Scripta Materialia 154 (2018) 20–24

Contents lists available at ScienceDirect

Scripta Materialia

journal homepage: www.elsevier.com/locate/scriptamat

Regular article

Hierarchical microstructure and strengthening mechanisms of a

CoCrFeNiMn high entropy alloy additively manufactured by selective
laser melting
Z.G. Zhu a, Q.B. Nguyen a, F.L. Ng a, X.H. An b,⁎, X.Z. Liao b, P.K. Liaw c, S.M.L. Nai a,⁎, J. Wei a,⁎
a
Singapore Institute of Manufacturing Technology, 73 Nanyang Drive, 637662, Singapore
b
School of Aerospace, Mechanical and Mechatronic Engineering, The University of Sydney, Sydney, NSW 2006, Australia
c
Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 37996, USA

a r t i c l e i n f o a b s t r a c t

Article history: A near-fully dense CoCrFeNiMn high entropy alloy was additively manufactured by selective laser melting suc-
Received 26 March 2018 cessfully. The as-built samples exhibit a hierarchical structure, including melt pools, columnar grains, sub-micron
Received in revised form 5 May 2018 cellular structures, and dislocations. An outstanding combination of high strength and excellent ductility com-
Accepted 10 May 2018
pared to those fabricated by conventional methods was achieved in the as-built samples. Quantitative analysis
Available online 20 May 2018
revealed that the cellular structures contribute significantly to the enhanced strength through dislocation hard-
Keywords:
ening rather than cellular boundary strengthening. The excellent ductility is correlated with the steady strain
Selective laser melting hardening controlled by the sophisticated dislocation activities.
High entropy alloys © 2018 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Cellular structure
Mechanical property
Steady strain hardening

The discovery of high entropy alloys (HEAs) catalyzes significant ex-
citement in the materials community due to their attractive properties
[1–4]. As a representative HEA, the CoCrFeNiMn that forms a single-
phase solid solution with a face-centered-cubic (FCC) crystal structure
has been extensively investigated [5–8]. One intriguing characteristic
of this HEA is its exceptional fracture toughness at cryogenic tempera-
tures, making it a candidate as a low-temperature structural material
[7]. However, the current preparations of HEAs rely mainly on the con-
ventional melting or casting methods [9], imposing enormous limita-
tions to produce samples with complex geometry in terms of cost and
efficiency for practical applications.
The additive manufacturing (AM) has been recognized as a transfor-
mative technology across multiple industries [10,11]. Based on their ad-
vantages of net-shape manufacturing capability and design freedom, it
is feasible to harvest parts with complex geometries directly from com-
puter-aided design (CAD) models [12]. Among the various AM methods,
selective laser melting (SLM) has enjoyed ever-growing popularity and
has been employed to fabricate engineering materials [13–16]. Up to
now, the investigation of processing HEAs by SLM is quite rare due to
compositional complexity [17,18]. Brif et al. [17] printed a CoCrFeNi
HEA, yet the detailed microstructures and deformation mechanisms of
⁎ Corresponding authors.
E-mail addresses: xianghai.an@sydney.edu.au, (X.H. An), mlnai@simtech.a-star.edu.sg,
(S.M.L. Nai), jwei@simtech.a-star.edu.sg (J. Wei).
the fabricated samples are lacking to understand the structure−prop-
perty relationship. This study aims to explore printability of the proto-
type CoCrFeNiMn HEA, and uncover its hierarchical microstructure as
well as deformation behaviors.
The CoCrFeNiMn HEA powder was fabricated by gas atomization.
The chemical composition was analyzed by inductively coupled plasma
mass spectroscopy (ICP-MS) for metallic elements and Instrumental gas
analysis (IGA) for non-metallic light elements. As listed in Table 1, the
as-built sample shows a slightly lower Mn and higher oxygen content
due to the high vapor pressure of Mn and oxidation, respectively. The
samples were fabricated using a ProX 300 (3D systems) with a laser
beam diameter of 75 μm at a constant layer thickness (d: 40 μm) and
hatching space (h: 50 μm) while varying the laser power (P: 160–
290 W) and scanning speed (v: 1500–2500 mm/s). The scanning
angle between adjacent layers is 90°, as shown in Fig. 1(a). Samples
with two geometries were manufactured: cubic samples with a dimen-
sion of 10 mm × 10 mm × 6 mm for microstructural analysis and tensile
coupons with a gauge dimension of 25 mm × 6 mm × 4 mm. The heat
treatment (HT) of the as-built samples was conducted at 900°C for 1 h
in argon atmosphere, followed by furnace cooling. The tensile test was
conducted using a universal Instron-5982 testing system at a strain
rate of 10−3/s. To benchmark the as-built HEAs, the tensile properties
of the as recrystallized HEA samples fabricated by the conventional
method were also investigated.
Based on the Archimedes principle, the density of the samples was
measured (the density of the as recrystallized sample, 7.986 g/cm3, is

