Document [original]

Metal-Nitrogen-doped Carbon (M-N-C) Catalysts

for the Direct Electrochemical Reduction of CO2

to value-added Chemicals and Fuels -

Materials, Mechanisms and Cell Performance

vorgelegt von

M. Sc. Wen Ju

ORCID: 0000-0002-6485-1133

von der Fakultät II - Mathematik und Naturwissenschaften

der Technischen Universität Berlin

zur Erlangung des akademischen Grades

Doktor der Naturwissenschaften

Dr. rer. nat.

genehmigte Dissertation

Promotionsausschuss:

Vorsitzender: Prof. Dr. Michael Gradzielski

Gutachter: Prof. Dr. Peter Strasser

Gutachter: Prof. Dr. Yong Lei (TU Ilmenau)

Tag der wissenschaftlichen Aussprache: 09. April. 2019

Berlin 2019

天之道损有余而补不足

P a g e | I

Acknowledgement

I would like to express my first gratitude to my supervisor Prof. Dr. Peter Strasser, not only for the guidance

to study my work but also for the opportunity to learn my world.

I also thank Prof. Dr. Yong Lei and Prof. Dr. Michael Gradzielski, the committee of my defense, for the

insightful feedback and valuable advice.

Sincere thank goes to Dr. Ana Sofia Varela for mentoring and advising. These make my research strict but

interesting. Same appreciation goes to Dr. Alexander Bagger and Prof. Dr. Jan Rossmeisl, your theoretical

contribution leads me to search the in-depth answer. Special acknowledge moves toward Prof. Dr. Guang-

Ping Hao, partly for the kind collaboration partly for the warm friendship. I am also grateful to Dr. Ilya

Sinev and Prof. Dr. Beatriz Roldan Cuenya, for the fantastic ex-situ- and operando- techniques.

I appreciate the CO2 team partners - Xingli, Tim, Yulin, Trung, Cheonghee and Jorge. Together with you,

I am able to over the challenge. I also like to deliver my thanks to Fang, Julian, Mathias and Nate. During

my daily learning, you are always important and great. Much enjoy the time with Huan, Xingli, Xiaojia,

and Chengyue - the “Entertaining Gang”, you fill the weekdays with a lot of fun.

Further, I would also like to thank Mrs. Annette Wittebrock and Mr. Benjamin Paul for ordering chemicals

and lab wares, Mrs. Astrid Müller-Klauke for measuring ICP-OES and Mrs. Andrea Kluge for managing

the lab material especially the gas bottles.

I would like to thank everybody at the Strasser’s group for sharing success and happiness. Being one of

you is really my luckiness. I also thank the people aid me during my master study. With your help, every

difficulty turns easy.

However, real life could not be always breezy. Besides the days with sunshine, man should pass through

the days gloomy. I want to thank Weisen, Lebing, and Li, the friends went along with me. You did light the

season of my Rhine-Memory.

Importantly, I would like to thank my family - my parents, my wife and Nova my sweety. Always being

with me, making me stronger and better than I can be.

Lastly, I am grateful to the China Scholarship Council (CSC) for financing my Ph.D. study.

P a g e | II

P a g e  | III 
 
 
Table of Contents 
Acknowledgement  I 
Table of Contents  III 
Abstract    V 
Zusammenfassung  VI 
Chapter 1. Introduction and Motivation  1 
Chapter 2. Candidates for CO2RR: from metals to M-N-Cs  4 
2.1 Fundamental background  4 
2.2 The M-N-C catalysts  7 
2.3 CO2RR over M-N-Cs  8 
Chapter 3. Experiment section  12 
3.1 Synthesis of M-N-C catalyst  12 
3.2 Physiochemical characterization  12 
3.3 Electrochemical Methods  15 
3.4 Products quantification  18 
3.5 Density functional theory (DFT) calculation  21 
Chapter 4. Understanding activity and selectivity of metal-nitrogen-doped carbon (M-N-C) catalysts for 
electrochemical reduction of CO2  22 
4.1 Synthesis and characterization  24 
4.2 Electrochemical CO2RR over various M-N-C catalysts  28 
4.3 Correlating the theoretical prediction and experiments  31 
4.4 Discussion  35 
Chapter 5. CO evolution at industrial current densities on Ni-N-C derived gas diffusion electrode  36 
5.1 Catalysts synthesis and characterization  38 
5.2 Liquid-electrolyte H-Cell screening tests  41 
5.3 DFT prediction of CO2RR into CO over various Ni-Nx-C motifs  43 
5.4 CO2RR electrolysis using GDE combined MFC  45 
5.5 Discussion  46 
Chapter 6. Tuning  the  active  site  density  of  poly  anlinie  derived  Fe-N-C  catalyst  using  a  secondary 
Nitrogen precursor  47 
6.1 PANI based Fe-N-C catalysts synthesis using a 2nd Nitrogen precursor  49 
6.2 Physiochemical Characterization  50 
6.3 Correlating CO2RR performance and physical properties  54 
6.4 Operando X-ray absorption spectra under CO2RR condition  57 
6.5 Discussion  59 
Chapter 7. Unraveling the mechanistic insight of Electrochemical CO2 Reduction to Methane on the Fe-N-
C Catalyst  60 
7.1 Catalysts preparation and regular characterization  62 
7.2 Products spectrum of prolonged CO2RR and CORR on Fe-N-C catalyst  63 

P a g e  | IV 
 
 
7.3 Electrochemical reduction of a set of different COxHy molecules  64 
7.4 Proton- coupled and decoupled reaction steps  67 
7.5 Density Functional Theory Calculation  69 
7.6 Discussion  72 
Chapter 8. Summary and outlook  74 
Reference    77 
Appendix    85 
A1.   Supplementary Information to Chapter 4  85 
A2.   Supplementary Information to Chapter 5  98 
A3.   Supplementary Information to Chapter 6  105 
A4.   Supplementary Information to Chapter 7  117 
Table of figures and schemes  IX 
Table of tables  XVII 
List of Abbreviations  XVIII 
List of Chemicals  XIX 
List of Publications during Ph.D. study  XX 
 
 
   

P a g e | V

Abstract

In the past decades, the surplus of atmospheric CO2 concentration has drawn tremendous political and

scientific attention for its negative impacts, such as the greenhouse effect and ocean carbonation. To

mitigate such CO2 issues, a combination of various strategies is required. The electrochemical CO2

reduction reaction (CO2RR) is a promising alternative to convert CO2 into carbon-based chemicals and

fuels, and electricity generated from the renewable sources (solar and wind) could be employed to sustain

this transformation. At the current moment, the technological viability of this process is still contingent on

finding affordable and efficient catalysts.

In this thesis, a family of catalyst materials composed of abundant elements, in particular, non-precious

metals, nitrogen, and carbon, typically referred to as precious group metal (PGM)-free “M-N-C” catalysts,

were synthesized and mechanistically investigated – both experimentally and computationally – as catalyst

candidates for the CO2RR. MNC catalysts feature hemoglobin-like single-site metallated porphyrin

moieties with great impact on the catalytic reactivity and selectivity of the CO2RR. Among our studied M-

N-C catalysts, the Ni- functionalized one exhibits great efficiency for CO yielding at high potentials and

current densities. In particular, employment of Ni-N-C-based gas diffusion electrodes (GDE) combined

with micro flow cells, allowed high CO evolution that could exceed 80% faradaic efficiency at 250 mA cm-

2 current density, outperforming the industry commonly used Ag benchmark. By coupling our experimental

observation and density functional theory (DFT) simulation, the reaction path from CO2 to CO over this

sort single site catalyst could be deduced.

Unlike the Ni-N-C catalyst, the Fe-N-C shows selective CO production only at low potentials. Further, due

to relatively strong interaction with CO*, it opens the chance for hydrocarbons formation, yet showing little

selectivity. To understand the mechanism behind this kind of selectivity, we carried out a series of studies,

discussing catalytic tests, in-operando spectroscopic analysis, and computational modeling. Towards

material research, operando-XAFS measurements identified an unusual Fe-N3, possibly a FeI-N3 state,

which appears to enable CH4 evolution. Further mechanistic studies included the electrocatalytic reduction

of CO and CH2O as possible reactive intermediates for CH4 production. By combining the experimental

and computational results, we suggest a reaction network for CO2 reduction into a variety of carbon-based

products over the Fe-N-C catalyst. This contributes to the overall mechanistic understanding of CO2RR

over the M-N-C catalysts and delivers perspectives to evolve and design novel catalysts to produce

hydrocarbons of high value.

P a g e | VI

Zusammenfassung

In den letzten Jahrzehnten ist die atmosphärische CO2 Konzentration deutlich angestiegen. Wegen seiner

negativen Auswirkungen bzw. dem Treibhauseffekt und Versauerung der Meere, zog dieser CO2-Anstieg

enorme politische und wissenschaftliche Aufmerksamkeit auf sich. Um diese CO2-Problematik zu

vermindern, ist eine Kombination verschiedener Strategien vielversprechend. Die elektrochemische CO2-

Reduktionsreaktion (CO2RR) ist eine potentielle Alternative, um CO2 in Kohlenstoff-basierte Chemikalien

und Kraftstoffe umzuwandeln. Strom aus erneuerbaren Quellen wie Solar und Wind könnte diese

Transformation unterstützen. Zum gegenwärtigen Zeitpunkt ist die technologische Durchführbarkeit dieses

Verfahrens noch von verfügbaren und effizienten Katalysatoren abhängig.

In dieser Arbeit werden Materialien, die aus häufig vorkommenden Elementen – (Nicht-Edel-) Metall,

Stickstoff und Kohlenstoff - (M-N-C) aufgebaut sind, synthetisiert und mechanistisch für und während der

elektrokatalytischen CO2RR untersucht. Dabei beeinflusst das Hämoglobin-ähnliche „Single-Site“ Metall-

Stickstoff-Zentrum die katalytische Reaktivität stark. Unter unseren untersuchten M-N-C-Katalysatoren

zeigt der Nickel-funktionalisierte Katalysator große Effizienz für die CO-Ausbeute bei hohen Potentialen

und Stromdichten. Insbesondere die CO-Entwicklung bei einer Gas-Diffusion-elektroden- (GDE-) Micro-

Flow-Cell weist einen Faraday’sche Wirkungsgrad von über 80% bei einer hohen Stromdichte von 250 mA

cm-2 auf. Durch die Verknüpfung unserer experimentellen Beobachtungen und einer

Dichtefunktionaltheorie (DFT) -Simulation konnte der Reaktionsmechanismus von CO2 zu CO über diesen

Katalysator abgeleitet werden.

Im Gegensatz zu Ni-N-C funktioniert Fe-N-C als CO-selektiver Katalysator im niedrigen Potentialbereich.

Aufgrund der relativ starken Wechselwirkung mit CO* eröffnet sich außerdem die Möglichkeit der Bildung

von Kohlenwasserstoffen, die jedoch nur eine geringe Selektivität aufweist. Wir führen daher eine Reihe

von Studien durch, um tiefgreifende mechanistische Einblicke in die Reaktionen nach die CO-Bildung,

bzw. zu den Kohlenwasserstoffen zu erhalten. Unsere Operando-XAFS-Messungen zeigen einen

ungewöhnlichen Fe-N3-Zustand, wahrscheinlich FeI-N3, der für die Methan-Entwicklung selektiv sein

könnte. Bei Zwischenproduktstudien führen wir die Reduktion von CO und CH2O als mögliche

Zwischenprodukte für die CH4-Produktion ein. Durch die Kombination der experimentellen und

rechnerischen Ergebnisse haben wir für den Fe-N-C Katalysator ein Reaktionsnetzwerk der CO2-Reduktion

zu verschiedenen Kohlenstoffprodukten aufgebaut. Dies trägt zum mechanistischen Verständnis der

CO2RR für die M-N-C Katalysatoren bei und bietet Perspektiven für die Entwicklung neuartiger

Katalysatoren zur Herstellung von Kohlenwasserstoffen.

P a g e | VII

Die vorliegende Arbeit wurde unter der Leitung von Herrn Prof. Dr. Peter Strasser in der Zeit vom

01.04.2016 bis zum 09.04.2019 im Fachbereich Chemie am Institut für Technische Chemie der

Technischen Universität Berlin angefertigt. Die praktischen Arbeiten wurden unter der Leitung von Herrn

M.Sc. Wen Ju.

P a g e | VIII

P a g e | 1

Chapter 1. Introduction and Motivation

The fossil fuels, namely coal, oil, and natural gas, are playing significant roles on the stage of modern

human society. For more than half a century, the fossil energy contributes over 80% of the power to allow

the prosperity of the modern society, and, will remain as the dominant power source in the coming decades.

On the backside of the profit given by the fossil fuels, the CO2 emission, due to their combustion, causes

the drastic anthropogenic greenhouse gas issue, threatening the climate and ecosystem of our planet.1,2

To mitigate these carbon issues, while maintaining our current living standard, a great variety of techniques

are needed. By cutting to the source of CO2 emission, gradual transformation towards low-carbon and high-

efficiency energy generation, storage and utilization cycle is necessary. In order to replace conventional

fossil fuels, sources of renewable energies, as solar and wind are getting considerable attention, focusing

on the challenges of efficient conversion and storage. Thanks to the progress in scientific research and

production, usage of renewable energies has significantly increased in the past years and is likely going to

pose one major power supply in the upcoming decades.

Besides the energy-related strategies, other techniques to directly reduce the CO2 are also required, which

essentially rely on the technological combination of carbon dioxide capture, storage, and conversion. Apart

from the importance of the capture and storage techniques, this work focuses on an electrochemical

approach to transform the carbon dioxide into carbon-based fuels and manifold chemicals.3

Figure 1- 1 Schematic electrochemical CO2 reduction accompanied with industrial plants and renewable

electricity.

The investigation on electrochemical CO2 reduction (CO2RR) could be traced back to 80s last century. Hori

et. al. utilized typical metals as the electrocatalysts for CO2RR and found, this reaction allows to convert

waste CO2 into useful carbon-based chemicals.4 Technologically, the CO2RR could be accompanied with

industrial CO2 plants, water, and renewable electricity (see Figure 1-1), which could provide this eco-

P a g e | 2

beneficial transformation in a sustainable manner. However, knowledge on CO2 electrolysis is still at a very

preliminary stage, and its technologic viability meets multiple distinct limitations. One of the key-factors

relies upon affordable, efficient, and scalable catalysts to sustain this electrochemical process, which could

be potentially fulfilled by a sort of functionalized carbon materials.

In this dissertation, a novel single site coordinative metal and nitrogen co-doped carbon-based (M-N-C)

materials are synthesized and employed as the catalyst for the electrochemical CO2 reduction. One of the

main goals of this work is to determine the relation between the catalytic CO2RR performance (selectivity

/ activity) and the nature of the coordinative metal-nitrogen (M-Nx) site. This contribution could 1) help to

select the optimal M-Nx active motifs, as well as the operation conditions for CO2 electrolysis; 2) deliver

in-depth mechanistic understanding of CO2 reduction into CO and other carbon-based products; 3) provide

ideas to up-scale the electrolysis to meet the industrial level; 4) offer perspective to evolve and design novel

M-N-C catalysts for producing hydrocarbons. This work is structured as follows.

In Chapter 2, the theoretical background and state-of-knowledge of the electrochemical CO2 reduction are

stated. The single site M-N-C catalysts, as well as their function for CO2RR are discussed. In addition, basic

knowledge referred to the synthesis and characterization of the M-N-C catalysts is presented. This section

aims to put the results from this work into perspective to the broader scientific context.

In Chapter 3, the experimental procedures of this work, including material synthesis and important

physicochemical characterization are listed. Moreover, Online Gas-Chromatograph (Online-GC) test-

station for CO2RR products quantification, equipment as liquid-GC, HPLC for liquid products analysis are

shown. Further, the density functional theory (DFT, performed by Prof. Rossmeisl’s group at University

Copenhagen) calculation method is presented.

In Chapter 4, a fundamental work to determine the impact of the coordinative M-Nx site of the M-N-C

catalysts on CO2RR catalytic reactivity is shown. In this work, five M-N-C catalysts (M: Mn, Fe, Co, Ni,

Cu. Materials are synthesized by Dr. Guang-Ping Hao in Prof. Kaskel’s group at TU Dresden) were

prepared as the catalysts for CO2 reduction. Experiments, in a homemade H-cell, deliver the faradaic

activity / efficiency for CO evolution, whereas DFT calculation predicts the corresponding reaction steps.

Correlating the experimental and computational data, we contribute the first-of-its-kind mechanistic insight

into the rate- and selectivity- determining processes on the single-site M-Nx centers. We show that the

binding energies of intermediates to the M-Nx moieties provide excellent descriptors to describe, predict,

and understand the mechanistic details of the CO2RR activity and selectivity of this family of catalysts over

a wide potential range.

In Chapter 5, we utilize a Ni-N-C Gas Diffusion Electrode (GDE) combined Micro-Flow-Cell (MFC) to

technologically up-scale the CO2 electrolysis achieving industry relevant current densities. Unlike the

performance in regular H-type cell, the catalyst poses a faradaic efficiency exceeding 80% even at elevated

currents of 250 mA cm-2, which result in a partial current of above 200 mA cm-2 for CO yielding. This study

P a g e | 3

displays the potential of the earth abundant and affordable Ni-N-C materials to replace precious metal

catalyst in industrial scale electrolyzers for the electrochemical CO2RR to CO.

In Chapter 6, we perform a special synthesis strategy to improve the site density of the iron-based catalysts,

since the Fe-N-C catalyst show superior CO2RR performance in the low potential range. A secondary

chemical N-source is added in the poly-aniline (PANI) derived Fe-N-C, which make the catalyst not only

catalytically more active, but allow a variation of the physico-chemical properties of the resulting Fe-N-C

materials. Particular emphasis is given to the chemical state and local structure of the coordinative Fe-Nx

center during the CO2 electrolysis. In collaboration with Prof. Roldan Cuenya’s group, Operando X-ray

absorption spectroscopy is applied to identify the active Fe-Nx motif under the CO2 electrolysis condition

and an unusual Fe(I)-N3 state is found, which is suggested to be responsible for the formation of CH4.

In Chapter 7, we further unravel the mechanistic details of CH4 evolution from CO2RR over the Fe-N-C

catalyst, since it provides the possibility of converting CO2 into hydrocarbons. Analogous reduction

reactions involving distinct reactants, namely, CO2, CO, CH2O, CH3OH and formate, are systematically

conducted to investigate the reactivity of these molecules. By linking our experimental results and DFT

prediction under consideration of our previous results, we establish a full and deep reaction network from

CO2 into various carbon-based outputs over the Fe-N-C catalyst with clarifying the mechanistic role of the

proton in these key reactions.

In Chapter 8, the results of this work are summarized, and general conclusions are drawn. Additionally,

perspectives for further investigations on the basis of the findings of this work and their integration into the

state-of-knowledge are given.

P a g e | 4

Chapter 2. Candidates for CO2RR: from metals to M-N-Cs

In this chapter, the fundamental principles of electrochemical CO2 reduction are described. First, the

reaction mechanism of catalytic CO2RR is stated. For this, descriptors for the faradaic selectivity are

discussed. Thereafter, a sort of novel single-site coordinative M-N-C catalyst as the candidate of CO2

reduction is introduced. Furthermore, the state-of-art knowledge of M-N-C catalysts preparation,

characterization and their application as CO2RR electrocatalysts are stated.

2.1 Fundamental background

The CO2 electrolysis, as cathodic reactions, are normally coupled with the electrochemical water splitting

(eq. 2-1). Oxygen Evolution Reaction (OER, eq. 2-2) occurs at the anodic side, whereas the Hydrogen

Evolution Reaction (HER, eq. 2-3) emerges as a competitive cathodic reaction.

2H2O ⇆ O2+ 2H2 E0= +1.23 VRHE eq. 2‐ 1

2H2O ⇆ O2+ 4e−+ 4H+ E0= +1.23 VRHE eq. 2‐ 2

2H++ 2e−⇆ H2 E0= 0.0 VRHE eq. 2‐ 3

To date, a vast range of catalysts, such as metals, metal oxides, metal (oxide) nanoparticles, and molecular

catalysts, have been investigated for the electrochemical CO2 reduction. The first ground-laying work was

achieved by Hori et. al.. They pioneered this field by testing a number of metals as the electrocatalysts for

CO2 reduction and found the products distribution of CO2RR is highly dependent on the nature of the

metals.4 In detail, as well being agreed in other studies, metallic Pt, Fe, and Ni are primarily active for HER

(thus should be avoided as the candidate for CO2RR), while noble metals as Ag and Au could preferentially

yield CO as the major product.4-11 Ultimately, the metallic Cu was particularly interesting due to its unique

capability of producing a broader range of carbon-based products (eq. 2.4 to 2.9), including formate, CO,

hydrocarbons, and alcohols.4,12 Therefore, numerous works were implemented on Cu-based catalysts to

enhance the yield for value-added products as hydrocarbons and oxygenates. 4,12-25

CO2+ 2(H++ e−) ⇆ HCOOH E0= −0.17 VRHE eq. 2‐ 4

CO2+ 2(H++ e−) ⇆ CO + 2H2O E0= −0.11 VRHE eq. 2‐ 5

CO2+ 8(H++ e−) ⇆ CH4 + 2H2O E0= +0.17 VRHE eq. 2‐ 6

2CO2+12(H++ e−) ⇆ C2H4+ 4H2O E0= +0.08 VRHE eq. 2‐ 7

2CO2+12(H++ e−) ⇆ C2H5OH + 3H2O E0= +0.08 VRHE eq. 2‐ 8

3CO2+18(H++ e−) ⇆ C3H7OH + 6H2O E0= +0.09 VRHE eq. 2‐ 9

P a g e | 5

To understand the specific catalytic selectivity of Cu towards hydrocarbons, A.A. Peterson and co-workers

utilizes density functional theory (DFT) to simulate the reaction path from CO2 to CH4. Their contribution

pinpoints that the carboxyl (COOH*) is the first key intermediate of CO2 reduction into CO (and formate),

while the further adsorption and protonation of CO are relevant for hydrocarbons formation, which is

allowed on the metallic copper surface.13 This implies, free energy of COOH* and CO* are crucial factors

to determine the reactivities of the CO2RR candidates.

Figure 2- 1 a) The CO2RR products spectrum classified by ΔEH* descriptor. Metals prefer HER (marked in

red) due to strong H* binding (having HUPD), while metals favor CO2RR owing to weak H* binding (not

having HUPD). b) The binding strength diagram of CO* and H*. Hydrocarbons formation occurs via

protonation of intermediate CO (CO*), which require moderated binding energy to CO* and H* as Cu.

Products spectrum is obtained from reference 4.

To further understand other metallic surfaces, establishing a theoretical “selectivity criteria”, the adsorption

energy of H*, the key intermediate for the competitive HER is taken into consideration. Recently, a

computational study was done by Bagger et. al. utilizing the binding strength towards H* as the descriptor

to traditionally classify the typical metals into four distinct groups according to their main products, namely,

hydrogen, CO, hydrocarbons, and formate.15 Here, we simply refer them as two general classes, either

selective for HER, or for CO2RR. Figure 2-1a shows the products map as a function of hydrogen binding

(ΔEH*). Strong binding towards H* leads to hydrogen underpotential deposition (HUPD), resulting in the

HER favoring catalysts. On the contrary, CO2RR is preferred on the metals with weak H* interaction.

Figure 2-1b models the free energy map for hydrocarbons. Based on Peterson’s prediction,13 three critical

factors are claimed, 1) proper binding to H* to avoid HUPD, 2) proper binding strength enabling the CO*

adsorption while avoiding CO* poisoning, 3) proper binding strength enabling CO* protonation. This leads

to a very narrow room to select the hydrocarbons favoring catalyst (the region is marked as “Beyond CO”

in Figure 2-1b.15

P a g e | 6

Figure 2- 2 a) Illustration of proposed active M-N4-C motif. b) Relation between free energy of H* (ΔG*H)

and COOH* (ΔG*COOH) (gray circles) as well as H* (ΔG*H) and OCHO* (ΔG*OCHO) (red diamonds) on

various metal-Porphyrins. The diagonal line (black dashed) separates selectivity towards HER and CO2RR

Free energy diagram of first binding via COOH* (red) or H* (blue) and then binding a second H* at a

nearby site for the relevant c) Cu metal surface and d) Fe-porphyrin like motif. Figures are reproduced

The aforementioned predictions indicate that ΔEH*, ΔECOOH*, and ΔECO* pose as relevant selectivity criteria

of the CO2 electrolysis, which is also valid to investigate other types of CO2RR candidates. Recently, a sort

of metal functionalized porphyrin-like M-Nx motifs is predicted as the responsible active site for CO2

reduction. The proposed active M-N4-C motif is illustrated in Figure 2-2a. Taking these descriptors (binding

energies to H*, COOH*, OCHO*, see Figure 2-2b) into consideration, both CO2RR and HER are allowed

on this type catalysts,26 and the further CO reduction towards various oxygenates and hydrocarbons are

predicted in Tripokovic’s study.27 Additionally, in comparison to typical metallic catalysts, the coordinative

M-Nx moieties hold the structural advantage to prevent the competitive hydrogen evolution, further

benefiting the faradaic selectivity towards CO2RR (see Figure 2-2 c and d).15 Above is the preliminary

predictions on M-N-C catalysts. We thereby move forward to discuss the practical strategies to investigate

this type catalyst.

P a g e | 7

2.2 The M-N-C catalysts

The metalorganic complex is an optional candidate for CO2 reduction. Since the 1980s, various complexes,

like the metal- macrocyclics,28 cyclams,29 bipyridines,30,31 as well as various porphyrins32-35 have been

reported for their catalytic performance. CO and formate were found as the major product over these

molecular catalysts. However, in practice, these molecular complexes are commonly utilized as homo-

catalysts or immobilized on the electrode surface, thus suffer the drawbacks such as mass / electron transfer

limitation. Their technologic applicability is therefore restricted by low working current densities.36

Figure 2- 3 Synthesis strategy of metal nitrogen-doped carbon (M-N-C) catalysts.

Inspired by this metal-contained complexes, a sort metal and nitrogen functionalized carbon (M-N-C) was

designed and performs as a promising alternative to the Pt catalyst for oxygen reduction reaction (ORR).37-

42 Recent studies show their potential for CO2RR utilization.43,44 Compare to the original metal-complex,

these functionalized carbon catalysts own excellent conductivity, large surface, mechanical stability, low-

cost and especially good scalability, thus hold the great promise for industrial level applications.36 Regular

synthesis approach of these catalysts generally is shown in Figure 2-3. Usually, by mixing and carbonizing

the carbon, nitrogen and metal precursors, man could formulate and embed nitrogen functionalities

(including pyrrolic, pyridinic, metal-coordinated, quaternary, graphitic and oxidized ones, see Figure 2-3)

in the carbon structure. Other procedures are also of importance during specific synthesis approaches. For

instance, using a soft template allows to control the structure and surface area of the final product,

accordingly, the leaching procedure is necessary to remove the remaining template material.

P a g e | 8

In previous studies, the nitrogen coordinated metal ions (M-Nx) are demonstrated as the active sites for

CO2RR,26,27,43,44 although the metal free N-C catalysts were also reported for their CO2RR performance.45,46

In practice, following the normal M-N-C synthesis procedure, various functionalities could exist in the

obtained material, such as the inorganic particles and a number of nitrogen moieties (see Figure 2-3). For

the purpose to identify the coordinative M-Nx motifs that we expect, a combination of characteristic

methodologies is required. Based on the state-of-knowledge, the X-ray absorption spectra (XAS, including

XANES and XAFS, usually performed in synchrotron facilities for intense and tunable beam source) is

able to accurately determine the electronic structure of the matter. To characterize the M-N-C catalysts, the

coordination number and bond distance could be analyzed. Reported in previous studies, the XAFS profile

of the M-N-C (FeN4,42,47 CoN448 and NiN448) catalysts could nicely match the fingerprint of the porphyrin

type references, and coordination number of metal center is roughly 4 with bond-length around 1.93 Å.

This confirms that the M-Nx motif is the dominant state of the metal composition. Beside the XAS

technique, the X-ray photoelectron spectra (XPS) is more widely used to analyze the state of the sample

surface. Recently, Atanassov’s group focused on the M-N-C catalysts to address the assignments - electron

binding energy (BE) - of various distinct nitrogen functionalities. In particular, the nitrogen 1s electron

holding binding energy around 399.7 eV is reported as metal-coordinated ones, which is the evidence of

M-Nx motifs.49,50

2.3 CO2RR over M-N-Cs

Previously predicted using the DFT calculations, the metal-porphyrins and porphyrin-like metal-nitrogen

motifs in M-N-C catalysts play as the active centers for CO2 reduction, controlling the intrinsic activity and

selectivity.26,27 However, these simulations remain preliminary, and the free energy of the key intermediates

(H* vs. COOH*) strongly scale with each other. Hence, corresponding experimental demonstrations are of

high importance, and linking those with the computational trends could deliver in-depth perspectives to

achieve practically effective catalysts for CO2 electrolysis.

Selectivity criteria (HER versus CO2RR)

On metallic CO2RR candidates, the binding strength towards H* plays as a key descriptor to determine the

catalytic CO2RR selectivity, since HER occurs as the predominate catalytic process due to HUPD.15 Towards

M-N-C catalysts, such rule also makes sense.

Ju et al. carried out the CO2 electrolysis on a series pyrolyzed M-N-C catalysts in presence of the first row

transition metals (Mn, Fe, Co, Ni, Cu), bringing insight to this issue by combining their experimental and

theoretical investigations (which is in detail described in chapter 4 of this thesis).51 As found by them, the

coordinative Ni-Nx sites, who possess weak interaction with protons, could deny the unwanted HER, further

benefiting the selectivity to CO evolution. Thus, among these series catalysts, the Ni-N-C shows the best

maximum selectivity for CO2RR into CO. On the contrary, the cobalt sites with strong H* adsorption,

accordingly incline to better HER performance. Others stand in between, and selectivity to CO2RR order is

P a g e | 9

given as Ni > Fe > Mn > Co = Cu. In a more recent study, Jiang et al. compared the CO2RR performance

of various M-NG (metal atom coordinated in N doped graphene vacancies) using gas diffusion layer, and

the same selectivity trends (Ni > Fe > Mn > Co = Cu) is shown.52 Hu et al. reported the use of pyrolyzed

Fe-, Co- and Ni- functionalized nitrogen-doped porous carbon for CO2 electrolysis. In particular, selectivity

for CO2-to-CO conversion in water was found to be Ni > Fe ≫ Co.52 Hence, the interaction to H* could be

used as a descriptor to control CO2RR selectivity, and the superior CO selectivity on Ni-based ones is

attributed to prohibited H* binding.

