New Peritectoid Reaction Identified at the MnSb Alloy

The tunable Tc region of MnSb alloy (44 49 % at. Sb) was analyzed using OM, VSM, DSC, XRD, EDS and XRF characterization techniques. Thermal and magnetic analysis suggests the existence of a non reported irreversible reaction on heating, compatible to a reverse peritectoid transition MnxSb ➔ Mn2Sb + Mnx'Sb.


Introduction
The MnSb alloy presents an interesting phase, on atomic percentage of Sb (%at.Sb) from 45% to 48%, with Curie Temperature (Tc) varying from 90 °C to 314 °C. The composition limits of the MnSb phase and the corresponding Tc varies significantly at the reported phase diagrams (1)(2)(3)(4)(5)(6), these divergences are summarized at Table 1. The minimum limit of the MnSb phase varies from 40.0 to 46.0 %at.Sb, and the maximum limit goes from 49.0 to 50.5 %at.Sb. Also the temperature of peritectic (from 840 to 853 °C) and eutectic (from 568 to 574 °C) reactions are reported at different temperatures. These divergences can be in part related to temperature range evaluated, presence of oxidation, melting and heat treatment methods, and/or chemical composition analysis method.

Materials and Method
Eight samples with atomic percentage of Sb (%atSb), from 43% to 50%, were prepared with high purity elements (Sb -Alfa Aesar 99.99% and Mn Alfa Aesar 99.98%), were weighed on precision equipment and melted, under argon atmosphere, at electric arc furnace (EAF). Each sample was melted 6 times, inverting its position after each melt. The ingots were annealed at 500°C for 5 days, slowly cooled (1°C min -1 ), named as "Rec", a part of the ingots were cut and annealed at 820 °C for 10 days and quenched and named as "TT". The ingots' chemical composition was checked at X-Ray Fluorescence (XRF) equipment, and the phases' stoichiometry at Scanning Electron Microscopy with Energy Dispersive Spectroscopy (EDS) equipment.
Also XRD (X-Ray Difractometer) with Rietveld method analysis, DSC (Differential Scanning Calorimeter), VSM (Vibratory Sample Magnetometer) and OM (Optical Microscope) equipments were used to characterize the samples.

Results and Discussion
The samples were identified in accord to its nominal composition, from "1 = 43% at Sb" to 8 = 50% at Sb, and respective heat treatment (TT or Rec). This study was focused on the region of "tunable" Curie Temperature, the partial phase diagram, Fig. 1., locates the quenched (TT) and annealed (Rec) samples analyzed in this study. Images obtained from OM, Fig. 1, evidenced a light gray phase named as "a" corresponding to Mn x Sb present on all samples, a dark phase (granular and stripes) named as "b" corresponding to Mn 2 Sb present on TT_2, TT_3 and Rec_2 samples, and an almost white phase named as "c" -predominantly Sb, present on TT_8 and Rec_8 samples. EDS analysis nominally confirmed the composition of the phases. The other micrographics' images show basically the same phases.
The reported Curie temperature (Tc) x stoichiometry, on known phase diagrams 1-3 , indicates higher Tc as the percentage of Sb increases. The measurement of composition at XRF and precision balance and their respective Tc do not converge to reported data, but do with EDS measures of the individual's phase, this can be explained by the first two are general measurements and the third is specific at MnSb phase. Giving evidences the Tc registered is exclusively related to Mn x Sb phase ("x" is the stoichiometry variation).
The chemical composition of Mn x Sb measured at EDS, from 49.4% to 52.3 %at.Sb meaning "x" has a value between 1.02 and 0.91 (1.02>x>0.91) are not completely agreeing with phase diagrams 1-3 , where Mn x Sb varies from 45.5% to 50.0% at Sb (1.20>x>1.00). The samples Rec_1 and Rec_8 presents a secondary phase (Mn 2 Sb or Sb), meaning it exceeds the Mn x Sb phase limits. Extrapolating the data Tc x composition as a straight line from closer compositions, the range limit should be 49.5% at Sb for 90 °C, and 53.0 % at Sb for 316 °C, and "x" factor should vary from 0.883 to 1.020 on Rec samples. The Table 2 provides the collected data but is ordered in accord to "x" value of Mn x Sb phase, exclusively where it has only one phase (Rec_1 and Rec_2 presents two phases).
The magnetic analysis (MxT) of TT and Rec samples, plotted at Fig. 2"a- Teramoto 1 reported an irreversible reaction when analyzing a 46 %at.Sb quenched sample, after each measurement of five cicles, the susceptibility curve moved to upper temperatures, he attributed this behavior to the "precipitation of Mn 2 Sb phase from MnSb". Chen 2 also mentioned the precipitation of Mn 2 Sb phase.
Only the sample Rec_8 didn't present this hysteresis, suggesting, at this stoichiometry and temperature range, probably the peritectoid transition does not occur anymore. On the other hand, the events M 4 and M -4 were not expected. A similar event was reported by Nwodo 3 as a FOMT (first order magnetic transition) AFM-FI (Anti-ferromagnetic ➔Ferrimagnetic) reaction, attributed to a spin reorientation of Mn 2 Sb dopped with Sn (Mn 2 Sb 0.9 Sn 0.1 ). At Rec_8 and TT_8 the present phases are "Mn x Sb + Sb", where "Sb" is apparently providing an AFM behavior to the alloy up to 256.69 °C on heating, when a FOMT occurs (event M 4 -AFM ➔ FM), followed by an M 3  The thermal analysis (DSC), Fig. 3, of Rec samples evidenced two endothermic peaks, the first identified as E1 from 65.98 to 79.77 °C, which is compatible to the reported precipitation of Mn 2 Sb from Mn x Sb, probably related to a crystaline structure change from Mn x Sb (hexagonal) to form Mn 2 Sb (tetragonal). And the second, E2, near 573 °C, compatible to a reverse peritectic reaction "Mn x Sb ➔ Mn x' Sb + Liquid", reported by Okamoto 4 at 570 °C, and by Kainzbauer at 566 °C, but in disagreement with lower levels of Sb (Rec_3 to Rec_8), where it was expected to be near to 840 °C.

Conclusions
The magnetic hysteresis between heating and cooling suggests a reaction with compositional change. The thermal analysis revealed two endothermic peaks E1 ~ 80 °C compatible to the Mn 2 Sb precipitation mentioned previously, and E2 ~ 573 °C on samples Rec_3 to Rec_8 samples agreeing to peritectic reaction at the limit border of Mn 0.883 Sb phase and "Mn 0.883 Sb + Sb" solid solution. The absence of E2 on samples Rec_1 and Rec_2, a two phase region of "Mn 2 Sb + Mn x Sb" is probably because at this stoichiometry there are no more exceeding Mn atoms to diffuse and permit the precipitation of Mn 2 Sb phase from MnSb.
The results suggests the existence of a reverse peritectoid reaction (Mn x Sb ➔ Mn 2 Sb + Mn 0.883 Sb) on heating at samples Rec_3 to Rec_6 (or more precisely 1.020≥x>0.883), also the existence of the two phase solid (Mn 2 Sb + Mn x Sb) at high temperature region (above E1 ~ 79 °C) from 33% to near 53 %at.Sb, but limited from 33% to 49.7 %at.Sb under this temperature.