Copper-based Nanostructured Catalysts for Efficient
and Selective CO2 Electroreduction –
Synthesis, Catalytic Performance, and Mechanistic
Analysis
Vorgelegt von
MSc. Chem. Xingli Wang
ORICD: 0000-0003-2785-9707
an der Fakultät II-Mathematik und Naturwissenschaften
der Technischen Universität Berlin
zur Erlangen des akademischen Grades
Doktorin der Naturwissenschaften
-Dr. rer. nat.
genehmigte Dissertation
Promotionsausschuss:
Vorsitzende: Prof. Dr. Maria Andrea Mroginski
Gutachter: Dr. Frédéric Jaouen (U i e i de M e ie )
Gutachter: Prof. Dr. Peter Strasser
Tag der wissenschaftlichen Aussprache: 05. Juni. 2019
Berlin 2020
I
Acknowledgement
First and foremost, I would like to thank my supervisor, Prof. Dr. Peter Strasser, for offering
me the opportunity to work in his group. I appreciate all his contributions of time, ideas to
make my Ph.D. study successful.
I also would like to thank Dr. Frédéric Jaouen and Prof. Dr. Maria Andrea Mroginski, the
committee of my defense, for the insightful feedback and valuable advice.
It would not have been possible to graduate without the support and guidance that I received
from many people. I would like to deliver my sincere thank to Dr. Ana Sofia Varela, who
supported and guided me in the beginning of my Ph.D. I am greatly thankful to our CO2 team,
Wen, Tim, Jorge, Cheonghee, Lujin, Yulin, and Trung. My research would have been
impossible without the aid and support of you. I would like to thank Jorge for the grate
collaboration in DEMS and teaching me how to read/analyze the data, which are one of the
most complicated data I have ever processed. Heartfelt thanks go to our in situ TEM team,
Stefanie and Malte, for every failure and success we got during the experiments. I would like
to express my deep gratitude to Henrik, Fabio, Elisabeth, Thomas, Malte, Dr. Isaac Martens
(ESRF) and Dr. Jakub Drnec (ESRF) for their kind help with beam time, data processing and
useful discussions. I am profoundly grateful to Dr. Katharina Klingan, Shan Jiang and Prof.
Holger Dau for the helpful discussions and collaboration in operando XAS and in situ Raman
experiments. I would like to thank Dipl. Ing. Sören Selve, Dr. Ing. Dirk Berger, Dipl. Ing.
Ulrich Gernert, Dr. Tore Niermann, Jan from ZELMI and Dr. Marc Heggen from
Forschungszentrum Jülich, for their great help and support for microscopy and all discussions.
Many thanks to Vera, Camillo, Manuel, Mikaela for their kind help in both scientific work
and daily life.
I would like to thank e e y c ege f mi e i S a e ’ g up, for the stimulating discussions,
for every enjoyable moment we were working together, and for all the fun we have had in the
last four years.
II
Further, I would also like to thank Mrs. Annette Wittebrock and Benjamin Paul for ordering
chemicals and lab wares, Benjamin Paul for taking care of SEM, Mrs. Astrid Müller-Klauke
for measuring ICP-OES and Mrs. Kluge for managing the gas bottles.
I am also very grateful to all of my BIG-NSE friends Huan Sun, Yanyan, Fang, Shuang,
Chengyue, Xiaojia, Weichao, Yuwen, Huan Wang, and Yu Peng.
I especially thank my mom, dad, my grandpa, grandma, and my sister. I would not have made
it this far without them. I know I always have my family to count on when times are rough.
I would also like to say a heartfelt thank to my beloved husband, Yu, for passing through
many fine times together with me in Germany. He has been a true and great supporter and has
unconditionally loved me during my good and bad times. He has faith in me and my intellect
e e whe I fe de e ed becau e I did ’ ha e fai h i my e f. His support makes me
stronger and better than I could be.
Lastly, I am grateful to China Scholarship Council (CSC) and for financing my Ph.D. study.
III
Abstract
In the past decades, the atmospheric CO2 emissions increased with the unrestrained
combustion of fossil fuels to meet the growing energy demand, leading to serious
environmental pollution and climate issues. The electrochemical CO2 reduction reaction
(CO2RR) is a promising alternative to convert CO2 into value-added products, and has the
potential to contribute to a carbon-neutral energy cycle by using surplus electricity generated
from renewable sources (e.g., solar and wind). Among various materials, copper-based
catalysts are most studied, given their unique capability to make hydrocarbons in considerable
amounts and at reasonable overpotentials.
In this thesis, in-depth understanding of the CO2 electroreduction process was firstly
established by adjusting the local reaction environment of a CuOx nanoparticle (NP) model
catalyst. A tunable product distribution during catalytic CO2 electroreduction could be
achieved by varying areal particle densities and co-feeding CO2 reactant with CO. With
higher areal density and lower mean interparticle distances, a shift in faradic efficiency
towards C2H4 over CH4 was observed, which was attributed to enhanced CO(g) re-adsorption
on catalyst surface sites in close proximity. Furthermore, the electroreduction of CO2 feed
with CO-bleeding showed enhanced ethylene production over a broad potential range with
various co-feed ratios. The origin of the carbon-atoms in C2H4 under co-feed conditions was
traced and quantified by a custom-designed operando differential electrochemical mass
spectrometry (DEMS) flow cell, giving unique mechanistic insight in the CO2-CO co-feed
system. The co-feed mechanism was extended to novel tandem catalysts, in which a CO
producer (Ag, or NiNC) works as local CO feeding source and combines with Cu-based
catalysts, showing obvious enhancement in ethylene production compared to purely
copper-based systems.
Furthermore, a sheet-shaped CuOx catalyst was designed, developed and systematically
investigated. In H-Cell measurements, a high activity for CO2RR could be observed,
followed by tests in a micro flow cell, demonstrating a record C2H4 partial current density of
229 mA cm-2 and a C2+ partial current density of ~ 410 mA cm-2. Moreover, a combination of
IV
operando/(quasi) in situ XAS, WAXS, DEMS, S/TEM techniques were employed to discover
structure-activity-selectivity relations under catalytic CO2RR operating conditions, delivering
perspectives to design novel catalysts to produce hydrocarbons as the value-added products.
V
Zusammenfassung
CO2-Emissionen sind in den letzten Dekaden immer rasanter gestiegen, da der erhöhte
Energiebedarf der wachsenden Weltwirtschaft vor allem durch fossile Energieträger gedeckt
wurde und wird. Dies führt zu immer stärkerer Umweltverschmutzung und hat Auswirkungen
auf das weltweite Klima. Die elektrochemische CO2-Reduktionsreaktion (CO2RR) bietet
hierin eine vielversprechende Chance um CO2 aus der Atmosphäre oder als direkte
Abgasverwertung zu entfernen und in höherwertige Produkte zu wandeln. Wird dazu
überschüssige Elektrizität aus erneuerbaren Energien wie z.B. Solarenergie oder Windkraft
genutzt, hat die CO2RR das Potenzial zu einer CO2-neutralen Energiewirtschaft beizutragen.
Unter den bekannten Katalysatoren wurden kupferbasierte Materialien am häufigsten
untersucht, da sie die einzigartige Eigenschaft besitzen zu großem Anteil Kohlenwasserstoffe
zu produzieren ohne das Überpotential zu stark zu erhöhen.
Diese Arbeit stellt zum ersten Mal ein tieferes Verständnis der CO2RR her, indem die lokale
Umgebung eines CuOx-Nanopartikel (NP) Modellkatalysators systematisch variiert wurde.
Durch Kontrolle der Partikelbeladungsdichte und der Zusammensetzung des Eduktstroms aus
CO2 und CO kann die Produktverteilung gezielt eingestellt werden. Mit steigender
Beladungsdichte, d.h. sinkendem Interpartikelabstand, wurde eine Verschiebung der
Produkteffizienzen von CH4 zu C2H4 beobachtet, die hier auf verstärkte CO(g)-Readsorption
auf nahe zusammenliegenden Katalysatorzentren zurückgeführt werden konnte. Des Weiteren
ermöglicht der co-feed aus CO2/CO verstärkte Ethylenproduktion in einem breiten
Potentialbereich indem das CO2/CO-Verhältnis angepasst wird. Zur Aufdeckung des
Reaktionsmechanismus unter CO2/CO co-feed wurde die Herkunft der C-Atome im
produzierten Ethylen über Isotopenmarkierung in einer eigens dafür designten Flusszelle für
operando differentielle elektrochemische Massenspektrometrie (DEMS) untersucht und
quantifiziert. Der vorgestellte co-feed Mechanismus konnte auf neuartige
Tandemkatalysatoren übertragen werden. In diesem wird ein CO-produzierender Katalysator
(z.B. Ag oder NiNC) mit einem kupferbasierten Katalysator kombiniert, was ebenfalls zu
erhöhtere Ethylenproduktion gegenüber rein kupferbasierten Systemen führt.
VI
Außerdem, wurden in dieser Arbeit ein neuartiger, plattenförmiger CuOx Katalysator
entworfen, synthetisiert und systematisch untersucht. Aufgrund herausragender
CO2RR-Aktivität in H-Zellmessungen wurde der Katalysator in einer anwendungsnahen
Mikroflusszelle getestet. Dort konnten für C2H4 und C2+ industrierelevante
Rekordproduktstromdichten von 229 mA cm-2 bzw. ~410 mA cm-2 gezeigt werden. Weiterhin
wurden durch Kombination von operando/(quasi) in situ XAS, WAXS, DEMS und S/TEM
Techniken Struktur-Aktivitäts-Selektivitätsbeziehungen für die katalytische CO2RR entdeckt,
die neue Wege zum Design von hochaktiven und –selektiven Katalysatoren aufzeigen, um
gezielt höherwertige Kohlenwasserstoffe aus CO2 zu produzieren.