Moreover, Zn based M-N-Cs are also studied, showing impressive CO selectivity. The remarkable CO

selectivity of Zn-N4 motifs might be due to the weak H* interaction since its “ECOOH* versus EH*” free

energy diagram stands close to that of Ni-N4 motif with respect to Wannakao’s DFT calculation 26 (see

Figure 2-2b). Experimentally performed by Yang et al,53 single site Zn-Nx-C catalyst was prepared by

pyrolyzing the mixture of urea and Zn-acetate, and faradaic efficiency to CO could reach 95 % at -0.43

VRHE. Similar performance (FECO = 91 % at 0.39 V overpotential) was observed on the Zn-N-C catalyst in

another work using a different synthesis procedure.54

Other transition metal centers have also been tested to expand the knowledge on M-N-C catalysts. Roy et

al. synthesized a set of M-N-C catalysts for the CO2RR, and in this contribution, the transition metals Cu,

Mo, Ce, and Pr were studied.55 Interestingly, the Cu-, Mo- containing catalysts were highly selective to the

competing HER process, whereas the Ce-, Pr- based catalysts perform slightly better towards CO

production, but provide no advances than their metal free reference one. Furthermore, in Birdja’s work,

porphyrin-comprised In, Sn and Rh were tested for their CO2RR performance and formate was found to be

the major product.34

The Cu-Nx motifs should be specially discussed, since the free energy simulation of copper-nitrogen site

could fulfill the binding strength requirements (comparable to Ni-Nx), however, considerable performance

was not observed.51,55,56 This could be presumably attributed to poor chemical stability of the Cu-Nx sites

under cathodic conditions, as the d-orbit of Cu0 / Cu1+ / Cu2+ is filled by 9 (one SOMO) or 10 (fully occupied)

electrons, therefore the ionic copper could be reduced into nano-particles and therefore contribute low

reactivity in their studied working potential regime.

Selectivity criteria (towards hydrocarbons)

Methane could also be formed over M-N-C catalysts during the CO2 electrolysis, however, showing

extremely low selectivity. In normal reaction conditions (ambient pressure, neutral solution) only Fe- and

Mn- based NC catalysts enable the CH4 production due to strong CO* binding energy13,27,44,51. Moreover, it

has also been reported the presence of Fe-Nx centers facilitates CH4 production even in the presence of

other heteroatoms.57 On the contrary, CH4 evolution is hard to happen on Co-N433 and Ni-N458 sites, since

the CO* is more likely to be released due to its weak binding. This is a clear indication that the CO* binding

is crucial for methane formation over M-N-C catalyst. Despite Fe-N-C catalysts’ ability to produce

P a g e | 10

hydrocarbons, it should be noticed that the selectivity is dramatically lower than that observed on copper

catalysts, which will be discussed in Chapter 7 on the fine structure of these single-site motifs.

Activity descriptor

According to the Arrhenius equation, k = A ∗ exp (− Ea RT

⁄), the activation energy Ea is the key factor

determining the reaction rate at a certain pressure and temperature (applied potential as well). Given that

the CO2 reduction to CO is via the path: CO2(g)→COOH*→CO*→CO(g), thus the free energy of the key

transition states, such as COOH* and CO*, could in principle be seen as the descriptor to determine the CO

formation activity.51,56,59

Figure 2- 4 a) Relation of CO formation current densities at -0.9 VRHE and COOH* binding at 0 VRHE.

Figure is reproduced from reference56 with Copyright permission 2018, Royal Society of Chemistry. b)

Fitted *CO desorption (*CO→CO, red line) and *COOH formation (CO2→*COOH, black line) trends as

a function of CO* binding energy over all five Metal-Pc electrodes. Figure is adapted from reference 59,

In a recent work carried by Su et. al., two types ligand bases, CTF (covalent-triazine-framework) and TPP

(tetraphenylporphyrin), were employed to hold the ionic metal (Co, Ni, Cu) sites for the further CO2RR

utilization.56 Apart from the ligand effects given by the CTF and TPP, the partial CO formation current

densities at -0.9 VRHE clearly elevated when the COOH* binding energy turned stronger, which is an

indication that adsorbing the COOH* on the active sites is of high importance to determine the reactivity

of CO yielding (see Figure 2-4a). Further in the work done by Zhang et al., binding energy of both CO*

and COOH* were taken into account to fit the reactivity of CO2RR (into CO) over metal phthalocyanines

(Pc) complexes (Mn, Fe, Co, Ni, Cu).59 It is suggested that for Fe-, Mn-, and Co- Pcs, the *CO desorption

is the most endergonic step that determines the overall reaction rate for a strong CO* interaction. On the

contrary, on Ni- and Cu- Pcs, the *COOH formation becomes the most endergonic and thus limits the

overall reaction rate (shown in Figure 2-4b).

P a g e | 11

The binding energy to COOH* and CO* were evaluated in Ju’s work based on a wide range of M-N-Cs

(M: Mn, Fe, Co, Ni, Cu, Chapter 4 of this dissertation). In their fitting, three distinct reaction dynamics

were found in different potential ranges.51 In low potential range, the onset of CO evolution relies on the

free energy of COOH* transition state. Further in the middle potential regime, the reaction dynamic is

controlled by the CO* binding. Finally, in the high potential region, the CO* interaction causes a desorption

issue. For this reason, the CO production on Fe-N-C meets a drastic downhill. On the contrary, the Ni-Nx

motifs do not suffer the CO* desorption issue, simultaneously denies the competitive HER due to weak H*

binding, thus holds the promise as the candidate to selectively yield CO at high overpotentials and current

densities.48,51,52,58,60,61

Briefly summarizing this section, the binding energies of the active M-Nx motifs towards H*, COOH* and

CO* control the faradaic reactivity for CO2 electrolysis. Weaker H* interaction is beneficial for the CO2RR

selectivity, and relatively strong COOH* binding benefits the activity. In particular, the CO* binding plays

a special role during the electrochemical CO2RR process. It, on one hand provides, the positive effect to

boost the dissociation of hydroxide group from COOH* into CO* and further release the CO, but on the

other hand turns to a limiting factor, causing the mass transfer issue. The DFT computation also shows that

the binding strength of CO* could be tuned by electrode potential. Therefore, not only the catalytic active

sites, but the applied potential window should also be taken into account to avoid the CO* desorption

barrier.51,56,58

P a g e | 12

Chapter 3. Experiment section

3.1 Synthesis of M-N-C catalyst

The Non-PGM coordinative metal nitrogen functionalized carbon (M-N-C) has been widely utilized for

electrochemical CO2RR in recent years. The general preparation strategy involves 1) mixing and the

reaction of the precursors, 2) carbonization (heat treatment) and 3) leaching steps. In this dissertation, two

types of synthesis approaches are used to prepare the M-N-C catalysts. CO2RR study in Chapter 4 is done

on copper-MOF derived M-N-Cs, while the works in Chapter 5 to 7 are based on PANI-family. In this

section, the respective synthesis procedures of each series catalysts are described in Table 3-1.

Table 3- 1 Amount of chemical substances used in the synthesis protocols to prepare the M-N-C catalysts

and the corresponding synthesis protocols for the respective studies described, investigated and discussed

later on in Chapters 4-7.

3.2 Physiochemical characterization

Powder X-ray diffraction (PXRD) was employed to analyze the crystallinity of the samples based on the

reflection of the ordered lattice. In our contribution, the XRD patterns were recorded on a Bruker D8

Advance instrument with Cu Kα radiation (λ = 1.54056 Å) in the 2θ range of 10° ~ 90°. Normally, the non-

PGM M-N-C catalysts involve no significant ordered crystal phases, thus rarely showing broader diffraction

peaks at 25° and 43° for carbon plane (002) and (100), respectively.

N2 physisorption (BET) measurements using Brunauer–Emmett–Teller (BET) theory are conducted on

Autosorb-1 (Quantachome Instruments) to determine the specific surface area of the catalysts. The Pore

size distributions were calculated from a nonlocal density functional theorem (NLDFT) pore model based

on carbon pores with both slit and cylindrical geometries.

Precursors

Chapter 4

Chapter 5, 6 and 7

Carbon

Co-Bpys derived Cu-Bpys

Ketjen 600

Nitrogen

PANI (2nd Nitrogen is tuned in Chapter 6)

Metal

Mn, Fe, Co, Ni, Cu- Clx

FeCl3 (NiCl2 is used in Chapter 5)

1st Heat treat

500 °C for 2 h in Ar

900 °C for 1 h in N2

1st Acid washing

4M HNO3 for 24 h

2M H2SO4 for overnight

2nd Heat treat

900 °C for 2 h in Ar

900 °C for 1 h in N2

2nd Acid washing

2M H2SO4 for 24 h

2M H2SO4 for overnight

3rd Heat treat

900 °C for 2 h in Ar, 1 h in H2

900 °C for 1 h in N2

Special Note

N/A

Ni-N-C takes an additional AW-HT

P a g e | 13

The morphologies of the catalysts were investigated using microscopy techniques. Scanning Electron

Microscope (SEM) was done using JEOL 7401F, for resolution between 10 and 100 µm. For higher

resolution (5 ~ 50 nm) images, Transmission Electron Microscopy (TEM) was performed using a FEI

Tecnai G2 Microscope 20 S-Twin with a LaB6-cathode at 200 kV accelerating voltage (ZELMI Centrum,

Technical University of Berlin, carried out by Ms. Xingli Wang). The TEM samples were sonicated in

ethanol solution and drop-dried onto Cu-grids and the analysis was done using software from ImageJ.

Inductively Coupled Plasma (ICP) was done to determine the overall metal content in the as-prepared

catalysts. 50 mg catalyst powder was firstly dispersed in aqua-regia and then solved using a microwave at

180 °C for 20 mins. The residual mixture was diluted to 50 ml and tested using a Varian 715-ES-inductively

coupled plasma (ICP) analysis system with optical emission spectroscopy detection (OES). Elemental

Analysis (EA) was used for the determination of bulk elemental distribution (mostly the non-metallic

species).

Figure 3- 1 Operando XAS cell used in this work: a) schematic illustration of the cell, 1 – working electrode

(GDE with the ink sample) sealed with Kapton tape, 2 – Pt gauze counter electrode, 3 – leak-free Ag/AgCl

reference electrode. Designed by Roldan Cuenya’s Group at FHI. b) The cell during measurements at

SAMBA beamline of SOLEIL synchrotron light source (Paris, France).

X-ray absorption fine-structure (XAFS) spectroscopy is carried in collaboration with Prof. Beatriz Roldan

Cuenya’s group at Fritz Haber Institute. The data were acquired at the undulator beamline P65 of PETRA

III storage ring (DESY, Hamburg, Germany) operating at 6 GeV in top-up mode. All XAS measurements

are performed in collaboration with Prof. Beatriz Roldan Cuenya’s group at Fritz Haber Institute. The

experiments were carried out in transmission mode at the Fe K absorption edge (7112 eV). Operando XAFS

measurements were performed in fluorescence mode at the SAMBA beamline of the SOLEIL synchrotron

(Saint-Aubin, France) using a 35-element solid-state Ge detector. A home-built operando electrochemical

P a g e | 14

cell was used, with Pt gauze (MaTek) serving as a counter electrode and a leak-free Ag/AgCl reference

electrode (Innovative Instruments Inc, shown in Figure 3-1, built up by Roldan Cuenya Group, FHI). CO2-

saturated KHCO3 was employed as the electrolyte. Raw data reduction was performed using the program

Athena.62 Analysis of extended X-ray absorption fine structure (EXAFS) spectra was conducted in Artemis

by using the FEFF8 code to extract the coordination numbers (CN), interatomic distances (r), disorder

parameters (Debye-Waller factor, σ2), and edge energy shift ∆E.63

X-ray photoelectron spectroscopy (XPS) was measured in an ultrahigh vacuum (UHV) system equipped

with a monochromatic Al Kα source (hν = 1486.5 eV) operated at 14.5 kV and 300 W, and Phoibos 150

(SPECS GmbH) analyzer. For each sample, a survey and high-resolution C 1s, O 1s, N 1s, Fe 2p, and S 2p

regions were measured. The C 1s signal of graphitic-like carbon was used for binding energy calibration

and assigned to 285 eV. The CasaXPS software with pseudo-Voight Gaussian-Lorentzian product functions

and Shirley background was used for peak deconvolution. Atomic ratios were calculated from XPS

intensities corrected by the corresponding sensitivity factors provided by the manufacturing company

(SPECS). In this work, most of XPS analysis was done in collaboration with Dr. Ilya Sinev and Prof. Beatriz

Roldan Cuenya.

CO chemisorption measurements were carried out to quantify the Fe-Nx sites per mass of the as-prepared

Fe-N-C catalysts (Thermo Scientific TPD/R/O 1110). Each experiment was performed on 100−150 mg of

the as-prepared catalyst in helium condition (He flow: 20 ccm). As a cleaning pretreatment step, the sample

was heated up to 600 °C and kept for 15 min. After the sample was cooled down to -80 °C, 6 times CO

pulses were carried out to perform the CO chemisorption. The CO-uptake mole amount per gram of catalyst

obtained from this measurement could be utilized to evaluate the FeNx site density (μmol g-1) of each Fe-

N-C catalyst.

P a g e | 15

3.3 Electrochemical Methods

Electrochemical characterization methods such as double layer capacitance measurements for

Electrochemical Surface Area (ECSA), Potentiostant Electrochemical Impedance Spectroscopy (PEIS),

Linear Sweep Voltammetry (LSV) and Chronoamperometry (CA) are of fundamental importance for this

work as they allow the evaluation of overall faradaic reactivity of as-prepared materials for during the

electrochemical operation.

For electrochemical characterization, an ink was produced containing 15 mg catalyst, 150 μl isopropanol,

800 μl deionized water, and 50 μl 5 wt.% Nafion perfluorinated resin solution (Sigma Adlrich). The ink

was sonicated for 8 minutes using an ultrasonic horn, and 50 μl of ink were dropped cast onto a 1 cm2 glassy

carbon electrode resulting in loading of 0.75 mg cm-2. The procedure is presented in Figure 3-2a).

Figure 3- 2 a) Preparation of the electrode: from catalyst powder via ink to layer. Catalyst loading: ~0.75

mg cm-2 on GC plate. b) Schematic H-type two compartments cell divided by a polymer membrane. In

operation, the CO2RR as well as other cathodic reactions occur at the WE (working electrode), whereas the

OER happens at the CE (count electrode).

The resulting electrode was inserted into a CO2-saturated, 0.1 M KHCO3 solution. The electrochemical

reduction of CO2 was carried out in a two compartments cell (see Figure 3-2b) divided by a polymer

electrolyte membrane (NR212, NR117, and Selemion were used, showing no performance difference). The

electrochemical data were acquired using a SP-300 potentiostat (Biologic).

P a g e | 16

Cyclic voltammetry was firstly carried out on the Fe-N-C catalysts in CO2-saturated 0.1 M KHCO3 at

various scan rates (dV

dt , 20 mV s-1, 15 mV s-1, 10 mV s-1, 5 mV s-1, 1 mV s-1) to estimate the double layer

capacitance (CDL), which is in principle proportional to the Electrochemical Surface Area (ECSA).

jDL = dQDL

dt = CDL ∗dV

dt eq. 3‐ 1

and

CDL ∝ ECSA ∝ BETSA eq. 3‐ 2

The potential cycling was performed between -0.1 and +0.42 VRHE to avoid the faradaic process. By

extracting the double layer current densities (jDL) at +0.16 V vs. RHE (middle of the E scanning range), we

were able to quantify the double layer capacitance of each catalyst under electrochemical conditions, see

eq. 3-1. In principle, the ECSA should be in agreement with the BET surface area.

A catalyst ink was produced with 15 mg catalyst, 150 μl isopropanol, 800 μl DI water, and 50 μl 5 wt.%

Nafion perfluorinated resin solution (Sigma Adlrich). After the dosing and mixing, the ink suspension was

sonicated using an ultrasonic horn for 8 minutes. 50 μl of ink were deposited onto glassy carbon with 1 cm2

area resulting in a catalyst loading of 0.75 mg cm-2. The prepared electrode was inserted into a CO2-

saturated, 0.1 M KHCO3 solution (Honeywell) in a two-compartments home-made H-cell, divided by an

anion exchange membrane (Selemion AMV, AGC Engineering Co., LTD). The electrochemical reduction

reaction is controlled using a SP-300 potentiostat (Biologic). 50% of the ohmic drop was automatically

corrected and the rest was corrected manually (see eq. 3-3). Before the bulk CO2 electrolysis, a Linear

Sweep Voltammetry step (LSV) was performed at the scan rate 5 mV s-1 from -0.1 V vs. RHE towards the

desired working potential and then kept the potential constant for 60 minutes.

Applied working potential against reversible hydrogen electrode

ERHE = ERef+EAg/AgCl + 0.059 ∗pH − R ∗ I eq. 3‐ 3

ERHE: Working potential against reversible hydrogen electrode/ VRHE

ERef: Applied potential against the reference electrode / V

EAg/AgCl: Potential of reference electrode measured against normal hydrogen electrode (0.21 V) / V

pH: pH-value of the electrolyte

R: Ohmic resistance between working and reference electrode / Ω

I: Total Current of the experiment / A

Measurements at high current densities were performed by Mr. Tim Möller (TU Berlin) in a commercial

Micro Flow Cell (MFC) supplied by ElectroCell. In all Flow-Cell experiments, a commercial Ir-MMO plate

(ElectroCell) was used as anode. The catalyst-inks were spray-coated on the microporous layer (MPL) of

a Freudenberg C2 gas diffusion layer (GDL) on an area of 3 cm2 to achieve a catalyst loading 1 mg cm-2.

P a g e | 17

Nafion (Sigma-Aldrich, 5 wt% resin solution) was used as binder and for ionic conductivity of the catalyst-

layer. For usual fabrication of the catalyst inks 15 mg of the M-N-C powders were dispersed with 60 μl

Nafion solution in a mixture of Milli-Q water and isopropanol. An aqueous solution of 1 M KHCO3 was

used as anolyte and catholyte, which were separated by an anion exchange membrane (Selemion AMV,

AGC Engineering Co., LTD.). Both electrolytes were cycled through each respective compartment at 100

mL min-1. The CO2 feed was supplied at rate of 50 mL min-1 to the cathode and was flown from the back

of the carbon paper through the catalyst-layer. Measurements were performed galvanostatically for two

hours at each respective current during the catalytic tests, sweeping the current from low to high values.

P a g e | 18

3.4 Products quantification

Online gas products quantification. The gas flow was continuously purged into the cell and the exhaust

was directly introduced in the Online GC (Shimadzu GC 2014, shown in Figure 3-3) for H-Cell

experiments, whereas the catholyte/gas mixture was returned to the catholyte reservoir first, in the Flow-

Cell, from where sampling was performed. Gas products were detected at a certain time (using H-Cell) or

every 30 min (using Flow-Cell) of bulk electrolysis.

Figure 3- 3 Schematic of online gas chromatograph testing platform. Detailed schematic of each part, 1.

Loop valve, 2. Column, 3. Thermal conductivity detector, 5. Flame ionization detector, is displayed in

Figure 3-4 below.

During the testing procedure, the gas purge out from the electrolyzer is continuously flowing in the GC,

and controlled by a 10-Valve, see Figure 3-4 a and b. At the valve setting “-91”, the gas sample fills in, and

passes through the Loop (sample loading), and then leaves from the GC setup as exhaust. By setting the

Valve from “-91” to “+91”, the gas sample in the Loop is carried by the carrier gas argon (sample dosing),

flowing into the column for separation. Thereafter, the valve is turned back to “-91” for the next set.

The gas sample, the mixture of various compounds, is carried by the argon flow. Different mobility (due to

polarity, mass, and size) leads to different retention time in the column. Usually, small and light compounds

like H2 could pass the column quickly, whereas the larger ones like CO2, take longer moving in the column.

The movements of these three compounds in the column as a function of retention time is illustrated in

Figure 3-4c.

P a g e | 19

Figure 3- 4 a) and b) Schematic of the 10-Valve for gas sampling in the GC, a) loading the sample in the

loop and b) dosing and transporting the loaded sample into the column. c) Schematic of gas sample (mixture

of various compounds) carried by Argon flow, flowing in the column (colored in green). For instance, H2

(blue) with better mobility, moves faster in the column and accordingly takes less retention time, therefore

could be detected as the first compound. On the contrary, CO2 molecules (grey) with lower mobility, suffers

longer retention time in the column. Schematic of the detectors e) thermal conductivity detector, and e)

flame ionization detector.

Thermal Conductivity Detector (TCD) is equipped to analyze the column effluent by sensing the thermal

conductivity changes (marked in blue, shown in Figure 3-4d) using the constant Argon flow as a reference

(marked in green, shown in Figure 3-4d). Application of the TCD consists of an electrically heated filament

in a temperature-controlled cell. Under normal conditions, the reference Ar flow and Ar carrier gas (as

column effluent) flow by the filaments, holding a constant filaments temperature and accordingly constant

resistance (R1 and R2). Once other gas, like H2, N2, and CO2, exists in the column effluent, the temperature

of the filament changes, causing the resistance (R2) shift, and this resistance change could be sensed by a

Wheatstone-bridge, further quantifying the sample amount.

To detect the carbon-based gases, such as the hydrocarbons, a more sensitive detector, flame ionization

detector (FID) is applied. As shown in Figure 3-4 e), the gas sample flows through the FID jet and into a

flame maintained by the hydrogen and air. The flamed ions could be collected by the electric field,

delivering a current signal. It is notable that the FID is rarely available for hydrocarbons and alcohols. To

detect carbon-oxygenates (CO or CO2), a methanizer is needed to first convert them into methane (see part

4 shown in Figure 3-3).

P a g e | 20

Data given by the GC is the volume percentage of one individual gas compound in the mixture sample

purge. Thus, the production rate of each is calculated with eq. 3-4. The respective faradaic efficiency and

partial current density are obtained using eq. 3-5 and eq. 3-6.

Reaction rate of gas products

n󰇗 = V

󰇗∗ C

A ∗ VM

eq. 3‐ 4

n󰇗 : Generation rate of the product / mol s-1 cm-2

󰇗: CO2 gas flow rate / L s-1

C: Concentration of the product detected by GC / Vol%

A: Geometric area of the electrode / cm-2

VM: Ideal gas molar volume / 22.4 L mol-1

Faradaic efficiency of gas Products

FE = n󰇗 ∗ z ∗ F

jtotal

∗100% eq. 3‐ 5

n󰇗 : Generation rate of the product / mol s-1 cm-2

FE: Faradaic Efficiency of the product / %

z: Charge transfer of each product

F: Faradaic Constant / 96500 C mol-1

jtotal: Total current density during CO2 bulk electrolysis / A cm-2

Partial current density

jx= FEx∗ jtotal eq. 3‐ 6

jx: Partial current density of individual gas product / A cm-2

FE: Faradaic Efficiency of the product / %

jtotal: total current density / mA cm-2

Liquid products quantification. After a certain period of electrolysis, the remained electrolyte was

injected in HPLC (Agilent 1200, Detector: RID, for formate and aldehydes) and liquid GC (Shimadzu 2010

plus, Detector: FID, for alcohols) for liquid products quantification. Faradaic efficiency of each is

calculated using eq. 3-7.

Faradaic efficiency of liquid Product

FE = V ∗ ΔC ∗ z ∗ F

Q∗100% eq. 3‐ 7

FE: Faradaic Efficiency of the liquid product / %

V: Volume of the electrolyte / L

ΔC: Accumulated concentration of the product detected by HPLC or liquid GC / mol L-1

z: Charge transfer of each product

P a g e | 21

F: Faradaic Constant / 96500 C mol-1

Q: Total charge transfer during the electrolysis / C

3.5 Density functional theory (DFT) calculation

In this thesis, all theoretical works are contributed by Mr. Alexander Bagger and Prof. Jan Rossmeisl at

University Copenhagen. The M-N-C model was created in ASE64 by a 3 × 5 unit cell of graphene with a

functionalized Fe-N4 site by removing carbon atoms. The outmost carbon atoms were fixed in position and

periodic boundaries were applied. Further, the metal (111) model was built by a standard 3 x 3 x 4 slab

including a vacuum region and the two lower layers fixed. The electronic calculations were carried out with

the GPAW software65 with the projector augmented wave method, spin polarization (Fe-N-C) and the

revised Perdew–Burke–Ernzerhof (RPBE) functional.66 In detail, the DFT free energy (chemisorption

energy) of each species is calculated according to equation 3-8. ΔGA displays the chemisorption energy of

matter A on M surface. EA and EM represent the energy of the isolated adsorbate and free surface, whereas

EAM stands as the total energy of the adsorbed system.

ΔGA= EAM − EA− EM eq. 3‐ 8

ΔGA: Binding energy of A species on M surface / eV

EAM: Total free energy of A species and M surface under adsorption condition / eV

EA: Free energy of un-adsorbed free A species / eV

EM: Free energy of clean M surface / eV

We applied a 0.18 grid spacing together with a (2 × 2 × 1) k-point sampling for M-N-C and (3 x 3 x 1) k-

point sampling for Cu(111) and all the structure were relaxed to a force below 0.1 eV/Å. The free energy

diagrams were calculated using the hydrogen electrode67 and thermodynamic values from reference.68 The

functional error of the calculated CO2 RPBE energy was corrected by 0.45 eV together with a -OH water

correction of 0.25 eV and a *CO water correction of 0.1 eV.13

P a g e | 22

Chapter 4. Understanding activity and selectivity of metal-nitrogen-doped

carbon (M-N-C) catalysts for electrochemical reduction of CO2

Most parts of this chapter are reproduced from the journal article:

Understanding activity and selectivity of metal-nitrogen-doped carbon catalysts for electrochemical

reduction of CO2

Link: https://doi.org/10.1038/s41467-017-01035-z

with permission from Nature Communications, 2017, 8, 944, as Reference 51 in this dissertation. Copyright

2017 Springer Nature (CC BY 4.0).

Author list: Wen Ju, Alexander Bagger, Guang-Ping Hao, Ana Sofia Varela, Ilya Sinev, Volodymyr Bon,

Beatriz Roldan Cuenya, Stefan Kaskel, Jan Rossmeisl and Peter Strasser

Contributions: P.S., W.J. and A.S.V. conceived and designed the experiments. J.R. and A.B. performed the

DFT calculation. S.K., B.R.C., G-P.H., I.S. and B.V. carried out the chemical synthesis, microscopic and

spectroscopic characterization of the carbon-based materials and subsequent in depth data analysis. W.J.

and A.S.V carried out the electrocatalytic tests and analyzed the results. G-P.H., P.S. and W.J. aggregated

the figures and co-wrote the manuscript. All authors discussed the results, drew conclusions and commented

on the manuscript.

P a g e | 23

Au and Ag the noble metals showing great performance of selectively producing CO as the major product

during the electrochemical CO2 reduction, however, due to high-cost and low-abundance, the precious

metals could hardly be considered as the top-prior candidate for industrial scale investment. Thanks to the

progress in Non-Platinum-Group-Metal (Non-PGM) for ORR investigation, a sort of non-precious metal,

nitrogen doped carbon (M-N-C) catalysts was designed, exhibiting excellent catalytic property for the fuel

cell utilization. Recent studies proof their great potential for CO2RR, outperforming the Ag/Au candidates

in CO yielding. Nonetheless, fundamental understanding in depth is still needed for further catalyst-

materials optimization.

In this contribution, we present advances in the understanding of trends in the CO2 to CO electrocatalysis

of the metal- and nitrogen–doped porous carbons containing catalytically active M-Nx moieties (M=Mn,

Fe, Co, Ni, Cu). We investigate their intrinsic catalytic reactivity, CO turnover frequencies, CO faradaic

efficiencies and demonstrate that Fe-N-C and especially Ni-N-C catalysts rival Au- and Ag-based catalysts.

We model the catalytic active M-Nx moieties using Density Functional Theory and correlate the theoretical

binding energies with the experiments to give reactivity-selectivity descriptors. This gives an atomic-scale

mechanistic understanding of potential-dependent CO and hydrocarbon selectivity from the M-Nx moieties

and it provides predictive guidelines for the rational design of selective carbon-based CO2 reduction

catalysts. 51

P a g e | 24

4.1 Synthesis and characterization

Figure 4- 1 Visualization, porosity and illustration of the M-N-C catalyst. a) Typical SEM image of the

family of Nitrogen-coordinated metal-doped (M-N-C) carbon electro-catalysts, scale bar: 4 m; b) CO2

physisorption isotherm (273 K); inset: the pore size distribution; c) Materials model and a schematic local

structure.

We have synthesized a family of M-N-C electrocatalysts starting with bipyridine-based coordinated

polymers and a variety of transition metals such as Mn, Fe, Co, Ni, and Cu. All chemicals were used as

received. Typically, 4,4-Dipyridyl hydrate (bipy, 1.114 g, Sigma-Aldrich Co.) was dissolved in 100 mL

ethanol solution. A certain amount of CoCl2·6H2O (1.2 g) was dissolved in 900 mL DI water solution. Then

the bipy solution was mixed with a CoCl2·6H2O solution and left standing for 24 h without stirring. Then,

50 mL CuCl2·2H2O (0.1 M) solution was rapidly mixed with the bpy-Co2+ solution and aged for 4 h.

Subsequently, the resultant product was collected by centrifugation with the speed of 4200 rpm for 12 min.

After drying, the polymer product was carbonized at 500 °C for 2 h at a heating rate of 60 °C h−1 in Ar

atmosphere. Finally, hydrophilic N-doped porous carbons (N-C) with trace amounts of Cu were obtained

after leaching in 4 M HNO3 for 24 h.

Subsequently, additional transition metal species (MMn, Fe, Co, Ni) were introduced in N-C through

incipient impregnation of MClx solutions. The nominal weight concentration of M respective to N-C was

set to 25 wt%. The dried M-N-C composite was re-pyrolyzed at 900 °C for 2 h at a heating rate of 2.0 °C

min−1 in Ar atmosphere. The carbonized M-N-C was dispersed in aqueous sulfuric acid (ca. 2.0 mol L−1)

P a g e | 25

and refluxed at 100 °C for 1 day. The leached sample was collected and washed with DI water until pH

value close to neutral. Finally, the leached sample was treated at 900 °C first in Ar for 2 h, and change to

H2 for another 1 h, then change back to Ar, let it cool down and harvest the final M-N-C electrocatalysis

(MMn, Fe, Co, Ni). The Cu-N-C material was obtained directly after the reductive annealing procedure

without any additional acid leaching.

Materials characterization started with morphological and gas adsorption experiments (Figure 4-1a,b). The

M-N-C electrocatalysts showed hierarchical chemical structures with visible macropores (Figure 4-1a,

Figure S1-1). The pore size distribution peaks narrowly at ca. 0.7-0.8 nm (2.5-2.9 times of the dynamic

diameter of CO2 molecules, Figure 4-1b inset), enabling this family M-N-C materials a remarkable 4.0-4.5

mmol g-1 capacity for CO2 capture at atmospheric pressure (Figure 4-1b) due to their high adsorption

potential to trap CO2 molecules.69 This could result in CO2 enrichment within a local environment despite

the low CO2 solubility in the working electrolyte. Figure 4-1c displays a structural illustration of the

interconnected macropore walls, composed of thin carbon branches with highly accessible micropores, all

over which the coordinated metal sites as well as N-containing carbon lattice are homogeneously

distributed.

The N2 physisorption isotherms (Figure S1-2) are essentially type I for Cu, Co, Ni, or Mn-N-C samples,

indicating their microporous nature, while the visible hysteresis of Fe-N-C material reveals the presence of

a small fraction of mesopores in addition to micropores. Notably, a significant increase of gas uptake was

observed at higher relative pressure (P/P0>0.9) for all M-N-C samples, indicating their rich macroporosity,

which is consistent with the SEM images (Figure 4-1a). The specific surface area based on Brunauer–

Emmett–Teller (BET) theory is in range of 615-938 m2g-1, while the Ni-N-C and Mn-N-C show the lowest

and the highest BET surface area, respectively, and the others are in between (Table S1-1). This shows a

roughly linear relationship with the double layer capacity under the electrochemical condition (Figure S1-

3, Table S1-1). The M-N-C samples showed a moderate hydrophilic character (Figure S1-4). The XRD

patterns (Figure S1-5) reflect the predominant amorphous carbon support, particularly for Mn, Co, Ni or

Cu-N-C; while the presence of Fe, to some extent, led to graphitic domains. Some residual Fe, Co, Ni in a

metallic state was detectable after the H2 reduction at 900 oC. The STEM elemental mappings (Figure S1-

6) are fully consistent with the XRD findings showing presumably carbon-encapsulated metal particles as

well as coordinated metal ion sites for the three catalysts.

