Re-Assessment of the Strength and Deformation Behavior of Binary B2-ordered Iron Aluminides

1 Re-Assessment of the Strength and Deformation Behavior of Binary B2-ordered Iron Aluminides Jan Lars Riedela, Alexander Kauffmannb*, Stefan Gutha, Marcel Müncha, Georg Winkensa, Daniel Schliephakea, Jung Soo Leec, James P. Bestc, Frank Steinc and Martin Heilmaiera a Institute for Applied Materials (IAM), Karlsruhe Institute of Technology (KIT), Engelbert-ArnoldStr. 4, D-76131 Karlsruhe, Germany b Institute for Materials (IM), Ruhr University Bochum (RUB), Universitätsstraße 150, D-44801 Bochum, Germany c Max Planck Institute for Sustainable Materials GmbH, Max-Planck-Straße 1, D-40237 Düsseldorf, Germany * corresponding author mail: alexander.kauff[email protected] (A. Kauffmann) phone: +49 234 32 18430 Abstract Iron aluminides are potential candidates for structural applications at temperatures up to 700 °C. The 1 status of literature lacks a consistent and comprehensive understanding of the mechanical properties of 2 binary iron aluminides within the concentration range of B2-ordered FeAl with 30 to about 50 at.% Al. 3 The present study addresses this gap by systematically characterizing the compositionand temperature4 dependent mechanical behavior of a series of alloys with 30 to 53 at.% Al from room temperature (RT) 5 up to 700 °C. Particular emphasis was placed on achieving a consistently low impurity level and on 6 establishing comparable heat-treatment conditions to reduce the strengthening effect of point defects. 7 At RT, both hardness and yield strength exhibit a distinct trend after a vacancy minimizing heat 8 treatment at 400 °C for 120 h, first decreasing towards a minimum at 42 at.% and thereafter increasing 9 again to 53 at.% Al. RT compression tests reveal high strain hardening capability for alloys with < 50 10 at.% Al that diminishes as the Al content increases. Contrary to earlier reports, the increasing yield 11 strength with increasing temperature was only observed for alloys with 30 and 35 at.% Al. Post12 deformation SEM-BSE and SEM-EBSD analyses reveal uniform plastic deformation depending on 13 alloy composition below 400–600 °C, while kink band formation at higher temperatures significantly 14 reduces strain hardening. Dynamic recrystallization was observed at 700 °C in alloys with < 47 at.% Al. 15 Keywords Iron aluminides, B2 ordered structure, compression testing, Vickers hardness, Nanoindentation, 16 deformation, strength 17 2 1. Introduction Binary iron aluminides Fe𝑥Al1−𝑥 possess an outstanding corrosion and sulfidation resistance [1,2], a 18 high density-normalized Young’s modulus and low density compared to Ni-based superalloys [2,3]. 19 Since they are made of affordable elements [4], they are suitable for applications like coal 20 gasification [1], heat exchanger tubes, and protective coatings [3,5,6]. Binary body-centered cubic (bcc)- 21 based iron aluminides form different, disordered and ordered crystal structures depending on the Al 22 content, which determines strength, strain hardening capability, hardness and plastic deformability. At 23 room temperature (RT) [2], the A2-disordered crystal structure forms up to 18 at.% Al, the D03-ordered 24 structure appears between 18 and 35 at.% Al, and the B2-ordered crystal structure forms at Al contents 25 up to 50 at.% Al. For high-temperature application, B2-ordered iron aluminides between 35 and 50 at.% 26 Al are of particular interest as they are exceptionally oxidation and sulfidation resistant [1,2]. 27 The RT strength of iron aluminides is rather low (~ 300 MPa for up to 48 at.% Al [7]) and strongly 28 depends on the Al content [8] as well as on point defects like impurity solute atoms and vacancies [7]. 29 The temperature-dependent yield strength σy of iron aluminides, which is divided into four regions, is 30 shown in Figure 1 with: 31 - Region I: Strength decreases with increasing temperature by thermal activation of dislocation 32 motion [11,12] 33 - Region II: Strength is dependent on the intrinsic lattice resistance 34 - YSA Region (“Yield strength anomaly”): σy can increase with increasing temperature for alloys 35 with < 47 at.% Al by more than 100 MPa compared to σy at RT (YS peak) [2,10,13–16] 36 - Region III: A drop of σy occurs due to the onset of diffusion-controlled plastic deformation [12] 37 Figure 1: Schematic temperature dependence of 𝜎𝑦 and 𝜎𝑓 of B2-ordered iron aluminides adopted from Refs. [2,9,10]. The transition temperatures between two regions are visualized with yellow areas. Mechanical strength of iron aluminides in the regions I, II and the YS peak is sensitive to the 38 concentration of vacancies [16], which in turn is influenced by the heat treatment and cooling 39 conditions [17]. Retained vacancies from heat treatment at high temperatures, for example 1000 °C, raise 40 σy within the regions I, II and YSA [2]. Observation of the YS peak is dependent on a vacancy-reducing 41 heat treatment at 400 °C. Thermally generated vacancies at temperatures below the YS peak are 42 3 considered to be the reason for the increase of σy. They exhibit a low formation (~1 eV) and high 43 migration enthalpy (up to 2.3 eV) [9,10]. The low temperature heat treatment at 400 °C is intended to 44 remove these excess vacancies, enabling “close to thermodynamic equilibrium conditions” [17,18]. 45 Durations vary in the range of 100–120 h across different literature reports [9,17,19,20]. A YS peak has 46 been reported for example for 120 h in Ref. [9]. 