Evolution of Fe environments and phase composition during mechanical amorphization of Fe70Zr30 and Fe70Nb30 alloys

Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 1 Evolution of Fe environments and phase composition during mechanical amorphization of Fe70Zr30 and Fe70Nb30 alloys A. F. Manchón-Gordón, J. J. Ipus, J. S. Blázquez*, C. F. Conde, A. Conde Dpto. Física de la Materia Condensada, ICMSE-CSIC, Universidad de Sevilla, P.O. Box 1065, 41080 Sevilla, Spain * The corresponding author e-mail: [email protected] ABSTRACT: Amorphous Fe-Zr and Fe-Nb alloys with 70 at. % Fe have been prepared by ball milling a mixture of elemental powders. The combination of Mössbauer spectroscopy, X-ray diffraction, scanning electron microscopy and differential scanning calorimetry techniques supplies detailed information about the composition as well as microstructural parameters of remaining crystalline and amorphous phases developed during milling. Detailed analysis of the Fe environments allows us to distinguish Fe atoms in paramagnetic sites between those incorporated to crystalline Zr or Nb rich phases and those Fe atoms in amorphous phase. KEYWORDS: mechanical amorphization, mechanical alloying, amorphous alloys 1. INTRODUCTION Advanced materials can be defined as those that represent superior properties over the traditional materials for demanding applications. Typically, progress on improving the properties of these materials is given thanks to the systematic synthesis and control of the structure of the materials [1] processing them under non-equilibrium conditions [2]. Among the different processes in commercial use, mechanical alloying (MA) [2], plasma processing [3] or rapid solidification from the liquid state [4] can be signed. MA technique, originally developed to produce oxide-dispersion strengthened nickel and iron base superalloys for applications in the aerospace industry [5], has a unique Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 2 capability of producing thermodynamically disfavored reactions and destabilizing equilibrium systems [2]. In fact, the large diversity of metastable systems that can be produced by MA, including amorphous alloys, supersaturated solid solutions or metastable intermetallics, makes this technique to be present in many research fields [2, 6]. Generally, high energetic milling of alloys leads to the formation of a nanocrystalline structure and, for some compositions, to amorphization [2]. Iron-based systems are among the most suitable materials for studying the mechanical milling as a process and the properties of the final products. The cheap character of this element makes it desirable and ubiquitous in any technological field. The magnetic nature of many compounds of iron significantly increases the amount of information that can be obtained for these systems. Moreover, the existence of a Mössbauer isotope, 57Fe, in a non-negligible natural abundancy [7] makes possible the study of hyperfine interactions of Fe-compounds. Therefore, all this leads to a more comprehensive understanding of the material. Particularly, the Fe-Zr alloys present extremely different magnetic properties depending on Fe content [8, 9] and the local atomic order [10], as well as a wide range of the existence of the amorphous phase, FexZr100-x (20≤x≤93 at. %) [11]. In the case of the Fe-Nb alloys, although still some research is needed to determine a complete phase diagram [12], amorphous phase is found in a wide compositional range [13, 14]. These phases show interesting peculiarities due to singular aspects of their nanostructures [15]. In a previous study [16] we used Fe70Zr30 and Fe70Nb30 (at. %) compositions