Magnetic ordering suppressed phase transformation of a TRIP-HEA during thermal cycling

Transformation-induced plasticity (TRIP) is an effective mechanism to enhance the mechanical properties of metallic engineering materials, as it enables additional work hardening and more uniform deformation through activation of phase transformation under mechanical loads.^1,2 Recently, TRIP has been introduced to high-entropy alloys (HEAs) and achieved exceptional combinations of strength and ductility.^3–8 Li et al.³ reduced the atomic ratio of Mn to Fe in the FeMnCoCr quandary system and produced a dual-phase TRIP-HEA, which had a portion of hexagonal close-packed (HCP) phase embedded in a metastable face-centered cubic (FCC) phase at room temperature (RT). When under tension at room temperature, the FCC-to-HCP transformation was found easily triggered upon yielding and continued to near exhaustion at fracture, contributing to the persisting work-hardening of the material.⁹ Subsequently, TRIP-HEAs started to attract much attention of exploration and optimization.

Magnetic ordering has been recently found to pose significant impact on the deformation mechanisms of TRIP-HEAs. Niu et al.¹⁰ revealed the magnetically driven FCC-to-HCP phase transformation strengthening mechanism in a CoCrNi HEA. They also concluded that such mechanism is suppressed by the magnetically frustrated Mn element in a CrMnFeCoNi HEA as Mn eliminates the FCC and HCP energy difference. Wu et al.¹¹ investigated the role of magnetic ordering for the design of quinary Cr₂₀Mn_xFe_yCo₂₀Ni_z (x + y + z = 60, at. %) HEAs. They found that the stabilization of antiferromagnetic ordering is mainly governed by the Mn content, which changes the magnetic critical temperature and alters the stability of the FCC phase, thus introducing different deformation mechanisms (twining-induced plasticity, TWIP, and/or TRIP) and affecting the macroscopic deformation behavior. However, those studies focused on single-phase HEAs and did not consider the effect of magnetic ordering on the phase composition of dual-phase HEAs that can more directly and significantly affect the deformation behavior of a TRIP-HEA. Moreover, the kinetic connection between thermally induced phase transformation and magnetic ordering, which might affect the application temperature range of a TRIP-HEA, is not revealed.

In this study, the effect of magnetic ordering on the phase transformation of a dual-phase TRIP-HEA during thermal cycling was probed by in situ neutron diffraction. The results show how magnetic ordering kinetically affects the FCC-to-HCP phase transformation during cooling from a single FCC state and thus shed light on the strategy of tuning initial dual phase fraction by manipulation of magnetic properties of the TRIP-HEA.

The TRIP-HEA studied in this work had a nominal composition of Fe₅₀Mn₃₀Co₁₀Cr₁₀ (at. %). The alloy was first mixed with pure metals by vacuum arc melting and then drop-cast to form a rectangular bar. The as-cast bar was homogenized at 1423 K for 4 h in vacuum and then rolled at room temperature (RT) to a thickness reduction of 75%. The cold-rolled plate was cut into small plate samples and then annealed at 1223 K for 1 h, followed by air cooling to RT and then quenching in liquid nitrogen. The microstructure of the annealed TRIP-HEA was observed by electron back scattering diffraction (EBSD), presenting a morphology of lath-shaped HCP grains embedded within equiaxed FCC matrix, as shown in Fig. 1(a).

In situ neutron diffraction of the annealed TRIP-HEA during thermal cycling between RT and 425 K was carried out on the engineering materials diffractometer VULCAN^12,13 at the Spallation Neutron Source of Oak Ridge National Laboratory (ORNL). The in situ thermal cycling tests were performed using induction heating. A type K thermocouple was directly attached to the sample by spot welding. Unlike the postmortem localized measurements, the in situ time-of-light neutron diffraction enables a real-time acquisition of bulk diffraction patterns under thermal treatments. It can provide the real-time evolution of bulk average mesoscale information of structures from the atomic level, including the crystal and magnetic structures. The collected neutron diffraction data were reduced and analyzed by the VDRIVE,¹⁴ GSAS,¹⁵ and FullProf¹⁶ programs. The nuclear phase fraction and ordered magnetic structure were analyzed by using the Rietveld refinement method.