47 The determination of the temperature-dependent σy as the onset of macroscopic plastic deformation can 48 be obstructed in the case of brittle materials by premature normal fracture (cleavage) in tension. To 49 circumvent the cleavage and premature failure of a specimen, compression tests can be utilized. As 50 illustrated in Figure 1, the fracture stress σf is less sensitive to a temperature increase, just following the 51 trend of elastic stiffness. For brittle materials, it may be lower than the yield strength σy, specifically at 52 low test temperatures. Thus, a transition from brittle to ductile material behavior is observed, which 53 defines the so-called brittle-to-ductile transition temperature (BDTT) [21]. Consequently, σy is not 54 directly accessible in tensile tests at temperatures below BDTT, as is for B2-ordered iron aluminides, 55 because they suffer from weak grain boundary cohesion [22,23] and inherent brittleness [24]. Notably, 56 the BDTT for binary B2-ordered iron aluminides with up to 45 at.% Al has only been determined in 57 four-point bending tests for alloys after casting [8], where a continuous increase with increasing Al58 content up to 42 at.% Al and a strong increase for higher Al-concentrations was found. 59 Interpreting temperature-dependent mechanical properties in iron aluminides can be challenging when 60 the distinction between σy and σf in tension is not made explicit in the literature, specifically when the 61 test temperature is close to or below BDTT. In several reports, the YS peak is associated with stress 62 values that may potentially reflect σf rather than the intrinsic onset of plasticity. For the temperature 63 range in which the YSA is discussed, stress–strain curves are often not available, e.g. in Refs. [10,13– 64 16], which makes it difficult to confirm yielding. This view is consistent with the fracture surfaces 65 described in Refs. [13,15,16,25–39], which frequently exhibit brittle characteristics. Together, the 66 observations raise the possibility that some reported stresses capture fracture limits rather than yield 67 strengths. Accordingly, the mechanistic understanding of the YS peak and the temperature dependence 68 of strength in these alloys remains partially incomplete. In addition, the mesoscopic microstructural 69 changes induced by plastic deformation have been less systematically characterized. 70 The present study aims at generating appropriate datasets on σy with the respective stress-strain curves, 71 the analysis thereof and the microstructure after deformation over the entire Al concentration range of 72 B2-ordered iron aluminides. Specifically, it is focused on (i) keeping consistent impurity concentrations 73 between the different alloys and (ii) establishing comparable, defect-lean microstructural conditions by 74 appropriate heat treatments. By utilizing the combination of compression tests, Vickers hardness 75 measurements and nanoindentation experiments on B2-ordered iron aluminides cast from the same batch 76 of raw materials, a comprehensive overview over the mechanical properties on the mesoand 77 macroscale is provided. The experimental approach aims to clarify the following scientific questions: 78 1. How do mechanical properties, such as hardness, yield strength and strain hardening at RT up 79 to 700 °C, of single-phase, B2-ordered FeAl alloys with Al concentrations between 30 and 80 53 at.% Al depend on temperature and composition? 81 2. What changes occur in the mesoscale microstructure depending on test temperature and alloy 82 composition? 83 3. How does the temperature-dependent evolution of the mesoscale microstructure correlate with 84 the mechanical properties of the alloys? 85 4 2. Materials and Experimental Methodology The alloys used in this study with nominal compositions of 30, 35, 42, 47, 50 and 53 at.% Al were 86 produced in a vacuum induction furnace (Balzers and Co., Bad Schönborn, Germany) from pure Fe 87 (99.9 % purity) and Al (99.95 % purity) under Ar and cast into rectangular Cu molds with dimensions 88 of (200 ∙ 40 ∙ 180) mm3. For all alloys, the same raw materials were used to keep the impurity contents 89 comparable among the alloys. 90 The ingots were cut to cuboid testing samples with the nominal size of (3 ∙ 3 ∙ 5) mm3 by utilizing a 91 wire electric discharge machine (EDM), model BA24 supplied by Mitsubishi Electric 92 Corporation (Tokyo, Japan). After EDM processing, the samples were ground to grit P1000 SiC paper 93 to remove the oxidized surface layer followed by ultrasonic cleaning in ethanol. For the two-stage heat 94 treatment process, the compression testing samples were encapsulated into fused silica ampoules, which 95 were evacuated and back-filled with Ar of 99.996 % purity for five times. To prevent Si diffusion from 96 the fused silica into the specimens during heat treatment [40], the samples were wrapped in thin Mo-foil 97 (ThermoFisher Scientific, Waltham Massachusetts, USA) before encapsulation. The heat treatment 98 consisted of homogenization at 1000 °C (0.73 … 0.93 ∙ 𝑇S, solidus temperature 𝑇S) for 48 h (except for 99 Fe-53Al with 168 h) with subsequent rapid cooling by dropping the ampoules into water. This procedure 100 is followed by a low temperature heat treatment at 400 °C for 120 h also in ampoules to reduce excess 101 vacancies [9] and again subsequent rapid cooling. A laboratory chamber furnace CWF1300 by Carbolite 102 Gero GmbH & Co. KG (Neuhausen, Germany) was used for all heat treatments. The oxide scale on the 103 samples was removed by grinding to P1000 SiC paper. 