https://doi.org/10.1016/j.scriptamat.2018.05.015
1359-6462/© 2018 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
 Z.G. Zhu et al. / Scripta Materialia 154 (2018) 20–24 21

Table 1 results in balling [20]. Herein, the samples with the maximum relative
Chemical composition of the CoCrFeNiMn powder and as-built HEA (at. %). density of 99.2% (E: 60 J/mm3, P: 240 W, v: 2000 mm/s, named as as-
Sample Co Cr Fe Ni Mn O N H built V2000) and the minimum relative density of 97.1% (E: 48 J/mm3,
Powder 20.4 19.2 19.7 20.5 20.1 0.036 0.009 0.0007
P: 240 W, v: 2500 mm/s, named as as-built V2500) marked by arrows
As-built 20.6 19.3 20.0 20.3 19.6 0.050 0.017 0.0011 were selected for the following studies. The XRD patterns in Fig. S2 con-
firm that the as-built samples remain a single phase with an FCC
structure.
used as the reference). The detailed microstructures were characterized Fig. 2 exhibits the hierarchical microstructure of the as-built V2000.
using X-ray diffraction (XRD), optical microscopy (OM), scanning elec- The etched surface [Fig. 2(a)] presents typical semi-elliptical scan tracks
tron microscopy (SEM) equipped with energy dispersive spectrum of melt pools (two melt pool boundaries are marked by arrows) arising
(EDS), electron backscattered diffraction (EBSD), and scanning trans- from the heat which flows radially away from the center [21]. The width
mission electron microscopy (STEM). All microstructural characteriza- and height of the melt pools are about 80 μm and 50 μm, corresponding
tions were performed by observing the front view [Fig. 1(a)]. To reveal to the laser beam diameter and layer thickness, respectively. The SEM
the melt pool boundary, the samples were etched in a solution with image in Fig. 2(b) shows the formation of columnar grains along the
HNO3:HCl:H2O = 1:3:2. In the EBSD grain boundary (GB) maps, the cri- building direction (BD), signalizing the epitaxial growth [21]. Fig. 2(c)
terion of 15° was used to differentiate low angle GBs (LAGBs) and high indicates the formation of cellular structure that has been correlated
angle GBs (HAGBs). with the solidification conditions [15,22,23]. Besides, some nanovoids
The SEM and EBSD inverse pole figure (IPFs) images shown in Figs. 1 tend to distribute along cell walls. The EBSD IPF map along the BD in
(b) and (c) respectively, reveal spherical powders with an average size Fig. 2(d) reveals that many grains grow along the preferred growth di-
of 36 μm and a polycrystalline microstructure. The EDS mapping of the rection of 〈001〉 for cubic materials during the solidification process
powder (Fig. S1) indicates that Mn and Ni segregate at the GB areas. [24]. The grain size is estimated to be about 12.9 μm based on the
This is in accordance with Laurent-Brocq's finding [19] that the HAGBs [15]. The GB map in Fig. 2(e) manifests that the as-built sample
interdendrites are enriched in Mn and Ni for the cast samples. It is contained a high fraction (~ 30%) of LAGBs while the Kernel average
known that the density is dependent on the laser energy density, E, cal- misorientation (KAM) map [Fig. 2(f)] delineates local misorientations
culated by E = P / (hdv) [20]. As illustrated in Fig. 1(d), the density firstly on the order of 0.2° − 0.9° across grains. The forescatter electron detec-
increases with E, reaching its maximum of about 99.2%, due to a more tor (FSD) image and corresponding elemental distribution in Fig. S3 in-
complete melting. The density decreases with further increase of E, dicate that the elemental distribution is relatively uniform with a minor
since excess energy could be detrimental to the surface finish and segregation of Mn along the melt pool boundaries. Further fine