P a g e | 26

Figure 4- 2 High-resolution XPS characterization. N-1s XPS core level region of (a) Co, (b) Mn, (c) Ni and

(d) Fe doped M-N-C catalyst. The 2p3/2 spectra of the corresponding metal peaks (Co-2p, Mn-2p, Ni-2p,

Fe-2p) is shown in Supplementary Figure S1-8.

The catalyst surface chemical composition and state were investigated using X-ray photoelectron

spectroscopy (XPS). Fitted high-resolution N1s spectra (Figure 4-2 for Co, Mn, Ni, Fe, and Figure S1-7 for

Cu, detailed fitted parameters in Table S1-2) evidenced the presence of the porphyrin-like metal-

coordinated M-Nx moieties (399.7 eV), as well as pyrrolic (401.3 eV), pyridinic (398.6 eV), graphitic (402.5

eV), and N-Ox (403.9 eV) species (Figure 4-2).49,50 Additionally, a weak and broad peak can be fitted at

higher binding energies, centered at 405.9 eV, which is likely assigned to trace amounts of non-decomposed

nitrogen precursors.70 The N1s spectra of all samples are dominated by pyrrolic nitrogen (see Table S1-2),

whereas the M-Nx moiety gives rise to the most intense core level for the Ni-doped sample. A detailed

analysis of the metal 2p3/2 shake-up photoemission lines (marked in Figure S1-8) offers insight in the

chemical state of the metallic species. Combined, our materials characterization confirmed the prevalent

presence of N-coordinated metal single-site moieties, M-Nx, near the surface in all catalyst samples, except

for the Cu sample that exhibited evidence of near-surface metallic Cu particles (Figure S1-7a).

Better understanding of the type of metallic species present in these samples can be obtained by a detailed

analysis of the corresponding 2p3/2 photoemission lines, shown as insets in Figure S1-8. Due to the low

metal loading in our samples, we cannot fit the data with the multiple splitting features typical for oxides

and hydroxides.71 Instead, in order to gain insight into whether oxidized metal species or nitrogen-ligated

metals prevail in our samples we have compared the main 2p3/2 photoemission peak positions with data

reported for divalent metal species and metalloporphyrins, and a ratio of its area to the area of the shake-up

P a g e | 27

satellites, which serves as a fingerprint of 3d metals in 2+ state. Muralidharan and Hayes have reported

shake-up structures for Co(II), Ni(II) and Cu(II) porphyrins to be significantly weaker than the

corresponding satellites of simple oxides of the same elements.72 Thus, Ni(II) porphyrin exhibits an XP

spectrum with 2p3/2 peak centered at 854.8 eV, which is similar to Ni(OH)2, but unlike the latter it has a

shake-up structure of only 11% intensity as compared to the main line. In fact, the Ni-N-C sample in our

study has the main Ni 2p3/2 photoemission peak at 854.9 eV and a shake-up peak just 18% of its area,

suggesting a dominant presence of Ni bound to nitrogen in our samples. In contrast, the Co-N-C sample

displayed a significantly more intensive shake-up structure, ca. 20% of the main Co 2p3/2 peak area, which

is considerably higher than a value reported for Co(II) porphyrin (6.5%), but also clearly lower than CoO

and Co(OH)2 (38% and 54% correspondingly). Mn 2p3/2 of Mn-N-C has its maximum at 641.7 eV, similar

to MnO, and lower than compounds with Mn3+ and higher oxidation states.73 At the same time, the shake-

up structure is as high as 11% of the main photoemission peak, which is slightly higher than the one reported

for MnO (9%). Unfortunately, there is no reliable XPS reference for the manganese porphyrin structure.

One can speculate however that Mn-N-C clearly contains divalent Mn species with a ligand structure

different from common Mn2+ compounds. The Cu 2p XP spectrum (Figure S1-7a) shows a clear presence

of divalent copper, as evident from the well-developed shake-up. Cu 2p3/2 XP spectrum shows two sharp

peaks at 932.7 and 934.8 eV and a shake-up satellite structure between 938.4 and 947.3 eV. A binding

energy of 934.8 eV is similar to the value reported by Biesinger at al71 for Cu(OH)2, but also to Cu

porphyrin.72,73 The observed intensity of the shake-up satellite structure is 59% of the main line, which is

only slightly higher than the 53% for both, Cu-porphine72 and Cu(OH)2 reported by Biesinger71. The peak

at 932.7eV can be assigned either to metallic copper or to Cu2O, however a detailed analysis is not possible

due to weak CuLMM Auger line.71 It appears however reasonable to assign this peak to the Cu+ state, since

metallic copper in the near-surface region is unlikely to remain after multiple acid treatments. The Fe 2p3/2

spectrum of the Fe-N-C sample shows a broad peak at 710.8 eV with a shoulder at ca. 708 eV. The position

of the former and absence of distinct satellite structure points out the presence of significant amount of

Fe2O3, while the latter is in an agreement with the values reported earlier for iron phthalocyanine.74

According to Fe 2p3/2 XP spectrum deconvolution, the nitrogen-coordinated iron comprises only 1.6% of

the total Fe content observed by XPS.

P a g e | 28

4.2 Electrochemical CO2RR over various M-N-C catalysts

Figure 4- 3 CO2 reduction reaction activities. Linear sweep voltammetry of a) Mn-N-C, b) Fe-N-C, c) Co-

N-C, d) Ni-N-C and e) Cu-N-C in CO2-saturated 0.1 M KHCO3 (solid lines) and in N2-saturated 0.1 M

KH2PO4/K2HPO4 (dashed lines) with a catalyst loading of 0.76 mg cm-2 at 5 mV s-1 in cathodic direction.

As a first test of the total faradic reactivity of our single-site solid catalysts in CO2-saturated 0.1 M KHCO3,

comprising both the HER and CO2RR, Linear Sweep Voltammetry (LSV) were performed between 0.0 and

-0.7 V vs. RHE, blue solid curves in Figure 4-3. Comparison with LSVs performed in absence of CO2 (red

dashed curves) revealed substantial CO2RR activity of the Mn, Fe, Ni and Cu-doped catalysts. Furthermore,

the Mn, Fe, Ni and Cu -doped samples exhibited an earlier onset (smaller overpotential) for the CO2RR

than HER, suggesting that, at least in a small potential window, they are selective towards CO2RR. By

contrast, Co-N-C presented a comparable activity suggesting that the HER may be the dominant faradic

process over the investigated potential range.

Longer-term catalytic performance testing was conducted using potential-controlled 1 hour electrolysis.

The geometric electrode area-normalized (jgeo) and the active interfacial area-normalized (double-layer

capacity-normalized) faradaic currents (jDL) after 15 min and 60 min are compared in Figure S1-9. The Co-

N-C catalyst generated the most overall faradaic current, while the Cu-N-C displayed the poorest overall

reactivity at larger overpotentials, in accordance with Figure 4-3.

P a g e | 29

Figure 4- 4 Catalytic performance and product analysis. (a-c) Faradaic Efficiencies (FE) vs. applied, iR-

corrected electrode potential of a) H2, b) CO and c) CH4. d) Catalyst mass-normalized CO partial currents

(mass activity) vs. applied potential for the five M-N-C catalysts compared to state-of-art Au catalysts

(performance ranges of Au-nanoparticle and Au-nanowires are shown by filled areas 10,11. Lines to guide

the eye. Conditions: 60 min at constant electrode potential in CO2-saturated 0.1 M KHCO3 with 0.76 mg

cm-2 M-N-C catalysts loading. Faradaic efficiencies and CO yields after 15 min are shown in Figure S1-10.

The stationary faradaic efficiencies (FE) of the three principal CO2RR products after 60 min electrolysis

are displayed in Figure 4-4. H2 and CO accounts for up to 95% of the transferred charge on the single-site

catalysts. Remarkably, small amounts of methane were detected, however only on Fe and Mn catalysts,

while no liquid product could be detected. Despite the low number of active surface single-sites on the M-

N-C catalysts, their mass-based partial CO currents (production rate) towards CO meets and are comparable

to that of Au-based catalysts10,11, especially at technologically interesting higher currents (Figure 4-4d).

These results highlight the significance of this family of compounds as non-precious, earth-abundant low-

cost and efficient CO2RR catalyst alternatives for the electrochemical production of CO in CO2 -consuming

electrodes.

To arrive at a fundamental mechanistic understanding of the CO2 catalysis on the single-site materials, we

focus on reactivity trends among the M-N-C catalysts at different applied overpotentials. The CO2RR

performance exhibits a strong dependence on the nature of the transition metal, not only in terms of the

molar CO/H2 ratio, but also in the experimental overpotential at maximum CO efficiency (see Figure 4-4 a

P a g e | 30

and b). H2 FE on Co-N-C remains above 80% over the entire electrode potential range, making it a catalyst

with poor selectivity towards CO2RR. On the other hand, Fe-N-C and particularly Ni-N-C catalysts clearly

act as highly promising catalyst for selective CO production, however, the maximum CO FE is obtained at

a smaller overpotential on Fe-N-C (VRHE = -0.55V, FECO = 65%) than on Ni catalyst (VRHE=-0.78V,

FECO=85%). Note that the selectivity of these two single site catalysts is drastically different from that of

metallic Ni and Fe catalysts, which yield H2 as virtually the only major product.4 We have conducted a

number of control measurements to confirm that the M-Nx site is indeed the most significant active center

for CO2 reduction into CO. First, the Nitrogen Free M-C as well as the polymer precursor before pyrolysis

(Cu-Bpy) contributes negligibly to the CO activity during the CO2RR process. Secondly, we could not

exclude some catalytic activity of the Nitrogen functionalities. However, based on their finite CO2RR

catalytic reactivity and the rough similarity of the distribution in the M-N-C catalysts, their effect could be

seen as a known weak background signal for all cases (Supplementary Table S1-2, Figure S1-11 and Figure

S1-12). This finding strongly suggests that the CO2RR (into CO) reactivity trends purely originate from the

differences in intrinsic catalytic activity of the various M-Nx moieties.

P a g e | 31

4.3 Correlating the theoretical prediction and experiments

Figure 4- 5 Experimental correlation to simulations. Experimental CO production turnover frequency

(TOF) of the M-N-C catalysts versus applied iR-corrected electrode potential. The a) catalytic reactivity

trends and b) reaction pathway split into three potential regions with distinctly different rate-determining

mechanistic features. Free energy diagrams for HER and CO2RR at 0.0 and -0.6 VRHE are given in Figure

S1-13. Insets: Region 1: Low overpotentials, the experimental onset potentials of CO production (better

seen on the log (CO TOF) – E plot in Figure S1-14) correlate with the binding energy of the reaction

intermediate COOH* taken from Figure S1-13. Region 2: Intermediate over-potentials, CO production TOF

at -0.6 VRHE correlates with the free energy of adsorbed CO, CO* taken from Figure S1-13; Region 3: High

overpotentials, free energy diagrams for the HER (dashed paths) and CO2RR (solid paths) at -0.8 VRHE for

each M-N-C catalyst. HER barriers are high for Ni and Cu, while CO2RR is downhill making these

materials favorable CO producing catalysts.

P a g e | 32

To bring theoretical mechanistic insight, DFT simulations pertaining to the CO2 reduction process on N-

coordinated metal-doped M-N-C catalysts were carried out. For this purpose, we took the single-site motif

M-N4 as active site to calculate the binding energy of the different reaction intermediates (see Figure S1-

13). We note that there exist other M-Nx functionalities75-77, however, previously we computationally found

the metal be the dominating factor compared to other M-Nx functionalities.78 Thus, the M-N4 site appears

to be a reasonable single active site model for our analysis here. For the model we calculate the binding

energies without electrolyte, which is reasonable for the trends and conclusions drawn here. While activity

can often be associated with a single descriptor, selectivity can obviously not, as it is related to competition

between different possible reaction paths. The different reaction paths show different dependence on metal

center and potential.

Figure 4-5 compares the trends in the experimental CO-specific turnover frequencies (TOF) of the five M-

N-C catalysts. The TOF values were derived from the absolute CO production rates normalized by the

respective BET surface area-weighted surface M-Nx concentration (see Equation 4-1).

TOF = V

󰇗∗ C

R ∗ M ∗ ABET ∗ xMetal−N

eq. 4‐ 1

TOF: BET Area and XPS Surface Metal-Nitrogen mole fraction Normalized CO generation rate

󰇗: CO2 gas flow rate / L s-1

C: Concentration of the product detected by GC / Vol%

R: Ideal Gas Volume / 22.4 L mol-1

M: Catalyst loading / g m-2

ABET: N2 physiosorption BET surface area/ m2 g-1

xMetal-N: XPS surface Mole fraction of Metal-Nitrogen

By correlating the experimental TOF trends and predicted DFT theoretical energy diagrams, we were able

to identify three regions of distinct reaction dynamics that control the electrocatalysis. Firstly, region 1, a

dynamic regime at low overpotentials near the onset of the CO production, then, region 2, a dynamic regime

at intermediate overpotentials and finally, Region 3, at larger overpotentials where the CO2 reduction

current densities approach technologically relevant levels. The reason for this division is that the order in

catalytic activity change in the different regions indicating that the rate in the different regions is determined

by a distinctly different surface chemistry.

The low overpotential regime around -0.45 VRHE (Region 1). Defining the CO production onset potential to

be the applied electrode potential at which the CO TOF exceeds 0.2 mmol h-1 m-2active, the Fe, Mn and Co-

N-C catalysts start producing CO at around -0.4 V vs. RHE, while the Cu and Ni samples require

considerable higher overpotentials, see Region 1 insert in Figure 4-5. The onset potential is determined by

the mechanistic elementary step that is the last to become exergonic as the overpotential is increased

P a g e | 33

(limiting potential). Simulations suggest that this potential-determining step is the first proton-coupled

electron transfer reduction of CO2 to adsorbed COOH* according to:

CO2+ H++ e− ⇆ COOH∗ eq. 4‐ 2

In agreement with electrochemical measurements, in the simulations the catalyst falls into two groups: Co,

Fe and Mn requiring only a small overpotential, whereas Cu and Ni need a larger thermodynamic driving

force for that step.

The intermediate overpotential regime around -0.6 VRHE (Region 2). Here, the Fe- and Co-N-C catalysts

approach their maximum CO2RR reactivity, while the Ni-N-C catalyst has barely passed above its CO

production onset. With the electrode potential being now past the limiting electrode potential, the overall

CO2 reduction reaction invariably becomes limited by a non-faradaic chemical reaction step. The larger the

thermodynamic driving force of this step, the faster the overall reaction rate. Correlating experiments to

DFT calculations reveals that the logarithm of the experimental CO TOF is now linearly related to the CO*

binding energy descriptor, see Region 2 insert in Figure 4-5. This suggests that the rate-controlling

intermediate has shifted from COOH* to CO*. As a result of this, the overall reaction rate appears limited

by the process

COOH∗+ H++ e− ⇆ CO(g) + H2O eq. 4‐ 3

While DFT predictions do not allow us to unambiguously pinpoint the exact rate-limiting point along the

reaction coordinate of step (2), we hypothesize that it is the chemical dissociative formation of H2O

according to

COOHH∗⇆CO∗+ H2O eq. 4‐ 4

The stronger CO* binds, the more driving force is available for this step. An evidence for the hypothesis

can be considered by comparing the Fe- and Mn-N-C, which have almost similar COOH* and H* binding.

However, these descriptors cannot explain the experimental different CO TOF from the two, while the

logarithm to the CO* descriptor can. (Figure S1-14)

The large overpotential regime < -0.7 VRHE (Region 3). Here, the experimental CO formation TOF in Figure

4-5 as well as the faradaic CO efficiencies of the Fe-N-C and Co-N-C catalysts have passed their maximum

and trend downward (see Figure 4-4b). That of Mn-N-C is levelling off, while the Ni-N-C catalyst continues

to increase its CO production rate at a very high faradaic CO efficiency, significantly outperforming all

other single-site catalysts as well as Au catalysts.

Our mechanistic DFT analysis shown in inset, Region 3, of Figure 4-5 is able to consistently explain all

these experimental findings. The free energy diagrams of the HER (dashed) and CO2RR (solid) evidences

that the Fe-, Co- and Mn-based catalysts start to strongly catalyze the hydrogen evolution (H++e- → H* →

P a g e | 34

H2(g)) illustrated by the all downhill reaction energy pathway. Among them, Co-N-C is the most efficient

HER catalyst and, thus, displays the highest faradaic efficiency for hydrogen evolution, see Figure 4-4a. In

contrast, the Ni- and Cu-based catalysts exhibit very weak binding of H* which makes the HER

thermodynamically unfavorable at -0.8 VRHE, giving rise to low faradaic hydrogen efficiencies.

The DFT predictions of the CO2RR pathway (CO2 → CO* → CO(g)) at -0.8 VRHE complete the mechanistic

picture. While the Ni-Nx and Cu-Nx moieties stand out as the single sites with the weakest binding to

COOH* and therefore with the largest overpotential to start CO2 reduction (see Region 1), their weak

binding of CO* prevents the potential-independent chemical CO-desorption process (CO* → CO(g)) to

become rate-limiting. This is in contrast with the Mn, Fe, Co-Nx sites whose CO TOF is controlled by the

CO* → CO(g) step due to their strong CO* binding (solid pathways in Region 3) leading to a positive ∆G

of CO desorption. Indeed, experimentally, the CO TOF values of the Mn, Co and Fe-N-C level off or slow

down, while the hydrogen evolution accelerates.

We note that the relatively strong binding of CO* on Fe-Nx and Mn-Nx single sites predicted for Region 3

is fully supported by the experimentally confirmed exclusive ability of these two catalysts to produce the

hydrocarbon CH4, see Figure 4-4c. In simple terms, one could say that to produce subsequent reaction

products from CO during the CO2RR, the CO molecule must be bound strong and long enough to undergo

subsequent dissociation and hydrogenation steps to arrive at CH4. For the Ni and Cu catalysts, the CO*

detaching is energetically all downhill reaction which prevents further transformations. This makes Ni-Nx

and Cu-Nx single-site catalysts ideal electrochemical CO producers. We note that the experimentally

observed reactivity trend of the Cu-Nx catalysts in region 3 does not closely follow that of Ni. This is due

to a DFT-predicted thermodynamic instability (not shown) of the Cu-Nx moiety under the strongly reducing

conditions of < -0.7VRHE in region 3. As a result of this, the N-coordinated Cu ions spontaneously reduce

to metallic Cu nanoparticles – as confirmed by our XPS results – which show lower CO efficiency and

lower TOF values at electrode potentials of region 3.12

P a g e | 35

4.4 Discussion

In this work, we found a family of solid, single site, N-coordinated transition metal-functionalized

nanoporous carbons that show very high electrocatalytic reactivity and selectivity with respect to the direct

CO2 reduction to CO (CO2RR). A technical challenge in these M-N-C catalysts is to achieve a high density

of active M-Nx sites, while minimizing effects of other nitrogen moieties and inorganic metal impurities,

which, for this class of materials cannot be completely excluded. However, based on our experimental

observation, we could confirm that the M-Nx site play the dominant role during the CO2RR process into

CO. For instance, the Co-Nx sites were efficient hydrogen producers whereas the Fe- and Ni-Nx single site

catalysts showed a unique reactivity and faradaic efficiency for reducing CO2 into CO, meeting and

exceeding the mass-based activity of state-of-art Au catalysts at a fraction of their cost.

To understand the trends in reactivity and selectivity of the single site catalysts we have correlated our

experimental results with DFT simulations of the energetics of the competing reaction pathways involved.

Our results demonstrate that the binding energies can be used as descriptors to predict the CO2RR activity

and selectivity of this class of catalysts. This is why we find a good agreement between the DFT predictions

and the catalytic experiments offering a detailed mechanistic understanding of the role of the metal centers

in the considered catalytic processes.

Consistent with experiments, Co-Nx sites displayed all-downhill energetics for hydrogen, but severe

energetic barriers to CO formation. By contrast, the low H* and CO* binding energy of the Ni-Nx single

site required larger overpotentials to jump start the reactions. At larger overpotentials, however, Ni-Nx

catalysts displayed all-downhill energetics toward CO, while hydrogen evolution is hindered.

The high CO efficiency at medium and large overpotentials of the Fe-N-C and Ni-N-C materials combined

with their earth-abundant constituents, compared to standard Au catalysts, makes them attractive catalysts

for deployment in future industrial CO2-consuming CO cathodes for use as counter electrode process in the

chlorine production industry. The choice of catalyst thereby becomes a tradeoff between voltage efficiency

(Fe produces most CO at lower potentials) and turnover frequency/current density (Ni makes most CO at

higher overpotentials). In particular for Ni-N-C catalysts, high CO efficiencies at current densities

approaching industrial levels make them suitable candidates for CO2-consuming gas-diffusion cathode

(CCC) designs to be deployed in next-generation chloralkaline electrolyzers. By eliminating the need for

fossil fuel-based steam reforming toward purified CO feed streams in the Chlorine-mediated polymer

industry, CO2 reuse in chlorine–CO co-electrolyzers would significantly contribute to a lowering of

industrial CO2 emissions worldwide.

P a g e | 36

Chapter 5. CO evolution at industrial current densities on Ni-N-C derived

gas diffusion electrode

Most parts of this chapter are reproduced from the journal article:

Efficient CO 2 to CO electrolysis on solid Ni–N–C catalysts at industrial current densities

Link: https://doi.org/10.1039/C8EE02662A

Society of Chemistry.

Author list: Tim Möller, Wen Ju (co-first), Alexander Bagger, Xingli Wang, Fang Luo, Trung Ngo Thanh,

Ana Sofia Varela, Jan Rossmeisl and Peter Strasser

Contributions: T.M, and W.J. conceived and designed the experiments. A.B. performed the DFT calculation.

W.J. and F.L. carried out the synthesis and regular characterization. X.W. carried out the HRTEM

measurements and data analysis. Electrochemical CO2RR tests were done by T.M. (GDE-MFC) and W.J.

(H-cell). All authors discussed the results, drew conclusions and commented on the manuscript.

P a g e | 37

The electrochemical CO2 reduction reaction (CO2RR) to pure CO streams in electrolyzer devices is posed

to be the most likely process for near-term commercialization and deployment in the polymers industry.

Delivered by the work presented in Chapter 4, the Ni-Nx and Fe-Nx hold the promise to convert CO2 into

CO at high- and low- overpotentials, respectively. We thus employ the PANI-based M-N-C candidates and,

initiate the electrochemical testing in regular small-scale CO2-saturated liquid electrolyte H-cell screening,

and obtain parallel performance of that given by the Cu-MOF derived M-N-C catalysts. However,

electrolysis conducted in normal H-type electrolyzer only provides limited current densities below 50 mA

cm-2, which is far not sufficient to commercializing level.

To overcome this bottleneck, we deploy a type of Gas Diffusion Electrode (GDE) combined Micro-Flow-

Cell (MFC) to drive the CO2 electrolysis on M-N-C catalysts, the catalytic performance of Ni-based one

rivals or exceeds the state-of-art electrocatalysts under industry equivalent conditions. tests and moved to

larger-scale CO2 electrolyzer cells, where the catalysts were deployed as GDEs to create a reactive three-

phase interface. We compared the faradaic CO yields and CO partial current densities of Ni-N-C catalysts

to a Ag-based benchmark, and to its Fe-containing Fe-N-C analogue under ambient pressures, temperatures

and neutral pH bicarbonate flows. Prolonged electrolyzer tests were conducted at industrial current densities

of up to 700 mA cm-2. Ni-N-C electrodes are demonstrated to provide CO partial current densities above

200 mA cm-2 and stable faradic CO efficiencies around 85 % for up to 20 hours (at 200 mA cm-2), unlike

their Ag benchmarks. Density Functional Theory-based calculations of catalytic reaction pathways help

offer a molecular mechanistic basis of the observed selectivity trends on Ag and M-N-C catalysts.

Computations lend much support to our experimental hypothesis as to the critical role of a N-coordinated

metal ion, Ni-Nx, motifs as the catalytic active site for CO formation. Apart from being cost effective, the

Ni-N-C powder catalysts allows flexible operation under acidic, neutral, and alkaline conditions. This study

demonstrates the potential of Ni-N-C and possibly other members of the M-N-C materials family to replace

precious group metal catalysts in CO2-to-CO electrolyzers. 58

P a g e | 38

5.1 Catalysts synthesis and characterization

In this approach, a series of poly aniline (PANI) derived M-N-C catalysts are synthesized as the candidates

for electrochemical CO2 reduction. The synthesis is done by mixing and pyrolyzing PANI as the Nitogen-

precursor and a high surface area Ketjen 600 as the carbon source (hard template), whereas Fe- and Ni-

chloride are used as metal precursors. High-temperature pyrolysis of the precursor mixtures is performed

in pure N2 atmosphere at 900°C for 1 hour, after which residual metal species were leached with H2SO4.

This catalyst synthesis is based on a proven recepie for solid PGM-free metal/nitrogen doped carbon

powder catalysts previously tested for the oxygen reduction reaction.37,39,41 Synthesis procedure of the

regarding M-N-C catalysts is listed below (schematic is shown in Figure 5-1).

Figure 5- 1 Synthesis procedure of our studied polyaniline (PANI) based M-N-C catalysts.

Preparation of carbon support. Treated Ketjen EC 600JD (AzkoNobel) was used as the hard template of the

PANI based catalyst. It was firstly stirred in 0.5 M HCl for 24 hours and then rinsed with DI water to reach

neutral pH. Afterwards, the dried carbon was refluxed in concentrated nitric acid for 8 hours at 90 °C

followed by rinsing with DI water till the pH value turned to 7. After drying at 90 °C in the oven, the carbon

will be referred as carbon support. Before the further utilization for catalyst synthesis, the carbon support

was sonicated in 50 mL DI-water till homogenously dispersed.

Synthesis of Fe-N-C. Preparation of the Fe-N-C follows a procedure published in Ref97. 3 ml of aniline, 5

g iron chloride (FeCl3) and 5 g ammonium persulfate (APS, (NH4)2S2O8) was added to 0.5 L of 1 M HCl

and stirred for one hour. Then, the suspension was stirred for 48 hours along with 0.4 g of dispersed carbon

support. Afterwards, the suspension was dried at 95 °C for 24 hours. After drying, the solid mixture was

ball-milled in a Zr2O3 container with Zr2O3 balls (ball diameter 1 cm). Heat treatment (HT) was performed

in a furnace with a ramp of 30°C min-1 to 900 °C and kept at this temperature for 1 hour within N2

atmosphere (flow rate: 30 ccm). After cooling down, the material was washed in 2 M sulfate acid (AW)

overnight and rinsed to neutral pH by use of vacuum filtration. Usually, to obtain high purity active M-Nx

coordinative moieties, multiple times acid-washing is a necessary process to remove the excessive inorganic

species and the XRD profiles in terms of acid washing amount is recorded in Figure S2-1. An HT-AW-

HT-AW-HT-procedure was performed for to obtain the final Fe-N-C catalyst.

Synthesis of Ni-N-C. The synthesis of Ni-N-C is analogous to the Fe-N-C preparation but NiCl2 is used as

metal precursor and one additional AW-HT cycle is performed compared to Fe-N-C. It is necessary to note

P a g e | 39

that the metallic nickel is distinct and has a special interaction with carbon species. This is why nickel is

commonly used as a catalyst for nanotube growth.79,80 Taking this into consideration, Ni particles formed

during the synthesis are likely to be further covered and encapsulated by the carbon matrix. This results in

the formation of acid-unsolvable Ni particles, protected by a carbon layers and strong signals in the XRD

from a crystalline phase (Figure S2-2). In Figure S2-3, we measured that the thickness of the dense carbon

layer is over 10 nm, denying the corrosion by the acid-treatments during the synthesis steps, while

simultaneously blocking any catalytic reactivity of such particles as well as the XPS detection (on the

crystalline).

Synthesis of metal free N-C. The synthesis of the metal free PANI (N-C) starts with an identical route to

the Fe-N-C preparation but does not involves any metal precursor. The N-C catalyst is obtained after the

first HT and does not involve any AW steps.

Figure 5- 2 Illustration of M-N-C catalysts. Representative TEM images of a) N-C, b) Fe-N-C and c) Ni-

N-C catalysts. Inserts: HR-TEM images of as-prepared catalysts. Figure S2-3 represents the thickness (over

10 nm) of carbon layers encapsulating the inorganic Nickel species.

Figure 5-2a schematizes the typical various chemical states of carbon-embedded nitrogen atoms as reported

in decade-long past work, including pyridinc, pyrrolic, quaternary, graphitic, oxidized nitrogen, and the

Metal-N4 moiety. A metal-free nitrogen-doped carbon powder catalyst (referred to as “N-C”) was

synthesized as control. Physico-chemical characterization of the degree of crystallinty and crystalline

P a g e | 40

phases was carried using X-Ray Diffraction (XRD). Experimental patterns in Figure S2-2 reflect a largely

amorphous character of the N-C and Fe-N-C catalysts, whereas residual crystalline inorganic Ni species

were detected for the Ni-N-C. The Tranmission Electron Microscopy (TEM) images (Figure 5-2 b-d) and

the bulk compositional analysis by ICP-OS (see Table S2-1) agree with our XRD findings, revealing a

predominantly amorphous character of the N-C and Fe-N-C catalysts, while carbon-encapsulated particles

were visible in the Ni-N-C powder (Figure S2-3). Specific surface areas of the three catalysts were assessed

by N2 physiosorption isotherms (see Figure S2-4). Table S2-1 compares the experiment-derived, calculated

Brunauer-Emmett-Teller (BET) surface areas: Fe-N-C displayed more than 600 m2 g−1, almost 3 times

larger than that of the Ni-N-C and metal free N-C catalysts. A significant rapid N2 uptake at relative

pressures above 0.9 for the Fe-N-C sample (see Figure S2-4a) suggested rich mesoporosity confirmed by

the calculated pore distribution (Figure S2-4b). All these ex-situ surface area tests stand in line with the in-

situ electrochemical double layer capacity measurements used to independently confirm the trend in the

real surface areas among the three carbonous catalysts (see Figure S2-5). Chemical analysis and nitrogen

speciation of the catalyst’s surface was conducted using X-Ray photoeelectronmission spectroscopy (XPS)

using survey scannings (Figure S2-6a) and high resolution analysis (Figure S2-6 b-d). The surface atomic

mole fractions of metal M, nitrogen, oxygen, sulphur, and carbon are given in Table S2-1. It is noteworthy

that even though more than 10wt.% Ni was detected in the bulk, XPS data evidence that the Ni species

show a molar fraction of only 0.38% in the catalyst surface, confirming the succees of the acid leaching.

