doi.org/10.26434/chemrxiv.8289812.v1

Ordered Intermetallic Nanoparticles with High Catalytic Activity Preparedby an Electrochemically Induced Phase TransformationDu Sun, yunfei wang, Kenneth Livi, chuhong wang, ruichun luo, zhuoqun zhang, hamdan alghamdi, chenyangli, fufei an, bernard gaskey, Tim Mueller, Shoji Hall

Submitted date: 18/06/2019 • Posted date: 19/06/2019Licence: CC BY-NC-ND 4.0Citation information: Sun, Du; wang, yunfei; Livi, Kenneth; wang, chuhong; luo, ruichun; zhang, zhuoqun; et al.(2019): Ordered Intermetallic Nanoparticles with High Catalytic Activity Prepared by an ElectrochemicallyInduced Phase Transformation. ChemRxiv. Preprint.

The synthesis of alloys with long range atomic scale ordering (ordered intermetallics) is an emerging field ofnanochemistry. Ordered intermetallic nanoparticles are useful for a wide variety of applications such ascatalysis, superconductors, and magnetic devices. However, the preparation of nanostructured orderedintermetallics is challenging in comparison to disordered alloys, hindering progress in materials development.We report a process for converting colloidally synthesized ordered intermetallic PdBi2 to ordered intermetallicPd3Bi nanoparticles under ambient conditions by an electrochemically induced phase transition. The lowmelting point of PdBi2 corresponds to low vacancy formation energies which enables the facile removal of theBi from the surface, while simultaneously enabling interdiffusion of the constituent atoms via a vacancydiffusion mechanism under ambient conditions. The resulting phase-converted ordered intermetallic Pd3Biexhibits 11x and 3.5x higher mass activty and high methanol tolerance for the oxygen reduction reactioncompared to Pt/C and Pd/C, respectively,which is the highest reported for a Pd-based catalyst, to the best ofour knowledge. These results establish a key development in the synthesis of noble metal rich orderedintermetallic phases with high catalytic activity, and sets forth guidelines for the design of ordered intermetalliccompounds under ambient conditions.

File list (3)

download fileview on ChemRxivSun-phase-convert-VF.docx (7.51 MiB)

download fileview on ChemRxivSun-phase-convert-VF.pdf (1.18 MiB)

download fileview on ChemRxiv2019-Sun-phase-convert-SI-.docx (4.46 MiB)

Ordered Intermetallic Nanoparticles with High CatalyticActivity Prepared by an Electrochemically Induced Phase

TransformationDu Sun,a‡ Yunfei Wang,a‡ Ken Livia, Chuhong Wang,a Ruichun Luo,b Zhuoqun Zhang,a

Hamdan Alghamdi,c Chenyang Li,a Fufei An, a,b Bernard Gaskey,a Tim Mueller,a Anthony Shoji Halla*

a Department of Materials Science and Engineering, Johns Hopkins University, Baltimore, Maryland 21218b School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai 200030, P. R. Chinac Department of Chemical and Biomolecular Engineering, Johns Hopkins University, Baltimore, Maryland 21218.‡Denotes co-first authors ORCID ID: A.S. Hall - 0000-0003-4134-4160

KEYWORDS Oxygen reduction reaction • Palladium • Ordered intermetallic • Bismuth • Phase conversion • Electrocatalysis• Alloy • dealloying

ABSTRACT: The synthesis of alloys with long range atomic scale ordering (ordered intermetallics) is anemerging field of nanochemistry. Ordered intermetallic nanoparticles are useful for a wide variety ofapplications such as catalysis, superconductors, and magnetic devices. However, the preparation ofnanostructured ordered intermetallics is challenging in comparison to disordered alloys, hinderingprogress in materials development. We report a process for converting colloidally synthesized orderedintermetallic PdBi2 to ordered intermetallic Pd3Bi nanoparticles under ambient conditions by anelectrochemically induced phase transition. The low melting point of PdBi2 corresponds to low vacancyformation energies which enables the facile removal of the Bi from the surface, while simultaneouslyenabling interdiffusion of the constituent atoms via a vacancy diffusion mechanism under ambientconditions. The resulting phase-converted ordered intermetallic Pd3Bi exhibits 11x and 3.5x higher massactivty and high methanol tolerance for the oxygen reduction reaction compared to Pt/C and Pd/C,respectively,which is the highest reported for a Pd-based catalyst, to the best of our knowledge. Theseresults establish a key development in the synthesis of noble metal rich ordered intermetallic phaseswith high catalytic activity, and sets forth guidelines for the design of ordered intermetallic compoundsunder ambient conditions.

IntroductionNoble metal rich phases of alloys have recently emerged as

high performance catalysts for a variety of industrially relevantcatalytic applications. The mixing of two different metals canresult in a solid solution that exhibits short-range ordering ofatoms (disordered alloy) or an ordered intermetallic that exhibitsboth short-range and long-range ordering of atoms. Orderedintermetallic compounds (OICs) can have superior catalyticperformance and stability compared to disordered alloys becauseof the atomically uniform active sites provided by long-rangestructural periodicity and stronger interatomic bonding from amore negative enthalpy of mixing.1-7

OICs are difficult to prepare, often requiring annealing totemperatures of ~700 oC to facilitate inter-diffusion andequilibration of atoms.7 Direct synthesis of some ordered

intermetallic nanostructures has been achieved at a relativelymild temperature by using either harsh reducing agents, specialreaction parameters such as microwave or solvothermalconditions, or electrochemical deposition.7-10 Moreover, noble-metal rich phases (>50% noble metal) are particularly challengingto synthesize in mild conditions, even if the ordered phases arethermodynamically favored, given the high melting points of thenoble metals and low diffusion rates in crystal lattices whichinhibit equilibration.6

The preparation of highly dispersed OIC catalysts byannealing at high temperatures is challenging because sinteringoccurs during annealing, reducing the surface area to volumeratio. The most common method for reducing sintering is tophysically isolate the particles by encapsulation with a templatebefore annealing. After crystallization, the templates can be

chemically etched away.11-13 The ability to control the shape ofthe particle is lost since the most thermodynamically stablefacet is typically expressed from surface reconstructions whichoccur during the phase transformation.14 Additionally,contamination can occur at elevated temperatures since thiscan allow atoms from the support or template to diffuse intothe alloy. The preparation of catalytically active OICnanostructures under mild conditions remains a greatchallenge in catalysis science.

Studies focusing on methods for preparing of OIC

nanoparticles are limited in comparison to disordered alloys.Recently, we reported that ordered intermetallic Pd31Bi12 canbe prepared directly by electrodeposition under ambientconditions; the resulting materials exhibit the highest specificactivity for the oxygen reduction reaction measured on a Pd-based material in alkaline electrolytes.10 However, the massactivity (an important technological metric for fuel celldevices) of electrochemically deposited Pd31Bi12 is low becausethis method affords a thin-film morphology, which has as a lowsurface area to volume ratio. Moreover, controlling thecomposition of a noble/non-noble metal alloys byelectrochemical deposition is challenging because of largedifferences in reduction potential between the twoconstituent elements. If OICs are to be practically utilized fordevices, a method which allows the preparation of OICnanoparticles under ambient conditions is needed.

It is known in the corrosion literature that the atoms ofalloys with low melting points exhibit high diffusivity duringde-alloying at room temperature.15-16 In contrast, materialswith high melting points (m.p. > 1000 oC) have low diffusivity;during de-alloying a thin noble metal film will form over thesurface, protecting the surface against further de-alloyingacross a wide range of voltages. This has been observed inbulk alloys such as Cu-Pd, Cu-Au, etc; and on nanoparticlealloys in which a noble metal shells forms over the alloy coresin systems such as Pt3Co, Cu3Pt, Pt3Ni, FeAu, etc.4, 15-19 Fromthese design principles, it can be inferred that the controlledcorrosion of a non-noble element from a low melting pointalloy nanoparticle could permit conversion to a noble metalrich OIC.

Here, we validate this concept by illustrating the kineticallycontrolled conversion of ordered intermetallic PdBi2 (m.p.~480 oC) nanoparticles to ordered intermetallic Pd3Binanoparticles via electrochemical de-alloying at roomtemperature. The low melting point of PdBi2 indicates that ithas low vacancy formation energies, which enables the facileremoval of the Bi from the surface while simultaneouslyenabling interdiffusion of the constituent atoms via a vacancydiffusion mechanism at room temperature.20-21 This methodallows for the entire particle to reconstruct to a noble metalrich OIC phase which has never been prepared under ambient

conditions. The phase converted-Pd3Bi exhibits record

breaking mass activity for the oxygen reduction reaction

(ORR), reaching activities as high as 1.2 A mgPd−1

at 0.9 V

vs the Reversible Hydrogen Electrode (RHE), which is ~3.4xthan Pd/C, and 11x higher than Pt/C, respectively.

Results and Discussion

Synthesis and characterization of PdBi2. The PdBi2

nanostructures were synthesized by reduction of Pd(II)

acetylacetonate and Bi(III) acetate in oleylamine at 300 0C.

PdBi2 has two known layered phases denoted as and

PdBi2 and PdBi2.22-23 The peaks in the experimental X-

ray diffraction (XRD) pattern exhibit broadening (Figure

1a), suggesting the crystallite size is in the nanoscale. The

pattern can be indexed to PdBi2 with a small amount of

PdBi2. The (011) deflection is more intense in

synthesized PdBi2 than the simulated pattern, indicating

preferential orientation along this direction. Note that

there is a shift of some of the peaks, which may be caused

by surface tension; this is commonly seen in

nanomaterials with small crystallites.24-25

The morphology and crystalline phases of PdBi2 were

probed by high-resolution transmission electron

microscopy (HRTEM). The sizes of the PdBi2 particles

ranged from ~10 to 50 nm in diameter (Figure 1b and

Figure SI1). The particles have a core-shell structure

where the core is single crystalline and the shell is

amorphous (Figure 1c). The atomic arrangement of the

core matches the (100) plane of PdBi2, the amorphous

shell is Bi (Figure SI2). The selected area electron

diffraction (SAED) was presented as a 1D line plot (Figure

1d) and compared to the 1D diffraction pattern of

PdBi2, both patterns match well, indicating that the

crystalline phases present in the material are PdBi2.

Figure 1. (a) Powder X-ray diffraction pattern of the as-prepared sample. The crystal planes that are corresponding to alpha-PdBi2 are marked in red, while the ones to beta-PdBi2

are in green. (b) TEM shows a representative image of the as-prepared nanoparticles. (c) High Angle Annular Dark Field-Scanning Transmission Electron Microscopy (STEM-HAADF)

image reveals the atomic arrangement of the nanoparticles, the atoms match with (100) plane of beta-PdBi2. (d) Line plot of intensity vs reciprocal lattice spacing derived from the

Electrochemical Phase Conversion of PdBi2. To

interrogate the stability of PdBi2 we cycled the material

from 0.1 V to 1.1 V vs Reversible Hydrogen Electrode

(RHE, henceforth all potentials will be reported vs this

reference electrode) under O2 at 1600 RPM in 0.1 M KOH.

Two redox waves appear between ~0.5 to 0.75 V in the

ORR voltammogram, indicating a background corrosion

process occurs in parallel with ORR (Figure 2a). Upon

repeated potential cycling, the activity continuously

increases, and redox waves decrease, finally reaching

stability after ~500 cycles. The features in the CVs

collected under Ar change significantly before and after

cycling, indicating a change in surface structure of PdBi2

(Figure2b). The loss of intensity of Bi related surface redox

waves and cathodic shift of Pd surface redox waves

indicate that the surface of PdBi2 is enriching with Pd. The

absence of hydrogen underpotential deposition (H-UPD)

from 0 to 0.4 V indicates that the surface does not evolve

to pure Pd, but rather a Pd-rich alloy.

To assess the evolution of this structural change, we

performed HRTEM after cycling (Figure 2c and Figure SI3).

Single-crystalline particles were observed, and the fast

Fourier transform (FFT) pattern matches with the

simulated electron diffraction pattern of Pd3Bi along the

[120] direction (Figure 2d & e). SAED patterns were

acquired on an ensemble of particles, and presented as a

1D line scan (Figure 2f). The SAED line plots of

Figure 2. (a) Linear sweep voltammograms (LSVs) the PdBi2 in O2-saturated 0.1 M KOH before and after electrochemical etching recorded at 1600 rpm and 20 mV s−1 scan rate.