The magnetic property of the annealed TRIP-HEA was further measured by the superconducting quantum interference device (SQUID) at the Center for Nanophase Materials Sciences (CNMS) of ORNL. The magnetic susceptibility of the material was measured between 4 and 390 K with the decreasing and increasing temperature in the magnetic field of 1 × 10³ Oe. The magnetization measurements up to 4 × 10⁴ Oe were then performed on the annealed sample near the magnetic transition temperature.

The evolution of diffraction patterns in Fig. 1(b) reveals the occurrence of austenitic and martensitic transformation and magnetic transition (disordering/ordering) during thermal cycling. At RT, the diffraction pattern presented FCC and HCP crystal peaks and two minor standalone magnetic peaks with the greater one at 2.546 Å. Rietveld refinement on the full pattern revealed that the collinear 1Q ordered magnetic structure with a k = (001) propagation vector, as shown by the sketch aside the diffraction pattern at RT, can well fit the standalone magnetic peaks. When the temperature increased to ∼326 K, the magnetic peaks disappeared. The FCC phase presented a disordered magnetic state, as schematically illustrated by the middle sketch. As temperature further increased to 422 K, only the FCC phase was present, indicating the complete austenitization. Under subsequent cooling, martensitic transformation occurred as manifested by the re-emerged HCP peaks, while the FCC phase was still in a disordered magnetic state as no magnetic peak was present at 326 K. When cooling to RT, all FCC, HCP, and magnetic peaks almost restored to as prior to thermal cycling, which demonstrates that the crystal phase transformation and magnetic transition are reversible upon thermal cycling.

The real-time kinetic phase transformation during thermal cycling is clearly visualized through the HCP fraction evolution with temperature, as shown in Fig. 1(c). The TRIP-HEA consisted of ∼31% HCP at RT before heating. Upon heating, the HCP fraction began to decrease at ∼391 K, indicating the start temperature of austenite (A_s). The HCP fraction rapidly reduced to zero as temperature increased to ∼422 K, indicating the finish temperature of austenite (A_f). Under cooling, the HCP fraction started to increase evidently at ∼347 K, indicating the start temperature of martensite (M_s). When cooled to RT, the HCP fraction restored to ∼27%, very close to the value prior to thermal cycling, indicating that the phase transformation during thermal cycling is almost fully reversible.

Figure 1(c) also compares the intensity evolution of the magnetic peak at 2.546 Å with the HCP fraction evolution and reveals the interaction of crystal phase transformation and magnetic transition. Unlike the different turning points of HCP fraction due to austenitic transformation upon heating at ∼391 K and martensitic transformation under cooling at ∼347 K, the magnetic peak intensity showed the same turning point at ∼326 K due to disordering/ordering transition upon heating and cooling. Apparently, the magnetic disordering occurred prior to the austenitic transformation start and the magnetic ordering occurred after the martensitic transformation start. While the magnetic disordering could hardly affect the austenitic transformation progress as without occurrence overlapping, the magnetic ordering transition posed impact on the martensitic transformation as manifested by a kink of the HCP fraction evolution.

The impact of magnetic ordering on the martensitic transformation is revealed as a suppression effect from the comparison of the transformation rate and normalized magnetic peak intensity as a function of temperature, as shown in Fig. 2. The martensitic transformation rate, defined as the increase in HCP fraction per unit K and derived from Fig. 1(c), first increased with the decreasing temperature and then started to decrease when cooling to ∼326 K, yielding a bell-shaped curve. The temperature at the peak of the transformation rate coincides with the starting point of magnetic peak intensity rise, which is the critical temperature for antiferromagnetic ordering, i.e., the Néel temperature T_N, as further confirmed by the SQUID measurements in Fig. 3. Below T_N, the magnetic scattering intensity increased with the decreasing temperature, indicating the increasing ordering domains. In the meantime, the transformation rate decreased simultaneously. Thus, it is reasonable to interpret the effect as that magnetic ordering suppresses the martensitic transformation.

