-annealing films, weak oscillations indicate that interface between layers are rough. This is because during the post-annealing process the whole sample is heated, causing the diffusion at the interface between each layer. The 25 nm thick Mn 2 VSi film grown at 723 K shows the smoothest interfaces. By fitting the result, the estimated thicknesses are Cr (2.80 ± 1.80 nm)/Ag (17.23 ± 1.16 nm)/ Mn 2 VSi (24.00 ± 0.26 nm)/CoFe (3.7 ± 0.18 nm)/Al (2.28 ± 0.20 nm). On the other hand, the 80 nm thick Mn 2 VSi film grown at 723 K shows rougher interface. These results indicate that the thinner films minimizes the interfacial diffusion during the deposition. The increased roughness in the 80 nm thick Mn 2 VSi films show increase in the corresponding corecivity in the CoFe film to be 200 Oe as compared with 114 Oe for the 25 nm thick Mn 2 VSi films. However, this did not affect the exchange bias as the degradation of the antiferromagnetic ordering in the Mn 2 VSi film with increasing thickness dominates the magnetic properties. However the thinner samples show less crystallinity (no (2 2 0) peak shown at 46.1° for 25 nm) as compared with the thicker layer (FWHM = 0.396° for 80 nm), as shown in figure 2(c). This is because the thickness of Mn 2 VSi is too thin for the XRD signal to be detected. Therefore, we used the heated growth for magnetic characterization and measured the thickness dependence of their magnetic properties.
In order to verify if the Mn 2 VSi thin films are antiferromagnetic, a 3 nm ferromagnetic CoFe layer was deposited on the top of the Mn 2 VSi layer. When materials with ferromagnetic (F) ordering and antiferromagnetic (AF) ordering are in close contact and are field-cooled below the Neel temperature (T N ) of the AF, an exchange bias is induced in the F layer. The exchange bias introduces a shift in the hysteresis loop along the field axis, which is in the opposite direction of the setting field. The loops also increase their coercivities H C with the increase in the thickness of AF layer. It should also be noted that a significant decrease in remanence is observed as the AF layer thickness increases, indicating the thicker AF layer may form rough surface and induces the formation of magnetically dead layer at the Mn 2 VSi/CoFe interface. If the measurement is performed at a temperature close to T N of the AF, the changes in the hysteresis loop become smaller. In Mn 2 VSi, we use the setting temperature of 498 K for 30 min to make the AF layer paramagnetic. Under an external field of 20 kOe, the spin moments in the CoFe layer are aligned along the direction of the field. The samples with different thicknesses of the Mn 2 VSi layer are measured at 100 K. The samples with the 80 nm Mn 2 VSi layer are found to induce the largest exchange bias of about 30 Oe, whereas the rest of the samples show their exchange biases of less than 10 Oe (see figure 4). This indicates that the poor crystallinity of the Mn 2 VSi cannot align their spin moments antiferromagnetically and hence cannot induce a large exchange bias. It should be noted that the Mn 2 VSi films themselves do not show any magnetic hysteresis within the order of 2 emu cm −3 , confirming that these films are antiferromagnetic.
The activation temperature (T ACT ) is also measured to determine the blocking temperature (T B ), where the exchange  bias becomes zero. Normally T B is determined by increasing the activation temperature until the loop shift becomes zero. In polycrystalline systems, individual grains have their own T B and hence the average T B is measured using the York Protocol [9]. The samples are uniformly set by thermal activation to the same setting state as described above (498 K for 30 min) before being cooled to T NA (a temperature at which no thermal activation occurs), thereby removing any magnetic history. By using a saturating magnetic field (20 kOe) at a setting temper ature (T SET ) that is above the T N of the AF Mn 2 VSi film (estimated to be below 498 K) but below the Curie temperature (T C ) of the F CoFe film (1273 K). A period of 30 min would reverse any activated grains to their original 'set' state. These times also negate any thermal activation that may occur during the temperature rise and fall [9]. The sample is then cooled down to T NA , we have set T NA to be 100 K because this is the lowest temperature VSM can achieve. The magnetization orientation of the F layer is reversed. The sample is then heated for 30 min to the activation temperature T ACT , following by cooling back to T NA . This procedure removes the first loop training effect and measuring at T NA ensures that slow thermal training does not occur. The average exchange bias is measured to be 34 Oe at 100 K as shown in figure 5. Here, the exchange bias did not go to zero as temperature changes. This result indicates that the blocking temperature is below 100 K. These characteristic temperatures can be engineered by substituting some constituent atoms in Mn 2 VSi with the other atoms as similarly demonstrated in ferromagnetic Heusler alloy [10] which would increase the anisotropy of AF layer.

Conclusion
Mn 2 VSi is confirmed to be antiferromagnetic, where the corre sponding exchange bias of 34 Oe at 100 K is measured when a 3 nm ferromagnetic CoFe layer is deposited on top of Mn 2 VSi layer. The film with 80 nm thick AF layer is found to be optimized for the growth at 723 K. Post-annealed samples provided large surface roughness and causing the diffusion at the interface between each layers, resulting in no exchange bias induced at the Mn 2 VSi/CoFe interfaces. The blocking temperature of Mn 2 VSi grown at 723 K is estimated to be below 100 K. These magnetic properties can be improved by substituting the constituent atoms with the other elements, suggesting a potential of Mn 2 VSi to be used as an antiferromagnet in a spintronic device.