Oxygen Binding to Active Sites of FeâNâC ORR Electrocatalysts

Subscriber access provided by Fudan University

Article

Oxygen Binding to Active Sites of Fe-N-C ORR Electrocatalysts Observed by Ambient-Pressure XPS Kateryna Artyushkova, Ivana Matanovic, Barr Halevi, and Plamen Atanassov J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b11721 • Publication Date (Web): 17 Jan 2017 Downloaded from http://pubs.acs.org on January 19, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Oxygen Binding to Active Sites of Fe-N-C ORR Electrocatalysts Observed by Ambient-Pressure XPS Kateryna Artyushkova,*,a Ivana Matanovica,b, Barr Halevic and Plamen Atanassova a.

University of New Mexico, Chemical & Biological Engineering Department, UNM,

Center for Miroengineered Materials, University of New Mexico, Albuquerque, NM 87131 (USA). *E-mail: [email protected] b.

Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico

87545, (USA). c.

Pajarito Powder, LLC. Albuquerque, NM 87102 (USA)

Abstract

We report the first in-situ Ambient Pressure X-Ray Photoelectron Spectroscopy(APXPS) study of the binding of oxygenated species to the active sites of iron-nitrogen-carbon Oxygen Reduction Reaction (ORR) electrocatalysts. To better interpret the results DFT calculations were used to calculate absorption energies of reactants and intermediates on potential active sites and calculate the core level shifts for those. The observed oxygen binding to nitrogen 1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

coordinated to iron centers correlates with the enhanced measured ORR fuel cell activity of these materials with respect to metal-free analogs and sheds light on the ORR mechanism on PGM-free electrocatalysts.

Introduction Development of active and durable non-platinum group metal electrocatalysts based on metal-nitrogen-carbon (M-N-C) structures is dependent on an understanding of the active sites responsible for the ORR activity observed. This is a challenging task that has recently become possible using a combination improved spectroscopic techniques, coupled with analysis that relies on Density Functional Theory (DFT) to more accurately interpret the spectroscopic analysis.

Computational approaches based on DFT have become widely accepted for

studying the absorption energies of reactants and intermediates on possible active site structures in its relationship to oxygen reduction mechanism.1 Experimental confirmation of the possible active sites structures is essential to elucidating the fundamental structureproperty relationships in these catalytic materials. 2 The structure-property relationships in M-N-C materials are usually established through correlations of electrocatalytic activity with such spectroscopic methods as XPS, XANES, and Mossbauer spectroscopy.3-7 Based on both XPS and

57

Fe Mossbauer spectroscopy, a direct

correlation between the amount of Fe-Nx sites and the kinetic current density of ORR has been established.7-10 Building direct structure-to-property relationships for electrocatalysts based on spectroscopic tests done under near-realistic reaction-like ambient conditions bridges the gap between the surface science and realistic fuel cell operation. Accordingly, the in-situ X-ray spectroscopic analysis identified a mechanism of ORR in acidic media with the Fe-N4 site serving as the primary oxygen adsorption site.6, 11 To address the importance of nitrogen functionalities in ORR, particularly those not directly coordinated to the metal, such 2 ACS Paragon Plus Environment

Page 2 of 26

Page 3 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


as pyridinic, pyrrolic and graphitic nitrogen, other in-situ spectroscopic methods have to be utilized. XPS is a very effective surface spectroscopic technique for the analysis of nitrogencontaining electrocatalysts because it allows quantification of the types of carbon, nitrogen and iron species present.8,

12-13

Synchrotron-based ex-situ XPS was previously used to

examine oxygen reduction intermediates which attach to active sites on nitrogen-doped graphene samples.14 XPS was previously limited to ultra high vacuum (UHV) regimen, but recent advances in XPS instrumentation allow XPS measurement from samples under as much as 25 torr of reactive gasses.15-16 In a previous study, we used Ambient Pressure XPS (APXPS) to probe the changes in the chemistry of nitrogen and cobalt in oxidizing (oxygen/water) and reducing atmospheres.17 The sensitivity of XPS towards change in oxidation state under in-situ conditions was demonstrated for other catalytic systems. 18-19 In the current study, we are focusing on a family of recently developed nano-structured non-platinum catalysts based on Fe and nitrogen-containing precursors such as carbendazim20 and nicarbazin6 that has been shown to have high and stable performance.21 To address the role of nitrogen and metal in the active sites in ORR mechanism occurring in M-N-C materials, we combined theoretical and experimental approaches: 1) APXPS characterization of Fe–containing electrocatalyst and a metal-free analogue in a thermal and gas environment that simulates fuel cell operating conditions; 2) calculation of oxygen and –OH adsorption energies on possible active sites; 3) calculations of core level shift (CLS) caused by possible adsorbate/active sites configurations and 4) fuel cell Membrane Electrode Assembly (MEA) tests of the Fe-containing and metal-free analogue electrocatalysts studied by APXPS. A plurality of sites energetically attractive for oxygen binding exist in both metal-free and Fecontaining electrocatalysts promoting multiple pathways of oxygen reduction. Atomically dispersed atom coordinated with nitrogen is the most energetically beneficial site catalyzing 3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 26