to show the validity of the cubic law [17] that correlates the progress of milling with the frequency of the mill. Microstructural and compositional characterization of amorphous structures is an essential point to optimize technological properties in which they are based on. However, this characterization is critical for nanoscale systems, for which the Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 3 characteristic size to be studied is so tiny that could presumably only be attainable to nanoprobe techniques. Otherwise, due to the extremely local character of atom-scale techniques, statistics using tiny probes are not straightforward. For example, in the analysis of a nanocrystal using atom probe, the sensitivity in the composition is not as good as for systems with a well-defined flat interface. Assuming a 5 nm sphere, the range not affected by the surface geometry is limited to a cube of 5/√3 nm in size. In the case of a bcc-Fe phase, this implies a length of only ~10 unit cells. On the other hand, methods based on global techniques (X-ray diffraction, Mössbauer spectroscopy, etc.) present the advantage of a much better statistics to obtain average values and are experimental techniques accessible to most of the scientific community. In fact, the capability of Mössbauer spectroscopy to analyze nanoscale systems have been pointed out, particularly when combining with other techniques, such as X-ray diffraction [18, 19]. Currently, there are many investigations on amorphous Fe-Zr and Fe-Nb phases obtained by different techniques and on the structural analysis of the products of mechanical alloying in these systems [10, 14, 20-24]. However, no systematic study of the composition of both amorphous and remnant bcc nanocrystals in Fe-based supersaturated solid solutions have been undertaken. In our case, we combined XRD and Mössbauer data to estimate the compositions of nanocrystals in FeNbB supersaturated solid solution [19]. It is worth mentioning that the knowledge of the composition of the different phases is very helpful to propose accurate models to describe the behavior of any physical system. In the present work, amorphous Fe-Zr and Fe-Nb alloys with 70 at. % Fe have been prepared using mechanical alloying. The microstructural evolution of both alloys was followed by Mössbauer spectroscopy, X- ray diffraction, scanning electron microscopy and differential scanning calorimetry. The Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 4 combination of these techniques supplies very detailed information about the composition of nanocrystals and the amorphous phase developed during mechanical alloying. Moreover, from the evolution of the compositions of nanocrystals core it has been possible to estimate the interface thickness. 2. EXPERIMENTAL Fe-Zr and Fe-Nb alloys with 70 at. % Fe were prepared by ball milling of elemental powders mixture in a planetary mill Fritsch Pulverisette Vario 4 using hardened steel balls and steel vials. The initial powder mass was 30 g, the ball to powder ratio 10:1 and the ratio between the rotational speeds of the vials and the main disk -2. Continuous milling steps of 0.5 h were performed in Ar atmosphere and after selected times, approximately 0.2 g of sample was extracted in a Saffron Omega glove box with oxygen and humidity levels below 2 ppm under Ar atmosphere. The microstructure was studied by X-ray diffraction (XRD) using Cu-Kα radiation in a Bruker D8 I diffractometer. Differential scanning calorimetry (DSC) was performed in a TA-Instruments Q600 SDT under nitrogen flow. The powder samples were pressed in silver capsules and a baseline was obtained using a standard Ag sample. Particle size distribution and morphology was studied by scanning electron microscopy (SEM) in secondary electrons mode performed in a Jeol 6460 LV microscope operated at 30 kV. The local environment of Fe atoms was analyzed by Mössbauer spectrometry (MS) in transmission geometry at room temperature (RT) using a 57Co(Rh) source. Values of the hyperfine parameters were obtained by fitting with NORMOS program [25] and the isomer shift (𝐼𝑆) was quoted relative to the Mössbauer spectrum of an α-Fe foil at RT. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 5 3. RESULTS and DISCUSSION 3.1. X-ray diffraction results XRD patterns of Fe70Zr30 and Fe70Nb30 alloys after selected milling times are presented in figure 1. The amorphous phase of the as-milled samples is evidenced by a wide amorphous halo in samples milled for at least 𝑡 =6 and 9 h, for Fe70Zr30 and Fe70Nb30, respectively. XRD does not clearly show the expected displacement to lower angles of bcc Fe diffraction maxima that should indicate an increase of the lattice parameter, 𝑎, as Zr or Nb atoms become integrated into the bcc Fe(Zr) and Fe(Nb) solid solution, respectively, although a slight effect can be observed at short times. The bcc Fe maxima get broader as milling time increases due to the reduction of the crystal size and the increase of microstrains during milling. Peaks attributed to the hcp Zr and bcc Nb phases are only detected for short milling times (𝑡 <6 and 𝑡 <12 h for Fe70Zr30 and Fe70Nb30, respectively). The evolution of the amorphous phase with milling time was estimated from a deconvolution of the main X-ray peak to separate the (110) line of the 𝛼-Fe nanocrystals (narrow Lorentzian contribution) and the amorphous halo (broad Gaussian contribution). Therefore, the amorphous volume fraction can be estimated as the area of the Gaussian contribution (amorphous halo) divided by the whole area of the experimental diffraction peak. In the case of the Fe70Nb30 alloy, for short milling times (𝑡 <12 h), it is necessary to take into account the contribution of the (110) peak of the bcc Nb nanocrystals. Figure 2 shows the evolution of the crystal size, 𝐷, estimated from Scherrer’s formula from the full width at half maximum of crystalline (110) diffraction peak, as a function of milling time for both alloys along with that of Nb phase for Fe70Nb30 alloy. Final 𝐷 Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 6 of bcc-Fe type phase is clearly larger for Fe70Zr30 alloy, for which a stable value ~10 nm is rapidly achieved, meanwhile 𝐷 continuously decreases down to ~5 nm for the Nb-containing alloy. The evolution of the lattice parameter of the 𝛼-Fe(X), being X=Zr, Nb, as a function of milling time is shown in figure 3. As commented above, the expected increase in lattice parameter is clearer for Fe70Nb30 than for Fe70Zr30 alloy. The position of the amorphous halo shifts to higher 2𝜃 values as milling progresses (figure 4a). This indicates a contraction of the average metal-metal distance (figure 4b), which can be estimated as [26]: 𝛿𝑀−𝑀 =5𝜆 8sin𝜃 (1) where 𝜆 is the wavelength of the used radiation. 3.2. Differential scanning calorimetry results In order to verify the formation of the amorphous structure, DSC curves of both studied alloys were recorded in non-isothermal regime. Figure 5 shows the DSC scans at 20 K/min of both studied samples after ~50 h of milling. All the features found in the curves are irreversible as they vanish in a second heating. The well-defined exothermic peaks around 950 and 1000 K for Zr and Nb containing alloys, respectively, correspond to crystallization of the amorphous phase. Previous deviations from baseline