Tailoring out-of-plane magnetic properties of pulsed laser deposited FePt thin films by changing laser energy fluence

Applied Surface Science 315 (2014) 37–44

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Tailoring out-of-plane magnetic properties of pulsed laser deposited FePt thin films by changing laser energy fluence Ying Wang a , T.L. Tan a , K.S. Tan a , P. Lee a , Hai Liu b , Boluo Yadian b , Ge Hu b , Yizhong Huang b , R.V. Ramanujan b , R.S. Rawat a,∗ a b

NSSE, NIE, Nanyang Technological University, 1 Nanyang Walk, Singapore 637616, Singapore School of Material Science and Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore

a r t i c l e

i n f o

Article history: Received 11 March 2014 Received in revised form 16 June 2014 Accepted 17 July 2014 Available online 24 July 2014 Keywords: FePt Pulsed laser deposition Laser energy fluence Magnetic data storage

a b s t r a c t Magnetic properties of pulsed laser deposited (PLD) FePt thin films are investigated at three different laser energy fluences of 51, 136 and 182 J/cm2 . Deposition at lower laser energy fluence (51 J/cm2 ) yields softer out-of-plane coercivity (≤0.4 kG), whereas deposition at higher laser energy fluence (136 and 182 J/cm2 ) results in harder out-of-plane coercivity (≥5.0 kG). The improved coercivity is found to be attributed to the formation of vacancy defects in thin films, which is indicated by stress change from tensile to compressive form with increasing laser energy fluence. Maximum out-of-plane saturated magnetization of 615 emu/cm3 and remanent squareness ratio of 0.88 are achieved for 16 nm thick FePt thin films deposited at moderate laser energy fluence of 136 J/cm2 , making them suitable for high density perpendicular data storage applications. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Magnetic nanoparticles (NPs) like FePt are of high interest both in basic research and applications. They have been extensively studied due to their potential applications in magnetic data storage, drug deliveries, biomedical imaging, etc. [1]. Face centered tetragonal (fct) FePt nanoparticles are especially of immense technological importance in high density magnetic storage media applications because of its high uniaxial magnetocrystalline anisotropy (Ku = 6.6 × 107 J/m3 ) and thermal stability [2]. FePt nanoparticles can be magnetically stable up to the size of 4 nm. They can also exhibit large coercivity of 10–20 kOe, a value of which, however, is unsuitable for recording media due to the limitation of the writing head [2–4]. For magnetic storage media, FePt nanoparticles should have coercivity lower than 10 kOe (for data bit to be easily written) and high magnetization (for better data readability) [2,3]. This requires a composite of hard and soft magnetic phases being strongly exchange coupled in FePt nanoparticles. Moreover, perpendicular orientation is preferred for high density storage application as compared to longitudinal orientation, because perpendicular magnetic recording has narrower transition region between recording bits and therefore a higher recording density [5]. Shen et al. [6] reported an achievement of out-of-plane

∗ Corresponding author. Tel.: +65 6790 3930/3908; fax: +65 6896 9414. E-mail address: [email protected] (R.S. Rawat). http://dx.doi.org/10.1016/j.apsusc.2014.07.094 0169-4332/© 2014 Elsevier B.V. All rights reserved.

squareness (∼1) and out-of-plane coercivity (9–15 kOe) for FePt thin films with a thickness of 5–15 nm prepared by dc magnetron sputtering. Chen et al. also obtained perpendicular magnetic anisotropy in FePt films of 10 nm thickness, while in-plane magnetic anisotropy are found to be preferred in FePt films with thickness over 20 nm [7]. Thus, films thinner than 20 nm are deposited in this work. In pulsed laser deposition (PLD), the orientation texture of the thin films is found to be associated with the energetic nature of the ablation plume. Nakano et al. [8–10] and Chang et al. [11] reported the dependence of composition, structure and magnetic properties on ablation laser power for FePt films deposited by pulsed laser deposition (PLD). Nakano et al. reported a transition from isotropic films to perpendicular anisotropic films with increasing laser power from 5 to 7 W [9]. Chang et al. varied the laser photon energy by changing laser wavelength (1064, 532, 355 to 266 nm) [11], but laser pulse energy and laser energy fluence was kept constant. They report an evolution of deposition mechanism from thermal evaporation (for wavelength of 1064 and 532 nm) to ablation (for wavelength of 266 nm) [11]. Compared to composition of target, Pt-rich FePt films are deposited in thermal evaporation mode while stoichiometric FePt films are prepared in ablation mode [11]. The changed composition for different wavelength leads to different structural and magnetic properties [11]. Our previous work [12] also showed that the ablation laser fluence plays a key role for the formation of impurities in FePt thin films and consequently affects their magnetic properties. All these studies [8–10] demonstrate the importance of ablation laser energy to