full reduction of oxygen to water. Hydrogenated and graphitic nitrogens which are present in both metal-free and metal-containing electrocatalysts are also favorable sites for binding of oxygenated species, but they catalyze partial oxygen reduction to hydrogen peroxide.

Experimental Section Sample preparation. Sacrificial support method (SSM) was used to synthesize M-N-C catalyst from iron and nitrogen containing precursors. In this method, a sacrificial support, Cab-O-Sil® fumed silica is combined with iron nitrate and mixture of nicarbazin (1,3-bis(4nitrophenyl)urea;

4,6-dimethyl-1H-pyrimidin-2-one)

and

carbendazim.

After

high

temperature heat treatment in a nitrogen atmosphere, the silica template and iron-rich phase are leached by concentrated HF and HNO3. The details of synthesis are presented in a recent report by Serov et.al.6, 20 Different ratios of nicarbazin and carbendazim were utilized in the scaled up by PPC from UNM technology to produce three M-N-C materials of variable performance and composition. A metal-free analog was synthesized using the same procedure from the same mixture of nicarbazin and carbendazim as sample M-N-C #1 but without Fecontaining salt. The suspensions of catalyst powders in IPA were sprayed onto gas diffusion layers for analysis. XPS. Kratos Axis Ultra DLD spectrometer was used to characterize samples for overall elemental composition. Monochromatic Al Kα source at 225 MW power was used to acquire high resolution spectra from three areas per sample without charge neutralization. The operating pressure was 3 × 10-10 mbar. Ambient Pressure XPS: In situ XPS were carried out at ISISS (Innovative Station for In Situ Spectroscopy), the catalysis beamline of the Fritz Haber Institute at the 3rd generation synchrotron BESSY II (Helmholtz-Zentrum Berlin). Samples were mounted, heated to 60°C, N 1s and C 1s high resolution spectra were first acquired at UHV at a pressure of 2 × 10-9 4 ACS Paragon Plus Environment

Page 5 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


mbar. The sample was then exposed to 1:4 O2:H2O at 0.5mbar and the same spectra were acquired. The spectra were acquired at primary X-ray source energy of 550 eV. This corresponds to a sampling depth of nitrogen 1s photoelectron of ~2.5 nm. Data analysis and quantification from both XPS and APXPS were performed using CasaXPS software. High resolution N 1s spectra were fitted with a 70% Gaussian/30% Lorentzian line shape with fixed full-width half max of 1.3-1.5 eV. No calibration was necessary for XPS spectra, while Fermi edge was used to calibrate all spectra acquired at APXPS end station. DFT calculations. All Density functional theory calculations were performed using gradient approximation (GGA) with the vdW-DF functional proposed by Dion et al. 22-24 and projector augmented-wave pseudopotentials 25-26 as implemented in Vienna Ab initio Software Package (VASP) 27-28. In all cases, the surface of a graphene sheet was represented by a 9.84 Å x 8.52 Å supercell containing 32 carbon atoms. Vacuum region of 14 Å was introduced in order to eliminate interactions between the graphene sheet and its periodic images. The electronic energy was calculated using 8x8x1 k-point Monkhorts-Pack

