Nanoscale mapping of phase-transformation pathways in medium-Mn TRIP steel by multimodal STEM

TL;DR

The study tackles nanoscale mapping of phase-transformation pathways in a medium-Mn TRIP steel by developing a correlative STEM workflow that combines NBED and EDS to simultaneously map lattice structure, orientation, and phase at ~10 nm. This approach yields quantitative, phase-resolved information on ferrite, austenite, and martensite fractions and their lattice parameters, revealing deformation-induced transformation to martensite with significant grain-size and texture changes. The Mn fingerprint in EDS enables robust ferrite–martensite separation, while kernel average misorientation highlights higher local distortion in martensite. The framework is transferable to other multiphase alloys and can be extended toward mesoscale diffraction and in-situ loading to connect nanoscale transformations with macroscopic mechanical behavior.

Abstract

The mechanical response of third-generation advanced high-strength steels is governed by phase transformations at the nanoscale, yet the coupled evolution of chemistry and crystallography remains poorly resolved. Here we apply a correlative scanning transmission electron microscopy approach that enables simultaneous mapping of lattice structure, crystallographic orientation, and phase distribution at 10 nanometre resolution in a medium-manganese TRIP steel. We combine nano-beam electron diffraction and energy-dispersive X-ray spectroscopy maps to characterize an industrial medium-manganese steel containing 7.15 weight percent Mn. Tensile testing of a rolled steel sample was performed, and lamellae were extracted from deformed and undeformed regions. Manganese-resolved energy-dispersive X-ray spectroscopy provides a chemical fingerprint that, when combined with nano-beam electron diffraction based phase segmentation, enables robust ferrite-martensite separation and phase-resolved lattice-parameter refinement. The phase fractions of ferrite, austenite, and martensite are quantified together with their corresponding lattice parameters, accompanied by measurable shifts in grain-size distributions and crystallographic texture in the deformed regions. Kernel average misorientation maps reveal systematically lower local misorientation in ferrite than in martensite. This multimodal workflow provides a transferable framework for quantitative, phase-resolved analysis of complex multiphase alloys at the nanoscale.

Figures (5)

• Figure 1: Workflow from sample preparation to selection of ROIs. a Schematic of the MMn1000 industrially rolled medium-manganese steel sheet. b Double-edge notched tensile (DENT) specimen extracted transverse to the rolling direction. c Scanning electron microscope images of the undeformed and deformed regions selected for TEM lamella preparation. d Dark-field scanning transmission electron microscopy (STEM) images of the lamellas, showing ROIs identified for further correlative analysis. e Corresponding bright-field STEM images indicating two of the ROIs studied in this work, corresponding to ROIs 2 and 4.
• Figure 2: Bright field images, spectral maps and phase weights. a Bright-field images showing intensity variations in electron transmission. b Energy-dispersive X-ray spectroscopy (EDS) signals corresponding to the Mn peaks aligned with the nano-beam electron diffraction data through affine transformations. c and d Map of the relative concentrations of body-centered cubic (BCC) and face-centered cubic (FCC) phases. There is a spatial overlap between ROIs 1 and 2 (undeformed lamella) and ROIs 3 and 4 (deformed lamella).
• Figure 3: Phase, orientation, grain and misorientation maps. a Phase maps showing the spatial distribution of ferrite, austenite, and martensite, generated by overlap of EDS and phase maps from Fig. \ref{['fig:EDS']}. b In-plane crystal orientations maps with inverse pole figure indicating the lattice plane with respect to the vertical direction. c Grain maps obtained by clustering neighboring pixels from the orientation map with misorientations below 5°, illustrating the grain morphology for each phase. d Kernel average misorientation (KAM) maps showing the average misorientation angle of each pixel with respect to its eight closest neighbors. The KAM colormap spans 045° and is centered at 5° (visualization only) to enhance discrimination between low-angle and high-angle boundaries.
• Figure 4: Quantitative analysis for each phase of the ROIs. a Lattice calibration of the ferrite regions (BCC). b Lattice calibration of the austenite regions (FCC). c Lattice calibration of the martensite regions (BCT). d Grain size distributions for each of the phases (legend in ROI 4) obtained from the clustered grain maps from Fig. \ref{['fig:Merit']}c .
• Figure 5: Reciprocal space calibration with Au-Pd. a Virtual dark-field image of the calibrant with 6 randomly selected pixels highlighted. b Diffraction patterns from the selected pixels and Bragg peaks harvested. c Virtual Bragg map combining all the peaks harvested from the calibrant dataset. d Integrated 1D diffraction pattern from c with the best lattice fit found during calibration.