Correlation of structural ordering with magnetic properties of pulsed laser deposited Co2FeGa Heusler alloy thin films

Accepted Manuscript Correlation of structural ordering with magnetic properties of pulsed laser deposited Co2FeGa Heusler alloy thin films N. Patra, C.L. Prajapat, Rajnarayan De, K.D. Rao, P.D. Babu, A.K. Sinha, Siju John, H.C. Barshilia, S.N. Jha, D. Bhattacharyya PII:

S0925-8388(18)31025-9

DOI:

10.1016/j.jallcom.2018.03.157

Reference:

JALCOM 45387

To appear in:

Journal of Alloys and Compounds

Received Date: 24 October 2017 Revised Date:

29 January 2018

Accepted Date: 12 March 2018

Please cite this article as: N. Patra, C.L. Prajapat, R. De, K.D. Rao, P.D. Babu, A.K. Sinha, S. John, H.C. Barshilia, S.N. Jha, D. Bhattacharyya, Correlation of structural ordering with magnetic properties of pulsed laser deposited Co2FeGa Heusler alloy thin films, Journal of Alloys and Compounds (2018), doi: 10.1016/j.jallcom.2018.03.157. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Correlation of structural ordering with magnetic properties of Pulsed Laser Deposited Co2FeGa Heusler alloy thin films

RI PT

N. Patra1,2, C.L. Prajapat3, Rajnarayan De4, K.D. Rao4, P.D. Babu5, A.K. Sinha6, Siju John7, H.C. Barshilia7, S.N. Jha1 and D. Bhattacharyya1,* 1

M AN U

SC

Atomic & Molecular Physics Division, Bhabha Atomic Research Centre, Mumbai-400 085, India 2 Homi Bhabha National Institute, Mumbai-400 094, India 3 Technical Physics Division, Bhabha Atomic Research Centre, Mumbai-400 085, India 4 Atomic & Molecular Physics Division, Bhabha Atomic Research Centre, VIZAG Centre, Visakhapatnam-530 012, India 5 UGC-DAE Consortium for Scientific Research, Mumbai Centre, Mumbai-400 085, India 6 Indus Synchrotron Utilisation Section, Raja Ramnna Centre for Advanced Technology, Indore-752013, India 7 CSIR-National Aerospace Laboratory, Bangalore-560 017, India Abstract:

In the present contribution, structural and magnetic properties of Pulsed laser deposited (PLD) Co2FeGa (CFG) thin films have been studied as a function of substrate temperature. Structural investigation carried

TE D

out by X-Ray Diffraction (XRD) measurement reveals mixed phase cubic structure of the films which changes from a nearly ordered to highly disordered phase with an increase in substrate temperature. Grazing Incidence X-Ray Diffraction (GIXRD) study shows the presence of disordered A2 phase similar to the bulk

EP

target. Grazing Incidence X-Ray Reflectivity (GIXR) measurement reveals the formation of bi-layer structure in the films with different density and thickness though the overall thickness remains same. Field

AC C

Emission Scanning Electron Microscopy (FESEM) study shows the formation of a droplet-like morphology of the films. Extended X-ray Absorption Fine Structure (EXAFS) study serves a major role in complementing the XRD results and also shows a strong hybridization of Co with Ga maintaining the half metallicity. A novel approach in analyzing the EXAFS data additionally gives quantitative estimation of the different kinds of disorders present in the samples at the atomic level. In addition to this, magnetization result suggests that the films grown at lower substrate temperatures acts as proper ferromagnet following the 1

ACCEPTED MANUSCRIPT

well known spin wave theory with higher value of spontaneous magnetization in comparison to other samples. A lesser value of saturation magnetization in the films compared to bulk also supports the presence of antisite disorders.

RI PT

Introduction:

In the recent years the increasing interest on the materials having high spin polarization have attracted lot of attention due to their possible applications as smart material in various fields such as

SC

Magnetic Tunnel Junctions (MTJ), Spin Transistor, spin injector, localized and itinerant ferromagnet, Magnetic Random Access Memory (MRAM) etc. [1–3]. The half metallic ferromagnets are the only

M AN U

candidates which have the potential to generate energy gap at Fermi level for one spin direction, yielding spin currents having ~100% polarization. However, half metallic ferromagnets, which are obtained in various forms such as oxides (Fe3O4, CrO2), magnetites (La1-xSrxMnO3) [4], double perovskites (Sr2FeReO6) [5], pyrites (CoS2) [6] etc., give more than 90% spin polarization at low temperature and hence

TE D

are not suitable for application in ambient temperature. Co based full Heusler alloys, on the other hand show half metallicity due to their high spin polarization, large minority spin gap, highest magnetic moment and have very high Curie temperature, facilitating their use in spintronic devices at room temperature [7]. Co

EP

based Heusler alloys like Co2MnSi and Co2FeAlxSi1-x have shown excellent performance when used as electrodes in magnetic tunneling junction devices [8]. Ziebeck and Webster [9] were first to synthesize Co

AC C

based full Heusler alloy compounds following which several researchers have worked on both theoretical and experimental aspects of the fabrication and characterization of Co based Heusler alloys in bulk [9,10], thin film and multilayer [11], single crystal [12] and nanoparticle [13] forms. One of the most interesting Co based Heusler alloys is Co2FeGa which have a higher Curie temperature around 1100K [8]. Moreover it has been observed that the half metallicity of this Heusler alloy is stable within the typical variation of lattice parameters of 3–5% [14]. This kind of Heusler alloy generally 2

ACCEPTED MANUSCRIPT

crystallizes in cubic L21 structure in Fm-3m space group with four interpenetrating f.c.c sub-lattices with a stoichiometric composition of X2YZ, where X and Y atoms are the transition metal (3d-group) and Z atom is the sp element (4p-group). This structure is known as the full Heusler alloy structure and when one of the

RI PT

X atoms are missing then it is known as the half Heusler alloy which crystallizes in C1b structure [15]. Apart from this structure these alloys can crystallizes in other cubic structures with reduced symmetry as B2 (Pm3m), A2 (Im-3m) and DO3 structure.

SC

Apart from crystallographic orientation, there are also various kinds of atomic disorders that can be present in a Heusler alloy sample. In case of a X2YZ type of Heusley alloy, if X atoms occupy their assigned

M AN U

position and Y and Z randomly share the other positions B2 structure is obtained, random mixing between X and Y atoms gives rise to DO3 disorder, while in A2 structure all X, Y and Z atoms randomly share their positions [7]. Several traditional analysis approaches are reported to evaluate the degree of atomic ordering in the Full Heusler alloy for bulk samples [16, 17]. This kind of antisite disorder between the atoms leads to

TE D

a decrease in the ferromagnetic ordering and creates additional states at the Fermi level resulting in reduction in spin polarization [18]. In the Co2FeGa Heusler alloy both the Fe and Ga atoms are situated at the octahedral symmetry co-ordinated with 8 Co atoms while the Co atoms is co-ordinated with 4Fe and

EP

4Ga atoms in the 1st co-ordination shell and 6 Co atoms in the 2nd co-ordination shell reducing the structural symmetry to tetrahedral one. In Co2FeGa Heusler alloy the magnetization is controlled by Fe atoms.

AC C

However, whenever there is an antisite disorder between Fe and Co atoms it reduces the spin magnetic moment of Fe and reduces the ferromagnetic interaction and spin polarization. In such systems, interaction between the Co and Fe/Ga atoms are the most important tool for explaining their magnetic interactions. A strong hybridization between the transition metals and the sp elements gives rise to the formation of the band gap at the Fermi level. [15]. Thus study of structural and magnetic properties of these Heusler alloys and the correlations among these properties is an important subject of current research interest. 3

ACCEPTED MANUSCRIPT

It is obvious that for any spintronic applications like in magnetic tunnel junction (MTJ) devices, the above alloys are needed to be prepared in thin film form. Therefore, a number of papers have been concerned with both theoretical aspects of half-metallic behavior of the Heusler Alloy thin films as well as

RI PT

with technological aspects aimed at fabricating these films with proper crystallographic structure and high spin polarization [19, 20, 21]. However, preparation of the Heusler alloy in thin film form is a much more challenging task from the perspective of material chemistry and chemical preparation. Besides chemical

SC

stoichiometry, a well ordered Heusler phase is also expected to maintain its half metallicity and hence high spin polarization, through L21 and B2 structures. In a Heusler alloy thin film the structural disorder due to

M AN U

lattice mismatch between the substrate and sample and the phase separation are frequently encountered. The problem of ordering in Heusler alloy thin films is one of the key issues for the application in spintronics devices. Apart from structural orientation, surface or interface of the thin films are also important as the half metallicity is critically connected with the surface and interface properties of the sample [22]. Difficulties

TE D

are encountered in preparation of these films with good ordering and without any trace of oxidation. Several deposition processes of the Co based Heusler alloy thin films have been reported such as RF magnetron Sputtering [23], Co-sputtering [24], Ion beam deposition [25] and Pulsed laser deposition (PLD) [12, 22,

EP

26].