(b) Cyclic voltammograms (CV) of the PdBi2 in Ar-saturated 0.1 M KOH before and after the electrochemical etching. (c) HRTEM image of PdBi2 after conversion and (d) its

corresponding FFT pattern (e) which matches with the simulated Pd3Bi down the [120] direction. (f) Line plot of intensity vs reciprocal lattice spacing derived from the

rotationally-averaged SAED pattern shown in the inset (black) and simulated Pd3Bi (green) ICSD 58839. (g) HAADF-STEM image of PdBi2 after conversion and its corresponding

EDS map of (h) Pd, (i) Bi, and (j) Bi + Pd. All reciprocal distances in parts d & e are in units of nm−1

electrochemically etched-PdBi2 and simulated Pd3Bi,

match well. Within the SAED pattern, both a ring and spot

pattern are observed, indicating that the region probed

consists of both polycrystalline materials and single

crystalline materials. The spot pattern in the inset of

Figure 2f matches well with the simulated electron

diffraction pattern of Pd3Bi [16 3 ]. The atomic ratio of

Pd and Bi after cycling was acquired by XPS and ICP-MS

(Figure SI4 and Table S1), both indicate a Pd:Bi atomic

ratio of ~ 77%:23% which corroborates well with the

composition of the observed crystal structure. HAADF-

STEM

with EDS mapping indicates that the Pd and Bi were mixed

homogenously throughout the particles (Figure 2g-j.)

Taken together, this data indicates that PdBi2

nanoparticles convert to Pd3Bi nanoparticles (henceforth,

denoted as converted-Pd3Bi) by electrochemical etching.

Electrochemical Stability of Phases The removal of Bi

from PdBi2 upon cycling is consistent with the calculated

Bi-Pd Pourbaix diagram (Figure 3a). In 0.1M KOH solution

(pH 13), PdBi2 is predicted to become thermodynamically

unstable above 0.03 V vs RHE. Under equilibrium

conditions, as the voltage increases, Bi will dissolve from

PdBi2 to produce Bi2O3, leaving behind PdBi, then Pd5Bi2,

then Pd3Bi, before finally converting to Pd and Bi2O3 above

0.35 V vs RHE. The experimental observation of converted

Pd3Bi after cycling up to 1.1 V vs RHE indicates that the

Pd3Bi is kinetically stabilized.

The kinetic stabilization may be due to the slow

formation of oxides upon Bi dissolution; for example,

Bi2O3 is known to precipitate slowly from solutions of

Bi3+.26 If the formation of oxides such as Bi2O3 is kinetically

hindered, and conversion to Pd is assumed to proceed via

the increase in concentration of aqueous Bi3+ (Figure 3b)

or BiOH2+ (Figure SI5) ions, Pd3Bi can be expected to

persist up to approximately 0.90 V or 0.67 V respectively

vs RHE. The formation of surface oxides on the

nanoparticles may further contribute to their kinetic

stability. Moreover, the formation of crystalline Pd shells

over Pd3Bi (orthorhombic lattice) may also be kinetically

hindered as the crystal structure and lattice constants vary

considerably from the ground state of elemental Pd (FCC).

Taken together these data indicate that Pd3Bi is kinetically

stable and forms by a combination of Bi corrosion from

PdBi2 and atomic restructuring.

To probe the evolution of Pd3Bi from PdBi2, we collected

high resolution TEM images on a sample which was not

transformed to completion (Figure 4). The image displays

particles with a crystalline core, and an amorphous shell.

HAADF-STEM imaging indicated that the crystalline cores of

the particles were denser than the amorphous shells (Figure

4b-c), and energy dispersive x-ray spectroscopy (EDS) mapping

indicated that the atomic ratio (Pd/Bi) of the crystalline core is

1.08 ± 0.19 and shell is 0.84 ± 0.05 (2 sigma). EDS mapping of

Pd and Bi, and line scans of the elemental composition of the

particles shown in Figure 4c indicates that the shell of the

particles are slightly richer in Bi than the core (Figure 4d-g).

Taken together, this data indicates that the removal of Bi from

the surface of the particle leaves behind a defect rich

amorphous material which allows facile transport of Bi out of

Figure 3. (a) The calculated Pd-Bi Pourbaix diagram. (b) The Pourbaix diagram with oxides removed, so that dissolved Bi only forms aqueous Bi3+. Two-phase regions are marked

with a color and a symbol. The colors indicate the equilibrium Bi-oxide or solvated Bi-containing ion, and the symbol indicates the equilibrium phase that contains Pd. Single

phase regions are shaded white. The dashed red lines indicate the electrochemical stability window for water, with the lower line corresponding to 0 V vs RHE. The experimental

pH, 13, is marked by a dotted grey line.

Figure 4. (a) TEM image of PdBi2 after 200 cycles, the crystalline core is crystalline with

a lattice constant of 0.23 nm, while the shell is amorphous (b) HAADF-STEM images of

PdBi2 after 200 cycles indicating that the crystalline core is denser than the amorphous

shell. The large low density particles in the background are Vulcan carbon. c) HAADF-

STEM imaging of the region in part (d) highlighted in the red box. EDS mapping of (d)

Bi, (e) Pd, and (f) Bi + Pd. (g) EDS line scans of the boxed regions are shown in part (f).

the core, while the core reconstructs to a crystalline

intermetallic phase. This observation is consistent from our

prediction from the Pourbaix diagram that as Bi dissolves from

PdBi2, intermetallic phases which are more Pd rich are left

behind as the particle is finally converted to Pd3Bi.

Mobility of Atoms During Corrosion To investigate the

rate of phase conversion between the ordered

intermetallic phases, we used density functional theory27

to calculate the minimum diffusion activation energy of Bi

atoms in PdBi2, assuming a vacancy-mediated diffusion

mechanism. Minimum energy paths and saddle point

energies were determined with the climbing image

nudged elastic band method.28-29 To identify possible

paths, we have taken into account all possible jumps

between an atom and a vacant neighbouring site. The

lowest barrier for Bi diffusion in PdBi2 is 0.44 eV,

corresponding to a two-dimensional migration pathway

(Fig. S6). The mean square displacement of a particle

undergoing a two-dimensional random walk via a

vacancy-mediated mechanism is

⟨∆r2 ⟩=4 xv Γ a2t , where xv is the vacancy

concentration, Γ is the hopping frequency from a

filled to a vacant site, a is the hopping length, and t is

the time. We estimate the vacancy concentration using

xv≈ exp (−E f /k BT ) , where Ef is the vacancy

formation energy. We find that diffusion of Bi through

PdBi2 is likely facilitated by high vacancy concentrations,

as the calculated vacancy formation energy for Bi along

the diffusion path is only 0.17 eV relative to bulk Bi. The

vacancy formation

energy will therefore be even lower relative to states in

which the chemical potential of Bi is lower than that of

bulk Bi, such as those that are predicted to be stable at

higher potentials (Figure 3). This low calculated vacancy

formation energy is consistent with the low melting point

of PdBi2. Transition state theory30 gives us

Γ ≈ν¿exp (−Ea/k BT ) , where ν¿ is generally

in the range of phonon frequencies and Ea is the

activation energy for a single hop. Using the reported

vibrational frequency of around 1012 Hz for Bi in β-PdBi2 31,

we estimate that at room temperature the root mean

square displacement of Bi in PdBi2 reaches a typical

particle width in well less than 1 second. This result

indicates that Bi diffusion in PdBi2 is facile, supporting the

potential for fast removal of Bi from PdBi2 under

experimental conditions.

ORR Activity and Stability of Converted-Pd3Bi. The ORR

activity of the converted-Pd3Bi was compared with

commercial Pt/C (TKK, 40% Pt on high surface area

carbon) and Pd/C (Premetek,

40% Pd on Vulcan XC-72) catalysts (Figure 5a and SI-7).

The onset potential (the potential in which the current

exceeds the background by 10%) of ORR on both Pt/C was

1.0 V and Pd/C was 1.05 V, which agreed with literature

reports32; while the converted-Pd3Bi was more negative

(~0.97 V). At potentials more positive than the onset

potential the Pd and Pt sites are covered with a surface

oxide, rendering them catalytically inactive. At more

negative potentials (starting at ~0.95 V), we observed a

rapid increase in current of the converted-Pd3Bi, while the

current for Pd/C and Pt/C increased gradually. The half-

wave potential of converted-Pd3Bi was more positive by

60 mV and 32 mV relative to Pt/C and Pd/C, respectively,

indicating that it may possess superior catalytic activity

relative to the commercial catalysts.

The Koutecky-Levich (K-L) plot was constructed by

measuring the steady-state current density at a constant

voltage at various rotation rates to calculate the kinetic

current density (Figure SI 8-9). The specific activity (SA)

and mass activity (MA) were measured by normalizing the

kinetic current density to the ECSA (Figure SI-10) or to the

total mass of precious metal, respectively. The SA of

converted-Pd3Bi was 2.3 ±0.19 mA/ cmPd2

at

0.9V vs RHE, which is 19x higher than Pt/C (0.12

±0.03 mA/ cmPt2

) and 4.4x higher than Pd/C

(0.52 ±0.07 mA/ cmPd2

) (Figure 5b). The MA of

converted-Pd3Bi was 1.2±0.08 A mgPd−1

, which

is 3.4x higher than Pd ( 0.36±0.04 A mgPd−1¿

and 11x higher than Pt ( 0.11±0.02AmgP t−1¿

(Figure 5b).6 The SA and MA of Pd3Bi is among the highest

reported for alkaline ORR (Table S2). The high MA is

remarkable given the relatively modest specific surface

area of Pd3Bi (Figure 5C), decreases in particle size could

permit even higher mass activities.

We evaluated the stability of our catalysts by

performing accelerated durability testing by cycling the

samples from 0.6 to 1.0 V at a sweep rate of 100 mV/s.

Converted-Pd3Bi retained 86% and 69% of the initial

specific and mass activity, respectively, after 10,000 cycles

at 0.9V, while the specific surface area decreased by 20%

(Figure 5b-c and SI11). In comparison, Pd/C retains 73%

and 71% of the initial specific activity and mass activity,

respectively, and Pt/C retains 63% and 54% of the initial

specific activity and mass activity, respectively (Figure 5b).

The specific surface area of Pt/C and Pd/C decreased by

14% and 3% respectively. This indicates that the decreases

in catalytic activity of the commercial samples is

predominately caused by removal of highly active sites by

surface reconstructions and/or Ostwald ripening (Figure

SI12) .33 The decrease of activity for Pd3Bi was coincident

with the emergence of Ostwald ripening, as indicated by a

Figure 5. (a) LSVs of Pt/C, Pd/C, and converted-Pd3Bi in O2-saturated 0.1 M KOH. (b) Mass activity and specific activity at 0.9 V before and after stability test calculated from

the Kouteky Levich equation. Negative values indicate the current is cathodic. (c) Electrochemically active surface area (ECSA) and mass loading before and after stability test.

ECSA was measured by CO stripping, and the mass loading was measured by inductively coupled plasma mass spectrometry (ICPMS). (d) STEM-HAADF image of a Pd 3Bi

catalyst particle after 10,000 cycles of ADTs. Inserts are FFT of the image and simulated single-crystal diffraction pattern down the [100] direction of Pd 3Bi, respectively. All

voltammetry was collected on a 5mm diameter rotating disk electrode (RDE) at1600 rpm, and a voltage sweep rate of 20 mV s−1. The mass loading of all samples onto the 5

mm glassy carbon disk was between ~13 to 15μg Pd or Pt cm−2geo

decrease in the specific surface area, and the emergence

of a small amount of metallic Pd. The metallic Pd probably

formed via redeposition of dissolved Pd ion onto the

electrode, as evident from the appearance of H-UPD in

the CVs collected under Ar after the ADTs (Figures SI13

and 5c, and Table S2). This was further supported by ICP-

MS and XPS (Table S3) which showed the composition of

the electrode was ~78% Pd and ~22% Bi after 10K cycles,

indicating 1% higher Pd content relative to the freshly

converted samples. HAADF-STEM images of the sample

after ADT test confirmed that the particles retained

intermetallic crystallinity (Figures 5d and Figure SI14).

Taken together, the data indicates that that the corrosion

of Pd3Bi to Pd was kinetically slow, hence it was able to

sustain remarkable performance after 10,000 cycles, even

outperforming freshly prepared Pt/C and Pd/C.

Fuel crossover and poisoning of the cathode, where

ORR takes place, drastically decreases the activity and

durability of commercial direct methanol fuel cells.34-35

This is largely due to the high catalytic activity of Pt and

Pd for alcohol oxidation;36 when fuel crosses over from

the anode to the cathode, the cathode experiences a

mixed potential from both oxidation and reduction

reactions taking place simultaneously, eroding the current

density and voltage of the device. To examine the

tolerance of converted-Pd3Bi towards fuel crossover, we

measured the ORR activity in the presence of methanol.

The ORR performance of Pt/C, Pd/C, and converted Pd3Bi

in O2-saturated 0.5 M CH3OH + 0.1 M KOH were measured

(Figure 6). Pd/C and Pt/C show significant increase in the

overpotential for ORR in the presence of methanol due to

the competitive methanol oxidation over ORR. The anodic

peak at 1.02 V for Pt/C and the 0.85 V peak for Pd/C are

from the oxidation of methanol, suggesting that ORR is

largely suppressed by the competing reaction. However,

only a small 10 mV decrease of the E1/2 for ORR on

converted Pd3Bi is observed. This suggests that the

converted-Pd3Bi has superior methanol tolerance and is a

promising cathode catalyst for direct methanol fuel cells.37

Conclusions

In summary, we have demonstrated kinetically-controlled

corrosion as an effective approach to convert a base-

metal rich ordered intermetallic nanoparticles (PdBi2) to

catalytically active ordered intermetallic nanoparticles

(Pd3Bi). We have employed the converted Pd3Bi

nanoparticles as electrodes for ORR with high methanol

tolerance, the resulting material exhibited superior mass

activity relative to commercial Pd/C and Pt/C. We found

that a low melting point ordered intermetallic

nanoparticle can convert to a noble metal rich ordered

intermetallic with high catalytic activity due to facile

formation of vacancies and high mobility of interdiffusion.