are ascribed to relaxation phenomena in agreement with the literature [9, 17]. The obtained values of the onset temperatures 𝑇𝑜𝑛𝑠𝑒𝑡 and peak temperatures 𝑇𝑝𝑒𝑎𝑘 of crystallization process are shown in Table 1. The values obtained for the Fe70Zr30 alloy are similar to those reported by Mishra et al. [9]. However, 𝑇𝑜𝑛𝑠𝑒𝑡 of powder samples are slightly higher than those obtained for ribbon samples [27]. In the case of the Nb containing alloy, the values obtained are close to those reported by El-Eskandarany et al. [22] for amorphous Fe52Nb48 alloy powder. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 7 3.3.Scanning electron microscopy results Microstructural modifications induced by milling can be followed from a direct comparison of the secondary electron images (figure 6) for the two studied alloys. These show the existence of agglomerates at early milling times (𝑡 =6 and 8.5 h for Fe70Zr30 and Fe70Nb30, respectively), meanwhile at longer milling times (𝑡 >22 h) almost spherical isolated powder particles are found. Although those agglomerates are present in both samples, for Fe70Nb30 the size of particles is larger and the number of particles per agglomerate is smaller than for Fe70Zr30 samples for equivalent milling times (e.g. after ~20 h milling, Fe70Nb30 alloy shows few tens of particles per agglomerate, meanwhile Fe70Zr30 shows few hundreds of particles per agglomerate). Insets of figure 6 show the corresponding grain size distribution histograms. The particle size after different milling times was obtained by averaging over a minimum of 50 particles for each sample. Particle size decreases with milling until 22 h, achieving sizes about d~12±4 and 33±13 𝜇m for Fe70Zr30 and Fe70Nb30, respectively. After this time of milling, the particle size increases, finally reaching sizes about d~23±11 and 44±18 𝜇m for Fe70Zr30 and Fe70Nb30, respectively. For agglomerates there is not a clear tendency, being their sizes between 50-200 𝜇m for both alloys when they are present. 3.4. Mössbauer spectroscopy results Figure 7 shows the evolution of the room temperature Mössbauer spectra with milling time. Two ferromagnetic sites were used to fit the residual α-Fe(X) phase: a sextet with hyperfine field, 𝐻𝐹=33 T (no X atoms in the neighborhood), and another with 𝐻𝐹=30.5 T (one X atom in the neighborhood). A quadrupolar distribution was used to describe the paramagnetic contributions, a priori assigned to the amorphous phase. Moreover, a distribution of hyperfine fields from 5 to 30 T was used to show the contribution from Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 8 the interface region and richer X neighborhoods. Figure 8 shows the fraction of the Fe atoms contributing to each environment type. The fraction of the paramagnetic contributions shows a continuous increase as milling time progresses. Figure 9 shows the average isomer shift (<𝐼𝑆>) of the paramagnetic contributions as a function of milling time for both studied alloys, either Zr or Nb play a similar role and values of this parameter show a common behavior for the two alloys. Initially, for short times of milling, for which hcp Zr and bcc Nb phases are detected by XRD, a reduction of <𝐼𝑆> from 0 to ~−0.18 mm/s is observed. Values show a common behavior for both alloys after the peaks attributed to the hcp Zr or bcc Nb phase in XRD patterns disappear. After further milling, <𝐼𝑆> reaches a stable value ~−0.15 mm/s for both samples. The values obtained for the Fe70Zr30 alloy agree with those reported for the amorphous alloys Fe65Zr35 (𝛿 = −0.16 mm/s) [20] and Fe70Zr30 (𝛿 = −0.16 mm/s) [16], but slightly higher values were reported for Fe73Zr27 (𝛿 = −0.13 mm/s) [21] and