Among these, PLD is a technique that makes it possible to ablate the target material instantaneously

AC C

and impinge it on the substrate with high kinetic energy which helps in maintaining the stoichiometry of the target in the films. Moreover, the high kinetic energy of the adatoms reaching on the substrate assists migration of the film precursors on the surface and thereby makes the film growth at low temperature possible with desired crystalline structure and morphology. In PLD it is also possible to control the growth of the films by varying parameters such as the laser energy, the pulse repetition rate, fluency (i.e., energy density), the target to substrate distance etc. PLD is now established as one of the simple and versatile 4

ACCEPTED MANUSCRIPT

methods of growing thin films of a very wide range of materials (including numerous complex oxide systems, ceramics, ferroelectrics, high-Tc superconductors and materials exhibiting giant magnetoresistance) on a wide variety of substrates [27–29]. Good stoichiometry, morphology and crystalline quality

RI PT

are essential for the films to conserve the ferromagnetic and half-metallic properties of the bulk material. To achieve these requirements in the present work we have used PLD to prepare the Heusler alloy films. To obtain the structurally ordered Heusley alloy thin films generally two approaches have been explored, viz.,

SC

(i) employing post deposition annealing and (ii) depositing at elevated substrate temperature [30]. In the present communication, the properties of a series of PLD grown Co2FeGa (CFG) thin films

M AN U

prepared at different substrate temperatures have been discussed. The goal in this investigation is to find out the correlation between the long and short range structural order in the samples with their magnetic properties. A combined approach of lab source XRD, GIXRD and GIXR measurements alongwith synchrotron based XRD have been used to characterize the bulk structure and surface of the thin films,

TE D

while synchrotron based element specific EXAFS technique has been utilized to provide the information on short range order in the samples. The above structural characterizations have been used to explain the magnetic ordering present in these films. These Heusler alloy samples are prone to have antisite disorders

EP

between the constituent metallic species, however lab based XRD measurement cannot determine this quantitatively due to similar X-ray scattering factors of the elements. Here we have used a novel

AC C

methodology in analsysing the EXAFS data measured at the absorption edges of the different species to find out the antisite disorders quantitatively.

Experimental details: Sample preparation:

5

ACCEPTED MANUSCRIPT

CFG thin films of ~1200Ǻ thickness were deposited from a smoothly polished commercially obtained circular stoichiometric target of 3 inch diameter and 3 cm thickness in a PLD system using a KrF excimer laser (248 nm, 10 Hz pulse) with output energy of 0.8J/pulse. Prior to depositions 10 mm×10 mm

RI PT

size Si (111) substrates were cleaned for 5 min. by placing them in a crucible inside of an ultrasonic bath followed by drying with blowing Argon gas. The target was placed at 45̊ to the incident laser while the substrate was heated to a high temperature for 30 min. to allow the surface reconstruction after which the

SC

temperature was set to the desired temperature for deposition. The thin films were grown at substrate

M AN U

temperatures of RT, 473K, 673K, 873K and 1073K and at a base pressure of 3×10-6 mbar.

Characterization techniques:

For the structural characterization of the CFG films room temperature XRD and GIXRD measurements have been carried out on a Rigaku 3 kWatt Smart Lab X-ray diffractometer equipped with a

TE D

Cu Kα X-ray source operated at 40 kV, 30 mA. For both the measurements the experimental data have been taken within the 2θ range of 20-90 ̊ with a step size of 0.01̊. For GIXRD measurement the angle of incidence was kept at 0.5̊ to avoid substrate contribution. Along with this for a better understanding of the

EP

structural phases of the thin films, synchrotron radiation based hard X-ray diffraction study at two different energies has been carried out at the Angle Dispersive X-Ray Diffraction beamline (BL-12) at Indus-2

AC C

synchrotron source (2.5GeV, 200 mA) at Raja Ramanna Centre for Advance Technology (RRCAT), Indore. The microstructure of the metallic films has been characterized by FESEM measurements with different accelerating voltage and magnification for better understanding about the surface morphology. The elemental composition of the films has been characterized by Energy Dispersive X-ray Spectroscopy (EDS) tool attached to the FESEM setup. The local atomic structure of the samples have been investigated by Xray Absorption Spectroscopy (XAS) measurement comprising of Extended X-ray Absorption Fine Structure 6

ACCEPTED MANUSCRIPT

(EXAFS) and X-ray Absorption Near Edge Structure (XANES) measurements at the Fe (7112 eV), Co (7709 eV) and Ga (10367 eV) K-edges. The XAS measurements have been carried out in fluorescence mode at the Energy Scanning EXAFS beamline (BL-09) at the Indus-2 Synchrotron Source. The details of

RI PT

the beamline which comprises of a double crystal monochromator and two meridional cylindrical mirrors have been described in the previous communications [31, 32]. For EXAFS measurement in the fluorescence mode the thin films were placed at 45̊ angle to the incident beam and the fluorescence signal (If) were

SC

collected using the Si drift detector placed at 90̊ to the incident beam. A 30 cm long ionization chamber filled with a predefined gas mixture to absorb 10-20% of the incident beam was placed prior the sample

the relation ( µ =

If I0

M AN U

stage to measure the incident beam (I0). Thus the absorption co-efficient of the sample was determined by ) and the spectrum was obtained as a function of energy by scanning the double crystal

monochromator of the beamline over the specified energy range. Finally in order to characterize the magnetic behaviour of the samples with varying temperature and external magnetic field, the magnetic

TE D

measurement were carried out in a temperature range of 5–350 K using a PPMS-9T platform with the Vibrating Sample Magnetometer (VSM) module by placing the sample surface parallel to the direction of

EP

magnetic field.

AC C

Result and Discussions: X-ray diffraction:

The room temperature X-ray diffraction patterns using laboratory based Cu Kα source, of the CFG Heusler alloy thin films grown at different substrate temperatures (TS) along with that of the bulk (target) material have been shown in Fig.1. In order to estimate about the value of the lattice parameter the structural refinement of the target material was performed using the FullProff code [33] based on the Rietveld analysis

7

ACCEPTED MANUSCRIPT

process. The refinement depicted in Fig.1 has been carried out assuming the cubic crystal structure of X2YZ (or XX’YZ) Co2FeGa with space group Fm-3m, where the cubic structure consists of four interpenetrated f.c.c lattice in an unit shell where Co (X) atom occupies the Wyckoff position at 4c (¼, ¼ , ¼ ), Fe occupy

RI PT

at 4a (0, 0, 0) and the Ga atoms at the 4b (½, ½, ½) [34]. The refinement result shows the value of lattice parameter a = 5.748Å, which is in good agreement with the reported result for the bulk sample [25]. There are different kinds of order structures present in a Co2FeGa Heusler alloy and the only way to

SC

identify the various disorder is to monitor the relative intensity of the (111) and (200) super lattice reflections along with the fundamental reflection (220). (220) peak should always be present in the Co2FeGa

M AN U

full Heusler alloy system irrespective of whether atomic disorder is present or not in the sample, whereas the appearance of the super lattice (111) peak confirms the presence of L21 phase and that of the (200) peak additionally with (220) peak confirms the presence of B2 phase. From the XRD pattern of the target, presence of three reflections of (220), (400), (422) planes are clearly observed where the (220) peak, known

TE D

as the fundamental diffraction peak of a full Heusler alloy (X2YZ) structure appears to be the most intense. All these diffraction patterns correspond to the rule (h+k+l) = 4n reflection, where n is a positive integer. However, absence of the (111) and (200) super lattice reflections in the XRD pattern of the target indicates

EP

the absence of the fully ordered L21 and B2 phases [35] and presence of the A2 phase in the target which occurs due to the atomic disorder between Co (X)-Fe(Y), Ga(Z) atoms [7, 24]. Broad diffraction peaks in

AC C

the XRD pattern of the target material manifests small crystallite size of the samples. In the room temperature XRD pattern of the thin films, however, the (111) peak appears to be the most significant along with weak (220) peak which signifies the absence of fully ordered L21 phase and presence of significant amount of atomic disorder in the films. According to the analysis approach introduced by Webster, the degree of order of the L21 and B2 structure is evaluated from the ratio of the intensity of the odd and even super lattice reflections to the intensity of the fundamental reflection peak [9]. 8

ACCEPTED MANUSCRIPT

The fundamental diffraction peak which follows h+k+l = 4n reflection is independent of atomic ordering. The even reflection which is defined by h+k+l = 4n+2 e.g. (200) suggests the formation of the B2 phase. The above mentioned structural phases occur due to the exchange of the site occupancy of the constituent

RI PT

atoms from their respective lattice sites. However, the absence of (200) reflection in the present set of CFG films manifests the inability of lab XRD measurements to identify the B2 phase arising due to disorder in atomic positions. Also due to similar atomic scattering factors of Fe and Co atoms, at the Cu Kα energy of

SC

8046 eV or 1.54 Ǻ the Co/Fe (DO3) type disorder and Co/Ga (A2’) type disorder merges and could not be distinguished from the above measurements.

M AN U

XRD pattern for all the samples shows that the intensity of the (220) peak remains the same throughout the series though the intensity of the (111) peak decreases with an increase in the substrate temperature and becomes negligibly small for the film deposited at the highest substrate temperature. This observation suggests that with an increase in the substrate temperature the fraction of the ordered L21 phase

TE D

decreases manifesting significant increase of disorder in the samples. Isber et.al. have also observed similar degradation of crystallinity of their PLD grown Co2MnAl Hesler alloy films with increase in substrate temperature [36]. Yadav et.al. have pointed out that the absence of the (200) and (111) peaks is suggestive

EP

of the fact that the coherence length of the B2 and L21 types of ordering does not exceed the sensitivity limit of the lab source XRD [24]. In case of the present CFG films the intensity ratio of the (220) and (422) i.e.,

AC C

I220/I422, varies from 0.81 to 1.05, which is much smaller than the value 2.0 for a polycrystalline film. Therefore the XRD pattern does not suggest formation of any textured growth [27]. The (111) reflection peak appears to be much broader in comparison to the other significant peaks. The broadness of this peak gradually decreases with an increase in the substrate temperature manifesting an increase in crystallite size. The crystallite size (D) of the samples shown in the inset to the Fig.1 has been estimated by considering (111) reflection as the reference peak using the well known Debye-Sherrer 9

ACCEPTED MANUSCRIPT

formula, D = 0.9λ/βCosθ, where β be the FWHM of the peak, λ be the wavelength of the X-ray and θ be the position of the peak. The result shows a linear increase in the crystallite size with the substrate temperature and reaches to a maximum value of 152 Å for the TS = 673K sample. However, the grain size abruptly

RI PT

decreases when the substrate temperatures is increased beyond 873K. Similar kind of observations has also been made by Yadav and Chaudhary for the case of the Co2FeAl thin films [24]. It has been suggested that with an increase in substrate temperature the diffusion of the ad atoms increases promoting the island-

SC

coalescence and resulting in an increase of the crystallite size. However, at still higher substrate temperature a higher growth rate in the transverse direction may cause the reduction in the lateral grain size [37]. It may

M AN U

be noted here that for the films grown at 873K and 1073K additional small peaks appear within the 2θ range of 34.14–35.46° and 37.73–38.41° respectively which might be due to experimental artifacts [38] and also there is a peak at 25̊ due to Si (111) substrate which is present in the samples grown at RT, 473 and 673 K

TE D

and absent in the films deposited at 873 and 1073 K.