Fig. 1. (a) Schematic illustration of the scanning strategy. The building direction, top view, front view, and side view are indicated. (b) SEM image and (c) EBSD IPF of the powder. (d)
Dependence of the relative density on the laser energy density.
22 Z.G. Zhu et al. / Scripta Materialia 154 (2018) 20–24

Fig. 2. (a) OM and (b) SEM images of the front surface. (c) SEM image of the cellular structure. (d) The EBSD IPFs of the front surface. (e) EBSD image quality (IQ) map with HAGBs (blue
lines) and LAGBs (red lines) superimposed. (f) KAM image of the front surface. (g) Bright-field STEM image of the cellular structure with the corresponding SAED pattern. (h) Bright-field
STEM image of the selected area illustrated by a square in (g) and the elemental distribution maps of the area.

microstructure observations, as shown in the bright-field STEM images
in Figs. 2(g, h), demonstrate a typical sub-micron cellular structure with
an average size of 0.34 ± 0.04 μm. Similar to those in additively
manufactured 316L stainless steel (SS) [15,25], the cell walls are deco-
rated with a high density of dislocations with clean interiors. The inset
corresponding selected area electron diffraction (SAED) pattern in Fig.
2(g) indicates small misorientation among the cells [15]. The elemental
distribution from EDS analysis associated with the area of Fig. 2(h) [il-
lustrated by a square in Fig. 2(g)] is fairly uniform without obvious seg-
regation since the fast cooling process may make the segregation
kinetically suppressed [26]. It is clear that the as-built sample exhibits
a hierarchical structure, including melt pools, columnar grains, sub-mi-
cron cellular structures, and dislocations. After HT (Fig. S4), the cellular
structure almost disappears and exhibits a much lower dislocation den-
sity, implying the thermodynamically metastable nature of the cellular
structure [23].
Representative engineering stress−strain curves of the fabricated
samples are shown in Fig. 3(a). The results of the yield strength (σy), ul-
timate strength (σUTS), and tensile elongation (εf) are summarized in
Table S1. The as-built V2000 exhibits an excellent combination of
strength and ductility, comparable to the as recrystallized and as cast
CoCrFeNiMn sample [27]. The as-built V2000 possesses a much higher
σy of 510 MPa, which nearly doubles compared with the as recrystal-
lized counterparts. After HT, although the σy decreases due to the recov-
ery of the cell structures, the ductility increases. For the V2500, the
existence of pores could act as critical sites for stress concentration
that leads to premature fracture [28]. Fig. 3(b) exhibits a typical fracture
surface of the as-built V2000, which is dominated by fine dimples with
similar sizes to the cellular structure. This similarity may imply that the
ductile fracture initiates at the cell walls [22]. Besides the dimples,
nanoparticles marked by the arrows were detected on the fracture sur-
face. The EDS results confirm that the particles are a Cr-rich compound
with the composition (at. %): Cr = 59.12, Mn = 33.43, Fe = 2.49, Ni =
2.47, and Co = 2.49. Since the oxygen content increases for the as-built
sample, the Cr-rich particles might be the oxide that has been reported
