Materials Today Nano 1 (2018) 41e46

Contents lists available at ScienceDirect

Materials Today Nano

j o u r n a l h o m e p a g e : h t t p s : / / w w w . e v i s e . c o m / p r o fi l e / # / M T N A N O / l o g i n

Full Length Article

PdePt nanoalloy transformation pathways at the atomic scale*

Min Tang a, c, Beien Zhu b, c, Jun Meng b, Xun Zhang a, Wentao Yuan a, Ze Zhang a,
Yi Gao b, **, Yong Wang a, *
a
State Key Laboratory of Silicon Materials and Center of Electron Microscopy, School of Materials Science and Engineering, Zhejiang University, Hangzhou
310027 China
b
Division of Interfacial Water and Key Laboratory of Interfacial, Physics and Technology, Shanghai Institute of Applied Physics, Chinese Academy of Science,
Shanghai 201800 China

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

Article history: Nanoalloys have attracted considerable attention for their wide applications in materials, optics, catal-
Available online 14 April 2018 ysis, and biomedicine, which largely rely on their composition-, size-, and shape-dependent properties.
Because these properties change dynamically with working conditions, the knowledge of the complex
Keywords: transformation pathway of nanoalloys is highly demanded. Herein, we combined the in-situ aberration-
Transformation pathways corrected scanning transmission electron microscopy and multiscale modeling to fully resolve the whole
Palladiumeplatinum core-shell
transformation trajectory of a bimetallic nanoalloy (PdePt) at the atomic level. The transformation from
nanoparticles
core-shell to solid-solution structure is a multistep and temperature-dependent pathway, which includes
In-situ STEM
Alloying
monometallic-surface refacetting, bimetallic-surface refacetting, and alloy mixing, owing to the different
atomic activation barriers of surface diffusion and bulk migration. The critical role of shell thickness in
determining the transformation pathway was also revealed and explained. In particular, a unique partial
core-shell structure with an anisotropic surface pattern was observed in the annealing process of ul-
trathin core-shell nanoparticles. This study offers a fundamental insight into the structure evolution of
nanoalloys, which is beneficial for the development of functionalized nanoparticles with kinetic stability.
© 2018 Elsevier Ltd. All rights reserved.

Nanoalloys have attracted extensive attention for their potential
applications in nanocatalysis, nano-engineering, and nanomedicine
[1,2]. The unique properties of nanoalloys are determined not only
by their morphologies and sizes, as for other nanomaterials, but
also by their unique chemical compositions and atomic distribu-
tions [1e9]. For Pt-based bimetallic alloys, the solid solution
structures and core-shell structures exhibit distinct properties. The
PdePt solid-solution alloy nanoparticles (NPs) have a higher
hydrogen-storage capacity than the core-shell-type NPs [10], and
their core-shell NPs have been widely employed as model catalysts
for oxygen reduction reaction in fuel cells [11e15]. In particular, the
core-shell PdePt NPs with controlled shell thickness and
morphology have attracted great interests because of their distinct
catalytic properties [16e18]. Interestingly, recent in-situ
* et al. presented a breakthrough work for mapping the short-lived
Given his role as Editor-in-Chief, Ze Zhang had no involvement in the peer-
review of this article and has no access to information regarding its peer-review.
Full responsibility for the editorial process for this article was delegated to
Zhifeng Ren of Materials Today Physics.
* Corresponding author.
** Corresponding author.
E-mail addresses: gaoyi@sinap.ac.cn (Y. Gao), yongwang@zju.edu.cn (Y. Wang).
c
These authors give the same contributions to this work.
experiments have shown that both structures and compositions of
nanoalloys may considerably transform during thermal treatment
and under real reaction conditions [19e26]. For instance, Vara et al.
recently reported that both the shape and the mixing pattern of
PdePt core shell NPs change at a high temperature, consequently
changing their catalytic activities [27]. As a result, understanding
how the structure of a controlled core-shell nanoalloy changes
under a given condition is crucial for the effective tuning of prop-
erties, which requires detailed information about the complex
transformation pathways.
In the past decade, much effort has been contributed to study
the complex pathways of formation [28], transformation [29e32],
and dissolution [33,34] of individual nanoscale building blocks. Chi
et al. identified the distinct stages of surface element rearrange-
ments of Pt3Co NPs under thermal treatment [35]. Very recently, Ye
intermediate state of single gold NPs [36] to provide the possibil-
ity of designing a kinetically stabilized surface. However, despite
the above achievements, only limited information is available on
the transformation pathways of nanoalloys because it remains a
challenge to monitor the structure evolution of a single NP during
the long-time annealing process at the atomic scale.