38

Y. Wang et al. / Applied Surface Science 315 (2014) 37–44

the properties of PLD deposited FePt films. However, films reported by Nakano et al. and Chang et al. are of tens m and 100 nm thick, respectively, while thin films of less than 20 nm are of interest in this work. Our previous work [12] focused on elimination of impurities in FePt thin films instead of systematical studies of effect of laser energy fluences. Since crystalline structures of FePt films also depend on film thickness as reported by Shen et al. [6] and Chen et al. [7], it is interesting to investigate the effect of ablation laser energy on thinner films as it might be different from that on thicker films (100 nm and tens m ones). Thus, the effects of laser energy fluences on morphological, crystalline and magnetic properties of PLD synthesized FePt thin films (of 5–16 nm thick) are worth studying in detail. In this work, FePt films thinner than 20 nm are deposited using PLD to form perpendicular orientation in samples. Various laser energy fluences (51, 136 and 182 J/cm2 ) are used to study their effect on the morphological, crystalline and magnetic properties of FePt thin films.

2. Experimental details FePt thin films of thickness 5, 10 and 16 nm are produced at room temperature in vacuum at background pressure better than 5.7 × 10−6 mbar by PLD using a Lotis-TII (LS-2137U) 2nd harmonic Nd:YAG laser (532 nm) with laser energy fluences (LEF) of 51, 136 and 182 J/cm2 and a pulse repetition rate of 10 Hz. The LEF on the target surface is varied using the method reported in our previous paper [13] in which neutral density filters of different transmission values were used for fixed laser pump energy and fixed laser spot size. The diameter of the laser spot size on the target surface is kept constant at 264 m. This method of controlling the LEF on target surface is most reliable as by keeping the laser pump energy same the laser pulse width also remains the same as it is well known that changing the laser pump energy also changes the laser pulse length. The average deposition rate for LEF of 51, 136 and 182 J/cm2 is estimated to be about 0.14, 0.18 and 0.22 nm/min, respectively. At fixed ablation LEF, the thickness of thin films is controlled by the number of laser shots. The as-deposited samples are uniform thin films with a few laser droplets as the scanning electron microscopy (SEM) image shown in Fig. 1(a). The compositions of as-deposited FePt thin films tend to reduce the amount of Fe by 2–9 at.% compared with that of target material (Fe50 Pt50 ). As shown in Fig. 1(b), the average Fe at% for as-deposited samples deposited at LEF of 51, 136 and 182 J/cm2 is 45.4, 48.2 and 40.9 at%,

respectively. This indicates Pt-rich composition for all the samples. The as-prepared samples were post-annealed at 600 ◦ C for 60 min in a 95% Ar/5% H2 flowing gas atmosphere. The silicon (Si) substrates were placed in the deposition system at a distance of 5 cm normal to the ablation target. The FePt target (50:50 at% with 99.99% purity, Ø 2.5 cm), which is purchased from Kurt J. Lesker, is polished and ultrasonically cleaned before being fixed in the PLD chamber. The Si substrates were cleaned sequentially in acetone, ethanol and deionized water for 10 min each in an ultrasonic bath. The commercial purchased Si substrates were of N type and with 1 0 0 orientation. The structure, morphology, composition and magnetic properties of samples were investigated by various characterization techniques. The surface morphology and thickness of samples were measured by field emission scanning electron microscopy (FE-SEM, JEOL JSM-6700F). The compositions of thin films were obtained by FE-SEM attached with Oxford Instrument’s energy dispersive X-ray (EDX) spectroscopy. Magnetic properties of the films were measured by the vibrating sample magnetometer (VSM, Lake Shore 7400) with a maximum field up to 14 kG. The phase composition of the coatings was identified using thin-film X-ray diffractometer (XRD, Shimadzu 6000) with Cu K␣ radiation of 1.5418 A˚ wavelength with a scan speed of 0.016◦ /s. The microstructure and images of magnetic nanoparticles were obtained on a JEOL 2100F transmission electron microscope (TEM).

3. Results and discussion The in-plane and out-of-plane hysteresis loops of samples deposited at different laser energy fluences (LEF) (51, 136 and 182 J/cm2 ) are shown in Fig. 2(a) and (b) for film thickness (t) of 5 nm and in Fig. 2(c) and (d) for t = 10 nm. Samples deposited at LEF of 51 J/cm2 show small coercivity values for both in- and out-ofplane hysteresis loops indicating low perpendicular and longitudinal anisotropy (Ku ). The samples synthesized with LEF of 136 and 182 J/cm2 , on the other hand, exhibit much larger perpendicular anisotropy (Ku ) than the longitudinal samples, as the area under the out-of-plane hysteresis loops is much larger than that of in-plane loops. The larger Ku is attributed to the strong (0 0 1)-preferred orientation induced by thinner thin films (t < 15 nm) [6] and increased ablation laser energy density [14]. The presence of longitudinal Ku grains in samples, indicated by the observed in-plane hysteresis loops, may originate from the relaxation of stress that favors the formation of perpendicular Ku FePt [15,16]. However, there

Fig. 1. (a) SEM image of as-deposited sample at LEF of 51 J/cm2 ; (b) Fe and Pt concentrations of as-deposited FePt thin films as a function of laser energy fluences. The composition of the target is Fe50 Pt50 .