This strategy opens a new synthetic method for accessing

high-performance heterogeneous catalysts in ambient

conditions.

Experimental Methods

Materials. Palladium (II) acetylacetonate (Pd(acac)2, 99%, Strem Chemicals), Bismuth(III) acetate (Bi(oac)3, 99.99%, Alfa Aesar), and oleylamine (OLAM, technical, 70%, Aldrich), 40% Pd on Vulcan XC-72 (Premetek Co), O2 gas (UHP grade, Airgas), 40% Pt on High Surface Area Carbon (TEC10E40E, Tanaka Kikinzoku Kogyo TKK), were used as received without purification. All syntheses were carried out under Ar (g) using standard Schlenk techniques, and work-up proce-dures were performed in air. Electrolyte solutions were prepared with deionized water (resistance: 18.2 MΩ) and potassium hydroxide (KOH, Semiconductor Grade 99.99%, Alfa Aesar).

Synthesis of PdBi2 nanoparticles (NPs). In a typical synthesis, 61 mg of Pd(acac)2 (0.2 mmol), 154 mg of Bi(oac)3 (0.4 mmol), and 10 mL of OLAM were first added into a 20-mL scintillation vial with a magnetic stir bar. After stirring for 20 mins, the vial was sonicated for another 10 mins to ensure all the solid precursors were dissolved and the mixture formed a pale-yellow solution. Another 5 mL of OLAM were added into a 100-mL 3-neck round-bottom flask with a condenser, thermometer adapter, thermometer, and rubber septum. Magnetic stirring was started and the solution was degassed under vacuum at 120 °C for ~5-10 minutes. The flask was backfilled with Ar and heated to ~300 °C at a rate of 10 °C/min. Once the temperature reached 300 °C, the 10 mL of precursor solution was slowly injected into the flask using a syringe pump at a rate of 0.3 mL/min. The solution in the flask turned from colorless to light brown and finally formed a dark brown colloidal mixture when the injection was

Figure 6. Linear sweep voltammograms of converted-Pd3Bi, Pd/C, and Pt/C in

O2-saturated 0.1 M KOH + 0.5 M CH3OH. Rotation rate: 1600 rpm. Sweep rate: 20

mV s−1. The mass loading of all samples onto the 5 mm glassy carbon disk was

between ~13 to 15μg Pd or Pt cm−2geo

completed. The solution was then cooled rapidly by removing the flask from the heating mantle. The nanoparticles were precipitated by adding 10 mL of ethanol and then centrifuged at 8000 rpm for 5 min. The precipitate was washed three timesusing a 1:1 ethanol/toluene mixture (with centrifugation in between washes) and could then be suspended in ethanol for further characterization and electrode preparation.

Physicochemical characterization. The ethanol suspension of the nanoparticles was dropcasted onto a zero-background Si wafer substrate and examined with a Philips X’Pert Pro Powder X-Ray diffractometer (XRD) with CuK radiation (K1, =1.5406 Å and K2, =1.5444 Å). Transmission electron microscopy (TEM) was performed on a FEI F200C Talos or FEI Tecnai TF30operation at 200 kV. The materials on the glassy carbon disk after electrochemical test was dissolved by concentrated nitric acid and then diluted with water. The metal concentration in the nitric acid solution is analyzed by the PerkinElmer NexION 300D with ICP.

Oxygen Reduction Measurements. Electrochemical measurements were carried out with a potentiostat/galvanostat (Metrohm Autolab Potentiostat or Nuvant EzStat Pro) and a three-electrode Teflon cell with 0.1 M KOH aqueous solution as the electrolyte. A Hg/HgO electrode in 1 M NaOH was used as the reference electrode. A graphite rod was used as the counter electrode. The Hg/HgO reference electrode was periodically checked with a Reversible Hydrogen Electrode (RHE) to verify the stability of the reference potential. All data measured with IR compensationon the potentiostat. ORR was performed by sweeping the voltage from -0.9 V to 0.2 V vs. Hg/HgO at a sweep rate of 20 mv/s, or by constant potential chronoamperometery within the same potential range. All measurements were performed on a 5 mm diameter rotating disk electrode at 1600 rpm. All ORR measurments were IR-corrected, the IR drop was measured by the positive feedback or current interrupt method.

CO Stripping Measurements. We utilized the CO stripping method to determine the electrochemically active surface area (ECSA) of Pt or Pd atoms on the surface of the electrode. The electrodes are first held at 0.15V vs RHE for 20minin 0.1M KOH saturated with CO to form a surface adsorbed CO monolayer, followed by removal of CO (aq) by saturating the electrolyte with Ar for 15 min. Finally, a the voltage was swept from 0.15V to 1.15 V vs RHE to oxidize the adsorbed CO layer, the surface area of the sample can be calculated by integrating the oxidation peak assuming one CO adsorbed per Pd or Pt atom on the surface. The ratio of charge transferred for COoxidation and metal surface area is estimated as 484 μC/cm2 for Pt and 420 μC/cm2 for Pd and Pd3Bi. We expect that Bi sites will not adsorb CO,

therefore CO stripping will not account for catalytically inactive Bi sites on the surface.

Conversion of PdBi2 to Pd3Bi. The electrode was prepared by directly dropcasting 4 L of the ethanol suspension of the as-prepared nanoparticle onto the glassy carbon disk. After it dried out, the electrode was then immersed in 0.1 M acetic acid solution for 20 mins followed by DI water wash, acetone wash, and ethanol wash. Finally, the electrode was placed in a UV-ozone chamber and treated with UV and ozone at room temperature for 20 mins. A graphite rod and Hg/HgO electrode were used as the counter and reference electrodes, respectively. The working electrode was then conditioned by cycling the potential from -0.9 V to 0.2 V vs. Hg/HgO reference at a scan rate of 0.1 V s-1 and at a rotation rate of 1600 rpm for ~500 cycles.

Preparation of Pt/C and Pd/C electrodes. To prepare Pd/C materials, 30 mg of 40% Pd on Vulcan XC 72(Premetek Co) was dispersed into a solution consisting of a mixture of 8μl Nafion Ionomer, 1ml DI water and 1ml IPA. The resulting solution was sonicated for 20 min. 1μl of the mixture was drop casted onto the 5 mm diameter disk. The Pt/C sample is prepared with the same procedure except with Pt/C (TKK).

Sample preparation for ICP-MS analysis. After the electrochemical measurement, the glassy carbon electrode was taken out of the electrolyte and rinsed with 18.2 MΩ water. After dried in air, the electrode was placed in a plastic tube and 1 mL concentrated TraceMetal grade nitric acid was added into the tube. The tube was sonicated for a few minutes to fully dissolve the catalyst on the surface. Additional 9 mL of 18.2 MΩ water was added to the tube and the solution is ready for testing.

Computational Methods. The Pourbaixdiagrams were generated using PyMatGen withdata provided by the Materials Project1. To bestview the possible formation of solid-state phasesthe concentrations of aqueous species in thePourbaix diagrams were set to the upper limit thatcould be achieved if all PdBi2 were dissolved. As itis now believed that the aqueous speciespreviously identified as BiO+ is in fact Bi6O4(OH)4

6+,the labels on the generated Pourbaix diagramhave been changed accordingly.2

XPS Analysis

Binding energies have been corrected by shiftingthe peak of C 1s to 284.8 eV. Palladium peakpositions have been assigned according to thevalues in Handbook of X-ray PhotoelectronSpectroscopy3 and reported data in NISTdatabase whereas for Bi to that of Casella andContursi.4 The splitting of peaks usually occursasymmetrically for metals in their lowest oxidation

state and mostly symmetrically for higheroxidation states and the symmetry of peaks iscontrolled by the peak area. The FWHM isassumed to be in the range of (0.9-1.1 eV) formetals and in the range of (1.7-2.0) for oxides. Toproperly fit the spectra, the separation betweenthe peaks is set to 5.2 and 5.3 eV for Pd and Bi,respectively. By utilizing relative sensitivity factors(R.S.F), the peak areas of the Pd and Bi arecorrected, and their ratio is calculated to yieldsurface atomic ratio as

surface atomic ratio=Region peak areaPd /R .S . FPd

Region peak areaBi /R .S . FBi

ASSOCIATED CONTENT

Full experimental details of particle dealloying,nanoparticle film preparation, andcharacterization of the materials used in thisarticle.

AUTHOR INFORMATION

Corresponding Author

* e-mail: shoji@jhu.edu

Author Contributions‡These authors contributed equally.

NotesThe authors declare no competing financial interest.

ACKNOWLEDGMENT A.S.H. acknowledges financial support fromthe National Science Foundation under awardno. CHE-1764310 and Johns HopkinsUniversity start-up funding. T.M. acknowledgesfinancial support from the National ScienceFoundation under award no. CHE-1437396. Wegratefully acknowledge Prof. Mingwei Chen(JHU), and Prof. Jonah Erlebacher (JHU) forhelpful discussions. .

REFERENCES

1. Coleman, E. J.; Chowdhury, M. H.; Co, A. C., Insights Into the Oxygen Reduction ReactionActivity of Pt/C and PtCu/C Catalysts. ACS Catal. 2015, 5 (2), 1245-1253.2. Mukerjee, S.; Srinivasan, S.; Soriaga, M. P.; McBreen, J., Role of Structural and ElectronicProperties of Pt and Pt Alloys on Electrocatalysis of Oxygen Reduction: An In Situ XANES and EXAFSInvestigation. J. Electrochem. Soc. 1995, 142 (5), 1409-1422.3. Casado-Rivera, E.; Volpe, D. J.; Alden, L.; Lind, C.; Downie, C.; Vázquez-Alvarez, T.; Angelo,A. C. D.; DiSalvo, F. J.; Abruña, H. D., Electrocatalytic Activity of Ordered Intermetallic Phases forFuel Cell Applications. J. Am. Chem. Soc. 2004, 126 (12), 4043-4049.4. Wang, D.; Xin, H. L.; Hovden, R.; Wang, H.; Yu, Y.; Muller, D. A.; DiSalvo, F. J.; Abruña, H.D., Structurally ordered intermetallic platinum–cobalt core–shell nanoparticles with enhanced activityand stability as oxygen reduction electrocatalysts. Nat. Mater. 2013, 12 (1), 81-87.5. Wencka, M.; Hahne, M.; Kocjan, A.; Vrtnik, S.; Koželj, P.; Korže, D.; Jagličić, Z.; Sorić, M.;Popčević, P.; Ivkov, J.; Smontara, A.; Gille, P.; Jurga, S.; Tomeš, P.; Paschen, S.; Ormeci, A.;Armbrüster, M.; Grin, Y.; Dolinšek, J., Physical properties of the InPd intermetallic catalyst.Intermetallics 2014, 55 (Supplement C), 56-65.6. Cui, Z.; Chen, H.; Zhao, M.; DiSalvo, F. J., High-Performance Pd3Pb Intermetallic Catalyst forElectrochemical Oxygen Reduction. Nano Letters 2016, 16 (4), 2560-2566.7. Yan, Y.; Du, J. S.; Gilroy, K. D.; Yang, D.; Xia, Y.; Zhang, H., Intermetallic Nanocrystals:Syntheses and Catalytic Applications. Advanced Materials 2017, 29 (14), 1605997.8. Johannes, T.; Martin, H.; Jen-Hui, C.; Michael, R., Refinement of the Microwave-Assisted PolyolProcess for the Low-Temperature Synthesis of Intermetallic Nanoparticles. European Journal ofInorganic Chemistry 2017, 2017 (42), 4930-4938.9. Cable, R. E.; Schaak, R. E., Low-Temperature Solution Synthesis of Nanocrystalline BinaryIntermetallic Compounds Using the Polyol Process. Chemistry of Materials 2005, 17 (26), 6835-6841.10. Wang, Y.; Sun, D.; Chowdhury, T.; Wagner, J. S.; Kempa, T. J.; Hall, A. S., Rapid Room-Temperature Synthesis of a Metastable Ordered Intermetallic Electrocatalyst. Journal of the AmericanChemical Society 2019, 141 (6), 2342-2347.11. Ji, X.; Lee, K. T.; Holden, R.; Zhang, L.; Zhang, J.; Botton, G. A.; Couillard, M.; Nazar, L. F.,Nanocrystalline intermetallics on mesoporous carbon for direct formic acid fuel cell anodes. Nat. Chem.2010, 2 (4), 286-293.12. Chen, H.; Wang, D.; Yu, Y.; Newton, K. A.; Muller, D. A.; Abruña, H.; DiSalvo, F. J., ASurfactant-Free Strategy for Synthesizing and Processing Intermetallic Platinum-Based NanoparticleCatalysts. J. Am. Chem. Soc. 2012, 134 (44), 18453-18459.13. Qi, Z.; Xiao, C.; Liu, C.; Goh, T. W.; Zhou, L.; Maligal-Ganesh, R.; Pei, Y.; Li, X.; Curtiss, L.A.; Huang, W., Sub-4 nm PtZn Intermetallic Nanoparticles for Enhanced Mass and Specific Activities inCatalytic Electrooxidation Reaction. J. Am. Chem. Soc. 2017, 139 (13), 4762-4768.14. Chi, M.; Wang, C.; Lei, Y.; Wang, G.; Li, D.; More, K. L.; Lupini, A.; Allard, L. F.; Markovic,N. M.; Stamenkovic, V. R., Surface faceting and elemental diffusion behaviour at atomic scale for alloynanoparticles during in situ annealing. Nature Comm. 2015, 6, 8925.15. Kaiser, H., Selective dissolution of high and low diffusivity alloys—A comparison of kineticaland micromorphological aspects. Corrosion Science 1993, 34 (4), 683-699.

