High temperature thermal and mechanical stability of high-strength nanotwinned Al alloys

Accepted Manuscript High temperature thermal and mechanical stability of high-strength nanotwinned Al alloys Qiang Li, Jaehun Cho, S. Xue, X. Sun, Y.F. Zhang, Z. Shang, H. Wang, X. Zhang PII:

S1359-6454(18)30887-5

DOI:

https://doi.org/10.1016/j.actamat.2018.11.011

Reference:

AM 14954

To appear in:

Acta Materialia

Received Date: 30 June 2018 Revised Date:

29 September 2018

Accepted Date: 5 November 2018

Please cite this article as: Q. Li, J. Cho, S. Xue, X. Sun, Y.F. Zhang, Z. Shang, H. Wang, X. Zhang, High temperature thermal and mechanical stability of high-strength nanotwinned Al alloys, Acta Materialia, https://doi.org/10.1016/j.actamat.2018.11.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

High temperature thermal and mechanical stability of high-strength nanotwinned Al alloys

RI PT

Qiang Lia, Jaehun Choa, S. Xuea, X. Suna, Y. F. Zhanga, Z. Shanga, H. Wanga,b and X. Zhanga,* a

School of Materials Engineering, Purdue University, West Lafayette, IN 47907

b

School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN 47907

SC

Corresponding author: X. Zhang, [email protected]

M AN U

Abstract

Al alloys have widespread applications but often suffer from low yield strength. Recently, nanotwinned Al-Fe solid solution alloys have shown high flow stress (> 1.5 GPa), ascribed to nanograins with abundant incoherent twin boundaries and solute-stabilized 9R phase. However, the high temperature mechanical behaviors of high-strength twinned Al-Fe alloys remain

TE D

unknown. In this study, we show that nanotwinned microstructures are stable up to 280 ˚C, followed by recrystallization at 300 ˚C. In-situ uniaxial compression tests in a scanning electron

EP

microscope show that the nanotwinned Al alloys retain their high flow stress of ~1.3 GPa when tested up to 200 ˚C, and substantial softening occurs at a test temperature of 300 ˚C. This work

AC C

reveals the superb thermal stability and high temperature mechanical behaviors of the nanotwinned Al-Fe alloys, and offers a new perspective to design future strong and thermally stable nanostructured Al alloys. Keywords: Nanotwins, Al alloys, thermal stability, high-strength, nanomechanical testing

1

ACCEPTED MANUSCRIPT

1. Introduction The significant interest in development of high-strength, ductile Al alloys is driven by the

RI PT

industrial needs for high strength, fuel efficient materials. Al alloys, however, have been outdistanced by the development of advanced high-strength steels with flow stress of 1-2 GPa [1, 2]. A majority of commercial Al alloys have yield strength of ~ 700 MPa or less [3, 4]. Grain refinement has been proven effective in strengthening Al alloys. However, the strength of

SC

nanostructured Al alloys rarely exceeds 1 GPa. Meanwhile it has been shown that when grain

M AN U

sizes of nanocrystalline (nc) metals reduce to certain values, softening, as opposed to Hall-Petch strengthening effect, often occurs [5-7]. In addition, nc metals often have diminished tensile ductility due to limited dislocation activities in nanograins [8-10].

Certain nanotwinned (nt) metals with an average twin spacing of 100 nm or less have a

TE D

combination of high strength, ductility and good thermal stability [11-15]. Most prior studies of nt metals focused on systems with low stacking fault energy (SFE), such as Cu, Ag, and stainless steels [11, 16-18]. In comparison, twin formation in Al is very difficult due to its high SFE (γsf =

EP

120–144 mJ m-2), the high ratio of SFE over unstable SFE (γsf/γusf) [15], and a large disparity between perfect and twin nuclei size during physical vapor deposition (PVD) [17, 19]. There are

AC C

limited successes in introducing twin boundaries into Al via thin film growth [20-22], quasistatic and high strain rate deformation [23-26], or annealing [27]. Recent studies show that Fe solutes promote the formation of high-density twin boundaries and 9R phase in Al-Fe solid solution alloys with flow stress exceeding 1.5 GPa [28]. However, the thermal stability and high temperature mechanical behavior of these high-strength nt Al-Fe alloys remain largely unknown. Al alloys with strength > 1 GPa have been reported in Al-TM (transition metal)-RE (rare earth metal)-based amorphous/nanocrystal composites [29], and complex alloys of Al84Ni7Gd6Co3 2

ACCEPTED MANUSCRIPT

with > 50 vol.% of intermetallics [30]. In parallel to the scientific findings, high-strength and multifunctional metallic materials prepared by far-from-equilibrium processes, such as electrodeposition and PVD, have been adopted in applications, including aluminide (β-AlNi)

RI PT

coatings on turbine blades [31], Al alloy-based spatial light regulator [32], and micro- and nanoelectromechanical systems (MEMS/NEMS) that require nanometer/atomic-level

smoothness. For instance, amorphous/nc Al-Mo thin films with > 30 at.% Mo for MEMS/NEMS

SC

resonators had a hardness > 5 GPa [33, 34] and smooth surfaces.

M AN U

Various Al-Fe alloys have been synthesized with a broad range of Fe compositions (0-40 at.%) using both equilibrium and far-from-equilibrium approaches. Equilibrium processing routes, such as casting, mostly generated sizable equilibrium phase, i.e. Al13Fe4, that has insignificant strengthening effect and causes steep decline in ductility [35, 36]. Some far-fromequilibrium processing methods, such as mechanical alloying [37-39] and e-beam evaporations

TE D

[40], would produce nc alloys with non-equilibrium phase, such as Al6Fe and Al5Fe2 [41], and excessive Fe addition would lead to amorphous phase [41, 42]. The coarse-grained and nc Al-Fe

EP

alloys investigated previously often have a mechanical strength of ~ 0.2 to 1 GPa. Many nc metals have poor thermal stability as the excess free energy of GBs serves as the

AC C

driving force for grain coarsening. A series of ultra-fine grained (ufg) Al alloys processed through severe plastic deformation experienced softening after heat treatment to less than ~ 200 ˚C due to dislocation recovery, grain growth and/or dissociation of Guinier–Preston (G–P) zones [43-45]. Rapidly solidified and cryomilled Al alloy powders also went through grain coarsening during hot extrusion [46, 47]. With the assistance of impurity, two routes have been proven effective to stabilize nc metals, including thermodynamic (GB energy relaxation) and kinetic (Zener pinning) stabilizations. The thermodynamic approach relaxes excess GB energy via GB 3

ACCEPTED MANUSCRIPT

segregation, and has been applied in nc W-Ti [48], Fe-Zr [49], Cu-Ta [50], Ni-P [51, 52], Ni-W [53], Ni-Mo [54] and Al-Pb alloys [55]. The kinetic approach uses the pinning effect of particulate second phase precipitation to restrict GB migration. Al3Sc and Al3Zr precipitates have

RI PT

been applied to kinetically stabilize grains in ufg, nc Al and Al-Mg alloys [56-58].

In this study, we report on the thermal stability and high temperature mechanical

behavior of high-strength nt Al-Fe solid solution alloys. The hardness of Al-Fe alloys remains

SC

exceeding 5 GPa when the annealing temperature is < 280 ˚C. Recrystallization occurs when

M AN U

annealing was performed at 300 ˚C. In situ micropillar compression tests in a scanning electron microscope show that the flow stress of Al-Fe alloys remains greater than 1 GPa when tested at 200 oC, and significant softening takes place at 300 ˚C. Extensive transmission electron microscopy (TEM) studies were performed to investigate the temperature dependent evolution of

TE D

microstructures. 2. Experimental techniques

Al-xFe (x=1–10 at.%) alloys were co-sputtered by using Al (99.999%) and Fe (99.98%)

EP

targets onto HF etched Si (111) wafers in an AJA ATC-2200-UHV sputtering system with a base pressure of 0.5–1 × 10-8 torr. Ar pressure was kept at 2 mtorr during sputtering. The sputtered

AC C

films have a thickness of 3–4 µm. Annealing of the films were performed in the same AJA chamber at 100 - 400 ˚C for 1 hour with a ramping rate of ~ 20 ˚C/min. X-ray diffraction (XRD) profiles were acquired on a Panalytical Empyrean X’pert PRO

MRD diffractometer with a 2 × Ge (220) hybrid monochromator to select Cu Kα1. TEM specimens were prepared by mechanically grinding, dimpling and low-energy Ar ion milling using a Gatan ion milling machine. An FEI quanta 3D FEG dual beam SEM/FIB microscope

4

ACCEPTED MANUSCRIPT

equipped with an Omniprobe manipulator was used to lift out and thin cross-section TEM samples (2kV low energy Ga+ ion beam polishing was eventually carried out to minimize ion damage). TEM and EDS experiments were conducted in an FEI Talos 200X microscopes

RI PT

operated at 200 kV, equipped with Fischione ultra-high resolution high angle annular dark field (HADDF) detectors and supper X electron-dispersive x-ray spectroscopy (EDS) detectors. Hardness and reduced moduli of the Al-Fe alloys were measured on a Hysitron TI