This also suggests that the residual encapsulated Ni particles that have resisted the repeated acid leaching

procedure, must be densely encapsulated by ≥ 10 nm thick carbon overlayers, indeed as shown by the

experimental microgaph in Figure S2-3. These Ni particles are therefore unlikely, probably completely

unable to participate in the catalytic CO2-to-CO reaction process, apart from the fact that Ni particles due

to their electronic surface structure are unable to reduce CO2 to CO.4 Looking at the high resolution

photoemission spectra, the N1s core level region (Figure S2-6b for Ni-N-C) evidenced the presence of the

porphyrin-like N-coordinated Ni-Nx moieties at 399.7 eV.50,61,81

P a g e | 41

5.2 Liquid-electrolyte H-Cell screening tests

Figure 5- 3 Catalytic performance and product analysis on N-C (black), Fe-N-C (red), Ni-N-C (blue) and

AgOx (Cyan) catalysts. a) Absolute geometric current densities; b) geometric CO production current

densities; c) CO faradaic efficiency as a function of applied iR-corrected electrode potential at 15 min of

each electrolysis (CO partial current densities and faradaic efficiency at 60 min and faradaic CH4 yield are

shown in Figure S2-7, S2-8). d) Geometric CO production current densities and e) CO faradaic efficiency

during the long-term stability testing as a function of stationary electrolysis time. Lines to guide the eye.

Conditions: CO2-saturated 0.1 M KHCO3 (pH 6.8) with 0.75 mg cm-2 catalysts loading.

The catalytic performance of the three Metal Nitrogen co-doped carbon electrocatalysts with respect to the

electrochemical CO2 reduction was first evaluated in a three-electrode, two-compartment liquid-electrolyte

H-Cell equipped with an ion exchange membrane between anode and cathode chamber. The catalyst was

immobilized on a mirror-flat glassy carbon plate with a geometric loading of 0.75 mg cm-2. Additionally, a

commercial AgOx powder (Sigma Aldrich, 223638) catalyst was measured under identical electrochemical

conditions as a state-of-art metallic benchmark. We note that we continue to refer to this benchmark catalyst

as “AgOx” even though we are aware that the surface of this catalyst reduces to metallic Ag under reaction

conditions. The total CO2RR current densities as a function of iR-corrected working electrode potentials

after 15 minutes of stationary constant-potential electrolysis are reported in Figure 5-3a. These curves

compare and contrast the overall catalytic activity of the four catalysts. Clearly, Fe-N-C and Ni-N-C

electrocatalysts exhibited much larger overall current densities over the entire potential range. Online Gas

Chromatography was used to quantify gaseous products and to assess the faradaic product efficiencies.

Without exception, CO was the major detectable CO2RR product, while all residual faradaic charge

contributed to the HER process. Due to stronger chemisorption of CO on the Fe-Nx moieties51 compared to

Ni-N-C, only the Fe-N-C samples catalyzed the consecutive protonation of CO to methane, as evidenced

in Figure S2-8. This makes Fe-N-C one extremely rare example of a non-Cu based CO2RR catalyst with

P a g e | 42

the capability to reduce CO2 into “beyond CO” products such as hydrocarbons. No measurable liquid

products were found in the electrolyte using High Performance Liquid Chromatograph (HPLC) and liquid

GC.

Figure 5-3 b and c plot the partial CO current density and the CO efficiency as a function of iR-free

potential. The Fe-N-C carbon catalyst showed good CO2RR reactivity at lower potentials up to -0.6 VRHE

and approached a efficiency maximum43,51 By contrast, the metal-free N-C catalyst showed only negligible

CO2RR reactivity, again, indicating that the M-Nx moieties played a dominating mechanistic role in the

catalytic reactions. Despite comparable CO efficiencies of the Ni-N-C and the AgOx benchmark catalyst at

electrode potentials up to -1.0 VRHE, the CO production on Ni-N-C reached 12 mA cm-2 at -0.85 VRHE,

which was more than twice that of the AgOx catalyst (7 mA cm-2 at -0.97 VRHE).

To investigate the durability of the catalysts in the H cell configuration, 10-hour electrolysis tests were

carried out using the Fe-N-C, Ni-N-C and AgOx catalysts, each at the electrode potentials where their

respective maximum CO efficiency was observed. Partial CO currents and CO efficiency over time data

are plotted in Figures 5-3 d and e. Both CO partial currents and CO efficiency of the Ni-N-C catalyst showed

only a minor drop, maintaining more than 90% of its initial performance values. This behavior was

comparable to the AgOx catalyst, which, however, displayed a much lower absolute CO yield (see Figure

5-3d). In contrast, the Fe-N-C catalyst started to produce more and more hydrogen until its FE value

stabilized after 4 hours. Its absolute CO yield was lower than that of the AgOx catalyst. We conclude from

the screening tests in our H-Cell that the Ni-N-C catalyst met or exceeded the AgOx benchmark in terms of

CO yield, efficiency and stability.

P a g e | 43

5.3 DFT prediction of CO2RR into CO over various Ni-Nx-C motifs

Figure 5- 4 Free energy diagram of CO2 reduction to CO on Ni-N-C and Fe-N-C catalysts. a) Chemical

structure of the M-Nx moieties considered, b) influence of the Ni-coordination on the binding strength for

the *COOH and *CO intermediates, c) Free energy diagram of CO2 reduction to CO and d) hydrogen

evolution reaction, d) on Fe-N4-C (red), Ni-N4-C (blue) and Ag (111) catalyst (cyan).

To understand the experimental activity-selectivity trends, Density Functional Theory (DFT) simulations

were carried out for the catalytic reaction process on well-defined Fe-N4-C and Ni-N4-C moieties, and were

compared to results obtained with a single crystalline Ag (111) surface. The various metal-nitrogen binding

schemes of the moieties are illustrated in Figure 5-4a. Their corresponding free energies are plotted in

Figure 5-4b. Finally, a comparison of the free energy diagrams for the CO2RR and the competing HER

pathways on the single-site Metal-N4 motifs and the Ag (111) metal facet are displayed in Figure 5-4 c and

Three considerations can help to understand the performance of a catalyst for the CO2RR to CO in aqueous

conditions. First, the initial selectivity can be evaluated by comparing the intermediate binding energy for

*COOH (CO2RR) and *H (HER). Here, Ni-N4-C and Fe-N4-C are quite comparable, while Ag(111) shows

a relatively stronger *H binding, agreeing with previous theoretical observations.15 Second, the desorption

of *CO from the active site has to be fast in order to achieve high rates for CO2RR, therefore *CO binding

has to be weak. Both Ag (111) and Ni-N4-C do not bind *CO (Figure 5-4c), while *CO binds strongly on

P a g e | 44

Fe-N4-C, indicating problematic kinetics for the Fe catalyst, but potentially enabling further reduction15.

Finally, in order to reduce the required overpotential, binding of *COOH needs to be strong, which is the

strongest on the Fe-N-C, intermediate for the Ag(111) and weakest on the Ni-N-C. While we do

acknowledge the influence that the detailed coordination (Figure 5-4a) or hydrogenated environment

(Figure S2-9) could have on the binding properties of the M-N-C, we believe the M-N4 motif and the (111)

metal facet models are appropriate for discussing our experimental observations. In our H-Cell experiments

the Fe-N-C material showed the lowest onset-potential for CO formation, agreeing with the calculated

strong *COOH binding, and a decrease of FECO with ongoing time, which can be indicative of a *CO

correlated poising caused by strong *CO binding. In contrast, the remarkably high and stable partial current

of CO formation on the Ni-N-C catalyst is likely to originate from the readily desorption of CO and

sufficiently weak binding of *H, exacerbating HER. Both, the DFT calculation and initial H-Cell tests

indicate the Ni-N-C motif to be a fine catalyst for the CO2RR to CO.

P a g e | 45

5.4 CO2RR electrolysis using GDE combined MFC

To assess the technological potential of the Ni-N-C catalysts for industrial CO2 co-electrolysis, we then

turned to single two- and three electrode electrolyzer tests in multi-chambers cell set-ups allowing for

pressurized gas flow and circulated electrolyte flows in the cathodic chamber (see Figure 5-5a).

Figure 5- 5 Schematic of a) Ni-N-C Gas Diffusion Electrode (GDE) and b) typical H-type liquid cell. c)

Experimental faradaic CO efficiency as function of the applied electrolyzer current density in CO2 saturated

(50 ccm) 1 M KHCO3 solution. Prolonged flow cell CO2RR testing at 200 mA cm-2 working current density

is presented in Figure S2-10.

Little solubility of CO2 is a commonly problematic for CO2RR in regular CO2-saturated liquid electrolytes,

which cause mass transfer limitation of CO2 reactant (schematic of GDE and H-cell for CO2 / CO transfer

are presented in Figure 5-5 a and b).82 To minimize this drawback, the M-N-C catalysts are immobilized in

a Gas Diffusion Electrode (GDE) to be able to approach industrial current densities of hundreds mini

ampere per geometric electrode area, with a loading of 1 mg cm-2. In principle, the porous nature of the

GDE material enables to create a three-phase interface between gaseous CO2, catalyst layer, and ionic

electrolyte.

The electrochemical testing is operated in a current density range between 50 and 700 mA cm-2. Constant

current density was held for 2 h for the gaseous products quantification. Shown in Figure 5-5c, the

experimentally tested CO faradaic efficiency show distinct dependence on the operated working current

densities. Remarkably, the Ni-N-C catalyst GDE reached a maximum FECO of nearly 90% between 100 to

250 mA cm-2, delivering an over 200 mA cm-2 CO evolution activity. Furthermore, such impressive

catalytic performance could be maintained for over 20 hours at 200 mA cm-2 current density (see Figure

S2-10), clearly outperforming the commercial AgOx control sample.

However, neither the metal-free N-C catalyst nor the Fe-doped one show efficient behavior under the

identical operating conditions, which perform parallel with our observation referred to the liquid electrolyte

H-Cell tests presented in Figure 5-3. Both catalysts contribute reasonable faradaic efficiency towards CO,

but remain in limited working potential of positive than -0.6 VRHE. Predicted by DFT simulation, the Fe-N-

C, who have strong interaction with CO, probably poses a CO poisoning. The inferior CO2RR activity over

the metal-free one could be attributed to the absence of active M-Nx functionalities.

P a g e | 46

5.5 Discussion

In comparison to the regular liquid H-cell, the flow cell coupled with GDEs allows to create 3 phase catalyst

interfaces, minimizing mass transfer limitation of the CO2 reactant. Moreover, a fast CO desorption owing

to large interfacial CO2 could be proposed. Referred to the electrochemical cell design, we suggest utilize

the GDE combined MFCs for CO2 electrolysis, which could realize its commercialization.

Referring to the catalyst benchmark, AgOx, which reduces to catalytically active metallic Ag at the surface

under reaction conditions, is the most commonly deployed catalyst material for alkaline CO2 to CO

electrode. Schmid and co-workers utilized commercial silver-based gas diffusion electrodes and reported

FECO values of up to 90 % (50 to 300 mA cm-2), consistent to data presented in the Summary table 5-1

below.83 Kenis and co-workers have used a Ag catalyst ink to fabricate GDEs at high catalyst loadings. In

their investigations, they achieved 87 % of FECO at -0.91 V vs. RHE, however, the total current density of

50 mA cm-2 remained relatively low.84 In direct comparison to earlier studies, the Ni-N-C CO2RR catalysts

presented here meet or exceed previous reports of Ag-based catalysts in terms of faradaic CO yield. This

demonstrates the great potential of the family of non-metallic M-N-C carbon catalysts, in particular Ni-N-

C, for replacing expensive precious group metal catalysts as the benchmark in CO2 to CO electrolyzers.

Table 5- 1 Summary table of the catalytic performance towards CO2RR referred to Gas Diffusion Electrode.

Catalysts

Electrolyte

Reported condition

Potential (VRHE)

Product efficiency

Reference

NOBLE METAL

REFERENCE:

Ag-based GDE

Cat. Loading:

Not given

1.5M

KHCO3/ 0.1

M K2SO4

300 mA cm-2

Working current

FECO: ~80 %

jCO: 240 mA cm-2

83Schmid

NOBLE METAL

REFERENCE:

CD-Ag/PTFE

Cat. Loading:

Not given

1 M KHCO3

Over

150 mA cm-2

Working current

-1.2

FECO: 90 %

85 Sargent

1 M KOH

-0.8

FECO: 90 %

M-N-C Candidates:

Ni-GS

Cat loading:

0.2 mg cm-2

0.5 M

KHCO3

-0.75 V vs. RHE

Working current

below 50 mA cm-2

-0.75

FECO: 90%

jGeo: 60 A g-1

60 Jiang et

al.

M-N-C Candidates:

Ni-PANI

Cat loading:

1 mg cm-2

1 M KHCO3

50 to 700 mA cm-2

Working current

-0.9

CO: ~85%

jCO: > 200 mA

cm-2

(> 200 A g-1)

This work

P a g e | 47

Chapter 6. Tuning the active site density of poly anlinie derived Fe-N-C

catalyst using a secondary Nitrogen precursor

Most parts of this chapter are reproduced from the journal article:

The chemical identity, state and structure of catalytically active centers during the electrochemical CO2

reduction on porous Fe–nitrogen–carbon (Fe–N–C) materials

Link: https://doi.org/10.1039/C8SC00491A

with permission from Chemical Science, 2018, 9, 5064-5073, as Reference 47 in this dissertation. Copyright

2018 Royal Society of Chemistry (CC BY 3.0).

Author list: Nathaniel Leonard, Wen Ju (co-first), Ilya Sinev, Julian Steinberg, Fang Luo, Ana Sofia Varela,

Beatriz Roldan Cuenya, and Peter Strasser

Contributions: N.L., W.J. and I.S. designed the experiments, synthesized and characterized (regularly) the

catalysts. I.S. and B.V. carried out the advanced characterization as XPS, ex-situ / operando XAS and

subsequent in-depth data analysis. W.J. carried out the electrocatalytic tests and analyzed the results. N.L,

W.J. and P.S. aggregated the figures and co-wrote the manuscript. All authors discussed the results, drew

conclusions and commented on the manuscript.

P a g e | 48

Rather than the Ni-N-C catalysts, which could industrial level CO2 electrolysis, the Fe-N-C based one allows to

selectively reduce CO2 into CO in a lower potential range. In this scenario, we locate our attention on Fe- based

M-N-C catalysts and report novel findings. Here, the structure–activity relationships, the chemical state and fine

structure of catalytically active sites under operando conditions during the electrochemical CO2 reduction reaction

(CO2RR) catalyzed by a series of porous iron–nitrogen–carbon (Fe-N-C) catalysts are investigated.

The Fe-N-C catalysts were synthesized from different nitrogen precursors and, as a result of this, exhibited quite

distinct physical properties, such as BET surface areas and distinct chemical N-functionalities in varying ratios.

The chemical diversity of the Fe-N-C catalysts was harnessed to set up correlations between the catalytic CO2RR

activity and their chemical nitrogen-functionalities, which provided a deeper understanding between catalyst

chemistry and function. XPS measurements revealed a dominant role of porphyrin-like Fe–Nx motifs and

pyridinic nitrogen species in catalyzing the overall reaction process. Operando EXAFS measurements revealed

an unexpected change in the Fe oxidation state and associated coordination from Fe2+ to Fe1+. This redox change

coincides with the onset of catalytic CH4 production around −0.9 VRHE. The ability of the solid state coordinative

Fe1+–Nx moiety to form hydrocarbons from CO2 is remarkable, as it represents the solid-state analogue to

molecular Fe1+ coordination compounds with the same catalytic capability under homogeneous catalytic

environments. This finding highlights a conceptual bridge between heterogeneous and homogenous catalysis and

contributes significantly to our fundamental understanding of the Fe-N-C catalyst function in the CO2RR. 47

P a g e | 49

6.1 PANI based Fe-N-C catalysts synthesis using a 2nd Nitrogen precursor

In this contribution, the synthesis of the carbon precursor is identical with that presented in Chapter 5, Ketjen

EC 600JD (AzkoNobel) was stirred in 0.5 M HCl for 24 hours and vacuum filtered with DI water till

completely neutralized. The washed and dried carbon was refluxed in HNO3 for 8 hours at 90 °C and again

vacuum filtered with DI water to neutral pH.

Similar synthesis approach as Chapter 5 is proceed, and 3 ml of Aniline are added into 0.5 liter of 1 M HCl

along with 5 g FeCl3 and 5 g Ammonium Persulfate. Specific behavior was done at this point. A secondary

nitrogen precursor was added for the purpose to tune the C-Nx cavities amount, to further increase the Fe-Nx

site density. For the control catalyst (referred as CTRL), no secondary nitrogen precursor was added during

the synthesis steps. The quantity of the secondary precursor is calculated to add 0.333 moles of nitrogen.

Various secondary nitrogen precursors (presented in Table 6-1) were chosen to represent common nitrogen

precursors86-93 with varying size and nitrogen contents as summarized in Table 6-1 in the supplemental

information. This resulted in 7 g melamine (MEL) or cyanimide (CM), 10 g Urea (UREA), or 23.6 g Nicarbazin

(NCB). After one hour of stirring, 0.4 g of pretreated carbon was added. This pretreated carbon has been

ultrasonically dispersed in 50 ml of DI water. The resulting mixture was stirred for 48 hours and then dried.

After drying, the mixture was ball-milled and heat treated with a ramp of 30°C per minute to 900 °C and kept

at this temperature for one hour in a nitrogen atmosphere. After heat treatment, the material was refluxed in 2

M H2SO4 overnight and rinsed to neutral via vacuum filtration. After this acid wash, a second identical heat

treatment was performed. At least a second acid wash and third heat treatment were performed on each sample.

After this third heat treatment, XRD was used to determine whether the sample had been cleaned of excess

residual Fe (usually in the form of FeS). If the sample is not clean, a third acid wash and fourth heat treatments

were performed (this was the case for CTRL and NCB). These materials have also been explored as oxygen

reduction catalysts in a related paper.

Table 6- 1 Characteristics of Secondary Nitrogen Precursors Used in this Work

Nitrogen Precursor

Formula

MW / g mol-1

N/C

Cyanamide

CH2N2

2/1

Melamine

C3H6N6

126

6/3

Urea

CH4N2O

2/1

Nicarbazin

C19H18N6O6

426

6/19

P a g e | 50

6.2 Physiochemical Characterization

Figure 6- 1 Comparison of Fe-N-C catalysts based on different secondary nitrogen precursors showing bulk

iron, nitrogen, and sulfur content as measured by ICP and Elemental Analysis and surface content (ca. 2-3 nm)

as measured by XPS. Catalysts ordered by increasing surface nitrogen content (XPS).

This range of PANI-derived Fe-N-C catalysts were synthesized by varying the secondary nitrogen precursor

while keeping the iron, carbon, and primary nitrogen precursor (polyaniline, PANI) the same. The secondary

nitrogen precursors investigated were melamine (MEL), cyanimide (CM), urea (UREA), and nicarbazin

(NCB). These secondary nitrogen precursors cause slight differences in chemistry and morphology. The basic

chemical compositions can be seen in Figure 6-1, with a comparison of bulk (EA and ICP) and surface (XPS)

measurements. By comparing the bulk (open) and surface (hashed) bars, the surface chemistry can be

contrasted with that of the bulk. In general, catalysts tend to be surface sparse in Fe, N, and S. This suggests

that the surface is rich in carbon. The only exceptions to this observation are the sulfur content of MEL, UREA,

and CM. This difference suggests significant sulfur surface functionalization for these catalysts. It is also

interesting that these catalysts also contain the highest surface nitrogen contents. For the other two catalysts

(NCB and CTRL), the large discrepancies between bulk and surface iron and sulfur contents are ascribed to

FeS that can be detected even after an additional acid wash by XRD (Figure S3-1). These particles cannot be

easily washed because they lie below the catalyst surface. The small, bulk/surface discrepancy regarding the

nitrogen content may also be attributed to similar iron nitride particles below the catalyst surface. SEM images

of MEL and CTRL are shown in Figure S3-2, indicating the similarity of the catalyst morphology. Ex-situ

BET measurements and in-situ double layer capacitance of this series catalysts are carried out, and the both

techniques nicely agree with each other (Figure S3-3 to S3-5). Further, we performed the CO chemisorption

P a g e | 51

to quantify Fe-Nx site density in the as-prepared Fe-N-C catalysts, and the data stands in line with the XPS

detection (Figure S3-6) according to eq. S3-1 and eq. S3-2. the Analysis of S 2p spectra (Figure S3-7, table

S3-1) however shows three doublets at 164.1, 166.5 and 167.7 eV (for 2p3/2), which can be assigned to thiol,94

sulfoxide,95 and sulfone96 species correspondingly, indicating thus exclusive presence of organic sulfur on the

surface formed during acid washing.

Figure 6- 2 (a) High resolution N 1s XPS data of various PANI-derived Fe-N-C samples with different N

precursors. (b) Example of the deconvolution of a N 1s spectrum acquired for the MEL sample. N1s assignment

of Fe-PP ref. sample is presented in Figure S3-8, showing identical BE (399.8eV) as the Nx-Fe moiety in MEL

sample.

Table 6- 2 Distribution of nitrogen species (in at%) in PANI samples as seen from N 1s XPS spectra

deconvolution.

sample

Pyridinic

Nx-Fe

pyrrolic

graphitic

N-Ox

NCB

25.0

12.8

44.1

13.0

5.1

CTRL

28.4

12.6

45.9

10.5

2.6

33.8

12.7

42.1

8.6

2.8

MEL

31.8

17.7

36.9

10.1

3.5

UREA

32.7

12.4

40.3

10.7

3.9

Fe-PP(Sigma-Aldrich)

100

Nitrogen 1s XPS region scans are shown in Figure 6-2a. All samples show a similar structure with two

dominating peaks at 398.7 and 401.3 eV, indicating prevalence of pyridinic and pyrrolic nitrogen in the

structure. A spectral valley between those peaks, where Nx-Fe species are reported,81 is shallower in the PANI-

406 404 402 400 398 396 406 404 402 400 398 396

pyrrolic

graphitic

N-Ox

Nx-Fe

UREA

MEL

CTRL

Intensity / arb. units

BE / eV

NCB

pyridinic b)

BE / eV

pyrrolic

graphitic

N-Ox

Nx-Fe

pyridinic

P a g e | 52

MEL sample, pointing to a higher concentration of Fe-porphyrin moieties in those samples. Indeed, a more

detailed analysis of the N 1s regions, exemplified by the PANI-MEL sample in Figure 6-2b is summarized in

Table 6-2. The N1s spectra of Fe-Protoporphyrin from Sigma-Aldrich is shown in Figure S3-8. Fe 2p spectra

of all PANI samples show a similar structure with Fe 2p3/2 having an intense peak centered at 711 eV and a

weak satellite observed around 715.5 eV (Figure S3-9). The structure detected is similar to the shape of Fe

2p3/2 previously reported for ferrous oxide (FeO).97,98 It is noteworthy that there are no hints of Fe-N moieties

reported at 708 eV (Figure S3-10).74 Altogether, the analysis of the Fe 2p3/2 spectra indicates that the most iron

seen by XPS is oxidize state Fe(II), while Fe-Nx species, probed indirectly in N 1s spectra (Figure 6-2b and

Figure S3-9), must be lying either in the deeper layers or in pores, not accessible to XPS at Fe 2p due to the

lower value of the corresponding inelastic mean free path of photoelectrons.99

As discussed previously, the discrepancy between surface and bulk iron contents (detected by ICP and XPS as

shown in Figure 6-1) indicates the presence of iron species not seen by surface sensitive techniques, e.g.

covered by a carbon layer or isolated in pores of the support. To investigate the nature of those species, X-ray

absorption spectroscopy (XAFS) measurements were carried out ex situ. X-ray absorption near edge structure

(XANES) spectra of selected samples (Figure 6-3a) indicate similar chemical state and coordination of iron.

The spectra show a pre-edge feature at ca. 7114 eV, corresponding to a 1s → 3d electronic transition typical

for Fe3+ in an octahedral local environment. An intense feature above the absorption edge, between 7126 and

7159 V (so-called white line) has however no similarities with the most common iron oxides (Figure S3-11

for comparison), but is well in line with the results published for similar materials.42,100 The EXAFS spectra

plotted on Figure 6-3b, despite looking somewhat alike, have distinct differences in both, peak positions and

intensities. The first peak, originating from a light backscatterer, is observed at 1.46 Å (uncorrected for a phase

shift) in the PANI-MEL sample and shifts towards shorter distances in PANI and PANI-CM, 1.42 and 1.39 Å

(uncorrected) correspondingly. At the same time, the peak intensity is similar in PANI-MEL and PANI-CM

samples, while the PANI sample shows a considerably smaller peak. The second backscattering feature

between ca. 2.0 and 3.0 Å (uncorrected) is worth special attention. Its location is somehow similar to Fe-Fe

backscattering in both common iron oxides with bcc structure (Figure S3-11b), although neither matches

exactly in peak position. Zitolo et al assigned a similar structure to crystalline Fe2N, formed during pyrolysis

in NH3.42 In our case, formation of nitrides was not observed by any other method. To obtain further details on

the local Fe environment, the EXAFS spectra were fitted using an Fe-porphyrin structure101 as model and the

results are summarized in Table 6-3. It is seen that the first coordination shell around Fe can be well described

by the FeN4 moiety. Slight deviations from 4-fold coordination are explained by the interference with iron

oxide species which were detected by our surface sensitive XPS method and should be present in the as-

prepared ex situ measured samples. The second next neighbor peak observed at ca. 2.4-2.5 Å (uncorrected) is

indeed well described by carbon from a porphyrin structure with the real bond distance close to the reference

of 3.0 Å. The corresponding coordination number however is significantly lower than 8 in crystalline

porphyrin, indicating a highly disordered structure. The latter is also supported by the substantially higher

Debye-Waller factors obtained for the PANI samples as compared to the reference iron protoporphyrin sample.

P a g e | 53

Figure 6- 3 a) Fe K-edge XANES and b) EXAFS spectra of selected Fe-PANI samples, dotted lines in b) show

fitted models. XANES and EXAFS spectra of referenced FeOx, Fe foil and FePP (Sigma-Aldrich) are shown

in Figure S3-11. Fe K-edge k2-weighted EXAFS data of Fe-PANI samples in k-space and analysis as

exemplified by CM sample are shown in Figure S3-12, S3-13.

Table 6- 3 Best-fit parameters for the Fe K-edge EXAFS spectra of the Fe-PANI samples shown in Figure 6-

3. Included are the coordination numbers (CN) for Fe-N and Fe-C species, and the bond lengths for the same

species (r) and Debye-Waller factor (σ2). The values in parenthesis are the standard errors in the last digit.

CNFe-N

rFe-N, Å

σ2Fe-N∙10-3, Å2

CNFe-C

rFe-C, Å

σ2Fe-C∙10-3, Å2

3.8(3)

1.98(1)

7.5(6)

1.3(2)

3.0(1)

10.3(9)

CTRL

3.5(3)

1.99(1)

8.5(7)

1.3(2)

2.8(1)

9.2(8)

MEL

3.8(2)

2.01(1)

7.3(6)

0.8(1)

2.7(2)

9.8(9)

Fe-PP (Sigma-Aldrich)

4.0

1.93(1)

4.7(6)

8.0

3.2(1)

4.6(9)

P a g e | 54

6.3 Correlating CO2RR performance and physical properties

Now that the catalysts have been described structurally and chemically, this information can be used to better

understand the catalysts electrochemical performance. Figure 6-4 shows the CO production rate (a) and CO

faradaic efficiency (b) of the various catalysts. From the CO production rate a kinetic region can be identified

by the strong increase in performance with decreasing potential between -0.45 and -0.6 V vs RHE. At lower

potentials this kinetic region gives way to a plateau with maximum production rates of over 5 mA cm-2 for the

melamine and cyanamide catalysts. The lack of potential dependence indicated by this plateau suggests that

the rate limiting step has shifted to some non-electrochemical process. The faradaic efficiency towards CO

production, shown in Figure 6-4b, shows peak faradaic efficiencies occurring around -0.6 V vs RHE with the

top performing catalysts being Melamine with 85 % maximum CO efficiency.

Figure 6- 4 CO2 reduction data for various Fe-N-C catalysts based on different secondary nitrogen precursors:

(a) CO generation rate, (b) faradaic efficiency towards CO production. Experimental conditions: CO2 saturated

0.1 M KHCO3, catalyst loading: 0.75 mg cm-2 on Glassy Carbon.

Comparing Figure 6-4 with the chemical characterizations, it is evident that the addition of the secondary

nitrogen precursor has impacted both chemistry and catalyst performance. In order to understand this

P a g e | 55

performance better, surface area effects must be understood. There is a strong correlation between BET surface

area and performance as shown in Figure 6-5. This relationship is not surprising considering the fundamental

role of real surface area in heterogeneous catalysis. To reach a deeper understanding of catalysts behavior, the

current can be normalized to the real surface area. This will allow a comparison of specific current densities

with surface chemistries obtained from XPS results.

Figure 6- 5 Trends of CO current density on the plateau and in the kinetic region (-0.53 V vs RHE) varying

with BET specific surface area.

For a better understanding of the intrinsic catalytic activity of these Fe-N-C catalysts for CO2-to-CO

conversion, specific current densities were calculated by normalizing the current in the kinetic region by the

BET surface area, subsequently correlated to the chemical make-up of the surface as measured by XPS. The

data show particularly good correlations with N-Fe species (black squares) and pyridinic nitrogen species

(open circles) as shown in Figure 6-6a. The N-Fe correlation suggests potentially active Fe-Nx sites, and the

pyridinic nitrogen trend is in accordance with its catalytic properties hypothesized by Wu et al. (both mentioned

in the introduction).43,46 Comparing the N-Fe and N-pyridinic correlations, it can be observed that some

catalysts have relatively more N-Fe (MEL and CTRL) and some relatively more N-Pyridinic (UREA and CM).

This observation leads to the hypothesis that both constituents contribute to the catalytic activity. For this

reason, the authors also include a correlation of the specific current density with the sum of N-pyridinic and

P a g e | 56

N-Fe content (black triangles). The linear fit of the summed data set shows a higher R2 than either of the other

fits. This fit improvement suggests that both sites are likely active.

Figure 6- 6 Trends of CO current densities in the kinetic region (-0.53 V vs RHE) varying with (a) various

pyridinic nitrogen and N-Fe and (b) surface Fe content. Current density is normalized to the specific surface

area as calculated by the BET method. BET-normalized CO current densities in the kinetic region as a function

of other functionalities are shown in Figure S3-14 and Figure S3-15. Free Energy Diagrams from CO2 to CO

over FeNx and Pyridinic-N sites are shown in Figure S3-16, data are adapted from Refs.45,51

In addition to the hypothetical nitrogen active sites, literature results have also suggested that trapped iron

content near the surface may be an active site for this type of catalysts.44,102,103 For comparison, Figure 6-6b

shows the correlation of specific activity with total iron from XPS. The iron has a poorer correlation than either

of the nitrogen constituents, suggesting that this data does not support the hypothesis that encapsulated Fe is

an active site for CO2RR. This uncertainty is compounded by the fact that the Fe peak is small and hard to

quantify from XPS and that the samples contained not only Fe-N species, but also FeOx species at/near the

surface. The alternative theory that the metallic Fe content is a H2-generation site is also hard to prove.44,103 In

this case, the melamine based catalyst would be expected to show the highest H2 faradaic efficiencies (lower

CO efficiencies), but this is certainly not the case. In fact, MEL has the highest CO efficiency even though it

also has the highest surface metal content.

