The crystallization processes for t

The crystallization processes for the amorphous metallic alloys Fe74B17Si2Ni4Mo3 and Fe86B6Zr7Cu1 (at. %) were investigated using X-rays diffraction measurements performed in-situ during Joule-heating, with simultaneous monitoring of the electrical resistance. We determined the main structural transitions and crystalline phases formed during heating, and correlated these results to the observed resistance variations. As the annealing current is increased, the resistance shows an initial decrease due to stress relaxation, followed by a drop to a minimum value due to massive nucleation and growth of a-Fe nanocrystals. Further annealing causes the formation of small fractions of Fe-B, B2Zr or ZrO2, while the resistance increases due to temperature enhancement. In situ XRD measurements allowed the identification of metastable phases, as the g-Fe phase which occurs at high temperatures. The exothermal peaks observed in the differential scanning calorimetry (DSC) for each alloy corroborate the results. We also have performed DSC measurements with several heating rates, which allowed the determination of the Avrami exponent and crystallization activation energy for each alloy. The obtained activation energies (362 and 301 kJ/mol for Fe-B-Zr-Cu; 323 kJ/mol for Fe-B-Si-Ni-Mo) are comparable to reported values for amorphous iron alloys, while the Avrami exponent values (n = 1.0 or n = 1.2) are consistent with diffusion controlled crystallization processes with nucleation rates close to zero.

Keywords: crystallization kinetics, iron alloys, X-rays, DSC

1. Introduction

Great scientific effort has been developed in last years in order to predict and control the crystallization processes of metallic glasses, since several attractive properties of the resulting material are strongly related to the final attained microstructure. For instance, it has been shown that the nanocrystalline structure presented by Fe-Cu-Nb-Si-B alloys annealed at 813 K during 1 h is responsible for their excellent soft-magnetic properties1,2. A similar behavior is observed for alloys of the Fe-B-Zr-Cu family3,4. Several techniques have been applied to the study of crystallization processes in glassy metals, such as ferromagnetic resonance5,6, differential scanning calorimetry7,8, Mössbauer spectroscopy9, small angle X-rays scattering10,11, and electrical resistance variations12,13. On the other hand, there is an increasing number of studies which use synchrotron radiation sources, whose high intensity allow in-situ measurements to follow structural evolutions in real time.

Joule heating (JH) is a fast and highly reproducible annealing technique for amorphous metallic ribbons. It is worth noticing that it is a non-isothermal treatment, since the temperature in the sample depends on its resistivity, which, in turn, depends on the sample’s microstructure, that undergoes transformations upon heating, thus changing the temperature itself. A theoretical model that relates the structural transformations to the electrical resistance variations during Joule heating at a constant applied dc current was proposed by Allia et al., and used to estimate the temperature variations in an Fe-Ni-P-B alloy14,15. Measurements of the temperature evolution of Fe-B-Zr-Cu alloys during Joule annealing have been recently reported, showing a continuous temperature increase as the annealing current is linearly enhanced16.

As regards the crystallization kinetics during isothermal treatments, the time evolution of the crystalline volume fraction, fv, is given by the Johnson-Mehl-Avrami equation:


where K and n are time-independent parameters. A simple manipulation of Eq. (1) shows that, K and n can be determined by plotting the measured data, fv, as:


In the case of continuous heating, it can be shown that for E >> RT the following relations are true7,17,18:



where b is the heating rate, E is the activation energy, R is the gas constant, n is the Avrami exponent and the index p refers to the variables at a crystallization peak, i.e., Tp is the temperature of a peak, Kp is the rate constant at the peak and (dx/dt)p is the crystallization rate at the peak. If Tp2/b is measured in a series of exothermals taken at different heating rates, the plot of ln (Tp2/b) as a function of 1/Tp should be a linear function with slope E/R. The set of equations given by Eqs. 3 and 4 can then be used to determine the Avrami exponent n as an average of the set of parameters obtained for different heating rates.
In the present work, we investigate the crystallization process of two Fe-based amorphous metallic alloys, using X-rays diffraction (XRD) measured in-situ during Joule annealing, and simultaneously monitoring the electrical resistance variations. DSC measurements with different heating rates were also performed in order to determine important crystallization parameters, such as the activation energy and the Avrami exponent.