https://doi.org/10.1016/j.mtnano.2018.04.003
2588-8420/© 2018 Elsevier Ltd. All rights reserved.
42 M. Tang et al. / Materials Today Nano 1 (2018) 41e46
In this study, we employed in situ scanning transmission elec-
tron microscopy (STEM) annealing experiments to fully monitor
the whole pathway of the transformation of PdePt NPs from
segregated core-shell structures to randomly mixed structures. The
dynamic evolution of morphology, elemental diffusion, and alloy
mixing for individual NPs during annealing was tracked as a
function of temperature and time, by high-angle annular dark-field
(HAADF) STEM imaging. The temperature-dependent trans-
formation pathways were identified and rationalized by the density
functional theory calculations and multiscale modeling. This study
facilitates the tailoring of the physical and chemical properties of
nanoalloys for catalytic applications.
The core-shell PdePt cubic NPs were prepared by a chemical
method reported in a previous work [37] (see Supplemental
Material for details). The highly monodispersed PdePt NPs (to
prevent the nanoparticle coalescence) were heated on a silicon
nitride film using a double-tilt DENS TEM holder [38]. Because the
core-shell structures are not the equilibrium states of NPs, the
structures will transform during the annealing process. The in situ
annealing experiments were performed in an aberration-corrected
STEM (Titan ChemiSTEM, FEI). During the long-time heating pro-
cess, special attention has been paid to obtain atomic-resolution
HAADF-STEM images as NPs rotate frequently at a high tempera-
ture. Besides, EDX mapping has also been performed at the same
temperature to acquire in-situ information of the elemental dis-
tribution of PdePt alloys.
The steps in the annealing process are shown in Fig. 1(a), where
the temperature was increased to 600
C over 400 min. Fig. 1(b)
shows the atomic HAADF-STEM image of a corner of a typical PdePt
Fig. 1. (a) Annealing sequence for in situ experiments, plotted as the thermal annealing temperature with respect to time. The specific annealing conditions for the HAADF-STEM
images shown in Fig. 2 are labeled with red stars in this plot. (b) Atomic-resolution HAADF-STEM image of a corner of a typical PdePt NP at room temperature. (ced) HAADF-STEM
image and corresponding EDX elemental map of a PdePt NP at RT.
NP, which reveals the continuous lattice fringes from the Pd core to
the Pt shell, indicating an epitaxial relationship between them. As
shown in Fig. 1(b), the interplanar spacing of 1.945 and 1.415 Å on
the edge and corner of the cubic NP can be indexed to FCC Pt {100}
and {110} planes, respectively. No misfit dislocation was observed
because the lattice mismatch between Pt and Pd is 0.8% only. The
distribution of chemical components in the bimetallic PdePt cubic
NPs at room temperature was assessed using STEM/EDX mapping,
as shown in Fig. 1(c and d), respectively. As there is a relatively large
difference in the atomic number of 78Pt and 46Pd, the contrast
present in the atomic-scale HAADF-STEM images can be used to
analyze the elemental distribution of Pt and Pd [39,40]. Further
quantification of the distribution is very challenging for large-sized
nanoalloy particles. It is out of the scope of this work and needs
further studies in the future [41,42].
Fig. 2(a) shows the STEM image of a PdePt cube kept at 300
C
for 140 min. No obvious structural change was observed. With the
increase in temperature to 400
C, a refacetting of the core-shell NP
without alloy mixing was observed. As shown in Fig. 2(b, d), the
morphology of the monitored NP became more and more round
with time. The ultra-high resolution images (see the second and
third rows of Fig. 2) clearly show that the surface refacetting
resulted from the changes in the Pt shell thickness on the {110} and
{100} surfaces. In the beginning, the thicknesses of the atomic
layers of Pt shell were the same on the {110} and {100} facets. The Pt
shell thickness on the {100} facet increased from 4 atomic layers at
300
C to 6 layers, 9 layers, 11 layers after 30, 35, and 55 min at
400
C, respectively. Furthermore, the Pt shell thickness on the