X-ray diffraction with Synchrotron radiation:

We have observed above that the laboratory based XRD (Cu Kα energy) measurement was unable to

EP

distinguish the disorder between the Co/Fe and Co/Ga because of the same atomic scattering factors of Co (fCo) and Fe (fFe) atoms. However, there is a huge contrast in the atomic scattering factors of the atoms at

AC C

their respective absorption edge energy which is 7.112 KeV for Fe and 7.709 KeV for Co atoms due to the phenomena known as Anomalous X-ray Scattering (AXS) [39]. Therefore the superlattice diffraction due to the DO3 and B2 phase will be reflected in the XRD spectra. In our case this has been done by carrying out the measurements at energies of 15.47 KeV and 7.11 KeV using a Synchrotron based X-ray source. Here the XRD data measured at higher energy of 15.47 KeV is equivalent to the lab source based XRD measurement, where both Fe and Co atoms have almost similar atomic scattering factors and the XRD pattern obtained 10

ACCEPTED MANUSCRIPT

with lower X-ray energy of 7.112 KeV shows the anomalous scattering having different atomic scattering factors. Fig.2 shows the XRD patterns of the film deposited at Room Temperature at both the energies. All the figures show identical nature of the XRD pattern as that of the lab source based data manifesting

RI PT

preferred orientation of the films along (111) direction and the (200) peak which represents the presence of B2 phase could not been observed in the samples in this measurement also. From these studies it can be concluded that in the CFG films the order L21 phase acts as major phase at the low growth temperatures and

SC

it diminishes with increase in the growth temperature.

M AN U

Grazing Incidence X-ray Diffraction:

In order study about the surface structure of the CFG films, GIXRD measurements within the 2θ range of 20-90° have been carried out with a constant grazing angle of incidence of 0.5°. Fig.3 shows the GIXRD spectra of the CFG thin films deposited at different substrate temperatures. The two peaks for (220)

TE D

and (422) planes appearing within the 2θ range 44.2-45.2° and 81.2-83.2° agree with that of the target material, which indicates the dominant presence of the A2 phase at the surface. However the (111) reflection could not been observed in the GIXRD pattern of the samples, which does not agree with the

EP

XRD result of the samples. Though for the film deposited at the highest substrate temperature the (220) peak almost vanishes, a gradual increase in the intensity and decrease in the FWHM of the peak upto the

AC C

film deposited at 873K shows the increase in the crystallinity at the surface due to reduction of the internal strain in comparison to the bulk. GIXRD measurement shows that the surface structure of the thin film samples are quite different from their bulk structure which is possibly due to the bi-layer structure of the samples as revealed by the GIXR measurement discussed below.

11

ACCEPTED MANUSCRIPT

Grazing Incidence X-ray reflectivity (GIXR) measurements: In order to estimate the density, thickness and roughness of the CFG films grown at different temperatures, specular grazing incidence X-ray reflectivity measurement of the thin film samples have been

RI PT

carried out using a diffractometer by varying the 2θ angle from 0-4˚. The values of the parameters have been estimated by fitting the experimental data with a theoretical model using the IMD software. The experimentally obtained data show good fitting with the simulated GIXR plot with a bi-layer model of

SC

CFG1/CFG2/Si with different density and thickness of the two layers as shown in Fig.4(a). During the fitting process the values of the thickness of the bottom layer was kept constant at 1200Ǻ for all the

M AN U

samples, which was obtained from the ex-situ thickness calibration of the films with different times of laser ablation, while the density of the bottom layer and thickness of the upper layer, density and surface roughness have been varied independently. The theoretical reflectivity plots fit well with the experimental data except for the film deposited at 873K. No distinct Kissig fringes have been observed in the GIXR data

TE D

of the films due to their large thickness except for the film deposited at 673K and the data becomes noisier for the films deposited at 873K and 1073K manifesting an increase in surface roughness of the films deposited at higher temperatures. The increase in surface roughness of the films deposited at higher

EP

substrate temperature can be attributed to high temperature induced inter-grain agglomeration of the films [25]. These observations are quite consistent with observed FESEM back scattering images of the films as

AC C

discussed below. The variation in the estimated value of the GIXR parameters for the two layers has been shown in Fig.4(b)–(d). The density of the lower layer remains ~ 9.25 gm cm-3 for the film deposited at RT, 473K and 673 K, which is close to the value of a standard heusler alloy compound, however it drastically falls below 6 gm cm-3 for the samples deposited at substrate temperatures of 873K and 1073K. The bi-layer structure of the samples with different growth morphology explains the difference in the structural information obtained for the films from XRD and GIXRD measurements as discussed above. 12

ACCEPTED MANUSCRIPT

Field Emission Scanning Electron Microscopy (FESEM): The back scattering electron micrographs of the ~1200Å thick CFG films grown at various substrate temperatures have been presented in Fig.5. The morphologies of the films show droplet-like structure grown

RI PT

on an otherwise smooth surface with average size of the droplet varying from 1 to 6 µm. Similar droplettype morphology has been observed by other workers also for the PLD grown films [26]. The magnified image of the circular droplet has been shown in the inset of Fig.5. As the substrate temperature is increased

SC

beyond 473K, densities of the droplets increase drastically and the droplets are agglomerated to each other to form a homogeneous microstructure, though it assumes a rough surface as observed from the GIXR data.

M AN U

The formation of the grain structure is observable for the film deposited at 873K and it also shows a surface with deep trenches with the size of the grains varying from 100 nm to 300 nm. Microstructure of the present samples match with the microstructure of the epitaxially grown films of CoCrFeAl reported by Jacob et.al.

TE D

[40].

Energy Dispersive X-ray Spectroscopy (EDS) measurements: In order to get an estimation of the composition of the CFG films, the EDS spectra have been taken

EP

by averaging the data collected over different regions of the film. Figs.6 (a) and (b) show the EDS spectra of the samples grown at room temperature and at a substrate temperature of 1073K, while the variation in the

AC C

stoichiometry as a function of substrate temperature has been shown in Fig.6(c). The results indicate formation of off-stoichimetric films in which ratio of the metallic components slightly deviates from their ideal stoichiometric ratio of (C:F:G=2:1:1) except for the films deposited at RT and 673K where the Co and Fe maintain the ratio of 2:1. However the films are rich in Fe and Co and deficient in Ga content throughout the series with a drastic variation in the ratio of Co and Ga being observed for the films grown at the highest substrate temperature. Such kind of deficiency in Ga atom is likely to happen in PLD grown systems beyond 13

ACCEPTED MANUSCRIPT

certain temperature limit. This happens due to preferential scattering or sputtering of the ablated atoms at that particular substrate temperature [41]. Riet et.al. have explained that Ga atoms have a higher driving force for being segregated from the surface which enhances the preferential sputtering of Ga atoms from the

RI PT

film surface [42]. No significant peak of oxygen which can correspond to the oxide formation has been observed in the EDS spectra of the samples, rather a large Si peak around 1.83 KeV is observed as shown in the Fig.6 (c). This manifests significant diffusion of Si atoms from substrate to the film [43] with an increase

SC

in the substrate temperature, which may lead to silicidation reaction due to high reactivity of Co with Si atoms. This observation correlates with the drastic decrease in the magnetization value of the sample with

X-ray absorption spectroscopy (XAS):

M AN U

an increase in substrate temperature as discussed later.

Further, XANES and EXAFS measurements have been carried out at Co, Fe and Ga K-edges on the

TE D

CFG thin films grown at different substrate temperatures. The main goal of this XAS analysis is to provide the information about the local structure around the respective atoms and to investigate the disorder effects on the atomic structure which could not been well understood from the X-ray diffraction study. Several

EP

other authors have also indicated that determining structures of Heusler alloys, particularly when disorders are present, by employing only X-ray diffraction is not very fruitful [13,44,45] and XAS can be a

AC C

complementary method for determining structural information unambiguously. Figs.7 (a)-(c) show the normalized XANES spectra ( µ (E ) versus E ) of the films at Co, Fe and Ga K-edges deposited at different substrate temperatures. XANES results show that that in all the samples the absorption edges match with their respective metallic states indicating metallic nature of the samples throughout the substrate temperature range. However a significant change above the absorption edge can be noticed for the films deposited at 873 K and 1073K. At both the Co and Fe K-edges the appearance of a shoulder and the 14

ACCEPTED MANUSCRIPT

variation in the oscillations in the XANES feature above ~30 eV of the absorption edge ( E 0 ) may suggest a variation in the structural symmetry in these samples. In order to get a clear understanding about the local structure around the different metallic atom

RI PT

species present in the CFG samples, the experimentally obtained EXAFS absorption data have been analyzed. To extract only the oscillatory part from the µ (E ) vs. E absorption spectra, the theoretically simulated spline function for the bare atom background is removed from the spectra and the energy

SC

dependent absorption coefficient µ (E ) is subsequently converted to the wave number dependent absorption

co-efficient χ (k ) and finally χ (k ) was weighted by k 2 to amplify the oscillations in the higher k region.