VII
Table of Content
Acknowledgement ...................................................................................................................... I
Abstract……………………………………………………………………………………….III
Zusammenfassung ..................................................................................................................... V
Chapter 1 Introduction and Motivation .................................................................................. 1
1.1 Electrochemical CO2 Reduction Reaction ................................................................... 2
1.2 Electrocatalysts for CO2RR ........................................................................................ 4
1.3 Mechanistic pathways, research progress on Cu-based catalysts for CO2RR ............ 5
Chapter 2 Goals and Objectives ............................................................................................. 9
Chapter 3 Aspects and Experimental Procedures ................................................................ 13
3.1 Synthesis of shaped copper/copper oxides nanoparticles ......................................... 13
3.1.1 Synthesis of spherical CuOx nanoparticles (CuOx NPs) ................................. 14
3.1.2 Synthesis of sheet-like CuOx nanoparticles (CuOx NS) .................................. 14
3.1.3 Synthesis of Nickel-Nitrogen-Functionalized Carbon Material (NiNC) ........ 15
3.2 Physicochemical Characterization ............................................................................. 15
3.2.1 (Scanning/Transmission) Electron Microscopy ............................................. 16
3.2.2 X-ray Diffraction (XRD) .............................................................................. 18
3.2.3 Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES) ..... 20
3.3 Electrochemical Characterization .............................................................................. 21
3.3.1 Electrochemical Methods ............................................................................... 21
3.3.2 H-Cell configuration ....................................................................................... 22
3.3.3 Micro-Flow-Cell configuration ...................................................................... 23
3.4 Products quantification ............................................................................................. 24
3.4.1 Online gasous products quantification ........................................................... 24
3.4.2 Liquid products analysis ................................................................................. 25
3.5 (quasi) in situ/operando Characterization ................................................................. 26
3.5.1 (in situ) Synchrotron Wide-Angle X-Ray Scattering (WAXS) ...................... 26
3.5.2 in situ (Scanning/Transmission) Electron Microscopy ................................... 30
3.5.3 operando Differential electrochemical mass spectrometry (DEMS) ............. 31
VIII
3.3.4 operando X-ray absorption spectroscopy (XAS) ........................................... 39
Chapter 4 Catalyst particle density controls hydrocarbon product selectivity in CO2
electroreduction on CuOx ........................................................................................................ 43
4.1 Synthesis and characterization ................................................................................. 45
4.2 Electrochemical CO2RR over CuOx catalysts with various areal densities .............. 46
4.3 Discussions ................................................................................................................ 50
4.4 Conclusions ............................................................................................................... 54
Chapter 5 Enhanced Ethylene Yields in the electroreduction of CO2/CO co-feeds on Cu and
Cu-Tandem Electrocatalysts: Time-resolved quantitative deconvolution of mechanistic
reaction pathways .................................................................................................................... 57
5.1 Synthesis and characterization ................................................................................. 59
5.2 Feeding gases control ................................................................................................ 60
5.2.1 In H-cell design .............................................................................................. 60
5.2.2 In DEMS flow cell .......................................................................................... 61
5.3 Electrocatalytic rates and product yields under pure and mixed CO2/CO co-feeds .. 61
5.4 Discussions ................................................................................................................ 64
5.5 Extended co-feed mechanism on bifunctional tandem catalyst design ..................... 72
5.6 Conclusions ............................................................................................................... 74
Chapter 6 Sheet-like Copper Oxides with Stable and Selective Ethylene Production for
Direct CO2 Electroreduction .................................................................................................... 77
6.1 Synthesis and characterization ................................................................................. 79
6.2 Electrochemical CO2RR over sheet-like CuOx catalysts .......................................... 81
6.3 Discussions ................................................................................................................ 83
6.4 Conclusions ............................................................................................................... 88
Chapter 7 Achieving high C2H4 Evolution at industrial current densities on CuOx nanosheet
derived gas diffusion electrode ................................................................................................ 89
7.1 CO2RR electrolysis using GDE combined MFC ...................................................... 91
7.2 Discussions ................................................................................................................ 93
7.3 Conclusions ............................................................................................................... 96
IX
Chapter 8 Summary and Outlook ......................................................................................... 97
8.1 Manipulating *CO behavior by local condition ........................................................ 97
8.2 Shaped CuOx nanoparticles for CO2RR .................................................................... 99
8.3 Outlook .................................................................................................................... 101
Reference ...... ………………………………………………………………………………..103
Appendix ...... ………………………………………………………………………………..113
A1. Supplementary Information to Chapter 4 ............................................................... 113
A2. Supplementary Information to Chapter 5 ............................................................... 117
A3. Supplementary Information to Chapter 6 ............................................................... 127
A4. Supplementary Information to Chapter 7 ............................................................... 137
Table of figures and schemes ........................................................................................ 140
Table of tables ............................................................................................................... 146
List of Abbreviations ..................................................................................................... 147
List of Chemicals ........................................................................................................... 149
List of Publications during Ph.D. Study ........................................................................ 151
X
1
Chapter 1 Introduction and Motivation
Chapter 1 Introduction and Motivation
Population increase, urbanization, living standards upturn, as well as the world-wide industry
development, are always accompanying with the increase of energy demand. Fossil fuels,
which include coal, oil and natural gas, have been playing a significant role in energy
production ever since the industrial revolution.1-6 As a significant instance, in the U.S., the
largest energy consuming country, 62% of the electricity generation directly relies on fossil
fuels. It has been expected, the fossil fuels, serving as the main power source in the last
century, and, will continue playing the same role at least for the upcoming decades. We
should also keep in mind, combustion of these fossil fuels yields tremendous CO2 emissions
into the nature, which is one of the greenhouse gases that allows radiative forcing and
contributes to the global warming.7
Figure 1.1 Total global renewable energy capacity from Hydropower and marine (orange),
Wind (green), Solar (violet), Bioenergy (yellow), Geothermal (blue) sources. Data were taken
from ref 8. Copyright International Renewable Energy Agency.
Thus, low-carbon and high efficiency energy generation are highly desired than ever. The U.S.
Energy Information Administration (EIA) estimated that electricity generation from
2
Chapter 1 Introduction and Motivation
renewables (i.e. hydropower, solar and wind) would increases to 31% of total generation by
2050, from an 18% share in 2018.7 A statistical report from the International Renewable
Energy Agency (IRENA) reveals the constitution of global renewable energy capacity various
a lot during the past seven years.8 As shown in Figure 1.1, the share of hydropower and
marine capacity decreases from 70% in 2012 to 52% in 2018.8 Assumptions of declining costs
and improving performance make wind and solar increasingly competitive compared with
other renewable resources, in particular, the U.S. adds 72 gigawatts of new wind and solar
photovoltaic capacity between 2018 and 2021.7 However, these resources (solar radiation
intensity as well as the wind speed) are normally intermittent and heavily depend on the
weather. Thus, how to integrate these intermittent renewable capacities and store them
cost-effectively remains a challenge.
Among the different alternatives, electrochemical reduction of CO2 using the electricity from
renewable sources as the driving force is one promising solution.9-16 This process could
effectively reverse the carbon combustion process, using the CO2 and water as the feed,
delivering carbon-containing fuels and chemicals as the product, schematically realizing a
CO2 neutral process. It is necessary to mention, unlike the Fischer-Tropsh process, operation
condition of the electrochemical CO2 reduction is mild, i.e. room temperature and pressure.
The fast response time of electrochemical systems could excellently couple to intermittent
renewable energy sources,17 providing the driving force for such electrochemical
transformation. Towards the potential for industrial level utilization, the modular design of
electrolyzer enables people to scale up the reaction to kton/day easily,18, 19 with properly
separating and storing different products yielded from the one-shot reaction. The CO2RR
(electrochemical CO2 reduction reaction) sector could be populated based on the current
manufacture capability.
1.1 Electrochemical CO2 Reduction Reaction
The studies on CO2 conversion significantly explode due to the rising of public awareness for
environmental protection and the decreasing price of renewable electricity.20-36 A general
scheme for electrochemical process of CO2RR is given in Figure 1.2.
3
Chapter 1 Introduction and Motivation
Figure 1.2 General scheme of a typical electrolyzer for CO2RR.
In a typical CO2 electrolyzer, the cathode is separated from the anode by a membrane
separator. Under external potential, oxygen evolution reaction (OER, eq. 1-1) occurs on the
anode and oxygen is released. Proton, which is another product of water oxidation, could pass
the proton-transfer membrane to the cathodic side. Meanwhile, CO2 gas flows into the
cathode chamber and get reduced. As depicted in equations 1-2 and 1-3, the first possible
reduction products, which are thermodynamically more favorable, are HCOOH and CO.
Reduction of CO2 at high overpotentials could also yield alcohols and hydrocarbons (eq. 1-4
to eq. 1-7). A competitive reduction reaction in the cathodic side is the hydrogen evolution
reaction (HER, eq. 1-8). Choosing the proper catalyst and working condition could suppress
the HER to a large extent.
4
Chapter 1 Introduction and Motivation
The electrode potentials above are given as SHE at pH 7.
1.2 Electrocatalysts for CO2RR
Two major groups of materials have been screened for catalytically selective CO2RR, namely
carbon-based catalysts and metal-based catalysts. Carbon materials have many essential
advantages such as high surface area and excellent conductivity. However, the neutral carbon
atoms show negligible CO2-activation ability. Many methods have been developed to improve
the catalytic activity and selectivity toward carbon-based materials, for instant, doping of
heteroatoms (i.e., nitrogen, boron, and sulfur) into carbon matrix,37-42 or co-doping non-noble
metal atoms and nitrogen atoms (known as M-N-C).14, 43, 44 The M-N-C materials, which is
regarded as CO maker, has been well studied.45-48 For example, Möller et al. found that CO2
could be selectively converted to CO on nickel and nitrogen-doped porous carbon catalyst
(Ni-N-C) at industrial relevant current densities in flow cell setup.49
Figure 1.3 Classification of various metals depending on the formation of major products in
electrochemical CO2 reduction. a) Periodic table and faradaic efficiency of major product
from experimental data by Hori. H2 (red), CO (blue), formate (yellow), hydrocarbon (green) b)
The experimental product classification of H2, CO, a d HCOOH by he ΔEH*-ΔECOOH* and the
ΔEH* c) Faradaic efficiency of CO2 educ i eac i by ΔEH*. a-c) Modified and reproduced
from ref 51 with permission. Copyright 2018, John Wiley and Sons.
5
Chapter 1 Introduction and Motivation
Pioneered by the Hori, the elemental metallic electrocatalysts for CO2RR are usually
classified with four groups depending on the intermediate affinity and targeting products (see
Figure 1.3).20, 50, 51 More specifically, for metals (i.e., Pt, Ni, Fe, and Ti) that have high *H
affinity and favor HER; for the metals (i.e., Sn, Cd, Hg, In, Tl and Bi) effectively convert CO2
into formate; for metals (i.e. Au, Ag, Zn, Pd and Ga) that bound CO weakly are known as
CO-makers. While copper (Cu) is unique that binds CO neither too strong nor too week,
leading to a considerable amount of valuable ethylene at reasonable overpotentials.52-70
1.3 Mechanistic pathways, research progress on Cu-based catalysts for CO2RR
1.3.1 Mechanistic pathways
The unique property of Cu-based catalysts is attributed to the appropriate adsorption energy
of CO species (*CO),71-73 which is commonly accepted as a key reactive intermediate on Cu
surfaces. Formation of reactive *CO is followed by a protonation pathway to C1 products,
such as CH4, or by a dimerization pathway to C2 products, such as C2H4.74
CO evolution by desorption is known to occur prior to the formation of hydrocarbons and is
strong function of the detailed chemisorption properties of CO. Kortlever et al. reported that
the dimerization step is the rate-determining step in the C2 pathway.74 Huang et al.
demonstrated the interfacial *CO coverage is a key factor in determining the extent of the
C2H4 formation.75 In their simulation, higher *CO covered surface could decline the energy
barrier for *CO dimer, further enhancing the C2H4 formation rate.
Figure 1.4 Possible C1 species for C-C coupling.
6
Chapter 1 Introduction and Motivation
For the C2H4 formation, C-C coupling by different C1 species is proposed (see Figure 1.4). It
is noted that some mechanistic studies are based on CO reduction, since CO2RR is essentially
a self CO co-feeding system. Kortlever et al. reported a mechanism consisting of the coupling
of two *CO molecules based on DFT results.74 This pathway follows: *C2O2 → *C2O2- →
*CO-COH → *CCO → *CHCO → *CHCHO → *CH2CHO → C2H4. Cheng et al.
demonstrated that the hydrogenation of *CO-COH to *COH-COH is more favored than its
dehydration to form *CCO.76 Besides *CO-*CO dimerization, Goodpaster et al. found that
the dimerization of *CO and *CHO, forming *COCHO, is more favorable at high
overpotential.77, 78 Cheng et al. suggested that the coupling of *CO and *COH favors the
formation of *CO-COH instead of *COCHO.79, 80 Montoya et al. demonstrated that the
dimerization of hydrogenated species (*CH2O and *CHO) is kinetically more favorable than
*CO dimerization in an electrochemical environment.81
Many parameters can be used for tuning the product selectivity on Cu electrocatalysts.72-86 For
example, it is found that the products distribution during CO2RR is sensitive to the exposed
Cu facets.87-91 With Cu (111) facet, CH4 formation is enhanced, while (100) facet favors C2H4
formation.82, 84-86
1.3.2 Research progress on Cu-based catalysts for CO2RR
Many investigations have shown that oxide-derived copper catalyst (OD-Cu), which are
formed by in situ reduction of CuOx during catalytic CO2RR, are promising due to their
enhanced C2 products yield.92-94 It is suggested that several parameters, such as the effects of
grain boundaries, local pH, under-coordinated sites, CO coverage, and subsurface oxygen,
have been attributed to the enhanced activity and selectivity towards “beyond CO”
productd.93-105 (see Figure 1.5)
7
Chapter 1 Introduction and Motivation
Figure 1.5 a) Tuning the product distribution of CO2 electroreduction on OD-Cu foam
catalysts. b) Optimizing C-C coupling on oxide-derived copper. c) Effects of CO* coverage
on the selective formation of ethylene. d) Nature and distribution of stable subsurface oxygen
in OD-Cu. a) to d) are adapted from ref 97, 102, 75 and 112, respectively, with the permission,
copyright American Chemical Society.
Surface roughness and grain boundary
The roughened surfaces of OD-Coppers are composed of a large number of
under-coordination sites (edges, steps, and defects) which are more active for CO2RR. The
sites at the grain boundary are suggested to change the binding energies of different
adsorbates and decrease the energy barriers for the formation of key intermediates, thus
leading to particular products (Figure 1.5a).87, 94, 96, 106-108
Local pH
As CO2RR involves multi- protons and electrons transfer processes, it is suggested that the
local pH near the electrode surface need to be considered.60 Xiao et al. found that the donation
of proton from water nearby for the dehydroxylation to ethylene is more favorable at neutral
pH.109 Lum et al. demonstrated a moderate pH near OD-Cu electrode enhanced C2+ products,
suppressing the formation of C1 product (Figure 1.5b).102
CO coverage
It is also demonstrated that the activation barrier for the dimerization is dependent on the
coverage of *CO on the OD-copper surface (Figure 1.5c).75, 86 High coverage of adsorbed
*CO decreases the energy barrier of the formation of C-C bond and favors the C-C coupling
of C1 intermediates to ethylene.