Magnetization measurements shown in Fig. 3 confirm the critical temperature for antiferromagnetic ordering of the annealed TRIP-HEA, i.e., the Néel temperature T_N ≃ 326 K. A sharp peak is observed in both M(T) curves at ∼326 K under field cool (FC) and field warm (FW) conditions, as shown in Fig. 3(a). At temperatures higher than T_N, the magnetization M increased with the decreasing temperature. Below T_N, the magnetization M generally decreased with the decreasing temperature. Such features are typical of antiferromagnetic materials, and the sharp peak indicates the onset of antiferromagnetic ordering with the decreasing temperature. The linear field dependence of magnetization at 300 and 350 K, as shown in Fig. 3(b), demonstrates the antiferromagnetic and paramagnetic nature of the material below and above T_N, respectively.

The underlying mechanism of magnetic ordering suppressed martensitic transformation can be understood from the perspective of energy. In general, the entropy change associated with FCC magnetic ordering counteracts the cooling induced chemical driving force for the martensitic transformation.¹⁷ More specifically, the FCC-to-HCP transformation is driven by the Gibbs free energy difference between the HCP and FCC, ΔG^hcp/fcc = G^hcp − G^fcc (expressed in molar quantity). A negative ΔG^hcp/fcc serves as the chemical driving force to overcome the energy barrier associated with the formation of HCP nucleus in the parent FCC matrix. It is derived from the energy differences related to the lattice stability and the excess energy of mixing due to nonmagnetic chemical interaction ΔG^nmg,hcp/fcc as well as magnetic contribution ΔG^mg,hcp/fcc of the two phases.¹⁸ For a given chemical composition, the competitive energy state of the two phases is essentially temperature dependent. Figure 4 shows the variation of ΔG^hcp/fcc with temperature for the TRIP-HEA estimated by thermodynamic calculation in the supplementary material. The ΔG^hcp/fcc decreases as temperature decreases and becomes negative at a temperature near the measured (A_s + M_s)/2 ≃ 369 K. The negative ΔG^hcp/fcc indicates the metastability of the FCC phase. While the nonmagnetic contribution ΔG^nmg,hcp/fcc almost linearly decreases with temperature in the calculation range, the magnetic contribution ΔG^mg,hcp/fcc rapidly increases with the decreasing temperature below T_N. The ΔG^hcp/fcc in total is, thus, lifted upward by the increasing ΔG^mg,hcp/fcc when the magnetic ordering acts below T_N. Accordingly, when cooling just below M_s, under every decreasing K, the continuous increase in the chemical driving force (−ΔG^hcp/fcc) facilitates more and more volume of FCC to be triggered for transformation and yields the increasing HCP phase per decreasing K, i.e., the increasing transformation rate in Fig. 2. However, as magnetic ordering stabilizes the FCC structure below T_N, the augment of the driving force per decreasing K is reduced. Subsequent transformation of the remained FCC becomes less active, corresponding to the reduction in the transformation rate below T_N in Fig. 2. Such lowered transformation rate results in the inflection point in Fig. 1(c) as an indicator of the stunted evolution of HCP fraction under further cooling. Therefore, the facilitation of cooling on transformation is gradually weakened due to magnetic ordering and the transformation rate was then reduced to low values and limited the HCP fraction at RT.

The interaction between magnetic ordering and crystal transformation in the TRIP-HEA depends on the chemical composition in principle. High content of Mn above ∼15% could significantly stabilize the FCC crystal and ordered magnetic structure, which is manifested by the decrease in M_s and the increase in T_N of the FCC structure with the increasing Mn content.^19–21 When the Mn content is relatively low, M_s is typically greater than T_N, i.e., M_s > T_N. The start of martensitic transformation is not interfered by the magnetic ordering. When the Mn content increases to a relatively high level, the M_s would significantly drops below the T_N, i.e., M_s < T_N, demonstrating the delay of transformation start by magnetic ordering.¹⁹ The typical threshold for M_s < T_N of Fe–Mn systems is ∼23%.¹⁹ In contrast, the TRIP-HEA with a much higher Mn content of 30% shows M_s > T_N. This is attributed to the additions of Co and Cr, which are known to lower the T_N of the FCC structure while only slightly lower the M_s.^17,21,22 Collectively, the stabilization effect of high Mn on the FCC structure and magnetic ordering is balanced by the addition of Co and Cr in the TRIP-HEA so that the start of transformation remains above RT and not delayed by the magnetic ordering.