{110} facet decreased from 4 atomic layers to 0 atomic layers. The
 M. Tang et al. / Materials Today Nano 1 (2018) 41e46
Fig. 2. (aef) Atomic-resolution HAADF-STEM images of a single PdePt NP acquired at different temperatures and annealing times. (gel; mer) Corresponding details of {110} and
{100} facets, respectively. It should be noted that the electron beam may induce some contamination (the bright area in the center of the particle) when we focus the beam on the
particle to adjust the zone axis to obtain high-resolution HAADF images (for more details, please refer Figure S5).
STEM images and the EDX elemental mapping in Fig. 3(aec) show
that during refacetting, there was no obvious migration of the Pt
atoms into the bulk, which indicates that the surface refacetting
results from the surface diffusion of Pt atoms. Note that the Pt
atoms seen in the middle of the particle in Fig. 3(c) resulted from
the overlapping between the Pt(100) surface parallel to the STEM
Fig. 3. HAADF-STEM images and corresponding EDX elemental maps of a PdePt NP,
which was annealed at different temperatures: 400
C (aec), 500
C (def), and 600
C
(gei).
43
image and the Pd bulk. When the sample temperature was
increased to 500
C and held for 75 min, the migration of Pt atoms
into the core of the NP was observed as evidenced by the EDX
mapping (Fig. 3(def)), where more Pt atoms are observed in the
middle of the particle, compared with those in Fig. 3(c). Interest-
ingly, on the {100} facets, Pt shell with 11 atomic layers still re-
mains, while the area of the Pd {110} facets increases. This further
refacetting makes the morphology of the nanoalloy more round
(shown by yellow arrows in Fig. 2(e, q)). When the temperature was
increased to 600
C, the transformation of the monitored NP from a
core-shell structure to a random mixed structure was completed.
The EDX mapping in Fig. 3(gei) shows that Pt and Pd atoms were
homogeneously mixed. To exclude the effect of beam irradiation,
the morphology of other NPs (only exposed to electron beam when
taking images) at a high temperature was also acquired (refer to
Figures S1 and S3, Supplemental Material), showing a similar
phenomenon as observed in Fig. 2, which implies that beam irra-
diation is not a dominant factor in the refacetting and alloying
process.
To understand the experimental results, the refacetting process
observed at 400
C is interpreted first. We divide the refacetting
process into two stages. In the first stage, the Pt shell becomes
thinner on the {110} facets, while it becomes thicker on the {100}
facets, indicating the directional movement of Pt atoms from the
{110} facets to {100} facets. The Pt shell remains complete until a
monolayer Pt is present on the {110} facets (Fig. 2(aec)). Thus, this
stage is defined as the monometallic-surface refacetting stage. It can
be understood that by chamfering the cube, the surface energy re-
duces. The initial cubic structure is obviously much different from
its equilibrium structure, which is predicted to be a truncated oc-
tahedron composed mainly of {111} and {100} facets according to
the well-known Wulff rule [43e45]. Because the STEM image can
offer only 2-D structure information but not the {111} facet fraction,
it would be difficult to directly compare the refacetted structures
with the Wulff constructed structure. However, we can still find the
44 M. Tang et al. / Materials Today Nano 1 (2018) 41e46
reason for refacetting by measuring the ratio of d100 to d110 (R100/110),
where d100 and d110 are the distances from the center of the particle
to the {100} and {110} facets, respectively. When the NP reaches its
equilibrium structure, the R100/110 should be equal to the ratio of
g100 to g110 (g the surface tension) as stated by George Wulff. The
measured R100/110 from Fig. 2(aec) is 0.712, 0.793, and 0.839
respectively, while the ideal R100/110 calculated by the surface ten-
sion of Pt (100) and (110) surfaces is 0.917. Thus, it can be concluded
that the NP tends to reach its equilibrium shape by directional
surface atom diffusion. Interestingly, the second refacetting stage
named as the bimetallic-surface refacetting stage is found in our
