C Catalyst in Acetate

Feb 25, 2019 - Non-precious metal catalysts (NPMCs) have gained significant attention over the past decade as realistic alternatives to platinum-based...
1 downloads 0 Views 831KB Size
Subscriber access provided by AUBURN UNIV AUBURN

Article

Enhanced Oxygen Reduction Reaction on Fe/ N/C Catalyst in Acetate Buffer Electrolyte Kaspar Holst-Olesen, Luca Silvioli, Jan Rossmeisl, and Matthias Arenz ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b04609 • Publication Date (Web): 25 Feb 2019 Downloaded from http://pubs.acs.org on February 27, 2019

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 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 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.

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 22 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

ACS Catalysis

Enhanced Oxygen Reduction Reaction on Fe/N/C Catalyst in Acetate Buffer Electrolyte Kaspar Holst-Olesen1†, Luca Silvioli1†, Jan Rossmeisl1*, Matthias Arenz2* Department of Chemistry, Nano-Science Center, University of Copenhagen, Universitetsparken 5, 2100 Ø, Copenhagen, Denmark. 1

2

Department of Chemistry and Biochemistry, University of Bern, Freiestrasse 3, 3012, Bern, Switzerland.

Keywords: ORR, Non-Precious Metal Catalyst, RDE, DFT

1. Abstract

Graphical abstract

Non-Precious Metal Catalysts (NPMCs) have gained significant attention over the past decade as realistic alternatives to platinum-based catalysts for the oxygen reduction reaction (ORR) in fuel cells. An interesting feature of NPMCs is that the active site can be regarded as a 2D structure where both sides influence catalytic processes. Such a 2D structure enables different possibilities to alter the active site through the specific chemical environment as compared with typical 3D materials such as Pt based catalysts. In this work, we focus on the effect of carboxylate buffer species in the vicinity of iron-nitrogen-carbon (Fe/N/C) catalytic sites. The catalytic interface is studied with respect to the ORR activity using the rotating disk electrode (RDE) in aqueous electrolyte as well as theoretical modelling using density functional theory (DFT) calculations. We find that the ORR

1 ACS Paragon Plus Environment

ACS Catalysis 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 2 of 22

activity on Fe/N/C catalyst is promoted by carboxyl species in general and increases 4-fold in acetate buffer as compared to 0.1 M aqueous HClO4 electrolyte.

2. Introduction Hydrogen fuel cells are considered a promising clean technology for automotive purposes, back-up power supply and decentralized heating and power supply. The technology is however still facing several obstacles for entering the commercial market on a large scale. One of the main challenges is related to the use of platinum-based catalysts to promote the oxygen reduction reaction (ORR) taking place at the cathode of the fuel cell. The high cost and scarcity of platinum significantly increase the price of fuel cell stacks and put a limit to the amount of fuel cells that can be produced. Hence, there has been significant research for alternative catalysts using only non-precious metals and abundant materials. Jasinski was the first to show that the ORR could be catalyzed by porphyrin structures with a metallic center1. Yeager later found that these structures could be incorporated into carbon supports by simple procedures (pyrolysis) from inexpensive starting materials2. In the wake of this discovery, attention has been focusing on improving the activity of these catalysts and recent works have stated great improvements in this direction3-5. Due to the structure and components of these catalysts, they are often referred to as Me/N/C or NPMCs in the literature6. However, the exact structure of the catalytically active site has been elusive and is still a matter of debate. Several different synthesis strategies are reported but for the transition metal centered types, the proposed structures Me-N2+2 and Me-N4 appear to be valid candidates3, 7-15. Interestingly, these catalysts have been reported to act differently towards chemical species (or compounds) that are regarded poisonous to platinum-based catalysts16-17. This suggests that the active site(s) of NPMCs is (are) fundamentally different from the bulk metallic Pt-site, which opens up new

2 ACS Paragon Plus Environment

Page 3 of 22 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

ACS Catalysis

perspectives, when investigating these catalysts18-20. The main challenge for both Pt-based and NPM catalysts is the substantial overpotential needed to drive the ORR, eventually leading to a decrease in overall fuel cell performance. In a recent study it was proposed, based on theoretical modelling, that certain chemical groups such as -COOH in the vicinity of the NPMC active site promote the ORR21. Similar effects have also been reported for macrocyclic compounds, both based on experiments and based on theoretical calculations22-24. In order to investigate the performance enhancing effect of the interaction of chemical groups with the NPMC active site further, we studied the effect of several buffer electrolytes on the ORR on a commercially available NPMC of the type Fe/N/C (Pajarito Powder, NPC-2000). The study is based on a series of RDE measurements in combination with DFT-calculations. The influence of the buffer groups on the ORR is benchmarked against the ORR performance of the same catalyst film in an electrolyte that is assumed not to interact with the active sites of the catalyst, i.e. perchloric acid solution. The electrochemical measurements reveal that carboxylate buffer electrolytes improve the ORR significantly as compared to a “non-interacting” electrolyte. Based on DFT-calculations we discuss possible mechanistic reasons behind these experimental findings.