0 1 2

0.0 0.1 0.2 0.3

N-Fe R2 = 0.75

N-Pyridinic R2 = 0.73

N-Fe+N-Pyr R2 = 0.95

CO Current Density / mA m-2

BET

Nitrogen via XPS / at. %

UREA

MEL

CTRL

NCB

CO Current Density / mA m-2

BET

Iron via XPS / at. %

UREA

MEL

CTRL

NCB

R2=0.67

P a g e | 57

6.4 Operando X-ray absorption spectra under CO2RR condition

Figure 6- 7 Fe K-edge XANES (a) and EXAFS (b) spectra taken under operando conditions in CO2-saturated

0.1 M KHCO3 at -0.5 V (solid blue curves), -0.9 V (dashed green curves) and -1.1 V (red dot-dashed curves)

vs. RHE a – XANES. (c) CH4 faradaic efficiency from CO2RR varying with applied working potential. Lines

are added to indicate points representative of spectra in (a) and (b).

Table 6- 4 The best-fit parameters for Fe K-edge EXAFS spectra of the Fe-PANI measured under operando

conditions are shown in Figure 6-7. Included are the coordination numbers (CN) for Fe-N and Fe-C species,

and the bond lengths for the same species (r) and Debye-Waller factor (σ2). The values in parenthesis are the

standard errors in the last digit.

Potential

CNFe-N

rFe-N / Å

σ2Fe-N∙10-3, Å2

CNFe-C

rFe-C / Å

σ2Fe-C∙10-3, Å2

-0.5VRHE

4.3(8)

2.00(2)

6.8(3)

2.6(8)

3.0(2)

8.5(7)

-0.9VRHE

4.2(8)

2.00(2)

7.0(4)

2.4(5)

3.0(2)

8.7(7)

-1.1VRHE

3.9(9)

2.00(3)

7.2(4)

1.8(4)

3.1(2)

9.2(8)

P a g e | 58

Operando XAFS data were collected to elucidate possible changes in the oxidation state and local coordination

of Fe. Scheme of operando-XAS cell is presented in Figure 3-1. Fe K-edge XANES spectra taken under

reaction conditions in CO2 saturated KHCO3 shown in Figure 6-7a display a shift in edge position at -0.9 V

vs. RHE (red trace). This shift is similar to that found on related catalysts at around 0.7 V vs. RHE under acidic

conditions.104 That shift was connected to 2+/3+ active site redox behavior which had important implications

concerning adsorbate bond strength.104 Similarly, it is likely that the shift observed in the operando XANES

data is correlated to a 1+/2+ redox transition. This redox behavior has been observed for various iron-based

macrocycles at similar potentials.35,105-108 These changes in active site oxidation state and coordination are also

supported by changes observed in EXAFS as seen in Figure 6-7b. With decreasing potential, the corresponding

spectrum shows a slight decrease of both backscattering features at 1.44 Å (uncorrected), assigned to N/O, and

shoulder at 2.35 Å (uncorrected), previously shown to correspond to Fe-C. Thus, the Fe-N/O coordination

number decreases from 4.3 to 3.9, while the Fe-C coordination number decreases from 2.6 to 1.8 (see Tables

6-3 and 6-4 for details). The changes observed can be assigned to the reduction of surface iron oxides, detected

by XPS, and increasing disorder in the material under reaction conditions.

The redox transition could have a significant impact on binding energies and reaction mechanisms. One of the

interesting mechanistic questions concerning these catalysts is the role and prevalence of CO poisoning. It has

been suspected that strong CO binding is one of the inhibitors of higher CO2RR performance as well as the

cause of CH4 production. For that reason, the CO2 consumption as a function of potential was considered in

the hope of seeing significant changes in catalyst behavior. This was accomplished by calculating the ratio of

CH4 production to total CO2 consumed as shown in Figure 6-7c. Observable in this figure is a strong increase

in CH4 production between -0.9 and -1.1 V vs. RHE. The correspondence of this increase with the changes in

coordination and redox behavior indicated by EXAFS suggests that the binding behavior of reactants may

indeed be modified during this potential change. This observation is consistent with a proposed mechanism

for photochemical methane formation on a potentially similar iron-based macrocycle catalyst.107 In the

proposed mechanism the Fe 1+/2+ transition plays an integral role in the conversion of CO into CH4 via a

formyl intermediate.107 This theory is consistent with our observations of increased CH4 production, changed

oxidation state, and decreased coordination. All together, we hypothesize that these FeNx sites would provide

enough room and proper binding energy for proton binding, facilitating the intermediate CHO*, which might

be the rate-limiting step for CH4 formation. Despite the lack of systematic theoretical simulations, this

mechanistic change has significant implications for future catalysts development. Specifically, for actives sites

that are nitrogen-coordinated iron complexes, it is possible to adjust the iron center redox potential by adjusting

ligand number/strength. This adjustment of iron center redox behavior or coordination number could be used

to synthesize catalysts with higher CH4 yields. Conversely, it may be possible to inhibit CO-poisoning by

tuning iron complexes to have lower Fe 1+/2+ redox potentials. Unfortunately, such a study is difficult on the

present set of catalysts and would require a following work that looked at CO2RR on a set of Fe-macrocycles

with varying metal center electron density of states. Such a study is outside the scope of this work.

P a g e | 59

6.5 Discussion

In the present work five different FeNC CO2RR catalysts have been explored with two goals: 1. synthesizing

high performance, inexpensive CO2 reduction catalysts for aqueous media, 2. increasing our fundamental

understanding of the active state and structure of FeNC catalysts during the CO2RR process. Towards the first

goal, the melamine based Fe-PANI catalyst achieved a CO efficiency of 85% and a two-fold improvement in

CO production rate resulting in current densities of over 5 mA cm-2.

Towards our second goal of increasing understanding of M-N-C catalysts we have three additional conclusions.

Firstly, high specific surface areas are important to catalytic activity. This suggests that reaction rates are

limited by adsorption and/or kinetics (i.e. surface events). Secondly, the comparison of specific current density

with XPS data on the surface indicates that both N-Fe and pyridinic species are likely actives sites. Finally,

operando EXAFS results indicate a reduction in metallic content between -0.9 and -1.1 V vs. RHE which

corresponds with the redox potential of Fe 1+/2+. This event coincides with an increase in CH4 production,

which suggests a change in active site behavior. The authors hypothesize that this change in behavior is a

change in reaction mechanism that results in the onset of CH4 production.

P a g e | 60

Chapter 7. Unraveling the mechanistic insight of Electrochemical CO2

Reduction to Methane on the Fe-N-C Catalyst

Most parts of this chapter are reproduced from the journal article:

Unraveling Mechanistic Reaction Pathways of the Electrochemical CO2 Reduction on Fe-N-C Single Site

Catalysts

Link: https://doi.org/10.1021/acsenergylett.9b01049

with permission from ACS Energy Letters, 2019, , 5064-5073, as Reference 109 in this dissertation. Copyright

2019 American Chemical Society.

Author list: Wen Ju, Alexander Bagger, Xingli Wang, Yulin Tsai, Fang Luo, Tim Möller, Huan Wang, Ana

Sofia Varela, Jan Rossmeisl, and Peter Strasser

Contributions: W.J. and A.B. designed and combined the experiments and simulations. W.J. and A.B. co-wrote

the article. Other co-authors contributed characterization data and analysis.

P a g e | 61

In this work, we report an experimental-computational study of mechanistic reaction pathways during the

electrochemical reduction of CO2 to CH4, catalyzed by solid-state, single-site Fe-N-C catalysts. Fe-N-C

catalysts feature molecularly dispersed catalytically active Fe-N motifs and represent a type of non-Cu-based

catalysts that yield “beyond CO” hydrocarbon products. The various multi-step mechanistic pathways toward

hydrocarbons with these catalysts has never been studied before and is the focus of this study. A number of

different reactant molecules with varying formal carbon redox states, more specifically CO2, CO, CH2O,

CH3OH and formate were electrochemically converted at the Fe-N sites, yet only CO2, CO and CH2O could

be protonated into methane. Also, we observed a distinctly different pH dependence of the catalytic CH4

evolution from CO and CH2O, suggesting differences in the proton participation of rate determining steps. In

comparing the experimental observations with Density Functional Theory (DFT) -derived Free Energy

Diagrams of reactive intermediates along the reaction coordinates, we unraveled the distinctly different

dominant mechanistic pathways and roles of CO and CH2O along the catalytic CO2-to-CH4 cascade and their

rate-determine-steps (RDS). We close with the first comprehensive reaction network of the CO2

electroreduction on a M-N-C catalyst. Our findings offer valuable insights in the catalysis of the CO2RR on

single site Fe-N-C catalysts that may prove useful in developing efficient, non-Cu-based catalysts for direct

electrochemical hydrocarbons production. 109

P a g e | 62

7.1 Catalysts preparation and regular characterization

The poly-aniline-derived single-site Fe-N-C catalysts with atomically dispersed Fe-Nx moieties employed here

are identical to those described and characterized in our previous work.41,47 To ensure the role of the atomically

dispersed Fe-Nx motifs, an iron-free, yet otherwise identical N-C catalyst was prepared as used as a control.

Conventional ex-situ characterization of physico-chemical properties of the Fe-N-C material is shown in

Appendix 4 (Table S4-1, Figure S4-1, S4-2, and S4-3). X-ray Diffraction patterns (XRD, Figure S4-1) and

Transmission Electron Microscopy (TEM, Figure S4-2) indicated that the Fe-N-C catalyst had similar

amorphous carbon structure as the metal-free N-C one. N2 specific adsorption and double layer capacity

measurements were carried out, that suggested comparable surface areas (Figure S4-3) of Fe-N-C and N-C.

Elemental analysis revealed the same or similar nitrogen content in the two catalysts, and absence of Fe in the

control (ICP). CO sorption confirmed absence of CO adsorbing centers in the control (Table S4-1). Moreover,

the dispersed single site Fe-Nx motifs were identified and confirmed using N1s X-ray photo electron emission

spectroscopy,43,47 Mossbauer spectroscopy,41 and also X-rays absorption spectroscopy47 in our previous

approaches (Chapter 5 and Appendix 2). All these confirm the dominant character of the coordinative Fe-Nx

motifs in our Fe-N-C model catalyst, whereas the inorganic iron species remain in trace portion, catalytically

not responsible for methane evolution.4

P a g e | 63

7.2 Products spectrum of prolonged CO2RR and CORR on Fe-N-C catalyst

Upon electrochemical reduction of CO2 using CO2-saturated electrolytes in an H-cell, CO, H2 and CH4 were

identified as the main CO2RR products over the Fe-N-C catalyst.47,48 The overall faradaic efficiency reached

95% during bulk electrolysis tests at constant electrode potentials. Liquid products such as alcohols, aldehydes

and formate were below the detection limit for the typical present electrolysis time of 75 min. However, we

also performed a few longer-term, 1000 min CO2RR experiments (see Table S4-2). Similarly, prolonged CO

reduction reaction (CORR) was carried out in buffered potassium phosphate solution for 480 min at the same

pH. Now, very small, yet clearly detectable amounts of both methanol and formaldehyde were found during

these longer term electrolysis tests (see Figure S4-4 and S4-5), demonstrating that these two compounds are

apparently involved in the mechanistic network of the CO2RR as well as of the CORR over the Fe-N-C catalyst.

P a g e | 64

7.3 Electrochemical reduction of a set of different COxHy molecules

To get insight in the electrochemical CO2-to-CH4 reaction pathways, CO2, CO, CH2O, CH3OH and formate

were used separately as feed reactants in order to investigate their relative reaction rates and resulting product

spectra. The choice of these feeds was based on the fact that all of them may constitute reactive intermediates

of the CO2RR process. In particular, we were interested whether and to what degree these potential reactive

intermediates can be electrochemically reduced to CH4 on the Fe-N-C catalysts, which carries useful insight

in the catalytic CO2-to-CH4 reaction cascade. Table S4-2 lists the reaction parameters used in the experiments.

The formal chemical transformations and their standard potentials read:

CO2+ 8(H++ e−) → CH4 + 2H2O E0= +0.17 VRHE eq. 7‐ 1

HCOOH + 6(H++ e−) → CH4 + 2H2O E0= +0.28 VRHE eq. 7‐ 2

CO + 6(H++ e−) → CH4 + H2O E0= +0.26 VRHE eq. 7‐ 3

CH2O + 4(H++ e−) → CH4 + H2O E0= +0.45 VRHE eq. 7‐ 4

CH3OH + 2(H++ e−) → CH4 + H2O E0= +0.63 VRHE eq. 7‐ 5

Figure 7- 1 Overall electrochemical performance under different conditions and in presence of various

reactants in neutral 0.05M K3PO4 + 0.05M H3PO4 solution. a) Linear sweep voltammetry at -5 mV s-1 potential

scan rate and b) geometric current density during each bulk electrolysis. Presented dots data are averages

calculated from 15 min, 45 min and 75 min of the stationary electrochemical reaction. Line to guide the eye.

Catalyst loading: 0.75 mg cm-2 on glassy carbon.

The overall electrocatalytic polarization behavior comprising all cathodic processes were studied for each

individual feed molecule using transient linear sweep voltammetry (LSV) (Figure 7-1a) and stationary bulk

electrolysis (Figure 7-1b). Reactivity trends of transient and stationary measurements matched well. CO2

outperformed the HER control activity (N2 as feed in Figure 7-1a), largely thanks to significant CO evolution

P a g e | 65

(Figure S4-6) below -0.4 VRHE. CO2 was followed by formic acid, while the other reactants exhibited faradaic

currents comparable to the HER background. CO feeds, likely due to its site blocking nature, displayed lower

currents than the background (Figure 7-1a). Liquid products remained below the detection limit. Methane

formation, however, was observed for CO2, CO and CH2O feeds (Figure 7-2) and became the primary focus

of subsequent kinetic studies.

Figure 7- 2 Methane a) production rate and b) faradaic efficiency as a function of applied IR-free potential

during the electrochemical CO2 (saturated, 30 mM), CO (saturated, 1 mM) and CH2O (1 mM) reduction

reactions on Fe-N-C (solid dots) and metal free N-C (empty) catalysts in neutral 0.05 M K2HPO4 + 0.05 M

KH2PO4 buffer solution. Data points are averages obtained from 15 min, 45 min and 75 min of each bulk

electrolysis. Line to guide the eye. Catalyst loading: 0.75 mg cm-2 on glassy carbon plate.

First, we analyzed how the CH4 formation rate changed with applied potential. Figure 7-2 shows the potential-

dependent electrocatalytic methane production rate and the faradaic methane efficiency for each feed on Fe-

N-C catalysts (solid symbols) and Fe-free “N-C” control (open symbols) catalysts. The Fe-free catalysts

displayed little to none CH4 yield, evidencing the catalytic role of the Fe-Nx single sites in the CO protonation

process. Closer inspection of Figure 7-2 reveals more anodic CH4 onset potentials under CH2O feeds compared

to CO or CO2 feeds. This suggests that CH2O activation and subsequent protonation to methane is fast, while

activation barriers in the catalytic cascade from CO2 or CO to CH2O appeared to delay their reduction kinetics.

Indeed, the fact that the CH4 yields under CO2 and CO feeds track each other so closely is an indication that

they are kinetically limited by a shared elementary process, more specifically the protonation of adsorbed CO,

*CO. Note that we will abbreviate surface adsorbed species with an asterisk on the left to symbolize a surface

site that binds to the element adjacent to it. All three feeds showed steadily increasing catalytic CH4 formation

rates with increasing applied overpotential up to -0.65 VRHE. Beyond -0.65 VRHE, catalytic CH4 rate hikes with

potential slowed down for CO2, yet increased sharply for CH2O. For CO feeds, the CH4 formation rates actually

peaked and subsequently dropped slightly at more cathodic potentials. We attribute this distinct kinetic

behavior to the poisoning effect of adsorbed *CO on the FeN4 sites due to their strong binding, which will be

supported later by computational analysis.13,15,17 For CO2, even though its saturated bulk concentration is about

P a g e | 66

30 times that of CO, local depletion in COx concentration at the double layer will limit sustained CH4 formation

rates at sufficiently cathodic potentials. Cathodically of -0.65 VRHE, the CH4 formation rate from CH2O

displayed a sudden growth suggesting a shift in the rate-determining reaction step. We want to point out that

previous kinetic studies invariably showed that the electroreduction of aldehydes on metallic electrocatalysts

exclusively generated the respective primary alcohols.15,110,111 Here, however, the solid non-metallic, single

site Fe-N-C catalysts with graphene-embedded molecular sites generated selectively the respective

hydrocarbon (CH4). This is consistent with results on molecular catalysts and suggests mechanistic analogies

between molecular and solid state catalysts.33

Next, we investigated the CH4 formation rate and its faradaic efficiency as a function of the initial CH2O

concentration in the electrolyte, see Figure S4-7. The CH4 yield and its FE scaled nearly linearly with the CH2O

concentration at a given applied potential, with the slope varying with the applied potential. This suggests the

CH4 rate law is approximately first order with respect to CH2O, which was confirmed in kinetic log-log analysis

(see Figure S4-8).

P a g e | 67

7.4 Proton- coupled and decoupled reaction steps

The generation of CH4 during the electrochemical reduction of CO2 on Fe-N-C (and similarly on molecular

Cobalt-Protoporphyrin) is known to show a Nernstian dependence on pH,33,47 suggesting that the rate-

determining protonation of adsorbed *CO does involve a concerted proton-coupled electron transfer (PCET).

To verify this, we studied the pH dependence of the CORR on the single site catalysts. Results in Figure 7-3a

and 7-3c confirmed the Nernstian behavior of the CH4 onset potentials and CH4 formation rates, evidenced by

the 59 mV shift per pH unit on the NHE scale. Thus, on the present Fe-N-C single site catalysts, experiments

suggest the protonation of *CO at the carbon atom to *CHO to be the slowest and thus rate-determining step

in the CORR pathway to methane. This will be compared to computational predictions further below.

Figure 7- 3 Production rate of CH4 at various pH as a function of iR-corrected applied electrode potentials. a)

CO reduction and b) CH2O reduction plotted on the RHE scale. c) CO reduction and d) CH2O reduction plotted

on the NHE scale. e) logarithm of the CH4 formation rate from CH2O at different pH versus applied potential.

Electrolytes are 0.05 M K2HPO4 + 0.05 M K3PO4 (pH = 11.9), 0.05 M K3PO4 + 0.05 M H3PO4 (pH = 6.9), and

0.05 M KH2PO4 + 0.05 M H3PO4 (pH = 2.25) for pH variation. Data are averages over 75 min electrolysis.

Line to guide the eye. Catalyst loading: 0.75 mg cm-2 on glassy carbon.

P a g e | 68

Similar pH tests were then conducted using CH2O as reactant feed (see Figure 7-3b, 7-3d, and 7-3e). Now, the

CH4 onset potentials exhibited close to none pH dependence on the NHE scale (Figure 7-3d), yet did so on the

RHE scale (Figure 7-3b). We conclude that the catalytic pathway from CH2O to CH4 is limited by a proton-

decoupled electron transfer (PDET) step resulting in an experimental rate law following

𝑅𝑎𝑡𝑒𝐶𝐻4 ∽ 𝑘 𝑒−𝛼𝜂[𝐶𝐻2𝑂]1[𝐻+]~0 eq. 7‐ 6

where k is a heterogeneous rate constant, the exponential term describes the rate dependence on the applied

cathodic overpotential 𝜂 [𝑉] and 𝛼[𝑉−1] denotes a parameter related to the inverse Tafel slope.

To summarize our experimental findings on Fe-N-C catalysts, the CO2RR as well as the CORR generally

exhibit a wider range of products, primarily CO, methane, as well as some formaldehyde and methanol. The

CH2ORR exclusively yields CH4. Both the CO2 to CO reaction and the CH2O to CH4 reactions appear to be

rate-limited by a slow proton-decoupled electron transfers (PDET), while the CO to CH4 reaction features a

concerted proton-coupled electron transfer (PCET) as its slowest step. Mechanistic DFT calculations are now

needed to rationalize the experimental findings and to establish a full mechanistic network.

P a g e | 69

7.5 Density Functional Theory Calculation

DFT predictions of possible mechanistic pathways from CO2 to CH4 were carried out by calculating the Free

Energy of possible intermediates on single-site Fe-N4-C motifs at 0 VRHE. This particular metal coordination

was chosen, because iron sites generally prefer four coordinative nitrogen (DFT calculation details see

Experiment Section).47

Figure 7- 4 Free energy diagram towards CH4 from CO2, CO and CH2O on Fe-N-C at 0 VRHE. The three limiting

potential steps are shown by V1, V2 and V3, with the reduction of CH2O having the smallest limiting potential

step in line with the experiments.

Figure 7-4 presents the free energy network diagram of the individual reaction pathways from the different

reactants to methane. The calculations are assuming concerted proton electron transfer, are evaluated at 0 VRHE

and with reference made to CH4 (g), as this reference is common for the reactions investigated. It shows that

Fe-N4-C binds *CO relatively strongly and the most difficult step from CO2/CO to CH4 is the protonating *CO

to *CHO (V2: ΔG*CO to ΔG*CHO). Indeed, the *COH has also been proposed as the first reduced intermediate

following *CO on metallic Cu facets.112,113 However, considering that the M-N-C type catalysts contribute

isolated active sites, the free energy of *COH intermediate is about 1.5 eV higher than that of *CHO, possibly

accommodating an unfavourable triple bond between Fe-Nx and the *COH intermediate.

In comparison, the reduction of CO2 to CO (V1: ΔGCO2 to ΔG*COOH) and CH2ORR to CH4 (V3: ΔGCH2O to

ΔG*CH2OH/*OCH3) exhibit lower energetic pathways, leading to less potential requirement to drive these two

conversions, which is in line with the experiment observations (see Figure 7-2 and Figure S4-6).

We firstly focus on a discussion of the selectivity of the reduction of CH2O, CH2ORR. Notably, no methanol

was observed over Fe-N-C catalysts, which is distinctly different from metals (Cu, Ag and Au), which majorly

P a g e | 70

produces methanol. Previously, for metal catalyst we tried to classify the two types of products from aldehyde

reductions to be a matter of choice: oxygen bonding (*OCH3) gives alcohols while the carbon bonding

(*CH2OH) leads to fully reduced hydrocarbons.15 Nevertheless, we do observe that the calculated *CH2OH

and *OCH3 for the Fe-N-C on free energy scale are similar (see Figure 7-4). Given the fact that CH2O reduction

on Fe-N-C only produces CH4, we thus propose that uniquely this catalyst offers a special reaction path or has

a very different water stabilization as compared to the normal metal catalyst. Water indeed highly influence

the stabilization of CO2RR intermediates for Cu facets114 and for ORR/OER intermediates water solvation has

been shown to be different on M-N-C systems as compare metals.115 Investigating water dynamics on the M-

N-C for this analysis was found to be challenging due to the spin-polarized nature of the calculations. This

issue could be addressed with ab initio molecular dynamics (AIMD) of water on the M-N-C system, which is

out of scope of the present study.

Secondly, we turn to the pH dependency of these reactions, where, the CO2RR into CO (performed in our early

work116, data is presented in Figure S4-9) and CH2ORR into CH4 showed a non-Nernstian behavior, while the

CO-to-CH4 process followed a Nernstian one. We hypothesize that a proton-decoupled electron transfer

(PDET) is the key to rationalize this pH dependence. A proton-decoupled electron transfer was reported to be

more likely if the interaction of the intermediates with the catalyst is weak.117 Based on this, for weakly bound

species, the adsorption becomes a slow rate determining step (RDS) and the reaction is limited by a proton

decoupled electron transfer (ET) as shown in the equations below:

CO2+ ∗ + e− → *CO2

− eq. 7‐ 7

CH2O + ∗ + e− → *CH2O− eq. 7‐ 8

accounting for the experimentally observed pH independent catalytic rates on the NHE scale. However, our

DFT calculations using the Computational Hydrogen Electrode (CHE) assumes a concerted proton-electron

transfer (equivalent to PCET), thus could not capture the feature of the PDET. Previous investigations have

indeed shown the simulation of stable M-N-C-CO2- transition states,33,118 however, without clearly pointing the

reference potential for such proton decoupled electron transfer. This implies that both models have their

limitations to clearly describe the electrocatalytic process. The experiments, however, provide important

information about the reaction mechanism.

We now turn to the discussion of the relatively low (yet finite compared to methanol) methane selectivity on

the Fe-N-C catalyst. The Fe-N4 motifs show similar “binding” properties as copper (the well-known excellent

hydrocarbons producer), holding the ability to bind *CO without having *H under potential deposited. Upon

this, to better understand the performance difference in CH4 selectivity of these two types candidates, we

compared the proton transfer details of the *CO-to-*CHO steps on the single-site Fe-Nx motif and the extended

metal Cu (111) facet.119 Illustrated in Figure 7-5, comparable PCET barriers (marked as V2 in both Figure 7-

5a and 7-5b, ΔG*CO to ΔG*CHO, ~1eV) could be observed on these two candidates. For the Cu (111) facet, a pre

*H / *CO co-adsorption (schematic is displayed in Figure 7-5a) is energetically favored, leaving V2 (~1eV) as

the major limiting step. This is in agreement with a recent work focusing on CH4 evolution via CORR on Cu

P a g e | 71

based catalyst,120 suggesting that the protonation of adsorbed *CO into *CHO (RDS of CH4 formation) is

preferentially via the “Langmuir-Hinshelwood” reaction channel (*CO + *H → *CHO). On the contrary, on

Fe-N-C motif, the co-adsorption of *CO / *H step (V marked in Figure 7-5b) sets the main dynamic barrier,

emerging as the limiting step and resulting in low hydrocarbon activity and forcing to a “Eley-Rideal” type

protonation step (*CO + H+sol + e- → *CHO).

Figure 7- 5 The detailed protonation steps of *CO towards *CHO via the co-adsorption of *CO + *H on a) Cu

(111) facet and on b) Fe-N-C, respectively.

In principle, the single site Fe-N-C catalyst constitutes an excellent catalytic CO producer, and thus could

contribute sufficient reactive *CO intermediates for higher hydrocarbon yields. Unfortunately, indicated by

our simulations, the isolated active site structure on the other hand causes a difficulty in *CO and *H co-

adsorption, precluding a dominant formation of hydrocarbons. This finding suggests that the high-active

CH4(g) product formation may require an extended surface or a nearby proton source to lower the barrier of

protonation. Strategies to circumvent this would be to increase the number of active sites and thus reducing

their mutual distance, or else the introduction of suitable hydrogen adsorption / COx reactive twin sites may be

a way toward enhanced hydrocarbon yields.

P a g e | 72

7.6 Discussion

The present combined experimental and theoretical mechanistic study has addressed the kinetics of the

electroreduction of various small molecules that may constitute reactive intermediates on the way to methane,

in particular CO2, CO and CH2O. The study aimed at i) identifying the role of these reducible reactants during

the overall CO2-towards-CH4 pathway; ii) understanding the rate limiting factor of these key reactions; iii)

reasoning the obstructed hydrocarbons selectivity of the Fe-N-C catalyst. As this knowledge will be crucial in

the design of novel catalysts to reduce CO2 into hydrocarbons.

Figure 7- 6 Hypothesized paths of methanol and methane formation on single-site Fe-N-C catalyst.

To achieve the first two goals, unfortunately, neither our experiments nor computation deliver direct evidence

to address the CH2O reduction details. By tracking the methanol trace (see Figure 7-6), which is produced in

terms of CO2/CORR electrolysis (Figure S4-5) while denied in CH2O reduction (Figure S4-7 and S4-8), we

could speculate, methanol is more likely originated via a non-CH2O channel, and the *CH2OH is the key

intermediate (*CO →*CHO →*CHOH →*CH2OH →CH3OH). Upon this hypothesis, the *CH2OH – relevant

for methanol formation – could be filtered out from the pure CH2ORR, making methane as the only end-

product (CH2O →*OCH3 →*O + CH4, shown in Figure 7-6). This shows that the Fe-N-C catalyst provides

different CH2ORR paths than the metals,15 and better understanding requires more systematical investigation

upon a broader catalysts benchmark.

Based on the afore discussions, we could track the overall reaction paths of electrochemical CO2 reduction

into CH4 with clearly addressing the contribution of CO2, CO and CH2O. Here, CO plays as the key

intermediate towards methane, formaldehyde and methanol. Moreover, the produced CH2O in this reaction

network, could be further reduced and open up an extra reaction channel towards methane. On the contrary,

methanol is more likely yielded from the non-CH2O reaction channel, poses as a by-product aside the methane.

More interestingly, our experimental studies have shown that some of the reduction steps involve proton-

coupled electron transfers (PCET, namely, CO-to-CH4) steps, whereas others, in particular such that involved

rather weakly bounded species, featured proton-decoupled electron transfer (PDET, namely, CO2-to-CO and

CH2O-to-CH4) steps. This provided a broader and deeper understanding of the reaction mechanism and

reaction network of the CO2 reduction into a wide range of single-carbon chemical on the Fe-N-C single site

catalysts. In this regard, an overall schematic display in Chapter 8, Scheme 8-2.

As to the third goal, our simulations revealed that the low hydrocarbons selectivity is primarily due to the

isolated nature of the active site of the Fe-Nx motifs, which limits the rate with which further reduction and

protonation of adsorbed *CO can occur. To overcome this drawback, synthesis approach to achieve catalysts

P a g e | 73

with novel site densities and site structure are needed. We speculate, that introducing dual twin metal sites,

where two metal centers are located in atomic proximity, would mechanistically allow a more rapid protonation

of CO, boosting the hydrocarbons formation on Fe-N-C catalysts.

P a g e | 74

Chapter 8. Summary and outlook

The M-N-C group catalysts have attracted increasing attention in recent years due to their remarkable CO2RR

performance, abundance and low-cost, making them promising candidates with great potential for future

commercial applications. However, as early-stage research, concerns such as reaction / process mechanism,

detailed understanding of nature-activity relationships need to be addressed in order to establish novel

perspectives for evolving the effective CO2RR catalysts and finding the efficient reaction conditions. This

work presented novel and valuable approaches to overcome some of these challenges for M-N-C catalysts,

which are summarized below (the corresponding structure is displayed in Scheme 8-1).

Scheme 8- 1 Summarized structure of this dissertation.

The mechanistic understanding of CO evolution over the M-N-C catalysts is addressed in chapter 4. Our

synthesis provided 5 different M-N-C (M: Mn, Fe, Co, Ni, Cu) for the broader testing, and linked the

experimental observation with DFT prediction. So we could suggest the nature of these porphyrin-like active

motifs. In this contribution, the Ni-based one showed remarkable performance for CO evolution and the great

potential for future industrial application. Explained by our theoretical simulation, the nature of the Ni-Nx

allowed to deny the unwanted HER and avoid the desorption issue due to weak interaction to H* and CO*,

with the drawback of a large overpotential requirement due to weak binding to COOH*, the first key

intermediate for CO2RR into CO.

To reach the industrial level CO2 electrolysis, performance exhibited in common liquid cell is far from

sufficient. Little solubility of CO2 is a common problem for CO2RR in regular CO2-saturated liquid electrolytes,

causing mass transfer limitation of the CO2 reactant. Thus, an engineering progress was done in a following-

up work. In chapter 5, the gas diffusion electrode was deployed to overcome the CO2 solubility issue. The

GDE could create a special CO2(g) / CO2(Sol.) / Catalyst interface, boosting the CO2 transfer for the

P a g e | 75

electrochemical reaction. Moreover, indicated by our DFT prediction, the Ni-N-C active site could intrinsically

deny CO desorption problem in this operation, thus delivering the CO evolution partial current density over

200 mA cm-2. However, the CO selectivity could not be maintained at higher working current densities. To

overcome this limitation, strategies in catalysts synthesis and GDL-interface modification could be

implemented. Towards catalysts synthesis, removable (soft) template such as the SiO2 substrate could be

introduced during the synthesis approach to physically increase the active site density of the catalyst.