2. Experimental Procedures
The amorphous alloys of nominal composition Fe74B17Si2Ni4Mo3 and Fe86B6Zr7Cu1 (at. %) were rapidly quenched from the melt by planar flow on a fast rotating cylinder, producing high quality metallic ribbons of thickness 25 + 1 mm. Samples cut from each amorphous alloy were submitted to JH with an effective length of 10 mm between electrical contacts. The sample was firmly clamped at both ends by two pairs of electrical contacts, using a sample holder specially designed to minimize conduction thermal losses and keep the ribbon perpendicular to the X-ray beam, allowing for sample expansion during heating. The applied dc current was varied from 0 to 3.0 A while the voltage drop across the sample was measured via two independent contacts. The heating procedure was computer controlled, thus allowing the on-line analysis of the R(I) curve. During annealing, the sample was kept in vacuum (10-2 mbar) to avoid oxidation and minimize convective thermal losses.

It was shown in a previous work that for a constant applied current the structural changes occur during the initial 10 s (see Fig. 1 in Ref. 12). Consequently, in the case of a step by step enhancement of the annealing current each increment DI should be applied after a minimum waiting time Dtmin = 10 s. For longer time intervals one obtains reproducible R(I) curves, suggesting that the crystallization process is also reproduced. Using the characteristic R(I) curve as a guide, different current values were selected for in situ measurements of X-ray diffraction intensities. For each thermal treatment we performed 12 sequential exposures, where XRD patterns were registered on a curved Imaging Plate in the angular range 10° < 2q < 152°. The experimental setup for in situ XRD measurements requires detector displacement between film expositions, so we selected long time intervals Dt = 450 s for each current increment DI = 0.1 A. After waiting 20 s for temperature stabilization in the sample, the remaining 430 s were used for recording XRD spectra. These measurements were carried out at the SAXS beam-line of the LNLS – National Synchrotron Light Laboratory, Brazil19,20. The wavelength l = 1.7574 Å was selected in order to minimize the attenuation factor and fluorescence from Fe atoms in the samples, and the beam dimensions over the sample were 0.8 x 3.0 mm. Diffraction curves from a standard Al2O3 powder with controlled grain size (1 mm) were measured at room temperature in the same experimental geometry for 2q calibration. The average grain size was calculated from the full width at half maximum of the 110 a-Fe reflection, using the Scherrer formula (for details see Ref. 12).








DSC measurements were performed on a Shimadzu D-50 calorimeter from room temperature to 1000 K, using heating rates of 2, 5, 10, 20 and 30 K/min. Some treatments were interrupted at particular temperatures, and XRD patterns were obtained at room temperature using a Siemens D5000 diffractometer, in order to evaluate the corresponding stage of crystallization. There is a great similarity between JH and DSC heating procedures, since both are non-isothermal treatments. As the applied current is increased the sample temperature continuously increases16. On the other hand, the connection between the applied JH heating rate DI / Dt and the resulting temperature rate DT / Dt should be experimentally determined, since it depends on the balance between the heating power (I 2R) and the thermal losses. In Ref. 16, for example, it was shown that a Joule heating rate of 0.003 A/s produced an average temperature rate of 50 K/min for Fe-B-Zr-Cu ribbons.

3. Results and Discussion
Figure 1 presents XRD spectra obtained at different annealing currents. Figure 2 shows the corresponding R(I) curves, where solid dots represent the points selected for exposure to X-rays. The electrical resistance R showed a typical behavior as a function of the applied current, and very similar R(I) curves were obtained for both alloys. For the Fe-B-Zr-Cu alloy, the typical R(I) curve initially decreases, then presents a slow increase up to about 1.1 A. In this range the sample is still amorphous, thus the initial resistance decrease is interpreted as internal stress relaxation in the as-quenched material, followed by a resistance increase due to a temperature enhancement. Further annealing causes a resistance drop to a minimum at 1.7 A, while XRD spectra show the nucleation and growth of a-Fe nanocrystals, with grain sizes of about 10 nm, and increasing crystalline volume fraction in this current range. During the massive Fe crystallization the resistance is reduced down to a minimum, and then the samples presents a new metallic behavior with a rapid temperature increase due to a larger thermal coefficient of resistance (TCR)13,16. A similar behavior was observed for the Fe-B-Si-Ni-Mo alloy, with the beginning of structural ordering shifted to a higher current value (1.4 A), and the formation of larger a-Fe grains, about 40 nm in size.