3. Experimental Electrochemical details All RDE measurements were performed in an electrochemical half-cell with a 3-electrode setup. The setup consisted of a double-walled glass cell, enabling temperature control. The electrochemical cell was fitted with several glass joints on top, which supported the rotating disk electrode (RDE) in a central position and additionally a direct gas inlet made of glass, a closed glass tube containing a thermo-logger, a glassy carbon rod counter electrode (CE) and a reference electrode (RE). The RE

3 ACS Paragon Plus Environment

ACS Catalysis 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 22

was a reversible hydrogen electrode (RHE) in a closed glass tube equipped with a sealed in ceramic frit (Mettler) as liquid junction to the cell. The volume of electrolyte in the electrochemical cell was ~225 ml. The temperature was controlled through a thermostat (Lauda, Germany) connected to the outer jacket of the cell. The electrolyte solutions were prepared from >99 % CH3COOH/CH3COONa (Sigma-Aldrich), 85 % ortho-phosphoric acid (Suprapur, Merck), NaOH (Emsure, Merck) and 70 % HClO4 (Suprapur, Merck) and ultra-pure water (Milli-Q, ρ = 18.2 MΩ cm, Total organic carbon ≤ 5 ppb). Argon gas (purity ≥ 99.999 %), oxygen gas and hydrogen gas (purity ≥ 99.995 %) were supplied by Air Liquide, Denmark. The RDE (EDI101, Radiometer) was combined with a software-controlled rotation control unit (CTV101, Radiometer) and the 3-electrode system was interfaced to an ECi200 potentiostat operated with the EC4U software package (Nordic Electrochemistry). In all measurements the solution resistance (iR drop) was online recorded with the potentiostat by superimposing a 5 kHz, 5 mV AC signal and compensated for by an analogue positive feedback scheme. The effective solution resistance was adjusted to 3 Ω. The WE was a 5 mm (ø) polycrystalline platinum or glassy carbon disk, mounted into an 11 mm (ø) PEEK or PTFE cylinder respectively. The WE was polished to a mirror finish on polishing pads with alpha alumina suspensions of 1.0 and 0.3 µm grain size respectively (Struers, Denmark). The Fe/N/C ink was produced by adding 23 µl 10 % aqueous Nafion solution (Sigma-Aldrich) and 490 µl absolute ethanol (99.8 %) to 4 mg dry Fe/N/C powder (Pajarito Powder). This results in a Nafion/Carbon ratio of 0.6095 (with the approximation that all catalyst powder consists of carbon). The mixture was treated in an ultrasonic bath for one hour keeping the suspension cold. Catalyst films were made by applying 20 µl of ink solution onto a GC tip and drying for approximately two hours in air. Uniform films with a loading of 800 µg cm-2 were used exclusively. The catalyst films were introduced into oxygen saturated electrolyte (25 °C) and conditioned by cycling between 0.1 and 1.0 VRHE at a scan 4 ACS Paragon Plus Environment

Page 5 of 22 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

ACS Catalysis

rate of 100 mV s-1 and a rotation rate of 900 rpm until a stable behavior was observed (typically around 25 cycles). The ORR polarization curves were thereafter measured applying a scan rate of 10 mV s-1 and a rotation rate of 900 rpm unless otherwise stated. For separating catalytic and double layer charging processes the measurements are background corrected, i.e. the corresponding curves recorded in inert electrolyte are subtracted from the ORR curves. From the linear sweep voltammetry (LSV) curves, which are provided in the supplementary information (SI), the kinetic currents are calculated using the Koutecky-Levich equation25. The stated pH values were verified by an electronic pH-meter.

4. Results and Discussion Electrochemical measurements At first, we compare the ORR performance of the Fe/N/C catalyst in aqueous HClO4 electrolyte solution, which is usually considered as non-interacting, and in an acetate buffer electrolyte (pH 5.1), respectively. In Figure 1, RDE measurements conducted on the same catalyst film of NPM catalyst first in 0.1 M HClO4, then in acetate buffer electrolyte and finally in 0.1 M HClO4 again, are shown. A clear difference is observed between the ORR in the two electrolytes and the Tafel plot indicates that the ORR mass activity increases ~4 times in the acetate buffer as compared to the HClO4 electrolyte. This is a significantly higher enhancement factor than previously observed for phosphoric acid25. Furthermore, the Tafel slopes in the acetate buffer and the “blank” 0.1 M HClO4 electrolyte are the same (parallel Tafel plots) over a wide potential range indicating that the ORR reaction pathway does not change in the two electrolytes. The reproducibility of this behavior in the measurements conducted both before and after subjecting the catalyst layer to the acetate buffer (see Figure S1 in the SI for the LSV curves) demonstrates that this is not an artifact due to an

5 ACS Paragon Plus Environment

ACS Catalysis 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 6 of 22

imperfect catalyst film. Furthermore, the measurements demonstrate that the catalyst film has not undergone any permanent chemical changes during the measurements in the buffer solution. This is also supported by the cyclic voltammetry (CV) performed in argon purged electrolyte, where the two CVs recorded in 0.1 M HClO4 show a complete overlap. The CV recorded in the acetate buffer by comparison exhibits slightly less capacitive currents (appearing narrower) as compared to the behavior in 0.1 M HClO4 electrolyte, indicating that the acetate/acetic acid does not enhance the electrochemical active surface area (ECSA) of the catalytic layer. In fact, the observed behavior could be interpreted as a slightly reduced ECSA in the buffer solution. As the change is reversible, however, we favor the interpretation of the slightly reduced polarization currents due to small, reversible changes in the polarizability of the catalyst-electrolyte interface. Last but not least, it should be mentioned that the observed effect is not limited to the Fe/N/C catalyst from Pajarito but has also been confirmed for a non-commercial “Dodelet-type” Fe/N/C catalyst provided by Kramm et al.26 (not shown). For the sake of reproducibility, however, all presented measurements were conducted with the commercial catalyst.