104 For microstructural investigations, a standard metallographic grinding and polishing procedure was 105 applied. The sample surfaces were first ground with SiC paper to grit P4000. Subsequently, mechanical 106 polishing was carried out utilizing a non-crystallizing oxide polishing suspension with pH = 7, particle 107 size 50 nm (OP-S, Sommer Diamant Abrasive GmbH, Euenheim, Germany). This step was followed by 108 chemo-mechanical polishing with a non-crystallizing oxide polishing suspension with pH = 9.8, particle 109 size 40 nm (OP-S, Struers ApS, Ballerup, Denmark) to further remove the deformed layer after grinding. 110 Finally, chemo-mechanical vibratory polishing was applied for 8 h utilizing a non-crystallizing oxide 111 polishing suspension with pH = 9.8 (OP-S NonDry, Struers ApS, Ballerup, Denmark). For 112 nanoindentation, the samples were mechanically polished up to 1 µm diamond suspension, followed by 113 electropolishing using A2 electrolyte (Struers ApS, Ballerup, Denmark). 114 The crystal structure of all alloys was determined by X-ray diffraction (XRD) with a D2 Phaser device 115 supplied by Bruker Corporation (MA, USA). The chemical composition was determined in the as-cast 116 condition by means of inductively coupled plasma optical emission spectroscopy (ICP-OES) with an 117 iCAP 7600 DUO analyzer (ThermoFisher Scientific, Waltham Massachusetts, USA), carrier gas hot 118 extraction (CGHE) and combustion analysis (CA) with TC 600 and CS 600 devices, both supplied by 119 LECO (St. Joseph, Michigan, USA). Backscattered electron (SEM-BSE) and secondary electron 120 imaging (SEM-SE) were performed utilizing a Leo 1530 field emission gun scanning electron 121 microscope (SEM, Carl Zeiss AG, Oberkochen, Germany). SEM-BSE imaging was used for obtaining 122 information about homogeneity by using atomic number contrast. In addition, SEM-SE contrast imaging 123 was used for identification of topographic features such as pores and cracks. The acceleration voltage 124 was set to 20 kV in all cases. Mesoscale deformation mechanisms were investigated by using an Auriga 125 60 focused ion beam field emission gun SEM (Carl Zeiss AG, Oberkochen, Germany) equipped with 126 an EDAX DigiView electron backscatter diffraction system (EBSD, AMETEK Inc., Berwyn, USA). 127 Data were acquired on an area of (800 ∙ 800) µm2 in size at a step size of 2 µm. EBSD data was evaluated 128 with MTex [41] in MATLAB. 129 5 Quasi-static compression tests were performed on a Z100 universal testing machine with an electro130 mechanical drive supplied by ZwickRoell GmbH & Co. KG (Ulm, Germany) equipped with a three131 zone vacuum furnace and temperature controller by Maytec GmbH (Singen, Germany). The heating rate 132 was 10 K/min and the sample was held at the testing temperature for 30 min prior to testing. The initial 133 engineering strain rate 𝜀˙ was set to 10−4 s−1. Hexagonal BN was used as a lubricant to reduce friction 134 forces between samples’ faces and punches. The temperature of the sample was measured during testing 135 with a type S thermocouple, which was applied to the center of the sample. The atmospheric pressure 136 inside the furnace was kept lower than 1 ∙ 10−4 mbar during all tests. For each temperature step, three 137 compression tests were conducted to check reproducibility of test data. 138 The hardness of the alloys was evaluated by using a Q10A+ semi-automatic Vickers hardness (HV) 139 indenter from ATM Qness GmbH (Mammelzen, Germany) with a load of 1kg (HV1). A minimum 140 number of ten indents within a random selection of grains were evaluated for statistical reasons. The 141 minimum distance between two indents were at least three times the largest diagonal of the indents, the 142 distance to the sample edges were at least six times the largest diagonal, according to DIN EN ISO 6507143 1:2024 [42]. Nanoindentation was performed on different grains of the specimens with a distance of at 144 least 30 µm from one indent to another (more than 25 times the maximum indentation depth) using a 145 G200 nanoindenter by KLA Corporation (Milpitas, CA, USA). A diamond Berkovich indenter was used 146 with a maximum indentation depth of 1 µm at a constant strain rate of 0.02 s−1. Nine indents were 147 performed on each grain. The analysis developed by Oliver and Pharr [43,44] was used to determine the 148 nanohardness. 149 3. Results 3.1 Chemical Analysis and Initial Microstructure The chemical composition of all alloys after casting was verified using ICP-OES, CGHE and CA. The 150 results presented in Table 1 show that the actual alloy compositions closely align with the desired ones. 151 Additionally, impurity levels are consistently low across the different alloys, which has previously been 152 reported for this processing route in Ref. [14]. N and S were below the detection limit. 153 Table 1: Chemical composition of the investigated alloys by ICP-OES (Al, Mn, Si) and CGHE (O, N) and CA (C, S). Fe is balanced. Desired Al at.% Al at.% Al wt.