experiments. At the end of the first stage, the refacetting process is
not finished because the R100/110 is considerably different from the
ideal one. However, because there is only a monolayer Pt on the
{110} facets, further chamfering exposes the Pd core on the {110}
facets. Because Pd has a lower surface tension than Pt [44,45], this
exposure of Pd core triggers an expansion of Pd atoms on the {110}
facets, which squeezes the Pt {100} facets, resulting in a more round
shape.
As observed from our experiments, the whole transformation
process of the cubic PdePt core-shell nanoalloy involves two indi-
vidual processes, i.e. the surface refacetting process and the alloying
process. Two energy barriers play important roles in these two
processes, i.e. the diffusion energy barriers (DED) of the Pt atoms on
the {110} and {100} surfaces and the migration energy barriers
(DEM) of the Pt atom from Pt subsurface layer to the Pd core. DFT
calculations were performed to calculate the energy barriers by
using the Vienna ab initio simulation package (VASP) [46]. The GGA-
PBE functional and the projector augmented-wave methods were
adopted [47,48]. The transition state during the diffusion or
migration was found by the nudged elastic band method imple-
mented in VASP. More details can be seen in the Supplemental
Material. The results showed that DED is 0.86 eV, while DEM is
approximately 0.94 eV. The corresponding energy pathways can be
found in Figure S4 in the Supplemental Material. At 400
C, DED can
be overcome, while it is difficult to overcome DEM. Therefore, we
consider that the refacetting process occurs without alloy mixing.
When the temperature is relatively higher (500
C and 600
C), both
the energy barriers can be overcome. Thus, it can be concluded that
the alloying began at 500
C (Fig. 3(def)) and finished completely at
600
C (Fig. 3(gei)). It can be predicted that if a high temperature
treatment is applied without an annealing process, the trans-
formation process could thus be a one-step process. This was
confirmed by an experiment where the NPs were heated directly to
600
C, where we observed the surface refacetting and alloying
simultaneously (Figures S5 and S6).
The above analysis indicates that (1) the difference between the
surface diffusion barrier and the bulk migration barrier makes the
transformation process a selective multistep process. (2) The sur-
face energy plays an important role in the surface refacetting pro-
cess. To confirm our understanding, an effective kinetic Monte
Carlo (KMC) simulation model is established to reproduce the
experimental observations. In each KMC trial, a random atom i is
picked to move it to its nearby empty crystal lattice site. The
acceptance rate of this trial is controlled by both the energy barrier
and the energy balance of the movement. The possibility of over-
coming the diffusion/migration energy barrier is determined first.
DED is used if it is surface diffusion, while DEM is used if it is bulk
migration. By doing this, the surface diffusion and the bulk
migration are selected. If this overcomes the energy barrier, then
the energy change for moving the atom is further determined. To
reduce the computational cost to make the simulation of large-
sized NPs feasible, we adopted a simplified model to describe the
change in energy. The model assumes that the contributions to the
surface energy are from the missing first-neighbor bonds [49]. The
PtePt (PdePd) first-neighbor bonding energy (εptpt : 0.42 eV,
εpdpd : 0.34 eV) was fitted using the surface energies calculated at
DFT level (Table S1). Moreover, the PdePt bonding energy
(εptpd : 0.38 eV) was fitted using the surface segregation energy of
Pt on the Pd surface. A series of model systems were tested using
both the DFT calculations and fitting parameters. The difference
between the DFT results and the fitted ones was reasonably small
(Table. S2). All details are provided in the Supplemental Material.
The change in the energy caused by the KMC trial is considered by
counting the number of broken and formed nearest-neighbor
bonds. For example, if a Pt atom jumps from position 1 to posi-
tion 2, then
   