Moreover, tuning the active motifs (coordination number, doping of other elements) using specific

methodologies could also be considered to achieve higher catalytic reactivities. Towards interface

modification, the hydrophobicity, pore-size of the diffusion electrode, receipt of catalyst coating as well as the

electrolyte pH value could play a significant role, which requires systematic research in the coming scenarios.

Scheme 8- 2 Overall reaction network form CO2 reduction towards CH4 via CO and CH2O over the Fe-N-C

catalyst. PCET: Proton Coupled Electron Transfer; PDET: Proton Decoupled Electron Transfer.

Unlike the Ni-based catalyst, the Fe- doped one suffers CO* poisoning problem at high potential / current

range. On the positive site, the strong binding allows to operate the CO2RR with less potential input, further,

could produce hydrocarbons. In chapter 6, we perform the synthesis (PANI-based Fe-N-C) using a secondary

nitrogen precursor and find melamine could help to introduce more Fe-Nx functionalities, improving the

CO2RR reactivity. Moreover, operando-XAS was employed to identify the Fe-Nx moieties in our operation

potential range. An unusual Fe(I)-N3 state was observed at high potential range (more negative than -0.9VRHE),

which could be responsible for methane formation.

P a g e | 76

In chapter 7, we provide the knowledge to draw an overall CO2RR reaction network over the Fe-N-C catalyst,

contributing the Scheme 8-2. The CO2-to-CO and CH2O-to-CH4 happens via a proton decoupled electron

transfer, whereas the proton is of great importance for CO further protonation. Our simulation points that the

CO* and H* co-adsorption is blocked at the singe site single site Fe-Nx motifs, thus methane remains as a

minor product. We suspect, novel M-N-C catalysts with specific bridging metal-metal-dimer active centers

could boost the hydrocarbons production.

P a g e | 77

Reference

1 Cox, P. M., Betts, R. A., Jones, C. D., Spall, S. A. & Totterdell, I. J. Acceleration of global

warming due to carbon-cycle feedbacks in a coupled climate model. Nature 408, 184,

doi:10.1038/35041539 (2000).

2 Hoegh-Guldberg, O. et al. Coral Reefs Under Rapid Climate Change and Ocean

Acidification. Science 318, 1737-1742, doi:10.1126/science.1152509 (2007).

3 Whipple, D. T. & Kenis, P. J. A. Prospects of CO2 Utilization via Direct Heterogeneous

Electrochemical Reduction. The Journal of Physical Chemistry Letters 1, 3451-3458,

doi:10.1021/jz1012627 (2010).

4 Hori, Y. in Modern Aspects of Electrochemistry (eds Constantinos G. Vayenas, Ralph E.

White, & Maria E. Gamboa-Aldeco) 89-189 (Springer New York, 2008).

5 Mistry, H. et al. Exceptional Size-Dependent Activity Enhancement in the Electroreduction

of CO2 over Au Nanoparticles. Journal of the American Chemical Society 136, 16473-

16476, doi:10.1021/ja508879j (2014).

6 Lu, Q. et al. A selective and efficient electrocatalyst for carbon dioxide reduction. Nature

Communications 5, 3242, doi:10.1038/ncomms4242

https://www.nature.com/articles/ncomms4242#supplementary-information (2014).

7 Kim, C. et al. Achieving Selective and Efficient Electrocatalytic Activity for CO2

Reduction Using Immobilized Silver Nanoparticles. Journal of the American Chemical

Society 137, 13844-13850, doi:10.1021/jacs.5b06568 (2015).

8 Kim, C. et al. Insight into Electrochemical CO2 Reduction on Surface-Molecule-Mediated

Ag Nanoparticles. ACS Catalysis 7, 779-785, doi:10.1021/acscatal.6b01862 (2017).

9 Rosen, J. et al. Mechanistic Insights into the Electrochemical Reduction of CO2 to CO on

Nanostructured Ag Surfaces. ACS Catalysis 5, 4293-4299, doi:10.1021/acscatal.5b00840

(2015).

10 Wenlei Zhu, R. M., Önder Metin, Haifeng Lv, Shaojun Guo, Christopher J. Wright,

Xiaolian Sun, Andrew A. Peterson, Shouheng Sun. Monodisperse Au Nanoparticles for

Selective Electrocatalytic Reduction of CO2 to CO. J. Am. Chem. Soc 135, 16833−16836

(2013).

11 Wenlei Zhu, Y.-J. Z., Hongyi Zhang, Haifeng Lv, Qing Li, Ronald Michalsky, Andrew A.

Peterson, Shouheng Sun. Active and Selective Conversion of CO2 to CO on Ultrathin Au

Nanowires. J. Am. Chem. Soc 136, 16132−16135 (2014).

12 Hori, Y., Murata, A. & Takahashi, R. Formation of hydrocarbons in the electrochemical

reduction of carbon dioxide at a copper electrode in aqueous solution. Journal of the

Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases 85,

2309-2326, doi:10.1039/F19898502309 (1989).

13 Peterson, A. A., Abild-Pedersen, F., Studt, F., Rossmeisl, J. & Norskov, J. K. How copper

catalyzes the electroreduction of carbon dioxide into hydrocarbon fuels. Energy &

Environmental Science 3, 1311-1315, doi:10.1039/C0EE00071J (2010).

14 Kuhl, K. P., Cave, E. R., Abram, D. N. & Jaramillo, T. F. New insights into the

electrochemical reduction of carbon dioxide on metallic copper surfaces. Energy &

Environmental Science 5, 7050-7059, doi:10.1039/C2EE21234J (2012).

15 Bagger, A., Ju, W., Varela Ana, S., Strasser, P. & Rossmeisl, J. Electrochemical CO2

Reduction: A Classification Problem. ChemPhysChem 18, 3266-3273,

doi:10.1002/cphc.201700736 (2017).

16 Wang, L. et al. Electrochemical Carbon Monoxide Reduction on Polycrystalline Copper:

Effects of Potential, Pressure, and pH on Selectivity toward Multicarbon and Oxygenated

Products. ACS Catalysis, 7445-7454, doi:10.1021/acscatal.8b01200 (2018).

P a g e | 78

17 Hori, Y., Takahashi, R., Yoshinami, Y. & Murata, A. Electrochemical Reduction of CO at a

Copper Electrode. The Journal of Physical Chemistry B 101, 7075-7081,

doi:10.1021/jp970284i (1997).

18 Murata, A. & Hori, Y. Product Selectivity Affected by Cationic Species in Electrochemical

Reduction of CO2 and CO at a Cu Electrode. Bulletin of the Chemical Society of Japan 64,

123-127, doi:10.1246/bcsj.64.123 (1991).

19 Schouten, K. J. P., Pérez Gallent, E. & Koper, M. T. M. The influence of pH on the

reduction of CO and CO2 to hydrocarbons on copper electrodes. Journal of

Electroanalytical Chemistry 716, 53-57, doi:https://doi.org/10.1016/j.jelechem.2013.08.033

(2014).

20 Mistry, H. et al. Highly selective plasma-activated copper catalysts for carbon dioxide

reduction to ethylene. Nature Communications 7, 12123, doi:10.1038/ncomms12123

https://www.nature.com/articles/ncomms12123#supplementary-information (2016).

21 Gao, D., Scholten, F. & Roldan Cuenya, B. Improved CO2 Electroreduction Performance on

Plasma-Activated Cu Catalysts via Electrolyte Design: Halide Effect. ACS Catalysis 7,

5112-5120, doi:10.1021/acscatal.7b01416 (2017).

22 Tang, W. et al. The importance of surface morphology in controlling the selectivity of

polycrystalline copper for CO2 electroreduction. Physical Chemistry Chemical Physics 14,

76-81, doi:10.1039/C1CP22700A (2012).

23 Li, C. W., Ciston, J. & Kanan, M. W. Electroreduction of carbon monoxide to liquid fuel on

oxide-derived nanocrystalline copper. Nature 508, 504, doi:10.1038/nature13249 (2014).

24 Dinh, C.-T. et al. CO<sub>2</sub> electroreduction to ethylene via hydroxide-mediated

copper catalysis at an abrupt interface. Science 360, 783-787, doi:10.1126/science.aas9100

(2018).

25 Wang, X., Varela, A. S., Bergmann, A., Kühl, S. & Strasser, P. Catalyst Particle Density

Controls Hydrocarbon Product Selectivity in CO2 Electroreduction on CuOx.

ChemSusChem 10, 4642-4649, doi:doi:10.1002/cssc.201701179 (2017).

26 Wannakao, S., Jumpathong, W. & Kongpatpanich, K. Tailoring Metalloporphyrin

Frameworks for an Efficient Carbon Dioxide Electroreduction: Selectively Stabilizing Key

Intermediates with H-Bonding Pockets. Inorganic Chemistry 56, 7200-7209,

doi:10.1021/acs.inorgchem.7b00839 (2017).

27 Tripkovic, V. et al. Electrochemical CO2 and CO Reduction on Metal-Functionalized

Porphyrin-like Graphene. The Journal of Physical Chemistry C 117, 9187-9195,

doi:10.1021/jp306172k (2013).

28 Fisher, B. J. & Eisenberg, R. Electrocatalytic reduction of carbon dioxide by using

macrocycles of nickel and cobalt. Journal of the American Chemical Society 102, 7361-

7363, doi:10.1021/ja00544a035 (1980).

29 Beley, M., Collin, J.-P., Ruppert, R. & Sauvage, J.-P. Nickel(II)-cyclam: an extremely

selective electrocatalyst for reduction of CO2 in water. Journal of the Chemical Society,

Chemical Communications, 1315-1316, doi:10.1039/C39840001315 (1984).

30 Hawecker, J., Lehn, J.-M. & Ziessel, R. Electrocatalytic reduction of carbon dioxide

mediated by Re(bipy)(CO)3Cl (bipy = 2,2′-bipyridine). Journal of the Chemical Society,

Chemical Communications, 328-330, doi:10.1039/C39840000328 (1984).

31 Reuillard, B. et al. Tuning Product Selectivity for Aqueous CO2 Reduction with a

Mn(bipyridine)-pyrene Catalyst Immobilized on a Carbon Nanotube Electrode. Journal of

the American Chemical Society 139, 14425-14435, doi:10.1021/jacs.7b06269 (2017).

32 Behar, D. et al. Cobalt Porphyrin Catalyzed Reduction of CO2. Radiation Chemical,

Photochemical, and Electrochemical Studies. The Journal of Physical Chemistry A 102,

2870-2877, doi:10.1021/jp9807017 (1998).

P a g e | 79

33 Shen, J. et al. Electrocatalytic reduction of carbon dioxide to carbon monoxide and methane

at an immobilized cobalt protoporphyrin. Nature Communications 6, 8177,

doi:10.1038/ncomms9177

http://www.nature.com/articles/ncomms9177#supplementary-information (2015).

34 Birdja, Y. Y., Shen, J. & Koper, M. T. M. Influence of the metal center of

metalloprotoporphyrins on the electrocatalytic CO2 reduction to formic acid. Catalysis

Today 288, 37-47, doi:https://doi.org/10.1016/j.cattod.2017.02.046 (2017).

35 Costentin, C., Drouet, S., Robert, M. & Savéant, J.-M. A Local Proton Source Enhances

CO<sub>2</sub> Electroreduction to CO by a Molecular Fe Catalyst. Science 338, 90-94,

doi:10.1126/science.1224581 (2012).

36 Varela, A. S., Ju, W. & Strasser, P. Molecular Nitrogen–Carbon Catalysts, Solid Metal

Organic Framework Catalysts, and Solid Metal/Nitrogen-Doped Carbon (MNC) Catalysts

for the Electrochemical CO2 Reduction. Advanced Energy Materials 8, 1703614,

doi:doi:10.1002/aenm.201703614 (2018).

37 Wu, G., More, K. L., Johnston, C. M. & Zelenay, P. High-Performance Electrocatalysts for

Oxygen Reduction Derived from Polyaniline, Iron, and Cobalt. Science 332, 443-447,

doi:10.1126/science.1200832 (2011).

38 Kramm, U. I. et al. Structure of the catalytic sites in Fe/N/C-catalysts for O2-reduction in

PEM fuel cells. Physical Chemistry Chemical Physics 14, 11673-11688,

doi:10.1039/C2CP41957B (2012).

39 Ranjbar Sahraie, N., Paraknowitsch, J. P., Göbel, C., Thomas, A. & Strasser, P. Noble-

Metal-Free Electrocatalysts with Enhanced ORR Performance by Task-Specific

Functionalization of Carbon using Ionic Liquid Precursor Systems. Journal of the American

Chemical Society 136, 14486-14497, doi:10.1021/ja506553r (2014).

40 Sahraie, N. R. et al. Quantifying the density and utilization of active sites in non-precious

metal oxygen electroreduction catalysts. Nature Communications 6, 8618,

doi:10.1038/ncomms9618

https://www.nature.com/articles/ncomms9618#supplementary-information (2015).

41 Leonard, N. D. et al. Deconvolution of Utilization, Site Density, and Turnover Frequency of

Fe–Nitrogen–Carbon Oxygen Reduction Reaction Catalysts Prepared with Secondary N-

Precursors. ACS Catalysis 8, 1640-1647, doi:10.1021/acscatal.7b02897 (2018).

42 Zitolo, A. et al. Identification of catalytic sites for oxygen reduction in iron- and nitrogen-

doped graphene materials. Nature Materials 14, 937-+, doi:10.1038/Nmat4367 (2015).

43 Varela, A. S. et al. Metal-Doped Nitrogenated Carbon as an Efficient Catalyst for Direct

CO2 Electroreduction to CO and Hydrocarbons. Angew Chem Int Edit 54, 10758-10762,

doi:10.1002/anie.201502099 (2015).

44 Li, X. et al. Exclusive Ni–N4 Sites Realize Near-Unity CO Selectivity for Electrochemical

CO2 Reduction. Journal of the American Chemical Society 139, 14889-14892,

doi:10.1021/jacs.7b09074 (2017).

45 Wu, J. et al. Achieving Highly Efficient, Selective, and Stable CO2 Reduction on Nitrogen-

Doped Carbon Nanotubes. ACS Nano 9, 5364-5371, doi:10.1021/acsnano.5b01079 (2015).

46 Wu, J. et al. A metal-free electrocatalyst for carbon dioxide reduction to multi-carbon

hydrocarbons and oxygenates. 7, 13869, doi:10.1038/ncomms13869

https://www.nature.com/articles/ncomms13869#supplementary-information (2016).

47 Leonard, N. et al. The chemical identity, state and structure of catalytically active centers

during the electrochemical CO2 reduction on porous Fe-nitrogen-carbon (Fe-N-C) materials.

Chemical Science 9, 5064-5073, doi:10.1039/C8SC00491A (2018).

P a g e | 80

48 Zitolo, A. et al. Identification of catalytic sites in cobalt-nitrogen-carbon materials for the

oxygen reduction reaction. Nature Communications 8, 957, doi:10.1038/s41467-017-01100-

7 (2017).

49 Artyushkova, K., Serov, A., Rojas-Carbonell, S. & Atanassov, P. Chemistry of

Multitudinous Active Sites for Oxygen Reduction Reaction in Transition Metal–Nitrogen–

Carbon Electrocatalysts. The Journal of Physical Chemistry C 119, 25917-25928,

doi:10.1021/acs.jpcc.5b07653 (2015).

50 Kabir, S., Artyushkova, K., Kiefer, B. & Atanassov, P. Computational and experimental

evidence for a new TM-N3/C moiety family in non-PGM electrocatalysts. Physical

Chemistry Chemical Physics 17, 17785-17789, doi:10.1039/C5CP02230D (2015).

51 Ju, W. et al. Understanding activity and selectivity of metal-nitrogen-doped carbon catalysts

for electrochemical reduction of CO2. Nature Communications 8, 944, doi:10.1038/s41467-

017-01035-z (2017).

52 Jiang, K. et al. Isolated Ni single atoms in graphene nanosheets for high-performance CO2

reduction. Energy & Environmental Science 11, 893-903, doi:10.1039/C7EE03245E (2018).

53 Yang, F. et al. Highly Efficient CO2 Electroreduction on ZnN4-based Single-Atom

Catalyst. Angewandte Chemie 130, 12483-12487, doi:doi:10.1002/ange.201805871 (2018).

54 Chen, Z., Mou, K., Yao, S. & Liu, L. Zinc-Coordinated Nitrogen-Codoped Graphene as an

Efficient Catalyst for Selective Electrochemical Reduction of CO2 to CO. ChemSusChem

11, 2944-2952, doi:doi:10.1002/cssc.201800925 (2018).

55 Roy, A. et al. Nanostructured metal-N-C electrocatalysts for CO2 reduction and hydrogen

evolution reactions. Applied Catalysis B: Environmental 232, 512-520,

doi:https://doi.org/10.1016/j.apcatb.2018.03.093 (2018).

56 Su, P., Iwase, K., Harada, T., Kamiya, K. & Nakanishi, S. Covalent triazine framework

modified with coordinatively-unsaturated Co or Ni atoms for CO2 electrochemical

reduction. Chemical Science 9, 3941-3947, doi:10.1039/C8SC00604K (2018).

57 Dembinska, B., Kiciński, W., Januszewska, A., Dobrzeniecka, A. & Kulesza, P. J. Carbon

Dioxide Electroreduction at Highly Porous Nitrogen and Sulfur Co-Doped Iron-Containing

Heterogeneous Carbon Gel. Journal of The Electrochemical Society 164, H484-H490,

doi:10.1149/2.0721707jes (2017).

58 Möller, T. et al. Efficient CO2 to CO electrolysis on solid Ni–N–C catalysts at industrial

current densities. Energy & Environmental Science, doi:10.1039/C8EE02662A (2019).

59 Zhang, Z. et al. Reaction Mechanisms of Well-Defined Metal–N4 Sites in Electrocatalytic

CO2 Reduction. Angewandte Chemie International Edition 57, 16339-16342,

doi:doi:10.1002/anie.201808593 (2018).

60 Jiang, K. et al. Transition-Metal Single Atoms in a Graphene Shell as Active Centers for

Highly Efficient Artificial Photosynthesis. Chem,

doi:https://doi.org/10.1016/j.chempr.2017.09.014 (2017).

61 Yan, C. et al. Coordinatively unsaturated nickel–nitrogen sites towards selective and high-

rate CO2 electroreduction. Energy & Environmental Science 11, 1204-1210,

doi:10.1039/C8EE00133B (2018).

62 Ravel, B. & Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray

absorption spectroscopy using IFEFFIT. J Synchrotron Radiat 12, 537-541,

doi:10.1107/S0909049505012719 (2005).

63 Ankudinov, A. L., Ravel, B., Rehr, J. J. & Conradson, S. D. Real-space multiple-scattering

calculation and interpretation of x-ray-absorption near-edge structure. Phys Rev B 58, 7565-

7576, doi:DOI 10.1103/PhysRevB.58.7565 (1998).

64 Ask Hjorth, L. et al. The atomic simulation environment—a Python library for working with

atoms. Journal of Physics: Condensed Matter 29, 273002 (2017).

P a g e | 81

65 Enkovaara, J. et al. Electronic structure calculations with GPAW: a real-space

implementation of the projector augmented-wave method. Journal of Physics: Condensed

Matter 22, 253202 (2010).

66 Hammer, B., Hansen, L. B. & Nørskov, J. K. Improved adsorption energetics within

density-functional theory using revised Perdew-Burke-Ernzerhof functionals. Phys Rev B

59, 7413-7421, doi:10.1103/PhysRevB.59.7413 (1999).

67 Nørskov, J. K. et al. Origin of the Overpotential for Oxygen Reduction at a Fuel-Cell

Cathode. The Journal of Physical Chemistry B 108, 17886-17892, doi:10.1021/jp047349j

(2004).

68 Chan, K., Tsai, C., Hansen Heine, A. & Nørskov Jens, K. Molybdenum Sulfides and

Selenides as Possible Electrocatalysts for CO2 Reduction. ChemCatChem 6, 1899-1905,

doi:10.1002/cctc.201402128 (2014).

69 Guang-Ping Hao, W.-C. L., Dan Qian, Guang-Hui Wang, Wei-Ping Zhang, Tao Zhang, Ai-

Qin Wang, Ferdi Schüth, Hans-Josef Bongard, An-Hui Lu. Structurally Designed Synthesis

of Mechanically Stable Poly(benzoxazine-co-resol)-Based Porous Carbon Monoliths and

Their Application as High-Performance CO2 Capture Sorbents. J. Am. Chem. Soc 133,

11378–11388 (2011).

70 C.D. Wagner, A. V. N., A. Kraut-Vass, J.W. Allison, C.J. Powell, J.R. Jr. Rumble. NIST

Standard Reference Database 20. (2003).

71 M.C. Biesinger, B. P. P., L.W.M. Lau, A.R. Gerson, R.St.C. Smart. Resolving surface

chemical states in XPS analysis of first row transition metals, oxides and hydroxides: Cr,

Mn, Fe, Co and Ni. Appplied Surface Science 257, 887-898 (2010).

72 S. Muralidharan, R. G. H. Satellite in the x‐ray photoelectron spectra of metalloporphyrins.

Journal of Chemical Physics 71, 2970-2974 (1979).

73 H.W. Nesbitt, D. B. Interpretation of XPS Mn(2p) spectra of Mn oxyhydroxides and

constraints on the mechanism of MnO2 precipitation. American Mineralogist 83, 305-315

(1998).

74 Bai, Y. Photoelectron Spectroscopic Investigations of Porphyrins and Phthalocyanines on

Ag(111) and Au(111): Adsorption and Reactivity. (2010).

75 Li, Y., Su, H., Chan, S. H. & Sun, Q. CO2 Electroreduction Performance of Transition

Metal Dimers Supported on Graphene: A Theoretical Study. ACS Catalysis 5, 6658-6664,

doi:10.1021/acscatal.5b01165 (2015).

76 Back, S., Lim, J., Kim, N.-Y., Kim, Y.-H. & Jung, Y. Single-atom catalysts for CO2

electroreduction with significant activity and selectivity improvements. Chemical Science 8,

1090-1096, doi:10.1039/C6SC03911A (2017).

77 Shen, H., Li, Y. & Sun, Q. CO2 Electroreduction Performance of Phthalocyanine Sheet with

Mn Dimer: A Theoretical Study. The Journal of Physical Chemistry C 121, 3963-3969,

doi:10.1021/acs.jpcc.7b00317 (2017).

78 Michael Busch, N. B. H., Ulrike I. Kramm, Samira Siahrostami, Petr Krtil, Jan Rossmeisl.

Beyond the top of the volcano? – A unified approach to electrocatalytic oxygen reduction

and oxygen evolution. Nano Energy, 2211-2855 (2016).

79 Atchudan, R., Joo, J. & Pandurangan, A. An efficient synthesis of graphenated carbon

nanotubes over the tailored mesoporous molecular sieves by chemical vapor deposition.

Materials Research Bulletin 48, 2205-2212,

doi:https://doi.org/10.1016/j.materresbull.2013.02.048 (2013).

80 Wang, Z. et al. Nanocarbons from rice husk by microwave plasma irradiation: From

graphene and carbon nanotubes to graphenated carbon nanotube hybrids. Carbon 94, 479-

484, doi:https://doi.org/10.1016/j.carbon.2015.07.037 (2015).

81 Artyushkova, K. et al. Density functional theory calculations of XPS binding energy shift

for nitrogen-containing graphene-like structures. Chem Commun 49, 2539-2541,

doi:10.1039/c3cc40324f (2013).

P a g e | 82

82 Gupta, N., Gattrell, M. & MacDougall, B. Calculation for the cathode surface concentrations

in the electrochemical reduction of CO2 in KHCO3 solutions. Journal of Applied

Electrochemistry 36, 161-172, doi:10.1007/s10800-005-9058-y (2006).

83 Haas, T., Krause, R., Weber, R., Demler, M. & Schmid, G. Technical photosynthesis

involving CO2 electrolysis and fermentation. Nature Catalysis 1, 32-39,

doi:10.1038/s41929-017-0005-1 (2018).

84 Verma, S., Lu, X., Ma, S., Masel, R. I. & Kenis, P. J. The effect of electrolyte composition

on the electroreduction of CO2 to CO on Ag based gas diffusion electrodes. Phys Chem

Chem Phys 18, 7075-7084, doi:10.1039/c5cp05665a (2016).

85 Dinh, C.-T., García de Arquer, F. P., Sinton, D. & Sargent, E. H. High Rate, Selective, and

Stable Electroreduction of CO2 to CO in Basic and Neutral Media. ACS Energy Letters,

2835-2840, doi:10.1021/acsenergylett.8b01734 (2018).

86 Nallathambi, V., Leonard, N., Kothandaraman, R. & Barton, S. C. Nitrogen Precursor

Effects in Iron-Nitrogen-Carbon Oxygen Reduction Catalysts. Electrochem Solid St 14,

B55-B58, doi:10.1149/1.3566065 (2011).

87 Leonard, N., Nallathambi, V. & Barton, S. C. Carbon Supports for Non-Precious Metal

Oxygen Reducing Catalysts. Journal of The Electrochemical Society 160, F788-F792,

doi:10.1149/2.026308jes (2013).

88 Leonard, N. D. & Barton, S. C. Analysis of Adsorption Effects on a Metal-Nitrogen-Carbon

Catalyst Using a Rotating Ring-Disk Study. Journal of the Electrochemical Society 161,

H3100-H3105, doi:10.1149/2.0161413jes (2014).

89 Chung, H. T. et al. Combining Nitrogen Precursors in Synthesis of Non-Precious Metal

ORR Catalysts with Improved Fuel Cell Performance. Meeting Abstracts MA2015-02, 1278

(2015).

90 Chung, H. T. et al. Cyanamide-derived non-precious metal catalyst for oxygen reduction.

Electrochem Commun 12, 1792-1795 (2010).

91 Chung, H. T., Won, J. H. & Zelenay, P. Active and stable carbon nanotube/nanoparticle

composite electrocatalyst for oxygen reduction. Nat Commun 4, 1922,

doi:10.1038/ncomms2944 (2013).

92 Tian, J., Birry, L., Jaouen, F. & Dodelet, J. P. Fe-based catalysts for oxygen reduction in

proton exchange membrane fuel cells with cyanamide as nitrogen precursor and/or pore-

filler. Electrochim Acta 56, 3276-3285, doi:10.1016/j.electacta.2011.01.029 (2011).

93 Workman, M. J. et al. Platinum group metal-free electrocatalysts: Effects of synthesis on

structure and performance in proton-exchange membrane fuel cell cathodes. J Power

Sources 348, 30-39, doi:10.1016/j.jpowsour.2017.02.067 (2017).

94 Beamson, G. & Briggs, D. High-Resolution Monochromated X-Ray Photoelectron-

Spectroscopy of Organic Polymers - a Comparison between Solid-State Data for Organic

Polymers and Gas-Phase Data for Small Molecules. Mol Phys 76, 919-936, doi:Doi

10.1080/00268979200101761 (1992).

95 Lindberg, B. J. et al. Molecular Spectroscopy by Means of ESCA II. Sulfur compounds.

Correlation of electron binding energy with structure. Physica Scripta 1, 286 (1970).

96 Smart, R. S. C., Skinner, W. M. & Gerson, A. R. XPS of sulphide mineral surfaces: metal-

deficient, polysulphides, defects and elemental sulphur. Surface and Interface Analysis 28,

101-105, doi:10.1002/(SICI)1096-9918(199908)28:1<101::AID-SIA627>3.0.CO;2-0

(1999).

97 Yamashita, T. & Hayes, P. Analysis of XPS spectra of Fe2+ and Fe3+ ions in oxide

materials. Appl Surf Sci 254, 2441-2449, doi:10.1016/j.apsusc.2007.09.063 (2008).

98 Grosvenor, A. P., Kobe, B. A., Biesinger, M. C. & McIntyre, N. S. Investigation of multiplet

splitting of Fe 2p XPS spectra and bonding in iron compounds. Surface and Interface

Analysis 36, 1564-1574, doi:10.1002/sia.1984 (2004).

P a g e | 83

99 Tanuma, S., Powell, C. J. & Penn, D. R. Calculations of Electron Inelastic Mean Free Paths

for 31 Materials. Surface and Interface Analysis 11, 577-589, doi:DOI

10.1002/sia.740111107 (1988).

100 Li, J. K. et al. Structural and mechanistic basis for the high activity of Fe-N-C catalysts

toward oxygen reduction. Energy & Environmental Science 9, 2418-2432,

doi:10.1039/c6ee01160h (2016).

101 Hu, C. J., Noll, B. C., Schulz, C. E. & Scheidt, W. R. Four-coordinate iron(II)

porphyrinates: Electronic configuration change by intermolecular interaction. Inorganic

Chemistry 46, 619-621, doi:10.1021/ic0620182 (2007).

102 Varnell, J. A. et al. Identification of carbon-encapsulated iron nanoparticles as active species

in non-precious metal oxygen reduction catalysts. Nature Communications 7, 12582,

doi:10.1038/ncomms12582

http://www.nature.com/articles/ncomms12582#supplementary-information (2016).

103 Bagger, A., Ju, W., Varela, A. S., Strasser, P. & Rossmeisl, J. Single site porphyrine-like

structures advantages over metals for selective electrochemical CO2 reduction. Catalysis

Today, doi:http://doi.org/10.1016/j.cattod.2017.02.028 (2017).

104 Jia, Q. et al. Experimental Observation of Redox-Induced Fe-N Switching Behavior as a

Determinant Role for Oxygen Reduction Activity. Acs Nano 9, 12496-12505,

doi:10.1021/acsnano.5b05984 (2015).

105 Costentin, C., Robert, M., Saveant, J. M. & Tatin, A. Efficient and selective molecular

catalyst for the CO2-to-CO electrochemical conversion in water. P Natl Acad Sci USA 112,

6882-6886, doi:10.1073/pnas.1507063112 (2015).

106 Azcarate, I., Costentin, C., Robert, M. & Saveant, J. M. Through-Space Charge Interaction

Substituent Effects in Molecular Catalysis Leading to the Design of the Most Efficient

Catalyst of CO2-to-CO Electrochemical Conversion. Journal of the American Chemical

Society 138, 16639-16644, doi:10.1021/jacs.6b07014 (2016).

107 Rao, H., Schmidt, L. C., Bonin, J. & Robert, M. Visible-light-driven methane formation

from CO2 with a molecular iron catalyst. Nature 548, 74-77, doi:10.1038/nature23016

(2017).

108 Bhugun, I., Lexa, D. & Saveant, J. M. Homogeneous catalysis of electrochemical hydrogen

evolution by iron(0) porphyrins. Journal of the American Chemical Society 118, 3982-3983,

doi:DOI 10.1021/ja954326x (1996).

109 Ju, W. et al. Unraveling Mechanistic Reaction Pathways of the Electrochemical CO2

Reduction on Fe-N-C Single Site Catalysts. ACS Energy Letters,

doi:10.1021/acsenergylett.9b01049 (2019).

110 Bertheussen, E. et al. Acetaldehyde as an Intermediate in the Electroreduction of Carbon

Monoxide to Ethanol on Oxide‐Derived Copper. Angewandte Chemie International Edition

55, 1450-1454, doi:10.1002/anie.201508851 (2015).