SC

premier with a diamond Berkovich indenter. The maximum indentation depth was kept below ~

M AN U

15% of the sample thickness to avoid substrate effect. At least 20 indents were conducted on each sample. The elastic modulus was deduced from the reduced modulus, Er, collected in nanoindentation measurements, using the equation:




=


  

+


  

(1)

TE D

where E and v stand for elastic modulus and Poisson’s ratio, respectively, and their subscript i and s denote indenter and sample, respectively. Prior to in-situ compression experiments, micropillars with a diameter-to-height aspect ratio of 1:3 – 1:2 were fabricated by FIB and a

EP

series of concentric annular milling/polishing with progressively de-escalated currents were adopted to minimize the tapering effect. The fabrication parameters were carefully chosen to

AC C

avoid any involvement of Si substrate as part of the micropillars. In situ compression tests were performed up to 300 ˚C inside the FEI quanta 3D FEG dual beam SEM/FIB microscope on micropillars by using a Hysitron PI 88×R PicoIndenter with a 5 µm diamond flat punch tip mounted to a piezoelectric actuator on the capacitive transducer. Moreover, the mechanical response was synchronized with the spontaneous micropillar morphology evolution during deformation. The influence of elastic compliance from Si substrate and the diamond indenter on

5

ACCEPTED MANUSCRIPT

the displacement was estimated by taking the pressed elastic half-space into account. The valid displacement is defined by Sneddon equation as [59]:
 



( ) − 








( ) 

(2)

RI PT

= −

where um and F are the measured displacement and force, respectively. dt and db are the top and bottom diameter of pillars, and E denotes the elastic modulus. The subscript i and Si denote

SC

diamond indenter and Si substrate, respectively.

For high temperature compression tests, a ceramic probe heater and stage heater were

M AN U

used. The two heating terminals were heated concurrently at a rate of 10 ˚C/min to the designated temperatures, and held isothermally for a minimum of 30 min prior to compression experiments to minimize thermal and image drifts due to temperature disparity between the two terminals. An average drift rate of 0.2 – 0.7 nm/s was measured in advance within preloading

TE D

contacts and a 45 s preloading could compensate drift-associated displacement errors. A strain

average. 3. Results

EP

rate of 5 × 10-3/s was used throughout this study. The force noise was measured to be ± 5 µN on

AC C

3.1. Microstructural and chemical characterizations The XRD profiles of the as-deposited nt Al-5.5 at.% Fe (denoted as Al-5.5Fe hereafter)

and the samples annealed at different temperatures are shown in Fig. 1a. The as-deposited specimen shows a strong (111) texture. In spite of a slight (111) peak shift, the structure remains the same for samples annealed at less than ~ 300 ˚C. When the annealing temperature is ≥ 300 ˚C, a small broad peak between 2θ = 40˚ to 41˚ arises due to convolution of multiple diffractions of

6

ACCEPTED MANUSCRIPT

Al6Fe (Fig. 1b), consistent with the annealed Al-Fe alloys reported previously [37, 60]. It is worth mentioning that the annealed specimens remain (111) textured.

RI PT

The plan-view and cross-section TEM micrographs (Fig. 2a-b) of the as-deposited Al5.5Fe specimen show that the average columnar grain size is ~ 4 nm. The inserted selected area electron diffraction (SAED) patterns in Fig. 2a-b confirm the (111) texture of the film. The

SAED pattern of the cross-section TEM specimen examined along <110> zone axis indicates the

SC

formation of twins. The diffraction spots reflecting (111), (111) and (200) evolve to arc shape,

M AN U

suggesting that the specimens undergo slight out-of-plane rotation among columnar structures. Fig. 2c-f show the formation of abundant diffused twin boundaries, i.e. 9R phase, in columnar grains. Furthermore, no Fe segregation is observed (Fig. S1).

The microstructural evolution of Al-3Fe and Al-5.5Fe annealed from 100 to 400 ˚C for 1

TE D

h is similar and hence only the microstructures of the annealed Al-5.5Fe will be presented here. As shown in Fig. 3a1-a4, after annealing at 200 ˚C, the average columnar grain size is ~ 4 nm and the intermingled 9R and narrow twin columns are seemingly identical to that of the as-

EP

deposited samples. Within some columns, coherent TBs exist, and low-angle grain boundaries (LAGBs) with Frank partials are frequently observed (Fig. S2). At 280 ˚C, the precipitation first

AC C

took place from the free surface of the films, while the matrix remains fcc structure with 9R phases as shown in Fig. 3b1-b4, and EDS mapping showed that Fe-rich particles formed close to free surface (Fig. S3). HRTEM micrograph in Fig. 3b3 shows that that the precipitate has superlattice structure with interplanar spacing of ~0.46 nm, and Fig. 3b4 displays a classic 9R phase bounded by matrix and twin columns. Severe precipitation and recrystallization occurred at 300 ˚C, in good agreement with the XRD results. The intermetallic precipitates and resultant Fe-deficient fcc Al crystals are equiaxed grains with an average grain size of 46 ± 13 and 50 ± 18 7

ACCEPTED MANUSCRIPT

nm, respectively (Fig. 3c1-c3), and they alternate with no obvious structural hierarchy according to the STEM and EDS compositional maps in Fig. S4. The HRTEM image in Fig. 3c4 shows an Al6Fe precipitate with orthorhombic cmcm structure, examined along <001>, and their (220)

RI PT

diffraction spots superimposed on the fcc (111) spot. Meanwhile, excessive superlattice

structures, a majority of which are adjacent to the newly formed Al6Fe, are identified with an average interplanar spacing of 0.44 to 0.55 nm (Fig. 3c2, 3 and Fig. S5). Except for the

SC

superlattice, it is expected that various transition structures exist concurrently.

M AN U

The plan-view HADDF-STEM micrographs alongside EDS maps suggest that the Fe distribution remains homogeneous at 200 ˚C as shown in Fig. 4a, and a transformation took place at ~ 280 ˚C in Fig. 4b. Fe segregation accounts for the faceted cellular contours and the curly diffusion paths within cells as evidenced by Fig. 4b2 and b3. In the meantime, micron-size particles were protruded out of free surface as shown in Fig. S6, and the linear EDS scan of SEM

TE D

shows that these particles are Fe deficient (less than 3 %). TEM-EDS analysis was conducted to confirm the Fe concentration of ~ 1.5 % in protrusions (Fig. S7). The crystallite size has not undergone apparent increment after annealing at 300 - 400 ˚C, but secondary recrystallization

EP

made the fcc Al(Fe) solid solution further decomposed to tiny nanoprecipitates and the grains

AC C

became elongated at comparably high temperature (Fig. 4c and 4d). After 400 ˚C annealing, the Fe concentration in fcc Al-Fe grains mostly approaches zero according to compositional analyses (Fig. 4d), and the precipitated second phase is primarily Al6Fe. 3.2 Mechanical behaviors of annealed specimens tested at room temperature To unveil the real-time mechanical response of the Al-Fe alloys, systematic in situ micropillar compression tests have been conducted inside a scanning electron microscope (Movie S1-S4), as demonstrated in Fig. 5a. Using constant strain rate tests at 5 × 10-3/s at room 8

ACCEPTED MANUSCRIPT

temperature, the representative engineering stress-strain curves of Al-5.5Fe annealed at different temperatures are plotted in Fig. 5b alongside SEM snapshots of the deformed pillars at different strain levels (Fig. 5c-f). Note that no true stress-strain curves are presented due to the difference

RI PT

in spontaneous contact area variation among different samples. Two partial unloading segments were deliberately incorporated to verify the indenter-pillar alignment. The Al-5.5% Fe specimens after 100 and 200 ˚C annealing have a respective flow stress of ~ 1.67 and 1.63 GPa (when ε =

SC

5%), slightly lower compared to ~ 1.73 GPa for the as-deposited sample. A significant drop of

M AN U

flow stress to ~ 0.88 GPa (ε = 5%) occurred for the alloys annealed at 300 ˚C. 3.3 Temperature-dependent in situ compression experiments

Engineering stress-strain curves obtained from in-situ compressions at elevated temperature up to 300 ˚C are compared in Fig. 6 alongside the SEM snapshots at various strains

TE D

(Movie S5-S7). In Fig. 6a, results of compressions on pillars with diameters of 0.5 and 1.3 µm are overlaid to rule out the size effect on mechanical properties tested at room temperature. The flow stress (at ε = 5%) of the deposited Al-5.5Fe tested at room temperature, 100 and 200 ˚C is ~

EP

1.73, 1.5 and 1.3 GPa, respectively. A significant stress drop to 0.1-0.15 GPa was observed for the sample tested at 300 ˚C, indicating a drastic shift of deformation mechanism despite the

AC C

strength reduction related to microstructure coarsening. For pillars tested at 200 ˚C and below, the preferential dilation near pillar tops and subsequent downward propagation of deformation zone were observed. However, SEM snapshots in Fig. 6b-c reveal globules that were extruded from deformation zones for specimens tested at 100 and 200 ˚C. While the globules were sporadic at 100 ˚C, their quantity considerably increased as the testing temperature rose to 200 ˚C.