Figure 1 - Difference between a) the ORR performance in oxygen purged electrolyte (Tafel plot) and b) the CV in argon purged electrolyte of the Fe/N/C catalyst in 0.1 M HClO4 and 0.5 M acetate buffer (pH 5.1) respectively. The same catalyst layer (WE) was used throughout the whole experiment and (i) and (ii) refer to before (i) and after (ii) the measurements in the acetate buffer. The electrode was gently rinsed in Milli-Q water between the measurements. All measurements were performed at 25 °C with a scan rate of 10 mV∙s-1 and 900 rpm rotation. The corresponding ORR LSV curves are shown in Figure S1.

6 ACS Paragon Plus Environment

Page 7 of 22 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

ACS Catalysis

Effect of pH The 0.5 M acetate buffer has a pH of 5.1, whereas the 0.1 M HClO4 electrolyte compared is significantly more acidic with a pH of approximately 1. To support the hypothesis that the enhanced ORR activity is due to the acetate buffer itself, we must exclude possible pH effects on the activity.

Figure 2 - ORR in acetate and phosphate buffer electrolyte both adjusted to pH 5.1. All measurements were performed at 25 °C with a scan rate of 10 mV∙s-1 and 900 rpm rotation. The corresponding LSV curves are shown in Figure S2.

For this purpose, a phosphate buffer was produced with a similar pH of 5.1 by adjusting the acid with NaOH. The results presented in Figure 2 clearly reveal that the ORR performance is not enhanced in the phosphate buffer. Phosphates are not known to inhibit Fe/N/C catalysts17, 25, 27 (see also Figure S2). Hence, we exclude that the improved ORR performance in the acetate buffer (as compared to 0.1 M HClO4) is due to the increased pH. We propose that the ORR activity enhancement is caused by the nature of the acetate buffer species (CH3COOH/CH3COO¯) and its interaction with the active site of the NPM catalysts. To support this hypothesis and to enable a comparison to computational modelling an experiment was conducted where buffers with different acid/base ratios were used as electrolyte to disclose wether the acid or the corresponding base is the dominant species improving the ORR. However, these experiments were not able to

7 ACS Paragon Plus Environment

ACS Catalysis 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 22

differentiate an activity improvement caused by one species over the other (see figure S3). Therefore a different approach was necessary and we decided to start with either the pure acid or the pure corresponding base and then gradually adjust the pH up or down. In the electrochemical experiments using the RDE, the pure acetic acid electrolyte unfortunately caused a very large internal resistance, which could not be compensated. Hence, the starting point was a solution of 0.5 M sodium acetate (CH3COONa) electrolyte. The results of this experiment are displayed in Figure 3. The effect of adding minute amounts of HClO4 to the electrolyte solution thereby creating an acetate buffer with trace amounts of the non-interacting ClO4¯ species is clearly seen. This suggests that the pure sodium acetate does not promote the ORR on the Fe/N/C catalyst (see also figure S4), whilst the addition of only 10 mM HClO4 (in solution) leads to a remarkable enhancement of the ORR.

Figure 3 - Tafel plots of Fe/N/C in 0.5 M sodium acetate with addition of minute amounts of HClO4. All measurements were performed at 25 °C with a scan rate of 10 mV∙s-1 and 900 rpm rotation. The corresponding LSV curves are shown in Figure S4.

This effect continues with further HClO4 addition but appears to be levelling off in the range of 2040 mM added HClO4. At an addition of 40 mM HClO4 the ORR activity is comparable to the one measured in the acetate buffer with a pH of 5.1. The ORR activity in the pure sodium acetate is considerably lower than in “blank” 0.1 M HClO4 despite the higher pH of 8.8. This could indicate that

8 ACS Paragon Plus Environment

Page 9 of 22 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

ACS Catalysis

the COO¯ unit might not be the promoting species in the buffer solution, but the COOH unit instead. It should however be noted that the proton (H+) concentration is comparetively low in pure sodium acetate and that the low activity might also be influenced by proton starvation28-29. Due to the high solution resistance of pure acetic acid, we decided to conduct an experiment, where acetic acid is added to a solution of 0.1 M HClO4, which in this case acts as a supporting electrolyte. The Tafel plot in Figure 4 demonstrates that the addition of minute amounts of acetic acid to the supporting electrolyte is insignificant regarding the ORR. In this case, no buffer is created during the addition of acetic acid because the strong supporting electrolyte keeps the COOH unit in its protonated form. Hence, this experiment indicates that the pure acetic acid alone is not sufficient to promote the ORR, i.e. for the promotional effect observed in the acetate buffer the presence of both the acetic acid and the acetate is required.

Figure 4 - Tafel plots of Fe/N/C in 0.1 M HClO4 supporting electrolyte with addition of acetic acid. All measurements were performed at 25 °C with a scan rate of 10 mV∙s-1 and 900 rpm rotation.