12

16. Marshakov, I. K., Anodic Dissolution and Selective Corrosion of Alloys. Protection of Metals2002, 38 (2), 118-123.17. Wang, D.; Yu, Y.; Zhu, J.; Liu, S.; Muller, D. A.; Abruña, H. D., Morphology and ActivityTuning of Cu3Pt/C Ordered Intermetallic Nanoparticles by Selective Electrochemical Dealloying. NanoLetters 2015, 15 (2), 1343-1348.18. Oezaslan, M.; Hasché, F.; Strasser, P., Pt-Based Core–Shell Catalyst Architectures for OxygenFuel Cell Electrodes. The Journal of Physical Chemistry Letters 2013, 4 (19), 3273-3291.19. Sun, K.; Cheng, T.; Wu, L.; Hu, Y.; Zhou, J.; Maclennan, A.; Jiang, Z.; Gao, Y.; Goddard, W. A.;Wang, Z., Ultrahigh Mass Activity for Carbon Dioxide Reduction Enabled by Gold–Iron Core–ShellNanoparticles. Journal of the American Chemical Society 2017, 139 (44), 15608-15611.20. Medasani, B.; Haranczyk, M.; Canning, A.; Asta, M., Vacancy formation energies in metals: Acomparison of MetaGGA with LDA and GGA exchange–correlation functionals. ComputationalMaterials Science 2015, 101, 96-107.21. Otake, S.; Ishii, Y.; Matsuno, N., Migration Energy of Vacancies in Bismuth. Japanese Journalof Applied Physics 1981, 20 (6), 1037-1040.22. Shao, M. H.; Huang, T.; Liu, P.; Zhang, J.; Sasaki, K.; Vukmirovic, M. B.; Adzic, R. R.,Palladium Monolayer and Palladium Alloy Electrocatalysts for Oxygen Reduction. Langmuir 2006, 22(25), 10409-10415.23. Jiang, J.; Gao, H.; Lu, S.; Zhang, X.; Wang, C.-Y.; Wang, W.-K.; Yu, H.-Q., Ni–Pd core–shellnanoparticles with Pt-like oxygen reduction electrocatalytic performance in both acidic and alkalineelectrolytes. Journal of Materials Chemistry A 2017, 5 (19), 9233-9240.24. Sun, D.; Feng, S.; Terrones, M.; Schaak, R. E., Formation and Interlayer Decoupling of ColloidalMoSe2 Nanoflowers. Chem. Mater. 2015, 27 (8), 3167-3175.25. Sun, Y.; Wang, Y.; Sun, D.; Carvalho, B. R.; Read, C. G.; Lee, C.-h.; Lin, Z.; Fujisawa, K.;Robinson, J. A.; Crespi, V. H.; Terrones, M.; Schaak, R. E., Low-Temperature Solution Synthesis ofFew-Layer 1T -MoTe2 Nanostructures Exhibiting Lattice Compression. ′ Angew. Chem. Int. Ed. 2016, 55(8), 2830-2834.26. Wiberg, E.; Wiberg, N.; Holleman, A. F., Inorganic chemistry. Academic Press ; De Gruyter:San Diego; Berlin; New York, 2001.27. Kohn, W.; Sham, L. J., Self-consistent equations including exchange and correlation effects.Phys. Rev. 1965, 140, A1133-A1138.28. Henkelman, G.; Jonsson, H., Improved tangent estimate in the nudged elastic band method forfinding minimum energy paths and saddle points. Journal of Chemical Physics 2000, 113 (22), 9978-9985.29. Henkelman, G.; Uberuaga, B. P.; Jonsson, H., A climbing image nudged elastic band method forfinding saddle points and minimum energy paths. Journal of Chemical Physics 2000, 113 (22), 9901-9904.30. Vineyard, G. H., Frequency Factors and Isotope Effects in Solid State Rate Processes. J PhysChem Solids 1957, 3 (1-2), 121-127.31. Zheng, J.-J.; Margine, E. R., Electron-phonon coupling and pairing mechanism in

β−Bi2Pd

13

centrosymmetric superconductor. Physical Review B 2017, 95 (1).32. Jiang, L.; Hsu, A.; Chu, D.; Chen, R., Oxygen Reduction Reaction on Carbon Supported Pt andPd in Alkaline Solutions. J. Electrochem. Soc. 2009, 156 (3), B370-B376.33. Wang, H.; Luo, W.; Zhu, L.; Zhao, Z.; E, B.; Tu, W.; Ke, X.; Sui, M.; Chen, C.; Chen, Q.; Li, Y.;Huang, Y., Synergistically Enhanced Oxygen Reduction Electrocatalysis by Subsurface Atoms inTernary PdCuNi Alloy Catalysts. Advanced Functional Materials 2018, 28 (15), 1707219.34. Sekol, R. C.; Li, X.; Cohen, P.; Doubek, G.; Carmo, M.; Taylor, A. D., Silver palladium core–shell electrocatalyst supported on MWNTs for ORR in alkaline media. Appl. Catal. B 2013, 138(Supplement C), 285-293.35. Heinzel, A.; Barragán, V. M., A review of the state-of-the-art of the methanol crossover in directmethanol fuel cells. J. Power Sources 1999, 84 (1), 70-74.36. Ataee-Esfahani, H.; Wang, L.; Nemoto, Y.; Yamauchi, Y., Synthesis of Bimetallic Au@PtNanoparticles with Au Core and Nanostructured Pt Shell toward Highly Active Electrocatalysts. Chem.Mater. 2010, 22 (23), 6310-6318.37. Gurau, B.; Smotkin, E. S., Methanol crossover in direct methanol fuel cells: a link betweenpower and energy density. J. Power Sources 2002, 112 (2), 339-352.

14

download fileview on ChemRxivSun-phase-convert-VF.docx (7.51 MiB)

Ordered Intermetallic Nanoparticles with High Catalytic

Activity Prepared by an Electrochemically Induced Phase

Transformation

Du Sun,a‡ Yunfei Wang,a‡ Ken Livia, Chuhong Wang,a Ruichun Luo,b Zhuoqun Zhang,a Hamdan Al-

ghamdi,c Chenyang Li,a Fufei An, a,b Bernard Gaskey,a Tim Mueller,a Anthony Shoji Halla*

a Department of Materials Science and Engineering, Johns Hopkins University, Baltimore, Maryland 21218

b School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai 200030, P. R. China

c Department of Chemical and Biomolecular Engineering, Johns Hopkins University, Baltimore, Maryland 21218.

‡Denotes co-first authors

ORCID ID: A.S. Hall - 0000-0003-4134-4160

KEYWORDS Oxygen reduction reaction • Palladium • Ordered intermetallic • Bismuth • Phase conversion • Electrocataly-

sis• Alloy • dealloying

ABSTRACT: The synthesis of alloys with long range atomic scale ordering (ordered intermetallics) is an emerging field of nano-

chemistry. Ordered intermetallic nanoparticles are useful for a wide variety of applications such as catalysis, superconductors, and

magnetic devices. However, the preparation of nanostructured ordered intermetallics is challenging in comparison to disordered al-loys, hindering progress in materials development. We report a process for converting colloidally synthesized ordered intermetallic

PdBi2 to ordered intermetallic Pd3Bi nanoparticles under ambient conditions by an electrochemically induced phase transition. The low melting point of PdBi2 corresponds to low vacancy formation energies which enables the facile removal of the Bi from the

surface, while simultaneously enabling interdiffusion of the constituent atoms via a vacancy diffusion mechanism under ambient conditions. The resulting phase-converted ordered intermetallic Pd3Bi exhibits 11x and 3.5x higher mass activty and high methanol

tolerance for the oxygen reduction reaction compared to Pt/C and Pd/C, respectively,which is the highest reported for a Pd-based catalyst, to the best of our knowledge. These results establish a key development in the synthesis of noble metal rich ordered inter-

metallic phases with high catalytic activity, and sets forth guidelines for the design of ordered intermetallic compounds under ambient

conditions.

Introduction

Noble metal rich phases of alloys have recently emerged as high performance catalysts for a variety of industrially relevant catalytic applications. The mixing of two different metals can re-sult in a solid solution that exhibits short-range ordering of atoms (disordered alloy) or an ordered intermetallic that exhibits both short-range and long-range ordering of atoms. Ordered interme-tallic compounds (OICs) can have superior catalytic performance and stability compared to disordered alloys because of the atomi-cally uniform active sites provided by long-range structural perio-dicity and stronger interatomic bonding from a more negative en-thalpy of mixing.1-7

OICs are difficult to prepare, often requiring annealing to temperatures of ~700 oC to facilitate inter-diffusion and equilibra-tion of atoms.7 Direct synthesis of some ordered intermetallic nanostructures has been achieved at a relatively mild temperature by using either harsh reducing agents, special reaction parameters such as microwave or solvothermal conditions, or electrochemical deposition.7-10 Moreover, noble-metal rich phases (>50% noble metal) are particularly challenging to synthesize in mild condi-tions, even if the ordered phases are thermodynamically favored, given the high melting points of the noble metals and low diffusion rates in crystal lattices which inhibit equilibration.6

The preparation of highly dispersed OIC catalysts by anneal-ing at high temperatures is challenging because sintering occurs during annealing, reducing the surface area to volume ratio. The

most common method for reducing sintering is to physically iso-late the particles by encapsulation with a template before anneal-ing. After crystallization, the templates can be chemically etched away.11-13 The ability to control the shape of the particle is lost since the most thermodynamically stable facet is typically ex-pressed from surface reconstructions which occur during the phase transformation.14 Additionally, contamination can occur at elevated temperatures since this can allow atoms from the sup-port or template to diffuse into the alloy. The preparation of cata-lytically active OIC nanostructures under mild conditions remains a great challenge in catalysis science.

Studies focusing on methods for preparing of OIC nanoparti-cles are limited in comparison to disordered alloys. Recently, we reported that ordered intermetallic Pd31Bi12 can be prepared di-rectly by electrodeposition under ambient conditions; the result-ing materials exhibit the highest specific activity for the oxygen re-duction reaction measured on a Pd-based material in alkaline elec-trolytes.10 However, the mass activity (an important technological metric for fuel cell devices) of electrochemically deposited Pd31Bi12

is low because this method affords a thin-film morphology, which

has as a low surface area to volume ratio. Moreover, controlling the composition of a noble/non-noble metal alloys by electro-chemical deposition is challenging because of large differences in reduction potential between the two constituent elements. If OICs are to be practically utilized for devices, a method which allows the preparation of OIC nanoparticles under ambient conditions is needed.

It is known in the corrosion literature that the atoms of alloys with low melting points exhibit high diffusivity during de-alloying at room temperature.15-16 In contrast, materials with high melting points (m.p. > 1000 oC) have low diffusivity; during de-alloying a thin noble metal film will form over the surface, protecting the surface against further de-alloying across a wide range of voltages. This has been observed in bulk alloys such as Cu-Pd, Cu-Au, etc; and on nanoparticle alloys in which a noble metal shells forms over the alloy cores in systems such as Pt3Co, Cu3Pt, Pt3Ni, FeAu, etc.4,

15-19 From these design principles, it can be inferred that the con-trolled corrosion of a non-noble element from a low melting point alloy nanoparticle could permit conversion to a noble metal rich OIC.

Here, we validate this concept by illustrating the kinetically controlled conversion of ordered intermetallic PdBi2 (m.p. ~480 oC) nanoparticles to ordered intermetallic Pd3Bi nanoparticles via electrochemical de-alloying at room temperature. The low melt-ing point of PdBi2 indicates that it has low vacancy formation en-ergies, which enables the facile removal of the Bi from the surface while simultaneously enabling interdiffusion of the constituent at-oms via a vacancy diffusion mechanism at room temperature.20-21 This method allows for the entire particle to reconstruct to a noble metal rich OIC phase which has never been prepared under ambi-ent conditions. The phase converted-Pd3Bi exhibits record break-ing mass activity for the oxygen reduction reaction (ORR), reaching activities as high as 1.2 A mgPd

−1 at 0.9 V vs the Reversible Hydrogen

Electrode (RHE), which is ~3.4x than Pd/C, and 11x higher than Pt/C, respectively.