9

ACCEPTED MANUSCRIPT

4. Discussions 4.1. The influence of Fe solute on lattice parameters of Al alloys

RI PT

Fe has limited equilibrium solid solubility (~ 0.3 at.%) in Al at room temperatures [37]. At an equilibrium state, Fe often exists in the large needle- or rod-shaped intermetallic

precipitates in Al, and consequently has insignificant strengthening effect in Al alloys [35].

SC

What’s more, the ductility of Al-Fe alloys after casting steeply declines as a consequence of cracks at phase boundaries as Fe concentration exceeds 1 wt.% [61]. Hence, Fe is often

M AN U

considered as a harmful addition in the casted and wrought Al alloys. On the other hand, at a constant solute composition, Al-Fe has a lower liquidus temperature than Al-Sc, Al-Ti and Al-Cr, and thus a higher supersaturation of Fe solid solution can be accomplished in Al-Fe at a fixed quenching rate [62]. Prior studies also suggest that minute Fe solutes appear to significantly

TE D

strengthen Al, although the mechanism for such anomalous strengthening is unclear [63]. The probability of forming supersaturated Al-Fe alloys is enhanced by magnetron sputtering, a rapid quenching process. It has been shown that the ratio of solubility at real

EP

quenching rate, Cv, over the solubility at equilibrium state, C0, Cv/C0 of Fe in Al is 0.1-0.2 at 103 K/s, and increases to ~9.6 at 107 K/s [62]. Therefore, the formation of supersaturated Al-5.5Fe

AC C

solid solution alloys benefited from the high quenching rate (~106 – 1010 K/s) of sputtering process [64-66]. Amorphous phase is identified in as-deposited Al-9.9Fe (Fig. S8). In several supersaturated binary alloy systems, such as Al-Mo [34], Al-Mg [67] and Cu-

Ta [68], abundant solute segregation occurs at GBs to minimize free energy and excess doping could produce amorphous phase. Consequently, the lattice constant in grain interiors stays constant or has a tendency to remain constant at higher impurity concentrations. Intriguingly, the

10

ACCEPTED MANUSCRIPT

excess addition of Fe into Al increases lattice constant as Fe exceeds ~ 2.5 % despite Fe possessing comparably smaller atomic radius (rFe = 0.124 nm vs. rAl = 0.143). The lattice constants deduced from (111) peaks on XRD profiles would represent the average lattice

RI PT

constants since the (111) texture allows mostly identical magnitude of Cauchy stress tensors on three facets of the fcc cube. We adopted an elasticity inclusion model [69] that takes the

electronic interactions at outermost quantum shells of solute and solvent atoms into account to

#($%& $'( )
/   !" ) * $ )

(3)

'( %& '(

M AN U

  = ( + 4

SC

interpret the lattice constant variation with Fe concentration and the model is described as

where R is the atomic radius of solute or solvent, c is Fe concentration, and ϑ = 1, - = 4 for ./'(

cubic lattice. The variable " = 1 + modulus, 2 = (


3

56 7

+

3

56 7

0%& 4

)
−

./ 

./'(

0%&

, where the effective bulk

. µ is the effective shear modulus, and can be obtained

TE D

0'( 4

0

, that makes "1 = 1 +

by employing the Hill’s self-consistent method [70]:
3

+


3


4.//0%&

− 59


3

6 6%&




+

3

6 6'(




:+2=0

(4).

EP


4.//0'(

Fig. 7a shows that the lattice parameter variation of Al-Fe in the regime of cFe ≤ 2.5 %

AC C

agrees well with the elasticity inclusion model. Beyond 2.5 %, the addition of Fe raises the lattice parameter of Al-Fe alloy linearly. It was reported that severe plastic deformation was able to extend the Fe solubility in Al up to ~ 1 at.% [71]. Multiple prior studies suggest that mechanically alloyed nc Al-Fe alloys experience compositional fluctuation, and the excess Fe atoms above ~ 2.5 % transfer the alloy to an amorphous phase or lead to Al6Fe segregation [37, 60]. However, the sputtered Al-5.5Fe remains a solid solution alloy, and amorphous heterogeneity emerges in Al-9.9% Fe (Fig. S8). The excess Fe solutes (beyond 2.5%) may 11

ACCEPTED MANUSCRIPT

unusually occupy interstitial sites and thus lead to lattice expansion in sputtered Al-Fe alloys. Consequently, the lattice parameter of Al-Fe increases with Fe solute concentration.

RI PT

The influences of annealing temperature on the average lattice parameter and elastic modulus show two distinct regimes. Fig. 7b shows that the lattice parameter of Al-Fe increases from 0.4052 nm to 0.4057 nm upon annealing until 280 ˚C, and decreases prominently thereafter with some fluctuations up to 400 ˚C. The temperature for sudden decrease of lattice parameter

SC

coincides with the formation of precipitates, and thus it is likely that the reduction of lattice

M AN U

parameters results from precipitation that drives interstitial Fe solutes out of Al lattice. It is worth mentioning that the lattice parameter of sample annealed at 400 ˚C is 0.4035 nm, smaller than 0.4049 nm of pure Al, for which we speculate that the residual substitutional Fe solutes in the matrix and stress state between intermetallic and fcc grains may account.

TE D

4.2 The evolution of microstructures after annealing

The equilibrium width of a SF in Al is typically less than 1 nm [72]. 9R phase is a complex stacking fault with a period of 9 repeating atomic layers. Although 9R has been

EP

observed in metals with low SFEs, such as Ag [73], Au [74] and Cu [72, 75], it is rare in Al. DFT calculations show that Fe solute significantly increases the energy barrier for SF recovery.

AC C

Hence 9R phase, once formed by rapid quenching (sputtering), cannot recover as a consequence of large kinetic energy barriers [28]. The current study shows that 9R and nanotwinned columnar grains are stable up to an annealing temperature of 280 ˚C, ~ 0.6 Tm. In contrast, ultrafine grained Al alloys prepared by mechanical alloying or severe plastic deformation experienced grain growth with dislocation recovery from 100 to 230 ˚C [44, 45, 76]. It is known that the activation energy for grain growth of Al is comparably low (around 40 kJ/mol) at low temperature [77]. The good thermal stability of the 9R and nanotwins in nt Al-Fe may be 12

ACCEPTED MANUSCRIPT

attributed to several aspects. First, Fe solute stabilizes the extended ITBs, i.e. 9R. As discussed previously, Fe solute increases the energy barrier for the removal of stacking faults. Second, it has been shown the driving force for detwinning is proportional to 2"; /<, where "; is twin

RI PT

boundary energy and t is twin spacing [16]. Nt Al-Fe alloys with long ITBs resulted in less pronounced driving force for detwinning. Molecular dynamics (MD) simulations show that ITBs are thermodynamically stable, but some of other types of GBs dissociate into terraced interfaces

SC

[78]. Third, the highly {111} textured film contains abundant LAGBs, which are more stable than conventional high-angle GBs [79]. Our prior MD simulations show that the dissociated

M AN U

ITBs in nt Al-Fe alloys are stable at ambient condition, and partially survive after compression [28].

Fig. 7c (and Fig. S9) clearly marks two microstructural regimes upon annealing. The average columnar grains size of fcc matrix remains less than 5 nm up to 280 ˚C, at which

TE D

precipitation occurs. The average grain sizes of fcc matrix and precipitates increase rapidly thereafter and become ~ 60 and 40 nm, respectively, at 400 ˚C. Note that at such a high annealing temperature, 0.7 Tm, although grain coarsening has taken place, the retention of grain

EP

size below 100 nm is still remarkable. The retention of nanograins is presumably due to the pinning of grain boundaries by Fe solutes. Furthermore, as both Al rich matrix and Al6Fe phase

AC C

appear to have similar volume fraction and are well intercalated and separated from one another, each phase retains its nanograins. 4.3 The temperature and grain size dependent evolution of mechanical properties The addition of Fe dramatically strengthens the Al solid solution alloys. The hardness of as-deposited nt Al-5.5Fe is ~ 5.7 GPa (Fig. 8a). After heat treatment to 250 ˚C, the hardness declines slightly to 5.3 GPa, and reduces further to 4.75 GPa at 280 ˚C (~ 0.6 Tm of pure Al; or ~ 13

ACCEPTED MANUSCRIPT

0.5 Tm according to the Al-Fe phase diagram). The hardness drops sharply to 2.5-3 GPa after annealing at 300-400 ˚C. The hardness of Al-3Fe film evolves in a similar way with temperature, although annealing at 280 ˚C induces significant softening. The retention of high hardness is

RI PT

associated with the stability of nanotwins. Fe solutes significantly improve the thermal stability of fine microstructures and extended ITBs. It is interesting to see that the hardness of fully recrystallized Al-Fe remains high, at 2.5-3 GPa, compared to a majority of crystalline and

SC

metallic glass Al alloys. The high hardness of recrystallized Al-Fe may arise from nc grains (40-

M AN U

60 nm).