8
Chapter 1 Introduction and Motivation
Subsurface oxygen
Combining experimental and theoretical results, several groups suggested that the presence of
subsurface oxygen (Figure 1.5d) favors both CO2 activation and C-C dimerization, leading to
an increased ethylene formation.69, 110, 111 Cavalca et al. proposed the subsurface oxygen
withdraws electrons from the sp- and d-bands of copper, thus the selectively decreases the
energy barrier of C-C coupling.112 However, the role of subsurface oxygen in CO2RR is in
debate and needs more careful discussion. Garza et al. found that the oxygen in the subsurface
of Cu (100) is unstable.101, 113 Agreeing with the role of subsurface oxygen, in increasing the
coverage of *CO and decreasing *H adsorption, Liu et al. suggested that the presence of
subsurface oxygen doesn’t necessarily change the free-energy activation barrier.114
In summary, although the above parameters are studied for obtaining a selective
electrocatalysts, detailed mechanistic studies are still needed for a better catalyst design.
Recent advanced in situ/operando techniques provide an excellent opportunity to gain
catalytic insight on both CO2RR mechanism and catalyst matrix. Besides, it should be noted
that these factors are correlated with each other rather than absolutely independent. For
example, changing the surface roughness also companying with the change of local pH.102
Comprehensive consideration is also required in the complex system.
9
Chapter 2 Goals and Objectives
Chapter 2 Goals and Objectives
This work is focused on the electrocatalytic reduction of carbon dioxide (CO2RR) towards
value-add products over copper-based nanoparticles. First goal (1) of this work is to improve
the catalytic activity and selectivity of the C2 products, the second (2) is to expand the state-of
mechanistic understanding, and the final (3) goal is to design novel catalysts with high C2
products yielding.
Using spherical CuOx nanoparticles as a model material, the influence on the nanoparticle
assembles was firstly investigated. The catalytic activities, as well as the C2 products
distribution, are studied and attributed to the re-adsorption behavior of key *CO intermediate.
To further uncover the contribution of the interfacial CO, CO2 with CO co-feeding system is
then investigated using advanced electrochemical methods and operando differential
electrochemical mass spectrometry (DEMS). Based on the co-feed mechanism, Cu-based
tandem catalysts are designed and measured for CO2RR. Moreover, the shape selected
CuOx-derived nanoparticles are also promising candidates. A detailed understanding of the
evolution behavior in both phase and morphology aspects is also the goal. Advanced in situ
(operando) wide-angle X-Ray scattering (WAXS), X-ray absorption spectroscopy and
(Scanning/Transmission) Electron Microscopy are used as most directive and best approaches
to investigate a chosen CuOx nanoparticle system under real operating condition.
Figure 2.1 Schematic overview of scientific questions and goals dealt with in the course of
this work.
In Chapter 3, the experimental procedures used in this work are described in detail, including
material synthesis and physicochemical characterizations. The quantification of CO2RR
10
Chapter 2 Goals and Objectives
products, which are detected by home-build on-line GC, liquid-GC, HPLC for liquid products
analysis, is explained as well. Furthermore, the description of advanced in situ (operando)
methods is shown at the end of this chapter.
In Chapter 4, the relationship between reactive *CO dimerization behavior and areal particle
density are investigated on a family of monodispersed copper oxide nanoparticles (CuOx NPs)
during the CO2RR. Rietveld-refined catalytically active CuOx NPs are correlated to
hydrocarbons production and efficiency under systematic variation of the particle density and
hence mean distance inside the 3D catalyst layers. We demonstrate that improved C2H4 yields
are achieved using high areal particle density (reduced mean-particle distance with the
increased real surface area) and attribute this to dynamically favored CO(g) re-adsorption at
elevated local interfacial pH.
In Chapter 5, we systematically investigate the effective hydrocarbon production rates during
the CO2RR on Cu-based nanocatalysts under variations of the CO2/CO feed ratio. The
emphasis is placed on the kinetics of the catalytic ethylene formation. The dissolved COx
species are carefully controlled by the feed ’ partial pressure in an H-type cell. To uncover
the mechanistic origin of the enhanced ethylene formation rates, we conducted mixed,
isotope-labeled (12C/13C) feed experiments in a newly developed CO2RR-specific capillary
electrochemical cell attached to our differential electrochemical mass spectrometer (DEMS).
Using our DEMS technique, we not only trace the chemical-mechanistic origin (from either
dissolved CO or CO2) of the two carbon atoms in the C2H4 product molecules, but also
succeed in quantifying the relative contribution of the competing mechanistic reaction
pathways toward C2H4. Even though the focus of this chapter is on the electrochemical
kinetics and dominant mechanistic pathways and not on the chemical state dynamics of the
Cu-based catalyst, high-energy X-ray diffraction (HE-XRD) and Rietveld refinement were
used to clarify the initial population of crystalline Cu phases and their time evolution.
Buildling on more mechanistic insights on local CO co-feeding, we close our report in
demonstrating a new family of hybrid carbonous/metal, metal/metal tandem CO2RR catalyst
11
Chapter 2 Goals and Objectives
concepts that succeed in realizing full local CO self-co-feeding benefits now even under pure
CO2 feeds.
In Chapter 6, the investigation is extended on the new catalyst design. With successfully
synthesized sheet-like copper oxide nanoparticles, regular H-cell measurement is firstly
applied for CO2RR. A combination of (in situ/operando) grazing incidence wide-angle
X-Ray scattering (WAXS), X-ray absorption spectroscopy are used to examine the
phase/local atomic structure evolution of highly active CuOx nanoparticles under operating
CO2RR electrolysis condition. The millisecond-resolved operando differential
electrochemical mass spectrometry (DEMS) was employed to determine the onset potential
shift of the products with gradually self-electroreduced CuOx electrocatalyst to purely metallic
Cu phase. SEM measurement is also performed to check the morphology evolution after CO2
electroreduction process.
In Chapter 7, further gears the as-prepared CuOx nanosheet into a Micro-Flow-Cell (MFC)
configuration equipped with a Gas Diffusion Electrode (GDE) and manipulated the current
density from lab-scale to industry relevant level. Remarkably, the catalyst poses a faradaic
efficiency exceeding 33% ethylene at high current (700 mA cm-2) with suppressed HER, since
the gas diffusion layer allows overcoming the CO2 reactant transport limitation compared to
H-cell design. More importantly, the morphological transformations of free-standing CuOx
nanosheets are also investigated in an electrochemical liquid TEM cell set up, comparing with
different devices. The present study corroborates the power of in situ electrochemical liquid
TEM studies for the understanding of activity-selectivity-morphology relations under
catalytic CO2RR operating conditions at the industry level.
In Chapter 8, summarizes the results of this thesis, and general conclusions are drawn.
Additionally, perspectives for further investigations on the catalysts design and device
development are given.
12
Chapter 2 Goals and Objectives
13
Chapter 3 Aspects and Experimental Procedures
Chapter 3 Aspects and Experimental Procedures
Experimental Procedures reproduced in part with permission from ChemSusChem, 2017, 10,
4642-4649 (Copyright of The John Wiley and Sons (2017))126 and from Nat. Nanotechnol., 2019, 14,
1063-1070 (Copyright of Springer Nature)127. Experimental Procedures reproduced in part from
manuscripts under preparation for submission.
A List of Chemicals listing Acronyms, Purity/Concentration and Supplier is attached to the
Appendix Section on page 147.
This chapter presents all synthesis routes of a series of shaped Cu-based nanoparticles that are
investigated in this work. Following, regarding methodologies for physicochemical
characterization as well as the procedures for electrocatalytic performance (activity,
selectivity, durability, and scalability) testing will be described. In the final session of this
chapter, the experimental details on the performed in situ analytical and X-ray/microscopy
methods will be given including a brief explanation on the theoretical fundaments of the most
important methods used in this work.
3.1 Synthesis of shaped copper/copper oxides nanoparticles
The development synthesis methods for copper/copper oxides are necessarily needed since
copper is not easily reduced due to its low standard reduction potential under mild condition
and high oxygen affinity. In order to prepare narrow size-distributed, homogeneously shaped
nanoparticles, different synthesis methods, including wet chemistry method and thermal
decomposition, are carefully adjusted and applied in this work. Table 3.1 gives an overview
of the Cu-based nanoparticles and in the corresponding chapters they are studied. The first set
of catalysts are spherical CuOx nanoparticles (CuOx NPs), which are used for catalysts areal
density study, co-feed mechanistic insight study and as well as hybrid tandem catalysts. The
second one is sheet-like CuOx nanoparticles (CuOx NS), which are employed for various
advanced operando studies and flow cell study.
14
Chapter 3 Aspects and Experimental Procedures
Table 3.1 Overview of different materials used in this work with the corresponding chapters
in which their characterization is discussed.
Material
Oxidation States
Chapter
CuOx nanosperes (CuOx NPs)
Cu (0,Ⅰ,Ⅱ)
4, 5
CuOx nanosheets (CuOx NS)
Cu (Ⅰ, Ⅱ)
6, 7
3.1.1 Synthesis of spherical CuOx nanoparticles (CuOx NPs)
All chemicals are used as commercial ones without further purification. The synthesis was
operated at all times under a rigorously protective atmosphere of N2. In a typical route, 430.35
mg CuBr, 9.7 g trioctylphosphine oxide (TOPO, 90%, Sigma Aldrich), and 10 mL oleylamine
(90%) were added in a three-necked flask. The mixture was then heated to 80 °C with
magnetic stirring. After being kept at this temperature for additional 15 min, 100 mg Borane
tert-butylamine complex (TBAB, 97%, Sigma Aldrich) was added to the resulting
homogeneous solution. The mixed solution was heated up to 200 °C and kept at this
temperature for 60 min with stirring. As the reaction progressed, the colour of the mixture
changed from light blue to red. The resulting red colloidal products were collected by
centrifugation, and washed several times with hexane and ethanol. Finally, the products were
re-dispersed in hexane and the content of copper was detected by ICP-OES. Or freeze dried
and stored as powders under inert atmosphere until use.
3.1.2 Synthesis of sheet-like CuOx nanoparticles (CuOx NS)
No purification was performed for chemicals before use. CuOx nanosheets (CuOx NS) were
obtained by thermal decomposition of the pre-synthesized Cu(OH)2 intermediate. To prepare
the Cu(OH)2 intermediate, 100 mg Cu(Ac)2 (Sigma Aldrich) was dissolved in 4 mL DMF (N,
N-dimethylformamide, Aldrich)/ Milli-Q wa e (> 18 mΩ cm), 2.5 mL 1M KOH solution was
then added into the solution dropwise into the solution. After 15-min stirring, 500 µL
ammonium hydroxide and 500 µL water were added. The resulting homogeneous light blue
solution was then transferred to a glass pressure vessel with a capacity of 47 mL. The sealed
15
Chapter 3 Aspects and Experimental Procedures
vessel was then heated from room temperature to 60 °C in 30 min and stayed at 150 °C for 12
h before cooling to room temperature. The products were precipitated by ethanol, separated
via centrifugation at 7350 rpm and further purified twice by ethanol and water. After
freeze-drying, the obtained powders were stored as powders under an inert atmosphere until
use.
3.1.3 Synthesis of Nickel-Nitrogen-Functionalized Carbon Material (NiNC)
Ni-N-C catalyst synthesis follows analogous procedure presented in ref 22. Ketjen EC 600JD
(AzkoNobel) was initially dispersed in 0.5 M HCl for one day stirring and vacuum filtered
with DI water till neutralized. Afterward, this neutralized carbon powder was leached in
concentrated HNO3 for 8 hours at 90 °C to modify the surface and thereafter vacuumed
filtered with DI water to neutral again. The obtained carbon powder will be referred to as
carbon support. For the following-up synthesis, the carbon support was sonicated in 50 mL
DI-water for 15 minutes.
For NiNC catalyst synthesis, 3 mL of Aniline, 5 g NiCl2 salt and 5 g Ammonium Persulfate
(APS, (NH4)2S2O8) were added into 500 mL of 1 M HCl, and kept stirred for one hour. 400
mg sonicated carbon support (mentioned above) was added into this suspension and stirred
for 48 hours at ambient condition. After that, the suspension was dried in the air at 95 °C for
24 hours. The obtained solid mixture was ball-milled and heat treated (HT) in a furnace under
N2 condition for carbonization. The heating temperature was controlled with a ramping of
30 °C min-1 to 900 °C and kept at this temperature for 60 min. The cooled down material
powder was washed in 2 M sulfate acid (AW) for overnight to remove the exposed inorganic
species and rinsed to neutral pH by using vacuum filtration. The final NiNC catalyst was
obtained with the protocol HT-AW-HT-AW-HT-AW-HT.