29

mesh and Fermi-Dirac

smearing method. In all cases, plane-wave basis cutoff was set to 800 eV. Core electron binding energies were calculated from the N 1s core state in the final state approximation. Binding energy was then determined as BE=BE(N 1s) – EF, where BE( N 1s) is the computed binding energy and EF is the energy of the Fermi level. In order to avoid the errors associated with the description of the core electrons, binding energies are always expressed relative to the binding energy of the 1s electron of the reference pyridinic-N defect. Experimental N1s binding energies can, thus, be obtained by adding DFT calculated CLSs to 398.8 eV, which is the experimentally determined N1s BE of the pyridinic-N. MEA testing. Membrane electrode Assemblies (5 and 50cm2) were made from Gas Diffusion Electrodes(GDE) pressed with 211 Nafion® membrane with PTFE-impregnated 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

glass-fiber sub-gaskets at 131oC for 10minutes (90psi for 5cm2 and 70psi for 50cm2 sized electrodes), then allowed to cool under 1psi pressure. Sub-gaskets thickness for the anode was 150 micrometers and 250 micrometers for the cathode. The gas diffusion electrode was sprayed using a Sono-tek Exacta-Coat automated spray system delivering 4 ml/min ink through a 25kHz ultrasonic nozzle onto SGL 25BC Gas Diffusion Layer(GDL) materials heated to 65oC. The ink was deposited at a rate of 40microgram/cm2 per deposition pass, for a total of 3mgcatalyst/cm2 and 75micrometer thick electrode. The inks were composed of 2:1 Isopropyl Alcohol:Deionized water (v:v), catalyst, and D2021 Nafion® dispersion mixed to a ratio of 3.5wt% solids. A 50mL vessel containing the ink ingredients was placed in a water bath and mixed for 30minutes using an IKA T-18 high shear mixer with the S18-19G dispersing element set for 18,000RPM. The MEAs were loaded into the cell testing assembly(Fuel Cell technologies) using single(5cm2) or triple(50cm2) serpentine pattern graphite flow plates and the cell hardware was assembled using 40inch-lbs torque for 5cm2 and 50inch-lbs for 50cm2 cells. The cell was allowed to come up to temperature under a feed 200sccm of 100% RH 2.5Air/1.5H2 and pressurized to a total of 1.70Atm total pressure. The cell temperature was raised to 80oC. After the temperature was achieved, the cell was held at 0.3V for 10 minutes, 0.6V for 10 minutes, and potentiostatic polarization curves were collected at different pressures. The polarization curves reported are the third curve, approximately 90 minutes after coming online, with the current reported after 60 seconds hold at each potential. Results and discussion. XPS Elemental composition of four samples, one of which is metal-free and the other three are containing Fe are presented in Table 1.

6 ACS Paragon Plus Environment

Page 6 of 26

Page 7 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. XPS elemental composition of catalysts powders C 1s %

N 1s %

O 1s %

Fe 2p %

MNC #1

95.4

2.7

1.9

0.18

MNC #2

94.4

2.2

3.4

0.15

MNC #3

94.8

2.3

2.8

0.18

NC

95.2

3.2

1.6

Metal-containing samples have between 2.2 and 2.7 at. % of nitrogen detected and between 0.15 and 0.18 at. % of iron detected. Thorough acid leaching ensures removal of iron-rich phase such as iron oxide nanoparticles with only atomically dispersed iron in coordination with nitrogen remaining. The metal-free sample has 3.2 at% of nitrogen detected. There is a variability in the elemental composition in metal-containing samples resulting from different formulations of nitrogen precursors used in the synthesis as discussed in the Experimental section. Using conventional laboratory-based X-ray source the information depth of nitrogen and iron is quite different, being ~11 nm and 3 nm, respectively. The amount of iron detected is very low due to intentional thorough leaching steps guaranteeing presence of only atomically dispersed iron

6, 20

and low attenuation length and large cross-section of Fe 2p

electrons. Therefore, for accurate determination of overall atomic % of iron within the surface of material, 3-4x longer acquisition times than those for % of nitrogen are necessary. Ambient Pressure XPS Using synchrotron sources with variable energy allows analysis of the chemistry of materials at depths lower than is possible using laboratory-based X-ray sources because the energy of the X-rays exciting the detected photoelectrons can be varied and therefore allowing only the outermost few layers of the surface and adsorbates to be analyzed. This allows unprecedented acuity in shedding light on possible oxygen binding to the surface of the material. As the change in chemical composition upon exposure to the oxidizing atmosphere 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 26