Fig. 8b shows the as-deposited Al-5.5Fe has a calculated Young’s modulus of 107 ± 4 GPa, higher than 76 GPa of (111)-textured Al. Upon annealing up to 280 ˚C, the modulus of Al5.5Fe gradually decreases to 99 ± 3 GPa, an indication of weaker interatomic bonding. The

precipitation of Al6Fe.

TE D

increase in modulus beyond an annealing temperature of 280 ˚C may be attributed to

Two major mechanisms are responsible for the high strength of nt Al-Fe, namely solid

EP

solution strengthening and Hall-Petch strengthening. Solid solution strengthening arises from the resistance of stress fields (around solutes) to dislocation propagation and can be expressed by

AC C

Fleischer formulation as [80]:

/D

∆> = α ∙ A ∙ BC

∙ !
/D

(5)

where α is a material-dependent factor, ∆τ is the shear stress increment derived from solute effect and ɛs is the interaction factor considering relations of shear modulus and lattice constant with solute concentration, which can be estimated by

14

ACCEPTED MANUSCRIPT

F G6 ∙ 6'( GH F G6
4
/DI ∙ I 6'( GH

BC = E

−3∙




K'(

∙

LK L3

E.

(6)

RI PT

α value is 1/700 for fcc Cu, whereas it was found that the empirical value of α for nc Ni-W alloys is 0.0235 [81], more than one order of magnitude higher than Cu. Al50Co50 and Al-Mo alloys have α values of 1/240 and 1/120, respectively [34, 82]. Thus, the α value is in the range of

1/700-1/100 for nt Al-Fe solid solution alloys. Both modulus and lattice constant are assumed to

SC

vary linearly with solute concentration [81, 83]. Consequently, the maximum stress increment

M AN U

from 5.5 % Fe solutes is estimated to be ~ 303 MPa, after a Taylor factor of ~ 3 is applied. It should be emphasized that the validity of Fleischer equation needs further study as it is generally applicable for solid solution systems with dilute impurity, and perhaps not all Fe solutes occupy the substitutional sites in the current study.

The Hall-Petch strengthening is expressed as:

TE D

MNO = MP +

QRS √

(7)

where d, σ0 denotes grain size (the transverse dimension of columnar grains) and lattice friction

EP

stress in single crystal, and KHP stands for Hall-Petch slope reflecting the barrier strength of GBs

AC C

against dislocation transmission. TBs in nt Al and Al-Fe alloys induce prominent strengthening, and TB-induced strengthening also follows the Hall-Petch relationship [20, 21, 83]. The KHP for nt Al-Fe alloys is estimated to be ~ 7.3 and ~ 6.05 GPa•√UV before and after the contribution of solid solution strengthening is subtracted [28]. Our previous study shows that Fe plays a much more pronounced role in grain refinement, compared to other solutes, such as Ag, W, Mo and Cr [28]. The nanoscale grain size may play a significant role in achievement of ultrahigh strength. The barrier strength of GBs for single dislocation transmission can be predicted using 15

ACCEPTED MANUSCRIPT

>∗ = (

QRS X(
)

/Y

)

(8),

where b is the magnitude of Burgers vector. The adoption of the equation assumes that the

RI PT

vertically aligned ITBs strengthen materials, similar to layer interface and horizontal TBs [84, 85]. Given Z = 0.35, A = 26 GPa, b = 0.286 nm and KHP = ~ 2 GPa√UV and after a Taylor factor of 3 is applied, > ∗ is estimated to be ~ 1.1 GPa. Hence the peak strength is estimated to be ~ 3

SC

GPa after solid solution effect is subtracted.

A combination of the solid solution and grain boundary strengthening mechanisms

M AN U

predicts a theoretical peak strength of ~ 3.3 GPa at an Fe composition of 5.5 %. This is greater than but comparable to the measured flow stresses, ~1.6-1.7 GPa (ε = 5%), from in situ micropillar compression tests on as-deposited samples. Also by using the Tabor equation, M = [/2.7, the measured maximum hardness of ~ 5.7 GPa for the nt Al-Fe alloys corresponds to

TE D

a flow stress of ~ 2.1 GPa, in agreement with flow stress measured from in situ micropillar compression tests. Also the flow stress of 1.7 GPa (lower bound) is ~ 46 % of the theoretical strength of Al [86].

EP

Recrystallization induced grain coarsening at 300 ˚C explains the degradation of

AC C

mechanical properties based on the Hall-Petch relation as shown in Fig. 8c. The flow stress of fully recrystallized Al-Fe measured at room temperature remains high, 0.9 – 1.1 GPa. The mechanical behaviors of recrystallized Al-Fe agree with Hall-Petch relation and the strength of nt Al-Fe outperforms other nanostructured Al and Al-Fe alloys [37, 87-91]. 4.4 Plasticity of pillars deformed at various temperatures

16

ACCEPTED MANUSCRIPT

Shear bands accompanied by drastic load-drop (on the stress-strain curve) have been frequently observed in deformed single-crystal micropillars [92, 93], metallic glass [94], high entropy alloys [95] and nanocrystalline grains [96]. However, all the nt Al-Fe samples deformed

RI PT

up to 15 % strain showed no shear bands or slip bands. And no dislocation burst appears on the stress-strain curves of nt Al-Fe as shown in Fig. 5 and Fig. 6. The plastic deformation of asdeposited and annealed pillars caused preferential dilation near the upper portion of the Al-Fe

SC

micropillars, and more importantly, the deformation and dilation can propagate gradually

downwards, distinct from the highly localized shear instability and catastrophic softening

M AN U

observed in some multilayers [97, 98]. Our prior post-compression microscopy studies of the asdeposited Al-Fe alloys show grain coarsening due to stress-induced partial dislocation migration [28].

TE D

4.5 High temperature mechanical properties

Fig. 9a shows the impacts of annealing on room-temperature mechanical behaviors of the nt Al-Fe as well as its stance in comparison with representative Al alloys. The upper and lower

EP

bound flow stresses are measured by using the mid-height and top diameters of micropillars, and the upper bound values are consistent with the flow stress converted from nanoindentation

AC C

hardness measurements. The strength of nt Al-Fe remains high up to 280 ˚C due to the thermal stability of nanotwinned columnar structure with 9R phase. Hardening derived from intermetallic formation is overshadowed by softening resulted from grain coarsening upon annealing above 280 ˚C. The fully recrystallized Al-Fe still shows a high flow stress of ~ 1 GPa. Al-Fe micropillars tested at elevated temperature show high flow stress, ~ 1.3 GPa, up to 200 ˚C, among one of the strongest reported to date. Such a high strength derives from the stability of nanotwins and Fe effects during high temperature mechanical testing. For 17

ACCEPTED MANUSCRIPT

micropillars tested at elevated temperatures (100 ˚C and 200 ˚C), the formation of particles was observed. It should be noted that ex situ studies also showed large Fe-deficient particles on nt AlFe surface when annealed at 280 ˚C (Fig. S6 and S7). It is likely that deformation accelerated the

RI PT

diffusion and subsequent formation of particles at elevated temperatures in the supersaturated AlFe alloys. However, the micropillars deformed in a chiefly uniform fashion at 300 ˚C. It is worth noting that significant softening occurs at 300 ˚C, as evidenced by the precipitous reduction of

SC

flow stress to ~ 0.13 GPa, much lower in comparison to ~ 0.88 GPa of Al-Fe annealed at the same temperature but tested at room temperature. As the nanocrystalline grains remain when

M AN U

tested at 300 ˚C, it is likely that significant softening occurs due to substantial reduction of elastic modulus of Al-Fe alloys, evidenced by the decrease of apparent modulus (Fig. S10), and possible involvement of other deformation mechanisms, such as grain rotation and sliding.

TE D

5. Conclusions

The thermal stability and temperature dependent mechanical properties of high-strength nt Al-Fe alloys were investigated through in-situ compression tests. Annealing studies clearly

EP

identify two regimes: 1) Fe solutes effectively stabilize the 9R phase and nanotwins in nt Al-Fe up to 280 ˚C, and 2) recrystallization and phase segregation occur at 300 ˚C, leading to the

AC C

formation of nanocomposite. The room-temperature flow stress of the Al-Fe alloys remains > 1.6 GPa after annealing at 200 ˚C, and is ~ 1.3 GPa when tested at 200 ˚C. A drastic softening occurs at a test temperature of 300 ˚C, presumably due to the substantial reduction of elastic modulus. The superb thermal and mechanical stability of nt Al-Fe alloys may promote their applications at elevated temperatures. The strategy of using solutes to stabilize nanotwins and 9R phase offers an opportunity to design nanotwinned metals for high temperature applications.

18

ACCEPTED MANUSCRIPT

RI PT

Acknowledgements We acknowledge financial support by DoE-BES under grant no. DE-SC0016337. Accesses to the Microscopy Facilities at Purdue University and Center for Integrated

SC

Nanotechnologies (managed by Los Alamos National Laboratory) are also acknowledged.

M AN U

Competing interests: The Authors declare no Competing Financial Interests.