% Mn wt.ppm Si wt.ppm O wt.ppm N wt.ppm C wt.ppm S wt.ppm 30 30.1 ± 0.7 17 ± 0.4 347 ±10 69 ±20 12 ±30 < 1 65 ±15 < 20 35 34.8 ± 0.9 20.3 ± 0.5 332 ±10 117 ±20 99 ±30 < 1 55 ±15 < 20 42 41.4 ± 1.2 25.1 ± 0.7 310 ±10 122 ±20 163 ±30 < 1 9 ± 15 < 20 47 46.6 ± 1.3 29.5 ± 0.8 323 ±10 140 ±20 95 ±30 < 1 46 ±15 < 20 50 49.9 ± 1.3 32 ± 0.8 311 ±10 138 ±20 14 ±30 < 1 52 ±15 < 20 53 52.4 ± 1.4 34.4 ± 0.9 276 ±10 151 ±20 65 ±30 < 1 31 ±15 < 20 To investigate the microstructure following the homogenization at 1000 °C and subsequent low154 temperature heat treatment at 400 °C, SEM-BSE micrographs were recorded and XRD measurements 155 were conducted. Representative micrographs for all alloys are shown in Figure 2, while the XRD data 156 are provided in the Supplementary Material (Figure S1). 157 6 Figure 2: The microstructure of iron aluminides after heat treatment at 1000 °C for 48 h (Fe-53Al: 168 h) and 400 °C for 120 h: a) Fe-30Al, b) Fe-35Al, c) Fe-42Al, d) Fe-47Al, e) Fe-50Al, f) Fe-53Al. The scale bar is the same for all images. Fe53Al shows artefacts from preparation. All alloys exhibit grain sizes in the range of 200–300 µm. The micrographs indicate that, apart from Fe158 53Al, all alloys possess a homogeneous, single-phase microstructure at the micrometer scale. Based on 159 the applied heat treatment at 400 °C and the phase diagram reported in Ref. [2], Fe-30Al is D0₃-ordered, 160 while Fe-35Al, Fe-42Al, Fe-47Al, and Fe-50Al are single-phase B2-ordered. In the case of Fe-53Al, a 161 secondary FeAl2 phase [2] is observed as lenticular features both within grains and along grain 162 boundaries after 400 °C/120 h (Supplementary Material Figure S2 for more details). 163 3.2 Mechanical Properties 3.2.1 Hardness tests The hardness is an indicator for vacancies retained in the alloys, if grain size and other microstructural 164 parameters remain constant. Thermal vacancies follow an Arrhenius-like relationship with 165 temperature [45,46]. The Vickers hardness (HV) at RT after the homogenization at 1000 °C (HT) for 166 48 h (single-phase Fe-53Al: 168 h, shown in Supplementary Material Figure S3) is presented in Figure 167 3. 168 Figure 3: HV as a function of alloy composition after a heat treatment at 1000 °C for 48 h (Fe-53Al: 168 h). For comparison, data from Ref. [17] are plotted as black hexagons. A dotted line is used as a guideline to the eyes. 7 For samples with this heat treatment, the HV continuously increases with increasing Al content, being 169 consistent with the data published in Ref. [17], included in Figure 3 for comparison. In order to reduce 170 the amount of thermal vacancies present in the alloys after HT as much as possible [17], the alloys 171 discussed here were subjected to a heat treatment at 400 °C for 120 h. The results are presented in Figure 172 4. 173 Figure 4: HV and NH as a function of alloy composition after a heat treatment at 400 °C for 120 h. For comparison, data from Ref. [17] are plotted as black symbols. Dashed and dotted lines are used as a guideline to the eyes. After the additional heat treatment at 400 °C, the HV of Fe-30Al does not change in comparison to HT 174 and remains at (2.7 ± 0.1) GPa. As also no change in constituting phases nor in grain size was observed 175 for this alloy, it can be concluded that there is no or a negligible change in point defect concentration. 176 For all other alloys, the hardness after the heat treatment at 400 °C for 120 h is significantly lower 177 compared to HT. Fe-42Al shows the lowest HV among the alloys tested. By further increasing the Al 178 content, the hardness increases again. The results obtained from nanohardness (NH) measurements show 179 the same trend as HV for 400 °C after 120 h but are shifted to higher values because of the indentation 180 size effect [47]. 181 3.2.1 Compression tests To investigate the yield strength and strain hardening behavior, compression tests were performed. 182 Representative true stress – true strain curves for the alloys deformed at RT after a heat treatment at 183 400 °C for 120 h are presented in Figure 5a. All data from the compression tests are available via Zenodo 184 under CC BY-SA 4.0 license . 185 The compression tests for the alloys with < 50 at.% Al content have been deliberately interrupted at (15 186 ± 1) % true strain, as marked by arrows, whereas the samples for Fe-50Al and Fe-53Al failed after 187 reaching true strains of (15 ± 1) % and (8 ± 1) %, respectively. The room-temperature strain hardening 188 behavior of the alloys is plotted in Figure 5b as Kocks-Mecking plots by showing true strain hardening 189 d𝜎t/d𝜀t as a function of true stress 𝜎t [48]. If there are no other crack-initiating mechanisms active, the 190 corresponding strain, where d𝜎t/d𝜀t= 𝜎t is fulfilled, marks the upper limit of strain that the material 191 can achieve via uniform plastic deformation in tension as the engineering relevant loading condition. 192 Thus, compression tests serve as a first, relevant screening criterion for potential tensile ductility. The 193 alloys with < 50 at.% Al have the potential to uniformly deform up to minimum 15 % true strain. In 194 contrast, the alloys with 50 and 53 at.% Al suffer from low strain hardening, showing a similar behavior 195 8 to what has been published in Ref. [49]. The strain for Fe-47Al, where the d𝜎t/d𝜀t= 𝜎t criterion is met, 196 is (15 ± 1) % true strain (marked with a red dot in the figure). Fe-50Al and Fe-53Al fulfill the d𝜎t/d𝜀t= 197 𝜎t criterion at true strains of (13 ± 1) % and (8 ± 1) %, respectively. 