Carboxylic group Because the effect of the acetic acid is quite remarkable, we found it intriguing to study additional carboxylate buffer electrolytes in an effort to unravel the possible effect of the carboxyl group. The

9 ACS Paragon Plus Environment

ACS Catalysis 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 10 of 22

two closest neighbours to acetic acid is formic acid (HCOOH) and propionic acid (CH3CH2COOH), which are both soluble in water. From Figure 5 it is obvious that the carboxylate buffers all contribute positively to the ORR activity. Nonetheless, the acetic acid performs significantly better than both the formic and propionic acid buffers. Considering that the carboxyl group (protonated or not) is here assumed to be the promoter of the ORR, the overall increase of the reaction in carboxylate buffers is expected. It is however not straightforward to explain why the acetate buffer improves the ORR performance more than the other buffers. Again, the pH values do not provide any obvious trend (3.9, 5.1 and 5.1 for formate, acetate and propionate buffer, respectively) and the pKa values of the carboxylic acids are also quite similar: 3.75 (HCOOH), 4.75 (CH3COOH) and 4.88 (CH3CH2COOH).

Figure 5 - Effect of three carboxylate buffer electrolytes (0.1 M HClO4 for comparison). 25 °C, rotation of 900 rpm, scan rate of 10 mV s-1.

Electrolyte effect modelling To further elucidate the factors responsible for the observed ORR enhancement in carboxylic acid buffer electrolyte, we modelled the buffer interaction with the Fe/N/C catalyst by means of density functional theory (DFT), aiming for a thermodynamic explanation of the electrolyte effect. For the

10 ACS Paragon Plus Environment

Page 11 of 22 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

ACS Catalysis

computational study on Fe/N/C catalyst, we employed our previously reported porphyrin-like motif (M/N/C)30. We allow both sides of the catalytic site to be available for adsorption and reactivity. Studies on Fe/N/C support the hypothesis of the metal to be grafted within defects and cracks of the surface, presumably allowing access from both sides10, 31. Further, it has been shown that FexNy defects are most stabilized at edge positions in graphitic matrices, suggesting facile accessibility for double adsorption32-33. More information on the catalyst structure (figure S5) and the computational details can be found in the SI. The reaction in HClO4 electrolyte is modelled neglecting any interaction between the electrolyte and the catalytic site, as HClO4 is considered a non-interacting electrolyte, i.e. does not inhibit nor promote the reaction activity34. For the carboxylate buffer electrolytes (formic, acetic and propionic acid), we investigated any possible buffer molecules interaction with the catalytic surface which could be thermodynamically viable and increase the calculated activity, by influencing the energetics of the ORR catalytic steps. Thus, we screened a number of carboxylic acid-catalyst surface configurations, as reported in table S1. The preferred interactions occur by adsorption at the metal center, via oxidative deprotonation of the carboxylic acid which coordinates through the carboxylate moiety at the Fe site. We then modelled the reaction thermodynamics through the ORR associative pathway35, including the identified carboxylic acid interactions with the catalyst, and the catalytic site accessibility from both sides.

11 ACS Paragon Plus Environment

ACS Catalysis 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 12 of 22

Figure 6 - reaction coordinate for the proposed associative pathway in: a-b) non-interacting electrolyte, c-e) acetate buffer, f-h) formate buffer and i-k) propionate buffer. The calculated overpotential is found applying the computational hydrogen electrode reference system35. Full notation for the reaction intermediates is given only in a), while in the other panels the first, last and one illustrative steps are reported.

Figure 6 shows the ORR energetics in the four electrolytes tested in experiments. The reaction

coordinates in Figure 6 (x axis) refer to the charge transfer steps occurring during O2 reduction. Figure

12 ACS Paragon Plus Environment

Page 13 of 22 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

ACS Catalysis

6a shows the 4e- pathway, from 4.92eV (energy level of O2 referenced to water and hydrogen) to

0.0eV (water), yielding a computational overpotential of 0.83V (1.23V - potential of rate limiting step, Ulimit). Figure 6b reports the energetics of the thermodynamically most favored catalytic cycle in a non-interacting electrolyte. The reaction proceeds with *OH stably anchored on one side, while the ORR intermediates follows each other on the other side. The pathway still involves the exchange of four (H+/e-), though for ease of comparison we report the reaction steps shifted along the x axis due to the extra possible reduction step for the *OH back adsorbed. The reaction overpotential decreases to 0.65V, approaching the experimentally measured value. This is an electronic effect, for back adsorbed *OH weakens the binding energy of all ORR intermediates. As all reaction steps shift upwards in energy, the ΔG for the rate limiting step (OH desorption) becomes higher, yet the catalyst limitation remains the too strong affinity with the intermediates. Weaker intermediate bindings through electronic effect can also be achieved in presence of carboxylic acid, which adsorbs on the metal back side and results in a similar enhancement of the overpotential. However, the buffer molecules influence further the catalytic cycle, inducing an additional energy variation on the catalyst resting states (initial and final). In equilibrium with the electrolyte, carboxylic acid reversibly adsorbs on the Fe site, yielding a different catalyst initial configuration shifted downwards in energy. The extent of such stabilization corresponds to the binding energy of the carboxylic acid at the onset potential. For the ORR to begin, the carboxylic acid is displaced by O2 adsorption as *OOH. In the last catalytic step, the carboxylic acid re-adsorb on Fe in a concerted step with *OH desorption. The final catalyst configuration is stabilized to the same extent as the initial state.