Reference

AC C

EP

TE D

[1] D.K. Matlock, J.G. Speer, Third Generation of AHSS: Microstructure Design Concepts, Microstructure and Texture in Steels and Other Materials (2009) 185. [2] M. Song, C. Sun, Z. Fan, Y. Chen, R. Zhu, K.Y. Yu, K.T. Hartwig, H. Wang, X. Zhang, A roadmap for tailoring the strength and ductility of ferritic/martensitic T91 steel via thermo-mechanical treatment, Acta Mater. 112 (2016) 361-377. [3] P.V. Liddicoat, X.Z. Liao, Y.H. Zhao, Y.T. Zhu, M.Y. Murashkin, E.J. Lavernia, R.Z. Valiev, S.P. Ringer, Nanostructural hierarchy increases the strength of aluminium alloys, Nat. Commun. 1 (2010). [4] H.N. Kou, J. Lu, Y. Li, High-Strength and High-Ductility Nanostructured and Amorphous Metallic Materials, Adv. Mater. 26(31) (2014) 5518-5524. [5] J. Schiotz, K.W. Jacobsen, A maximum in the strength of nanocrystalline copper, Science 301(5638) (2003) 1357-1359. [6] C.A. Schuh, T.G. Nieh, T. Yamasaki, Hall-Petch breakdown manifested in abrasive wear resistance of nanocrystalline nickel, Scripta Mater. 46(10) (2002) 735-740. [7] J.R. Trelewicz, C.A. Schuh, The Hall-Petch breakdown in nanocrystalline metals: A crossover to glass-like deformation, Acta Mater. 55(17) (2007) 5948-5958. [8] S.X. McFadden, R.S. Mishra, R.Z. Valiev, A.P. Zhilyaev, A.K. Mukherjee, Low-temperature superplasticity in nanostructured nickel and metal alloys, Nature 398(6729) (1999) 684-686. [9] R.S. Mishra, R.Z. Valiev, S.X. McFadden, R.K. Islamgaliev, A.K. Mukherjee, High-strain-rate superplasticity from nanocrystalline Al alloy 1420 at low temperatures, Philos. Mag. A 81(1) (2001) 3748. [10] R.S. Mishra, M.W. Mahoney, S.X. McFadden, N.A. Mara, A.K. Mukherjee, High strain rate superplasticity in a friction stir processed 7075 Al alloy, Scripta Mater. 42(2) (1999) 163-168. [11] L. Lu, Y.F. Shen, X.H. Chen, L.H. Qian, K. Lu, Ultrahigh strength and high electrical conductivity in copper, Science 304(5669) (2004) 422-426. [12] J. Wang, X.H. Zhang, Twinning effects on strength and plasticity of metallic materials, MRS Bull. 41(4) (2016) 274-285. 19

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[13] X. Zhang, A. Misra, Superior thermal stability of coherent twin boundaries in nanotwinned metals, Scripta Mater. 66(11) (2012) 860-865. [14] D.C. Jang, X.Y. Li, H.J. Gao, J.R. Greer, Deformation mechanisms in nanotwinned metal nanopillars, Nat. Nanotechnol. 7(9) (2012) 594-601. [15] I.J. Beyerlein, X.H. Zhang, A. Misra, Growth Twins and Deformation Twins in Metals, Annu. Rev. Mater. Res. 44 (2014) 329-363. [16] D. Bufford, H.Y. Wang, X.H. Zhang, Thermal stability of twins and strengthening mechanisms in differently oriented epitaxial nanotwinned Ag films, J. Mater. Res. 28(13) (2013) 1729-1739. [17] X. Zhang, A. Misra, H. Wang, T.D. Shen, M. Nastasi, T.E. Mitchell, J.P. Hirth, R.G. Hoagland, J.D. Embury, Enhanced ardening in Cu/330 stainless steel multilayers by nanoscale twinning, Acta Mater. 52(4) (2004) 995-1002. [18] O. Anderoglu, A. Misra, H. Wang, F. Ronning, M.F. Hundley, X. Zhang, Epitaxial nanotwinned Cu films with high strength and high conductivity, Appl. Phys. Lett. 93(8) (2008). [19] S. Xue, Z. Fan, Y. Chen, J. Li, H. Wang, X. Zhang, The formation mechanisms of growth twins in polycrystalline Al with high stacking fault energy, Acta Mater. 101 (2015) 62-70. [20] D. Bufford, Y. Liu, J. Wang, H. Wang, X. Zhang, In situ nanoindentation study on plasticity and work hardening in aluminium with incoherent twin boundaries, Nat. Commun. 5 (2014). [21] D. Bufford, Y. Liu, Y. Zhu, Z. Bi, Q.X. Jia, H. Wang, X. Zhang, Formation Mechanisms of Highdensity Growth Twins in Aluminum with High Stacking-Fault Energy, Mater. Res. Lett. 1(1) (2013) 5160. [22] S. Xue, W. Kuo, Q. Li, Z. Fan, J. Ding, R. Su, H. Wang, X. Zhang, Texture-directed twin formation propensity in Al with high stacking fault energy, Acta Mater. 144 (2018) 226-234. [23] M.W. Chen, E. Ma, K.J. Hemker, H.W. Sheng, Y.M. Wang, X.M. Cheng, Deformation twinning in nanocrystalline aluminum, Science 300(5623) (2003) 1275-1277. [24] V. Yamakov, D. Wolf, S.R. Phillpot, H. Gleiter, Deformation twinning in nanocrystalline Al by molecular dynamics simulation, Acta Mater. 50(20) (2002) 5005-5020. [25] X.Z. Liao, F. Zhou, E.J. Lavernia, D.W. He, Y.T. Zhu, Deformation twins in nanocrystalline Al, Appl. Phys. Lett. 83(24) (2003) 5062-5064. [26] S.C. Xue, Z. Fan, O.B. Lawal, R. Thevamaran, Q. Li, Y. Liu, K.Y. Yu, J. Wang, E.L. Thomas, H.Y. Wang, X.H. Zhang, High-velocity projectile impact induced 9R phase in ultrafine-grained aluminium, Nat. Commun. 8 (2017). [27] R.C. Pond, D.A. Smith, On the absorption of dislocations by grain boundaries, PHILOSOPHICAL MAGAZIN 36(2) (1977) 353-366. [28] Q. Li, S. Xue, J. Wang, S. Shao, A.H. Kwong, A. Giwa, Z. Fan, Y. Liu, Z. Qi, J. Ding, H. Wang, J.R. Greer, H. Wang, X. Zhang, High-Strength Nanotwinned Al Alloys with 9R Phase, Adv Mater. 30(11) (2018). https://doi.org/10.1002/adma.201704629 [29] G.S. Choi, Y.H. Kim, H.K. Cho, A. Inoue, T. Masumoto, Ultrahigh Tensile-Strength of Amorphous Al-Ni-(Nd,Gd)-Fe Alloys Containing Nanocrystalline Al Particles, Scripta Metall Mater 33(8) (1995) 1301-1306. [30] Z. Wang, R.T. Qu, S. Scudino, B.A. Sun, K.G. Prashanth, D.V. Louzguine-Luzgin, M.W. Chen, Z.F. Zhang, J. Eckert, Hybrid nanostructured aluminum alloy with super-high strength, Npg Asia Mater 7 (2015). [31] J.T. Demasimarcin, D.K. Gupta, Protective Coatings in the Gas-Turbine Engine, Surf Coat Tech 68 (1994) 1-9. [32] J.M. Neidrich, L.W. Barron, SPATIAL LIGHT MODULATOR MIRROR METAL HAVING ENHANCED REFLECTIVITY U.S. Patent No. WO2008030846. Dallas, TX: U.S. Patent and Trademark Office. (2007). [33] C. Ophus, E.J. Luber, M. Edelen, Z. Lee, L.M. Fischer, S. Evoy, D. Lewis, U. Dahmen, V. Radmilovic, D. Mitlin, Nanocrystalline-amorphous transitions in Al-Mo thin films: Bulk and surface evolution, Acta Mater 57(14) (2009) 4296-4303.