M AN U

The k2 weighted χ(k) plots at Co, Fe and Ga K-edges have been shown in Figs.8 (a)-(c) for all the samples. The k 2 χ ( k ) vs. k plots have subsequently been Fourier transformed to yield the radial distribution function or χ (r ) vs. r plots. The above exercise has been done using the ATHENA subroutine of the IFEFFIT software package [46] following the standard procedure mentioned earlier [47, 48]. The experimental χ ( r )

TE D

vs. r data were subsequently been fitted with theoretically generated plots using the ARTEMIS subroutine. To generate the theoretical FT-EXAFS spectra or the radial distribution function, the crystal

EP

structure i.e., lattice parameters, space group and Wyckoff positions of the constituent atoms obtained from the XRD study has been employed in the ATOMS [49] subroutine of the IFEFFIT code as an input

AC C

parameters. The atom to atom bond distance ( R ) and the Debye-Waller factor ( σ 2 ) which give the meansquare fluctuations in the bond distances of the respective scattering paths have been varied independently without any constraint during fitting. However the energy shift E 0 was varied for each edge independently for the proper alignment of the energy axis between the theory and the experimental data while the multi electron excitation factor ( S 02 = 0.9) was kept constant throughout the fitting process.

15

ACCEPTED MANUSCRIPT For the above exercise, the experimentally obtained k2 weighted χ(k) data over the k range of 2.5 – 10.5 Å-1 have been used for Fourier transform using Hanning window function and the fittings have been carried out within the R range of 1.0-4.1 Å. The goodness of the fit in the above process is generally factor

which is defined as:

R factor = ∑

where,

χdatand χ th

RI PT

R

[Im( χ dat ( ri ) − χ th ( ri )]2 + [Re( χ dat ( ri ) − χ th ( ri )]2 [Im( χ dat ( ri )]2 + [Re( χ dat ( ri )]2

refer to the experimental and theoretical χ ( r ) values respectively and Im and Re refer

to the imaginary and real parts of the respective quantities.

SC

expressed by the

M AN U

Initially the EXAFS data of the samples have been fitted with theoretical spectra generated assuming a structural model without considering any antisite disorder which might arise due to interchange in the site occupancies of the atoms. A schematic diagram of the atomic configuration surrounding the Co, Fe and Ga atoms in the L21 structure have been shown in Fig.9 where the arrows show the scattering paths of the 1st

TE D

co-ordination shell of all the atoms. In the crystal structure the Co atom is coordinated with an intermixing of the 4 Fe and 4 Ga atoms with exactly equal bond distances of 2.48Å in the 1st co-ordination sphere [50] while the 2nd and 3rd co-ordination shells at 2.87Å and 4.05Å are due to Co atoms. On the other hand as the

EP

Fe and the Ga atoms are situated at the equivalent site of each other, the Fe or the Ga atoms are surrounded by 8 Co atoms in the 1st co-ordination shell and 6 Ga or Fe atoms in the 2nd shell forming an octahedron.

AC C

The 3rd shell comes at 4.01 Å surrounded by 12 Fe or Ga atoms from the Fe or Ga absorber. The phase uncorrected χ(R) versus R plots at all the three edges fitted with the theoretically generated model (without antisite disorder) have been shown in the Figs. 10(a)-(c). The positions of the different back scattering paths for the different absorbers have also been shown in Fig.10. For the case of the Co edge the 1st intense peak that appears within the R range of 1.4-2.4 Å is due to the combined effect of the 4 Fe and 4 Ga atoms as explained above and the back scattering path due to 6 16

ACCEPTED MANUSCRIPT

Co atoms contributes to the shoulder within 2.5-2.8Å distance. Similarly for the case of the Fe or Ga edges the 1st peak within the R range of 1.5-2.5 Å appears due to the back scattering from 8 Co atoms only and the next lesser intense peak within the R range of 2.5-2.8 Å comes due to the back scattering from 6 Ga or

RI PT

Fe atoms respectively. It is worth noting that within the above mentioned atomic arrangements two multiple scattering paths with lesser scattering amplitudes also appear but these paths have been neglected in the fitting process due their non-colinearity and insignificant contributions. As the 1st co-ordination shell of the

SC

Co atom is contributed by 4 Fe and 4 Ga atoms and for Fe or Ga atom it is contributed by 8 Co atoms the ratio of intensities of the path due to the Fe or Ga shell to that of the path due to Co coordination shell

M AN U

should be 1:2 regardless to the fact that Fe and the Ga atoms have different scattering factors [51]. However, in our case none of the Fe or Ga paths at the Fe or Ga edges follows this relation which might be due to the high structural disorder present in the thin films. The splitting in the 1st peak at the Co and Fe edges has been observed for the films deposited at 873K and 1073 K and the 1st intense peak is shifted towards lower

TE D

R value by a little amount. The splitting of the 1st peak at Co edge is worth noting as the Fe and Ga atoms

having different atomic scattering factors contribute to the same co-ordination shell. To facilitate the comparison between the values of the EXAFS parameters of the cubic phase the

EP

results have been summarized in Table-1, 2 and 3 for the Co, Fe and Ga K-edges respectively. The results show little difference between the bond lengths of the Co-Fe and Co-Ga pairs. Similar difference in Co-Fe

AC C

and Co-Ga paths have been observed by other workers also in Co2FeGa Hesuler alloy samples due to deviation from a perfect ordered L21 structure [50]. It is to be noted that for an ideal cubic structure as the Co atom is situated at the body centre position so the Fe and the Ga atoms will be equidistant from the Co atom forming a Co-Fe-Ga triad. An important aspect that emerges from EXAFS data analysis at Co edge is that the Co-Ga bond length is slightly shorter than the Co-Fe bond length and have lesser values of σ 2 suggesting the rigidness of the Co-Ga bond in comparison to its counterpart. The rigidness of the bond may 17

ACCEPTED MANUSCRIPT

suggest to a stronger hybridization between the Co-Ga atoms, which is the characteristic of a half metallic alloy. This kind of hybridization between the transition metals leads to the formation of band gap near to the Fermi level, which is highly recommended for a half metallic Heusler alloy [15]. A negative shift in the Fe-

RI PT

Co bond length for films grown at 873K and 1073K may also suggests the decrease in the crystallite size as observed from the Debye-Sherrer formula [52]. The Fe K-edge result also does not show any dramatic change in the χ(R) vs. R plot and corresponds to the result obtained from the Co edge. The result shows that

SC

Co-Fe and Fe-Co bond lengths remain almost same maintaining the perfect cubic structure though the GaCo bond lengths deviate from the Co-Ga bond distance in the 1st co-ordination shell. This kind of

M AN U

observation is not expected for an ideal cubic structure and the difference in this correlation agrees with the argument that the atoms move from their crystallographic positions by little order of amplitudes [53]. We have seen from the XRD results that in the samples there is significant presence of A2 phase along with the ordered L21 phase though XRD measurement is unable to distinguish and quantify the amount of

TE D

disorder present between the Co and Fe atoms due to similar atomic scattering factors of the two elements. Next, we introduce a model which can give an idea about the amount of antisite disorder present between the atomic species and their effect to the crystal structure. In order to employ this model to our EXAFS

EP

analysis, the theoretical predictions that has been used by Takamura et.al. for explaining the anomalous Xray diffraction study of Co2FeSi thin films [44] based on the physical model (known as NRBB model)

AC C

proposed by Niculescu et.al. [54] has been adopted. This model includes few constraint relationships between the scattering paths by introducing three disorder parameters. The disorder parameters those have been used in the model are indicated as α, β and γ, where the parameter α gives the probability of occurrence of the Fe (Ga) atoms at the Ga(Fe) site, β gives the probability of occurrence of the Co(Ga) atoms at the Ga(Co) site and the γ gives the probability of occurrence of Co(Fe) atom at Fe(Co) site. These sorts of disorders are likely to occur in case of the half as 18

ACCEPTED MANUSCRIPT

well full Heusler alloy like Co2MnSi [52]. In order to introduce the above antisite disorders in the EXAFS analysis of the present samples, few FEFF input paths corresponding to the 1st, 2nd and 3rd shells have been added in the ATOMS program by giving them same bond length and Debye-Waller factor to their proper

RI PT

sited paths by replacing only the backscattered atoms. Thus the fitting standards were parameterized by multiplication of the disorder quantities with the co-ordination of the improperly sited paths at respective absorption edges. As an example, for the Co (Fe) and Co (Ga) disorder the disordered Co-Ga path is

SC

populated by CN×β and disordered Co-Fe path is populated by CN×γ. On the other hand the ordered Co-Fe path is populated with CN× (1-β) and the ordered Co-Ga path is populated with CN×(1-γ). As in the crystal

M AN U

structure the Co atoms are situated at X position with 2 Co atoms per atom the population of the proper Co– Co paths has been given as (CN/2)×(1-β-γ). In the fitting process α, β and γ have been used as fitting parameters. It should mentioned here that we have estimated the values of Co-Ga/Ga-Co and Co-Fe/Fe-Co and Ga-Fe/Fe-Ga disorder from the Co and Ga K-edge data and have used these values for the fitting at Fe

TE D

K-edge data. Similar methodology was used for the Fe(Ga) type of disorder at the Fe K-edge. It should also be noted that the fitting of the experimental data has been done by taking the data similar to the data range mentioned for the fitting without antisite disorder discussed above to have a comparison. Similar approach