13 ACS Paragon Plus Environment

ACS Catalysis 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 14 of 22

Figure 6c, f, i show the energetics if only one Fe side is available for the adsorption of acetic, formic

and propionic acid, respectively; in Figure 6d, g, j both side of Fe are occupied by the acid at the beginning of the catalytic cycle, while in Figure 6e, h, k the initial configurations have *OH stably adsorbed on one side and the respective carboxylic acid bound to the other side. In all cases, the carboxylic acid is displaced by the ORR intermediates for the reduction to occur, and it re adsorbs on Fe at the end of the cycle. Each adsorption of carboxylic acid implies the exchange of one (H+/e), so the catalytic cycle initial and final states are shifted along the x axis in figure 1; nevertheless, all the reactions consist of a total of four (H+/e-) exchanged. In the presence of carboxylic acid, the first catalytic step consists of a two e- reduction, one for carboxylate desorption as carboxylic and one for O2 adsorption as *OOH; in the last step, carboxylic acid oxidative adsorption replaces *OH in a potential independent step. Assuming this mechanism with acetic acid, the reaction steps are all downhill in energy up to a potential Ulimit of 1.01 V vs RHE (Figure 6d), significantly improving the reaction overpotential as compared to the best performance in non-interacting electrolyte (Figure 6b, Ulimit = 0.58 V vs. RHE). Even omitting both side accessibility (Figure 6c), the acetic acid interaction still improves the achievable potential (Ulim = 0.91 V vs. RHE). The same mechanism can be applied in presence of formic acid. However, formic acid (Figure 6c, d, e) has a stronger interaction with the Fe center than acetic acid, which according to our model – and in agreement with the experimental results – leads to a reduced enhancement, with a calculated onset potential of 0.76 V vs RHE in case of both side accessibility, or 0.67 V vs. RHE if neglected (Figure 6f). The presented variation of the associative mechanism, which we dubbed “substitution ORR”, is thermodynamically viable as long as the carboxylic acid binding energy is lower than *OH formation energy, to restore the initial catalyst configuration. For propionic acid, we predicted a weak adsorption strength: assuming the same reaction mechanism, the catalytic 14 ACS Paragon Plus Environment

Page 15 of 22 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

ACS Catalysis

cycle cannot be restored as the *OH formation energy is slightly lower than the adsorption energy of propionic acid. All the buffer adsorption energies are adjusted with a solvation correction, the same for all carboxylic acids, as their solubility in water is rather similar. An explicit solvent modelling would require significant assumptions and approximation, and is beyond the scope of this work. Herein instead, we report data incorporating the minimum solvation correction needed for the substitution ORR to be thermodynamically favored. We found that the model is validated with a solvent stabilization for the adsorbed carboxylic of 0.15eV for acetic and formic acid, while a higher solvation contribution (~0.4eV) would be required for propionic acid. Hence, by applying a conservative solvation correction - *OOH and *OH are stabilized in water by 0.3eV - the substitution ORR mechanism explains the overpotential reduction in formic and acetic acid. Figure 7 reports a graphical scheme for the mechanism.

Figure 7 - Proposed mechanism of substitution ORR with interacting electrolyte, here acetic acid omitting both side accessibility. The same mechanism is valid also with an adsorbate stably anchored on the catalyst back side.

15 ACS Paragon Plus Environment

ACS Catalysis 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 16 of 22

Experimental evidences in Figure 3 and Figure 4 exclude that the buffer enhancing effect is due exclusively to one of the two conjugated species. These observations are also rationalized in our computational model. In fact, acetic acid switches between protonated and deprotonated form to close the catalytic cycle. If the electrolyte is acidified so that it contains only protonated acetic acid (CH3COOH), then there would be a force opposing its deprotonation. This would make the adsorption step less favorable, reducing the competing effect with the ORR intermediates. On the other side, in a fully deprotonated (CH3COO¯) electrolyte, no acetic acid is left for adsorption, since is all converted to acetate. The acetate adsorption would proceed through simple chemisorption, through a potential independent step that does not scale favorably with applied potential respect to the ORR intermediates. We highlight some important features of the herein presented model. First, the substitution ORR mechanism does not rely on the assumption of both side accessibility, as can be seen comparing the overpotentials in Figure 6a, c, f. Second, the energetics variation shifts the reaction rate limiting step from *OH desorption to O2 adsorption, making Fe/N/C a catalyst limited by too weak binding (right leg of the ORR activity volcano). More importantly, the stabilization of the catalyst resting states does not affect the ORR intermediates binding energy. Consequently, the increased activity does not imply a reduced selectivity for the 4e- in favor of the 2e- pathway, as it would be if the overpotential was reduced through electronic or geometric effects that destabilize the ORR intermediates energy relative to H2O2 reversible potential36. In fact, if *O energy level is sufficiently close or above the peroxide reversible potential, 2e- pathway selectivity is enabled. In substitution ORR, the *O energy level is unaltered, therefore we preserve Fe/N/C catalyst selectivity towards 4eORR while improving the overpotential.