20

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[34] D.S. Gianola, Z. Lee, C. Ophus, E.J. Luber, D. Mitlin, U. Dahmen, K.J. Hemker, V.R. Radmilovic, Tensile behavior of Al1-xMox crystalline and amorphous thin films, Acta Mater 61(5) (2013) 1432-1443. [35] V.V. Stolyarov, R. Lapovok, I.G. Brodova, P.F. Thomson, Ultrafine-grained Al-5 wt.% Fe alloy processed by ECAP with backpressure, Mat Sci Eng a-Struct 357(1-2) (2003) 159-167. [36] N.A. Belov, A.N. Alabin, Use of Multicomponent Phase Diagrams for Designing High Strength Casting Aluminum Alloys, Aluminium Alloys 2014 - ICAA14 794-796 (2014) 909-914. [37] T.T. Sasaki, T. Ohkubo, K. Hono, Microstructure and mechanical properties of bulk nanocrystalline Al-Fe alloy processed by mechanical alloying and spark plasma sintering, Acta Mater 57(12) (2009) 3529-3538. [38] S.S. Nayak, M. Wollgarten, J. Banhart, S.K. Pabi, B.S. Murty, Nanocomposites and an extremely hard nanocrystalline intermetallic of Al-Fe alloys prepared by mechanical alloying, Mat Sci Eng a-Struct 527(9) (2010) 2370-2378. [39] T.T. Sasaki, T. Mukai, K. Hono, A high-strength bulk nanocrystalline Al-Fe alloy processed by mechanical alloying and spark plasma sintering, Scripta Mater 57(3) (2007) 189-192. [40] T. Mukai, S. Suresh, K. Kita, H. Sasaki, N. Kobayashi, K. Higashi, A. Inoue, Nanostructured Al-Fe alloys produced by e-beam deposition: static and dynamic tensile properties, Acta Mater 51(14) (2003) 4197-4208. [41] D.K. Mukhopadhyay, C. Suryanarayana, F.H. Froes, Structural Evolution in Mechanically Alloyed Al-Fe Powders, Metall Mater Trans A 26(8) (1995) 1939-1946. [42] B. Huang, K.N. Ishihara, P.H. Shingu, Metastable phases of Al-Fe system by mechanical alloying, Mat Sci Eng a-Struct 231(1-2) (1997) 72-79. [43] Z. Horita, T. Fujinami, M. Nemoto, T.G. Langdon, Equal-channel angular pressing of commercial aluminum alloys: Grain refinement, thermal stability and tensile properties, Metall Mater Trans A 31(3) (2000) 691-701. [44] Y.H. Zhao, X.Z. Liao, Z. Jin, R.Z. Valiev, Y.T. Zhu, Microstructures and mechanical properties of ultrafine grained 7075 Al alloy processed by ECAP and their evolutions during annealing, Acta Mater 52(15) (2004) 4589-4599. [45] M.R. Shankar, S. Chandrasekar, A.H. King, W.D. Compton, Microstructure and stability of nanocrystalline aluminum 6061 created by large strain machining, Acta Mater 53(18) (2005) 4781-4793. [46] H.M. Kimura, K. Sasamori, A. Inoue, Al-Fe-based bulk quasicrystalline alloys with high elevated temperature strength, J Mater Res 15(12) (2000) 2737-2744. [47] R.W. Hayes, R. Rodriguez, E.J. Lavernia, The mechanical behavior of a cryomilled Al-10Ti-2Cu alloy, Acta Mater 49(19) (2001) 4055-4068. [48] T. Chookajorn, H.A. Murdoch, C.A. Schuh, Design of Stable Nanocrystalline Alloys, Science 337(6097) (2012) 951-954. [49] K.A. Darling, R.N. Chan, P.Z. Wong, J.E. Semones, R.O. Scattergood, C.C. Koch, Grain-size stabilization in nanocrystalline FeZr alloys, Scripta Mater 59(5) (2008) 530-533. [50] T. Frolov, K.A. Darling, L.J. Kecskes, Y. Mishin, Stabilization and strengthening of nanocrystalline copper by alloying with tantalum, Acta Mater 60(5) (2012) 2158-2168. [51] R. Kirchheim, Grain coarsening inhibited by solute segregation, Acta Mater 50(2) (2002) 413-419. [52] B. Farber, E. Cadel, A. Menand, G. Schmitz, R. Kirchheim, Phosphorus segregation in nanocrystalline Ni-3.6 at.% P alloy investigated with the tomographic atom probe (TAP), Acta Mater 48(3) (2000) 789-796. [53] J.R. Trelewicz, C.A. Schuh, Grain boundary segregation and thermodynamically stable binary nanocrystalline alloys, Phys Rev B 79(9) (2009). [54] J. Hu, Y.N. Shi, X. Sauvage, G. Sha, K. Lu, METALLURGY Grain boundary stability governs hardening and softening in extremely fine nanograined metals, Science 355(6331) (2017) 1292. [55] R.O. Scattergood, C.C. Koch, K.L. Murty, D. Brenner, Strengthening mechanisms in nanocrystalline alloys, Mat Sci Eng a-Struct 493(1-2) (2008) 3-11.

21

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[56] H. Hasegawa, S. Komura, A. Utsunomiya, Z. Horita, M. Furukawa, M. Nemoto, T.G. Langdon, Thermal stability of ultrafine-grained aluminum in the presence of Mg and Zr additions, Mat Sci Eng aStruct 265(1-2) (1999) 188-196. [57] S. Lee, A. Utsunomiya, H. Akamatsu, K. Neishi, M. Furukawa, Z. Horita, T.G. Langdon, Influence of scandium and zirconium on grain stability and superplastic ductilities in ultrafine-grained Al-Mg alloys, Acta Mater 50(3) (2002) 553-564. [58] V.M.S. Muthaiah, S. Mula, Effect of zirconium on thermal stability of nanocrystalline aluminium alloy prepared by mechanical alloying, J Alloy Compd 688 (2016) 571-580. [59] I.N. Sneddon, THE RELATION BETWEEN LOAD AND PENETRATION IN THE AXISYMMETRIC BOUSSINESQ PROBLEM FOR A PUNCH OF ARBITRARY PROFILE, International Journal of Engineering Science 3(1) (1965) 47-57. [60] F. Zhou, R. Luck, K. Lu, E.J. Lavernia, M. Ruhle, Amorphous-to-crystalline transformation induced by thermal annealing of a metastable Al90Fe10 composite, Philos Mag A 82(5) (2002) 1003-1015. [61] N.A. Belov, A.A. Aksenov, D.G. Eskin, Iron in aluminium alloys: impurity and alloying element. London: Taylor & Francis; 2002. [62] N.A. Belov, D.G. Eskin, A.A. Aksenov, Multicomponent phase diagrams: applications for commercial aluminium alloys. Oxford: Elsevier; 2005. [63] G.P.M. Leyson, W.A. Curtin, L.G. Hector, C.F. Woodward, Quantitative prediction of solute strengthening in aluminium alloys, Nat Mater 9(9) (2010) 750-755. [64] O.V. V., Mössbauer Analysis of the Atomic and Magnetic Structure of Alloys, Cambridge International Science, Cambridge, 2004. [65] P. Bordeenithikasem, J.B. Liu, S.A. Kube, Y.L. Li, T.X. Ma, B.E. Scanley, C.C. Broadbridge, J.J. Vlassak, J.P. Singer, J. Schroers, Determination of critical cooling rates in metallic glass forming alloy libraries through laser spike annealing, Sci Rep-Uk 7 (2017). [66] Y.H. Liu, T. Fujita, D.P.B. Aji, M. Matsuura, M.W. Chen, Structural origins of Johari-Goldstein relaxation in a metallic glass, Nat Commun 5 (2014). [67] S.C. Pun, W.B. Wang, A. Khalajhedayati, J.D. Schuler, J.R. Trelewicz, T.J. Rupert, Nanocrystalline Al-Mg with extreme strength due to grain boundary doping, Mat Sci Eng a-Struct 696 (2017) 400-406. [68] C.M. Muller, S. Parviainen, F. Djurabekova, K. Nordlund, R. Spolenak, The as-deposited structure of co-sputtered Cu-Ta alloys, studied by X-ray diffraction and molecular dynamics simulations, Acta Mater 82 (2015) 51-63. [69] V.A. Lubarda, On the effective lattice parameter of binary alloys, Mech Mater 35(1-2) (2003) 53-68. [70] S. Nemat-Nasser, M. Hori, Micromechanics: Overall Properties of Heterogeneous Materials, (1999). [71] O.N. Senkov, F.H. Froes, V.V. Stolyarov, R.Z. Valiev, J. Liu, Microstructure and microhardness of an Al-Fe alloy subjected to severe plastic deformation and aging, Nanostruct Mater 10(5) (1998) 691-698. [72] J. Wang, A. Misra, J.P. Hirth, Shear response of Sigma 3{112} twin boundaries in face-centeredcubic metals, Phys Rev B 83(6) (2011). [73] L. Liu, J. Wang, S.K. Gong, S.X. Mao, High Resolution Transmission Electron Microscope Observation of Zero-Strain Deformation Twinning Mechanisms in Ag, Phys Rev Lett 106(17) (2011). [74] G. Lucadamo, D.L. Medlin, Geometric origin of hexagonal close packing at a grain boundary in gold, Science 300(5623) (2003) 1272-1275. [75] J. Wang, N. Li, O. Anderoglu, X. Zhang, A. Misra, J.Y. Huang, J.P. Hirth, Detwinning mechanisms for growth twins in face-centered cubic metals, Acta Mater 58(6) (2010) 2262-2270. [76] F. Zhou, X.Z. Liao, Y.T. Zhu, S. Dallek, E.J. Lavernia, Microstructural evolution during recovery and recrystallization of a nanocrystalline Al-Mg alloy prepared by cryogenic ball milling, Acta Mater 51(10) (2003) 2777-2791. [77] Kwiecinski J, Ryfka J, W. J., Early stages of recrystallization in aluminium, Mater Sci Forum 157 (1993) 113-115. [78] J. Wang, N. Li, A. Misra, Structure and stability of Sigma 3 grain boundaries in face centered cubic metals, Philos Mag 93(4) (2013) 315-327.