for this HEA [7]. In addition, the inset in Fig. 3(b) shows an open pore
associated with an unmelted powder marked by the arrow.
To essentially uncover the underlying deformation mechanisms of
the as-built samples, the microstructures of the as-built V2000 strained
to fracture were examined using EBSD and STEM. Based on the IPF map
in Fig. 4(a), the plastic deformation is mainly accommodated by disloca-
tion activities although several deformation twins marked by arrows
 Z.G. Zhu et al. / Scripta Materialia 154 (2018) 20–24
Fig. 3. (a) The engineering stress−strain curves of the samples with different conditions.
(b) The fracture surface of the as-built V2000. The arrows indicate the presence of Cr-rich
particles. The inset shows the presence of an unmelted powder marked by the arrow.
are occasionally found. According to a recent investigation [29], the crit-
ical stress for initiating deformation twins in this HEA is 720 ± 30 MPa.
In light of the true stress of about 820 MPa before necking, the formation
of deformation twins can be reasonably expected. However, the KAM
map associated with the geometrically necessary dislocations in Fig. 4
(b) indicates a local misorientation up to 1.6°, validating the dominance
of dislocation-mediated plasticity during deformation. The post-
mortem bright-field STEM image in Fig. 4(c) suggests a remarkable dis-
location trapping and retention mechanism within the cells [15], lead-
ing to an apparent increase of dislocation density inside the cell walls.
Another fingerprint of dislocation substructures is the planar slip
bands, as shown in Fig. 4(d). These slip bands substantially interact
with cellular structures to form a three-dimensional dislocation net-
work. In general, the hierarchical dislocation activities controlled the
deformation process in the as-built HEA, while deformation twinning
made additional contributions to the plastic flow.
The present research reveals that the CoCrFeNiMn HEA can not only
be additively manufactured by SLM, but also possesses enhanced me-
chanical properties. In fact, the strengthening effect is common for as-
built samples with SLM and has been correlated with particular micro-
structures [13,15,30]. Previous investigations on 316L SS proposed
that the sub-micron cellular structure is a major contributor (over
~90%) to the enhanced σy which was scaled with the cell size based
on the Hall-Petch relationship [15]. By employing this philosophy, the
estimated σy for the as-built V2000 is ~1034 MPa [31] that is much
higher than the observed value. Regarding the sophisticated dislocation
23
activities during deformation, the σy can be expressed as follows [32]:
pffiffiffi pffiffiffi
σ y ¼ σ 0 þ k= d þ αMGb ρ ð1Þ
where σ0 is the friction stress, the second term is GB strengthening (k is
the Hall–Petch slope, and d is the grain size), and the third term is dislo-
cation strengthening (α is a constant, M is the Taylor factor of 3.06, G is
the shear modulus, b is the Burgers vector of full dislocations, and ρ
stands for the dislocation density). In addition, the relationship between
dislocation density and cell size (λ) can be expressed as [32]:
pffiffiffi
ρ ¼ c=λ ð2Þ
where c is a constant [32]. Therefore, based on the typical values of σ0