Results and Discussion

Synthesis and characterization of PdBi2. The PdBi2

nanostructures were synthesized by reduction of Pd(II) acet-

ylacetonate and Bi(III) acetate in oleylamine at 300 0C. PdBi2

has two known layered phases denoted as and −PdBi2 and

−PdBi2.22-23 The peaks in the experimental X-ray diffraction

(XRD) pattern exhibit broadening (Figure 1a), suggesting the

crystallite size is in the nanoscale. The pattern can be indexed

to −PdBi2 with a small amount of −PdBi2. The (011) deflec-

tion is more intense in synthesized PdBi2 than the simulated

pattern, indicating preferential orientation along this direc-

tion. Note that there is a shift of some of the peaks, which

may be caused by surface tension; this is commonly seen in

nanomaterials with small crystallites.24-25

The morphology and crystalline phases of PdBi2 were

probed by high-resolution transmission electron microscopy

(HRTEM). The sizes of the PdBi2 particles ranged from ~10 to

50 nm in diameter (Figure 1b and Figure SI1). The particles

have a core-shell structure where the core is single crystalline

and the shell is amorphous (Figure 1c). The atomic arrange-

ment of the core matches the (100) plane of −PdBi2, the

amorphous shell is Bi (Figure SI2). The selected area electron

diffraction (SAED) was presented as a 1D line plot (Figure 1d)

and compared to the 1D diffraction pattern of −PdBi2, both

patterns match well, indicating that the crystalline phases

present in the material are −PdBi2.

Electrochemical Phase Conversion of PdBi2. To interrogate

the stability of PdBi2 we cycled the material from 0.1 V to 1.1

V vs Reversible Hydrogen Electrode (RHE, henceforth all po-

tentials will be reported vs this reference electrode) under O2

at 1600 RPM in 0.1 M KOH. Two redox waves appear between

~0.5 to 0.75 V in the ORR voltammogram, indicating a back-

ground corrosion process occurs in parallel with ORR (Figure

2a). Upon repeated potential cycling, the activity continu-

ously increases, and redox waves decrease, finally reaching

stability after ~500 cycles. The features in the CVs collected

Figure 1. (a) Powder X-ray diffraction pattern of the as-prepared sample. The crystal planes that are corresponding to alpha-PdBi2 are marked in red, while the ones to beta-PdBi2

are in green. (b) TEM shows a representative image of the as-prepared nanoparticles. (c) High Angle Annular Dark Field-Scanning Transmission Electron Microscopy (STEM-

HAADF) image reveals the atomic arrangement of the nanoparticles, the atoms match with (100) plane of beta-PdBi2. (d) Line plot of intensity vs reciprocal lattice spacing derived

from the rotationally-averaged selected area electron diffraction (SAED) pattern shown in the inset (black) and simulated beta-PdBi2 (green) ICSD 56280

under Ar change significantly before and after cycling, indi-

cating a change in surface structure of PdBi2 (Figure2b). The

loss of intensity of Bi related surface redox waves and ca-

thodic shift of Pd surface redox waves indicate that the sur-

face of PdBi2 is enriching with Pd. The absence of hydrogen

underpotential deposition (H-UPD) from 0 to 0.4 V indicates

that the surface does not evolve to pure Pd, but rather a Pd-

rich alloy.

To assess the evolution of this structural change, we per-

formed HRTEM after cycling (Figure 2c and Figure SI3). Single-

crystalline particles were observed, and the fast Fourier

transform (FFT) pattern matches with the simulated electron

diffraction pattern of Pd3Bi along the [120] direction (Figure

2d & e). SAED patterns were acquired on an ensemble of par-

ticles, and presented as a 1D line scan (Figure 2f). The SAED

line plots of electrochemically etched-PdBi2 and simulated

Pd3Bi, match well. Within the SAED pattern, both a ring and

spot pattern are observed, indicating that the region probed

consists of both polycrystalline materials and single crystal-

line materials. The spot pattern in the inset of Figure 2f

matches well with the simulated electron diffraction pattern

of Pd3Bi [163]. The atomic ratio of Pd and Bi after cycling was

acquired by XPS and ICP-MS (Figure SI4 and Table S1), both

indicate a Pd:Bi atomic ratio of ~ 77%:23% which corrobo-

rates well with the composition of the observed crystal struc-

ture. HAADF-STEM

with EDS mapping indicates that the Pd and Bi were mixed

homogenously throughout the particles (Figure 2g-j.) Taken

together, this data indicates that PdBi2 nanoparticles convert

to Pd3Bi nanoparticles (henceforth, denoted as converted-

Pd3Bi) by electrochemical etching.

Electrochemical Stability of Phases The removal of Bi from

PdBi2 upon cycling is consistent with the calculated Bi-Pd

Pourbaix diagram (Figure 3a). In 0.1M KOH solution (pH 13),

PdBi2 is predicted to become thermodynamically unstable

above 0.03 V vs RHE. Under equilibrium conditions, as the

voltage increases, Bi will dissolve from PdBi2 to produce

Bi2O3, leaving behind PdBi, then Pd5Bi2, then Pd3Bi, before fi-

nally converting to Pd and Bi2O3 above 0.35 V vs RHE. The ex-

perimental observation of converted Pd3Bi after cycling up to

1.1 V vs RHE indicates that the Pd3Bi is kinetically stabilized.

The kinetic stabilization may be due to the slow formation

of oxides upon Bi dissolution; for example, Bi2O3 is known to

precipitate slowly from solutions of Bi3+.26 If the formation of

oxides such as Bi2O3 is kinetically hindered, and conversion to

Pd is assumed to proceed via the increase in concentration of

aqueous Bi3+ (Figure 3b) or BiOH2+ (Figure SI5) ions, Pd3Bi can

be expected to persist up to approximately 0.90 V or 0.67 V

Figure 2. (a) Linear sweep voltammograms (LSVs) the PdBi2 in O2-saturated 0.1 M KOH before and after electrochemical etching recorded at 1600 rpm and 20 mV s−1 scan

rate. (b) Cyclic voltammograms (CV) of the PdBi2 in Ar-saturated 0.1 M KOH before and after the electrochemical etching. (c) HRTEM image of PdBi2 after conversion and (d)

its corresponding FFT pattern (e) which matches with the simulated Pd3Bi down the [120] direction. (f) Line plot of intensity vs reciprocal lattice spacing derived from the

rotationally-averaged SAED pattern shown in the inset (black) and simulated Pd3Bi (green) ICSD 58839. (g) HAADF-STEM image of PdBi2 after conversion and its correspond-

ing EDS map of (h) Pd, (i) Bi, and (j) Bi + Pd. All reciprocal distances in parts d & e are in units of nm−1

respectively vs RHE. The formation of surface oxides on the

nanoparticles may further contribute to their kinetic stability.

Moreover, the formation of crystalline Pd shells over Pd3Bi

(orthorhombic lattice) may also be kinetically hindered as the

crystal structure and lattice constants vary considerably from

the ground state of elemental Pd (FCC). Taken together these

data indicate that Pd3Bi is kinetically stable and forms by a

combination of Bi corrosion from PdBi2 and atomic restruc-

turing. To probe the evolution of Pd3Bi from PdBi2, we collected high

resolution TEM images on a sample which was not transformed to

completion (Figure 4). The image displays particles with a crystal-

line core, and an amorphous shell. HAADF-STEM imaging indicated

that the crystalline cores of the particles were denser than the

amorphous shells (Figure 4b-c), and energy dispersive x-ray spec-

troscopy (EDS) mapping indicated that the atomic ratio (Pd/Bi) of

the crystalline core is 1.08 ± 0.19 and shell is 0.84 ± 0.05 (2 sigma).

EDS mapping of Pd and Bi, and line scans of the elemental compo-

sition of the particles shown in Figure 4c indicates that the shell of

the particles are slightly richer in Bi than the core (Figure 4d-g).

Taken together, this data indicates that the removal of Bi from the

surface of the particle leaves behind a defect rich amorphous ma-

terial which allows facile transport of Bi out of the core, while the

core reconstructs to a crystalline intermetallic phase. This obser-

vation is consistent from our prediction from the Pourbaix dia-

gram that as Bi dissolves from PdBi2, intermetallic phases which

are more Pd rich are left behind as the particle is finally converted

to Pd3Bi.

Mobility of Atoms During Corrosion To investigate the rate

of phase conversion between the ordered intermetallic

phases, we used density functional theory27 to calculate the

minimum diffusion activation energy of Bi atoms in PdBi2, as-

suming a vacancy-mediated diffusion mechanism. Minimum

energy paths and saddle point energies were determined

with the climbing image nudged elastic band method.28-29 To

identify possible paths, we have taken into account all possi-

ble jumps between an atom and a vacant neighbouring site.

The lowest barrier for Bi diffusion in PdBi2 is 0.44 eV, corre-

sponding to a two-dimensional migration pathway (Fig. S6).

The mean square displacement of a particle undergoing a

two-dimensional random walk via a vacancy-mediated mech-

anism is ⟨∆𝑟2⟩ = 4𝑥𝑣 Γ𝑎2𝑡, where 𝑥𝑣 is the vacancy concen-

tration, Γ is the hopping frequency from a filled to a vacant

site, 𝑎 is the hopping length, and t is the time. We estimate

the vacancy concentration using𝑥𝑣 ≈ exp(−𝐸𝑓 /𝑘𝐵𝑇), where

𝐸𝑓 is the vacancy formation energy. We find that diffusion of

Bi through PdBi2 is likely facilitated by high vacancy concen-

trations, as the calculated vacancy formation energy for Bi

along the diffusion path is only 0.17 eV relative to bulk Bi. The

vacancy formation

energy will therefore be even lower relative to states in which

the chemical potential of Bi is lower than that of bulk Bi, such

as those that are predicted to be stable at higher potentials

(Figure 3). This low calculated vacancy formation energy is

Figure 3. (a) The calculated Pd-Bi Pourbaix diagram. (b) The Pourbaix diagram with oxides removed, so that dissolved Bi only forms aqueous Bi3+. Two-phase regions are marked

with a color and a symbol. The colors indicate the equilibrium Bi-oxide or solvated Bi-containing ion, and the symbol indicates the equilibrium phase that contains Pd. Single

phase regions are shaded white. The dashed red lines indicate the electrochemical stability window for water, with the lower line corresponding to 0 V vs RHE. The experimental

pH, 13, is marked by a dotted grey line.

Figure 4. (a) TEM image of PdBi2 after 200 cycles, the crystalline core is crystalline with

a lattice constant of 0.23 nm, while the shell is amorphous (b) HAADF-STEM images of

PdBi2 after 200 cycles indicating that the crystalline core is denser than the amorphous

shell. The large low density particles in the background are Vulcan carbon. c) HAADF-

STEM imaging of the region in part (d) highlighted in the red box. EDS mapping of (d) Bi,

(e) Pd, and (f) Bi + Pd. (g) EDS line scans of the boxed regions are shown in part (f).

consistent with the low melting point of PdBi2. Transition

state theory30 gives us Γ ≈ ν∗exp (−𝐸𝑎/𝑘𝐵𝑇), where ν∗ is

generally in the range of phonon frequencies and 𝐸𝑎 is the

activation energy for a single hop. Using the reported vibra-

tional frequency of around 1012 Hz for Bi in β-PdBi2 31, we es-

timate that at room temperature the root mean square dis-

placement of Bi in PdBi2 reaches a typical particle width in

well less than 1 second. This result indicates that Bi diffusion

in PdBi2 is facile, supporting the potential for fast removal of

Bi from PdBi2 under experimental conditions.

ORR Activity and Stability of Converted-Pd3Bi. The ORR ac-

tivity of the converted-Pd3Bi was compared with commercial

Pt/C (TKK, 40% Pt on high surface area carbon) and Pd/C

(Premetek,

40% Pd on Vulcan XC-72) catalysts (Figure 5a and SI-7). The

onset potential (the potential in which the current exceeds

the background by 10%) of ORR on both Pt/C was 1.0 V and

Pd/C was 1.05 V, which agreed with literature reports32; while

the converted-Pd3Bi was more negative (~0.97 V). At poten-

tials more positive than the onset potential the Pd and Pt

sites are covered with a surface oxide, rendering them cata-

lytically inactive. At more negative potentials (starting at

~0.95 V), we observed a rapid increase in current of the con-

verted-Pd3Bi, while the current for Pd/C and Pt/C increased

gradually. The half-wave potential of converted-Pd3Bi was

more positive by 60 mV and 32 mV relative to Pt/C and Pd/C,

respectively, indicating that it may possess superior catalytic

activity relative to the commercial catalysts.