22

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[79] X.C. Liu, H.W. Zhang, K. Lu, Strain-Induced Ultrahard and Ultrastable Nanolaminated Structure in Nickel, Science 342(6156) (2013) 337-340. [80] R.L. Fleischer, Substitutional Solution Hardening, Acta Metall Mater 11(3) (1963) 203-&. [81] T.J. Rupert, J.C. Trenkle, C.A. Schuh, Enhanced solid solution effects on the strength of nanocrystalline alloys, Acta Mater 59(4) (2011) 1619-1631. [82] R.L. Fleischer, Substitutional Solutes in Alru .1. Effects of Solute on Moduli, Lattice-Parameters and Vacancy Production, Acta Metallurgica Et Materialia 41(3) (1993) 863-869. [83] Y. Zhang, S. Xue, Q. Li, C. Fan, R. Su, J. Ding, H. Wang, H. Wang, X. Zhang, Microstructure and mechanical behavior of nanotwinned AlTi alloys with 9R phase, Scripta Mater 148 (2018) 5-9. [84] A. Misra, J.P. Hirth, R.G. Hoagland, Length-scale-dependent deformation mechanisms in incoherent metallic multilayered composites, Acta Mater 53(18) (2005) 4817-4824. [85] J. Li, Y. Chen, S. Xue, H. Wang, X. Zhang, Comparison of size dependent strengthening mechanisms in Ag/Fe and Ag/Ni multilayers, Acta Mater 114 (2016) 154-163. [86] S. Ogata, J. Li, S. Yip, Ideal pure shear strength of aluminum and copper, Science 298(5594) (2002) 807-811. [87] R.W. Hayes, D. Witkin, F. Zhou, E.J. Lavernia, Deformation and activation volumes of cryomilled ultrafine-grained aluminum, Acta Mater 52(14) (2004) 4259-4271. [88] X.K. Sun, H.T. Cong, M. Sun, M.C. Yang, Preparation and mechanical properties of highly densified nanocrystalline Al, Metall Mater Trans A 31(3A) (2000) 1017-1024. [89] H.J. Choi, S.W. Lee, J.S. Park, D.H. Bae, Positive Deviation from a Hall-Petch Relation in Nanocrystalline Aluminum, Mater Trans 50(3) (2009) 640-643. [90] N. Tsuji, Y. Saito, H. Utsunomiya, S. Tanigawa, Ultra-fine grained bulk steel produced by accumulative roll-bonding (ARB) process, Scripta Mater 40(7) (1999) 795-800. [91] P.G. Partridge, Subzero Tensile Properties of Vapor Quenched Aluminum-Alloys, J Mater Sci 21(9) (1986) 3211-3222. [92] J. Ding, Q. Li, J. Li, S. Xue, Z. Fan, H. Wang, X. Zhang, Mechanical behavior of structurally gradient nickel alloy, Acta Mater 149 (2018) 57-67. [93] C.P. Frick, B.G. Clark, S. Orso, A.S. Schneider, E. Arzt, Size effect on strength and strain hardening of small-scale [111] nickel compression pillars, Mat Sci Eng a-Struct 489(1-2) (2008) 319-329. [94] C.Q. Chen, Y.T. Pei, J.T.M. De Hosson, Effects of size on the mechanical response of metallic glasses investigated through in situ TEM bending and compression experiments, Acta Mater 58(1) (2010) 189-200. [95] A.M. Giwa, P.K. Liaw, K.A. Dahmen, J.R. Greer, Microstructure and small-scale size effects in plasticity of individual phases of Al0.7CoCrFeNi High Entropy alloy, Extreme Mech Lett 8 (2016) 220228. [96] D.C. Jang, J.R. Greer, Size-induced weakening and grain boundary-assisted deformation in 60 nm grained Ni nanopillars, Scripta Mater 64(1) (2011) 77-80. [97] S.M. Han, M.A. Phillips, W.D. Nix, Study of strain softening behavior of Al-Al3Sc multilayers using microcompression testing, Acta Mater 57(15) (2009) 4473-4490. [98] Z. Fan, Q. Li, J. Li, S.C. Xue, H.Y. Wang, X.H. Zhang, Tailoring plasticity of metallic glasses via interfaces in Cu/amorphous CuNb laminates, J Mater Res 32(14) (2017) 2680-2689.

23

ACCEPTED MANUSCRIPT

RI PT

Figure captions

Figure 1. (a) XRD profiles of Al-5.5Fe annealed up to 400 oC. (b) The corresponding magnified XRD profiles showing Al (111) peak shift and the formation of intermetallic phase, Al6Fe at higher annealing temperatures.

M AN U

SC

Figure 2. Plan-view and cross-section TEM (XTEM) studies on the microstructure of as-deposited Al5.5Fe. Low magnification (a) Plan-view and (b) XTEM micrographs with corresponding selected area electron diffraction (SAED) patterns. (c) The intermediate magnification XTEM and the corresponding fast Fourier Transform (FFT) showing diffuse ITBs, i.e. 9R phase. (d, e, f) High resolution TEM (HRTEM) images show columnar and twin structure where the 9R phase is bounded by matrix and twin variants. The 9R phase boundaries are marked by blue arrows; some low-angle grain boundaries are marked by white arrows.

TE D

Figure 3. XTEM micrographs of Al-5.5Fe annealed at (a) 200, (b) 280 and (c) 300 ˚C. (a1-a4) XTEM images showing the retention of nanoscale columnar grains, twin boundaries and 9R structure after annealing at 200 oC. (b1- b2) XTEM images of specimens annealed at 280 oC showing the formation of precipitates adjacent to free surface and fine columnar structure remains underneath free surface. (b3-b4) TEM micrographs showing the twin structure adjacent to one precipitate survives and the precipitate has superlattice structure with ~ 0.46 nm interplanar distance. (c1-c3) XTEM images showing a complete recrystallization of the alloy annealed at 300 ˚C, containing abundant superlattice and Al6Fe precipitates. (c4) An HRTEM image confirms the precipitate of Al6Fe with orthorhombic structure. Figure 4. STEM micrographs and corresponding EDS compositional maps of Al-5.5Fe annealed at (a) 200, (b) 280, (c) 330 and (d) 400 ˚C.

AC C

EP

Figure 5. Room temperature in-situ compression tests on Al-5.5Fe specimens annealed at various temperatures: (a) An SEM image demonstrates a typical in-situ experiment on the Al-Fe micropillars. (b) Representative engineering stress-strain curves of Al-5.5Fe treated at different temperatures. (c-f) Corresponding SEM snapshots captured at different strains during in-situ tests of as-deposited and annealed specimens. Partial unloading segments were deliberately incorporated at low strains to probe apparent elastic moduli and alignment conditions. No shear bands were observed for all cases. Ta and Ttest stand for annealing temperature and testing temperature, respectively. (See supplementary Movies S1-S4 for details). Figure 6. Elevated temperature in-situ compression tests on Al-5.5Fe specimens. Engineering stressstrain curves and corresponding SEM snapshots at various strains of pillars tested at (a) room temperature (data of pillars with ~ 500 nm and ~ 1300 nm in diameter are coupled), (b) 100, (c) 200 and (d) 300 ˚C. Flow stresses remain greater than 1 GPa when tested at 200 oC or less, and reduce substantially to ~ 0.13 GPa at 300 oC. Particles were extruded from walls of micropillars at 100 and 200 ˚C. At 300 ˚C, micropillars deform in a homogeneous mode and exhibit surface wrinkles. (See supplementary Movies S5-S7 for more details). Figure 7. (a) Lattice parameter versus iron composition. (b) Lattice parameter evolution: lattice parameter of fcc phase in annealed Al-5.5Fe specimens as a function of annealing temperature. The lattice constants

24

ACCEPTED MANUSCRIPT

calculated using elasticity inclusion model are present as dotted lines. (c) Temperature dependent evolution of grain sizes of fcc phase and precipitates in Al-5.5Fe specimens.

RI PT

Figure 8. (a) Nanoindentation hardnesses of Al-3.0Fe and Al-5.5Fe remain largely unchanged up to 250 o C, followed by precipitous softening thereafter. (b) Elastic moduli of Al-3.0Fe and Al-5.5Fe decrease gradually up to 250 oC, followed by increment at higher annealing temperatures. The moduli of Al(111) and Al6Fe are shown as horizontal dotted lines as references. (c) The Hall-Petch plot for the as-deposited Al-5.5Fe and specimens annealed at 200, 280, 300, 330 and 400 oC. The selected cg, ufg and nc pure Al and Al-xFe alloys processed by different techniques are collected from literature [53, 76-80].

M AN U

SC

Figure 9. Evolutions of flow stresses of Al-5.5Fe at different annealing and elevated temperatures. (a) Flow stress (at 5 % strain) of Al-5.5Fe remain >1.5 GPa up to annealing temperature of 200 oC, and decreases to ~ 1 GPa after annealing at 300 oC or greater. (b) Variations of flow stress at 5 % strain for alloys tested at elevated temperatures. Flow stresses of Al-5.5Fe remain 1.5 GPa at 200 oC, and then drop drastically to ~ 130 MPa at 300 oC. The upper and lower bound flow stresses were calculated by taking the contact and middle height diameters into account. Flow stress was also converted from nanoindentation hardness divided by the Tabor factor of 2.7.