EP

has been used by Ravel et.al. also for the EXAFS analysis of Co2MnSi thin films also [52]. The phase uncorrected χ(R) versus R plot along with its real and imaginary parts for the film

AC C

deposited at RT, fitted with the theoretically generated model at all the three edges have been shown in Figs. 11 (a)–(c). The fitting in both the real and imaginary parts demonstrate very good quality of fitting achieved using the above model. The values of the EXAFS best fitted parameter alongwith the disorder parameters have been shown in Tables 5-8. Table 5 shows that in the samples there is a presence of significant amount of chemical disorder and a phase mixture of them. It also shows a negligible amount of disorder between the Fe and Ga atoms (B2 disorder) which remains below 2% though a significant amount of the A2 (Co/Fe/Ga) 19

ACCEPTED MANUSCRIPT

and DO3 (Co/Fe) disorders are present in the sample where the DO3 (Co/Fe) disorder dominates over other (~48%) ordering. The values of the disorder parameters are found to increase gradually with an increase in substrate temperature which can be clearly correlated with the XRD results of the samples. From Tables 6-8

RI PT

we can see that distances between the various metal atom pairs follow the following trend: Co-Co’≥ Co-Fe and Fe-Fe’< Fe-Co, Co-Co’ ≥ Co-Ga and Ga-Ga’< Ga-Co, Fe-Fe’ ≥ Fe-Ga and Ga-Ga’>Ga-Fe, where the primed numbered atoms represent the atoms at the improper sited positions. It can be seen from above that

SC

among the different distances, the Fe–Ga disorder does not follow the proper correlation as per their atomic radii (atomic radii of Co, Fe and Ga are 1.25Å, 1.24Å and 1.22Å respectively). So it can be said that the Fe-

M AN U

Ga disorder or (B2 type disorder) obtained from the above EXAFS analysis is over-estimated and may not be present in the samples significantly, which corroborates with the XRD results also as discussed above. Thus from Figs. 10 and 11 it can be noticed that using the present methodology the quality of fitting of χ (r ) vs. r data improves significantly in comparison to the previous one, which is also reflected in the R

factor

in Tables 6-8 compared to in Tables 1-3 and thus the above EXAFS analysis of the

TE D

lower values

Co2FeGa thin films reveals the compositional dependent variation in the local structure which XRD measurements was unable to explain properly. The results obtained from the above analysis using the

EP

NRBB model is quite successful to describe the occurrence of antisite disorder which is consistent with magnetization properties of the samples discussed later. The results obtained from this study also manifests

AC C

that in the lattice Ga atoms are much strongly bound to the Co atoms than the Fe atoms though the structure remains otherwise cubic.

Magnetization Study:

Co2FeGa is known to have half metallic (X2YZ–type) ferromagnetic property where the main magnetic moment is localized at the Fe atom situated at Y position which is ferromagnetically coupled to 20

ACCEPTED MANUSCRIPT

the Co atom situated at X position for a properly sited L21 structure. To characterize the magnetic behavior of the PLD grown CFG/Si films the externally applied magnetic field dependent in-plane magnetization M (H) have been carried out at three different temperatures of 5K, 300K and 350K with varying the magnetic

RI PT

field upto 9T. The magnetization curves of the films with different substrate temperatures and measured at the above three temperatures have been depicted in Figs.12 (a)–(c), while Figs.13 (a)–(e) show the enlarged view of the in-plane hysteresis curves of the samples near the low magnetic field. The diamagnetic

SC

contribution of the c-Si substrate has been subtracted from the magnetization of the films prior to the calculations. Fig.13 shows a good correspondence of the magnetization property with the crystallinity of the

M AN U

film which decreases gradually with an increase in the substrate temperature. All the samples show soft ferromagnetic nature with a very small amount of coercive field (Hc) and started to saturate at low magnetic field. The saturation magnetization value of the films have been calculated by linearly fitting the M (H) curve at very high magnetic field and then extrapolating the line towards the H=0 line or the intercept of the saturation magnetization (Ms),

TE D

straight line. The values of the various magnetization parameters i.e.,

coercive field (Hc), remnant magnetization (MR) and the squareness of the curve which is the ratio between the remnant magnetization to the saturation magnetization (MR /Ms) have been obtained from Figs.13 (a)–

EP

(e) and the values have been depicted in Figs. 14 (a)-(d). It should be noted that no significant variations between the values of the parameters have been observed at temperatures of 300 K and 350 K with increase

AC C

in the substrate temperature which may be due to high Curie temperature of the films compared to the temperature of measurement. The result shows that the films deposited at RT and 473 K have very small amount of coercive field (~2.5 mT or 25 Oe) and high remnant magnetization in comparison to the others suggesting soft ferromagnetic behavior of these films. The coercive field increases for the film deposited at 673 K, however, it again decreases drastically as the substrate temperature is increased further. The decrease in the coercive field with an increase in the 21

ACCEPTED MANUSCRIPT

substrate temperature for a magnetic thin film has been observed by other workers also for PLD deposited Co2FeSi films [55]. The highest value of HC for this sample may be due to a change in the magnetic anisotropy [25]. It has been observed from the above XRD and EXAFS studies that the structural properties

RI PT

of the films deposited at 873K and 1073K are similar and are significantly different from the films deposited at relatively lower substrate temperature. The sudden decrease in the value of Hc for these films is thus attributed to the presence of relatively higher structural disorder in the films with a decrease in the crystallite

SC

sizes. The M ( H ) curves of these films at very low field region and at 300K and 350K show lesser ferromagnetic behavior with negligible amounts of remnant magnetization which is a characteristic of

M AN U

paramagnetic ordering.

The film deposited at RT and 473K have higher squareness value (~70%) suggesting lesser defect in these films. However with increase in substrate temperature the squareness ratio suddenly falls to 40-50% and later vanishes with higher defect.

TE D

Galanakis et.al. have shown by ab-initio calculations that electronic and magnetic properties of these compounds are intrinsically co-related with their minority spin density of states [15]. It has also been explained that according to the Slater-Pauling rule, the total spin magnetic moment (Mt) of the Co2FeGa

EP

types of full heusler alloy is defined by the relation Mt = Zt –24, where Zt is the number of valance electrons. The compounds containing Al and Ga have total spin magnetic moment ~ 4µ B from Slater-Pauling rule.

AC C

Experimentally it has been observed that the maximum value of the saturation magnetization (Ms) ~ 2.68µ B/f.u for the RT grown film, which is much smaller than the value obtained experimentally by Umetsu [8] et.al. (~5.2 µ B/f.u) and Galanakis et.al (~4.89µ B/f.u) for bulk polycrystalline samples [56]. With an increase in the substrate temperature it decreases to ~2.19 µ B/f.u for the film grown at 473K and suddenly falls in the following samples. For the thin film samples the decrease in the value of Ms with increase in the substrate temperature suggests the reduced amount of magnetic moment due to the presence of some 22

ACCEPTED MANUSCRIPT

structural disorder effect or the intrinsic chemical inhomogenity in the film [57]. Umetsu et.al. have also explained the reason of decrease of Ms from the electronic structure [8]. Significant presence of Stoner excitation in the RT and 473K grown samples can also play a responsible role for the suppression in the

RI PT

value of the Ms from their theoretically predicted value [55]. EXAFS results show that in the film there is a presence of atomic disorder between the constituent atoms giving rise to A2 or DO3 kind of structure along with very negligible amount of B2. Yadav et.al have mentioned that the B2 structure does not affect much to

SC

the suppression of the Ms, but the presence of A2 type disorder is very much able to diminish the value of saturation magnetization by decreasing the spin magnetic moment of Fe atoms [27,57]. The highly

M AN U

diminished value of Ms for the film deposited at the highest substrate temperature deposited may also be due to the formation of a silicide layer as observed from the EDAX study. Thus our magnetization study corroborate with the observation made from the structural characterizations. In order to understand the magnetic behavior of the metallic alloy thin films, the temperature

TE D

dependent magnetization M (T ) study has been carried out over the temperature range of 5-350 K in the presence of a constant external magnetic field parallel to the surface of the film as shown in the Fig. 15 (a) – (e). The external field (5T) was selected where the M ( H ) curve at constant temperature saturates. The

EP

M (T ) data has been collected in the Zero Field Cooled (ZFC) and Field Cooled Cooling (FCC) mode. Prior

to the measurement the samples were cooled to 5K from room temperature without application of any

AC C

external magnetic field and later the films were heated upto 350K by the application of a pre-determined magnetic field. In the samples no bifurcation between the ZFC and FCC curves has been observed at the applied magnetic field even at very low temperature (not shown in the figure), suggesting the reversibility of the magnetization due to the orientation of the magnetic dipoles. The M (T ) curves of the films deposited at RT and 473 K show good quality data with the variation of temperature. The films grown at RT and 473K, magnetization varies as T3/2 below 280 K and as T2 beyond 310 K as shown in the inset to the Fig. 15 (a) 23

ACCEPTED MANUSCRIPT

and (b), however magnetization of the other samples deviates from these relations. The magnetization value follows the spin wave theory with the Bloch’s theorem M (T) = M (0) [1-AT3/2] [55], where M (0) be the spontaneous saturation magnetization at 0 K. Linearly fitting the M (T ) vs T3/2data for the films grown at RT

RI PT

and 473K with the above mentioned relation yield the best fit values of M(0) = 63.73 emu/gm (2.70 µ B/f.u) and 54.04 emu/gm (2.35 µ B/f.u) respectively which is closer to the value of saturation magnetization (Ms) mentioned before. The value of the spin stiffness co-efficient (D) was estimated from the spin wave

SC

dispersion law ħω = Dq2, where the value of D can be calculated from the slope of the curve A via relation

A = 2.612V0S-1 × (kB/4πD)3/2, where V0 be the volume per magnetic atom and S be the average spin per atom

