The formation of nanocomposite Nd-Fe-B/Mo/FeCo multilayered magnet films using the infrared laser annealing technique. The perpendicular magnetic anisotropy of the film with Mo interlayer was larger than that of the film without Mo interlayer before annealing. Magnetic properties of both films were drasticly changed from soft magnetism to hard magnetism with in-plane magnetic anisotropy after the infrared laser annealing process at 760°C. Coercivity of in-plane annealed film with Mo interlayer was about 10 kOe which was double that of the film without Mo interlayer. In second quadrant, squareness of demagnetization curve of infrared laser annealing processed film with Mo interlayer was better than that of film without Mo interlayer. The infrared laser annealing process crystallizes the amorphous phase of the Nd-Fe-B layer as well as the multi-step annealing, even though the infrared laser annealing process time is only 1/100 of the process time of the multi-step annealing using halogen lamps reported previously.

Nanocomposite magnets composed of magnetic hard phase Nd2Fe14B nanoparticles and magnetic soft phase α-Fe nanoparticles are not only expected to increase the maximum energy product, (BH)max, compared to the Nd-Fe-B magnets,1,2 but also effective in reducing the use of heavy rare-earth elements, HREs. It is known that pulse-thermal processing, PTP, has been attempted to improve the properties of Nd-Fe-B magnet films and to form Nd-Fe-B nanocomposite magnets by using various heat sources and methods for rapid heating and cooling.3–7 PTP has been applied to a two-step method combining post-annealing with deposition below the crystallization temperature. This is one of the methods to form magnetic thin films which can give excellent magnetic properties while maintaining high film surface flatness.

Among them, the KrF excimer laser pulses with wavelength λ of 248 nm, which are selected as a high energy source, can crystallize the amorphous phase in a very short time (ps order). The emergence of hard magnetism with large perpendicular coercivity has been reported in Nd-Fe-B thin films of 2 μm thickness irradiated with KrF excimer pulses.6 However, excimer laser sources are very expensive and are used for special applications requiring non-thermal processing. In contrast, continuous wave laser diode, CWLD, have the advantage of being very low cost, although they cannot be pulsed like excimer lasers. The infrared lasers of CWLD with a wavelength of about 1 μm can rapidly heat up the material to be irradiated by resonant absorption and have a fast cooling rate. The laser spot size can be adjusted to the order of mm, which is convenient for annealing small samples. Recently, we have succeeded in producing hard magnetism and controlling the perpendicular magnetic anisotropy, PMA, and in-plane magnetic anisotropy, IMA, in the Nd-Fe-B/Fe-based multilayered films by using a two-step method.8,9 However, during the post-deposition annealing process in the ultra-high-vacuum, UHV, chamber, a common focused halogen lamp heater was used, which required an extremely long annealing process. As mentioned above, there are various known methods of post-annealing, but the use of highly efficient lasers is an excellent technique that can significantly reduce process time and improve magnetic properties.

In this work, we report on the effects of the infrared laser annealing on the magnetic properties of Nd-Fe-B/Mo/FeCo multilayered films.

The thin films were deposited on MgO(001) single crystal substrate with 0.5 mm thick during heating at 300°C by using an UHV magnetron/helicon multi-target sputtering system with base pressure less than 5 × 10−8 Pa. The stacking forms of thin films were as follows: MgO(001)/Mo(20 nm)/[Nd-Fe-B(30 nm)/Mo(tMo nm)/FeCo(5 nm)/Mo(0.3 nm)]5/Mo(10 nm). The periodic number of the multilayer, ML, part is five. The ML part without Mo interlayer (tMo = 0 nm) and with Mo interlayer (tMo = 0.3 nm) were prepared to evaluated the effect of Mo interlayer on magnetic properties. Details of the sputtering conditions of samples were denoted in previous paper.9 Size of the deposited film is approximately 4 mm × 4 mm on the substrate of 5 mm × 5 mm squarer. Post-deposition annealing was carried out using the infrared laser annealing (hereafter called iLA) system in a vacuum under 5 × 10−4 Pa. The iLA system is equipped with a CWLD with a wave length of 939.8 nm and a laser spot size of about 5 mm diameter on surface of the samples. Fig. 1 shows typical laser irradiation patterns at 4.5 W (7 A) and surface temperature of the films before correction. The iLA processed temperature of surface of the films was monitored using pyrometer which was correlated by platinum platinum-rhodium, PR, thermocouple thermometer. The temperature of the film during laser irradiation is not stable. Therefore, we considered that the annealing condition where the temperature-time coefficient, ΔTt, is less than about 0.2 (°C/sec) is close to the thermal equilibrium. The relationship between corrected annealing temperature and laser current value in the iLA process as shows in Table I. Crystal structures were characterized using x-ray diffraction, XRD, with Cu-Kα radiation for the out-of-plane and the in-plane configuration. Surface topography of the films was analyzed by atomic force microscopy, AFM. The magnetization curves were measured using a vibrating sample magnetometer, VSM, with external magnetic field, Hex, up to 15 kOe and SQUID magnetometer with Hex up to 50 kOe. Directions of applied filed are the out-of-plane, OOP, and the in-plane, IP, of surface of the thin film samples. All measurements were carried out at room temperature.