3.2 Physicochemical Characterization
Due to the variety of methods used in this work, an overview and experimental description of
all physicochemical characterization techniques used in this thesis will be summarized in the
following sessions, starting with essential methods such as SEM, TEM/STEM, XRD and
16
Chapter 3 Aspects and Experimental Procedures
ICP-OES followed by more advanced operando analytical techniques such as DEMS, WAXS,
XAS etc.. Table 3.2 gives references to the sub-sections with the experimental descriptions
and the corresponding chapters in which the results are presented and discussed.
Table 3.2 Overview of Methods for physicochemical characterization used in this thesis with
reference to the sections in which they are described. x indicates the application of the method
for the corresponding result chapter.
Method
Ch. 4 –
Ch. 5 –
Ch. 6 –
Ch. 7 –
SEM
x
x
x
x
TEM
x
x
x
x
STEM
x
x
XRD
x (laboratory)
x (synchrotron)
x (synchrotron)
ICP-OES
x
Elemental mapping
x
In situ
DEMS
x
x
WAXS
x
x
x
TEM
x
x (quasi)
XAS
x
3.2.1 (Scanning/Transmission) Electron Microscopy
Theoretical Aspects
The electron microscope is capable of much higher magnifications and has a greater resolving
power than a light microscope, allowing it to see much smaller objects in finer detail. The
minimum separation ( ) that can be resolved by any kind of a microscope is given by the
following formula:
Where is the refractive index (which is 1 in the vacuum of an electron microscope), is
the wavelength, and is the maximum half-angle of the cone of light that can enter the lens.
17
Chapter 3 Aspects and Experimental Procedures
According to the formula, one can improve resolution by using shorter wavelengths and
mediating with larger indices of refraction. Electron microscopy is designed to use extremely
short wavelengths of accelerated electrons to form high-resolution images. Figure 3.1 displays
how the electron beam interacts with the samples. By using different scattering beam and/or
transmitted beam, we could gain information of the sample from different aspects.
Figure 3.1 lists the interaction between electrons and samples used for different techniques.
3.2.1.1 Scanning electron microscopy (SEM)
SEM is a nondestructive method to obtain the information about morphology, particle size,
and surface topography. The SEM was measured using a JEOL 7401F instrument at an
acceleration voltage of 10 kV and a working distance of approximately 4.1 mm. Image J was
employed to analyze the SEM images.
3.2.1.2 (Scanning) Transmission electron microscopy (S/TEM)
The TEM is used to study nanoparticle morphology. TEM images were obtained using an FEI
Tecnai G2 20 S-TWIN equipped with a LaB6 cathode operating with 200 kV acceleration
voltage and a resolution limit of 0.24 nm. A Cu grid (200 mesh) coated with a holey carbon
18
Chapter 3 Aspects and Experimental Procedures
film was impregnated with the sample solution and air-dried at 60 °C. An FEI TECNAI G² 20
S-TWIN microscope, equipped with a GATAN MS794 P CCD-detector operated at 200 kV
was used. The mean particle size was determined from the TEM images by counting at least
200 particles.
The STEM-EDS elemental maps were acquired using an FEI Titan 80-200 (“ChemiSTEM”)
electron microscope, equipped with a Cs-probe corrector (CEOS GmbH) and a high-angle
annular dark field (HAADF) detector. The microscope was operated at 200 kV. In order to
achie e “Z-C a ” c di i , a be emi-angle of 25 mrad and an inner collection
semi-angle of the detector of 88 mrad were used. Compositional maps were obtained with
energy-dispersive X-ray spectroscopy (EDX) using four large-solid-angle symmetrical Si drift
detectors. The quantitative analyses were performed on Cu K and O K.
3.2.2. X-ray Diffraction (XRD)
Theoretical Aspects
Similar to other electromagnetic radiation, X-ray can be treated as either a wave or a particle.
The wave-particle duality of light is defined as
where i he e e gy f he h wi h h bei g P a ck' c a , a d ν he f eque cy a d
λ he wa e e g h f he wa e. F in eV and in nm,
⁄.
Figure 3.2 Schematic illustration of the Bragg equation.
19
Chapter 3 Aspects and Experimental Procedures
Diffraction is the effect following the interaction of an electromagnetic wave with an object
having a size dimension comparable with the period of the wave. The Bragg diffraction
occurs when the reflected beams are in phase and interfere constructively. Schematic
illustration of the Bragg equation is shown in Figure 3.2, and expressed mathematically in
B agg’ aw a f wi g
where θ i ca e i g a g e, λ i he wa e e g h f he i cide X-ray wave, and d is the
interplanar distance of the lattice planes, while n is a positive integer ( .
Experimental Description
The X-ray diffraction (XRD) technique is a powerful method to obtain the information about
the bulk structure of crystalline materials and can provide information on unit cell dimensions.
The diffraction patterns are based on constructive interference of monochromatic X-rays and
the sample.
The profiles were acqui ed a B uke D8 Ad a ce diff ac me e wi h Cu Kα adia i a d
a Lynx Eye detector (point detector). Samples were analyzed between 30-95° 2θ, wi h a e
size of 0.04° and collection time of 7s. Data analysis was carried out using MDI Jade 9 and
TOPAS 4.2.
Rietveld refinement
Rietveld refinement is a treatment of powder XRD data which is particularly useful in the
case of a multiphase material. Information about the structure of different phases, i.e., unit
cell dimensions, phase quantities, atomic coordinates/ bond length and so on, can be extracted
with this method. It was described by Hugo Rietveld in the 1960s.115 The Rietveld Method
involves: 1) recording the diffracted intensities with different crystal phases;) 2 creating a
crystal structure model, i.e., symmetry, atomic positions, unit-cell size, site occupancies; 3)
varying these and other parameters by least squares refinement to get the best fit between
calculated and observed data. Peak shape and the geometry of the diffractometer set-up are
also taken into account. The Rietveld refinement is generally used for solving an unknown
crystal structure, calculating the amount of disorder or mixing on a Wyckoff site,
20
Chapter 3 Aspects and Experimental Procedures
quantitatively determining the percentages of different phases in a sample and determining
the crystallite sizes.
XRD i e i ie b e ed a diff ac i (g i me e ) a g e (2θ) a e u sed to be constant
with the calculated intensities by using an appropriate model.
∑[ ∑( | | )
]
where,
– Calculated intensity of data point of a powder pattern that comprises of
phases
– Scale factor of the phase
– Multiplicity
– Absorption factor
– Preferred orientation factor
– Structure factor for particular reflection
– Lorentz-Polarization factor
– Profile function
– Background intensity of data point
While the experimentally observed powder XRD profile is an outcome of the combination of
many such scan intensities ( ) that are calculated using these math functions and
correction terms. Successful fitting procedures require high quality experimental data and a
large q-range, as well as reasonable structure model(s) to minimize residual function at
data point over all data points , which can be expressed as:
∑
[ ]
where,
– Measured intensity of data point
– The weight of the diff ac i a e ’ i e i y a d i gi e by equa i 3-4:
[ ]
3.2.3 Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES)
Theoretical Aspects
21
Chapter 3 Aspects and Experimental Procedures
ICP-OES is a method of optical emission spectrometry, which is developed by Fassel et al.
Iowa State University in the US and by Greenfield et al. at Albright & Wilson, Ltd. in the UK
in the 1960s.116 It is a powerful tool to determine trace elements in a myriad of sample types.
Figure 3.3 shows the Bohr model of an atom, which is depicted as a nucleus and surrounded
by electrons in discrete orbitals.
Figure 3.3 Bohr model of an atom. When energy is absorbed by an atom, the electron jumps
from its ground state orbital to an orbital with higher energy level. For the stabilizing, the
atom may decay back to a lower energy state by emitting a photon, .
In ICP-OES measurement, the component elements (atoms) are excited when plasma energy
is given to an analysis sample from outside. When the excited atoms return to low energy
position, emission rays ( ) that is corresponding to the photon wavelength are released and
measured. The element type is determined based on the position of the photon rays, and the
content of each element is determined based on the emission intensity.
Experimental Description
The metal content Cu was analyzed by inductively coupled plasma-optical emission
spectroscopy (ICP-OES) using a Varian 715-ES spectrometer with a CCD detector. In
Chapter 4, 400 µL catalyst/hexane solution was carried out directly after synthesis and dried
in N2. Then powders were digested in acid (HNO3: H2SO4: HCl in a 1:1:1ratio) and kept for
overnight. The samples were diluted with Milli-Q wa e (> 18 mΩ cm) each a
appropriate emission intensity. Standards with a known concentration were co-analyzed with
the samples.
3.3 Electrochemical Characterization
3.3.1 Electrochemical Methods
RHE Potential:
22
Chapter 3 Aspects and Experimental Procedures
ERHE: RHE potential / V
ERef: Applied WE potential against the Ag/AgCl reference electrode / V
EAg/AgCl: Potential of the reference electrode measured against NHE (0.21 V) / V
pH: pH-value of the electrolyte
R: Ohmic resistance between working and reference electrode / Ω
I: Total measured current / A
3.3.2 H-Cell configuration
The Schematic illustration and photograph of H-type cell are shown in Figure 3.4. In the
operating condition, the reduction reaction of CO2 and/or CO happens in the working part and
oxygen evolution reaction occurs in the counter compartment. Three-electrode system was
applied for H-type configuration. Polished glassy carbon or carbon paper are used as working
electrodes and measured with a Biologic SP 300 potentiostat. A platinum mesh 100
(Sigma-Aldrich 99.9%) was used as a counter electrode (CE) and a leak-free Ag/AgCl
electrode as a reference electrode (Multi Channel Systems MCS GmbH). A certain amount of
catalyst ink (a solution of 2 mg catalyst, 50 L of 5 % Nafion, 150 L of isopropanol and
800 L of ultrapure water) was drop-coated on 1 cm2 of working electrode to vary the
catalyst loading for each study.
Figure 3.4 a) Schematic and b) photograph of H-type two compartments cell divided by a
polymer membrane.
23
Chapter 3 Aspects and Experimental Procedures
The working/counter compartment was filled with 40 mL of electrolyte, respectively. Before
and during the electrochemical reaction the working compartment was purged continuously
with various feeds (30 sccm in total) from the bottom of the cell and the gas atmosphere was
controlled with an in situ mass flow controller. Every measurement was started with a linear
voltammetric sweep, performed with a scan rate of -5 mV/s between E = +0.05 VRHE and the
working potential (between -0.78 VRHE and -1.0 VRHE) followed by a chronoamperometric
step for a certain time. All reported potentials are corrected for ohmic drop determined by
electrochemical impedance spectroscopy (EIS). EC-Lab software was used to automatically
correct 50% of the Ohmic drop, the remaining 50% was corrected manually. For each
measurement, a fresh electrolyte was used to ensure that adsorbates from previous
experiments did not influence the result.
3.3.3 Micro-Flow-Cell configuration
The micro-flow-cell (MFC) is used for measuring the samples at high current densities, which
were performed in a commercial cell supplied by ElectroCell. The catalyst ink (a solution of 6
mg catalyst, a certain amount of 5 % Nafion, isopropanol and ultrapure water) was
spray-coated on the microporous layer (MPL) of a Freudenberg C2 gas diffusion layer (GDL).
A geometric area of 3 cm2 was coated to achieve a final metal loading of ~1 mg cm-2.
Additionally 30 wt% of Nafion was used as a binder and for ionic conductivity of the catalyst
layer. 1 M KHCO3 (500 mL, Sigma-Aldrich, BioUltra, ≥99.5%) 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 by using
a peristaltic pump (PMP Ecoline, Cole-Parmer). The CO2 (4.5N) was supplied at a rate of 50
mL min-1 to the cathode and was flown from the back of the GDL through the catalyst layer.
Measurements were performed galvanostatically for 2 h at each respective current during the
catalytic tests, changing the current from low to high values. Each galvanostatic step was
followed by a PEIS measurement to account for the ohmic drop in the calculation of RHE
potentials.
24
Chapter 3 Aspects and Experimental Procedures
3.4 Products quantification
As reported in the literature, more than 16 products are formed on Cu electrocatalyst during
CO2RR, including both gaseous products and liquid products. Table 3.3 is summarized the
products which are detectable in our lab and their corresponding detection method.
Table 3.3 Summary of the detectable products in the H2O/CO2RR electrolysis and the
corresponding detection methods.