Y. Wang et al. / Applied Surface Science 315 (2014) 37–44

39

Fig. 2. The in-plane and out-of-plane hysteresis loops of the FePt thin films deposited with different laser energy fluences (LEFs) for (a and b) 5 nm, (c and d) 10 nm and (e and f) 16 nm thickness.

is a clear magnetization drop (resulting in kink formation) at the soft-phase nucleation field in each of the out-of-plane hysteresis loops (Fig. 2(b) and (d)), which indicates that there still exist some nanoparticles with relatively low ordering factors in those samples. Such kinks in hysteresis loops originate from the exchange coupling between the grains of magnetically hard and soft phases [17]. The decrease of the percentage of magnetically soft fcc phase will weaken the exchange interactions between the two phases and the magnitude of kinks [17]. To further study the effect of film thickness and LEF on sample properties, two more samples deposited at LEF of 136 and 182 J/cm2 with t = 16 nm are characterized. Their in-plane and out-of-plane hysteresis loops are shown in Fig. 2(e) and (f). A much larger decrease of perpendicular Ku is observed as LEF is increased from 136 to 182 J/cm2 for 16 nm film as compared to the decrease observed for 5 and 10 nm thick samples. The variations of in-plane and out-of-plane coercivity (Hc,in and Hc,out ), saturated magnetization (Ms,in and Ms,out ) and remanent squareness ratio ((Mr /Ms )in and (Mr /Ms )out ) with the LEF are plotted in Fig. 3. Their values are listed in Table 1. The Hc,in and Hc,out

(Fig. 3(a)) mostly increase with increasing LEF, except the Hc,out of 16 nm sample. This is because of (i) the induced phase transition from face centered cubic (fcc) to face centered tetragonal (fct) by increasing LEF from 51 to 136 J/cm2 , and (ii) improved crystallization by increasing LEF from 136 to 182 J/cm2 as shown in the XRD patterns in Fig. 4. However, the Ms (Fig. 3(b)) and Mr /Ms (Fig. 3(c)) do not vary linearly with increasing LEF. For samples with fixed t, maximum Ms,out and minimum Ms,in are exhibited for samples deposited at LEF of 136 J/cm2 . As reported by Sahu et al. [2] and Sun et al. [3], the ideal magnetic properties for FePt applied in ultra-high data storage should be perpendicular orientation, of Hc,out less than 10 kG and of higher Ms,out . Thus, samples synthesized with LEF of 136 J/cm2 exhibit better magnetic properties compared to samples with LEF of 51 and 182 J/cm2 . In Fig. 3(c), the much larger (Mr /Ms )out than (Mr /Ms )in for all the samples further confirms that the preferred perpendicular magnetic anisotropy is obtained in all the deposited samples. However, the values of (Mr /Ms )in increased with increasing LEF, indicating an increased longitudinal Ku state in the samples. (Mr /Ms )out increases

40

Y. Wang et al. / Applied Surface Science 315 (2014) 37–44

Fig. 3. In-plane and out-of-plane (a) coercivity (Hc ), (b) saturated magnetization (Ms ) and (c) remanent squareness ratio (Mr /Ms ) values as a function of ablation LEF at a constant film thickness (t) of 5, 10 and 16 nm, respectively.

with increasing LEF for 5 nm sample, but decreases with increasing LEF from 136 to 182 J/cm2 for 10 and 16 nm samples. The larger difference between (Mr /Ms )in and (Mr /Ms )out values appears for FePt film deposited at LEF of 136 J/cm2 with film thickness of 10 and 16 nm. Their values of (Mr /Ms )in and (Mr /Ms )out are (0.25, 0.84) and (0.31, 0.88), respectively. As reported by McCalum et al. [15], a near zero (Mr /Ms )in and a value of (Mr /Ms )out as large as 1 are preferred for bit patterned media (BPM) to decrease the magnetic switching filed distribution and bit error rate. The greater difference between (Mr /Ms )in and (Mr /Ms )out values is attractive for BPM. Therefore, samples deposited with LEF of 136 J/cm2 shows the best magnetic properties among all the samples. Comparison of X-ray diffraction patterns for samples synthesized with different LEF of 51, 136 and 182 J/cm2 is shown in Fig. 4. For 5 nm thick films (Fig. 4(a)), an overlapped diffraction peak at about 2 = 53.2◦ of chemically disordered face centered cubic (fcc) A1(2 1 0) phase and ordered face centered tetragonal (fct)