DE ¼ εPtPt CNPt
2 2
þ εPdPt CNPd 1
 εPtPt CNPt 1
þ εPdPt CNPd (1)
where CNPt 1 /CN 1 and CN 2 /CN 2 are the numbers of first-neighbor
Pd Pt Pd
Pt/Pd atoms before and after the movement. A Metropolis-
Hastings algorithm was then carried out on the basis of the
energy barrier and the energy balance. If the trial is accepted
eventually, we refer it as an effective KMC step.
The simulated NP had a size of about 9 nm, containing 45,563
atoms and 2 Pt layers. A 60-million-KMC-step simulation at 400
C
was performed first to simulate the surface refacetting process. It
can be seen from Fig. 4(b and c) and Supplemental Material that our
simulation well reproduces the experimental observations: (1) The
Pt shell on {100} facets became thicker with time. (2) The direc-
tional diffusion of Pt atoms from {110} facets to {100} facets
exposed the Pd atoms on {110} facets. Besides, the {111} facets were
found at the truncated vertexes. Further statistical analysis showed
that the number of atoms on {110} facets increased with KMC steps,
thus representing the expansion of {110} facets. Furthermore, the
concentration of Pd atoms on the exposed {110} facets increased
from 9% at 10 M KMC steps to 76% at 60 M KMC steps. It can also be
observed that the partial core-shell structure showed some sta-
bility after 50 M KMC steps, and no large change was found in the
structure and composition. The annealing simulation was per-
formed by 50 M KMC steps at each temperature (400
C, 500
C,
and 600
C). As expected, the alloying process began at 500
C
(Figure S9). The final structure at 600
C was a random mixed round
structure (Fig. 4(d)). Using the simulation results, the whole
multistep PdePt core-shell alloy transformation is interpreted in
Fig. 4.
As discussed above, the goal of the refacetting process is to
change the controlled shape to the equilibrium shape. If the shell
thickness of the NP is relatively larger than its size, it may reach the
equilibrium structure without the bimetallic-surface refacetting
process. To confirm this idea, we first ran the developed KMC
simulation using a NP of the same size but with 4 Pt layers. The
results showed that the unique partial core-shell structure did not
appear for this case (Fig. 5, Figure S10 in the Supplemental
Material). Followed by the theoretical simulation results, we per-
formed similar experiments using NPs with thicker Pt shells.
Consistent results were obtained: in the surface refacetting process,
the shell was too thick to expose the subsurface Pd atoms. It is
known that the catalytic property of the core-shell NP is very
sensitive to its shell thickness. Besides the known effect of shell
thickness on the electronic structure, we show here that the shell
thickness could also change the intermediates appearing in the
transformations of a core-shell structure. Some unique structures
may appear in the case of an ultrathin core-shell structure. For
example, the partial core-shell structure found in our study may
have a special catalytic property for some kinds of reactions
because NPs have selective Pt and Pd facets on the surface
simultaneously.
 M. Tang et al. / Materials Today Nano 1 (2018) 41e46 45