Products
Detection Method
Detection limitation
H2
GC
1ppm
CO
GC
1ppm
CH4
GC
1ppm
C2H4
GC
1ppm
HCOO-
HPLC
0.1mM
CH3COO-
HPLC
0.1mM
EtOH
Liquid GC
1ppm
PrOH
Liquid GC
1ppm
Adehyde
Liquid GC
1ppm
3.4.1 Online gasous products quantification
Gas samples were analyzed with a gas chromatograph (Shimadzu GC 2014) equipped with a
thermal conductivity detector (TCD) and a flame ionization detector (FID). Argon (Air liquid
5.0) was employed as the carrier gas. The gaseous compounds H2, N2, O2, CH4 and CO were
separated in a molecular sieve column (Alltech, part no. 57732, 1.65 m × 1/8 in., molecular
sieve 13X, 60/80 mesh) while for C2-C3 hydrocarbons and CO2 in a HayeSep column (Alltech,
part no. 14487, 3.5 m × 1/8 in., HayeSep D, 80/100 mesh).
The production rate towards the gas products was calculated taking into account the
concentration obtained for the gas chromatography analysis and the feed flow (30 sccm)
according to the following equation:
25
Chapter 3 Aspects and Experimental Procedures
: Generation rate of the product / mol s-1 cm-2
: CO2 gas flow rate / L s-1
C: Volume/Molar fraction of detected product by GC (assuming ideal gas) / Vol%
A: Geometric area of the electrode / cm-2
VM: Gas Molar Volume / 22.4 L mol-1
The measured molar production rate was used to calculate the Faradaic efficiency and the
average current density measured during the last minute of reaction:
: Molar production rate of product per unit area / mol s-1 cm-2
FE: Faradaic Efficiency of the product / %
z: number of transferred electrons per mol of product
F: Faraday constant / 96500 C mol-1
: Total current density during CO2 bulk electrolysis / mA cm-2
3.4.2 Liquid products analysis
2 mL of the electrolyte after reaction was analyzed by high performance liquid
chromatograph (HPLC Agilent 1200, Zimmer Chromatography® Column, RID detector) to
measure formic acid concentration and analyzed by liquid injection gas chromatography
(Shimadzu GC 2010 plus, Fused-Silica-Capillary Column, REF 723060.30, FID Detector) for
alcohol products. Therefore, the production rate was calculated considering the total charge
transfer, according to the following equation:
: Generation rate of the product x / nmol s-1 cm-2
V: Volume of the electrolyte / mL
ΔCx: Accumulated concentration of the product x detected by HPLC or liquid GC / mmol L-1
A: Geometric area of the electrode / cm-2
26
Chapter 3 Aspects and Experimental Procedures
Δ : Reac i ime at const. current or potential / s
Faradaic efficiency of liquid products is calculated by the following equation:
FEx: Faradaic Efficiency of the product x / %
V: Volume of the electrolyte / mL
ΔCx: Accumulated concentration of product x detected by HPLC or liquid GC / mmol L-1
zx: electrons transferred for reduction to product x
ΔQ: T a cha ge a fe du i g he e ec y i at const. potential or current / C
The partial current density,
jx: Partial current density / mA*cm-2
FEx: Faradaic Efficiency of the product x / %
jtotal: Total current density / mA*cm-2
3.5 (quasi) in situ/operando Characterization
In this thesis, various in situ/operando methods were applied to analyze both catalysts
properties and clear products evolution under operating electrochemical reaction conditions.
These methods will be described in the following sections, beginning with in situ X-ray based
setup and methods, followed by the electron microscopy.
3.5.1 (in situ) Synchrotron Wide-Angle X-Ray Scattering (WAXS)
In this work, operando synchrotron WAXS was used to determine the phase evolution of
as-prepared CuOx nanoparticles under operating CO2 electroreduction condition. Synchrotron
facilities could provide higher energy due to the fact that an electron accelerator could
produce high-flux radiation with excellent properties. Such measurements allow to reach
a ge q, a d hu 2θ, a wider range of the diffraction pattern, which is necessary for advanced
data analysis (Rietveld refinement and pair distribution function (PDF) analysis) revealing the
fine structural details. Moreover, comparing to lab source X-ray diffractometer with a point
27
Chapter 3 Aspects and Experimental Procedures
detector, the large area detector (Pilatus 3X CdTe 2M) has advantages in accuracy for
detecting the materials with preferred orientation and considerably shorter acquisition times.
Azimuthally integrated line profiles provide us additional information for texture strain.
Results from ex situ WAXS measurement on CuOx NPs are discussed in Chapter 5, and in
situ WAXS measurements on CuOx NS during CO2RR are presented in chapter 6.
Theoretical Aspects
The structural investigation by X-ray scattering mainly depends on elastic scattering, without
any change of energy/wavelength of incoming/measured beam. Figure 3.5 shows the
geometry defining when a fraction of the incident beam (wavelength ) is scattered by the
sample in the experiment.
Figure 3.5 Geometry defining the incident beam and scattered beam with scattering vector .
The measured intensity of the scattered beam corresponds to the square of the Fourier
transform of the scattering length density distribution. The scattering vector is defined as
the difference between the wave-vector of the incident and the scattered beam (Figure 3.5).
| |
In the wide angle range, the observed scattering intensities are related to the interplanar
spacing between the crystallographic planes ( ) in the crystal, which can be expressed by a
modified Bragg equation,
Experimental Description
(in situ) synchrotron WAXS measurements were performed at ID31 at the European
Synchrotron Radiation Facility (ESRF) in Grenoble, France, at relatively high energy of ~70
keV. The working distance between the sample and the detector was calibrated using a CeO2
28
Chapter 3 Aspects and Experimental Procedures
standard (NIST SRM 674b). The diffraction patterns were corrected by the empty beam. The
dry sample was prepared by placing a small amount of catalyst powder (1 mg) between two
layers of Kapton tape for ex situ WAXS measurement. The samples for in situ measurements
were prepared by drop-casting CuOx electrocatalysts ink ( a solution of 50 µL 5 % Nafion and
800 µL ultrapure water (18.2 MΩcm, Pu eLab P u Sy em, E ga) and 150 µL isopropanol)
onto the glassy carbon cylinder (diameter, 2.5 mm) which was polished with 1/0.05 µm Al2O3
suspension and cleaned with ultrapure water before. The catalyst loading was kept as
20 ug cm-2 to avoid overloaded current density. Rietveld refinement was performed using the
TOPAS® software package (Bruker).
Figure 3.6a shows a schematic illustration of the setup as employed for the in situ
experiments operated at a synchrotron in this work. The side view and top view of the test
station was presented in Figure 3.6b and Figure 3.6c, respectively. The experimental setup
consists of an electrochemical cell with a three electrode configuration. The Ag/AgCl (3M
KCl, World Precision Instruments) reference electrode was placed in vicinity to the working
electrode via a polytetrafluoroethylene (PTFE) capillary. A Pt mesh and a glassy carbon
cylinder were used as counter electrode and a working electrode. The electrochemical cell
was then covered by an X-ray transparent, chemically-inert and -resistant Prolene foil
(Chemplex®, 4 µm). A PTFE tubing beside the reference electrode allows the
pre-CO2-saturated electrolyte flowing inside the sealed cell. The outlet of the electrolyte
shares the glass tubing with the counter electrode. The working electrode is placed on a
PCTFE rod and is held by a slight under pressure which is ensured by a membrane pump. A
demister unit is applied between the sample holder and membrane pump to safe the pump
from flooding in case of electrolyte leakage. The sealing between sample and sample holder is
realized by Viton® O-Ring and the sample is contacted from the backside using a Ti wire.
29
Chapter 3 Aspects and Experimental Procedures
Figure 3.6 a) Schematic illustration of the in situ WAXS setup under electrochemical
condition using at a synchrotron facility, showing the incident and scattered X-rays, the 2D
detector and the grazing incidence cell. b) The side view and c) the top view of the test station
at the European Synchrotron Radiation Facility (ESRF).
During characterization, the electrolyte was continuously pumped through the cell using a
peristaltic pump. We used CO2-saturated 0.1M KHCO3 as the electrolyte. The height of the
working electrode should be carefully adjusted to minimize the scattering and attenuation of
X-ray by the water molecules.
We collected WAXS spectra within 30 s data acquisition time. The experimental protocol was
the following:
1. Diffraction pattern of dry film
2. Diffraction pattern with electrolyte
3. LSV with 5mV/s from -0.02 to -0.84 VRHE
Diffraction patterns were taken between 0 min to 130 min every 10 min while -0.84 VRHE was
applied. Beam damage was controlled after each experiment.
30
Chapter 3 Aspects and Experimental Procedures
3.5.2 in situ (Scanning/Transmission) Electron Microscopy
In this thesis, conventional Transmission Electron Microscopy (TEM) was used in order to
check the morphology/size distribution of as-prepared nanoparticles. The liquid
electrochemical TEM was employed for investing the particles evolution, combining with the
ex situ test cell.
The results using liquid electrochemical TEM techniques and ex situ test cell are presened in
Chapter 7.
Experimental Description
Poseidon Select electrochemical cell holder (Protochips) was used to load the samples into the
microscope and maintain the liquid environment (see Figure 3.7) for the experiment. By using
E-chips, electrochemistry measurement can be performed on a three-electrode system. Both
big and small E-chips are delivered with a protective photoresist coating to prevent damage to
the SiN membrane. Acetone and methanol were used to remove the protective photoresist
coating. The three electrodes are placed on the big chip with glassy carbon as working
electrode, Pt as reference electrode and Pt as the counter electrode. The CuOx catalyst
nanoparticles were dispersed in ethanol and drop cast onto the center of the big E chip with
electrodes, which were plasma cleaned before use.
Figure 3.7 a) Photograph and dimension of the electrochemistry chips with enlarged big
E-chip, in comparison to 1 euro cent. b) Chip design and three-electrode system applied on
31
Chapter 3 Aspects and Experimental Procedures
the big E-chip, showing the glassy carbon working electrode in the center, surrounded by the
Pt reference electrode (grey) and the Pt counter electrode (light grey). c) Assembling of the
liquid E Chem holder with gasket, big/small E-chips and lid. d) Side view of the chip with the
corresponding bottom chip and the electrolyte layer between the chips. e) ex situ cell with
same chip assemble, flow mode and electrochemical connection. f) Gamry potentiostat for the
liquid E Chem holder and ex situ cell. Images are partly ada ed f m “W kf w & T ai i g
P eid Se ec Ve i 1.2; C y igh 2017, P chi , I c.” wi h e mission from
Protochips.
After assembling, solution of 0.1 M Na2HPO4/NaH2PO4 (pH=6.9) (Sigma Aldrich with
milli-Q water) was delivered by an external Hamilton syringe pump through the microfluidic
tubing into the tip of the Poseidon Select TEM holder at 30-300 μL h-1. Then the flow rate
was decreased to 30-50 μL h-1 for vacuum check and real experiments. Microscopy was
performed using an FEI TECNAI G² 20 S-TWIN microscope with a GATAN MS794 P
CCD-detector at 200 kV. Experiments taken with ex situ cell used the same procedure.
After using, a cleaning step needs to be performed. All lines were purged with water for
overnight and rinsed by ethanol. Dry air was used to purge the lines and flush the holder tip.
The chip design in ex situ cell is the same as in situ liquid holder. Thus, the chips used in the
ex situ cell can be re-checked in microscopy with liquid EChem holder after rinsing.
Specifications of Pt Ref (E-Chip ECT-45CR-10)
Calculation of potential setting in Gamry:
EApply = ERHE + Eadj
Eadj = E0-Pt + 0.059V pH [E0-Pt = -1.18 V]
Since the current is quite low in both ex situ cell and liquid E Chem holder, no IR correction
was considered.
3.5.3 operando Differential electrochemical mass spectrometry (DEMS)
For operando differential electrochemical mass spectrometry (DEMS) measurements, inks of
as-prepared CuOx electrocatalysts ink were prepared with a solution of 50 µL 5 % Nafion and
800 µL ultrapure water (18.2 MΩcm, Pu eLab P u Sy em, E ga) and 150 µL iso-propanol.
The ink was sonicated and subsequently certain amount of catalyst ink was deposited onto the
glassy carbon cylinder (diameter, 5 mm) which was polished with 1/0.05 µm Al2O3
32
Chapter 3 Aspects and Experimental Procedures
suspension and cleaned with ultrapure water before. The final catalyst loading was
100 µg cm-2. In this thesis, operando DEMS measurements are used for 1) investigating the
co-feed mechanism on CuOx NPs with (12C/13C) isotope labeling feeds; 2) determining the
onset potential for major products on electrochemically self-reducing CuOx NS. This part of
work is collaborated with Jorge Ferreira de Araújo, from the Technical University of Berlin.
Results from operando DEMS measurements on CuOx NPs for co-feed study are presented in
chapter 5 and on CuOx NS for tracing products onset potential is included in chapter 6.