(194 MPa) and k (490 MPa μm1/2) [31], G (80 GPa) and b (0.2536 nm)
[5], d (12.9 μm), λ (340 nm), and the multiplication value of α and c
to be approximately one [32], the calculated σy with the value of
~513 MPa is perfectly consistent with the experimental value. This sig-
nals that the cell size cannot be simply scaled with the σy of the as-
built sample using the Hall-Petch relationship. Instead, the enhanced
strength is mainly contributed by the dislocation strengthening in addi-
tion to the friction stress and grain boundary strengthening. The excep-
tionally high GB hardening efficiency of CoCrFeNiMn HEA may enable
the cell walls not to be effective as HAGBs in terms of strengthening
[8]. The dislocation strengthening effect could get further support
from the HT sample, which possesses a decrease of σy with a reduction
of the dislocation density.
The as-built HEA exhibits an excellent uniform elongation due to its
steady strain hardening capability at high stress levels. Since the cellular
structures are not traditional dislocation walls [15], the high density of
dislocations can be remarkably trapped and stored in the cell interior
during deformation. Besides, the extensive interactions between slip
bands and cellular structures result in the formation of a sophisticated
dislocation configuration, as exhibited in Fig. 4(d), which effectively
blocks the dislocation motion like convectional HAGBs. These elaborate
dislocation substructures played a crucial role in maintaining the strain
hardening ability. Moreover, owing to the microstructural heterogene-
ities, the back stress may exist during deformation to facilitate the strain
hardening behaviors [15]. Therefore, combined with improved strength,
the exceptional ductility of as-built HEA makes the strength−ductility
synergy comparable or even superior to those high-performance
CoCrFeNiMn HEAs [31]. Since the SLM process plays crucial roles in
the hierarchical structure at the micro-scale, the mechanical properties
of the as-built materials may be further improved feasibly through opti-
mizing the process parameters, scanning strategy, and fabrication
orientation.
In summary, a near-fully dense CoCrFeNiMn HEA was additively
manufactured using SLM successfully. The hierarchical structure, in-
cluding melt pools, columnar grains, sub-micron cellular structures,
and dislocations, endowed the as-built samples with an outstanding
combination of high strength and excellent ductility compared to
those fabricated by conventional methods. Quantitative analysis re-
vealed that cellular structures contributed remarkably to the enhanced
strength through dislocation hardening rather than cellular boundary
strengthening.
The work was supported by the A*STAR Additive Manufacturing
Centre (AMC) Initiative: Work Package I (High Temperature Materials
Development for 3D Additive Manufacturing) with project No.
1426800088 and the Australian Research Council (DE170100053 and
DP150101121). We would like to thank Dr. Zhang Xuan, Dr. Guo Jiang,
and Mr. Goh Minhao for the fruitful discussions and experimental assis-
tance. We would also like to acknowledge Dr. He Junyang for providing
us the tensile test data of the as cast HEAs.
24 Z.G. Zhu et al. / Scripta Materialia 154 (2018) 20–24