L10 (2 0 1) is observed for films coated at 51 J/cm2 . This indicates a small amount of fcc to fct phase transition in the sample, which is in agreement with the small Hc,in and Hc,out values (∼200–300 G) observed in Fig. 2(a). When LEF increases from 51 to 136 J/cm2 , the appearance of diffraction peak L10 (0 0 1) for out-of-plane Ku FePt as well as a shift of the overlapped peak at ∼53.2◦ to fct L10 (2 0 1) peak at ∼53.6◦ are observed. This indicates an improved phase transition from low Ku fcc to high Ku fct by raising LEF, which results in the obviously broadened out-of-plane hysteresis loop (Fig. 2(b)) and enhanced Hc,out and (Mr /Ms )out (Fig. 3). Further increase in LEF to 182 J/cm2 causes more intense L10 (0 0 1) phase peak and a less intense L10 (2 0 1) phase peak increasing I(001) /I(201) to 1.13 from 0.56 for 136 J/cm2 deposition (the expected value of I(001) /I(201) is 2.84 for fully transition fct-FePt phase; refer to PDF #03-0659121). This reveals a further improved phase transition in FePt thin films deposited at 182 J/cm2 . The calculated lattice parameters: (i) c = 3.7036 and c/a = 0.959 for 136 J/cm2 energy deposited

Table 1 In-plane and out-of-plane coercivity (Hc,in and Hc,out ), in-plane and out-of-plane saturated magnetization (Ms,in and Ms,out ), in-plane and out-of-plane remanent squareness ratio ((Mr /Ms )in and (Mr /Ms )out ) for samples deposited at different LEFs with various fixed film thickness (t). t (nm)

LEF (J/cm2 )

Hc,out (kG)

Ms,out (emu/cm3 )

(Mr /Ms )out

Hc,in (kG)

Ms,in (emu/cm3 )

(Mr /Ms )in

5

51 136 182

0.3 9.1 9.5

284 545 394

0.20 0.63 0.76

0.2 0.2 0.2

323 120 151

0.15 0.15 0.17

10

51 136 182

0.4 7.1 7.7

132 464 350

0.26 0.84 0.75

0.3 0.4 1.1

179 141 143

0.19 0.25 0.46

16

51 136 182

– 5.6 5.3

– 615 142

– 0.88 0.58

– 0.6 1.1

– 129 129

– 0.31 0.46

Y. Wang et al. / Applied Surface Science 315 (2014) 37–44

41

Fig. 4. XRD patterns for samples deposited at various LEF (51, 136 and 182 J/cm2 ) with thickness of (a) 5 nm, (b) 10 nm, and (c) 16 nm.

(throughout) films, and (ii) c = 3.6942 and c/a = 0.960 for 182 J/cm2 films of tetragonal cell are in good agreement with that of fct-FePt phase (PDF #03-065-9121, c = 3.7020 and c/a = 0.964). The degree of preferred orientation is studied by adopting Lotfactor (LOF) [16,18]. For a specific orientation gering orientation   such as 0 0 l , the Lotgering orientation factor LOF =

(p − p0 ) (1 − p0 )

where

 I(0 0 l)oriented p=  and I(h k l)oriented

p0 =

 I(0 0 l)non-oriented  . I(hkl)non-oriented

Normally LOF has values from 0 to 1; 0 for isotropic texture, where the orientation of grains is randomly distributed, and 1 for perfect alignment of a specific orientation. The LOF of c-axis orientation (Fig. 5(c)) increases from 0.2 to 0.4 with increasing LEF from 136 to 182 J/cm2 , indicating an improved perpendicular orientation in the sample with higher LEF, which is also confirmed by the increased values of Hc,out , (Mr /Ms )out and the difference of (Mr /Ms )in and (Mr /Ms )out for 5 nm thick films in Fig. 3. For the XRD patterns of 10 nm thick films (Fig. 4(b)), similar structural changes as that of 5 nm thick films with increasing LEF are observed. When the LEF increases from 51 to 136 J/cm2 , a much higher degree of transformation from fcc to fct is achieved,