Experimental Description
Operando Differential electrochemical mass spectrometry (DEMS) was done using a
custom-made TU Berlin electrochemical capillary DEMS flow cell (See Figure 3.8). DEMS
capillary flow cell is characterized by having a well-defined electrolyte profile flow over the
working electrode. The reaction products are transported into the interface liquid vacuum
throughout a 0.15 mm glass capillary. The collection of high concentrated aliquot near
catalyst surface and enhanced liquid-vacuum interface allows a fast and at high-intensity
detection of gaseous products. The high performance with DEMS capillary flow cell results in
the distributed flow over the hydrophobic membrane compartment also known as the cyclonic
flow. The interface liquid/vacuum promoted by PTFE hydrophobic membrane with a pore
size of 20 and thickness of 50 m (Cobetter®, Cat. No. PF-002HS) are commercially
available at Hangzhou Cobetter Filtration Equipment Co., Ltd. The reaction products after
been vaporized into the vacuum chamber from the flow cell were detected using a PrismaTM
quadrupole mass spectrometer (QMS 200, Pfeiffer-Vacuum). The vacuum chamber was
composed of two turbomolecular pumps (HiPace 80) that perform an ultimate pressure 10-6
mbar at MS detectors. Each turbomolecular pump has an independent baking system helped a
by membrane and oil pump (coupled with a molecular sieves oil trap).
33
Chapter 3 Aspects and Experimental Procedures
Figure 3.8 Schematic DEMS setup - custom-made TU Berlin electrochemical capillary
DEMS flow cell.
At the reaction compartment, the flow in of electrolyte should be pre-saturated with gas feeds.
Schematic view of the saturation setup used for this thesis is presented in Figure 3.9. The
system is composed of one/two saturation stack for different feed gases (CO2 and/or CO). The
degasification electrolyte is pumped into the stacks and saturated in parallel and
independently from each gas. After the saturation stack, the electrolyte saturated with CO2
and/or CO is mixed up at the desired ratio before entering the electrochemical flow cell. In
co-feed mechanism study (in Chapter 5), the pure CO2 measurement was performed similarly
but instead of CO was replaced by Argon feed.
Figure 3.9 Schematic view and photography of feed gas control system at DEMS setup -
saturation of electrolyte in the flow stream.
Isotopes labeled gas was also conducted during the co-feed mechanism study. The volume of
isotopes gas could be minimized in a closed system (high ratio of saturated on electrolyte for
34
Chapter 3 Aspects and Experimental Procedures
a volume of gas). To avoid mass fragment overlapping with CO2 fragmentation, we choose to
label 13C on CO instead of CO2 to have a true comparison with pure CO2 feeding and to avoid
carbon mixing with CO2 as KHCO3-derived. Pure 13CO measurement was not performed due
to ubiquitous of CO2 as KHCO3-derived in bicarbonate solutions. We tracked the origin
carbon sources by labeling one of the carbon within co-feed species (13CO) maintaining
constant the CO2 to CO ratio. Our DEMS isotopes experiments are performed in gas feed
ratio of 1 to 3, namely CO2/CO and CO2/13CO.
Onset potential determination methodology
The onset potential is defined as the potential where ion current signal reaches 1% from the
highest detected signal of a correspondent product between all gas feeding systems during a
cathodic voltammetric scan of 5 mV/s.
DEMS-based deconvolution of reactive pathways
The following calculation based on DEMS results is used for co-feed mechanism study.
Detailed discussion of ethylene pathways in the co-feed condition is included in Chapter 5.
The three reactive pathways toward ethylene via the dimerization of two *CO adsorbates
derived from CO2 only, CO only, and of CO2 and CO (cross-coupling) are denoted throughout
a “CO2-CO2”, “CO-CO”, a d “CO2-CO”, e ec i e y. Whi e he e ame abe d
include isotope labels at the carbon atoms, we emphasize that our experimental measurements
and subsequent analysis included and considered isotope labeled 13CO in the feed. A typical
mass spectrum of ethylene gas directly dissolved in 0.1 M KHCO3 and measured at DEMS
setup is presented in Figure 3.10.
35
Chapter 3 Aspects and Experimental Procedures
Figure 3.10 A typical mass spectrum of ethylene gas directly dissolved in 0.1 M KHCO3 and
measured at DEMS setup.
To arrive at a DEMS based kinetic deconvolution we focused on the hydrogen-abstracted
molecular ethylene fragment, denoted as (M - H+) henceforth. Its chemical sum formula
identity is C2H3.+. Under isotope labeling, the three distinct (M - H+) fragments of interest are
then 13CCH3.+ f he “CO2-CO” mecha i m (m/z =28), 12C2H3.+ f he “CO2-CO2”
(m/z=27), and 13C2H3.+ f he “CO-CO” (m/z =29). The e a i e c ibu i f each CO
dimerization pathways was evaluated based on the distribution of specific isotope-labeled
fragments. The ion mass intensities of interest, Im/z=27, Im/z=28, and Im/z=29, were, in part,
deconvoluted from direct measurements or, in case of ion mass overlaps of different
fragments, from relative intensity relations provided.
More specifically, the evaluation of the relative contributions of each mechanism started with
the experimentally measured ion current intensities Im/z 26 (CO2-CO2) (CO2-CO) (CO-CO), Im/z 27 (CO2-CO2)
(CO2-CO) and Im/z 30 (CO-CO), which, as shown by the subscripts, involved various fragment
overlaps from the three dimerization pathways. The extraction of Im/z=27, Im/z=28, and Im/z=29 was
achieved using the following analysis:
Mass 27 (CO2-CO2)
We note that the mass m/z =27 is composed of 50% of the CO2-CO2 and 50% of CO2-13CO
mechanism (see Table 3.4).
36
Chapter 3 Aspects and Experimental Procedures
Table 3.4 All main possible fragments of ethylene including CO and CO2 species. Relative
intensity (R.I.) of various mass fragments and their structural assignment (Assgn.) for
ethylene (C2H4, 13CH2CH2 and 13CH213CH2), carbon dioxide, carbon monoxide and 13C-carbon
monoxide (CO2 and CO and 13CO).
Ethylene
13/12C-Ethylene
13C-Ethylene
carbon
dioxide
carbon
monoxide
13C- carbon
monoxide
Molecule
C2H4
13CH2CH2
13CH213CH2
CO2
CO
13CO
C2H4 m/z 28
C2H4 m/z 29
C2H4 m/z 30
m/z 44
m/z 28
m/z 29
Masses
R.I.
Assgn.
R.I.
Assgn.
R.I.
Assgn.
R.I.
Assgn.
R.I.
Assgn.
R.I.
Assgn.
24
5
C2
.+
25
15
C2H+
5
13CC.+
26
70
C2H2
.+
15
13CCH+
5
13C2
.+
27
68
C2H3
+
70
13CCH2
.+
15
13C2H+
28
100
M.+
68
13CHCH2
+
70
13C2H2
.+
10
CO+
100
M.+
29
100
M.+
68
13C2H3
+
0.1
13CO+
1.2
13CO+
100
M.+
30
100
M.+
0.4
C18O+
44
100
M.+
45
1.2
13CO2
+
The latter results from control experiments that evidenced a 50% decrease of the m/z=26
signal intensity upon addition of labeled 13CO. To evaluate the contribution of CO2-CO to
m/z=26, that is Im/z 26 (CO2-CO), we used the fragment relation intensity factor 0.21 from our
experimental ethylene mass spectrum analysis (Table 3.5).
37
Chapter 3 Aspects and Experimental Procedures
Table 3.5 Resume of the main ethylene fragments used for the DEMS based deconvolution of
the contribution of each reaction pathways with relative intensity (r.i.) and fragment
assignments (Assgn.) of the three *CO dimerization mechanisms resulting in ethylene
formation (C2H4, 13CH2CH2 and 13CH213CH2). The three mechanisms are denoted as
“CO2-CO2” if b h ca b a m f e hy e e de i e f m 12CO2, ”CO2-CO” i ca e f mixed
origin and “CO-CO” if b h ca b a m de i e f m 13CO.
CO2-CO2
CO2-CO
CO-CO
Molecule
C2H4
13CH2CH2
13CH2
13CH2
C2H4 m/z 28
C2H4 m/z 29
C2H4 m/z 30
Masses
R.I.
Assgn.
R.I.
Assgn.
R.I.
Assgn.
26
70
C2H2
.+
15
13CCH+
5
13C2
.+
27
68
C2H3
+
70
13CCH2
.+
15
13C2H+
30
100
M.+
With Im/z 26 (CO-CO) being directed deconvoluted from experimental Im/z 30 (CO-CO), we are now
able to deconvolute Im/z 26 by simple subtraction of Im/z 26 (CO-CO), resulting in Im/z 26 (CO2-CO2).
Im/z 27 (CO2-CO)* = (Im/z 27 (CO2-CO2) (CO2-CO) (CO-CO) - Im/z 30 (CO-CO) x 0.15) / 2
Im/z 26 (CO2-CO)* = 0.21 x (Im/z 27 (CO2-CO)*)
Im/z 26 (CO-CO) = Im/z 30 (CO-CO) x 0.05
Then,
Im/z 26 → Im/z 26 (CO2-CO2)
Im/z 26 (CO2-CO2) = Im/z 26 - Im/z 26 (CO2-CO)* - Im/z 26 (CO-CO)
Im/z 27 (CO2-CO2) = Im/z 26 (CO2-CO2) x 0.97
38
Chapter 3 Aspects and Experimental Procedures
The calculated Im/z 27 (CO2-CO2) represents ion mass signal at m/z=27 form the ethylene molecule
(M) with loss of hydrogen atom; we denoted as M - H+.
Mass 28 (CO2-CO) correction intensity
Having already calculated the mass intensity Im/z 26 (CO2-CO2) and using the fragment intensity
factor 0.97. Now we discard the assumption of 50% CO2-CO2 and 50% CO2-CO as for Im/Z
26 (CO2-CO2) we deconvolute the signal mass 27 (CO2-CO) using the previously determined Im/Z
27 (CO2-CO2). Finally, we determine Im/z 28 (CO2-CO) using an extrapolation based on ethylene
fragment relative intensity signal.
Im/z 27 (CO2-CO) = Im/z 27 (CO2-CO2) (CO2-CO) (CO-CO) – Im/z 27 (CO2-CO2) – Im/z 27 (CO-CO)
Im/z 27 (CO-CO) = Im/z 30 (CO-CO) x 0.15
Im/z 28 (CO2-CO) = Im/z 27 (CO2-CO) x 0.97
The calculated Im/z 28 (CO2-CO) represents ion mass signal at m/z=27+1 (13C) form the ethylene
molecule (M) with loss of hydrogen atom, we denoted as M - H+.
Mass 29 (CO-CO) correction intensity
The signal at m/z 30 is exclusive to the mechanism (CO-CO), allowing us to directly
determine mass 29 based on ethylene fragment relative intensity signal from the experimental
measured Im/z 30 (CO-CO), then Im/z 29 (CO-CO) was determined. The relation factor is 0.68.
Im/z 29 (CO-CO) = Im/z 30 (CO-CO) x 0.68)
The calculated Im/z 29 (CO-CO) represents ion mass signal at m/z=27+2 (2 x 13C) form the
ethylene molecule (M) with loss of hydrogen atom, we denoted as M - H+.
39
Chapter 3 Aspects and Experimental Procedures
3.5.4 operando X-ray absorption spectroscopy (XAS)
The reduction behavior and local atomic information of CuOx NS during CO2RR are
followed by operando XAS at the Cu K-edge. This part of work is collaborated with Dr.
Katharina Klingan, from the Free University of Berlin.
Results from operando XAS measurement on CuOx NS are presented in Chapter 6.
Experimental Description
A few of materials are used as references. Reference powders of CuO (nanopowder, particle
size < 50 nm, surface area 29 m²/g, Sigma Aldrich) and Cu(OH)2 (Sigma Aldrich) were
mixed and grinded with boron nitride and measured at 20 °C in absorption mode at the Cu
K-edge with ionization chambers before and after the samples. Energy calibration was done
by measuring simultaneously a Cu foil (0.001 mm, 99.999%, Goodfellow) and shifting the
energy axis to the first fitted maximum of the derivative of the absorption of the Cu foil. The
metal foil reference was measured in absorption mode as well.
CuOx NS catalysts were prepared on 2x2.5 cm glassy carbon sheets (250 µm thickness,
Sigradur K) and mounted in an in-house made electrochemical Teflon cell. Operando spectra
of CuOx NS samples were collected in fluorescence geometry from the backside of the glassy
carbon electrode. We used a 13 element Si-drift energy resolving detector (RaySpec) which
we equipped with an Al-shielding having a 25x25 mm Ni foil (0.00125 mm, 99.999%,
Goodfellow) in front to suppress scattered light.