198 Figure 5:a) Representative (selected) true stress – true strain curves for the alloys measured at RT after heat treatment at 400 °C for 120 h. Crosses mark tests with failed samples. The other tests were interrupted at 𝜀𝑡=(15 ± 1) %, indicated by arrows. The initial engineering strain rate 𝜀˙ was set to 10−4 𝑠−1. b) Kocks-Mecking plots derived from the tests shown in a). Red dots mark where 𝜎𝑡=𝑑 𝜎𝑡 𝑑 𝜀𝑡 is met. The RT 1 % and 5 % offset yield strengths (Rp1 and Rp5) were determined for all alloys and are 199 displayed in Figure 6a. Rp1 was used to avoid scatter by localization of plastic deformation in the early 200 stages of compression [50]. 201 Given the extent of plastic deformation and strain hardening during hardness measurements, a direct 202 transformation into the initial yield strength is not possible. While the flow stress at 8 % plastic strain 203 can be employed for correlating hardness and strength [51], Fe-53Al exhibited premature failure prior 204 to reaching this strain level. Consequently, the flow stress at 5% true strain was utilized in Figure 6b to 205 remain consistent among all alloys [52]. The resulting correlation between Rp5 and HV (both for 400 °C 206 for 120 h) is Rp5 = 0.26 ∙𝐻𝑉, (adjusted coefficient of confidence 𝑅adj 2= 0.99) and the correlation 207 between Rp5 and NH is Rp5 = 0.20 ∙ 𝐻 (𝑅adj 2= 0.98). Both correlations are shown in the 208 Supplementary Material in Figures S4 and S5. 209 9 Figure 6: a) 𝑅𝑝1 and 𝑅𝑝5 at room-temperature as function of composition (closed squares for 𝑅𝑝1, open squares for 𝑅𝑝5). b) Comparison of true strain hardening 𝑑 𝜎𝑡 𝑑 𝜀𝑡 as slope between 𝜀𝑡= 1 % and 1.5 % plastic strain. The black squares are data taken from Ref. [53]. Error bars are smaller than the symbol size. Trendlines are shown in both figures as guidelines to the eyes. The strain hardening coefficient for each alloy has been determined in this work as slopes taken from 210 the true stress – true strain curves between 1 % and 1.5 % plastic strain to remain consistent with 211 Ref. [53]. A steady increase with increasing Al-content is found between 30 and 50 at.% Al, whereas a 212 steep increase is found at 53 at.% Al. A detailed discussion of this is provided in Section 4. 213 For the investigation of the high temperature behavior of the alloys, compression tests have been carried 214 out up to 700 °C. Representative true stress – true strain curves for Fe-35Al and Fe-50Al are shown in 215 Figure 7a and Figure 8a alongside Kocks-Mecking plots; for the other investigated alloys, the reader is 216 referred to the Supplementary Material (Figures S6-S9). 217 Figure 7: a) Representative (selected) true stress – true strain curves for Fe-35Al at different temperatures. The tests were interrupted at 𝜀𝑡=(15 ± 1) %, indicated by arrows. The initial engineering strain rate 𝜀˙ was set to 10−4 𝑠−1. b) KocksMecking plots for the Fe-35Al deformed at different temperatures. The stress-strain curves of Fe-35Al reveal strain hardening at temperatures up to 500 °C, which is 218 representative for the alloys up to 47 at.% Al. At 600 °C, a maximum in the stress-strain curve is 219 observed, indicating the onset of dynamic recovery or recrystallization [54]. Section 3.3 provides the 220 supporting evidence. At 700 °C, the curve flattens and the alloy does not show any strain hardening. 221 The strain where d𝜎t/d𝜀t= 𝜎t is fulfilled is not reached for the alloy deformed at temperatures below 222 500 °C, but it is reached at a true strain of (13 ± 1) % for deformation at 500 °C and at (3 ± 1) % for 223 deformation at 600 °C and 700 °C, respectively. The red dots in Figure 7b mark the points where the 224 16 onset of plasticity is expected to resemble similar trends as strength. This is indeed the case for the 368 dataset generated in this study when comparing Figure 6b and a with the most pronounced changes for 369 > 42 at.% Al. 370 The post-deformation microstructure observed in the present study, including the formation of kink 371 bands, has previously not been reported in literature for iron aluminides. Kink bands represent a 372 localization of plastic deformation and are observed at temperatures up to 700 °C. Apart from dynamic 373 recovery the appearance of kink bands might contribute to the continuously decreasing strain hardening 374 capability of iron aluminides with increasing temperature. However, deformation bands in general are 375 reported to disappear at elevated temperatures as more slip systems become activated. This enhances 376 homogeneous deformation and localized plastic deformation becomes less [54,59–61]. This is not the 377 case for iron aluminides in the tested composition and temperatures up to 700 °C. 378 After the deformation at 700 °C, the onset of dynamic recrystallization was observed for alloys with 379 < 47 at.% Al, even though these alloys possess lower homologous test temperatures than the ones with 380 higher Al-content. An entirely recrystallized microstructure is expected for larger total strains as 381 reported in Ref. [54]. The absence of dynamic recrystallization for alloys with higher Al contents might 382 be rationalized by the increasing vacancy concentration in 𝐵2-ordered FeAl [45] and with this, the 383 increasing probability for dynamic recovery by dislocation annihilation rather than recrystallization. 384 Another possible origin is the reduction in grain boundary mobility and therefore retarding 385 recrystallization when atomic ordering becomes more pronounced [54]. As dynamic recrystallization is 386 dependent on total strain and strain rate [54], either of both might be too low to achieve dynamic 387 recrystallization at the temperatures tested. 