Fe2Zr (𝛿 = −0.13 mm/s) [10]. The values obtained for the Fe70Nb30 alloy also agree with those reported for the amorphous alloy with the same composition (𝛿 = −0.16 mm/s) [16] and Fe80Nb20 (𝛿 = −0.17 mm/s) [14]. However, Eskandarany et al. observed an isomer shift 𝛿 = −0.19 mm/s in amorphous samples of Fe52Nb48 prepared by mechanical alloying [22]. In fact, the slightly increase in <𝐼𝑆> from the observed minimum could be ascribed to the Fe enrichment of the amorphous phase as milling progresses. The quadrupole splitting distributions, 𝑃(𝑄𝑆), are shown in figure 10. It shows the same behavior for both alloys, with a non-zero probability for 𝑄𝑆=0 mm/s. They also present two well resolved maxima for short times of milling (𝑡 <6 and 𝑡 <10 h for the Fe70Zr30 and Fe70Nb30 alloy, respectively) that become only one with the increase of milling Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 9 time. The first peak (𝑄𝑆~0.25 mm/s) may be ascribed to Fe atoms in environments similar to Fe2X. This environment, for X=Zr has 𝑄𝑆=0.35 mm/s (𝛿 = −0.13 mm/s) [10] and for X=Nb has 𝑄𝑆=0.34 mm/s (𝛿 = −0.19 mm/s) [24]. The second maximum found at low milling times (𝑄𝑆~1.25 mm/s) may be attributed to X richer environments, e.g. Vincze et al. [28] reported 𝑄𝑆=0.91 mm/s for FeZr3. With these data, we can admit that 𝑄𝑆 distributions for low milling times do not really correspond to amorphous phase, but to Fe present in Zr and Nb remaining crystals (the times at which this two-peak 𝑃(𝑄𝑆) are observed are in agreement with the observation of bcc Nb and hcp Zr phases by XRD). For larger milling times, a value of <𝑄𝑆>=0.53±0.07 mm/s is obtained for both alloys. This value is close to the values previously reported for Fe70Zr30 (<𝑄𝑆>=0.44 mm/s) and for Fe70Nb30 (<𝑄𝑆>=0.42 mm/s) [16]. Taking into account the nanocrystalline microstructure of our samples, some words on superparamagnetism are needed. An ensemble of magnetic nanoparticles shows superparamagnetic behavior above a certain temperature: the blocking temperature 𝑇𝐵. This temperature is defined as the temperature at which the relaxation time 𝜏 equals the experiment timescale 𝜏𝑚 [29]: 𝑇𝐵=∆𝐸 𝑘𝐵𝑙𝑛( 𝜏 𝑚 𝜏0) (2) where ∆𝐸 is the energy barrier that the magnetization flip has to overcome by thermal energy and can be considered to be proportional to the magnetic anisotropy constant K and particle volume V (∆𝐸 = 𝐾𝑉), 𝑘𝐵 is the Boltzmann constant and 𝜏0 is the length of time characteristic of the probed material. Although superparamagnetism should be expected for quasistatic measurements (𝜏𝑚~100 s) above Curie temperature 𝑇𝐶 of the amorphous phase, it is worth noting that in Mössbauer spectroscopy 𝜏𝑚~𝜏0. Moreover, Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 16 [19] Blázquez JS, Ipus JJ, Franco V, Conde CF, Conde A. Extracting the composition of nanocrystals of mechanically alloyed systems using Mossbauer spectroscopy. Journal of Alloys and Compounds 2014;610:92-9 https://doi.org/10.1016/j.jallcom.2014.04.195 [20] Pizarro R, Garitaonandia JS, Plazaola F, Barandiaran JM, Greneche JM. Magnetic and Mossbauer study of multiphase Fe-Zr amorphous powders obtained by high energy ball milling. Journal of Physics-Condensed Matter 2000;12:3101-12 https://doi.org/10.1088/0953- 8984/12/13/318/ [21] Povstugar IV, Butyagin PY, Dorofeev GA, Elsukov EP. Kinetics of the initial stage of mechanical alloying in the Fe(80)Zr(20) system. Colloid Journal 2002;64:178-85 https://doi.org/10.1023/A:1015248810332 [22] El-Eskandarany MS, Bahgat AA, Gomaa NS, Eissa NA. Kinetics and formation mechanism of amorphous Fe52Nb48 alloy powder fabricated by mechanical alloying. Journal of Alloys and Compounds 1999;290:181-90 