16 ACS Paragon Plus Environment

Page 17 of 22 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

ACS Catalysis

We correlate computational and experimental findings in Figure 8. The experimental activity of the system is compared to a descriptor, the buffer adsorption energy in respect to *OH formation energy. The choice of this descriptor is justified by the importance of the competition *buffer - *OH, which determines the degree of catalytic enhancement.

Figure 8 - Experimentally measured current density at a potential of 0.8 V vs. RHE as function of buffer adsorption strength (expressed relative to *OH binding).

We note formic acid binds stronger than acetic acid and this mitigates the beneficial effect as the catalytic cycle is limited to a greater extent by too weak *OOH binding. Contrarily, propionic acid binds more weakly the metal center than acetic acid. Although experimentally we observe an overpotential improvement in propionic buffer electrolyte, our model does not directly explain the cause. Notably, its adsorption energy fits the descriptor. Hence, we suggest that propionic acid also undergoes interactions with the metal center, and that these interactions are too weak to produce the same enhancement as observed for the acetate buffer, but still the overpotential improves to some extent.

17 ACS Paragon Plus Environment

ACS Catalysis 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 18 of 22

5. Conclusions Our experimental results demonstrate that the reduction of oxygen on Fe/N/C is significantly improved in carboxylate buffer electrolytes as compared to a non-interacting electrolyte. This effect is especially remarkable in the case of an acetate buffer electrolyte, where a fourfold activity increase is observed. We find no evidence that the buffer electrolyte causes any permanent chemical change to the Fe/N/C catalyst, nor do we observe any clear indications of a pH effect. Furthermore, our RDE measurements suggest that both COOH and COO¯ species must be present to improve the catalytic performance. To rationalize the experimental observations we developed a mechanistic model based on thermodynamics, dubbed substitution ORR. We highlight a generalized mechanistic consideration: the reversible adsorption of a competing adsorbate stabilizes the initial and final state of the catalytic cycle, improving the catalytic performances of materials affected by strong binding limitations. Furthermore, in substitution ORR the reaction intermediate energy levels are untouched, preserving the catalyst selectivity to 4e- pathway. We propose that an effective competition in the ORR on Fe/N/C can be obtained if the electrolyte has similar binding energy as *OH formation, thus the deviation from *OH formation energy is an appropriate choice of descriptor. We acknowledge the choice of solvation corrections influences this descriptor, hence we use it as study variable and we show that a conservative value (0.15 eV) is enough for the model to be consistent for two out of three buffers. The proposed model has limitations, as it can only indirectly explain the increased activity for propionic acid buffer, through the chosen reaction descriptor. The current study did not consider kinetic effects, such as proton shuttling or more complex solvent-adsorbate interactions37, which could play a significant role in determining the reaction energetics. We neglected electrolyte proximity effects, such as proton

18 ACS Paragon Plus Environment

Page 19 of 22 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

ACS Catalysis

transfer to the intermediates through proton acceptor/donor sites (e.g. COOH/COO¯), since this was previously modelled by our group in the work of Busch et al.21, making the herein presented model complementary to our prior studies. Nevertheless, the fundamental insights of this work can bolster further research focused on electrolyte and/or additives screening, both with computational and experimental approaches. To employ a catalyst limited by too strong intermediates binding in a reaction system with competitive adsorption species could prove to be a versatile way to improve the onset potential for ORR and beyond.

6. Associated content Supporting Information containing additional information on experimental and computational procedures and additional figures is available online.

7. Author information † These authors contributed equally and are listed alphabetically

* Corresponding authors: [email protected]; [email protected]

8. Acknowledgements The authors would like to thank the Innovations Fund Denmark for funding through the Initiative Towards Non-Precious Metal Polymer Fuel Cells (NonPrecious), project no. 4106-00012A as well as ProActivE project no. 5160-00003B and the Velux Foundation through The VILLUM Center for the Science of Sustainable Fuels and Chemicals (#9455).