The electrochemical cell was controlled by an SP-300 potentiostat (Biologic). We used an
Ag/AgCl reference electrode and a Pt coil as a counter electrode. The CuOx area exposed to
the electrode was 1.96 cm². The 0.1 M KHCO3 electrolyte was purged throughout the
ex e ime wi h ≈ 20 mL/mi -1. A e ia we e c m e a ed f 85% hmic d (R≈40
Ω).
40
Chapter 3 Aspects and Experimental Procedures
Figure 3.11 Photograph of operando XAS test station at Bessy (photo taken by Dr. Katharina
Klingan).
We collected EXAFS spectra to k12.2 within 6 min data acquisition time. The experimental
protocol was the following:
1. Spectrum of dry film
2. Spectrum of film at open circuit potential (OCP)
3. LSV with 5 mV/s from -0.02 to -0.84 VRHE
4. Spectra taken between 0 min to 130 min every 10 min while -0.84 VRHE was applied
During 4. the beam shutter was closed after each spectrum (after 6 min data acquisition time)
and three different sample spots have been used to protect the sample from radiation damage.
The experimental protocol was repeated two times, and the three individual data sets have
been averaged accordingly. Energy calibration was done by measuring before and after the
experimental protocol a Cu foil and shifting the energy axis to the first inflection point of the
fluorescence spectrum of Cu foil.
EXAFS simulations
The extracted spectra were weighted by k3 and simulated in k-space. All EXAFS simulations
were performed using in-house software (SimXLite) after calculation of phase functions with
the FEFF program (version 8.4, self-consistent field option activated). Phase functions were
calculated using geometries from open-access cif files of Cu, CuO, Cu(OH)2, and the
extracted cif files of CuOx NS and Cu(OH)2 from experimental XRD data. As usual, the
EXAFS phase functions did not depend strongly on the details of the used model. Cosine
windows covering 10% at the low-k and high-k side of the spectra were applied before
41
Chapter 3 Aspects and Experimental Procedures
calculation of the Fourier transforms. An amplitude reduction factor (S02) of 0.8 was used. The
data range used in the simulation was 34.3-461.1 eV (3-11 Å-1). The Debye-Waller parameters
for all shells were fixed to avoid overparameterization, and to emphasize the changes in
coordination number of the samples. The fixed Debye-Waller parameters have been chosen as
followed: CuO powder reference, Cu metal reference, and CuOx NS (at OCP) spectra have been
simulated with fixed coordination numbers. The so obtained Debye-Waller parameters have
been used for the simulation of the respective shells. The EXAFS simulation was optimized by
a minimization of the error sum obtained by summation of the squared deviations between
measured and simulated values (least-squares fit). The fit was performed using the
Levenberg-Marquardt method with numerical derivatives. The error ranges of the fit
parameters were estimated from the covariance matrix of the fit, and indicate the 68 %
confidence intervals of the corresponding fit parameters. The fit error was calculated as in
reference117 For calculation of the Fourier-filtered error (described in reference118), the range
from 1 to 6.5 Å on the reduced distance scale was used.
42
Chapter 3 Aspects and Experimental Procedures
43
Chapter 4 Particle Density Study
Chapter 4 Catalyst particle density controls hydrocarbon product
selectivity in CO2 electroreduction on CuOx
A key challenge of the carbon dioxide electroreduction (CO2RR) on Cu-based nanoparticles
is its low faradic selectivity towards higher-value products such as ethylene. Here, we
demonstrate a facile method for tuning the hydrocarbon selectivities on CuOx nanoparticle
ensembles by varying the nanoparticle areal density. The sensitive dependence of the
experimental ethylene selectivity on catalyst particle areal density is attributed to a
diffusional interparticle coupling which controls the de- and re-absorption of CO and thus the
effective coverage of COad intermediates. Thus, higher areal density constitutes dynamically
favored conditions for CO re-adsorption and *CO dimerization leading to ethylene formation
independent of pH and applied overpotential.
44
Chapter 4 Particle Density Study
Chapter 4 and its supplementary information section (Appendix A1) were reproduced with permission
from ChemSusChem, 2017, 10, 4642-4649. Copyright of The John Wiley and Sons (2017).126
Xingli Wang, Ana Sofia Varela, Arno Bergmann, Stefanie Kühl and Peter Strasser, “Catalyst
particle density controls hydrocarbon product selectivity in CO2 electroreduction on CuOx”,
ChemSusChem, 2017, 10, 4642-4649.
P.S., X.W and A.S.V. conceived and designed the experiments. X.W. performed the experiments and
analyzed the data, A.B. helped the Rietveld refinement of XRD patterns; S.K. recorded TEM images;
X.W. and P.S. wrote the manuscript; all authors contributed to the discussions.
45
Chapter 4 Particle Density Study
4.1 Synthesis and characterization
Figure 4.1 Physiochemical characterization of as-prepared CuOx NPs. a) Representative TEM
images of CuOx nanoparticles. The inset shows the size distribution of CuOx nanoparticles. b)
Powder X-ray diffraction (XRD) of the CuOx NPs catalysts between 30° to 95° degree. c)
Weight fractions of solid phases (solid symbols, left axis of ordinates) and evaluated
crystallite size (hollow symbols, the right axis of ordinates) for each phase.
We used a liquid-phase method to reduce Cu2+ precursors in organic solvents to obtain
catalytically active copper oxide nanoparticles (CuOx NPs) with controlled shape and narrow
size distribution. Figure 4.1a shows representative TEM images of the as-prepared CuOx NPs.
The analysis revealed a monodisperse spherical morphology with an average diameter of
10.4±0.97nm. To avoid agglomeration of the as-prepared CuOx NPs, they were stored as a
suspension in hexane at room temperature. The TEM analysis of aged particle suspensions
(Figure A1.1) indicated that the morphology (NP size: 10.23±1.02nm) and monodispersity
remained stable for at least seven consecutive days.
To learn more about the type, number and ratio of the individual Cu oxide phases present in
the NP ensemble, X-ray diffraction and Rietveld refinement were used. A typical XRD
pattern of CuOx NPs is shown in Figure A1.2 suggesting the presence of three crystal phases,
namely metallic Cu, CuO and Cu2O (see crystal structures of these phases in Figure A1.3).
Rietveld refinement was employed to fit the experimental pattern, calculate phase fractions,
lattice constants and oxygen occupancies (see Figure 4.1b, c). The results indicated the
coexistence of 22.3±1.1wt% fcc structured metallic Cu0, 13.1±1.1wt% cubic-Cu2O,
26.4±2.5wt% monoclinic-CuO, as well as 38.1±2.4wt% of oxygen-defective Cu2O1-x. While
the particle preparation was kept strictly air-free, the XRD analysis was performed under air
46
Chapter 4 Particle Density Study
and thus we suspect that oxygen exposure of the metallic Cu particle surfaces may have
contributed to the formation of the observed mix of passivating oxides. To explain the
formation of the non-stoichiometric oxide, we note that adsorbed oxygen atoms on the surface
of metal nanoparticles can enable the generation of oxidic surface compounds and facilitate
metal ion diffusion across the particles, even at low temperature. The diffusion of Cu ions
toward the particle bulk or surface may thus lead to non-stoichiometry and structural
distortion of the crystal parameters. In the case of cubic Cu2O1-x structure, revealed oxygen
occupancy of only 0.593 (Cu2O0.593), indicating an incomplete diffusion of oxygen atoms
inside the unit cells. Crystallite size and coherence length evaluations of the individual Cu,
Cu2O, CuO and Cu2O1-x phases (Figure 4.1c) revealed the presence of ~8nm metallic Cu, as
well as 2-3nm cubic Cu2O, monoclinic CuO and cubic Cu2O0.593 crystallites.
4.2 Electrochemical CO2RR over CuOx catalysts with various areal densities
To investigate their intrinsic catalytic CO2 reduction activity under varying areal catalyst
densities (reported in terms of catalyst mass per cm2 geometric electrode surface area), the
CuOx NPs were coated on a flat supporting glassy carbon working electrode using three
different loadings. All samples were tested in a two-compartment membrane-separated cell
("H-cell"), a three-electrode setup using CO2-saturated 0.1M KHCO3 as reactant and
electrolyte. The electrolyte was kept under CO2 bubbling throughout the experiments at a
constant flow of 30sccm (standard cubic centimeter per minute).
Figure 4.2 Catalytic activity represented by linear sweep voltammograms taken at 5 mV/s in
CO2-saturated 0.1 M KHCO3 for 4 µg cm-2, 15 µg cm-2, 31 µg cm-2 CuOx NPs areal densities,
catalyst layers were supported on glassy carbon. Catalytic activity expressed in terms of a)
47
Chapter 4 Particle Density Study
geometric area-normalized current densities and b) catalyst NP mass-normalized current
densities.
The current densities normalized by geometric surface areas of the supporting electrode are
reported in Figure 4.2a. For all catalyst densities, reductive catalytic currents associated with
hydrogen evolution and CO2 reduction emerged at potentials more negative than -0.6VRHE
(RHE: reversible hydrogen electrode) with an apparent exponential potential dependence.
Around -0.9VRHE, the small and medium areal-particle-density scans exhibited somewhat
declining j-E slopes (activity shoulder) before all three potentiodynamic scans return to rapid
exponential growth. At potentials more negative than -0.8VRHE, bubble formation and
detachment of H2, CH4, and C2H4 gases started to become significant and caused a noisy
current signal. Although geometric current densities represent the performance of the
electrode layer structures, real surface area-normalized or catalyst mass-normalized current
densities are more suitable to compare electrocatalyst performance on an intrinsic materials
characteristic. For monodisperse-size-distributed and homogeneous NP catalysts, mass- and
real surface area-based performance data are equivalent. This is why we shall consider
intrinsic catalytic activities and faradaic product yields on a catalyst-mass basis. Figure 4.2b
shows the corresponding mass-normalized current density scans from -0.1 to 1.0VRHE. Now
the scans suggested an inverted activity trend with the highest areal density (31 μg cm-2)
displaying the lowest catalyst mass-based current density. This evidences that the interfacial
charge transfer did not linearly scale with catalyst areal density. In other words, at higher
catalyst loadings, processes and conditions emerge that limit interfacial charge transfer. We
associate these with mass-transport limitations of CO2 to the electrified catalyst NP surface
possibly linked to local pH increases owing to the hydrogen evolution reaction from protons
or water, as discussed below, according to:
2H++2e-→H2
2H2O+2e-→H2+2OH-
To get insight in the Faradic competition between hydrogen evolution and CO2 reduction at
varying areal catalyst densities, we considered the absolute product yields (catalyst
mass-based production rates) and associated short-term Faradaic efficiencies (FE). To this
end, we performed bulk electrolysis at constant electrode potentials and analyzed the reaction
48
Chapter 4 Particle Density Study
products after 15 min. Details about the chromatograph based efficiency analysis method are
provided in the Supporting Information. Figure 4.3 displays the mass-normalized absolute
production rates of the four major gas products. As the areal catalyst density increased from 4
µg cm-2 to 15 µg cm-2 and 31 µg cm-2, so decreased the H2 formation rate exhibiting a rather
flat logarithmic production (black symbols in Figure 4.3a-c). More importantly for our
discussion here, the areal CuOx catalyst nanoparticle density evidenced a clear effect on the
hydrocarbon product yields during the CO2RR. In particular, the CH4 and C2H4 production
rates revealed opposite trends with particle density: the production rate of C2H4 strongly
increased, whereas that of CH4 decreased. Note also that the onset potential of C2H4
production at around -1.0 VRHE for the low areal density experiment shifted strongly
anodically (smaller overpotentials) by about 100 mV with higher areal density (15 µg cm-2
and 31 µg cm-2). Such a density-dependent onset potential shift was not observed for the CH4
formation.
Figure 4.3 Catalyst mass-normalized absolute product formation rates of major gaseous
products as a function of applied electrode potentials during CO2 electroreduction in
CO2-saturated 0.1 M KHCO3 at CuOx nanoparticle catalyst areal densities of a) 4 µg cm-2, b)
15 µg cm-2, c) 31 µg cm-2. d) Trends in C2H4/CH4 production ratio at varying catalyst areal
densities at -0.95 VRHE and -1.01 VRHE applied electrode potential. Potentials are IR corrected.
49
Chapter 4 Particle Density Study
Figure 4.4 shows the production rate-derived Faradaic efficiencies (FEs) of the major gas
products over the applied electrode potential. Regardless of the areal density, the H2 evolution
reaction is clearly preferred at low overpotentials but shows steadily decreasing FE values
with higher overpotentials up to -1.0 VRHE. Significant FE values of CH4 and C2H4 emerge at
applied potentials more negative of -0.84 VRHE. Consistent with the data in Figure 4.3, the FE
of CH4 drops with higher areal densities in favor of that of C2H4. Linked with these trends are
the FE values of CO that remained lower with catalyst density. The total FEs towards gaseous
products and liquid products for CO2RR and HER are shown in Figure A1.4.