Fig. 2 and Fig. 3 show IP and OOP magnetization curves of [Nd-Fe-B/Mo(tMo = 0 nm, 0.3 nm)/FeCo]5 multilayered films as-deposited and annealed at various laser currents, I. Both films with and without Mo interlayer as deposited state and below current value I = 5A showed soft magnetic properties with small coercivity and weak PMA. The PMA of the film with Mo interlayer was larger than that of the film without Mo interlayer. Magnetic properties of both films were drastic changed from soft magnetism to hard magnetism with IMA after iLA process above current value I = 6 A (646°C). The coercivity of in-plane direction of iLA processed film with Mo interlayer was about 10 kOe which was double of the film without Mo interlayer. In second quadrant, squareness of demagnetization curve of iLA processed film with Mo interlayer was better than that of film without Mo interlayer. The good hard magnetism resulting from the iLA process on the films with the Mo interlayer is similar to that obtained by conventional annealing with halogen lamp heaters.9 

Fig. 4 shows the annealing temperature, Ta, or laser current dependence of in-plane coercivity, Hc, and residual magnetization, Mr, for the [Nd-Fe-B/Mo(tMo = 0 nm, 0.3 nm)/FeCo]5 multilayered films. These graphs show that the values of Hc and Mr increase from soft to hard magnetism after iLA process of Ta 646°C. After the rapid increase of the values of Hc and Mr, these values remained almost constant up to 760°C. These results could be attributed to the change of microstructure of the films and crystallization of the Nd-Fe-B layer from the amorphous to the Nd2Fe14B by iLA process. The direction of the c-axis of the Nd2Fe14B, which is the easy axis of magnetization in Nd-Fe-B/Mo/Fe multilayers, is parallel to the in-plane direction of the film. The IMA of the multilayer film is the sum of the magnetic shape anisotropy and the magnetocrystalline anisotropy. Therefore, assuming that the c-axis is dispersed in the in-plane direction, the saturation magnetization Ms of this multilayer is about 500 emu/cc, and the value of IMA estimated from its shape anisotropy Ms2/2μ0 is estimated to be in the order of 106 erg/cm3.

The annealing temperature (irradiation laser current) dependence of average surface roughness, Ra, and peak-to-valley, P-V, of Nd-Fe-B/Fe-Co multilayered films and Nd-Fe-B/Mo/Fe-Co multilayered films obtained by AFM is shown in Fig.5. These results show that value of Ra varies in the range of 0.188 nm to 0.291 nm and value of P-V varies in the range of 1.288 nm to 2.986 nm in the films without Mo interlayer. In the films with Mo interlayer, value of Ra varies in the range of 0.279 nm to 0.483 nm and value of P-V varies in the range of 2.636 nm to 5.763 nm. The films annealed at 760°C (I = 7A) showed the largest changes in both the films without and with Mo interlayer. In spite of this slight change in the surface of the film. There was no change in the surface morphology of the films with or without the Mo interlayer and at different annealing temperatures, and the average roughness and maximum height difference of the film surface were almost the same, indicating that the film surface was very flat.

Fig.6 shows the out-of-plane XRD patterns of Nd-Fe-B/FeCo and Nd-Fe-B/Mo/FeCo multilayers. No peaks are observed in the as-deposited state of the multilayer films due to their amorphous nature. After annealing, the peaks of Nd2Fe14B, dhcp-Nd and Nd2Fe17 phases appeared in both films. The weak peak intensities are due to the precipitation of very small size Nd2Fe14B particles (<30 nm) with defects from the amorphous state. PDF data10 and simulated XRD patterns11 show that the peak intensity ratio of Nd2Fe14B is closer to (220) orientation than to random orientation. This result is similar to our previous report using a halogen lamp heater.9 On the other hand, a strong peak of the Nd2Fe17 phase appeared in the Mo interlayer introduced film, while the peak of Nd2Fe17 hardly appeared in the Mo interlayer introduced film. This result implies that the inter-diffusion between the Nd-Fe-B and FeCo layers during laser annealing process is suppressed by introducing a 0.3 nm Mo interlayer.