as indicated by the emergence of L10 (0 0 1) phase peak and sharp increase in Hc,out (Figs. 2(e) and 3(a)). When LEF is further increased to 182 J/cm2 , the intensity of both L10 (0 0 1) and L10 (1 1 1) peaks increases indicating enhanced chemically ordered fct phase, resulting in increased Hc,out as shown in Fig. 3(a). However, many more diffraction peaks corresponding to different orientations are also observed, demonstrating a relative increase of in-plane orientation. This is also indicated by the increase in Hc,in from 0.4 to 1.1 kG as shown in Figs. 2(f) and 3(a). It is reported that for materials with randomly oriented nanoparticles undergoing coherent magnetic moment rotations without interaction, the remanent squareness ratio is equal to 0.5. Thus, there is decreased difference of (Mr /Ms )in and (Mr /Ms )out due to the decreased (Mr /Ms )out and increased (Mr /Ms )in toward 0.5 value, as seen in Fig. 3(c). The calculated lattice parameters and tetragonality represented by c/a for the 10 nm thick films with different LEF are plotted in Fig. 5(a) and (b), respectively. With increasing LEF from 51 to 136 J/cm2 , the value of c decreases from 3.7217 to 3.6942, and the value of a increases from 3.8178 to 3.8479. This results in a decrease of c/a value from 0.975 to 0.960, the value of which matches better with that of fct-FePt phase (PDF #03-065-9121, c = 3.7020 and c/a = 0.964). The shrinkage of c axis and expansion of a axis, enabling grains with (0 0 1) orientation that are more stable energetically under biaxial tensile-stressed state, are due to the disorder-order phase transition. The expansion of a axis could cancel out the tensile strain to lower the total strain energy of the film. When LEF further increases, the value of c stays constant while the value of a increases sharply to 3.9940, leading to a sharp drop of c/a value from 0.960 to 0.925. The expansion of a axis might be due to the formation of antisite defect of Pt on the Fe site induced by larger ablated species energy at larger LEF. The Pt

42

Y. Wang et al. / Applied Surface Science 315 (2014) 37–44

Fig. 5. (a) The lattice constants of a and c plane and (b) the tetragonality represented by c/a for 10 nm thick films synthesized at different LEF of 51, 136 and 182 J/cm2 ; (c) Lotgering orientation factor, LOF, and (d) the tetragonality represented by c/a for 5, 10 and 16 nm thick films deposited at LEF of 136 and 182 J/cm2 ; SEM images for 10 nm thick films deposited at different LEF of (e) 51 J/cm2 , (f) 136 J/cm2 , and (g) 182 J/cm2 .

antisite defect on Fe site is reported to have the lowest formation energy among all kinds of defects [19]. The formation of Pt antisite defect also decreases Fe content in films, which might be the reason of the decreased Ms as shown in Fig. 3(b). This is because of the larger spin momentum of Fe (∼3.0B ) than that of Pt (∼0.3B ). For the XRD patterns of 16 nm thick films (Fig. 4(c)), increased LEF makes the thin films even more polycrystalline, resulting in a decrease of perpendicular LOF (Fig. 5(c)) and a closer value to 0.5 (for random orientation) for (Mr /Ms )in and (Mr /Ms )out . Compared with the XRD patterns of 5 and 10 nm thick films, it is also found that the number of diffraction peaks (besides the ones corresponding to perpendicular orientation) increases with increasing thickness, indicating an increasing longitudinal orientation in thicker films. This phenomenon is already reported by Chen et al. [7] and Shen et al. [6,20]. However, the changes of LOF and c/a with film thickness for different LEF are different as shown in Fig. 5(c) and (d). The calculated values of a and c are a = 4.0506 and c = 3.6889 for 136 J/cm2 coated films and a = 4.0540 and c = 3.6787 for 182 J/cm2

coated films. The value of c/a is 0.911 and 0.907, respectively. Compared to the lattice parameters of fct-FePt phase (PDF #03065-9121, c = 3.7020 and c/a = 0.964), the decrease of c/a is mainly due to the over expansion of a axis. The trends in variation of c/a ratio (shown in Fig. 5(d)) for different film thickness indicate greater changes for higher LEF. The average crystallite sizes of all the samples are calculated They from their XRD patterns.   are determined using Scherrer formula: D = 0.9/ B cos B , where D is the crystallite size,  is the radiation wavelength,  B is the Bragg angle, and B is the full width at half maximum (FWHM) of the most significant diffraction peak [21]. The calculated crystallite size for 5, 10 and 16 nm thick films deposited at different LEF were found to be in range of 7–14 nm. The typical SEM images for 10 nm annealed sample deposited at different LEF are shown in Fig. 5(e)–(g). The average particle size determined by ImageJ@ analysis software for SEM image shown in Fig. 5(e) for sample deposited at 51 J/cm2 is found to be 11.2 ± 0.3 nm. The particles seem to be uniform in size as the

Y. Wang et al. / Applied Surface Science 315 (2014) 37–44

43

Fig. 6. TEM images of annealed FePt thin films deposited at 51 J/cm2 ; (b)(d) HRTEM images of selected FePt nanoparticles; (c) FFT pattern for FePt nanoparticle shown in (b); (e) FFT pattern for FePt nanoparticle shown in (d).