19 ACS Paragon Plus Environment

ACS Catalysis 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 20 of 22

9. Conflicts of Interest The authors declare no conflicts of interest.

10. References 1. Jasinski, R., New Fuel Cell Cathode Catalyst. Nature 1964, 201 (492), 1212-&. 2. Gupta, S.; Tryk, D.; Bae, I.; Aldred, W.; Yeager, E., Heat-Treated Polyacrylonitrile-Based Catalysts for Oxygen Electroreduction. J Appl Electrochem 1989, 19 (1), 19-27. 3. Lefevre, M.; Proietti, E.; Jaouen, F.; Dodelet, J. P., Iron-Based Catalysts with Improved Oxygen Reduction Activity in Polymer Electrolyte Fuel Cells. Science 2009, 324 (5923), 71-74. 4. Sa, Y. J.; Seo, D. J.; Woo, J.; Lim, J. T.; Cheon, J. Y.; Yang, S. Y.; Lee, J. M.; Kang, D.; Shin, T. J.; Shin, H. S.; Jeong, H. Y.; Kim, C. S.; Kim, M. G.; Kim, T. Y.; Joo, S. H., A General Approach to Preferential Formation of Active Fe-N-x Sites in Fe-N/C Electrocatalysts for Efficient Oxygen Reduction Reaction. J Am Chem Soc 2016, 138 (45), 15046-15056. 5. Banham, D.; Kishimoto, T.; Zhou, Y. J.; Sato, T.; Bai, K.; Ozaki, J.; Imashiro, Y.; Ye, S. Y., Critical advancements in achieving high power and stable nonprecious metal catalyst-based MEAs for real-world proton exchange membrane fuel cell applications. Sci Adv 2018, 4 (3). 6. Zagal, J. H.; Bedioui, F.; Dodelet, J. P., N4-Macrocyclic Metal Complexes. Springer-Verlag New York: 2006. 7. Charreteur, F.; Jaouen, F.; Ruggeri, S.; Dodelet, J. P., Fe/N/C non-precious catalysts for PEM fuel cells: Influence of the structural parameters of pristine commercial carbon blacks on their activity for oxygen reduction. Electrochim Acta 2008, 53 (6), 2925-2938. 8. Zitolo, A.; Goellner, V.; Armel, V.; Sougrati, M. T.; Mineva, T.; Stievano, L.; Fonda, E.; Jaouen, F., Identification of catalytic sites for oxygen reduction in iron- and nitrogen-doped graphene materials. Nat Mater 2015, 14 (9), 937-+. 9. Jia, Q. Y.; Ramaswamy, N.; Tylus, U.; Strickland, K.; Li, J. K.; Serov, A.; Artyushkova, K.; Atanassov, P.; Anibal, J.; Gumeci, C.; Barton, S. C.; Sougrati, M. T.; Jaouen, F.; Halevi, B.; Mukerjee, S., Spectroscopic insights into the nature of active sites in iron-nitrogen-carbon electrocatalysts for oxygen reduction in acid. Nano Energy 2016, 29, 65-82. 10. Kramm, U. I.; Herranz, J.; Larouche, N.; Arruda, T. M.; Lefevre, M.; Jaouen, F.; Bogdanoff, P.; Fiechter, S.; Abs-Wurmbach, I.; Mukerjee, S.; Dodelet, J. P., Structure of the catalytic sites in Fe/N/C-catalysts for O-2reduction in PEM fuel cells. Phys Chem Chem Phys 2012, 14 (33), 11673-11688. 11. Guan, B. Y.; Yu, L.; Lou, X. W., A dual-metal-organic-framework derived electrocatalyst for oxygen reduction. Energ Environ Sci 2016, 9 (10), 3092-3096. 12. Guan, B. Y.; Lu, Y.; Wang, Y.; Wu, M. H.; Lou, X. W., Porous Iron-Cobalt Alloy/Nitrogen-Doped Carbon Cages Synthesized via Pyrolysis of Complex Metal-Organic Framework Hybrids for Oxygen Reduction. Adv Funct Mater 2018, 28 (10). 13. Zhang, J. T.; Xia, Z. H.; Dai, L. M., Carbon-based electrocatalysts for advanced energy conversion and storage. Sci Adv 2015, 1 (7). 14. Jin, H. H.; Zhou, H.; Li, W. Q.; Wang, Z. H.; Yang, J. L.; Xiong, Y. L.; He, D. P.; Chen, L.; Mu, S. C., In situ derived Fe/N/S-codoped carbon nanotubes from ZIF-8 crystals as efficient electrocatalysts for the oxygen reduction reaction and zinc-air batteries. J Mater Chem A 2018, 6 (41), 20093-20099. 15. Malko, D.; Kucernak, A.; Lopes, T., In situ electrochemical quantification of active sites in Fe-N/C nonprecious metal catalysts. Nature Communications 2016, 7. 16. Birry, L.; Zagal, J. H.; Dodelet, J. P., Does CO poison Fe-based catalysts for ORR? Electrochem Commun 2010, 12 (5), 628-631. 20 ACS Paragon Plus Environment