The suppressing effect of magnetic ordering of the TRIP-HEA acts after the martensitic transformation starts. Though not interfering M_s, the T_N of the TRIP-HEA is still high enough due to the high Mn content so that the magnetic ordering suppresses the thermal transformation from the very early stage and, thus, restricts the HCP phase fraction of the material at RT. Such overlapping of magnetic transition with martensitic transformation also leads to the temperature-sensitive phase composition between RT and a hundred K above, as shown in Figs. 1(c) and 2. It, thus, implies a worth-noting temperature sensitivity of the mechanical properties of the dual-phase TRIP-HEA, as the HCP phase was found to show much higher yield stress than the FCC phase and quite different strain-hardening behavior.⁹ In addition to the effect on thermally induced transformation and the resulting initial phase fraction, the Mn content is also found to affect the deformation behaviors of Mn-containing TWIP–TRIP HEAs.¹¹ The underlying mechanism is that Mn stabilizes the antiferromagnetic configurations that impact the stacking fault energy (SFE),¹¹ while low SFE is known to introduce TWIP (SFE ∼20–40 mJ/m²) and TRIP (SFE < 20 mJ/m²).^23,24 In general, the magnetic ordering suppresses the thermally and mechanically induced transformation as it stabilizes the FCC structure and lowers the driving force for transformation. By tuning the Mn content with balanced Co/Cr, the difference between M_s and T_N could be enlarged to increase the obtained HCP fraction so to enhance the yield strength of the dual-phase material. On the other hand, the magnetic ordering could elevate the SFE and postpone the start of mechanically induced transformation or TRIP to accommodate large deformation at high loads. Therefore, with more efforts on quantifying those effects of magnetic ordering in future, both the initial phase fraction and SFE can be designed and engineered by manipulating the Mn content to achieve better mechanical properties of Mn-containing alloys.

It should be noted that, besides the suppressing effect of magnetic ordering, the phase transformation of the TRIP-HEA during cooling can also be affected by microstructures, including dislocations, grain morphology, and texture through the kinetic process of nucleation and growth. Certain level of dislocations and grain size can favor HCP nuclei formation and growth, while high dislocation density and strong texture by cold working can significantly suppress the transformation. Such changes in microstructures can result in large variation in the transformation critical temperature and critical external stress as well as the phase fraction, shedding lights on potential ways to control the phase transformation by tuning microstructures other than tuning the chemical driving force. The kinetic and intertwined effects of microstructures on the phase transformation will be discussed in our on-going work with the focus on the microstructure dependence of phase metastability.

The kinetic phase transformation of a dual-phase TRIP-HEA Fe₅₀Mn₃₀Co₁₀Cr₁₀ (at. %) during thermal cycling between RT and 425 K was probed by in situ neutron diffraction. The results revealed the austenitic transformation start temperature A_s ≃ 391 K and finish temperature A_f ≃ 422 K and the martensitic transformation start temperature M_s ≃ 347 K. More importantly, the results revealed the suppression effect of magnetic ordering during the martensitic transformation under cooling, which is manifested by the concurrence of reduced transformation rate and increased magnetic scattering starting at ∼326 K. This critical temperature was further confirmed as the Néel temperature, T_N, by magnetization measurements. Thermodynamic calculations revealed that the high Mn content dominated antiferromagnetic ordering reduced the chemical driving force for the martensitic transformation and, thus, suppressed the phase transformation soon after it started. Such effect of magnetic ordering on thermally induced transformation provides insights into a potential strategy of designing and engineering the magnetic properties by tuning the Mn content to achieve better mechanical properties of Mn-containing alloys.

See the supplementary material for the thermodynamic calculation of Gibbs free energy of the studied material and the related parameters for the Gibbs expression.

Neutron scattering experiments were carried out at the Spallation Neutron Source (SNS) at Oak Ridge National Laboratory (ORNL). Microscopy experiments were performed at the Center for Nanophase Materials Sciences (CNMS) at ORNL. SNS and CNMS are national user facilities at ORNL sponsored by the Scientific User Facilities Division, BES, DOE. The authors gratefully thank Dr. Hongbin Bei for providing the tested material. S.F. is grateful to the financial support from China Scholarship Council for her visit at SNS, ORNL. This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains, and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).