https://doi.org/10.1016/S0925-8388(99)00222-4 [23] Velez GY, Alcazar GAP, Zamora LE. Influence of atomic ordering on the Fe2Nb antiferromagnetic alloy. Journal of Magnetism and Magnetic Materials 2014;354:333-5 https://doi.org/10.1016/j.jmmm.2013.11.031 [24] Velez GY, Alcazar GAP, Zamora LE, Tabares JA. Structural and Magnetic Study of the Fe2Nb Alloy Obtained by Mechanical Alloying and Sintering. Journal of Superconductivity and Novel Magnetism 2014;27:1279-83 https://doi.org/10.1007/s10948-013-2428-5 [25] Brand RA, Lauer J, Herlach DM. The evaluation of hyperfine field distribuions in overlapping and asymmetric Mossbauer spectra: A study of the amorphous alloy PD77.5- XCu6Si16.5FeX. Journal of Physics F-Metal Physics 1983;13:675-83 https://doi.org/10.1088/0305-4608/13/3/018 [26] Egami T. Rapidly Solidified Alloys: Processes, Structures, Properties, Applications. New York: Marcel Dekker, Inc.; 1993. [27] Zarubova N, Moser N, Kronmuller H. KINETICS OF CRYSTALLIZATION IN A FE90ZR10 AMORPHOUS ALLOY. Materials Science and Engineering a-Structural Materials Properties Microstructure and Processing 1992;151:205-15 https://doi.org/10.1016/0921- 5093(92)90209-J [28] Vincze I, Vanderwoude F, Scott MG. Local structure of amorphous Zr3Fe. Solid State Communications 1981;37:567-70 https://doi.org/10.1016/0038-1098(81)90135-6 [29] Morup S, Hansen MF. Handbook of Magnetism and Advanced Magnetic Materials: John Wiley & Sons; 2007. [30] Franco V, Blazquez JS, Conde CF, Conde A, Kiss LF, Kemeny T, et al. Frequency dependence of the superparamagnetic transition in a Finemet-type nanocrystalline alloy. Physica Status Solidi a-Applied Research 2004;201:3314-8 https://doi.org/10.1002/pssa.200405471 [31] Blazquez JS, Franco V, Conde CF, Conde A, Ferenc J, Kulik T. High temperature coercivity of Nb-containing HITPERM alloys: Effect of Cu addition. Materials Letters 2008;62:780-3 https://doi.org/10.1016/j.matlet.2007.06.067 [32] Alleg S, Hamouda A, Azzaza S, Bensalem R, Sunol JJ, Greneche JM. Solid state amorphization transformation in the mechanically alloyed Fe27.9Nb2.2B69.9 powders. Materials Chemistry and Physics 2010;122:35-40 https://doi.org/10.1016/j.matchemphys.2010.03.006 Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 17 FIGURES 30 40 50 60 70 80 90 100 110 120 30 40 50 60 70 80 90 100 110 120 Fe hcp-Zr 1.5 h 50 h 35 h 22 h 16 h 12 h 8.5 h 6 h 2(degrees) bcc Nb b) 49.5 h 21.5 h 34.5 h 15 h 12 h 9 h 6 h 2(degrees) 1.5 h a) Figure 1. XRD patterns of the amorphous samples after milling up to different times for a) Fe70Zr30 and b) Fe70Nb30. 0 2 4 6 8 10 12 14 16 18 20 22 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 18 20 22 0 5 10 15 20 25 30 D (nm) milling time (h) b) Fe(Nb) Nb (Fe) D (nm) milling time (h) a) Figure 2. Evolution of crystal size of the a) Fe(Zr) and b) Fe(Nb) and Nb(Fe) (right) bcc phases as a function of milling time. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 18 0 5 10 15 20 25 2.86 2.87 2.88 2.89 2.90 Fe(Zr) Fe(Nb) a (10-10 m) milling time (h) Figure 3. Evolution of lattice parameter of the Fe(Zr) and Fe(Nb) bcc phases as a function of milling time. 010 20 30 40 50 39.0 39.5 40.0 40.5 41.0 41.5 42.0 42.5 43.0 010 20 30 40 50 2.5 2.6 2.7 2.8 2.9 Fe70Zr30 Fe70Nb30 Fe70Zr30 Fe70Nb30 amorphous halo position (2 degrees) milling time (h) a) b) M-M (10-10 m) milling time (h) Figure 4. a) Evolution of the amorphous halo position and b) metal-metal distance as a function of milling time. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 19 400 600 800 1000 -1 0 1 2 3Tonset dH/dt (W/g) T (K) Fe70Nb30 exo Fe70Zr30 Tonset Tpeak Tpeak Figure 5. Non-isothermal DSC scans at 20 K/min for both studied samples after 50 h (Zr containing alloy) and 49.5 h (Nb containing alloy) of milling. Crystallization peak and onset temperatures are indicated. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 20 Figure 6. Scanning electron microscopy images in secondary electrons mode for the two studied alloys after selected milling times. Insets show the histograms for the powder particle size. 020 40 60 80 0 20 40 60 Fe70Nb30 t=21.5 h probability (%) d (m) 020 40 60 80 0 20 40 60 Fe70Zr30 t=22 h probability (%) d (m) 020 40 60 80 0 10 20 30 probability (%) d (m) Fe70Zr30 t=50 h 020 40 60 80 0 10 20 30 Fe70Nb30 t=49.5 h probability (%) d (m) Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 21 -8 -6 -4 -2 0 2 4 6 8 -8 -6 -4 -2 0 2 4 6 8 -8 -6 -4 -2 0 2 4 6 8 intensity (u.a.) 9 h6 h intensity (u.a.) velocity (mm/s) 16 h velocity (mm/s) 50 h 35 h velocity (mm/s) 3 h -8 -6 -4 -2 0 2 4 6 8 -8 -6 -4 -2 0 2 4 6 8 -8 -6 -4 -2 0 2 4 6 8 9 h6 h intensity (a.u.) 3 h 15.5 h intensity (a.u.) velocity (mm/s) 34.5 h velocity (mm/s) 49.5 h velocity (mm/s) Figure 7. Experimental Mössbauer spectra (symbols) and model fitting (lines) for several milling times: Fe70Zr30 (up), Fe70Nb30 (down). Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 22 0 5 10 15 20 25 30 35 40 45 50 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 30 35 40 45 50 0 10 20 30 40 50 60 70 80 90 100 fraction area (%) milling time (h) site 1: 33 T site 2: 30.5 T hiperfine distribution quadrupolar distribution a) b) fraction area (%) milling time (h) Figure 8. Phase contribution to Mössbauer spectra fitting for a) Fe70Zr30 and b) Fe70Nb30 alloys. 010 20 30 40 50 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 010 20 30 40 50 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 b) <IS> (mm/s) millig time (h) a) <IS> (mm/s) millig time (h) Figure 9. Average isomer shift (<IS>) of paramagnetic contributions as a function of milling time for a) Fe70Zr30 and b) Fe70Nb30 alloys. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 23 0.0 0.5 1.0 1.5 2.0 0 10 20 30 40 50 60 70 80 90 0.0 0.5 1.0 1.5 2.0 0 10 20 30 40 50 60 70 80 90 Q (mm/s) dP(Q)/dQ (% mm-1s) 3 h 6 h 9 h 16 h 35 h 50 h Fe70Nb30 Q (mm/s) dP(Q)/dQ (% mm-1s) 3 h 6 h 9 h 15.5 h 34.5 h 49.5 h Fe70Zr30 Figure 10. Probability distribution of quadrupole splitting for both samples after several milling times. 0 5 10 15 20 25 30 35 0 1 2 3 4 5 6 7 8 0 5 10 15 20 25 0 1 2 3 4 5 6 7 8 Zr content (%) milling time (h) a) b) From ratio A(30.5T)/A(33T) From Vegard's law Nb content (%) milling time (h) Figure 11. a) Content of Zr inside the nanocrystals from Mössbauer results and b) content of Nb inside the nanocrystals from Mössbauer results and Vegard’s law. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 24 020 40 60 80 100 0 20 40 60 80 100 020 40 60 80 100 0 20 40 60 80 100 Fe70Zr30 Fe70Nb30 XXRD AMS a) b) XXRD XMS Figure 12. Amorphous fraction from XRD patterns as a function of a) the relative area of paramagnetic contribution and b) the estimated amorphous phase fraction from Mössbauer spectra. The dashed lines correspond to the identity line. 0 5 10 15 20 25 30 35 0 20 40 60 80 100 120 140 510 15 20 25 30 0 20 40 60 80 Fe70Zr30 Fe70Nb30 R/ milling time (h) R/ DFe (nm) Figure 13. Evolution of 𝑅/𝛿 as a function of milling time for both alloys. Inset shows the relation between 𝑅/𝛿 and 𝐷𝐹𝑒 obtained from XRD data. Journal of Non-Crystalline Solids, 494 (2018) 78-85 https://doi.org/10.1016/j.jnoncrysol.2018.04.061 25 Table 1. Crystallization temperature of studied alloys after 50 h (Zr containing alloy) and 49.5 h (Nb containing alloy) of milling, respectively Sample 𝑻𝒐𝒏𝒔𝒆𝒕 (K) 𝑻𝒑𝒆𝒂𝒌 (K) Fe70Zr30 862 956 Fe70Nb30 980 1006