Fig. 4. Selective mechanism of the transformation from core-shell structure to mixed alloys. The 3-D structures were obtained from the kinetic Monte Carlo simulations. Orange
balls: Pt atoms; blue balls: Pd atoms.

In summary, by monitoring the transformation pathway of a monometallic-surface refacetting, bimetallic-surface refacetting,

single PdePt NP from a cubic core-shell structure to a round alloyed and alloy mixing. Depending on the heating treatments, these
structure, we revealed a selective mechanism for alloy trans- stages may occur in a stepwise manner or simultaneously. By the
formation. Three main stages may occur during the transformation: theoretical analysis and kinetic Monte Carlo simulations, we

Fig. 5. (1e6) KMC simulated results of similar transformation of the 9-nm PdePt NP with 4 Pt shells, 50 M effective KMC steps at each temperature. Orange balls: Pt atoms, blue
balls: Pd atoms. (aef) Experimentally observed alloy transformation of the PdePt NP with thicker shells.
46 M. Tang et al. / Materials Today Nano 1 (2018) 41e46
further explained that the refacetting process occurs with the
purpose of transforming a NP from a controlled structure to its
equilibrium structure. We also showed that the shell thickness
plays an important role in deciding the transformation pathway.
Several intermediates with anisotropic shell thickness were char-
acterized during the refacetting process in situ observations. In
particular, a unique partial core-shell structure with an anisotropic
surface pattern was observed with the ultrathin layer core-shell NP.
This work helps to understand and control the alloy transformation
pathway and reveals the potential usage of the intermediates in
future studies.
Acknowledgments
We acknowledge the support of the National Nature Science
Foundation of China (51390474, 11574340, 11604357, 91645103,
11234011, 11327901, and 21773287). B.Z. thanks the Natural Science
Foundation of Shanghai (16ZR1443200). The authors thank Prof.
Shengbai Zhang and Prof. Riccardo Ferrando for helpful discussions.
The computational resources utilized in this research were pro-
vided by Shanghai Supercomputer Center, National Super-
computing Center in Tianjin and Shenzhen, and Special Program for
Applied Research on Super Computation of the NSFC-Guangdong
Joint Fund (the second phase) under Grant No. U1501501.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.mtnano.2018.04.003.
References
[1] R. Ferrando, J. Jellinek, R.L. Johnston, Chem. Rev. 108 (3) (2008) 845e910.
[2] D.S. Su, B. Zhang, R. Schlogl, Chem. Rev. 115 (2015) 2818e2882.
[3] B.T. Sneed, C.N. Brodsky, C.H. Kuo, L.K. Lamontagne, Y. Jiang, Y. Wang, F.F. Tao,
W. Huang, C.K. Tsung, J. Am. Chem. Soc. 135 (2013) 14691e14700.
[4] L. Li, L. Zhou, S.O. Chikn, D.H. Anjum, M.B. Kanoun, J. Scaranto, M.N. Hedhili,
S. Khalid, P.V. Laveille, L.D. Souza, A. Clo, J.M. Basset, ChemCatChem 7 (2015)
819e829.
[5] B.T. Sneed, A.P. Young, D. Jalalpoor, M.C. Golden, S. Mao, Y. Jiang, Y. Wang,
C.K. Tsung, ACS Nano 8 (7) (2014) 7239e7250.
[6] Y. Yang, L. Lai, Y. Guo, Z. Dai, R. Zhang, C. Sun, X. Zhou, J. Electroanal. Chem.
783 (2016) 132e139.
[7] S.K. Konda, A. Chen, Mater. Today 19 (2016) 100e108.
[8] S.B. Vendelbo, C.F. Elkjær, H. Falsig, I. Puspitasari, P. Dona, L. Mele, B. Morana,
B.J. Nelissen, R. Rijn, J.F. Creemer, P.J. Kooyman, S. Helveg, Nat. Mater. 13
(2014) 884e890.
[9] Y. Yue, D. Yuchi, P. Guan, J. Xu, L. Guo, J. Liu, Nat. Commun. 7 (2016) 12251.
[10] H. Kobayashi, M. Yamauchi, H. Kitagawa, Y. Kubota, K. Kato, M. Takata, J. Am.
Chem. Soc. 132 (2010) 5576e5577.
[11] J. Park, L. Zhang, S.I. Choi, L.T. Roling, N. Lu, J.A. Herron, S. Xie, J. Wang,
M.J. Kim, M. Mavrikakis, Y. Xia, ACS Nano 9 (3) (2015) 2635e2647.
[12] K. Sasaki, H. Naohara, Y. Cai, Y.M. Choi, P. Liu, M.B. Vukmirovic, J.X. Wang,
R.R. Adzic, Angew. Chem. Int. Ed. 49 (2010) 8602e8607.
[13] J.X. Wang, H. Inada, L. Wu, Y. Zhu, Y. Choi, P. Liu, W.P. Zhou, R.R. Adzic, J. Am.