Page 21 of 22 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

ACS Catalysis

17. Gojkovic, S. L.; Gupta, S.; Savinell, R. F., Heat-treated iron(III) tetramethoxyphenyl porphyrin supported on high-area carbon as an electrocatalyst for oxygen reduction - I. Characterization of the electrocatalyst. J Electrochem Soc 1998, 145 (10), 3493-3499. 18. Herranz, J.; Jaouen, F.; Lefevre, M.; Kramm, U. I.; Proietti, E.; Dodelet, J. P.; Bogdanoff, P.; Fiechter, S.; Abs-Wurmbach, I.; Bertrand, P.; Arruda, T. M.; Mukerjee, S., Unveiling N-Protonation and Anion-Binding Effects on Fe/N/C Catalysts for O-2 Reduction in Proton-Exchange-Membrane Fuel Cells. J Phys Chem C 2011, 115 (32), 16087-16097. 19. Ramaswamy, N.; Tylus, U.; Jia, Q. Y.; Mukerjee, S., Activity Descriptor Identification for Oxygen Reduction on Nonprecious Electrocatalysts: Linking Surface Science to Coordination Chemistry. J Am Chem Soc 2013, 135 (41), 15443-15449. 20. Szakacs, C. E.; Lefevre, M.; Kramm, U. I.; Dodelet, J. P.; Vidal, F., A density functional theory study of catalytic sites for oxygen reduction in Fe/N/C catalysts used in H-2/O-2 fuel cells. Phys Chem Chem Phys 2014, 16 (27), 13654-13661. 21. Busch, M.; Halck, N. B.; Kramm, U. I.; Siahrostami, S.; Krtil, P.; Rossmeisl, J., Beyond the top of the volcano? - A unified approach to electrocatalytic oxygen reduction and oxygen evolution. Nano Energy 2016, 29, 126-135. 22. McGuire, R.; Dogutan, D. K.; Teets, T. S.; Suntivich, J.; Shao-Horn, Y.; Nocera, D. G., Oxygen reduction reactivity of cobalt(II) hangman porphyrins. Chem Sci 2010, 1 (3), 411-414. 23. Carver, C. T.; Matson, B. D.; Mayer, J. M., Electrocatalytic Oxygen Reduction by Iron Tetraarylporphyrins Bearing Pendant Proton Relays. J Am Chem Soc 2012, 134 (12), 5444-5447. 24. Baran, J. D.; Gronbeck, H.; Hellman, A., Analysis of Porphyrines as Catalysts for Electrochemical Reduction of O-2 and Oxidation of H2O. J Am Chem Soc 2014, 136 (4), 1320-1326. 25. Holst-Olesen, K.; Reda, M.; Hansen, H. A.; Vegge, T.; Arenz, M., Enhanced Oxygen Reduction Activity by Selective Anion Adsorption on Non-Precious-Metal Catalysts. ACS Catalysis 2018, 8 (8), 7104-7112. 26. Kramm, U. I.; Zana, A.; Vosch, T.; Fiechter, S.; Arenz, M.; Schmeisser, D., On the structural composition and stability of Fe-N-C catalysts prepared by an intermediate acid leaching. J Solid State Electr 2016, 20 (4), 969-981. 27. Strickland, K.; Pavlicek, R.; Miner, E.; Jia, Q. Y.; Zoller, I.; Ghoshal, S.; Liang, W. T.; Mukerjee, S., Anion Resistant Oxygen Reduction Electrocatalyst in Phosphoric Acid Fuel Cell. Acs Catalysis 2018, 8 (5), 3833-3843. 28. Wiberg, G. K. H.; Arenz, M., On the influence of hydronium and hydroxide ion diffusion on the hydrogen and oxygen evolution reactions in aqueous media. Electrochim Acta 2015, 159, 66-70. 29. Brocato, S.; Serov, A.; Atanassov, P., pH dependence of catalytic activity for ORR of the non-PGM catalyst derived from heat-treated Fe-phenanthroline. Electrochim Acta 2013, 87, 361-365. 30. Ju, W.; Bagger, A.; Hao, G.-P.; Varela, A. S.; Sinev, I.; Bon, V.; Roldan Cuenya, B.; Kaskel, S.; Rossmeisl, J.; Strasser, P., Understanding activity and selectivity of metal-nitrogen-doped carbon catalysts for electrochemical reduction of CO2. Nature Communications 2017, 8 (1), 944. 31. Deng, D.; Chen, X.; Yu, L.; Wu, X.; Liu, Q.; Liu, Y.; Yang, H.; Tian, H.; Hu, Y.; Du, P.; Si, R.; Wang, J.; Cui, X.; Li, H.; Xiao, J.; Xu, T.; Deng, J.; Yang, F.; Duchesne, P. N.; Zhang, P.; Zhou, J.; Sun, L.; Li, J.; Pan, X.; Bao, X., A single iron site confined in a graphene matrix for the catalytic oxidation of benzene at room temperature. Sci Adv 2015, 1 (11). 32. Holby, E. F.; Taylor, C. D., Control of graphene nanoribbon vacancies by Fe and N dopants: Implications for catalysis. Applied Physics Letters 2012, 101 (6), 064102. 33. Holby, E. F.; Wu, G.; Zelenay, P.; Taylor, C. D., Structure of Fe–Nx–C Defects in Oxygen Reduction Reaction Catalysts from First-Principles Modeling. The Journal of Physical Chemistry C 2014, 118 (26), 1438814393. 34. Rodes, A.; Pastor, E.; Iwasita, T., An Ftir Study on the Adsorption of Acetate at the Basal Planes of Platinum Single-Crystal Electrodes. J Electroanal Chem 1994, 376 (1-2), 109-118. 35. Nørskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J. R.; Bligaard, T.; Jónsson, H., Origin of the Overpotential for Oxygen Reduction at a Fuel-Cell Cathode. The Journal of Physical Chemistry B 2004, 108 (46), 17886-17892. 21 ACS Paragon Plus Environment

ACS Catalysis 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 22 of 22

36. Yang, S.; Verdaguer-Casadevall, A.; Arnarson, L.; Silvioli, L.; Čolić, V.; Frydendal, R.; Rossmeisl, J.; Chorkendorff, I.; Stephens, I. E. L., Toward the Decentralized Electrochemical Production of H2O2: A Focus on the Catalysis. ACS Catalysis 2018, 8 (5), 4064-4081. 37. Calle-Vallejo, F.; Krabbe, A.; García-Lastra, J. M., How covalence breaks adsorption-energy scaling relations and solvation restores them. Chem Sci 2017, 8 (1), 124-130.

22 ACS Paragon Plus Environment