The Koutecky-Levich (K-L) plot was constructed by meas-

uring the steady-state current density at a constant voltage

at various rotation rates to calculate the kinetic current den-

sity (Figure SI 8-9). The specific activity (SA) and mass activity

(MA) were measured by normalizing the kinetic current den-

sity to the ECSA (Figure SI-10) or to the total mass of precious

metal, respectively. The SA of converted-Pd3Bi was 2.3

± 0.19 mA/cmPd2 at 0.9V vs RHE, which is 19x higher than

Pt/C (0.12 ± 0.03 mA/cmPt2 ) and 4.4x higher than Pd/C (0.52

± 0.07 mA/cmPd2 ) (Figure 5b). The MA of converted-Pd3Bi

was 1.2 ± 0.08 A mgPd−1, which is 3.4x higher than Pd (0.36 ±

0.04 A mgPd−1) and 11x higher than Pt (0.11 ± 0.02 A mgPt

−1)

(Figure 5b).6 The SA and MA of Pd3Bi is among the highest

reported for alkaline ORR (Table S2). The high MA is remark-

able given the relatively modest specific surface area of Pd3Bi

(Figure 5C), decreases in particle size could permit even

higher mass activities.

We evaluated the stability of our catalysts by performing

accelerated durability testing by cycling the samples from 0.6

to 1.0 V at a sweep rate of 100 mV/s. Converted-Pd3Bi re-

tained 86% and 69% of the initial specific and mass activity,

respectively, after 10,000 cycles at 0.9V, while the specific

surface area decreased by 20% (Figure 5b-c and SI11). In

comparison, Pd/C retains 73% and 71% of the initial specific

activity and mass activity, respectively, and Pt/C retains 63%

and 54% of the initial specific activity and mass activity, re-

spectively (Figure 5b). The specific surface area of Pt/C and

Pd/C decreased by 14% and 3% respectively. This indicates

that the decreases in catalytic activity of the commercial sam-

ples is predominately caused by removal of highly active sites

by surface reconstructions and/or Ostwald ripening (Figure

SI12) .33 The decrease of activity for Pd3Bi was coincident with

the emergence of Ostwald ripening, as indicated by a de-

crease in the specific surface area, and the emergence of a

small amount of metallic Pd. The metallic Pd probably formed

via redeposition of dissolved Pd ion onto the electrode, as ev-

Figure 5. (a) LSVs of Pt/C, Pd/C, and converted-Pd3Bi in O2-saturated 0.1 M KOH. (b) Mass activity and specific activity at 0.9 V before and after stability test calculated from

the Kouteky Levich equation. Negative values indicate the current is cathodic. (c) Electrochemically active surface area (ECSA) and mass loading before and after stability test.

ECSA was measured by CO stripping, and the mass loading was measured by inductively coupled plasma mass spectrometry (ICPMS). (d) STEM-HAADF image of a Pd3Bi catalyst

particle after 10,000 cycles of ADTs. Inserts are FFT of the image and simulated single-crystal diffraction pattern down the [100] direction of Pd3Bi, respectively. All voltammetry

was collected on a 5mm diameter rotating disk electrode (RDE) at1600 rpm, and a voltage sweep rate of 20 mV s−1. The mass loading of all samples onto the 5 mm glassy carbon

disk was between ~13 to 15μg Pd or Pt cm−2geo

ident from the appearance of H-UPD in the CVs collected un-

der Ar after the ADTs (Figures SI13 and 5c, and Table S2). This

was further supported by ICP-MS and XPS (Table S3) which

showed the composition of the electrode was ~78% Pd and

~22% Bi after 10K cycles, indicating 1% higher Pd content rel-

ative to the freshly converted samples. HAADF-STEM images

of the sample after ADT test confirmed that the particles re-

tained intermetallic crystallinity (Figures 5d and Figure SI14).

Taken together, the data indicates that that the corrosion of

Pd3Bi to Pd was kinetically slow, hence it was able to sustain

remarkable performance after 10,000 cycles, even outper-

forming freshly prepared Pt/C and Pd/C.

Fuel crossover and poisoning of the cathode, where ORR

takes place, drastically decreases the activity and durability

of commercial direct methanol fuel cells.34-35 This is largely

due to the high catalytic activity of Pt and Pd for alcohol oxi-

dation;36 when fuel crosses over from the anode to the cath-

ode, the cathode experiences a mixed potential from both

oxidation and reduction reactions taking place simultane-

ously, eroding the current density and voltage of the device.

To examine the tolerance of converted-Pd3Bi towards fuel

crossover, we measured the ORR activity in the presence of

methanol. The ORR performance of Pt/C, Pd/C, and con-

verted Pd3Bi in O2-saturated 0.5 M CH3OH + 0.1 M KOH were

measured (Figure 6). Pd/C and Pt/C show significant increase

in the overpotential for ORR in the presence of methanol due

to the competitive methanol oxidation over ORR. The anodic

peak at 1.02 V for Pt/C and the 0.85 V peak for Pd/C are from

the oxidation of methanol, suggesting that ORR is largely sup-

pressed by the competing reaction. However, only a small 10

mV decrease of the E1/2 for ORR on converted Pd3Bi is ob-

served. This suggests that the converted-Pd3Bi has superior

methanol tolerance and is a promising cathode catalyst for

direct methanol fuel cells.37

Conclusions

In summary, we have demonstrated kinetically-controlled

corrosion as an effective approach to convert a base-metal

rich ordered intermetallic nanoparticles (PdBi2) to catalyti-

cally active ordered intermetallic nanoparticles (Pd3Bi). We

have employed the converted Pd3Bi nanoparticles as elec-

trodes for ORR with high methanol tolerance, the resulting

material exhibited superior mass activity relative to commer-

cial Pd/C and Pt/C. We found that a low melting point ordered

intermetallic nanoparticle can convert to a noble metal rich

ordered intermetallic with high catalytic activity due to facile

formation of vacancies and high mobility of interdiffusion.

This strategy opens a new synthetic method for accessing

high-performance heterogeneous catalysts in ambient condi-

tions.

Experimental Methods

Materials. Palladium (II) acetylacetonate (Pd(acac)2, 99%, Strem

Chemicals), Bismuth(III) acetate (Bi(oac)3, 99.99%, Alfa Aesar),

and oleylamine (OLAM, technical, 70%, Aldrich), 40% Pd on

Vulcan XC-72 (Premetek Co), O2 gas (UHP grade, Airgas), 40%

Pt on High Surface Area Carbon (TEC10E40E, Tanaka Kikinzoku

Kogyo TKK), were used as received without purification. All syn-

theses were carried out under Ar (g) using standard Schlenk tech-

niques, and work-up proce-dures were performed in air. Electro-

lyte solutions were prepared with deionized water (resistance:

18.2 MΩ) and potassium hydroxide (KOH, Semiconductor Grade

99.99%, Alfa Aesar).

Synthesis of PdBi2 nanoparticles (NPs). In a typical synthesis,

61 mg of Pd(acac)2 (0.2 mmol), 154 mg of Bi(oac)3 (0.4 mmol),

and 10 mL of OLAM were first added into a 20-mL scintillation

vial with a magnetic stir bar. After stirring for 20 mins, the vial

was sonicated for another 10 mins to ensure all the solid precur-

sors were dissolved and the mixture formed a pale-yellow solu-

tion. Another 5 mL of OLAM were added into a 100-mL 3-neck

round-bottom flask with a condenser, thermometer adapter, ther-

mometer, and rubber septum. Magnetic stirring was started and

the solution was degassed under vacuum at 120 °C for ~5-10

minutes. The flask was backfilled with Ar and heated to ~300 °C

at a rate of 10 °C/min. Once the temperature reached 300 °C, the

10 mL of precursor solution was slowly injected into the flask us-

ing a syringe pump at a rate of 0.3 mL/min. The solution in the

flask turned from colorless to light brown and finally formed a

dark brown colloidal mixture when the injection was completed.

The solution was then cooled rapidly by removing the flask from

the heating mantle. The nanoparticles were precipitated by adding

10 mL of ethanol and then centrifuged at 8000 rpm for 5 min. The

precipitate was washed three times using a 1:1 ethanol/toluene

mixture (with centrifugation in between washes) and could then

be suspended in ethanol for further characterization and electrode

preparation.

Physicochemical characterization. The ethanol suspension of

the nanoparticles was dropcasted onto a zero-background Si wafer

substrate and examined with a Philips X’Pert Pro Powder X-Ray

Figure 6. Linear sweep voltammograms of converted-Pd3Bi, Pd/C, and Pt/C in O2-

saturated 0.1 M KOH + 0.5 M CH3OH. Rotation rate: 1600 rpm. Sweep rate: 20 mV

s−1. The mass loading of all samples onto the 5 mm glassy carbon disk was between

~13 to 15μg Pd or Pt cm−2geo

diffractometer (XRD) with CuK radiation (K1, =1.5406 Å and

K2, =1.5444 Å). Transmission electron microscopy (TEM) was

performed on a FEI F200C Talos or FEI Tecnai TF30 operation at

200 kV. The materials on the glassy carbon disk after electro-

chemical test was dissolved by concentrated nitric acid and then

diluted with water. The metal concentration in the nitric acid solu-

tion is analyzed by the PerkinElmer NexION 300D with ICP.

Oxygen Reduction Measurements. Electrochemical measure-

ments were carried out with a potentiostat/galvanostat (Metrohm

Autolab Potentiostat or Nuvant EzStat Pro) and a three-electrode

Teflon cell with 0.1 M KOH aqueous solution as the electrolyte.

A Hg/HgO electrode in 1 M NaOH was used as the reference

electrode. A graphite rod was used as the counter electrode. The

Hg/HgO reference electrode was periodically checked with a Re-

versible Hydrogen Electrode (RHE) to verify the stability of the

reference potential. All data measured with IR compensation on

the potentiostat. ORR was performed by sweeping the voltage

from -0.9 V to 0.2 V vs. Hg/HgO at a sweep rate of 20 mv/s, or by

constant potential chronoamperometery within the same potential

range. All measurements were performed on a 5 mm diameter ro-

tating disk electrode at 1600 rpm. All ORR measurments were IR-

corrected, the IR drop was measured by the positive feedback or

current interrupt method.

CO Stripping Measurements. We utilized the CO stripping

method to determine the electrochemically active surface area

(ECSA) of Pt or Pd atoms on the surface of the electrode. The

electrodes are first held at 0.15V vs RHE for 20min in 0.1M KOH

saturated with CO to form a surface adsorbed CO monolayer, fol-

lowed by removal of CO (aq) by saturating the electrolyte with Ar

for 15 min. Finally, a the voltage was swept from 0.15V to 1.15

V vs RHE to oxidize the adsorbed CO layer, the surface area of

the sample can be calculated by integrating the oxidation peak as-

suming one CO adsorbed per Pd or Pt atom on the surface. The

ratio of charge transferred for CO oxidation and metal surface

area is estimated as 484 μC/cm2 for Pt and 420 μC/cm2 for Pd

and Pd3Bi. We expect that Bi sites will not adsorb CO, therefore

CO stripping will not account for catalytically inactive Bi sites on

the surface.

Conversion of PdBi2 to Pd3Bi. The electrode was prepared by di-

rectly dropcasting 4 L of the ethanol suspension of the as-pre-

pared nanoparticle onto the glassy carbon disk. After it dried out,

the electrode was then immersed in 0.1 M acetic acid solution for

20 mins followed by DI water wash, acetone wash, and ethanol

wash. Finally, the electrode was placed in a UV-ozone chamber

and treated with UV and ozone at room temperature for 20 mins.

A graphite rod and Hg/HgO electrode were used as the counter

and reference electrodes, respectively. The working electrode was

then conditioned by cycling the potential from -0.9 V to 0.2 V vs.

Hg/HgO reference at a scan rate of 0.1 V s-1 and at a rotation rate

of 1600 rpm for ~500 cycles.

Preparation of Pt/C and Pd/C electrodes. To prepare Pd/C ma-

terials, 30 mg of 40% Pd on Vulcan XC 72(Premetek Co) was dis-

persed into a solution consisting of a mixture of 8μl Nafion Iono-

mer, 1ml DI water and 1ml IPA. The resulting solution was soni-

cated for 20 min. 1μl of the mixture was drop casted onto the 5

mm diameter disk. The Pt/C sample is prepared with the same

procedure except with Pt/C (TKK).

Sample preparation for ICP-MS analysis. After the electro-

chemical measurement, the glassy carbon electrode was taken out

of the electrolyte and rinsed with 18.2 MΩ water. After dried in

air, the electrode was placed in a plastic tube and 1 mL concen-

trated TraceMetal grade nitric acid was added into the tube. The

tube was sonicated for a few minutes to fully dissolve the catalyst

on the surface. Additional 9 mL of 18.2 MΩ water was added to

the tube and the solution is ready for testing.