M AN U

and kB is the well known Boltzmann constant. The calculated value of the slope A were obtained as 5.458×10-6 K-3/2 and 6.8122×10-6 K-3/2 which gives rise to the value of D as 130 meVÅ2 and 113 meVÅ2 for the films grown at RT and 473K respectively, which are lesser than the values obtained from Co2FeSi alloy [35]. As has been mentioned earlier, beyond 300K the temperature dependence of magnetization changes

TE D

from a T 3 / 2 type to a T 2 type variation (Stoner excitation). The shifting of M (T ) vs. T relationship to a quadratic type with temperature has also been observed by other workers for the case of thin film [59] and single crystals [58] due to a moment ordering phase transition manifested by a hump around 300K in the

EP

M (T ) curve. The typical variation is attributed to the change in the nature of ferromagnetism from localized to itinerant-like ferromagnetism [58]. The value of the Curie temperature ( TC ) has been calculated

AC C

from the empirical relation [ M (T ) ]2 = [ M (0) ]2 [ 1 − TC2 / T 2 ] by linearly extrapolating the curve to the value [ M (T ) ] 2 = 0. The calculated value of the Curie temperature obtained for the sample grown at RT is 1117 K which is in very good agreement with the previously reported results of 1100 K for Co2FeGa based half metallic ferromagnets [8]. However for the CFG film grown at 473K the value decreases to 905 K. The decrease in the Curie temperature may arise due to the disorder between the Co and Fe atoms which modify the electronic structure near the Fermi level. 24

ACCEPTED MANUSCRIPT

Conclusions:

The above work presents detail structural and magnetic characterization of the PLD grown Co2FeGa thin films on c-Si substrates deposited at 5 different substrate temperatures of RT, 473K, 673K, 873K and

RI PT

1073K. XRD measurement with lab source ascertains that the target material used for the deposition of the thin films have only A2 (or DO3) type of disordered cubic phase. However, in case of thin films, PLD grown from the above target, L21 appears to be the dominating phase whose fraction decreases with increase

SC

in substrate temperature, manifesting significant increase of disorder in the samples. Synchrotron based Anomalous X-ray Scattering measurement also supports the above result. However GIXRD measurement

M AN U

shows a contradictory result with the observations made from the above lab source and synchrotron based XRD studies and shows the presence of the A2-like phase similar to the balk target. GIXR data of the samples are best fitted with a bi-layer model with a surface layer of ∼30-50 Å thickness of relatively lower density on a bulk-like thick layer. Such bi-layer structure of the films might give rise to the contradictory

TE D

results observed for the GIXRD and XRD measurements on the samples. EDS results show formation of slightly off-stoichiometric films with deficiency in Ga throughout the series. Synchrotron based XANES study shows that the metallic or zero oxidation states of the atoms are preserved in the films deposited at

EP

lower substrate temperatures, though there is an indication of structural changes in the films deposited at substrate temperatures of 873 and 1073 K. EXAFS measurements confirm L21 structure of the films and a

AC C

possibility of stronger Co-Ga (d-p) hybridization. EXAFS measurements have also been able to establish the presence of A2 and DO3 phases which occur due to Co/Fe/Ga and Co/Fe types of antisite disorders in the samples and could not be clearly detected by XRD measurements. Fitting of FT-EXAFS data also gives quantitative estimation of the disorder parameters which are found to increase gradually with an increase in substrate temperature. EXAFS results show negligible B2-type disorder phase in the samples, which was also confirmed by synchrotron based XRD measurement. Finally magnetization results show soft 25

ACCEPTED MANUSCRIPT

ferromagnetic nature of the films with a saturation value of magnetization less than the theoretically expected value which may be due to the presence of antisite atomic disorder in the films. Along with this it also manifests the existence spin wave nature with lesser stiffness constant in the films deposited at

AC C

EP

TE D

M AN U

SC

RI PT

relatively lower substrate temperatures.

26

ACCEPTED MANUSCRIPT

References:

I. Žuti´c, J. Fabian, S. Das Sarma, Rev. Mod. Phys. 76 (2004) 323 – 410.

2.

A. Hirohata, K. Takanashi, J. Phys. D Appl. Phys. 47 (2014) 193001 - 193040

3.

J.A. Caballero, Y.D. Park, J.R. Childress, J. Bass, W.C. Chiang, A.C. Reilly, W.P. Pratt Jr, F.

RI PT

1.

Petroff, J. Vac. Sci. Technol. A 16 (1998) 1801 – 1805. 4.

R. J. Soulen Jr.,J. M. Byers, M. S. Osofsky, B. Nadgorny, T. Ambrose, S. F. Cheng, P. R.

SC

Broussard,C. T. Tanaka J. Nowak, J. S. Moodera, A. Barry, J. M. D. Coey, Science 282 (1998) 8588.

H. Kato, T. Okuda, Y. Okimoto, Y. Tomioka, K. Oikawa, T. Kamiyama, Y. Tokura, Phys. Rev. B 69 (2004) 184412 - 184419

6.

M AN U

5.

T. Shishidou, A. J. Freeman, R Asahi, Phys. Rev. B 64 (2001) 180401- 180404. 7. M. Belmeguenai, H. Tuzcuoglu, M. S. Gabor,T. Petrisor, Jr., C. Tiusan, F. Zighem, S. M. Cherif, P. Moch, J.Appl.Phys, 115 (2014) 043918 – 043924.

R. Y. Umetsu, A. Okubo, R. Kainuma, J.Appl. Phy. 111 (2012) 073909 - 073915.

9.

P.J.Webster, J.Phys.Chem.Solids 32 (1971) 1221 – 1231.

10.

R. Y. Umetsu, K. Kobayashi, A. Fujita, K. Oikawa, R. Kainuma, K. Ishida, Phys. Rev. B 72 (2005)

EP

TE D

8.

214412 – 214418.

J. Shen, Z. Gai, J. Kirschner, Surface Science Reports 52 (2004) 163 – 218.

12.

W.H. Wang, M. Przybylskia, W. Kuch, L.I. Chelaru, J. Wang, Y.F. Lu, J. Barthel, J. Kirschner,

AC C

11.

Journal of Magnetism and Magnetic Materials 286 (2005) 336 – 339. 13.

C. H. Wang, Y. Z. Guo, F. Casper, B. Balke, G. H. Fecher, C. Felser, Y. Hwu, Appl. Phys. Lett. 97 (2010) 103106 - 103109.

14.

S. Picozzi, A. Continenza, A. J. Freeman, Phys. Rev. B 66 (2002) 094421- 094429. 27

ACCEPTED MANUSCRIPT

15.

I.Galanakis, Ph. Mavropoulos, P.H.Dederichs, J.Phys.D:Appl.Phys, 39 (2006) 765 – 775.

16.

Y. Takamura, R. Nakane, S. Sugahara, J. Appl. Phys. 105 (2009) 07B109 – 07B111.

17.

B.S.D.Ch.S. Varaprasad, A. Srinivasan, Y.K. Takahashi, M. Hayashi, A. Rajanikanth, K. Hono,

18.

RI PT

Acta Materialia 60 (2012) 6257–6265. S. Wurmehl, M. C Martins Alves, J. Morais, V. Ksenofontov, S. R Teixeira, G. Machado, G. H Fecher, C. Felser, J. Phys. D: Appl. Phys. 40 (2007) 1524–1533.

P. J. Brown, K. U. Neumann, P. J. Webster, K. R. A. Ziebeck, Journal of Condensed Matter, 12

SC

19.

(2000) 1827 – 1835.

C. Grigorescu,, E. Valerio, L. Tortet, R. Notonier, H.J. Trodahl, V.J. Kennedy, A. Markwitz, S.A.

M AN U

20.

Manea, M. Autric, Journal of Crystal Growth 275 (2005) 1183 – 1188. 21.

U.Geiersbach, A.Bergmann, K.Westerholt, Journal of Magnetism and Magnetic Materials, 240 (2002) 546 – 549.

H. Pandey, R.C.Budhani, J.Appl.Phys.113 (2013) 203918

23.

H. Schneider, Ch Herbort, G. Jakob, H Adrian, S. Wurmehl, C Felser, J. Phys. D: Appl. Phys. 40 (2007) 1548 – 1551.

TE D

22.

A. Yadav, S. Chaudhary, J.Appl. Phys. 115 (2014) 133916 – 133922.

25.

S. Husain, S. Akansel, A. Kumar, P. Svedlindh, S. Chaudhary, Sci. Rep. 6 (2016) 28692 - 286102

26.

C.E.A. Grigorescu , E. Valerio , O. Monnereau , L. Tortet , L. Argeme, G. Pavelescu, S.A. Manea,

AC C

EP

24.

C. Ducu, A. Malinovski, H.J. Trodahl, A. Bittar, N. Strickland, R. Notonier, W.R. Branford, M. Autric, Applied Surface Science 253 (2007) 8102 – 8106. 27.

W. Wang, W. Yang, Z. Liu, Y. Lin, S. Zhou, H. Qian, F. Gao, G. Li, Cryst Eng Comm 16 (2014) 4100.

28

ACCEPTED MANUSCRIPT

28.

D. Pergolesi, E. Fabbri, A. D’Epifanio, E. Di Bartolomeo, A. Tebano, S. Sanna, S. Licoccia, G. Balestrino, E. Traversa, Nat. Mater. 9 (2010) 846-852 G. Li, W. Wang, W. Yang, H. Wang, Surface Science Reports 70 (2015) 380–423.

30.

A.Yadav, S.Chaudhary; J.Appl.Phys.117 (2015) 083911 – 083919.

31.