388 5. Conclusion Based on the results gained in this study, the following conclusions are drawn: 389 1. After the point defect-reducing heat treatment at 400 °C for 120 h, no YS peak was found in the 390 present study for alloys with > 35 at.% Al. In the current study, a well-defined dwell time at the 391 testing temperature was used, leading to different thermal vacancy concentrations prior to 392 starting the test. These concentrations vary depending on Al-concentration and on temperature 393 due to the exponential temperature dependence of the kinetics. Previous reports on the YS peak 394 in literature for higher Al-contents can be explained by several factors, ranging from chemical 395 inhomogeneity, the specific (thermo-)mechanical treatments and the distinct differentiation of 396 σf and σy. 397 2. Kink band formation is the cause for the decreasing strain hardening capability with increasing 398 temperatures in iron aluminides. In general, deformation bands are found to occur at angles 399 close to 45° with respect to the compression direction [59]. The appearances of kink bands found 400 in the present study differ significantly from this angle, which is indicative of a superposition 401 of kink band formation and uniform rotation of grains. Alloys with ≥ 47 at.% Al show low grain 402 boundary strength and are not suitable for being subjected to tensile loads to probe the onset of 403 plastic flow σy. 404 Acknowledgements We gratefully acknowledge financial support by the Deutsche Forschungsgemeinschaft (DFG) within 405 the framework of the project no. 511095365 (HE1872-44/1 and KA4631-4/1 at KIT, STE1077-3/1 and 406 BE7628/1-1 at MPI SusMat). This work was partly carried out with the support of the Karlsruhe Nano 407 Micro Facility (KNMFi, www.knmf.kit.edu), a Helmholtz Research Infrastructure at Karlsruhe Institute 408 17 of Technology (KIT, www.kit.edu). We acknowledge the chemical analysis by ICP-OES and CGHE at 409 the Institute for Applied Materials (IAM-AWP) by Dr. Bergfeldt, Karlsruhe Institute of Technology 410 (KIT) and support by Dr. Anwesha Kanjilal (MPI SusMat) within the framework of this project. 411 Conflict of Interest The authors declare no conflict of interest. 412 Data Availability Statement The data presented in this study are available in Zenodo at https://doi.org/10.5281/zenodo.17510975, 413 https://doi.org/10.5281/zenodo.17512156 and https://doi.org/10.5281/zenodo.17517632 under CC BY414 SA 4.0 license. Further information is available upon request with [email protected]. 415 Author Contributions Jan Lars Riedel: Investigation, Formal analysis, Software, Data Curation, Visualization, Project administration, Writing - Original Draft, Writing - Review & Editing Alexander Kauffmann: Conceptualization, Investigation, Methodology, Formal analysis, Data Curation, Visualization, Supervision, Funding acquisition, Writing - Original Draft, Writing - Review & Editing Stefan Guth: Investigation, Methodology, Formal analysis, Visualization, Supervision, Writing - Original Draft, Writing - Review & Editing Marcel Münch: Methodology, Formal analysis, Visualization, Writing - Review & Editing Georg Winkens: Formal analysis, Writing - Review & Editing Daniel Schliephake: Formal analysis, Writing - Review & Editing Jung Soo Lee: Investigation, Data Curation, Formal analysis, Writing - Review & Editing James P. Best: Supervision, Resources, Methodology, Funding acquisition, Writing - Review & Editing Frank Stein: Supervision, Resources, Funding acquisition, Writing - Review & Editing Martin Heilmaier: Conceptualization, Supervision, Resources, Funding acquisition, Writing - Review & Editing References 416 [1] R.R. Judkins, U.S. Rao, Fossil energy applications of intermetallic alloys, Intermetallics 8 (2000) 417 1347–1354. https://doi.org/10.1016/S0966-9795(00)00110-2. 418 [2] M. Palm, F. Stein, G. Dehm, Iron Aluminides, Annual Review of Materials Research 49 (2019) 419 297–326. https://doi.org/10.1146/annurev-matsci-070218-125911. 420 [3] D.G. Morris, M.A. Muñoz-Morris, High temperature mechanical properties of iron aluminides, 421 Revista de Metalurgia 37 (2001) 230–239. https://doi.org/10.3989/revmetalm.2001.v37.i2.471. 422 [4] M. Zamanzade, A. Barnoush, C. Motz, A Review on the Properties of Iron Aluminide 423 Intermetallics, Crystals 6 (2016) 10. https://doi.org/10.3390/cryst6010010. 424 [5] M. Palm, Fe–Al materials for structural applications at high temperatures: Current research at 425 MPIE, International Journal of Materials Research 100 (2009) 277–287. 426 https://doi.org/10.3139/146.110056. 427 [6] S.C. Deevi, Advanced intermetallic iron aluminide coatings for high temperature applications, 428 Progress in Materials Science 118 (2021) 100769. https://doi.org/10.1016/j.pmatsci.2020.100769. 429 18 [7] I. Baker, P.R. Munroe, Mechanical properties of FeAl, International Materials Reviews 42 (1997) 430 181–205. https://doi.org/10.1179/imr.1997.42.5.181. 431 [8] D. Risanti, J. Deges, L. Falat, S. Kobayashi, J. Konrad, M. Palm, B. Pöter, A. Schneider, C. 432 Stallybrass, F. Stein, Dependence of the brittle-to-ductile transition temperature (BDTT) on the Al 433 content of Fe–Al alloys, Intermetallics 13 (2005) 1337–1342. 434 https://doi.org/10.1016/j.intermet.2005.02.007. 435 [9] E.P. George, I. Baker, A model for the yield strength anomaly of Fe-Al, Philosophical Magazine 436 A 77 (1998) 737–750. https://doi.org/10.1080/01418619808224080. 437 [10] E.P. George, I. Baker, Thermal vacancies and the yield anomaly of FeAl, Intermetallics 6 (1998) 438 759–763. https://doi.org/10.1016/S0966-9795(98)00063-6. 