Chem. Soc. 131 (2009) 17298e17302.
[14] S. Xie, S.I. Choi, N. Lu, L.T. Roling, J.A. Herron, L. Zhang, J. Park, J. Wang,
M.J. Kim, Z. Xie, M. Mavrikakis, Y. Xia, Nano Lett. 14 (2014) 3570e3576.
[15] J. Zhang, M.B. Vukmirovic, Y. Xu, M. Mavrikakis, R.R. Adzic, Angew. Chem. Int.
Ed. 44 (2005) 2132e2135.
[16] B. Lim, M. Jiang, P.H. Camargo, E.C. Cho, J. Tao, X. Lu, Y. Zhu, Y. Xia, Science 324
(5932) (2009) 1302e1305.
[17] P. Strasser, S. Koh, T. Anniyev, J. Greeley, K. More, C. Yu, Z. Liu, Nat. Chem. 2
(2010) 454e460.
[18] D. Wang, H.L. Xin, R. Hovden, H. Wang, Y. Yu, D. Muller, F. DiSalvo, H. Abruna,
Nat. Mater. 12 (2013) 81e87.
[19] F. Tao, M.E. Grass, Y. Zhang, D.R. Butcher, J.R. Renzas, Z. Liu, J.Y. Chung,
B.S. Mun, M. Salmeron, G.A. Somorjai, Science 322 (5903) (2008) 932e934.
[20] C. Cui, L. Gan, M. Heggen, S. Rudi, P. Strasser, Nat. Mater. 12 (2013) 765e771.
[21] H. Zheng, J. Wang, J.Y. Huang, J. Wang, Z. Zhang, S.X. Mao, Nano Lett. 13 (12)
(2013) 6023e6027.
[22] Y. Jiang, H. Li, Z. Wu, W. Ye, H. Zhang, Y. Wang, C. Sun, Z. Zhang, Angew. Chem.
Int. Ed. 55 (2016) 12427e12430.
[23] J. Villanova, R. Daudin, P. Lhuissier, D. Jauffres, S. Lou, C.L. Martin, S. Laboure,
R. Tucoulou, G.M. Criado, L. Salvo, Mater. Today 20 (2017) 354e359.
[24] Y.A. Wu, L. Li, Z. Li, A. Kinaci, M.K. Chan, Y. Sun, J.R. Guest, I. McNulty, T. Rajh,
Y. Liu, ACS Nano 10 (2016) 3738e3746.
[25] S. Prabhudev, M. Bugnet, G.Z. Zhu, C. Bock, G.A. Botton, ChemCatChem 7
(2015) 3655e3664.
[26] J. Sun, L. He, Y.C. Lo, T. Xu, H. Bi, L. Sun, Z. Zhang, S.X. Mao, J. Li, Nat. Mater. 13
(2014) 1007e1012.
[27] M. Vara, L.T. Roling, X. Wang, A.O. Elnabawy, Z.D. Hood, M. Chi, M. Mavrikakis,
Y. Xia, ACS Nano 11 (2017) 4571e4581.
[28] H. Zheng, R.K. Smith, Y.W. Jun, C. Kisielowski, U. Dahmen, A.P. Alivisatos,
Science 324 (5932) (2009) 1309e1312.
[29] S.H. Tolbert, A.P. Alivisatos, Science 265 (5170) (1994) 373e376.
[30] P. Buffat, J.P. Borel, Phys. Rev. A 13 (1976) 2287e2298.
[31] D.H. Son, S.M. Hughes, Y. Yin, A. Paul Alivisatos, Science 306 (5698) (2004)
1009e1012.
[32] Y. Sun, Y. Xia, Science 298 (5601) (2002) 2176e2179.
[33] M.J. Mulvihill, X.Y. Ling, J. Henzie, P. Yang, J. Am. Chem. Soc. 132 (2010)
268e274.
[34] M.N. O'Brien, M.R. Jones, K.A. Brown, C.A. Mirkin, J. Am. Chem. Soc. 136 (2014)
7603e7606.
[35] M. Chi, C. Wang, Y. Lei, G. Wang, D. Li, K.L. More, A. Lupini, L.F. Allard,
N.M. Markovic, V.R. Stamenkovic, Nat. Commun. 6 (2015) 8925.
[36] X. Ye, M.R. Jones, L.B. Frechette, Q. Chen, A.S. Powers, P. Ercius, G. Dunn,
G.M. Rotskoff, S.C. Nguyen, V.P. Adiga, A. Zettl, E. Rabani, P.L. Geissler,
A.P. Alivisatos, Science 354 (6314) (2016) 874e877.
[37] N. Lu, J. Wang, S. Xie, J. Brink, K. McIlwrath, Y. Xia, M.J. Kim, J. Phys. Chem. C
118 (2014) 28876e28882.
[38] Y. Jiang, Y. Wang, Y.Y. Zhang, Z. Zhang, W. Yuan, C. Sun, X. Wei, C.N. Brodsky,
C.-K. Tsung, J. Li, X. Zhang, S.X. Mao, S. Zhang, Z. Zhang, Nano Res. 7 (2014)
308e314.
[39] Y. Jiang, Y. Wang, J. Sagendorf, D. West, X. Kou, X. Wei, L. He, K.L. Wang,
S. Zhang, Z. Zhang, Nano Lett. 13 (2013) 2851e2856.
[40] S.I. Sanchez, M.W. Small, J. Zuo, R.G. Nuzzo, J. Am. Chem. Soc. 131 (2009)
8683e8689.
[41] A. Backer, G.T. Martinez, A. Rosenauer, S.V. Aert, Ultramicroscopy 134 (2013)
23e33.
[42] K.H.W. van den Bos, A. Backer, G. Martinez, N. Winckelmans, S. Bals, P. Nellist,
S.V. Aert, Phys. Rev. Lett. 116 (2016) 246101.
[43] G.Z. Wulff, Z. Kristallogr, Cryst. Mater 34 (1901) 449.
[44] B. Zhu, J. Meng, Y. Gao, J. Phys. Chem. C 121 (2017) 5629e5634.
[45] B. Zhu, Z. Xu, C. Wang, Y. Gao, Nano Lett. 16 (2016) 2628e2632.
[46] G. Kresse, J. Hafner, Phys. Rev. B 47 (1) (1993) 558e561.
[47] P.E. Blochl, O. Jepsen, O.K. Andersen, Phys. Rev. B 49 (23) (1994)
16223e16233.
[48] G. Kresse, D. Joubert, Phys. Rev. B 59 (3) (1999) 1758e1775.
[49] W.F. Hosford, Materials Science: an Intermediate Text, Cambridge University
Press, 2006.