A.K. Poswal, A. Agrawal, A.K. Yadav, C. Nayak, S. Basu, S.R. Kane, C.K. Garg, D. Bhattachryya, S.N. Jha, N.K. Sahoo, AIP Conf. Proc. 1591 (2014) 649 – 651.

S. Basu, C. Nayak, A.K. Yadav, A. Agrawal, A.K. Poswal, D. Bhattacharyya, S.N. Jha, N.K. Sahoo, J. Phys. Conf. Ser. 493 (2014) 012032 – 012035.

SC

32.

RI PT

29.

J. Rodrguez Carvajal, Physica B: Condensed Matter 192 (1993) 55 – 69.

34.

M. Yin, S. Chen, P. Nash, Journal of Alloys and Compounds 577 (2013) 49 – 56.

35.

A.Yadav, S.Chaudhary; J.Appl.Phys.118 (2015) 193902 – 193910.

36.

S. Isber, N. Skaff, M. Roumie, M. Tabbal, Web of conf. 75 (2014) 03008 – 03010.

37.

Y. Kajikawa, J. Cryst. Growth 289 (2006) 387 – 394.

38.

P.J Chen, G. Feng, R.D. Shull, IEEE Transactions on Magnetics, 49 (2013) 4379 – 4382.

39.

A.J.Bradley, J.W.Rodgers, Proc.R.Sco.Londo,Ser.A, 144 (1934) 340

40.

G. Jakob, F. Caspera, V. Beaumonta, S. Falk, N. Auth, H.-J. Elmers, C. Felser, H. Adrian, Journal of

EP

TE D

M AN U

33.

Magnetism and Magnetic Materials 290 (2005) 1104 – 1107. J. S. Horwitvitz, K. S. Grabowski, D. B. Chrisey, R. E. Leuchtnera, Appl. Phys. Lett. 59 (1991)

AC C

41.

1565. 42.

E. van de Ret, J. C. S. Kook, J. Dieleman, J. Appl. Phys. 73 (1993) 8290 – 8296.

43.

S.Yamada, K.Tanikawa, S.Oki, M.Kawano, M.Miyao, K.Hamaya, Appl.Phys.Lett., 105 (2014), 071601 – 071604

44.

Y. Takamura, R. Nakane, S. Sugahara, J.Appl Phys. 107 (2010) 09B111- 09B113. 29

ACCEPTED MANUSCRIPT

45.

B. Balke, S. Wurmehl, G.H. Fecher, C. Fesler, Appl. Phys. Lett 90 (2007) 172501 – 172503.\

46.

M. Newville, B. Ravel, D. Haskel, J.J. Rehr, E.A. Stern, Y. Yacoby, Physica B, 208 (1995) 154 – 156. X-Ray Absorption: Principles, Applications, Techniques of EXAFS, SEXAFS and XANES, edited by D.C. Konigsberger and R. Prince, 1988 Wiley, New York.

48.

RI PT

47.

S.D. Kelly, D. Hesterberg, B. Ravel, Analysis of soils and minerals using X-ray absorption

SC

spectroscopy, in Methods of Soil Analysis, Part 5–Mineralogical methods; Ulery A.L, Drees L.R, Eds.; Soil Science Society of America: Madison, WI, USA, (2008) 367 – 464. S.I. Zabinsky, J.J. Rehr, A. Ankudinov, R.C. Albers, M.J. Eller, Phys. Rev. B 52 (1995) 2995 –

M AN U

49.

3009. 50.

C. Wang, F. Casper, Y. Guo, T. Gasi, V. Ksenofontov, B. Balke, G. H. Fecher, C. Felser , Y.K. Hwu, J.J Lee, J.Appl. Phys. 112 (2012) 124314 - 124321.

S.Wurmehl, G. H. Fecher, H. C. Kandpal, V. Ksenofontov, C. Felser, H. Ji Lin, J. Morais, Phys.

TE D

51.

Rev. B 72 (2005) 184434 – 184442.

B.Ravel, M.P.Raphael, V.G.Harris, Phy.Rev.B, 65 (2002) 184431 – 184431.

53.

P.A.Bhobe, J.Phys.D.Appl.Phys, 41 (2008) 045004 – 045009.

54.

V.Niculescu, K,Raj, T.J.Burch, J.I.Budnick, Phy. Rev. B 13 (1976) 3167- 3174.

55.

Anupam, P.C.Joshi, P.K.Rout, Z. Hossain, R.C. Budhani, J.Phys.D:Appl.Phys, 43 (2010) 255002 –

AC C

EP

52.

255007. 56.

I. Galanakis, P. H. Dederichs, Phy.Rev.B 66 (2002) 174429 – 174437.

57.

D.Serrate, J.M.De.Teresa, R.Cordoba, S.M.Yusuf, Solid State Communication 142 (2007) 363 – 367.

30

ACCEPTED MANUSCRIPT

58.

L. Ritchie and G. Xiao, Y. Ji, T. Y. Chen, C. L. Chien, M. Zhang, J. Chen, Z. Liu, G. Wu, X. X. Zhang, Phys. Rev. B 68 (2003) 104430 – 104435.

EP

TE D

M AN U

SC

RI PT

D. Ristoiu, J.P.Nozieres, L. Ranno, J.Magn.Magn.Mater, 219 (2000) 97 – 103.

AC C

59.

31

ACCEPTED MANUSCRIPT

Caption of Figures: Fig.1. Room temperature XRD spectra of the (a) Co2FeGa thin films with varying substrate temperature, (b)

RI PT

Rietveld refinement of the target. (Inset shows the variation in the crystallite size with substrate temperature.

Fig.2. Synchrotron based XRD spectra of the Co2FeGa thin film grown at Room Temperature with different

SC

X-ray energy.

M AN U

Fig.3. Room temperature GIXRD spectra of the Co2FeGa thin films with varying substrate temperature. Fig.4. (a) Experimental data fitted with theoretical model, variation in (b) thickness of upper layer, (c) density and (d) surface roughness of both layers.

Fig.5. FESEM back scattering electron image of thin film samples grown at different substrate temperature.

TE D

Fig.6. (a) - (b) The EDXS spectra of the RT and 1073K deposited films, (c) Variation in the elemental stoichiometry of the film with varying the substrate temperature.

EP

Fig.7. Normalized XANES spectra (µ(E) vs E) of the CFG thin films at (a) Co, (b) Fe and (c) Ga K-edges.

AC C

Fig.8. k2 weighted χ(k) of the CFG thin film samples at (a) Co, (b) Fe and (c) Ga K-edges. Fig.9. Schematic diagram of the atomic configuration surrounding the Co, Fe and Ga atoms in the L21 structure. (The cubic shell is 1/8th of the unit structure) (Fe/Ga atoms are surrounded by 6 Ga/Fe atoms in an octahedron configuration; the arrows show the scattering paths of the 1st co-ordination shell of all the atoms.) Fig.10. Experimental χ(R) vs R data fitted with theoretically generated model (without antisite disorder) of the CFG thin film samples at (a) Co, (b) Fe and (c) Ga K-edges with varying the substrate temperature. 32

ACCEPTED MANUSCRIPT

(Position and contribution of the individual scattering paths have been shown with RT deposited sample). Fig.11. Experimental χ(R) vs R data along with Real and Imaginary part of the CFG thin film grown at TS =

RI PT

RT fitted with theoretically generated model (with antisite disorder) at (a) Co, (b) Fe and (c) Ga Kedges.

SC

Fig.12.Variation in the parallel magnetization (M(H) vs H) curves of the samples at three temperatures with varying the substrate temperature.

M AN U

Fig.13. Variation in the hysteresis (M(H) vs H) curves of parallel magnetization of different samples at very low field.

Fig.14. Summary of the value of the magnetization parameters as (a) saturation magnetization, (b) remnant magnetization, (c) coercive field and (d) sureness of the hysteresis curves of the samples with varying

TE D

the substrate temperature.

Fig.15. Thermo magnetization (M vs T) curves of the samples with varying the substrate temperatures. (In

AC C

EP

figure 15 (a) and (b) the spin wave and stoner excitation curves have been drawn).

33

ACCEPTED MANUSCRIPT

Table 1: Values of structural parameters obtained from EXAFS analysis at Co K-edge considering ordered

L21 phase TS = RT

TS = 473 K

TS = 673 K

TS = 873 K

TS = 1073 K

R (Å)

2.46 ± 0.02 0.019 ± 0.002

2.45 ± 0.02 0.027 ± 0.004

2.47 ± 0.01 0.018 ± 0.001

2.55 ± 0.01 0.003 ± 0.001

2.53 ± 0.02 0.002 ± 0.002

2.40 ± 0.01 0.005 ± 0.001 2.82 ± 0.02 0.014 ± 0.002 4.00 ± 0.03 0.021 ± 0.004

2.41 ± 0.01 0.006 ± 0.001 2.77 ± 0.03 0.029 ± 0.007 3.91 ± 0.02 0.021 ± 0.003

2.47 ± 0.01 0.007 ± 0.001 2.76 ± 0.01 0.015 ± 0.001 4.03 ± 0.01 0.020 ± 0.002

0.006

0.007

σ2 (Å2) Co-Ga× 4 Co-Co1× 6 Co-Co2× 12

R (Å) σ2 (Å2) R (Å) σ2 (Å2) R (Å) σ2 (Å2) Rfactor

RI PT

Parameters

2.38 ± 0.01 0.003 ± 0.001 2.72 ± 0.04 0.006 ± 0.005 3.90 ± 0.01 0.018 ± 0.002

SC

Scattering paths × C.N Co-Fe × 4

0.008

0.028

2.36 ± 0.01 0.004 ± 0.001 2.67 ± 0.04 0.009 ± 0.010 3.71 ± 0.05 0.028 ± 0.020 0.020

L21 phase Parameters

TS = RT

R (Å)