Captions for supplementary figures and movies

Figure S1. Plan-view and cross-section STEM micrographs and corresponding EDS maps of as-deposited Al-5.5Fe specimens show no indication of phase segregations. Figure S2. Microstructure of Al-5.5Fe annealed at 200 ˚C. (a) The low-to-intermediate magnification XTEM micrographs taken along <110> direction. (b, c) TEM image and filter image show that the vertical boundaries are low angle grain boundaries (LAGBs) decorated with Frank partials.

TE D

Figure S3. STEM micrograph and the corresponding EDS map of Al-5.5Fe annealed at 280 ˚C show that precipitation initiated at free surface of specimens and fine columnar grains were preserved underneath the free surface.

EP

Figure S4. STEM micrograph of Al-5.5Fe annealed at 300 ˚C shows equiaxed nanograins after recrystallization. (b) The corresponding EDS map reveals the existence of Fe-rich and -deficient regions. (c) Linear scan shows the alternating distribution of Fe. The Fe concentration is lower than stoichiometric Fe concentration of ~ 14.3 % in Al6Fe.

AC C

Figure S5. Microstructure of Al-5.5Fe annealed at 300 ˚C. (a) XTEM image exhibits ordering pattern, i.e. superlattice, a majority of which is parallel to growth direction. The superlattice structure has interplanar distance of ~ 0.44 to 0.55 nm. (b) TEM image showing the formation of Al6Fe intermetallic precipitates after recrystallization. (c) The intermetallic dispersoid is determined to be Al6Fe with an orthorhombic cmcm structure, which were frequently found adjacent to the superlattice structures. (d) HRTEM image showing fcc phase, which is Fe-deficient. Figure S6. SEM images show surface protrusion of particles from matrices takes place as annealing temperature reaches 300 ˚C or greater for Al-3Fe and Al-5.5Fe specimens. Linear scans indicate that the extruded particles have lower Fe composition in comparison with the one in matrix, indicating Al diffusion dominates the diffusion process at elevated temperature. Figure S7. STEM and corresponding EDS map of Al-5.5Fe specimen annealed at 280 ˚C present the cellular structure in matrix and surface protrusion of Fe-deficient giant particles from the matrix.

25

ACCEPTED MANUSCRIPT

Figure S8. XTEM images of Al-9.9Fe specimen showing heterogeneously distributed amorphous phase marked by white circles. Figure S9. Statistic distribution of grain sizes of (a) as-deposited Al-5.5Fe and specimens annealed at (b) 200, (c) 280, (d) 300, (e) 330 and (f) 400 ˚C. Grain size evolutions of both fcc phase and precipitates are presented.

RI PT

Figure S10. Apparent modulus of Al-5.5Fe versus testing temperatures. The apparent modulus derived from partial unloading segments of in-situ compressive stress-strain curves. Movie S1. In situ compression video showing Al-5.5Fe micropillar compressed at room temperature.

SC

Movie S2. In situ compression video showing Al-5.5Fe micropillar, annealed at 100 ˚C, compressed at room temperature. Movie S3. In situ compression video showing Al-5.5Fe micropillar, annealed at 200 ˚C, compressed at room temperature.

M AN U

Movie S4. In situ compression video showing Al-5.5Fe micropillar, annealed at 300 ˚C, compressed at room temperature. Movie S5. In situ compression video showing Al-5.5Fe micropillar compressed at 100 ˚C. Movie S6. In situ compression video showing Al-5.5Fe micropillar compressed at 200 ˚C.

AC C

EP

TE D

Movie S7. In situ compression video showing Al-5.5Fe micropillar compressed 300 ˚C.

26

ACCEPTED MANUSCRIPT

High temperature thermal and mechanical stability of high-strength nanotwinned Al alloys

RI PT

Qiang Lia, Jaehun Choa, S. Xuea, X. Suna, Y. F. Zhanga, Z. Shanga, H. Wanga,b and X. Zhanga,* a

School of Materials Engineering, Purdue University, West Lafayette, IN 47907

b

School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN 47907

EP

TE D

M AN U

SC

Corresponding author: X. Zhang, [email protected]

AC C

Figure 1. (a) XRD profiles of Al-5.5Fe annealed up to 400 oC. (b) The corresponding magnified XRD profiles showing Al (111) peak shift and the formation of intermetallic phase, Al6Fe at higher annealing temperatures.

1

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Figure 2. Plan-view and cross-section TEM (XTEM) studies on the microstructure of as-deposited Al5.5Fe. Low magnification (a) Plan-view and (b) XTEM micrographs with corresponding selected area electron diffraction (SAED) patterns. (c) The intermediate magnification XTEM and the corresponding fast Fourier Transform (FFT) showing diffuse ITBs, i.e. 9R phase. (d, e, f) High resolution TEM (HRTEM) images show columnar and twin structure where the 9R phase is bounded by matrix and twin variants. The 9R phase boundaries are marked by blue arrows; some low-angle grain boundaries are marked by white arrows.

2

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Figure 3. XTEM micrographs of Al-5.5Fe annealed at (a) 200, (b) 280 and (c) 300 ˚C. (a1-a4) XTEM images showing the retention of nanoscale columnar grains, twin boundaries and 9R structure after annealing at 200 oC. (b1- b2) XTEM images of specimens annealed at 280 oC showing the formation of precipitates adjacent to free surface and fine columnar structure remains underneath free surface. (b3-b4) TEM micrographs showing the twin structure adjacent to one precipitate survives and the precipitate has superlattice structure with ~ 0.46 nm interplanar distance. (c1-c3) XTEM images showing a complete recrystallization of the alloy annealed at 300 ˚C, containing abundant superlattice and Al6Fe precipitates. (c4) An HRTEM image confirms the precipitate of Al6Fe with orthorhombic structure.

3

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Figure 4. STEM micrographs and corresponding EDS compositional maps of Al-5.5Fe annealed at (a) 200, (b) 280, (c) 330 and (d) 400 ˚C.

4

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

Figure 5. Room temperature in-situ compression tests on Al-5.5Fe specimens annealed at various temperatures: (a) An SEM image demonstrates a typical in-situ experiment on the Al-Fe micropillars. (b) Representative engineering stress-strain curves of Al-5.5Fe treated at different temperatures. (c-f) Corresponding SEM snapshots captured at different strains during in-situ tests of as-deposited and annealed specimens. Partial unloading segments were deliberately incorporated at low strains to probe apparent elastic moduli and alignment conditions. No shear bands were observed for all cases. Ta and Ttest stand for annealing temperature and testing temperature, respectively. (See supplementary Movies S1-S4 for details).

5

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

Figure 6. Elevated temperature in-situ compression tests on Al-5.5Fe specimens. Engineering stress-strain curves and corresponding SEM snapshots at various strains of pillars tested at (a) room temperature (data of pillars with ~ 500 nm and ~ 1300 nm in diameter are coupled), (b) 100, (c) 200 and (d) 300 ˚C. Flow stresses remain greater than 1 GPa when tested at 200 oC or less, and reduce substantially to ~ 0.13 GPa at 300 oC. Particles were extruded from walls of micropillars at 100 and 200 ˚C. At 300 ˚C, micropillars deform in a homogeneous mode and exhibit surface wrinkles. (See supplementary Movies S5-S7 for more details).

6

RI PT

ACCEPTED MANUSCRIPT

TE D

M AN U

SC

Figure 7. (a) Lattice parameter versus iron composition. (b) Lattice parameter evolution: lattice parameter of fcc phase in annealed Al-5.5Fe specimens as a function of annealing temperature. The lattice constants calculated using elasticity inclusion model are present as dotted lines. (c) Temperature dependent evolution of grain sizes of fcc phase and precipitates in Al-5.5Fe specimens.

AC C

EP

Figure 8. (a) Nanoindentation hardnesses of Al-3.0Fe and Al-5.5Fe remain largely unchanged up to 250 o C, followed by precipitous softening thereafter. (b) Elastic moduli of Al-3.0Fe and Al-5.5Fe decrease gradually up to 250 oC, followed by increment at higher annealing temperatures. The moduli of Al(111) and Al6Fe are shown as horizontal dotted lines as references. (c) The Hall-Petch plot for the as-deposited Al-5.5Fe and specimens annealed at 200, 280, 300, 330 and 400 oC. The selected cg, ufg and nc pure Al and Al-xFe alloys processed by different techniques are collected from literature [53, 76-80].

7

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

Figure 9. Evolutions of flow stresses of Al-5.5Fe at different annealing and elevated temperatures. (a) Flow stress (at 5 % strain) of Al-5.5Fe remain >1.5 GPa up to annealing temperature of 200 oC, and decreases to ~ 1 GPa after annealing at 300 oC or greater. (b) Variations of flow stress at 5 % strain for alloys tested at elevated temperatures. Flow stresses of Al-5.5Fe remain 1.5 GPa at 200 oC, and then drop drastically to ~ 130 MPa at 300 oC. The upper and lower bound flow stresses were calculated by taking the contact and middle height diameters into account. Flow stress was also converted from nanoindentation hardness divided by the Tabor factor of 2.7.

8