Fig. 4. (a) EBSD IPF, (b) KAM map, and (c, d) Bright-field STEM images of an as-built sample strained to fracture. The arrows show the deformation twins.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.scriptamat.2018.05.015.
References
[1] B. Cantor, I.T.H. Chang, P. Knight, A.J.B. Vincent, Mater. Sci. Eng. A 213 (2004)
375–377.
[2] J.W. Yeh, S.K. Chen, S.J. Lin, J.Y. Gan, T.S. Chin, T.T. Shun, C.H. Tsau, S.Y. Chang, Adv.
Eng. Mater. 6 (2004) 299.
[3] Y. Zhang, T.T. Zuo, Z. Tang, M.C. Gao, K.A. Dahmen, P.K. Liaw, Z.P. Lu, Prog. Mater. Sci.
61 (2014) 1.
[4] X. Fu, C.A. Schuh, E.A. Olivetti, Scr. Mater. 136 (2017) 145.
[5] Q.Y. Lin, X.H. An, H.W. Liu, Q.H. Tang, P.Q. Dai, X.Z. Liao, J. Alloys Compd. 709 (2017)
802.
[6] F. Otto, A. Dlouhý, Ch. Somsen, H. Bei, G. Eggeler, E.P. George, Acta Mater. 61 (2013)
5743.
[7] B. Gludovatz, A. Hohenwarter, D. Catoor, E.H. Chang, E.P. George, R.O. Ritchie, Sci-
ence 345 (2014) 1153.
[8] W.H. Liu, Y. Wu, J.Y. He, T.G. Nieh, Z.P. Lu, Scr. Mater. 68 (2013) 526.
[9] D.B. Miracle, O.N. Senkov, Acta Mater. 122 (2017) 448.
[10] S.T. Williams, I. Todd, Scr. Mater. 135 (2017) 105.
[11] D.D. Gu, W. Meiners, K. Wissenbach, R. Poprawe, Int. Mater. Rev. 57 (2012) 133.
[12] Y. Kok, X.P. Tan, P. Wang, M.L.S. Nai, N.H. Loh, E. Liu, S.B. Tor, Mater. Des. 139 (2018)
565.
[13] W.E. Frazier, J. Mater. Eng. Perform. 23 (2014) 1917.
[14] X.P. Li, X.J. Wang, M. Saunders, A. Suvorova, L.C. Zhang, Y.J. Liu, M.H. Fang, Z.H.
Huang, T.B. Sercombe, Acta Mater. 95 (2015) 74.
[15] Y.M. Wang, T. Voisin, J.T. McKeown, J.C. Ye, N.P. Calta, Z. Li, Z. Zeng, Y. Zhang, W.
Chen, T.T. Roehling, R.T. Ott, M.K. Santala, P.J. Depond, M.J. Matthews, A.V. Hamza,
T. Zhu, Nat. Mater. 17 (2018) 63.
[16] W. Xu, M. Brandt, S. Sun, J. Elambasseril, Q. Liu, K. Lathan, K. Xia, Q. Ma, Acta Mater.
85 (2015) 74.
[17] Y. Brif, M. Thomas, I. Todd, Scr. Mater. 99 (2015) 93.
[18] X.P. Li, Adv. Eng. Mater. (2017)https://doi.org/10.1002/adem.201700874.
[19] M. Laurent-Brocq, A. Akhatova, L. Perrière, S. Chebini, X. Sauvage, E. Leroy, Y.
Champion, Acta Mater. 88 (2015) 355.
[20] X.J. Wang, L.C. Zhang, M.H. Fang, T.B. Sercombe, Mater. Sci. Eng. A 597 (2014) 370.
[21] H. Zhang, H.H. Zhu, X.J. Nie, J. Yin, Z.H. Hu, X.Y. Zeng, Scr. Mater. 134 (2017) 6.
[22] L.F. Liu, Q.Q. Ding, Y. Zhong, J. Zou, J. Wu, Y.L. Chiu, J.X. Li, Z. Zhang, Q. Yu, Z.J. Shen,
Mater. Tod. (2017)https://doi.org/10.1016/j.mattod.2017.11.004.
[23] K.G. Prashanth, J. Eckert, J. Alloys Compd. 707 (2017) 27.
[24] T. Mukherjee, J.S. Zuback, A.De.T. DebRoy, Sci. Rep. 6 (2016) 19717.
[25] K. Saeidi, X. Gao, Y. Zhong, Z.J. Shen, Mater. Sci. Eng. A 625 (2015) 221.
[26] S. Singh, N. Wanderka, B.S. Murty, U. Glatzel, J. Banhart, Acta Mater. 59 (2011) 182.
[27] J.Y. He, W.H. Liu, H. Wang, Y. Wu, X.J. Liu, T.G. Nieh, Z.P. Lu, Acta Mater. 62 (2014)
105.
[28] K. Kempen, L. Thijs, J.V. Humbeeck, J.P. Kruth, Phys. Procedia 39 (2012) 439.
[29] G. Laplanche, A. Kostka, O.M. Horst, G. Eggeler, E.P. George, Acta Mater. 118 (2016)
152.
[30] J. Suryawanshi, K.G. Prashanth, U. Ramamurty, Mater. Sci. Eng. A 696 (2017) 113.
[31] S.J. Sun, Y.Z. Tian, H.R. Lin, X.G. Dong, Y.H. Wang, Z.J. Zhang, Z.F. Zhang, Mater. Des.
133 (2017) 122.
[32] U.F. Kocks, H. Mecking, Prog. Mater. Sci. 48 (2003) 171.