439 [11] D. Rodney, J. Bonneville, Dislocations, in: Physical Metallurgy, Elsevier Amsterdam, 2014: pp. 440 1591–1680. https://doi.org/10.1016/B978-0-444-53770-6.00016-2. 441 [12] R.W. Cahn, P. Haasen, eds., Physical Metallurgy Vol. 3, 4th ed., Elsevier Science B.V., 442 Amsterdam, 1996. 443 [13] H. Xiao, I. Baker, The temperature dependence of the flow and fracture of Fe-40Al, Scripta 444 Metallurgica et Materialia 28 (1993) 1411–1416. https://doi.org/10.1016/0956-716X(93)90491445 A. 446 [14] U. Reimann, G. Sauthoff, Effects of deviations from stoichiometry on the deformation behaviour 447 of FeAl at intermediate temperatures, Intermetallics 7 (1999) 437–445. 448 https://doi.org/10.1016/S0966-9795(98)00104-6. 449 [15] I. Baker, H. Xiao, O. Klein, C. Nelson, J.D. Whittenberger, The effect of temperature and Fe: Al 450 ratio on the flow and fracture of FeAl, Acta Metallurgica et Materialia 43 (1995) 1723–1730. 451 https://doi.org/10.1016/0956-7151(94)00369-S. 452 [16] K. Yoshimi, S. Hanada, H. Tokuno, Effect of Frozen-in Vacancies on Hardness and Tensile 453 Properties of Polycrystalline B2 FeAl, Materials Transactions, JIM 35 (1994) 51–57. 454 https://doi.org/10.2320/matertrans1989.35.51. 455 [17] P. Nagpal, I. Baker, Effect of cooling rate on hardness of FeAl and NiAl, Metallurgical 456 Transactions A 21 (1990) 2281–2282. https://doi.org/10.1007/BF02647891. 457 [18] K.-M. Chang, Tensile and impact properties of directionally solidified Fe-40Al intermetallic, 458 Metallurgical Transactions A 21 (1990) 3027–3028. https://doi.org/10.1007/BF02647223. 459 [19] M.A. Morris, O. George, D.G. Morris, Vacancies, vacancy aggregates and hardening in FeAl, 460 Materials Science and Engineering: A 258 (1998) 99–107. https://doi.org/10.1016/S0921461 5093(98)00922-8. 462 [20] M. Zhao, K. Yoshimi, K. Maruyama, K. Yubuta, Thermal vacancy behavior analysis through 463 thermal expansion, lattice parameter and elastic modulus measurements of B2-type FeAl, Acta 464 Materialia 64 (2014) 382–390. https://doi.org/10.1016/j.actamat.2013.10.051. 465 [21] E. Orowan, Fracture and strength of solids, Reports on Progress in Physics 12 (1949) 185–232. 466 https://doi.org/10.1088/0034-4885/12/1/309. 467 [22] C.T. Liu, C.L. Fu, E.P. George, G.S. Painter, Environmental Embrittlement in FeAl Aluminides, 468 ISIJ International 31 (1991) 1192–1200. https://doi.org/10.2355/isijinternational.31.1192. 469 [23] L.M. Pike, C.T. Liu, Environmental and Strain Rate Effects on the Ductility and Yield Strength of 470 Fe-40Al, Scripta Materialia 38 (1998) 1475–1480. https://doi.org/10.1016/S1359-6462(98)00087471 6. 472 [24] C.T. Liu, Recent advances in ordered intermetallics, Materials Chemistry and Physics 42 (1995) 473 77–86. https://doi.org/10.1016/0254-0584(95)01546-9. 474 [25] I. Baker, D.J. Gaydosh, Flow and fracture of FeAl, Materials Science and Engineering 96 (1987) 475 147–158. https://doi.org/10.1016/0025-5416(87)90549-0. 476 [26] M.A. Crimp, K. Vedula, Effect of boron on the tensile properties of B2 FeAl, Materials Science 477 and Engineering 78 (1986) 193–200. https://doi.org/10.1016/0025-5416(86)90323-X. 478 [27] D.J. Gaydosh, S.L. Draper, R.D. Noebe, M.V. Nathal, Room temperature flow and fracture of Fe479 40at.%Al alloys, Materials Science and Engineering: A 150 (1992) 7–20. 480 https://doi.org/10.1016/0921-5093(90)90003-L. 481 [28] G. Hénaff, A. Tonneau, Environment-sensitive fracture of iron aluminides under monotonic 482 tensile loading, Metall Mater Trans A 32 (2001) 557–567. https://doi.org/10.1007/s11661-001483 0072-3. 484 19 [29] O. Klein, I. Baker, The effect of boron on the temperature dependence of the flow and fracture of 485 Fe45Al, Scripta Metallurgica et Materialia 30 (1994) 1413–1417. https://doi.org/10.1016/0956486 716X(94)90238-0. 487 [30] D.G. Morris, J. Chao, C.G. Oca, M.A. Muñoz-Morris, Obtaining good ductility in an FeAl 488 intermetallic, Materials Science and Engineering: A 339 (2003) 232–240. 489 https://doi.org/10.1016/S0921-5093(02)00108-9. 490 [31] P. Nagpal, I. Baker, The effect of strain rate on the room-temperature ductility of FeAl, Scripta 491 Metallurgica et Materialia 25 (1991) 2577–2580. https://doi.org/10.1016/0956-716X(91)90071-8. 492 [32] D. Wu, I. Baker, The effect of environment and strain rate on the room temperature tensile 493 properties of FeAl single crystals, Intermetallics 9 (2001) 57–65. https://doi.org/10.1016/S0966494 9795(00)00097-2. 495 [33] Y. Yang, I. Baker, E.P. George, Effect of Vacancies on the Tensile Properties of Fe-40Al Single 496 Crystals in Air and Vacuum, Materials Characterization 42 (1999) 161–167. 497 https://doi.org/10.1016/S1044-5803(98)00049-7. 498 [34] K. Yoshimi, S. Hanada, Positive temperature dependence of yield stress in B2 FeAl, in: 499 Proceedings of the First International Symposium on Structural Intermetallics, Publ by Minerals, 500 Metals & Materials Soc (TMS), Champion, PA, USA, 1993: pp. 475–482. 501 [35] K. Yoshimi, S. Hanada, M.H. Yoo, Tensile properties of B2-type Fe-39mol%Al single crystals at 502 elevated temperatures, Intermetallics 4 (1996) 159–168. https://doi.org/10.1016/0966503 9795(95)00032-1. 504 [36] K. Yoshimi, S. Hanada, M.H. Yoo, Yield Stress Anomaly in B2 FeAl, MRS Proceedings 460 505 (1996) 313. https://doi.org/10.1557/PROC-460-313. 506 [37] J.W. Cohron, Y. Lin, R.H. Zee, E.P. George, Room-temperature mechanical behavior of FeAl: 507 effects of stoichiometry, environment, and boron addition, Acta Materialia 46 (1998) 6245–6256. 