Computational Methods. The Pourbaix diagrams were generated using PyMatGen with data provided by the Materials Project1. To best view the possible formation of solid-state phases the concen-trations of aqueous species in the Pourbaix diagrams were set to the upper limit that could be achieved if all PdBi2 were dissolved. As it is now believed that the aqueous species previously identified as

BiO+ is in fact Bi6O4(OH)46+, the labels on the generated Pourbaix

diagram have been changed accordingly.2

XPS Analysis

Binding energies have been corrected by shifting the peak of C 1s

to 284.8 eV. Palladium peak positions have been assigned accord-ing to the values in Handbook of X-ray Photoelectron Spectros-copy3 and reported data in NIST database whereas for Bi to that of Casella and Contursi.4 The splitting of peaks usually occurs asym-metrically for metals in their lowest oxidation state and mostly symmetrically for higher oxidation states and the symmetry of peaks is controlled by the peak area. The FWHM is assumed to be in the range of (0.9-1.1 eV) for metals and in the range of (1.7-2.0)

for oxides. To properly fit the spectra, the separation between the peaks is set to 5.2 and 5.3 eV for Pd and Bi, respectively. By utiliz-ing relative sensitivity factors (R.S.F), the peak areas of the Pd and Bi are corrected, and their ratio is calculated to yield surface atomic ratio as

surface atomic ratio = Region peak areaPd/R. S. FPd

Region peak areaBi/R. S. FBi

ASSOCIATED CONTENT

Full experimental details of particle dealloying, nanoparticle film preparation, and characterization of the materials used in this arti-cle.

AUTHOR INFORMATION

Corresponding Author

* e-mail: shoji@jhu.edu

Author Contributions

‡These authors contributed equally.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT

A.S.H. acknowledges financial support from the National Sci-

ence Foundation under award no. CHE-1764310 and Johns

Hopkins University start-up funding. T.M. acknowledges fi-

nancial support from the National Science Foundation under

award no. CHE-1437396. We gratefully acknowledge Prof.

Mingwei Chen (JHU), and Prof. Jonah Erlebacher (JHU) for

helpful discussions.

.

REFERENCES

1. Coleman, E. J.; Chowdhury, M. H.; Co, A. C., Insights Into the Oxygen Reduction Reaction

Activity of Pt/C and PtCu/C Catalysts. ACS Catal. 2015, 5 (2), 1245-1253.

2. Mukerjee, S.; Srinivasan, S.; Soriaga, M. P.; McBreen, J., Role of Structural and Electronic

Properties of Pt and Pt Alloys on Electrocatalysis of Oxygen Reduction: An In Situ XANES and EXAFS

Investigation. J. Electrochem. Soc. 1995, 142 (5), 1409-1422.

3. Casado-Rivera, E.; Volpe, D. J.; Alden, L.; Lind, C.; Downie, C.; Vázquez-Alvarez, T.; Angelo,

A. C. D.; DiSalvo, F. J.; Abruña, H. D., Electrocatalytic Activity of Ordered Intermetallic Phases for Fuel

Cell Applications. J. Am. Chem. Soc. 2004, 126 (12), 4043-4049.

4. Wang, D.; Xin, H. L.; Hovden, R.; Wang, H.; Yu, Y.; Muller, D. A.; DiSalvo, F. J.; Abruña, H.

D., Structurally ordered intermetallic platinum–cobalt core–shell nanoparticles with enhanced activity and

stability as oxygen reduction electrocatalysts. Nat. Mater. 2013, 12 (1), 81-87.

5. Wencka, M.; Hahne, M.; Kocjan, A.; Vrtnik, S.; Koželj, P.; Korže, D.; Jagličić, Z.; Sorić, M.;

Popčević, P.; Ivkov, J.; Smontara, A.; Gille, P.; Jurga, S.; Tomeš, P.; Paschen, S.; Ormeci, A.; Armbrüster,

M.; Grin, Y.; Dolinšek, J., Physical properties of the InPd intermetallic catalyst. Intermetallics 2014, 55

(Supplement C), 56-65.

6. Cui, Z.; Chen, H.; Zhao, M.; DiSalvo, F. J., High-Performance Pd3Pb Intermetallic Catalyst for

Electrochemical Oxygen Reduction. Nano Letters 2016, 16 (4), 2560-2566.

7. Yan, Y.; Du, J. S.; Gilroy, K. D.; Yang, D.; Xia, Y.; Zhang, H., Intermetallic Nanocrystals:

Syntheses and Catalytic Applications. Advanced Materials 2017, 29 (14), 1605997.

8. Johannes, T.; Martin, H.; Jen-Hui, C.; Michael, R., Refinement of the Microwave-Assisted Polyol

Process for the Low-Temperature Synthesis of Intermetallic Nanoparticles. European Journal of

Inorganic Chemistry 2017, 2017 (42), 4930-4938.

9. Cable, R. E.; Schaak, R. E., Low-Temperature Solution Synthesis of Nanocrystalline Binary

Intermetallic Compounds Using the Polyol Process. Chemistry of Materials 2005, 17 (26), 6835-6841.

10. Wang, Y.; Sun, D.; Chowdhury, T.; Wagner, J. S.; Kempa, T. J.; Hall, A. S., Rapid Room-

Temperature Synthesis of a Metastable Ordered Intermetallic Electrocatalyst. Journal of the American

Chemical Society 2019, 141 (6), 2342-2347.

11. Ji, X.; Lee, K. T.; Holden, R.; Zhang, L.; Zhang, J.; Botton, G. A.; Couillard, M.; Nazar, L. F.,

Nanocrystalline intermetallics on mesoporous carbon for direct formic acid fuel cell anodes. Nat. Chem.

2010, 2 (4), 286-293.

12. Chen, H.; Wang, D.; Yu, Y.; Newton, K. A.; Muller, D. A.; Abruña, H.; DiSalvo, F. J., A

Surfactant-Free Strategy for Synthesizing and Processing Intermetallic Platinum-Based Nanoparticle

Catalysts. J. Am. Chem. Soc. 2012, 134 (44), 18453-18459.

13. Qi, Z.; Xiao, C.; Liu, C.; Goh, T. W.; Zhou, L.; Maligal-Ganesh, R.; Pei, Y.; Li, X.; Curtiss, L.

A.; Huang, W., Sub-4 nm PtZn Intermetallic Nanoparticles for Enhanced Mass and Specific Activities in

Catalytic Electrooxidation Reaction. J. Am. Chem. Soc. 2017, 139 (13), 4762-4768.

14. Chi, M.; Wang, C.; Lei, Y.; Wang, G.; Li, D.; More, K. L.; Lupini, A.; Allard, L. F.; Markovic,

N. M.; Stamenkovic, V. R., Surface faceting and elemental diffusion behaviour at atomic scale for alloy

nanoparticles during in situ annealing. Nature Comm. 2015, 6, 8925.

15. Kaiser, H., Selective dissolution of high and low diffusivity alloys—A comparison of kinetical

and micromorphological aspects. Corrosion Science 1993, 34 (4), 683-699.

16. Marshakov, I. K., Anodic Dissolution and Selective Corrosion of Alloys. Protection of Metals

2002, 38 (2), 118-123.

9

17. Wang, D.; Yu, Y.; Zhu, J.; Liu, S.; Muller, D. A.; Abruña, H. D., Morphology and Activity Tuning

of Cu3Pt/C Ordered Intermetallic Nanoparticles by Selective Electrochemical Dealloying. Nano Letters

2015, 15 (2), 1343-1348.

18. Oezaslan, M.; Hasché, F.; Strasser, P., Pt-Based Core–Shell Catalyst Architectures for Oxygen

Fuel Cell Electrodes. The Journal of Physical Chemistry Letters 2013, 4 (19), 3273-3291.

19. Sun, K.; Cheng, T.; Wu, L.; Hu, Y.; Zhou, J.; Maclennan, A.; Jiang, Z.; Gao, Y.; Goddard, W. A.;

Wang, Z., Ultrahigh Mass Activity for Carbon Dioxide Reduction Enabled by Gold–Iron Core–Shell

Nanoparticles. Journal of the American Chemical Society 2017, 139 (44), 15608-15611.

20. Medasani, B.; Haranczyk, M.; Canning, A.; Asta, M., Vacancy formation energies in metals: A

comparison of MetaGGA with LDA and GGA exchange–correlation functionals. Computational

Materials Science 2015, 101, 96-107.

21. Otake, S.; Ishii, Y.; Matsuno, N., Migration Energy of Vacancies in Bismuth. Japanese Journal

of Applied Physics 1981, 20 (6), 1037-1040.

22. Shao, M. H.; Huang, T.; Liu, P.; Zhang, J.; Sasaki, K.; Vukmirovic, M. B.; Adzic, R. R., Palladium

Monolayer and Palladium Alloy Electrocatalysts for Oxygen Reduction. Langmuir 2006, 22 (25), 10409-

10415.

23. Jiang, J.; Gao, H.; Lu, S.; Zhang, X.; Wang, C.-Y.; Wang, W.-K.; Yu, H.-Q., Ni–Pd core–shell

nanoparticles with Pt-like oxygen reduction electrocatalytic performance in both acidic and alkaline

electrolytes. Journal of Materials Chemistry A 2017, 5 (19), 9233-9240.

24. Sun, D.; Feng, S.; Terrones, M.; Schaak, R. E., Formation and Interlayer Decoupling of Colloidal

MoSe2 Nanoflowers. Chem. Mater. 2015, 27 (8), 3167-3175.

25. Sun, Y.; Wang, Y.; Sun, D.; Carvalho, B. R.; Read, C. G.; Lee, C.-h.; Lin, Z.; Fujisawa, K.;

Robinson, J. A.; Crespi, V. H.; Terrones, M.; Schaak, R. E., Low-Temperature Solution Synthesis of Few-

Layer 1T ′-MoTe2 Nanostructures Exhibiting Lattice Compression. Angew. Chem. Int. Ed. 2016, 55 (8),

2830-2834.

26. Wiberg, E.; Wiberg, N.; Holleman, A. F., Inorganic chemistry. Academic Press ; De Gruyter: San

Diego; Berlin; New York, 2001.

27. Kohn, W.; Sham, L. J., Self-consistent equations including exchange and correlation effects. Phys.

Rev. 1965, 140, A1133-A1138.

28. Henkelman, G.; Jonsson, H., Improved tangent estimate in the nudged elastic band method for

finding minimum energy paths and saddle points. Journal of Chemical Physics 2000, 113 (22), 9978-

9985.

29. Henkelman, G.; Uberuaga, B. P.; Jonsson, H., A climbing image nudged elastic band method for

finding saddle points and minimum energy paths. Journal of Chemical Physics 2000, 113 (22), 9901-

9904.

30. Vineyard, G. H., Frequency Factors and Isotope Effects in Solid State Rate Processes. J Phys

Chem Solids 1957, 3 (1-2), 121-127.

31. Zheng, J.-J.; Margine, E. R., Electron-phonon coupling and pairing mechanism in

β−Bi2Pd

centrosymmetric superconductor. Physical Review B 2017, 95 (1).

32. Jiang, L.; Hsu, A.; Chu, D.; Chen, R., Oxygen Reduction Reaction on Carbon Supported Pt and

Pd in Alkaline Solutions. J. Electrochem. Soc. 2009, 156 (3), B370-B376.

33. Wang, H.; Luo, W.; Zhu, L.; Zhao, Z.; E, B.; Tu, W.; Ke, X.; Sui, M.; Chen, C.; Chen, Q.; Li, Y.;

Huang, Y., Synergistically Enhanced Oxygen Reduction Electrocatalysis by Subsurface Atoms in Ternary

PdCuNi Alloy Catalysts. Advanced Functional Materials 2018, 28 (15), 1707219.

34. Sekol, R. C.; Li, X.; Cohen, P.; Doubek, G.; Carmo, M.; Taylor, A. D., Silver palladium core–

shell electrocatalyst supported on MWNTs for ORR in alkaline media. Appl. Catal. B 2013, 138

(Supplement C), 285-293.

10

35. Heinzel, A.; Barragán, V. M., A review of the state-of-the-art of the methanol crossover in direct

methanol fuel cells. J. Power Sources 1999, 84 (1), 70-74.

36. Ataee-Esfahani, H.; Wang, L.; Nemoto, Y.; Yamauchi, Y., Synthesis of Bimetallic Au@Pt

Nanoparticles with Au Core and Nanostructured Pt Shell toward Highly Active Electrocatalysts. Chem.

Mater. 2010, 22 (23), 6310-6318.