111 Ledezma-Yanez, I., Gallent, E. P., Koper, M. T. M. & Calle-Vallejo, F. Structure-sensitive

electroreduction of acetaldehyde to ethanol on copper and its mechanistic implications for

CO and CO2 reduction. Catalysis Today 262, 90-94,

doi:https://doi.org/10.1016/j.cattod.2015.09.029 (2016).

112 Nie, X., Esopi, M. R., Janik, M. J. & Asthagiri, A. Selectivity of CO2 Reduction on Copper

Electrodes: The Role of the Kinetics of Elementary Steps. Angewandte Chemie

International Edition 52, 2459-2462, doi:10.1002/anie.201208320 (2013).

113 Hussain, J., Jónsson, H. & Skúlason, E. Calculations of Product Selectivity in

Electrochemical CO2 Reduction. ACS Catalysis 8, 5240-5249,

doi:10.1021/acscatal.7b03308 (2018).

114 Bagger, A., Arnarson, L., Hansen, M. H., Spohr, E. & Rossmeisl, J. Electrochemical CO

Reduction: A Property of the Electrochemical Interface. Journal of the American Chemical

Society 141, 1506-1514, doi:10.1021/jacs.8b08839 (2019).

P a g e | 84

115 Calle-Vallejo, F., Krabbe, A. & García-Lastra, J. M. How covalence breaks adsorption-

energy scaling relations and solvation restores them. Chemical Science 8, 124-130,

doi:10.1039/C6SC02123A (2017).

116 Varela, A. S. et al. pH Effects on the Selectivity of the Electrocatalytic CO2 Reduction on

Graphene-Embedded Fe–N–C Motifs: Bridging Concepts between Molecular Homogeneous

and Solid-State Heterogeneous Catalysis. ACS Energy Letters 3, 812-817,

doi:10.1021/acsenergylett.8b00273 (2018).

117 Koper, M. T. M. Theory of the transition from sequential to concerted electrochemical

proton–electron transfer. Physical Chemistry Chemical Physics 15, 1399-1407,

doi:10.1039/C2CP42369C (2013).

118 Göttle, A. J. & Koper, M. T. M. Proton-coupled electron transfer in the electrocatalysis of

CO2 reduction: prediction of sequential vs. concerted pathways using DFT. Chemical

Science 8, 458-465, doi:10.1039/C6SC02984A (2017).

119 Hori, Y., Takahashi, I., Koga, O. & Hoshi, N. Electrochemical reduction of carbon dioxide

at various series of copper single crystal electrodes. Journal of Molecular Catalysis A:

Chemical 199, 39-47, doi:https://doi.org/10.1016/S1381-1169(03)00016-5 (2003).

120 Schreier, M., Yoon, Y., Jackson, M. N. & Surendranath, Y. Competition between H and CO

for Active Sites Governs Copper-Mediated Electrosynthesis of Hydrocarbon Fuels.

Angewandte Chemie International Edition 57, 10221-10225,

doi:doi:10.1002/anie.201806051 (2018).

121 Skála, R., Císařová, I. & Drábek, M. in American Mineralogist Vol. 91 917 (2006).

P a g e | 85

Appendix

A1. Supplementary Information to Chapter 4

Table S1- 1 Physical parameters of nitrogen-metal doped carbon (M-N-C) CO2RR electrocatalysts

SBET a

m2g-1

DL-Cap b

mF cm-2

Overall metal by ICP

wt. %

Metal c

Cu d

Mn-N-C

938

87.82

Mn: 0.44

0.16

Fe-N-C

724

66.94

Ni: 2.88

0.12

Co-N-C

758

80.39

Co: 1.11

0.17

Ni-N-C

615

38.82

Fe: 1.73

0.14

Cu-N-C

807

74.38

1.11

a) SBET: N2 ad/desorption-based surface area; b) Double layer capacity values (Geometric area of each electrode

is 1 cm2 with 0.76 mg cm-2 catalyst loading); c,d) overall metal content determined from ICP-OES

measurement. Notably, the detection limit of ICP-OES is as low as 0.1, 0.25, 0.4, 0.5 and 0.1 g/L for Fe, Co,

Cu, Ni and Mn, respectively, where the measured solution of the dissolved M-N-C sample is 5000, 1400, 900,

1700 and 1300 times higher than the limit, indicating a low uncertainty. Each sample is measured twice. The

value in the table is the average.

P a g e | 86

Scanning Electron Microscopy

Figure S1- 1 Typical SEM image of the M-N-C electrocatalysts. Scale bar: 10 m.

N2 physisorption

Figure S1- 2 N2 physisorption isotherms for this family of M-N-C electrocatalysts. As shown, the isotherms

are type I indicating the microporous structures; the abrupt increase in the higher pressure regime indicating

the existence of macropores. In short, this information confirms the rich of microporosity, and indicates the

hierarchical structure.

P a g e | 87

Double Layer Capacity Measurements – Relation to BET surface area

Cycle voltammetry was performed between 0.0 and 0.52 V vs. RHE to avoid the interference of the faradaic

process. Double layer current density was utilized to determine double layer capacity, which is proportional to

the double-layer interfacial area.

Figure S1- 3 a) to c) Cyclic voltammetry of the five M-N-C catalysts conducted in CO2 -saturated 0.1 M

KHCO3 at various scan rates for estimation of double layer (DL) capacity. d) current densities from a-c plotted

vs. electrode potential scan rate to extract the double layer capacity; e) Correlation of double layer capacity

and the N2 adsorption –derived BET surface area.

P a g e | 88

H2O adsorption measurements

Figure S1- 4 Water vapor adsorption isotherms for this family of M-N-C electrocatalysts (top) with highlighted

low pressure regime (bellow) together with typical benchmark samples, including hydrophilic zeolite 13X,

most hydrophilic carbon as well as porous carbons (mesoporous CMK-3, commercial activated carbon Norit

Super) that commonly show hydrophobic nature. As shown, the surface of this family of M-N-C is medium

hydrophilic or hydrophobic as compared with the typical benchmark materials.

P a g e | 89

X-ray Diffraction Measurements

Figure S1- 5 XRD patterns of this family of M-N-C catalysts.

P a g e | 90

STEM/EDX Dark field microscopy, z Contrast, and elemental mapping

Figure S1- 6 STEM images and elemental maps for Fe-N-C, Co-N-C and Ni-N-C; Scale bar in left column:

200 nm, middle column: 20 nm.

P a g e | 91

High Resolution X-ray photoemission spectroscopy

Figure S1- 7 High-resolution Cu 2p (a) and N 1s (b) XP spectra of the Cu-N-C sample. See main text for N 1s

peak assignment.

P a g e | 92

Figure S1- 8 High-resolution metal 2p3/2 of Co-N-C (a), Mn-N-C (b), Ni-N-C (c) and Fe-N-C (d). Blue lines

represent 2p3/2 main peaks, red – the corresponding shake-up satellites.

Table S1- 2 Metal atom, nitrogen content and assignment of the different N species in the M-N-C CO2RR

electrocatalysts extracted from XPS data. Total metal and nitrogen content are calculated from measured XP

spectra areas using instrument-specific relative sensitivity factors provided by manufacturer (SPECS).

Sample

Metala

atom %

Nitrogena

atom%

Fitting of Nitrogen moieties / atom%

Metal-N

Pyridinic

Pyrrolic

Graphitic

N-Ox

CH-Ox

Mn-N-C

Mn: 0.2

8.2

12.2

25.8

43.9

10.3

5.0

2.8

Fe-N-C

Fe: 0.3

8.2

10.4

24.2

44.6

11.7

6.0

3.1

Co-N-C

Co: 0.4

10.5

13.4

28.1

41.6

9.8

4.4

2.7

Ni-N-C

Ni: 0.7

9.5

20.2

17.1

43.6

10.7

5.3

3.2

Cu-N-C

Cu: 0.8

14.6

15.5

30.4

32.9

7.0

3.9

10.3

790 785 780 775 650 645 640

715 710 705865 860 855 850

CPS (arb. units)

a) Co 2p3/2

b) Mn 2p3/2

Binding Energy / eV Binding Energy / eV

d) Fe 2p3/2

Fe-N

CPS (arb. units)

c) Ni 2p3/2

P a g e | 93

Catalytic Activity Testing - Total Faradic current densities

Figure S1- 9 Stationary potentiostatic catalytic activity of Cu-N-C (grey), Fe-N-C (red), Ni-N-C (blue), Mn-

N-C (cyan) and Co-N-C (black) catalysts during bulk CO2 electrolysis. Geometric current densities at a) 15

min b) 60min. Double layer capacity -normalized stationary current densities: c) 15 min d) 60 min. Lines to

guide the eye. Conditions: CO2-saturated 0.1 M KHCO3, 0.76 mg cm−2 catalyst loading

P a g e | 94

Figure S1- 10 Product efficiencies and yields. CO a) Faradaic efficiency and Absolute value of geometric

reduction current density during bulk CO2 electrolysis on Cu-N-C (gold), Fe-N-C (red), Ni-N-C (blue), Mn-

N-C (cyan) and Co-N-C (black). Lines to guide the eye. Conditions: 15 min at constant electrode potential in

CO2-saturated 0.1 M KHCO3 at 0.76 mg cm−2 catalyst loading

P a g e | 95

Figure S1- 11 Catalytic performance of Ni-N-C, Ni-C, Fe-N-C, Fe-C and N-C for CO2RR in 0.1 M CO2

saturated KHCO3 electrolyte, a) Double layer capacity, b) Absolute total geometric current density, c) Faradaic

Efficiency towards CO and d) Mass normalized CO partial current density at 60 min of CO2 bulk electrolysis.

Guide for the eye lines are shown. Catalyst loading: 0.76 mg cm−2.

Figure S1- 12 Catalytic performance of Mn-N-C, Fe-N-C, Co-N-C, Ni-N-C, Cu-N-C and Cu-Bpy for CO2RR

in 0.1 M CO2 saturated KHCO3 electrolyte. Geometric partial current density for a) H2 and b) CO at 60 min of

CO2 bulk electrolysis. Lines to guide the eye. Catalyst loading: 0.76 mg cm−2.

P a g e | 96

DFT Calculations at 0VRHE and -0.6 VRHE

Figure S1- 13 DFT-calculated free energy diagram for the CO2 reduction reaction (CO2RR) and hydrogen

evolution reaction (HER) at 0 VRHE (a,c) and -0.6 VRHE.

P a g e | 97

CO production TOF trends – Correlation of experiments with theory

Figure S1- 14 Experimental CO production turnover frequency (TOF) of the M-N-C catalysts versus applied

iR-corrected electrode potential (see Equation SI 3). The catalytic reactivity trends split into 3 potential regions

with distinctly different rate-determining mechanistic features. Insets: Region 1: Low overpotentials, the

experimental onset potential of CO production correlates with the binding energy of the reaction intermediate

COOH*. Region 2: Intermediate over-potentials, CO production TOF at -0.6 VRHE vs. free energy of adsorbed

CO, CO*; Region 3: High overpotentials, free energy diagrams for the HER (dashed paths) and CO2RR (solid

paths) at -0.8 VRHE for each M-N-C catalyst. HER barriers are high for Ni and Cu, while CO2RR is downhill

making these materials favorable CO producing catalysts.

P a g e | 98

A2. Supplementary Information to Chapter 5

Table S2- 1 Physiochemical characterization

Sample

a) BET

m2 g-1

b) DL-Capa

μF cm-2

c) Metal

content

wt.% (ICP)

d) Metal ratio

%Mole (XPS)

N ratio

%Mole

O ratio

%Mole

S ratio

%Mole

C ratio

%Mole

e) CO

uptake

mmol g-1

Fe-N-C

634

72.4

1.05

0.82

5.93

3.55

0.53

88.92

Ni-N-C

238

22.5

10.2

0.38

3.67

1.91

1.17

92.88

N-C

174

25.3

6.68

3.39

1.03

88.9

a) BET: N2 ad/desorption-based surface area; b) Double layer capacity values, geometric area of each electrode

is 1 cm2 with 0.75 mg cm-2 catalyst loading; c) Overall metal content determined from ICP-OES measurement.

d) Interfacial atomic ratio quantified with X-ray photoelectron spectra. e) CO uptake at -80°C based on CO

chemisorption measurement.

P a g e | 99

X-Ray Diffraction profiles:

Figure S2- 1 Powder XRD patterns of the Fe-N-C materials without (1HT, 0AW), with one time (2HT, 1AW)

and with two times (3HT, 2AW) acid-washing during the synthesis approach.

Figure S2- 2 Powder XRD patterns of N-C, Fe-N-C and Ni-N-C catalysts.

10 20 30 40 50 60 70 80 90

Ni-N-C

N-C

Fe-N-C

2 / 

Intensity / a.u.

P a g e | 100

TEM image of encapsulated crystalline nanoparticles in Ni-N-C catalyst

Figure S2- 3 HR-TEM image of carbon-encapsulated crystalline nanoparticles in Ni-N-C catalyst.

BET surface area and Pore size distribution:

Figure S2- 4 a) N2 specific ad/desorption isotherm profile; b) pore size distribution of the N-C, Fe-N-C and

Ni-N-C catalysts.

P a g e | 101

Interfacial surface area determination: BETSA vs. ECSA

Figure S2- 5 Cyclic voltammetry of a) N-C, b) Fe-N-C and c) Ni-N-C catalysts conducted in N2-saturated 0.05

M K2HPO4 + 0.05 M KH2PO4 (pH=6.9) solution at scan rate 15 mV s-1, 10 mV s-1, 5 mV s-1, 1 mV s-1 to

determine the double layer capacity. Potential was scanned between -0.1 and 0.42 V vs. RHE. d) Double layer

current densities (extracted at +0.16 VRHE) on N-C, Fe-N-C and Ni-N-C catalysts at each scan rate. e)

Correlation of double layer capacity (ECSA) and the N2 adsorption derived BET surface area (BETSA).

Catalysts loading: 0.75 mg cm-2.

P a g e | 102

X-ray Photoelectron spectra:

Figure S2- 6 X-ray photoelectron spectra. a) Survey XPS spectra of the PANI derived materials with the main

features assigned and high resolution spectra of b) N1s of M-N-C catalysts according to the peak positions:

M-Nx moieties (399.7 eV), pyrrolic (401.3 eV), pyridinic (398.6 eV), quaternary (402.5 eV), and graphitic

(403.9 eV) according to ref50,61,81 c) Fe2p assignment of Fe-N-C and d) Ni2p of Ni-N-C. Please note, XPS data

on Fe-N-C and N-C catalysts were already reported in ref43, but measured with other spectrometers.

P a g e | 103

Electrochemical CO2RR in H-cell

Figure S2- 7 a) Geometric CO production current densities and b) CO faradaic efficiency as a function of

applied iR-corrected electrode potential. Lines to guide the eye. Conditions: 60 min at constant electrode

potential in CO2-saturated 0.1 M KHCO3 with 0.75 mg cm-2 catalysts loading.

Figure S2- 8 Faradaic efficiency of CH4 as a function of applied iR-corrected electrode potential. Lines to

guide the eye. Conditions: 15 min at constant electrode potential in CO2-saturated 0.1 M KHCO3 with

0.75 mg cm-2 catalysts loading.

P a g e | 104

DFT calculation of different Ni-Nx-C coordination motifs

Figure S2- 9 Free energy diagram of CO2 reduction to CO on the Ni-N-C and Fe-N-C catalysts. Influence of

hydrogenating the Ni-N4-C on the binding strength for the *COOH and *CO intermediate.

Stability test of Electrochemical CO2RR in the MFC:

Figure S2- 10 Stability test of Ni-N-C GDL for CO2RR on in MFC. a) Faradaic efficiency of CO and b) Cell

potential (voltage) as a function of stationary electrolysis time. Conditions: 20 hours at constant 200 mA cm-2

working current density in CO2-saturated 1 M KHCO3 with 1 mg cm-2 catalysts loading and 3 cm2 total

geometric electrode area. Line to guide the eye.

P a g e | 105

A3. Supplementary Information to Chapter 6

Figure S3- 1 Powder X-ray diffraction of catalysts from this project. At the bottom a diffraction pattern of

troilite from Skala et al121 is displayed.

P a g e | 106

Figure S3- 2 Typical Scanning Electron Microscopy images of a) CTRL and b) MEL PANI-Fe electro-

catalysts.

Figure S3- 3 Pore size distribution of support and catalysts.

P a g e | 107

Figure S3- 4 Cyclic voltammetry of the five Fe-N-C catalysts conducted in CO2-saturated 0.1 M KHCO3 at

various scan rates for estimation of the double layer (DL) capacity. a) CM, b) Mel, c) Urea, d) NCB and e)

Ctrl. Cycle voltammetry was performed between -0.1 and 0.42 V vs. RHE to avoid the interference of the

faradaic process and at the scan rate were 20 mV s-1, 15 mV s-1, 10 mV s-1, 5 mV s-1, 1 mV s-1. Double layer

current densities are utilized to determine the double layer capacitance, which is proportional to the double-

layer interfacial area. Catalysts loading: 0.75 mg cm-2.

Figure S3- 5 a) Double layer current densities on 5 different Fe-N-C catalysts as a function electrode potential

scan rate to extract the double layer capacity; b) Correlation of double layer capacity and the N2 adsorption

derived BET surface area.

P a g e | 108

Figure S3- 6 Combination of CO-Chemisorption measurements and interfacial Fe-Nx sites area.

020 40 60 80 100 120 140

0.0

0.5

1.0

1.5

2.0

NCB

Ctrl

Urea

Fe-Nx Active Area / m2 g-1

CO Uptake / umol g-1

Mel

P a g e | 109

Figure S3- 7 Survey XPS spectra of Fe-PANI materials with the main features assigned.

Table S3- 1 Atomic ratio of various elements determined using X-ray photoelectron spectra.

Sample

NCB

94.8

2.9

1.7

0.5

0.1

CTRL

93.3

3.3

3.1

0.2

0.1

93.1

3.9

1.6

1.3

0.2

MEL

90.9

3.6

4.8

0.5

0.2

UREA

92.8

3.7

2.0

1.3

0.2

P a g e | 110

Figure S3- 8 High resolution N 1s XPS spectra of a PANI-Mel sample and Fe-Protoporphyrin (Sigma-Aldrich).

Figure S3- 9 Fe 2p high resolution XPS spectra of Fe-PANI catalysts.

P a g e | 111

Figure S3- 10 Deconvolution of Fe 2p3/2 XPS spectrum representative of the NCB sample. Open dots show

experimental data, solid blue line – main Fe 2p3/2 peak, solid red line – shake up satellite. Vertical dashed line

shows position of Fe2p3/2 reported for Fe-porphyrin.

732 726 720 714 708 702

Fe 2p - NCB

Intensity (arb. units)

Binding energy (eV)

Fe-N

P a g e | 112

Figure S3- 11 Fe K-edge XANES (a) and EXAFS (b) spectra of Fe-PANI MEL sample, iron foil, commercial

iron protoporphyrin (FePP, Sigma Aldrich) and the most common iron oxides. The intensity of iron foil

spectrum in (b) is reduced by factor of 3 for better display.

P a g e | 113

Figure S3- 12 Fe K-edge k2-weighted EXAFS data of Fe-PANI samples in k-space.

P a g e | 114

Figure S3- 13 FeK-edge EXAFS analysis as exemplified by CM sample. Panel a: the fit in the Fourier

Transformed space, solid black line - raw data, dashed red line - fitting model. Panel b shows the corresponding

curves in the k-space along with the Fe-N and Fe-N single scattering components.

P a g e | 115

Figure S3- 14 BET normalized CO partial current density as a function of a) Graphitic N content and b) Pyrrolic

N content.

Figure S3- 15 BET-normalized CO partial current density as a function of a) Nitrogen content, b) Oxygen

content and c) Sulfate Content.

P a g e | 116

Figure S3- 16 Reaction mechanism and DFT free energy diagram of CO2RR to CO on Fe-Nx site and pyridinic

nitrogen. (Free Energy data are adapted from [45] and [51])

P a g e | 117

A4. Supplementary Information to Chapter 7

Table S4- 1 Physiochemical characterization

Sample

a)BET surface

area

(m2 g-1)

b)Double layer

capacity

(μF cm-2)

c)Fe content

(wt. %)

ICP

d)CO uptake

(mmol g-1)

CO-Chemisorption

e)N content (wt. %)

element analysis

Fe-N-C

634

25.3

1.05

~6.5

N-C

214

72.4

a) BET: N2 ad/desorption-based surface area; b) Double layer capacity values, geometric area of each electrode

is 1 cm2 with 0.75 mg cm-2 catalyst loading; c) Overall metal content determined from ICP-OES measurement.

d) CO uptake value examined according to CO chemisorption. e) Nitrogen content measured with elemental

analysis.

X-Ray Diffraction:

Figure S4- 1 Powder XRD patterns of Fe-N-C and N-C catalysts. Data of Fe-N-C and N-C has been reported

in Figure S2-1.

10 20 30 40 50 60 70 80 90

N-C

Fe-N-C

2 / 

Intensity / a.u.

P a g e | 118

TEM Images:

Figure S4- 2 Represented TEM images of as prepare a)-b) Fe-N-C and c)-d) N-C catalysts. Scale bar: left

column 100 nm, right column 20 nm. Identical catalysts have been presented in Chapter 5.

P a g e | 119

Interfacial surface area determination: BETSA vs. ECSA

Figure S4- 3 Cyclic voltammetry of a) Fe-N-C and b) N-C catalysts conducted in N2-saturated 0.05 M K3PO4

+ 0.05 M H3PO4 solution at scan rate 15 mV s-1, 10 mV s-1, 5 mV s-1, 1 mV s-1 to determine the double layer

capacity. Potential was scanned between -0.1 and 0.42 V vs. RHE. c) Double layer current densities (extracted

at +0.16 VRHE) on Fe-N-C and N-C catalysts at each scan rate. d) Correlation of double layer capacity (ECSA)

and the N2 adsorption derived BET surface area (BETSA). Catalysts loading: 0.75 mg cm-2 on glassy carbon.

P a g e | 120

Table S4- 2 Electrolysis parameters in presence of various reactants

Electrolysis

Electrolyte

Reactant

Tot. time

Ionomer

Products

Long-term

CO2RR

0.1 M KHCO3

(pH: 6.8)

20 ccm CO2

flow

1020 min

Selemion

CO, CH4,

CH3OH, CH2O

Long-term

CO2RR

0.05 M KH2PO4 + 0.05 M

K2HPO4 (pH: 6.9)

20 ccm CO

flow

480 min

Selemion

CH4, CH3OH,

CH2O

CO2RR

0.05 M KH2PO4 + 0.05 M

K2HPO4 (pH: 6.4)

20 ccm CO2

flow

75 min

Nafion 117

CO, CH4

CORR

0.05 M KH2PO4 + 0.05 M

K2HPO4 (pH: 6.9)

20 ccm CO

flow

75 min

Nafion 117

CH4

CH2ORR

0.05 M KH2PO4 + 0.05 M

K2HPO4 (pH: 6.9)

1 mM CH2O

6 ccm N2 flow

75 min

Nafion 117

CH4

Methanol RR

0.05 M KH2PO4 + 0.05 M

K2HPO4 (pH: 6.9)

1 mM CH3OH

6 ccm N2 flow

75 min

Nafion 117

N/A

Formate RR

0.05 M KH2PO4 + 0.05 M

K2HPO4 (pH: 6.7)

1 mM HCOOH

6 ccm N2 flow

75 min

Nafion 117

N/A

P a g e | 121

Figure S4- 4 Catalytic performance of long term CO2RR on Fe-N-C catalyst in CO2 purged 0.1 M KHCO3. a)

Geometric working current density and applied iR-free potential as a function of reaction time. b) Faradaic

efficiency of measured products. Catalyst loading: 0.75 mg cm-2 on glassy carbon electrode. Measurement

parameters see Table S4-2.

Figure S4- 5 Catalytic performance of long term CORR on Fe-N-C catalyst in CO purged phosphate solution

with neutral initial pH value. a) Geometric working current density and applied iR-free potential as a function

of reaction time. b) Faradaic efficiency of measured products. Catalyst loading: 0.75 mg cm-2 on glassy carbon

electrode. Measurement parameters see Table S4-2.

P a g e | 122

Figure S4- 6 Products distribution as a function of applied iR-free potentials on Fe-N-C and N-C catalysts. a)

CO production rate and b) faradaic efficiency during CO2 reduction. Data points are standard Mean and Error

obtained from 15 min, 45 min and 75 min of the bulk electrolysis. Catalyst loading: 0.75 mg cm-2 on glassy

carbon electrode.

P a g e | 123

Figure S4- 7 Catalytic performance of CH2O reduction on Fe-N-C catalyst with different initial CH2O

concentrations. a) Geometric current densities, b) CH4 production rate and c) faradaic efficiency. Data are

averages over 75 min electrolysis. Reaction conditions: 6 ccm N2 purged 0.05 M K3PO4 + 0.05 M H3PO4

electrolyte in presence of CH2O. Catalyst loading: 0.75 mg cm-2 on glassy carbon.

P a g e | 124

Figure S4- 8 Logarithm of CH4 formation rate during CH2ORR versus Logarithm of the CH2O concentration

in phosphate neutral buffer solution.

P a g e | 125

Figure S4- 9 Reaction rate of a) CO2 reduction to CO, b) CO2 reduction to CH4, c) CH2O reduction to CH4,

and d) CO reduction to CH4 as a function of iR-free potential in NHE scale. Data in a) and b) are adapted

from the work.116

P a g e  | IX 
 
 
Table of figures and schemes 
Figure 1- 1 Schematic electrochemical CO2 reduction accompanied with industrial plants and renewable electricity. 1 
 
Figure 2- 1 a) The CO2RR products spectrum classified by ΔEH* descriptor. Metals prefer HER (marked in red) due to 
strong H* binding (having HUPD), while metals favor CO2RR owing to weak H* binding (not having HUPD). b) The 
binding strength diagram of CO* and H*. Hydrocarbons formation occurs via protonation of intermediate CO 
(CO*), which require moderated binding energy to CO* and H* as Cu. Figures are adapted from reference 15 
with the permission of copyright 2017, John Wiley and Sons.    Products spectrum is obtained from reference 4.
 ........................................................................................................................................................................... 5 
Figure 2- 2 a) Illustration of proposed active M-N4-C motif. b) Relation between free energy of H* (ΔG*H) and COOH* 
(ΔG*COOH) (gray circles) as well as H* (ΔG*H) and OCHO* (ΔG*OCHO) (red diamonds) on various metal-Porphyrins. 
The diagonal line (black dashed) separates selectivity towards HER and CO2RR (into CO). Figure b) is adapted 
with permission from 26, Copyright 2017, American Chemical Society. c-d) Free energy diagram of first binding 
via COOH* (red) or H* (blue) and then binding a second H* at a nearby site for the relevant c) Cu metal surface 
and d) Fe-porphyrin like motif. Figures are reproduced according to reference 15. Copyright 2017, Elsevier. ..... 6 
Figure 2- 3 Synthesis strategy of metal nitrogen-doped carbon (M-N-C) catalysts. ..................................................... 7 
Figure 2- 4 a) Relation of CO formation current densities at -0.9 VRHE and COOH* binding at 0 VRHE. Figure is 
reproduced from reference56 with Copyright permission 2018, Royal Society of Chemistry. b) Fitted *CO 
desorption (*CO→CO, red line) and *COOH formation (CO2→*COOH, black line) trends as a function of CO* 
binding energy over all five Metal-Pc electrodes. Figure is adapted from reference 59, Copyright 2018, John 
Wiley and Sons. .................................................................................................................................................10 
 
Figure 3- 1 Operando XAS cell used in this work: a) schematic illustration of the cell, 1 – working electrode (GDE 
with the ink sample) sealed with Kapton tape, 2 – Pt gauze counter electrode, 3 – leak-free Ag/AgCl reference 
electrode. Designed by Roldan Cuenya’s Group at FHI. b) The cell during measurements at SAMBA beamline of 
SOLEIL synchrotron light source (Paris, France). ................................................................................................13 
Figure 3- 2 a) Preparation of the electrode: from catalyst powder via ink to layer. Catalyst loading: ~0.75 mg cm-2 on 
GC plate. b) Schematic H-type two compartments cell divided by a polymer membrane. In operation, the 
CO2RR as well as other cathodic reactions occur at the WE (working electrode), whereas the OER happens at 
the CE (count electrode). ...................................................................................................................................15 
Figure 3- 3 Schematic of online gas chromatograph testing platform. Detailed schematic of each part, 1. Loop valve, 
2. Column, 3. Thermal conductivity detector, 5. Flame ionization detector, is displayed in Figure 3-4 below. ...18 
Figure 3- 4 a) and b) Schematic of the 10-Valve for gas sampling in the GC, a) loading the sample in the loop and b) 
dosing and transporting the loaded sample into the column. c) Schematic of gas sample (mixture of various 
compounds) carried by Argon flow, flowing in the column (colored in green). For instance, H2 (blue) with 
better mobility, moves faster in the column and accordingly takes less retention time, therefore could be 
detected as the first compound. On the contrary, CO2 molecules (grey) with lower mobility, suffers longer 

P a g e  | X 
 
retention time in the column. Schematic of the detectors e) thermal conductivity detector, and e) flame 
ionization detector. ...........................................................................................................................................19 
 
Figure 4- 1 Visualization, porosity and illustration of the M-N-C catalyst. a) Typical SEM image of the family of 
Nitrogen-coordinated metal-doped (M-N-C) carbon electro-catalysts, scale bar: 4 m; b) CO2 physisorption 
isotherm (273 K); inset: the pore size distribution; c) Materials model and a schematic local structure. ...........24 
Figure 4- 2 High-resolution XPS characterization. N-1s XPS core level region of (a) Co, (b) Mn, (c) Ni and (d) Fe doped 
M-N-C catalyst. The 2p3/2 spectra of the corresponding metal peaks (Co-2p, Mn-2p, Ni-2p, Fe-2p) is shown in 
Supplementary Figure S1-8. ..............................................................................................................................26 
Figure 4- 3 CO2 reduction reaction activities. Linear sweep voltammetry of a) Mn-N-C, b) Fe-N-C, c) Co-N-C, d) Ni-N-C 
and e) Cu-N-C in CO2-saturated 0.1 M KHCO3 (solid lines) and in N2-saturated 0.1 M KH2PO4/K2HPO4 (dashed 
lines) with a catalyst loading of 0.76 mg cm-2 at 5 mV s-1 in cathodic direction. .................................................28 
Figure 4- 4 Catalytic performance and product analysis. (a-c) Faradaic Efficiencies (FE) vs. applied, iR-corrected 
electrode potential of a) H2, b) CO and c) CH4. d) Catalyst mass-normalized CO partial currents (mass activity) 
vs. applied potential for the five M-N-C catalysts compared to state-of-art Au catalysts (performance ranges of 
Au-nanoparticle and Au-nanowires are shown by filled areas 10,11. Lines to guide the eye. Conditions: 60 min at 
constant electrode potential in CO2-saturated 0.1 M KHCO3 with 0.76 mg cm-2 M-N-C catalysts loading. 
Faradaic efficiencies and CO yields after 15 min are shown in Figure S1-10. .....................................................29 
Figure 4- 5 Experimental correlation to simulations. Experimental CO production turnover frequency (TOF) of the M-
N-C catalysts versus applied iR-corrected electrode potential. The a) catalytic reactivity trends and b) reaction 
pathway split into three potential regions with distinctly different rate-determining mechanistic features. Free 
energy diagrams for HER and CO2RR at 0.0 and -0.6 VRHE are given in Figure S1-13. Insets: Region 1: Low 
overpotentials, the experimental onset potentials of CO production (better seen on the log (CO TOF) – E plot in 
Figure S1-14) correlate with the binding energy of the reaction intermediate COOH* taken from Figure S1-13. 
Region 2: Intermediate over-potentials, CO production TOF at -0.6 VRHE correlates with the free energy of 
adsorbed CO, CO* taken from Figure S1-13; Region 3: High overpotentials, free energy diagrams for the HER 
(dashed paths) and CO2RR (solid paths) at -0.8 VRHE for each M-N-C catalyst. HER barriers are high for Ni and 
Cu, while CO2RR is downhill making these materials favorable CO producing catalysts. ...................................31 
 