standard deviation in the particle size is small, which is also verified by TEM images shown later (in Fig. 5). The samples deposited at higher LEF show non-uniform particle size distribution as both smaller and bigger sized particles can be observed in images shown in Fig. 5(f) and (g). The bigger particles are actually particle agglomerates which are probably due to different surface mobility of ablated species upon their arrival at substrate surface governed by their kinetic energy [17,22]. The kinetic energy is related to the collision rate of ablated species and in turn can be tailored through deposition conditions, such as laser energy fluence, substrate to target distance and ambient gas pressure. It is reported by Mahmood et al. [23] that 70% of input laser energy is finally transferred to the ablated plasma species. Thus, larger laser energy fluence results in larger kinetic energy of plasma species (Fe and Pt atoms) which results in particle agglomeration on these samples. It has been reported by several research groups [24–28] that coercivity increases with increasing grain/particle size. The critical size (superparamagnetic limit) for FePt particles is 1.5–2 nm [24] above which the coercivity first increases with the increase in particles size until the upper limit of single domain size is reached; which is reported to be around 55 nm for 10 nm thick FePt thin films [28]. The particle size for samples deposited at higher LEF has become bigger compared to 51 J/cm2 deposition, but it is around the range of single domain size limit resulting in the increase in coercivity with increasing LEF as shown in VSM results. The validation of average particle size, film thickness and the microstructure of 10 nm annealed sample (deposited at LEF of 51 J/cm2 ) was also done using cross-sectional TEM analysis (shown in Fig. 6). Fig. 6(a) shows the nanoparticle morphology of the FePt thin film sample surface. The images shows very well defined highly uniform oval/elliptical shaped nanoparticles on annealed thin film surface. All nanoparticles are found to orientate with the major axis parallel to the substrate surface. The thickness of the film therefore can be taken as the average dimension of the nanoparticles along its minor axis and was found to be about 9.5 ± 0.5 nm. The average dimension of the nanoparticles along the major axis was calculated to be about 13.9 ± 0.8 nm. It is important to note that the nanoparticles are well isolated and do not seem to agglomerate even though the annealing was performed at high temperature of about 600 ◦ C. This is different from what has been normally reported for high temperature annealing where the nanoparticle size seems to grow significantly with annealing temperature due to particle agglomeration [29]. This may be due to very thin film (∼10 nm) being used in the present case. The high resolution TEM images of selected

nanoparticles are shown in Fig. 6(b) and (d) with the corresponding Fast Fourier Transform (FFT) pattern being shown in Fig. 6(c) and (e) respectively. The FFT patterns show that FePt nanoparticles are polycrystalline in nature. The FFT patterns in Fig. 6(c) and (e) match with fct FePt ICSD-659004 data using JEMS@JEOL. They show the formation of family of crystal planes corresponding to (1 1 1), (1 1 0), (2 1 0) and (1 1 2), two of which (planes (1 1 1) and (2 1 0)) are also detected in the XRD pattern labeled as 51 J/cm2 shown in Fig. 4(b). The FFT pattern in Fig. 6(c) has a slight mismatch with fct FePt ICSD-659004 (lattice a = 3.849 and c = 3.700) indicating the possibility of the presence of mixed fct and fcc phase. This is in line with XRD results which exhibited incomplete phase transition from fcc to fct phase with coexistence of A1(2 1 0) and L10 (2 0 1) diffraction peaks. Considering results of EDX, VSM, XRD, SEM and TEM together, the improved magnetic properties with increasing LEF might be attributed to the vacancy formation in samples deposited at higher LEF as discussed later. As discussed earlier, the composition of samples varies with increasing LEF as shown in Fig. 1(b). It is reported that a bit Pt-rich composition (especially for 50–60 at%) results in the harder coercivity [30,31]. All our samples are Pt-rich however the one deposited at higher fluence show higher coercivity while the one deposited at lower LEF of 51 J/cm2 shows lower coercivity. This shows mixed results with no clear indication of the dependence of magnetic properties on the composition of the samples. Thus, although composition is affected by LEF, we cannot conclude it to be the main reason for the ordering promotion and out-of-plane properties improvement. An estimation of strain from (1 1 1) diffraction peaks for 10 nm thick samples deposited at three different LEF (refer to XRD patterns in Fig. 4(b)) shows that strain changes from +0.0024 for 51 J/cm2 to −0.0004 for 136 J/cm2 and lowers further to a more negative value of −0.0019 for 182 J/cm2 . This indicates the change in stress from tensile to compressive form with the increasing LEF. The change from tensile to compressive stress indicates the formation of vacancies with increasing LEF. A theoretical study has shown that the vacancy defects formation leads to more magnetic anisotropic energy leading to better magnetic properties [19], as observed in our case also. It may however be noted that the formation energy for vacancy defect is quite high. In our previous experimental study, complemented by simulation results, it was demonstrated the fraction of input laser energy deposited to the ablated species is about 70% of the input laser energy at the target for all the laser energy fluences [13]. Thus, the ablated plasma species at higher LEF is