508 https://doi.org/10.1016/S1359-6454(98)00254-7. 509 [38] O. Klein, I. Baker, Effect of heat-treatment on the tensile behavior of iron-rich FeAl and FeAl + 510 B, Scripta Metallurgica et Materialia 30 (1994) 627–632. https://doi.org/10.1016/0956511 716X(94)90441-3. 512 [39] P. Nagpal, I. Baker, Room temperature fracture of FeAl and NiAl, Materials Characterization 27 513 (1991) 167–173. https://doi.org/10.1016/1044-5803(91)90060-H. 514 [40] R. Kerl, J. Wolff, Th. Hehenkamp, Equilibrium vacancy concentrations in FeAl and FeSi 515 investigated with an absolute technique, Intermetallics 7 (1999) 301–308. 516 https://doi.org/10.1016/S0966-9795(98)00118-6. 517 [41] F. Bachmann, R. Hielscher, H. Schaeben, Texture Analysis with MTEX – Free and Open Source 518 Software Toolbox, Solid State Phenomena 160 (2010) 63–68. 519 https://doi.org/10.4028/www.scientific.net/SSP.160.63. 520 [42] DIN EN ISO 6507-1:2024-01, Metallische Werkstoffe_Härteprüfung nach Vickers_Teil_1: 521 Prüfverfahren (ISO_6507-1:2023); Deutsche Fassung EN_ISO_6507-1:2023, (n.d.). 522 https://doi.org/10.31030/3503781. 523 [43] W.C. Oliver, G.M. Pharr, An improved technique for determining hardness and elastic modulus 524 using load and displacement sensing indentation experiments, Journal of Materials Research 7 525 (1992) 1564–1583. https://doi.org/10.1557/JMR.1992.1564. 526 [44] W.C. Oliver, G.M. Pharr, Measurement of hardness and elastic modulus by instrumented 527 indentation: Advances in understanding and refinements to methodology, Journal of Materials 528 Research 19 (2004) 3–20. https://doi.org/10.1557/jmr.2004.19.1.3. 529 [45] Y.A. Chang, L.M. Pike, C.T. Liu, A.R. Bilbrey, D.S. Stone, Correlation of the hardness and 530 vacancy concentration in FeAl, Intermetallics 1 (1993) 107–115. https://doi.org/10.1016/0966531 9795(93)90028-T. 532 [46] M. Kogachi, T. Haraguchi, Quenched-in vacancies in B2-structured intermetallic compound FeAl, 533 Materials Science and Engineering: A 230 (1997) 124–131. https://doi.org/10.1016/S0921534 5093(97)00016-6. 535 [47] W.D. Nix, H. Gao, Indentation size effects in crystalline materials: A law for strain gradient 536 plasticity, Journal of the Mechanics and Physics of Solids 46 (1998) 411–425. 537 https://doi.org/10.1016/S0022-5096(97)00086-0. 538 [48] H. Mecking, U.F. Kocks, Kinetics of flow and strain-hardening, Acta Metallurgica 29 (1981) 539 1865–1875. https://doi.org/10.1016/0001-6160(81)90112-7. 540 20 [49] Y. Yang, I. Baker, The influence of vacancy concentration on the mechanical behavior of Fe-40Al, 541 Intermetallics 6 (1998) 167–175. https://doi.org/10.1016/S0966-9795(97)00062-9. 542 [50] S. Nizamoglu, A. Gedsun, Kauffmann, A., M. Ghosh, H. Michels, T. Groten, C. Schulz, C. 543 Breuner, S. Seils, D. Schliephake, S. Laube, M. Palm, M. Heilmaier, Microstructure and 544 Mechanical Properties of a Precipitation-Strengthened Fe–Al–Nb Alloy, Advanced Engineering 545 Materials (2023) 2300148. 546 [51] D. Tabor, The hardness of solids, Reviews of Physics in Technology 1 (1970) 145–179. 547 https://doi.org/10.1088/0034-6683/1/3/I01. 548 [52] G. Winkens, A. Kauffmann, J. Herrmann, A.K. Czerny, S. Obert, S. Seils, T. Boll, C. Baruffi, Y. 549 Rao, W.A. Curtin, R. Schwaiger, M. Heilmaier, The influence of lattice misfit on screw and edge 550 dislocation-controlled solid solution strengthening in Mo-Ti alloys, Communications Materials 4 551 (2023) 26. https://doi.org/10.1038/s43246-023-00353-8. 552 [53] D.G. Morris, M.A. Muñoz-Morris, L.M. Requejo, Work hardening in Fe–Al alloys, Materials 553 Science and Engineering: A 460–461 (2007) 163–173. 554 https://doi.org/10.1016/j.msea.2007.01.014. 555 [54] F.J. Humphreys, M. Hatherly, Recrystallization and related annealing phenomena, 2nd ed., 556 Elsevier, Amsterdam, 2004. 557 [55] D.B. Miracle, O.N. Senkov, C. Frey, S. Rao, T.M. Pollock, Strength vs temperature for refractory 558 complex concentrated alloys (RCCAs): A critical comparison with refractory BCC elements and 559 dilute alloys, Acta Materialia 266 (2024) 119692. https://doi.org/10.1016/j.actamat.2024.119692. 560 [56] A. Schmitt, K.S. Kumar, A. Kauffmann, M. Heilmaier, Microstructural evolution during creep of 561 lamellar eutectoid and off-eutectoid FeAl/FeAl2 alloys, Intermetallics 107 (2019) 116–125. 562 https://doi.org/10.1016/j.intermet.2019.01.015. 563 [57] H. Chen, T. Hanemann, S. Seils, D. Schliephake, A.S. Tirunilai, M. Heilmaier, K.-P. Weiss, A. 564 Kauffmann, Influence of Temperature and Plastic Strain on Deformation Mechanisms and Kink 565 Band Formation in Homogenized HfNbTaTiZr, Crystals 11 (2021) 81. 566 https://doi.org/10.3390/cryst11020081. 567 [58] G. Gottstein, Physical Foundations of Materials Science, Springer Berlin Heidelberg, Berlin, 568 Heidelberg, 2004. https://doi.org/10.1007/978-3-662-09291-0. 569 [59] D. Kuhlmann-Wilsdorf, OVERVIEW No. 131 “Regular” deformation bands (DBs) and the LEDS 570 hypothesis, Acta Materialia 47 (1999) 1697–1712. https://doi.org/10.1016/S1359-6454(98)00413571 3. 572 [60] M.R. Drury, F.J. Humphreys, The development of microstructure in Al-5% Mg during high 573 temperature deformation, Acta Metallurgica 34 (1986) 2259–2271. https://doi.org/10.1016/0001574 6160(86)90171-9. 575 [61] F.J. Humphreys, P.B. Prangnell, J.R. Bowen, A. Gholinia, C. Harris, Developing stable fine-grain 576 microstructures by large strain deformation, Philosophical Transactions: Mathematical, Physical 577 and Engineering Sciences 357 (1999) 1663–1681. https://doi.org/10.1098/rsta.1999.0395. 578 579