37. Gurau, B.; Smotkin, E. S., Methanol crossover in direct methanol fuel cells: a link between power

and energy density. J. Power Sources 2002, 112 (2), 339-352.

download fileview on ChemRxivSun-phase-convert-VF.pdf (1.18 MiB)

Supporting Information

Ordered Intermetallic Nanoparticles with High Catalytic Activity Prepared by an ElectrochemicallyInduced Phase Transformation

Du Sun,a+ Yunfei Wang,a+ Ken Livia, Chuhong Wang,a Ruichun Luo,b Zhuoqun Zhang,a Hamdan Alghamdi,c

Chenyang Li,a Fufei An, a,b Bernard Gaskey,a Tim Mueller,a Anthony Shoji Halla*

a Department of Materials Science and Engineering, Johns Hopkins University, Baltimore, Maryland 21218b School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai 200030, P. R. Chinac Department of Chemical and Biomolecular Engineering, Johns Hopkins University, Baltimore, Maryland 21218

Table of Contents

Figures and Tables..............................................................................................................................2Figure S1. Size distribution of PdBi2 nanoparticles........................................................................................3Figure S2. STEM-EDS line and point scan reveal that the shell of the particle consists of pure Bi. The EDS was not calibrated during data collection, hence the Bi:Pd ratio from the core is not the accurate ratio..................................................................................................................................................................... 4Figure S3. Additional High-res TEM images of the converted Pd3Bi particles.............................................5Figure S4. XPS spectra with peak fitting for the converted Pd3Bi before (A-B) and after (B-C) ADTs for the Pd 3d and Bi 4f regions. Peak label attributions, positions, and FWHM are shown in Table below thefigure................................................................................................................................................................... 6Figure S6. Bi in PdBi2 migration trajectory (left) and the activation barrier (right). Pd is depicted by the white spheres, and Bi is depicted with purple spheres. As the vacancy moves through PdBi2, Bi atoms on the opposite side of the plane of Pd atoms relax towards the vacancy..................................................7Figure S7. Particle size distributions of (a) Pd/C and (b) Pt/C.......................................................................8Figure S8. (a) Steady state measurement performed at different rotation rates and its corresponding (b)Koutecky-Levich plot. Data recorded in O2-saturated 0.1 M KOH at 1600 rpm............................................8Figure S9. Number of e- involved in the ORR at different potentials calculated based on K-L plot...........8Figure S10. Representative CO stripping voltammograms of Pd/C, Pt/C and converted-Pd3Bi.................9Figure S11. Specific activity and mass activity retention at 0.9 V vs RHE after stability testing...............9Figure S12. Particle size distributions of (a) Pt/C and (b) Pd/C after 10K cycles.......................................10Figure S13. CVs of the PdBi2 in Ar-saturated 0.1 M KOH at different stage compared with Pd and Bi....10Figure S14. HAADF-STEM images of the converted Pd3Bi particles after a 10,000 cycle ADT test, with its corresponding FFT and simulated FFTs...................................................................................................10Table S1. Summary of the ICP-MS and XPS results at different stage........................................................11Table S2. Summary of the mass specific activities at 0.9 V vs RHE measured at pH 13...........................11Table S3. Summary ICP-MS results of a sample and the solution after 10k cycles...................................11

References.......................................................................................................................................12 Figures and Tables................................................................................................................................................2

Figure S1. Size distribution of PdBi2 nanoparticles.........................................................................................3Figure S2. STEM-EDS of PdBi2.......................................................................................................................4Figure S3. Additional High-res TEM images of the converted Pd3Bi particles.............................................5Figure S4. XPS spectra with peak fitting.........................................................................................................6Figure S5. Calculated Pourbaix diagram with oxides other than Bi(OH)2+ removed...................................7Figure S6. Bi diffusion trajectory and activation barrier for diffusion.........................................................7Figure S7. Particle size distributions of electrodes..........................................................................................8Figure S8. Koutecky-Levich plots....................................................................................................................8Figure S9. Number of e- involved in the ORR at different potentials calculated based on K-L plot...........8Figure S10. CO stripping voltammograms......................................................................................................9Figure S11. Specific activity and mass activity retention after stability testing..........................................10Figure S12. TEM images of Pt/C and Pd/C after ADTs................................................................................10

1

Figure S13. CVs of the PdBi2 in Ar-saturated 0.1 M KOH at different stage compared with Pd and Bi..10Figure S14. HAADF-STEM images of the converted Pd3Bi particles after a 10,000 cycle ADT test.........11Table S1. Summary ICP-MS and XPS on samples at different stages........................................................12Table S2. ICP-MS of Pd3Bi and the solution after ADT................................................................................12Table S3. Summary of the mass specific activities at 0.9 V vs RHE measured at pH 13............................11

References............................................................................................................................................................12

Figures and Tables

2

Figure S1. Size distribution of PdBi2 nanoparticles.

3

Figure S2. STEM-EDS line and point scan reveal that the shell of the particle consists of pure Bi. The EDS wasnot calibrated during data collection, hence the Bi:Pd ratio from the core is not the accurate ratio.

4

Figure S3. Additional High-res TEM images of the converted Pd3Bi particles.

5

(c) (d)

Converted Pd3Bi before ADTs Converted Pd3Bi After ADTsElement

Attribution 1st peak 2nd peak 1st peak 2nd peak

Pd

Pd, metallic I:(335.4, FWHM=1.0)

I’:(340.7, FWHM=1.0)

I:(335.7, FWHM=1.0)

I’:(340.9,FWHM =1.0)

Pd(OH)x II:(336.4, FWHM=1.8)

II’:(341.6, FWHM=1.8)

II:(336.3, FWHM=1.8)

II’:(341.5,FWHM =1.8)

PdO III:(337.5,FWHM =1.9)

III’:(342.7,FWHM =1.9)

III:(337.4,FWHM =1.8)

III’:(342.6,FWHM =1.8)

PdOx IV:(339.5,FWHM =2.0)

IV’:(344.7,FWHM =2.0)

IV:(338.8,FWHM =1.7)

IV’:(344.0,FWHM =1.7)

satellite/plasmon

V:(339.1, FWHM=2.6)

V’:(346.5, FWHM=2.6)

V:(338.7, FWHM=3.5)

V’: (346.1,FWHM =3.5)

Bi

Bi, metallic I:(157.5, FWHM=1.1)

I’:(162.8, FWHM=1.1)

I:(157.5, FWHM=1.1)

I’:(162.8,FWHM =1.1)

Bi(OH)3 II:(158.4, FWHM=1.5)

II’:(163.7, FWHM=1.5)

II:(158.3, FWHM=1.8)

II’:(163.6,FWHM =1.8)

BiOOH III:(159.6,FWHM =1.7)

III’:(164.9,FWHM =1.7)

III:(159.6,FWHM =1.7)

III’:(164.9,FWHM =1.7)

Bi2O3 IV:(160.7,FWHM =3.2)

IV’:(166.0,FWHM =3.2)

IV:(160.7,FWHM =3.2)

IV’:(166.0,FWHM =3.2)

Surface atomic ratio of (Pd/Bi)

3.0 3.7

Figure S4. XPS spectra with peak fitting for the converted Pd3Bi before (A-B) and after (B-C) ADTs for the Pd3d and Bi 4f regions. Peak label attributions, positions, and FWHM are shown in Table below the figure.

6

Figure S5. The calculated Pourbaix diagram with oxides other than Bi(OH)2+ removed, so that dissolved Bi onlyforms aqueous Bi3+ or Bi(OH)2+. Two-phase regions are marked with a color and a symbol. The colors indicatethe equilibrium Bi-oxide or solvated Bi-containing ion, and the symbol indicates the equilibrium phase thatcontains Pd. Single phase regions are shaded white. The dashed red lines indicate the electrochemical stabilitywindow for water, with the lower line corresponding to 0 V vs RHE. The experimental pH, 13, is marked by adotted grey line.

Figure S6. Bi in PdBi2 migration trajectory (left) and the activation barrier (right). Pd is depicted by the whitespheres, and Bi is depicted with purple spheres. As the vacancy moves through PdBi2, Bi atoms on the oppositeside of the plane of Pd atoms relax towards the vacancy.

7

Figure S7. Particle size distributions of (a) Pd/C and (b) Pt/C.

Figure S8. (a) Steady state measurement performed at different rotation rates and its corresponding (b)Koutecky-Levich plot. Data recorded in O2-saturated 0.1 M KOH at 1600 rpm.

Figure S9. Number of e- involved in the ORR at different potentials calculated based on K-L plot.

8

Figure S10. Representative CO stripping voltammograms of Pd/C, Pt/C and converted-Pd3Bi.

Figure S11. Specific activity and mass activity retention at 0.9 V vs RHE after stability testing

9

Figure S12. Particle size distributions of (a) Pt/C and (b) Pd/C after 10K cycles.

Figure S13. CVs of the PdBi2 in Ar-saturated 0.1 M KOH at different stage compared with Pd and Bi.

Figure S14. HAADF-STEM images of the converted Pd3Bi particles after a 10,000 cycle ADT test, with itscorresponding FFT and simulated FFTs.

10

Table S1. Summary of the ICP-MS and XPS results at different stage.

Sample ICP-MS(Pd:Bi)[a]

ICP-MS(Pd:Bi)[b]

XPS(Pd:Bi)[b]

As-synthesizedPdBi2

171.5:724.2 1:2.2 N/A

ConvertedPd3Bi 152.3:83.6 77%:23% 76%:24%

Post-10k cycleADTs 120.2:61.8 78%:22% 76%:24%

[a] ppb, [b] atomic ratio

Table S2. Summary of the mass specific activities at 0.9 V vs RHE measured at pH 13.

Materials Mass Activity at 0.9V(A/mgPt or Pd)

Reference

Converted-Pd3Bi 1.2 This workPdCuNi-AB-t/C 0.45 5

PdCu/C 0.28 5

PdNi/C 0.29 5

Pd4Fe 0.98 6

PtPd 0.33 6

Pd6Ni/C 0.22 8

PdCo@NPNCs 0.47 9

Pd/W18O49 0.22 10

Pd-N2H4 0.20 11

Au@Pd0.1/PEI 0.29 12

Pd3Pb TPs/C 0.56 13

PdCuCo/C 0.13 14

Pd/C 0.27 7

Pd/C 0.36 This workPt/C 0.14 7

Pt/C 0.11 This work

Table S3. Summary ICP-MS results of a sample and the solution after 10k cycles.

Mass Change after ADT

　 Atomic Ratio(Pd/Bi)

Mass Ratio(Pd/Bi)

Mass Bi(ug) Mass Pd(ug)

11

ADT 10k Solution

2.464 1.255 0.271 0.340

Converted-Pd3Bi ADT 10k Reserved

3.523 1.794 1.193 2.140

Solution & RDE Sample

3.327 1.694 1.464 2.480

References

1. Persson, K. A., Waldwick, B., Lazic, P. & Ceder, G. Prediction of solid-aqueous equilibria: Scheme to combine first-principles calculations of solids with experimental aqueous states. Physical Review B 85, 235438, doi:10.1103/PhysRevB.85.235438 (2012).

2. Wiberg, E., Wiberg, N. & Holleman, A. F. Inorganic chemistry. (Academic Press ; De Gruyter, 2001).3. Wagner, C. D., Riggs, W. M., Davis, L. E., Moulder, J. F., Mouilenberg, G. E., Eds. Handbook of Xray

Photoelectron Spectroscopy; Perkin Elmer Corporation: Eden Prairie, MN, 19784. Casella, I.G. and Contursi, M. Characterization of bismuth adatom-modified palladium electrodes: The

electrocatalytic oxidation of aliphatic aldehydes in alkaline solutions Electrochimica Acta, 52, 649-657, (2006)

5. Wang, H. et al. Synergistically enhanced oxygen reduction electrocatalysis by subsurface atoms in ternary PdCuNi alloy catalysts. Adv. Funct. Mater. 28, 1707219 (2018)

6. Lian, C. et al. Synthesis and electrocatalytic properties for oxygen reduction of Pd4Fe nanoflowers. Chem. Commun. 54, 7058-7061 (2018)

7. St John, S. et al. Segregated Pt on Pd nanotubes for enhanced oxygen reduction activity in alkaline electrolyte. Chem. Commun. 51, 16633-16636 (2015)

8. Feng, Y. et al. Surface-modulated palladium-nickel icosahedra as high-performance non-platinum oxygen reduction electrocatalysts. Sci. Adv. 4, eaap8817 (2018)

9. Zhang, Z. et al. Facile synthesis of n-doped porous carbon encapsulated bimetallic PdCo as a highly active and durable electrocatalyst for oxygen reduction and ethanol oxidation. J. Mater. Chem. A 5, 10876-10884 (2017)

10. Lu, Y, Jiang, Y, Gao, X, Wang, X, Chen, W. Strongly coupled Pd nanotetrahedron/tungsten oxide nanosheet hybrids with enhanced catalytic activity and stability as oxygen reduction electrocatalysts. J. Am. Chem. Soc. 136, 11687-11697 (2014)

11. Poon, KC. et al. Newly developed stepwise electroless deposition enables a remarkably facile synthesis of highly active and stable amorphous Pd nanoparticle electrocatalysts for oxygen reduction reaction. J. Am. Chem. Soc. 136, 5217-5220 (2014)

12. Xue, Q. et al. Au nanowires@Pd-polyethylenimine nanohybrids as highly active and methanol-tolerant electrocatalysts toward oxygen reduction reaction in alkaline media. ACS Catal. 8, 11287-11295 (2018)

13. Bu, L. et al. Coupled s-p-d exchange in facet-controlled Pd3Pb tripods enhances oxygen reduction catalysis. Chem 4, 359-371 (2018)

14. Jiang, K. et al. Ordered PdCu-based nanoparticles as bifunctional oxygen-reduction and ethanol-oxidation electrocatalysts. Angew. Chem. 128, 9176-9181 (2016)

12

download fileview on ChemRxiv2019-Sun-phase-convert-SI-.docx (4.46 MiB)