44

Y. Wang et al. / Applied Surface Science 315 (2014) 37–44

more energetic leading to possibility of more vacancy defect formation in the deposited sample with higher magnetic anisotropic energy. 4. Conclusions In summary, laser energy fluences of PLD are confirmed to have significant effect on magnetic, structural and morphological properties of FePt thin films. Hard out-of-plane Ku nanoparticles (Hc,out ∼ 5–9 kG, Ms,out ∼460–620 emu/cm3 and (Mr /Ms )out ∼ 0.8) are synthesized using LEF of medium value around 136 J/cm2 . The lower LEF (∼51 J/cm2 ) results in a soft magnetic behavior for FePt films even after 600 ◦ C post-deposition annealing. The larger LEF (∼182 J/cm2 ) leads to thin films of smaller Ms,out (350 emu/cm3 ) and larger in-plane coercivity (∼1.1 kG). Formation of vacancy defects in thin films, which is indicated by the change in stress from tensile to compressive form with increasing LEF, might contribute to the improved magnetic properties. It is expected that PLD grown FePt thin films can exhibit hard out-of-plane Ku nanoparticles with well isolated uniform morphology, matching well with the prerequisite for high density data storage application. Acknowledgments The authors are grateful to the National Institute of Education, Nanyang Technological University Singapore, for providing the AcRF Grants RI 7/08 RSR and RI 7/11 RSR. Ying Wang would like to thank NIE/NTU for providing the research scholarships. Ying Wang would also like to thank Dr. Ge Hu for his help in XRD, Dr. Hai Liu and Mr. Yadian Boluo for their help in TEM and Prof. R. V. Ramanujan and Ms. Swee Kuan for their help in VSM. References [1] Y. Liu, Y. Jiang, X. Zhang, Y. Wang, Y. Zhang, H. Liu, H. Zhai, Y. Liu, J. Yang, Y. Yan, Structural and magnetic properties of the ordered FePt3, FePt and Fe3Pt nanoparticles, Journal of Solid State Chemistry 209 (2014) 69–73. [2] N.K. Sahu, D. Bahadur, Influence of excess Fe accumulation over the surface of FePt nanoparticles: structural and magnetic properties, Journal of Applied Physics 113 (2013) 134303. [3] S. Sun, Recent advances in chemical synthesis, self-assembly, and applications of FePt nanoparticles, Advanced Materials 18 (2006) 393–403. [4] Y. Liu, T.A. George, R. Skomski, D.J. Sellmyer, Aligned and exchange-coupled FePt-based films, Applied Physics Letters 99 (2011) 172504. [5] S. Khizroev, D. Litvinov, Perpendicular magnetic recording: writing process, Journal of Applied Physics 95 (2004) 4521–4537. [6] C.L. Shen, P.C. Kuo, G.P. Lin, S.C. Chen, K.T. Huang, Effect of film thickness on magnetic properties of FePt thin films deposited on amorphous SiO2 substrate directly, IEEE Transactions on Magnetics 47 (2011) 3889–3892. [7] S.C. Chen, T.H. Sun, W.H. Hong, The effect of thickness on the texture and magnetic properties of single-layered FePt films by rapid thermal annealing, Journal of Nanoscience and Nanotechnology 11 (2011) 2623–2627. [8] M. Nakano, S. Shibata, T. Yanai, H. Fukunaga, Fe-Pt thick film magnets prepared by high-speed PLD method, IEEE Transactions on Magnetics 44 (2008) 4229–4231. [9] M. Nakano, S. Shibata, T. Yanai, H. Fukunaga, Anisotropic properties in Fe–Pt thick film magnets, Journal of Applied Physics 105 (2009) 07A732. [10] M. Nakano, W. Oniki, T. Yanai, H. Fukunaga, Magnetic properties of pulsed laser deposition-fabricated isotropic Fe–Pt film magnets, Journal of Applied Physics 109 (2011) 07A723.