2.46 ± 0.01 0.014 ± 0.001 2.79 ± 0.01 0.012 ± 0.001 4.02 ± 0.02 0.019 ± 0.004 0.015

σ2 (Å2) Fe-Ga× 6

R (Å)

σ2 (Å2) Fe-Fe× 12

R (Å) σ2 (Å2) Rfactor

TS = 473 K

TS = 673 K

TS = 873 K

TS = 1073 K

2.46 ± 0.01 0.012 ± 0.001 2.76 ± 0.01 0.013 ± 0.001 4.00 ± 0.02 0.019 ± 0.004 0.009

2.46 ± 0.01 0.016 ± 0.001 2.76 ± 0.02 0.010 ± 0.002 4.00 ± 0.03 0.016 ± 0.004 0.029

2.38 ± 0.01 0.010 ± 0.001 2.60 ± 0.01 0.007 ± 0.001 4.18 ± 0.04 0.009 ± 0.004 0.040

2.41 ± 0.02 0.008 ± 0.002 2.58 ± 0.02 0.005 ± 0.002 4.18 ± 0.05 0.015 ± 0.010 0.020

TE D

Scattering paths × C.N Fe-Co × 8

M AN U

Table 2: Values of structural parameters obtained from EXAFS analysis at Fe K-edge considering ordered

L21 phase Parameters

TS = RT

TS = 473 K

TS = 673 K

TS = 873 K

TS = 1073 K

2.45 ± 0.01 0.007 ± 0.001

2.47 ± 0.01 0.008 ± 0.001

2.48 ± 0.01 0.009 ± 0.001

2.44 ± 0.01 0.005 ± 0.002

2.47 ± 0.01 0.011 ± 0.002

σ2 (Å2)

2.83 ± 0.02 0.015 ± 0.003

2.81 ± 0.04 0.028 ± 0.009

2.74 ± 0.03 0.015 ± 0.003

2.82 ± 0.03 0.007 ±0.004

2.91 ±0.04 0.022 ±0.007

R (Å) σ2 (Å2)

3.97 ± 0.02 0.013 ± 0.002

3.98 ± 0.03 0.013 ± 0.003

4.04 ± 0.02 0.010 ± 0.002

4.01 ± 0.02 0.010 ± 0.002

4.14 ± 0.05 0.019 ± 0.013

Rfactor

0.003

0.011

0.022

0.003

0.053

AC C

Scattering paths × C.N Ga-Co × 8

R (Å)

σ2 (Å2)

Ga-Fe× 6 Ga-Ga× 12

EP

Table 3: Values of structural parameters obtained from EXAFS analysis at Ga K-edge considering ordered

R (Å)

34

ACCEPTED MANUSCRIPT

Table 4: Definition of the atomic disorder parameters in the constituent scattering paths considering proper L21 (Co2Fe1Ga1) structure + disorder structures

(Ordered Co site) C.N/2×

(Ordered Fesite) C.N×

(Ordered Ga site) C.N×

Disordered Co site) C.N×

2-β-γ

1-γ

1-β

…

Fe (1)

1-γ

1-α-γ

1-α

γ

Ga (1)

1-β

1-α

1-α-β

β

Table 5: Variation in the values of the disorder parameters Atomic disorder

TS = RT

α

Fe & Ga (B2)

0.004 ± 0.010

β

Co & Ga (A2’)

γ

Co & Fe (DO3)

TS = 473 K

β

…

α

α

…

TS = 673 K

TS = 873 K

0.011 ± 0.023

0.029 ± 0.021

0.021 ± 0.072

0.125 ± 0.112

0.144 ± 0.188

0.133 ± 0.142

0.149 ± 0.085

0.053 ± 0.301

0.111 ± 0.351

0.284 ± 0.260

0.484 ± 0.052

M AN U

Parameters

Disordered Ga site) C.N×

γ

SC

Co (2)

Disordered Fe site) C.N×

RI PT

K-edge absorption edge

disordered L21 phase Scattering paths × C.N Co-Fe × 4

Parameters

Co-Co’(Fe)× 4

TE D

Table 6: Values of structural parameters obtained from EXAFS analysis at Co K-edge considering

TS = RT

TS = 473 K

TS = 673 K

TS = 873 K

3.92 2.40 ± 0.02

3.67 2.52 ± 0.01

2.93 2.40 ± 0.01

2.15 2.49 ± 0.01

E.C.N R’ (Å) σ2 (Å2) E.C.N R (Å)

0.08 2.40 ± 0.02 0.005 ± 0.003 3.59 2.40 ± 0.01

0.33 2.53 ± 0.02 0.014 ± 0.002 3.54 2.40 ± 0.01

1.07 2.41 ± 0.03 0.022 ± 0.002 3.44 2.43 ± 0.01

1.85 2.49 ± 0.01 0.002 ± 0.001 3.48 2.31 ± 0.01

Co-Co’(Ga)× 4

E.C.N R’ (Å) σ2 (Å2)

0.41 2.42 ± 0.02 0.024 ± 0.001

0.46 2.41 ± 0.05 0.008 ± 0.001

0.56 2.43 ± 0.05 0.007 ± 0.001

0.52 2.32 ± 0.05 0.004 ± 0.001

Co-Co× 6

E.C.N R’ (Å) σ2 (Å2)

5.63 2.80 ± 0.01 0.017 ± 0.002

5.41 2.82 ± 0.04 0.022 ± 0.004

4.77 2.68 ± 0.01 0.011 ± 0.001

4.22 2.67 ± 0.01 0.003 ± 0.001

Rfactor

0.002

0.007

0.016

0.003

AC C

EP

E.C.N R (Å)

Co-Ga × 4

35

ACCEPTED MANUSCRIPT

Table 7: Values of structural parameters obtained from EXAFS analysis at Fe K-edge considering

Parameters

TS = RT

TS = 473 K

TS = 673 K

TS = 873 K

E.C.N R (Å)

7.79 2.46 ± 0.01

7.33 2.46 ± 0.01

5.85 2.47 ± 0.01

4.28 2.43 ± 0.03

Fe-Fe’(Co) × 8

E.C.N R’ (Å) σ2 (Å2)

0.21 2.46 ± 0.02 0.016 ± 0.001

0.67 2.42 ± 0.04 0.012 ± 0.001

2.15 2.44 ± 0.02 0.016 ± 0.001

3.72 2.44 ± 0.01 0.012 ± 0.002

Fe-Ga× 6

E.C.N R (Å)

6.00 2.79 ± 0.01

5.94 2.73 ± 0.02

5.83 2.76 ± 0.01

5.87 2.65 ± 0.02

Fe-Fe’(Ga) × 8

E.C.N R’ (Å) σ2 (Å2)

0.00 2.76 ± 0.02 0.012 ± 0.001

0.06 2.74 ± 0.05 0.014 ± 0.001

0.17 2.77 ± 0.05 0.011 ± 0.001

0.12 2.67 ± 0.05 0.007 ± 0.002

Fe-Fe× 12

E.C.N R’ (Å) σ2 (Å2)

11.68 4.01 ± 0.03 0.017 ± 0.003

10.87 3.98 ± 0.02 0.018 ± 0.003

8.44 3.98 ± 0.02 0.013 ± 0.002

6.17 3.86 ± 0.04 0.004 ± 0.002

Rfactor

0.012

0.019

0.030

M AN U

RI PT

Scattering paths × C.N Fe-Co × 8

SC

disordered L21 phase

0.006

disordered L21 phase Scattering paths × C.N (c)Ga-Co × 8

Parameters

(a)Ga-Ga’(Co) × 8

E.C.N R’ (Å) σ2 (Å2 )

(d)Ga-Fe× 8

TE D

Table 8: Values of structural parameters obtained from EXAFS analysis at Ga K-edge considering

TS = RT

TS = 473 K

TS = 673 K

TS = 873 K

7.16 2.45 ± 0.01

7.11 2.47 ± 0.01

6.98 2.48 ± 0.01

7.01 2.43 ± 0.01

0.84 2.41 ± 0.03 0.008 ± 0.001

0.88 2.41 ± 0.04 0.010 0.001

1.01 2.44 ± 0.04 0.009 ± 0.001

0.99 2.43 ± 0.01 0.007 ± 0.001

E.C.N R (Å)

6.00 2.82 ± 0.01

5.94 2.84 ± 0.05

5.82 2.74 ± 0.02

5.87 2.81 ± 0.02

(b)Ga-Ga’(Fe) × 8

E.C.N R’ (Å) σ2 (Å2 )

0.00 2.82 ± 0.01 0.015 ± 0.003

0.06 2.85 ± 0.06 0.029 ± 0.008

0.18 2.95 ± 0.05 0.014 ± 0.004

0.13 2.83 ± 0.01 0.007 ± 0.002

(e)Ga-Ga× 12

E.C.N R’ (Å) σ2 (Å2 ) Rfactor

10.74 3.96 ± 0.02 0.013 ± 0.002 0.002

10.54 4.00 ± 0.02 0.012 ± 0.022 0.007

10.13 4.04 ± 0.01 0.008 ± 0.002 0.015

10.25 4.00 ± 0.02 0.008 ± 0.002 0.002

AC C

EP

E.C.N R (Å)

* ECN = Effective Co-ordination Number = C.N × disorder factor 36

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT


Co2FeGa Heusler alloy films prepared by PLD at 5 substrate temperatures upto 1073K.
XRD shows L21 major phase which decreases with increase in substrate temperature.
EXAFS results confirm L21 structure and higher Co-Ga (d-p) hybridization.

RI PT


EXAFS measurements could estimate relative fraction of A2 and DO3 phases.

AC C

EP

TE D

M AN U

SC


Magnetization results show soft ferromagnetic nature due to antisite disorder.