is of primary interest in APXPS studies, O 1s and Fe 2p spectra were not acquired. The APXPS spectra were acquired at a primary X-ray energy of 550 eV. Only C 1s and N 1s spectra can be acquired at this primary energy of source. This energy is much lower than that of laboratory-based X-ray spectrometers operating Al monochromatic source and producing highly energetic N 1s photoelectrons from as deep as 11 nm. The 550 eV X-rays create N 1s photoelectrons from much shallower depths as low as 2-3 nm. Very low absolute amounts of Fe present, smaller attenuation length and much larger photoionization cross-section of Fe 2p electron creates difficulty in acquiring Fe 2p spectra with good enough signal to noise ratio for detecting the shift in spectra that occurs upon binding of oxygenated species. For this reason, we have focused on high resolution spectra of nitrogen photoelectron, as they have a high signal to noise and are sensitive to binding energy shifts as small as 0.1 eV. Moreover, comparison of changes in nitrogen chemistry of metal-free and metal-containing systems allows for clear understanding the role of metal in oxygen binding and subsequent reduction. Figure 1 shows N 1s spectra acquired from the metal-free (Figure 1d) and a metalcontaining MNC #1 (Figure 1a) samples in UHV. The spectra were fitted using 6 peaks representative of the chemical species typically found in M-N-C catalysts3,

10, 30

and in

conformity with our previous systematic studies.8, 31 The peak at the lowest BE of 398 is assigned to imine or cyano groups while the peak at 398.8 eV is due to pyridinic nitrogen and metal coordinated to nitrogen in disordered states such as N2-Fe and N3-Fe.

32

As was

reported previously based on the correlation between Mossbauer and XPS spectra as well as DFT calculations, different types of sites in which nitrogen is coordinated to Fe can contribute to the binding energy of ~398.8 eV where the pyridinic peak is observed.10, 32 The peak at 399.9 eV has a contribution from amine nitrogens in the metal-free sample.


Page 9 of 26

V H U C N

V H U C N M

d)

a)

c)

1.0

e)

O2

b)

H / 2 O C N

O2

H / 2 O C N M

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


f)

UHV O2 H2O

0.8 0.6 0.4 0.2 0.0 404

402

400

398

396

Binding energy, eV

Figure 1. N 1s spectra for MNC #1 and NC electrocatalysts: a) and d) in UHV; b) and e) in O2/H2O at 60oC. Overlay of spectra acquired in UHV and oxygen/water for c) MNC and f) NC electrocatalysts. In metal-containing samples, there is also a contribution from metal coordinated to nitrogen groups in a mesomeric N4-Fe configuration, as have been confirmed by previous studies and DFT calculations of binding energy shifts.17, 32 The peak at 400.7 eV is due to pyrrolic and hydrogenated N-H nitrogens, and peaks at 401.8 and 402.7 eV are due to graphitic 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

nitrogens.31 Namely, using DFT with the vdW-DF functional proposed by Dion et al 22-24 N 1s binding energy of the hydrogenated pyridinic and pyrrolic nitrogen was calculated as 401.0 eV and 401. 1 eV, respectively. 33 DFT helps confirm that both defects have very similar N 1s binding energies and contribute to a broad peak at 400.7 eV. Figure 1 b) and e) show N 1s spectra acquired in the presence of oxygen in solvated environment for metal-containing and metal-free samples, while Figure c) and f) show overlay between spectral signatures obtained in reducing and oxidizing atmosphere to highlight the spectral changes occurring. The peaks highlighted in the spectra in Figure 1 have largest changes in their relative abundance upon transition from reducing to the atmosphere with oxygen and oxygen in water vapor. Table 2 shows the distribution of nitrogen species in UHV conditions and percent change in each peak for three metal-containing and one metal-free catalyst. MNC #1 sample has the highest amount of nitrogen coordinated to metal among metal containing samples. It also has the smallest amount of graphitic nitrogen. Keeping in mind uncertainties of XPS measurements, only the change in the distribution of peaks that is larger than 5% can be considered statistically significant. For metal-containing electrocatalyst, major change happens in the range of peaks where different types of metal coordinated to nitrogens are located. These peaks increase in relative intensity by as much as 56% upon exposure to oxygen in humidified atmosphere. The change is smaller when samples are exposed to oxygen only. The intensity of peak due to hydrogenated pyridinic and pyrrolic nitrogen peak decreases slightly as well. The change that occurs in the metal-free sample is strikingly different. The sizes of the hydrogenated pyridinic and pyrrolic nitrogen peak decrease significantly while other nitrogen type peaks increase, with graphitic and pyridinic peaks increasing the most.


Page 10 of 26

Page 11 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Relative composition at UHV and percent change in nitrogen peak abundance upon exposure to O2 and O2/H2O with respect to rel.% detected at UHV conditions UHV,

N

rel %

imine

N

pyridine/

Nx-Fe (x

Oxygen Binding to Active Sites of FeâNâC ORR Electrocatalysts

Recommend Documents

Oxygen Binding to Active Sites of FeâNâC ORR Electrocatalysts