The in-plane XRD patterns of NFB/FeCo and NFB/Mo/FeCo multilayers are shown in Fig.7. The diffraction peaks at 43° and 94° for all the films originate from (200) and (400) of the MgO single crystal substrate. Therefore, the peaks at 42° and 88° are considered to be those of Mo(110) and Mo(220) from the epitaxial growth of the Mo(001) under-layer on MgO(001).12 In α-Fe, the peak of (210) does not appear. Therefore, the peaks at 44° and 74° originate from (110) and (210) of FeCo (Pm3̄m),13 and (210) is preferentially oriented in the in-plane direction of the film. Here, we focus on the lattice constant a of FeCo before and after annealing: the lattice constant a of the FeCo layer of the film without the Mo interlayer increased from a = 0.2853 nm before annealing to a = 0.2860 nm after annealing, approaching the value of α-Fe powder (a = 0.2866 nm).11 On the other hand, the FeCo layer of the film with Mo interlayer showed a slight decrease from a = 0.2856 nm before annealing to a = 0.2853 nm after annealing, approaching the literature value for FeCo (a = 0.2851 nm).13 This may be evidence that the Mo interlayer plays the role of a barrier to interdiffusion, similar to the discussion above that the Mo interlayer suppresses the precipitation of the Nd2Fe17 phase after annealing. The peak from the Nd2Fe14B phase is not visible in the in-plane XRD pattern of the film. Comparison of the simulated pattern of (100) orientation with the experimental results suggests that if (100) in Nd2Fe14B is oriented in the in-plane direction of the film, then the peaks of (400) and (800) overlap with the peaks of Mo(110) and Mo(220). In order to clarify the conditions under which the Nd2Fe14B particles are oriented during annealing, we are planning to carry out detailed observations of the microstructure of the thin films using transmission electron microscopy, TEM.

The hard magnetism obtained by various laser annealing and PTP processes reported in the past mainly shows perpendicular anisotropy or isotropic magnetic properties. However, the good hard magnetism with in-plane magnetic anisotropy resulting from the iLA process on thin films with Mo interlayer is similar to that obtained by annealing with halogen lamp heaters as we have reported. Therefore, it is clear that the in-plane magnetic anisotropy of the multilayered films is produced even by shortening the annealing time due to the difference in the annealing method, and that the insertion of the Mo interlayer improves the squareness of demagnetization curves and increases the coercivity at the same time.

We are grateful to Mr. Y. Mizuno of Yamagata University for his helpful support in XRR analysis. This work supported in partly by the KAKENHI Grant Number 16H04488, 20K05059 and 20H02425, Japan Science Promotion Society (JSPS), Japan, and the Joint Usage Research in Molecular Photoscience Research center in Kobe University, Japan.

The authors have no conflicts to disclose.

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

1.
R.
Skomski
and
J. M. D.
Coey
,
Phys. Rev. B
48
,
15812
(
1993
).
2.
E. F.
Kneller
and
R.
Hawig
,
IEEE Trans. Magn.
27
,
3588
(
1991
).
3.
T.
Harada
,
M.
Fujita
, and
T.
Kuji
,
Nuc. Inst. Meth. in Phys. Res. B
121
,
383
(
1997
).
4.
M.
Yu
,
Y.
Liu
,
S. H.
Liou
, and
D. J.
Sellmyer
,
J. Appl. Phys.
83
,
6611
(
1998
).
5.
Z. Q.
Jin
,
V. M.
Chakka
,
Z. L.
Wang
,
J. P.
Liu
,
P.
Kadolkar
, and
R. D.
Ott
,
JOM
58
,
46
(
2006
).
6.
T.
Okuda
,
A.
Sugimura
,
O.
Eryu
,
L. K. E. B.
Serrona
,
N.
Adachi
,
I.
Sakamoto
, and
A.
Nakanishi
,
Jpn. J. Appl. Phys.
42
,
6859
(
2003
).
7.
D.
Hopkinson
,
A.
Cockburn
, and
W.
O’Neill
, in
International Congress on Applications of Lasers and Electro-Optics
(
2011
), Vol. 11, p.
1233
.
8.
K.
Kobayashi
,
D.
Ogawa
,
K.
Koike
,
H.
Kato
,
M.
Oogane
,
T.
Miyazaki
,
Y.
Ando
, and
M.
Itakura
,
J. Phys.: Conf. Ser.
903
,
012015
(
2017
).
9.
K.
Koike
,
K.
Ohashi
,
T.
Suzuki
,
C.
Okita
,
N.
Inaba
,
H.
Kato
,
M.
Itakura
,
S.
Hara
,
Y.
Saito
,
S.
Okubo
, and
H.
Ohta
,
AIP Advances
11
,
015134
(
2021
).
10.
S.
Gates-Rector
and
T.
Blanton
, “
The powder diffraction file: A quality materials characterization database
,”
Powder Diffr.
34
(
4
),
352
360
(
2019
).
11.
D. C.
Palmer
, CrystalMaker and CrystalDiffract, CrystalMaker Software Ltd., Begbroke, Oxfordshire, England,
2014
.
12.
K.
Koike
,
T.
Kusano
,
D.
Ogawa
,
K.
Kobayashi
,
H.
Kato
,
M.
Oogane
,
T.
Miyazaki
,
Y.
Ando
, and
M.
Itakura
,
Nanoscale Res. Lett.
11
,
33
(
2016
).
13.
W. C.
Ellis
and
E. S.
Greiner
,
Trans. Am. Soc. Met.
29
,
415
434
(
1941
).