[11] H.W. Chang, F.T. Yuan, C.W. Yuan, C.H. Yu, C.R. Wang, W.C. Chang, Effect of magnetic field on the structure and magnetic properties of pulse-laser-deposited FePt films, Journal of Alloys and Compounds 584 (2014) 148–151. [12] Y. Wang, R. Medwal, N. Sehdev, B. Yadian, T.L. Tan, P. Lee, A. Talebitaher, U. Ilyas, R.V. Ramanujan, Y. Huang, R.S. Rawat, Elimination of impurity phase formation in FePt magnetic thin films prepared by pulsed laser deposition, Applied Surface Science 288 (2014) 381–391. [13] S. Mahmood, R.S. Rawat, Y. Wang, S. Lee, M. Zakaullah, T.L. Tan, S.V. Springham, P. Lee, Effects of laser energy fluence on the onset and growth of the Rayleigh–Taylor instabilities and its influence on the topography of the Fe thin film grown in pulsed laser deposition facility, Physics of Plasmas 19 (2012) 103504. [14] H. Kim, R.P. Camata, S. Lee, G.S. Rohrer, A.D. Rollett, Y.K. Vohra, Crystallographic texture in pulsed laser deposited hydroxyapatite bioceramic coatings, Acta Materialia 55 (2007) 131–139. [15] A.T. McCallum, D. Kercher, J. Lille, D. Weller, O. Hellwig, Prevention of dewetting during annealing of FePt films for bit patterned media applications, Applied Physics Letters 101 (2012) 092402. [16] J.K. Mei, F.T. Yuan, W.M. Liao, Y.D. Yao, H.M. Lin, H.Y. Lee, J.H. Hsu, Effect of initial stress/strain state on formation of (0 0 1) preferred orientation in L10 FePt thin films, Journal of Applied Physics 109 (2011) 07A737-707A737-733. [17] J.J. Lin, L.S. Loh, P. Lee, T.L. Tan, S.V. Springham, R.S. Rawat, Effects of target–substrate geometry and ambient gas pressure on FePt nanoparticles synthesized by pulsed laser deposition, Applied Surface Science 255 (2009) 4372–4377. [18] F.K. Lotgering, Topotactical reactions with ferrimagnetic oxides having hexagonal crystal structures—I, Journal of Inorganic and Nuclear Chemistry 9 (1959) 113–123. [19] L. Shen, Z.M. Yuan, J.Q. Goh, T.J. Zhou, B. Liu, Y.P. Feng, The effect of introduced defects on saturation magnetization and magnetic anisotropy field of L1(0) FePt, IEEE Transactions on Magnetics 47 (2011) 2422– 2424. [20] C.L. Shen, P.C. Kuo, G.P. Lin, Y.S. Li, K.T. Huang, S.C. Chen, Effect of film thickness on morphology and magnetic properties of CoPt thin films, Nanoscience and Nanotechnology Letters 3 (2011) 262–266. [21] A.L. Patterson, The Scherrer formula for X-ray particle size determination, Physical Review 56 (1939) 978–982. [22] J.J. Lin, S. Mahmood, T. Zhang, S.M. Hassan, T. White, R.V. Ramanujan, P. Lee, R.S. Rawat, Synthesis of Fe3 O4 nanostructures by backward plume deposition and influence of ambient gas pressure on their morphology, Journal of Physics D: Applied Physics 40 (2007) 2548–2554. [23] S. Mahmood, R.S. Rawat, M. Zakaullah, J.J. Lin, S.V. Springham, T.L. Tan, P. Lee, Investigation of plume expansion dynamics and estimation of ablation parameters of laser ablated Fe plasma, Journal of Physics D: Applied Physics 42 (2009) 135504. [24] T. Miyazaki, O. Kitakami, S. Okamoto, Y. Shimada, Z. Akase, Y. Murakami, D. Shindo, Y.K. Takahashi, K. Hono, Size effect on the ordering of L10 FePt nanoparticles, Physical Review B: Condensed Matter and Materials Physics 72 (2005) 144419. [25] Z.Y. Jia, S. Kang, S. Shi, D.E. Nikles, J.W. Harrell, Size effect on L1(0) ordering and magnetic properties of chemically synthesized FePt and FePtAu nanoparticles, Journal of Applied Physics 97 (2005) 10J310. [26] Y.K. Takahashi, T. Ohkubo, M. Ohnuma, K. Hono, Size effect on the ordering of FePt granular films, Journal of Applied Physics 93 (2003) 7166–7168. [27] I.A. Golovchanskiy, S.A. Fedoseev, A.V. Pan, Quantitative model for tunable microstructure in magnetic FePt thin films by pulsed laser deposition, Journal of Physics D: Applied Physics 46 (2013) 215502. [28] N. Kikuchi, S. Okamoto, O. Kitakami, The critical size between single domain and multidomain in L10-FePt particles, Journal of Applied Physics 103 (2008) 07D511. [29] J.J. Lin, Z.Y. Pan, S. Karamat, S. Mahmood, P. Lee, T.L. Tan, S.V. Springham, R.S. Rawat, FePt:Al2 O3 nanocomposite thin films synthesized by magnetic trapping assisted pulsed laser deposition with reduced intergranular exchange coupling, Journal of Physics D: Applied Physics 41 (2008) 095001. [30] Y.L. Li, T.F. Feng, Q. Chen, A.P. Huang, Z.Y. Chen, Promotion of L10 ordering in Pt-rich nonepitaxially grown FePt films, Materials Letters 65 (2011) 2589– 2591. [31] M. Weisheit, L. Schultz, S. Fähler, On the influence of composition on laserdeposited Fe–Pt films, Journal of Magnetism and Magnetic Materials, 290–291 (Part 1) (2005) 570–572.