Figure 5- 1 Synthesis procedure of our studied polyaniline (PANI) based M-N-C catalysts. ........................................38 
Figure 5- 2 Illustration of M-N-C catalysts. Representative TEM images of a) N-C, b) Fe-N-C and c) Ni-N-C catalysts. 
Inserts: HR-TEM images of as-prepared catalysts. Figure S2-3 represents the thickness (over 10 nm) of carbon 
layers encapsulating the inorganic Nickel species. ............................................................................................39 
Figure 5- 3 Catalytic performance and product analysis on N-C (black), Fe-N-C (red), Ni-N-C (blue) and AgOx (Cyan) 
catalysts. a) Absolute geometric current densities; b) geometric CO production current densities; c) CO faradaic 
efficiency as a function of applied iR-corrected electrode potential at 15 min of each electrolysis (CO partial 
current densities and faradaic efficiency at 60 min and faradaic CH4 yield are shown in Figure S2-7, S2-8). d) 
Geometric CO production current densities and e) CO faradaic efficiency during the long-term stability testing 

P a g e  | XI 
 
as a function of stationary electrolysis time. Lines to guide the eye. Conditions: CO2-saturated 0.1 M KHCO3 (pH 
6.8) with 0.75 mg cm-2 catalysts loading. ...........................................................................................................41 
Figure 5- 4 Free energy diagram of CO2 reduction to CO on Ni-N-C and Fe-N-C catalysts. a) Chemical structure of the 
M-Nx moieties considered, b) influence of the Ni-coordination on the binding strength for the *COOH and *CO 
intermediates, c) Free energy diagram of CO2 reduction to CO and d) hydrogen evolution reaction, d) on Fe-N4-
C (red), Ni-N4-C (blue) and Ag (111) catalyst (cyan)............................................................................................43 
Figure 5- 5 Schematic of a) Ni-N-C Gas Diffusion Electrode (GDE) and b) typical H-type liquid cell. c) Experimental 
faradaic CO efficiency as function of the applied electrolyzer current density in CO2 saturated (50 ccm) 1 M 
KHCO3 solution. Prolonged flow cell CO2RR testing at 200 mA cm-2 working current density is presented in 
Figure S2-10. .....................................................................................................................................................45 
 
Figure 6- 1 Comparison of Fe-N-C catalysts based on different secondary nitrogen precursors showing bulk iron, 
nitrogen, and sulfur content as measured by ICP and Elemental Analysis and surface content (ca. 2-3 nm) as 
measured by XPS. Catalysts ordered by increasing surface nitrogen content (XPS). ..........................................50 
Figure 6- 2 (a) High resolution N 1s XPS data of various PANI-derived Fe-N-C samples with different N precursors. (b) 
Example of the deconvolution of a N 1s spectrum acquired for the MEL sample. N1s assignment of Fe-PP ref. 
sample is presented in Figure S3-8, showing identical BE (399.8eV) as the Nx-Fe moiety in MEL sample. ..........51 
Figure 6- 3 a) Fe K-edge XANES and b) EXAFS spectra of selected Fe-PANI samples, dotted lines in b) show fitted 
models. XANES and EXAFS spectra of referenced FeOx, Fe foil and FePP (Sigma-Aldrich) are shown in Figure S3-
11. Fe K-edge k2-weighted EXAFS data of Fe-PANI samples in k-space and analysis as exemplified by CM 
sample are shown in Figure S3-12, S3-13. ..........................................................................................................53 
Figure 6- 4 CO2 reduction data for various Fe-N-C catalysts based on different secondary nitrogen precursors: (a) CO 
generation rate, (b) faradaic efficiency towards CO production. Experimental conditions: CO2 saturated 0.1 M 
KHCO3, catalyst loading: 0.75 mg cm-2 on Glassy Carbon. ..................................................................................54 
Figure 6- 5 Trends of CO current density on the plateau and in the kinetic region (-0.53 V vs RHE) varying with BET 
specific surface area. .........................................................................................................................................55 
Figure 6- 6 Trends of CO current densities in the kinetic region (-0.53 V vs RHE) varying with (a) various pyridinic 
nitrogen and N-Fe and (b) surface Fe content. Current density is normalized to the specific surface area as 
calculated by the BET method. BET-normalized CO current densities in the kinetic region as a function of other 
functionalities are shown in Figure S3-14 and Figure S3-15. Free Energy Diagrams from CO2 to CO over FeNx 
and Pyridinic-N sites are shown in Figure S3-16, data are adapted from Refs.45,51 .............................................56 
Figure 6- 7 Fe K-edge XANES (a) and EXAFS (b) spectra taken under operando conditions in CO2-saturated 0.1 M 
KHCO3 at -0.5 V (solid blue curves), -0.9 V (dashed green curves) and -1.1 V (red dot-dashed curves) vs. RHE a – 
XANES. (c) CH4 faradaic efficiency from CO2RR varying with applied working potential. Lines are added to 
indicate points representative of spectra in (a) and (b). ....................................................................................57 
 
Figure 7- 1 Overall electrochemical performance under different conditions and in presence of various reactants in 
neutral 0.05M K3PO4 + 0.05M H3PO4 solution. a) Linear sweep voltammetry at -5 mV s-1 potential scan rate and 
b) geometric current density during each bulk electrolysis. Presented dots data are averages calculated from 

P a g e  | XII 
 
15 min, 45 min and 75 min of the stationary electrochemical reaction. Line to guide the eye. Catalyst loading: 
0.75 mg cm-2 on glassy carbon. ..........................................................................................................................64 
Figure 7- 2 Methane a) production rate and b) faradaic efficiency as a function of applied IR-free potential during 
the electrochemical CO2 (saturated, 30 mM), CO (saturated, 1 mM) and CH2O (1 mM) reduction reactions on 
Fe-N-C (solid dots) and metal free N-C (empty) catalysts in neutral 0.05 M K2HPO4 + 0.05 M KH2PO4 buffer 
solution. Data points are averages obtained from 15 min, 45 min and 75 min of each bulk electrolysis. Line to 
guide the eye. Catalyst loading: 0.75 mg cm-2 on glassy carbon plate. ...............................................................65 
Figure 7- 3 Production rate of CH4 at various pH as a function of iR-corrected applied electrode potentials. a) CO 
reduction and b) CH2O reduction plotted on the RHE scale. c) CO reduction and d) CH2O reduction plotted on 
the NHE scale. e) logarithm of the CH4 formation rate from CH2O at different pH versus applied potential. 
Electrolytes are 0.05 M K2HPO4 + 0.05 M K3PO4 (pH = 11.9), 0.05 M K3PO4 + 0.05 M H3PO4 (pH = 6.9), and 0.05 
M KH2PO4 + 0.05 M H3PO4 (pH = 2.25) for pH variation. Data are averages over 75 min electrolysis. Line to 
guide the eye. Catalyst loading: 0.75 mg cm-2 on glassy carbon. ........................................................................67 
Figure 7- 4 Free energy diagram towards CH4 from CO2, CO and CH2O on Fe-N-C at 0 VRHE. The three limiting 
potential steps are shown by V1, V2 and V3, with the reduction of CH2O having the smallest limiting potential 
step in line with the experiments. .....................................................................................................................69 
Figure 7- 5 The detailed protonation steps of *CO towards *CHO via the co-adsorption of *CO + *H on a) Cu (111) 
facet and on b) Fe-N-C, respectively. .................................................................................................................71 
Figure 7- 6 Hypothesized paths of methanol and methane formation on single-site Fe-N-C catalyst. ........................72 
 
Scheme 8- 1 Summarized structure of this dissertation. ............................................................................................74 
Scheme 8- 2 Overall reaction network form CO2 reduction towards CH4 via CO and CH2O over the Fe-N-C catalyst. 
PCET: Proton Coupled Electron Transfer; PDET: Proton Decoupled Electron Transfer. ......................................75 
 
Figure S1- 1 Typical SEM image of the M-N-C electrocatalysts. Scale bar: 10 m. ......................................................86 
Figure S1- 2 N2 physisorption isotherms for this family of M-N-C electrocatalysts. As shown, the isotherms are type I 
indicating the microporous structures; the abrupt increase in the higher pressure regime indicating the 
existence of macropores. In short, this information confirms the rich of microporosity, and indicates the 
hierarchical structure. .......................................................................................................................................86 
Figure S1- 3 a) to c) Cyclic voltammetry of the five M-N-C catalysts conducted in CO2 -saturated 0.1 M KHCO3 at 
various scan rates for estimation of double layer (DL) capacity. d) current densities from a-c plotted vs. 
electrode potential scan rate to extract the double layer capacity; e) Correlation of double layer capacity and 
the N2 adsorption –derived BET surface area. ...................................................................................................87 
Figure S1- 4 Water vapor adsorption isotherms for this family of M-N-C electrocatalysts (top) with highlighted low 
pressure regime (bellow) together with typical benchmark samples, including hydrophilic zeolite 13X, most 
hydrophilic carbon as well as porous carbons (mesoporous CMK-3, commercial activated carbon Norit Super) 
that commonly show hydrophobic nature. As shown, the surface of this family of M-N-C is medium hydrophilic 
or hydrophobic as compared with the typical benchmark materials. ................................................................88 
Figure S1- 5 XRD patterns of this family of M-N-C catalysts. ......................................................................................89 

P a g e  | XIII 
 
Figure S1- 6 STEM images and elemental maps for Fe-N-C, Co-N-C and Ni-N-C; Scale bar in left column: 200 nm, 
middle column: 20 nm. ......................................................................................................................................90 
Figure S1- 7 High-resolution Cu 2p (a) and N 1s (b) XP spectra of the Cu-N-C sample. See main text for N 1s peak 
assignment. .......................................................................................................................................................91 
Figure S1- 8 High-resolution metal 2p3/2 of Co-N-C (a), Mn-N-C (b), Ni-N-C (c) and Fe-N-C (d). Blue lines represent 
2p3/2 main peaks, red – the corresponding shake-up satellites. .......................................................................92 
Figure S1- 9 Stationary potentiostatic catalytic activity of Cu-N-C (grey), Fe-N-C (red), Ni-N-C (blue), Mn-N-C (cyan) 
and Co-N-C (black) catalysts during bulk CO2 electrolysis. Geometric current densities at a) 15 min b) 60min. 
Double layer capacity -normalized stationary current densities: c) 15 min d) 60 min. Lines to guide the eye. 
Conditions: CO2-saturated 0.1 M KHCO3, 0.76 mg cm−2 catalyst loading ............................................................93 
Figure S1- 10 Product efficiencies and yields. CO a) Faradaic efficiency and Absolute value of geometric reduction 
current density during bulk CO2 electrolysis on Cu-N-C (gold), Fe-N-C (red), Ni-N-C (blue), Mn-N-C (cyan) and 
Co-N-C (black). Lines to guide the eye. Conditions: 15 min at constant electrode potential in CO2-saturated 0.1 
M KHCO3 at 0.76 mg cm−2 catalyst loading ........................................................................................................94 
Figure S1- 11 Catalytic performance of Ni-N-C, Ni-C, Fe-N-C, Fe-C and N-C for CO2RR in 0.1 M CO2 saturated KHCO3 
electrolyte, a) Double layer capacity, b) Absolute total geometric current density, c) Faradaic Efficiency 
towards CO and d) Mass normalized CO partial current density at 60 min of CO2 bulk electrolysis. Guide for the 
eye lines are shown. Catalyst loading: 0.76 mg cm−2. ........................................................................................95 
Figure S1- 12 Catalytic performance of Mn-N-C, Fe-N-C, Co-N-C, Ni-N-C, Cu-N-C and Cu-Bpy for CO2RR in 0.1 M CO2 
saturated KHCO3 electrolyte. Geometric partial current density for a) H2 and b) CO at 60 min of CO2 bulk 
electrolysis. Lines to guide the eye. Catalyst loading: 0.76 mg cm−2. .................................................................95 
Figure S1- 13 DFT-calculated free energy diagram for the CO2 reduction reaction (CO2RR) and hydrogen evolution 
reaction (HER) at 0 VRHE (a,c) and -0.6 VRHE.........................................................................................................96 
Figure S1- 14 Experimental CO production turnover frequency (TOF) of the M-N-C catalysts versus applied iR-
corrected electrode potential (see Equation SI 3). The catalytic reactivity trends split into 3 potential regions 
with distinctly different rate-determining mechanistic features. Insets: Region 1: Low overpotentials, the 
experimental onset potential of CO production correlates with the binding energy of the reaction intermediate 
COOH*. Region 2: Intermediate over-potentials, CO production TOF at -0.6 VRHE vs. free energy of adsorbed 
CO, CO*; Region 3: High overpotentials, free energy diagrams for the HER (dashed paths) and CO2RR (solid 
paths) at -0.8 VRHE for each M-N-C catalyst. HER barriers are high for Ni and Cu, while CO2RR is downhill 
making these materials favorable CO producing catalysts. ................................................................................97 
 
Figure S2- 1 Powder XRD patterns of the Fe-N-C materials without (1HT, 0AW), with one time (2HT, 1AW) and with 
two times (3HT, 2AW) acid-washing during the synthesis approach. ................................................................99 
Figure S2- 2 Powder XRD patterns of N-C, Fe-N-C and Ni-N-C catalysts. .....................................................................99 
Figure S2- 3 HR-TEM image of carbon-encapsulated crystalline nanoparticles in Ni-N-C catalyst............................. 100 
Figure S2- 4 a) N2 specific ad/desorption isotherm profile; b) pore size distribution of the N-C, Fe-N-C and Ni-N-C 
catalysts. ......................................................................................................................................................... 100 

P a g e  | XIV 
 
Figure S2- 5 Cyclic voltammetry of a) N-C, b) Fe-N-C and c) Ni-N-C catalysts conducted in N2-saturated 0.05 M 
K2HPO4 + 0.05 M KH2PO4 (pH=6.9) solution at scan rate 15 mV s-1, 10 mV s-1, 5 mV s-1, 1 mV s-1 to determine the 
double layer capacity. Potential was scanned between -0.1 and 0.42 V vs. RHE. d) Double layer current 
densities (extracted at +0.16 VRHE) on N-C, Fe-N-C and Ni-N-C catalysts at each scan rate. e) Correlation of 
double layer capacity (ECSA) and the N2 adsorption derived BET surface area (BETSA). Catalysts loading: 0.75 
mg cm-2. .......................................................................................................................................................... 101 
Figure S2- 6 X-ray photoelectron spectra. a) Survey XPS spectra of the PANI derived materials with the main 
features assigned and high resolution spectra of b) N1s of M-N-C catalysts according to the peak positions: M-
Nx moieties (399.7 eV), pyrrolic (401.3 eV), pyridinic (398.6 eV), quaternary (402.5 eV), and graphitic (403.9 eV) 
according to ref50,61,81 c) Fe2p assignment of Fe-N-C and d) Ni2p of Ni-N-C. Please note, XPS data on Fe-N-C and 
N-C catalysts were already reported in ref43, but measured with other spectrometers. .................................. 102 
Figure S2- 7 a) Geometric CO production current densities and b) CO faradaic efficiency as a function of applied iR-
corrected electrode potential. Lines to guide the eye. Conditions: 60 min at constant electrode potential in 
CO2-saturated 0.1 M KHCO3 with 0.75 mg cm-2 catalysts loading. .................................................................... 103 
Figure S2- 8 Faradaic efficiency of CH4 as a function of applied iR-corrected electrode potential. Lines to guide the 
eye. Conditions: 15 min at constant electrode potential in CO2-saturated 0.1 M KHCO3 with 0.75 mg cm-2 
catalysts loading. ............................................................................................................................................. 103 
Figure S2- 9 Free energy diagram of CO2 reduction to CO on the Ni-N-C and Fe-N-C catalysts. Influence of 
hydrogenating the Ni-N4-C on the binding strength for the *COOH and *CO intermediate. ............................ 104 
Figure S2- 10 Stability test of Ni-N-C GDL for CO2RR on in MFC. a) Faradaic efficiency of CO and b) Cell potential 
(voltage) as a function of stationary electrolysis time. Conditions: 20 hours at constant 200 mA cm-2 working 
current density in CO2-saturated 1 M KHCO3 with 1 mg cm-2 catalysts loading and 3 cm2 total geometric 
electrode area. Line to guide the eye. ............................................................................................................. 104 
 
Figure S3- 1 Powder X-ray diffraction of catalysts from this project. At the bottom a diffraction pattern of troilite 
from Skala et al121 is displayed. ....................................................................................................................... 105 
Figure S3- 2 Typical Scanning Electron Microscopy images of a) CTRL and b) MEL PANI-Fe electro-catalysts. .......... 106 
Figure S3- 3 Pore size distribution of support and catalysts. .................................................................................... 106 
Figure S3- 4 Cyclic voltammetry of the five Fe-N-C catalysts conducted in CO2-saturated 0.1 M KHCO3 at various scan 
rates for estimation of the double layer (DL) capacity. a) CM, b) Mel, c) Urea, d) NCB and e) Ctrl. Cycle 
voltammetry was performed between -0.1 and 0.42 V vs. RHE to avoid the interference of the faradaic process 
and at the scan rate were 20 mV s-1, 15 mV s-1, 10 mV s-1, 5 mV s-1, 1 mV s-1. Double layer current densities are 
utilized to determine the double layer capacitance, which is proportional to the double-layer interfacial area. 
Catalysts loading: 0.75 mg cm-2. ...................................................................................................................... 107 
Figure S3- 5 a) Double layer current densities on 5 different Fe-N-C catalysts as a function electrode potential scan 
rate to extract the double layer capacity; b) Correlation of double layer capacity and the N2 adsorption derived 
BET surface area. ............................................................................................................................................. 107 
Figure S3- 6 Combination of CO-Chemisorption measurements and interfacial Fe-Nx sites area. ............................ 108 
Figure S3- 7 Survey XPS spectra of Fe-PANI materials with the main features assigned. .......................................... 109 

P a g e  | XV 
 
Figure S3- 8 High resolution N 1s XPS spectra of a PANI-Mel sample and Fe-Protoporphyrin (Sigma-Aldrich). ........ 110 
Figure S3- 9 Fe 2p high resolution XPS spectra of Fe-PANI catalysts. ........................................................................ 110 
Figure S3- 10 Deconvolution of Fe 2p3/2 XPS spectrum representative of the NCB sample. Open dots show 
experimental data, solid blue line – main Fe 2p3/2 peak, solid red line – shake up satellite. Vertical dashed 
line shows position of Fe2p3/2 reported for Fe-porphyrin. ............................................................................. 111 
Figure S3- 11 Fe K-edge XANES (a) and EXAFS (b) spectra of Fe-PANI MEL sample, iron foil, commercial iron 
protoporphyrin (FePP, Sigma Aldrich) and the most common iron oxides. The intensity of iron foil spectrum in 
(b) is reduced by factor of 3 for better display. ............................................................................................... 112 
Figure S3- 12 Fe K-edge k2-weighted EXAFS data of Fe-PANI samples in k-space. .................................................... 113 
Figure S3- 13 FeK-edge EXAFS analysis as exemplified by CM sample. Panel a: the fit in the Fourier Transformed 
space, solid black line - raw data, dashed red line - fitting model. Panel b shows the corresponding curves in 
the k-space along with the Fe-N and Fe-N single scattering components. ....................................................... 114 
Figure S3- 14 BET normalized CO partial current density as a function of a) Graphitic N content and b) Pyrrolic N 
content............................................................................................................................................................ 115 
Figure S3- 15 BET-normalized CO partial current density as a function of a) Nitrogen content, b) Oxygen content and 
c) Sulfate Content. ........................................................................................................................................... 115 
Figure S3- 16 Reaction mechanism and DFT free energy diagram of CO2RR to CO on Fe-Nx site and pyridinic 
nitrogen. (Free Energy data are adapted from [45] and [51]) ............................................................................. 116 
 
Figure S4- 1 Powder XRD patterns of Fe-N-C and N-C catalysts. Data of Fe-N-C and N-C has been reported in Figure 
S2-1. ................................................................................................................................................................ 117 
Figure S4- 2 Represented TEM images of as prepare a)-b) Fe-N-C and c)-d) N-C catalysts. Scale bar: left column 100 
nm, right column 20 nm. Identical catalysts have been presented in Chapter 5. ............................................. 118 
Figure S4- 3 Cyclic voltammetry of a) Fe-N-C and b) N-C catalysts conducted in N2-saturated 0.05 M K3PO4 + 0.05 M 
H3PO4 solution at scan rate 15 mV s-1, 10 mV s-1, 5 mV s-1, 1 mV s-1 to determine the double layer capacity. 
Potential was scanned between -0.1 and 0.42 V vs. RHE. c) Double layer current densities (extracted at +0.16 
VRHE) on Fe-N-C and N-C catalysts at each scan rate. d) Correlation of double layer capacity (ECSA) and the N2 
adsorption derived BET surface area (BETSA). Catalysts loading: 0.75 mg cm-2 on glassy carbon. ................... 119 
Figure S4- 4 Catalytic performance of long term CO2RR on Fe-N-C catalyst in CO2 purged 0.1 M KHCO3. a) Geometric 
working current density and applied iR-free potential as a function of reaction time. b) Faradaic efficiency of 
measured products. Catalyst loading: 0.75 mg cm-2 on glassy carbon electrode. Measurement parameters see 
Table S4-2. ...................................................................................................................................................... 121 
Figure S4- 5 Catalytic performance of long term CORR on Fe-N-C catalyst in CO purged phosphate solution with 
neutral initial pH value. a) Geometric working current density and applied iR-free potential as a function of 
reaction time. b) Faradaic efficiency of measured products. Catalyst loading: 0.75 mg cm-2 on glassy carbon 
electrode. Measurement parameters see Table S4-2. ..................................................................................... 121 
Figure S4- 6 Products distribution as a function of applied iR-free potentials on Fe-N-C and N-C catalysts. a) CO 
production rate and b) faradaic efficiency during CO2 reduction. Data points are standard Mean and Error 

P a g e  | XVI 
 
obtained from 15 min, 45 min and 75 min of the bulk electrolysis. Catalyst loading: 0.75 mg cm-2 on glassy 
carbon electrode. ............................................................................................................................................ 122 
Figure S4- 7 Catalytic performance of CH2O reduction on Fe-N-C catalyst with different initial CH2O concentrations. 
a) Geometric current densities, b) CH4 production rate and c) faradaic efficiency. Data are averages over 75 
min electrolysis. Reaction conditions: 6 ccm N2 purged 0.05 M K3PO4 + 0.05 M H3PO4 electrolyte in presence of 
CH2O. Catalyst loading: 0.75 mg cm-2 on glassy carbon. ................................................................................... 123 
Figure S4- 8 Logarithm of CH4 formation rate during CH2ORR versus Logarithm of the CH2O concentration in 
phosphate neutral buffer solution. ................................................................................................................. 124 
Figure S4- 9 Reaction rate of a) CO2 reduction to CO, b) CO2 reduction to CH4, c) CH2O reduction to CH4, and d) CO 
reduction to CH4 as a function of iR-free potential in NHE scale. Data in a) and b) are adapted from the work.116
 ........................................................................................................................................................................ 125 
 
   

P a g e  | XVII 
 
Table of tables 
Table 3- 1 Amount of chemical substances used in the synthesis protocols to prepare the M-N-C catalysts and the 
corresponding synthesis protocols for the respective studies described, investigated and discussed later on in 
Chapters 4-7. 12 
 
Table 5- 1 Summary table of the catalytic performance towards CO2RR referred to Gas Diffusion Electrode. ...........46 
 
Table 6- 1 Characteristics of Secondary Nitrogen Precursors Used in this Work.........................................................49 
Table 6- 2 Distribution of nitrogen species (in at%) in PANI samples as seen from N 1s XPS spectra deconvolution. .51 
Table 6- 3 Best-fit parameters for the Fe K-edge EXAFS spectra of the Fe-PANI samples shown in Figure 6-3. Included 
are the coordination numbers (CN) for Fe-N and Fe-C species, and the bond lengths for the same species (r) 
and Debye-Waller factor (σ2). The values in parenthesis are the standard errors in the last digit. ....................53 
Table 6- 4 The best-fit parameters for Fe K-edge EXAFS spectra of the Fe-PANI measured under operando conditions 
are shown in Figure 6-7. Included are the coordination numbers (CN) for Fe-N and Fe-C species, and the bond 
lengths for the same species (r) and Debye-Waller factor (σ2). The values in parenthesis are the standard errors 
in the last digit. .................................................................................................................................................57 
 
Table S1- 1 Physical parameters of nitrogen-metal doped carbon (M-N-C) CO2RR electrocatalysts ...........................85 
Table S1- 2 Metal atom, nitrogen content and assignment of the different N species in the M-N-C CO2RR 
electrocatalysts extracted from XPS data. Total metal and nitrogen content are calculated from measured XP 
spectra areas using instrument-specific relative sensitivity factors provided by manufacturer (SPECS). ...........92 
 
Table S2- 1 Physiochemical characterization .............................................................................................................98 
 
Table S3- 1 Atomic ratio of various elements determined using X-ray photoelectron spectra. ................................ 109 
 
Table S4- 1 Physiochemical characterization ........................................................................................................... 117 
Table S4- 2 Electrolysis parameters in presence of various reactants ...................................................................... 120 

P a g e | XVIII

List of Abbreviations

CO2RR

(Electrochemical) CO2 reduction reaction

HER

Hydrogen reduction reaction

atm

atmospheres

(P)XRD

(Powder) X-ray diffraction

M-N-C

Metal-Nitrogen-Carbon

Hupd

Hydrogen underpotential deposition

Elemental analysis

ICP OES

Inductively Coupled Plasma Optical

Emission Spectrometry

Counter electrode

Working electrode

REF

Reference electrode

NHE

Normal hydrogen electrode

RHE

Reversible hydrogen electrode

LSV

Linear sweep voltammertry

Chronoamperometry

GDE / L

Gas diffusion electrode / layer

DFT

Density functional theory

MEA

Membrane electrode assembly

hours

min

minutes

ECSA

Electrochemical active surface area

Double layer capacity

BET

Brunauer-Emmett-Teller

ORR

Oxygen reduction reaction

EDX

Energy dispersive X-ray spectroscopy

PEIS

Potentiostatic electrochemical impedance

spectroscopy

GC-plate

Glassy carbon plate

(Non)-PGM

(Non)- Precious group metal

TEM

Transmission electron microscopy

HRTEM

High resolution TEM

XPS

X-ray photoelectron spectra

XAS

X-ray absorption spectra

XANES

X-ray absorption near edge spectroscopy

EXAFS

Extended X-ray absorption fine structure

Quadrupole splitting

TPD

Temperature programmed desorption

PCET

Proton coupled electron transfer

Gas chromatograph

HPLC

High performance liquid chromatograph

MFC

Micro flow cell

SEM

Scanning electron microscopy

Room temperature

TOF

Turn over frequency

Faradaic efficiency

Current times resistance (ohmic drop)

P a g e | XIX

List of Chemicals

Name

Acronym

Purity/Concentration

Supplier

Ultra-pure water

Milli-Q water

16.8 MΩ cm

Methanol

MeOH

anhydrous, 99.9 %

Alfa Aesar

Ethanol

EtOH

100 %

VWR Chemicals

Isopropanol

iPrOH

100 %

VWR Chemicals

n-Propanol

nPrOH

100 %

VWR Chemicals

Nafion

5 wt%

Sigma Aldrich

Nafion membrane

Sigma Aldrich

Selemion membrane

AGC Eng. Co.

Nickel chloride hexahydrate

NiCl2·6H2O

99.5 %

Sigma Aldrich

Iron chloride hexahydrate

FeCl3·6H2O

99.5 %

Sigma Aldrich

Perchloric acid

HClO4

70 % conc., 99.999 % trace metal bases

Sigma Aldrich

Sulfuric acid

H2SO4

95.0 %

VWR Chemicals

Hydrochloric acid

HCl

37.0 %

VWR Chemicals

Nitric acid

HNO3

69.0 %

Merck

Phosphoric acid

H3PO4

85 wt%

Sigma Aldrich

Potassium hydroxide

KOH

99.99%

Sigma Aldrich

Potassium bicarbonate

KHCO3

99.5%

Sigma Aldrich

Potassium dibasic phosphate

K2HPO4

99.95%

Sigma Aldrich

Potassium monobasic phosphate

KH2PO4

99.95%

Sigma Aldrich

Ammonium peroxodisulfate

(NH4)2S2O8

>98%

Merck

Formic acid

HCOOH

>95%

Sigma Aldrich

Formaldehyde

CH2O

37wt %

Merck

Carbon dioxide

CO2

99.999 %

Air Liquide

Hydrogen

99.999 %

Air Liquide

Carbon monoxide

99.997 %

Air Liquide

Nitrogen

99.999 %

Air Liquide

Oxygen

99.998 %

Air Liquide

Argon

99.999 %

Air Liquide

Helium

99.999 %

Air Liquide

P a g e | XX

List of Publications during Ph.D. study

 Page 22-35, as Chapter 4 of this dissertation:

Wen Ju, Alexander Bagger (co-first), Guangping Hao, Ilya Sinev, Ana Sofia Varela, Volodymyr Bon, Beatriz

Roldan Cuenya, Stefan Kaskel, Jan Rossmeisl, and Peter Strasser, Understanding activity and selectivity of

metal-nitrogen-doped carbon catalysts for electrochemical reduction of CO2, Nature Communications, 2017,

8, 944.

 Page 36-46, as Chapter 5 of this dissertation:

Tim Möller, Wen Ju (co-first), Alexander Bagger, Xingli Wang, Fang Luo, Trung Ngo Thanh, Ana Sofia

Varela, Jan Rossmeisl, and Peter Strasser, Efficient CO2 to CO Electrolysis on Solid Ni-N-C Catalysts at

Industrial Current Densities, Energy & Environmental Science, 2019.

 Page 47-59, as Chapter 6 of this dissertation:

Nathaniel D Leonard, Wen Ju (co-first), Ilya Sinev, Julian Steinberg, Fang Luo, Ana Sofia Varela, Beatriz

Roldan Cuenya, and Peter Strasser, The chemical identity, state and structure of catalytically active centers

during the electrochemical CO2 reduction on porous Fe–nitrogen–carbon (Fe–N–C) materials, Chemical

Science, 2018, 9, 5064–5073.

 Page 60-73, as Chapter 7 of this dissertation:

Wen Ju, Alexander Bagger (co-first), Xingli Wang, Yulin Tsai, Fang Luo, Huan Wang, Tim Möller, Jan

Rossmeisl, Ana Sofia Varela, and Peter Strasser, Unraveling mechanistic reaction pathways of the

electrochemical CO2 reduction on Fe-N-C single site catalysts, Submitted to ACS Energy Letters, accepted,

2019.