Figure 4.4 Faradaic efficiency for each product as a function of areal density at various
applied overpotentials. Faradaic efficiency of CO2 electrochemical reduction over a) 4 µg
cm-2, b) 15 µg cm-2, c) 31 µg cm-2. d) The Faradaic efficiency ratios of C2H4/CH4 versus
overpotential for the different areal densities.
The stability of the CuOx NP ensembles was evaluated at -1.0 VRHE over 400 minutes at the
areal catalyst density of 31 µg cm-2 which has the largest absolute C2H4 production rate. As
shown in Figure 4.5a, the overall faradaic current dropped 40% during the first 150mins
(Region 1) and then levelled out at -4 mA cm-2 (Region 2). The partial current densities of the
major CO2 reduction products, CH4 and C2H4 (see Figure 4.5b), followed that overall current
50
Chapter 4 Particle Density Study
trend. All three C1 and C2 based partial currents together accounted for about 50% of the total
faradaic charge over the test time. These trends were mirrored by the absolute molar
production rates of the gaseous products (see Figure 4.5c). The total Faradaic efficiencies for
both gaseous and liquid products are shown in Figure A1.5, which can reach 99.3% in the end.
Hence, the unbalanced charge at various overpotentials shown in Figure 4.4 is mainly
attributed to the liquid products.
Figure 4.5 a) Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NPs in CO2-saturated 0.1 M KHCO3 at -1.0 VRHE. b) Partial current densities of CO,
CH4, C2H4. c) Absolute product formation rates of gaseous products over time. d) The
consumption ratio of *CO for CO, CH4, C2H4, respectively. Catalyst areal density: 31 µg
cm-2.
4.3 Discussions
The observed correlation between catalyst particle areal density (geometric loadings) and
product yields can be plausibly rationalized in terms of the local concentrations of reactants at
the electrified particle surfaces, and their effect on the rates of elementary chemical reactions.
In light of prior work on the CO2 reduction mechanism adsorbed CO, denoted as *CO, serves
as a common intermediate of the reaction pathway toward methane and ethylene. After the 2e-
51
Chapter 4 Particle Density Study
reduction of CO2 to *CO, the latter may desorb, form gaseous CO(g) and diffuse into the
electrolyte. Alternatively, *CO may undergo subsequent stepwise proton-coupled electron
transfer (hydrogenation steps) to *CHO,
*CO + H+ + e- *CHO
Followed by reaction to *CHOH or *CH2O, *CH2OH or possibly to CH2O* and finally to
CH4. In a competing process, *CO was proposed to dimerize in a proton de-coupled electron
transfer according to
2 *CO + e- *CO-*CO-
Opening a C2 pathway toward C2H4 via a decouple proton-electron transfer step. Clearly, the
rate of this dimerization depends on the surface coverage of *CO. Now, a larger areal catalyst
density decreases the mean interparticle distance between individual CuOx NPs and increases
the absolute number of electrochemical surface active sites. This is because the CuOx NPs
were applied directly as unsupported metal oxide particles lacking any volume-controlling
carrier matric, such as for instance a porous carbon. The nonlinear scaling of catalyst particle
areal densities and geometric currents (see Figure 4.2a) is another indication that reactant
mass transfer rates were negatively affected by larger areal catalyst densities. At reduced
mean particle distances it is plausible that desorbed CO(g) molecules are more likely to
re-adsorb on the surface of nearby Cu particles rather than leaving the catalyst film as CO gas;
this increases the stationary mean coverage of adsorbed *CO on the catalyst surface. At
meanwhile, real-surface area in the dense areal particle density also increases the absolute
active sites for *CO. As a result of this, the reaction rate along the C2 pathway toward
ethylene would be kinetically preferred as it is second order in CO coverage compared to the
methane pathway (see Scheme 4.1). Indeed, as Figure 4.3d demonstrates, the ratio of C2H4 to
CH4 production rate more than doubles from 0.38 at areal density of 4 µg cm-2 to 0.84 at 15
µg cm-2, and finally almost triples to 1.06 at an areal density of 31 µg cm-2 and an
overpotential of -1.0 VRHE. This trend is even more obvious at -0.95 VRHE, where the ratio of
C2H4 to CH4 production rate increases dramatically from 0 to 2.44 and 5.65. Similar trends
were discernible in the areal density-dependence of the FE ratio of C2H4 and CH4
(FEC2H4/FECH4), a performance figure related to the charge efficiency of ethylene production.
As shown in Figure 4.4d, the (FEC2H4/FECH4) ratio rose with the areal catalyst density values
52
Chapter 4 Particle Density Study
within the given overpotential range. High areal catalyst densities associated with smaller
mean interparticle distances showed favorable charge efficiencies for C2H4 production.
Scheme 4.1 The favored pathway for *CO in low/high areal NP densities condition. In low
areal particle density, *CO tends to leave as CO(g); while in high areal NP density, the
desorbed CO(g) molecules are more likely to re-adsorb on the surface of nearby Cu particles.
As mentioned before, high areal particle density affects reactant mass transport to the
catalytic surface, as well, and typically decrease the local concentration in CO2 and protons,
reflected in observed reduced hydrogen yield per mass of catalysts. However, our data
suggest that even if local depletion of CO2 occurs at higher areal catalyst density, it does not
appear to be able to offset the kinetic benefit of enhanced *CO re-adsorption. The local
depletion of protons, or equivalently the accumulation of local OH- due to hydrogen evolution
and CO2 reduction, on the other hand, can result in increased local interfacial pH values. As
the reductive *CO dimerization reaction is believed to be proton-decoupled, it would not be
affected by a change in local pH while the competing reactions would. As a result, one would
expect a constant production rate of C2H4. As our experiments (Figure 4.3) revealed enhanced
C2H4 formation rates over CH4 (constant overpotential on SHE scale) with areal density, a
local pH effect can be ruled out. The selectivity of one additional catalyst areal density of 100
µg cm-2 is shown in Figure A1.6. The ratio of C2H4 to CH4 production rate (0.98) is similar to
the one of 31 µg cm-2 (1.05). Considering that most nanoparticles tend to overlap with each
other in dense condition leading to a steady mass transport state, the catalyst areal density
cannot be increased infinitely to improve the hydrocarbon selectivity.
53
Chapter 4 Particle Density Study
Recalling that the formation pathways of the three gas products CO(g), CH4 and C2H4 , involve
adsorbed CO molecules (*CO) as a common intermediate, a simple stationary mass balance
relation reads
n*CO = nCO + nCH4 + 2 × nC2H4
Where n denotes the consumed or produced molar amount per time of each product. From
this, a carbon reaction pathway selectivity (ni/n*CO) can be derived (Figure 4.5d) that
expresses the reactive preference of each product. It is evident that the *CO to C2H4 pathway
declines strongly, while more and more *CO simply desorbs and forms molecular CO. The
proportion of *CO for CH4 formation remains essentially constant. These trends suggest that
the catalyst or operation conditions during the test time vary such that the
dimerization/hydrogenation of *CO to ethylene is hindered, while that of generating methane
is unaffected.
Figure 4.6 a) TEM image of CuOx nanoparticles after the chronoamperometric stability test at
-1.0 VRHE from Figure 4.5. b) The mean particle size after 400 min significantly increased
compared to Figure 4.1a. Catalyst areal density: 31 µg cm-2.
In order to assess reasons for the stability trends, we investigated the particle morphology of
the CuOx NPs after the electrolysis at -1.0 VRHE after 400 min. As shown in Figure 4.6a,
particle growth and sintering to a mean size of (18.1±0.28 nm) was apparent (c.f. Figure
4.1a-b). This may occur by catalyst particle migration and coalescence and/or by free energy
difference-driven Ostwald ripening. In the former sintering mode, the translational motion of
entire nanoparticles leads to the collision and coalescence with neighboring particles. The
latter growth mode involves the dissolution of atomic monomers, diffusion and re-deposition
54
Chapter 4 Particle Density Study
on larger neighboring particles. The sintering process slows down as particles grow in size
and interparticle distances increase.119 We are inclined to associate the dramatic decline in
Faradic current during the first 150 mins of stability test with catalyst ripening and aging
processes, before the faradic current levelled out in region 2 where the particle size
distribution remained time stable. During the particle aging, the mean interparticle distance
must have increased to the point where the kinetic *CO-reabsorption and reductive
dimerization advantage on particles in close proximity was lost. *CO was now more likely to
desorb and diffuse away as CO(g) rather than re-absorb and dimerize to ethylene. Consistent
with this view, the methane pathway and production rate remained essentially unaffected
(note the logarithmic scale in Figure 4.5c). In summary, we believe that under our current
conditions the particle size-dependent interparticle distance played the key role for the
evolution of the faradaic product efficiency. At the same time, the real electrochemical active
surface area of the CuOx NPs declined and resulted in a decrease in the overall current density.
The influence of a lower local pH at lower current densities (the reverse pH-ethylene
correlation discussed above) appears limited, as it would affect all proton-consuming reaction
rates.
4.4 Conclusions
In summary, we prepared unsupported CuOx NP ensembles, deployed them in the form of NP
layers, and demonstrated that tuning of the product distribution during catalytic CO2
electroreduction was possible by simply adjusting the areal particle density. Increasing the
areal NP density decreased the mean interparticle distances and increased the total surface
active sites as deduced from geometric voltammetric current scans and TEM imaging. A shift
in faradic efficiency towards C2H4 over CH4 at higher areal density was observed. This
observation was attributed to higher reductive dimerization rates of adsorbed *CO at smaller
interparticle distances, thanks to enhanced CO(g) re-adsorption on NP in close proximity.
Enhanced re-adsorption caused higher mean surface coverages of *CO and higher kinetic
ethylene formation rates ensued. Catalytic performance tests over extended reaction times
corroborated the role of interparticle distances. Over other strategies to tune catalytic CO2
55
Chapter 4 Particle Density Study
reduction reaction product yields, such as varying the catalyst composition or electrolyte,
adjusting the catalyst density appears relatively facile.
56
Chapter 4 Particle Density Study
57
Chapter 5 Co-feed Study
Chapter 5 Mechanistic Reaction Pathways of Enhanced Ethylene
Yields during Electroreduction of CO2-CO co-Feeds on Cu and
Cu-Tandem Electrocatalysts
Unlike energy efficiency and selectivity challenges, the kinetic effects of impure or
intentionally mixed CO2 feeds on the catalytic reactivity of the direct electrochemical CO2
reduction reaction (CO2RR) have been poorly studied. Given that industrial CO2 feeds are
often contaminated with CO, a closer investigation of the CO2RR under CO2/CO co-feed
conditions is warranted. Here, we report mechanistic insights into the CO2RR reactivity of
CO2/CO co-feeds on Cu-based nanocatalysts. Kinetic isotope-labelling
experiments—performed in an operando differential electrochemical mass spectrometry
capillary flow cell with millisecond time resolution—showed an unexpected enhanced
production of C2H4, with a yield increase of almost 50%, from a cross-coupled 12CO2–13CO
reactive pathway. The results suggest the absence of site competition between CO2 and CO
molecules on the reactive surface at the reactant-specific sites. The practical significance of
sustained local interfacial CO partial pressures under CO2 depletion is demonstrated by
(non-)metallic/metallic tandem catalysts. Our findings show the mechanistic origin of
improved C2 product formation under co-feeding, but also highlight technological
opportunities of impure CO2/CO process feeds for H2O/CO2 co-electrolysers.
58
Chapter 5 Co-feed Study
Part of chapter 5 and its supplementary information section (Appendix A2) were reproduced with
permission from Nat. Nanotechnol., 2019, 14, 1063-1070. Copyright of Springer Nature.127
Xingli Wang, Jorge Ferreira de Araújo, Wen Ju, Alexander Bagger, Henrike Schmies,
Stefanie Kühl, Lujin Pan, Ja R mei a d Pe e S a e , “Mechanistic Reaction Pathways
of Enhanced Ethylene Yields during Electroreduction of CO2-CO co-Feeds on Cu and
Cu-Tandem Electrocatalysts”
X.W., J.A. and P.S. conceived and designed this project and co-wrote the manuscript. X.W. carried out
the materials synthesis, characterization and electrochemical evaluation. J.A. conducted the DEMS
measurement and analyzed the results. W.J. and A.B. participated in the discussion of electrochemical
and mechanism sections. H.S. performed the HE-XRD measurement and provided the help of data
analysis. X.W. and S.K. performed the TEM characterizations. L.P. contributes to the electrochemical
tests for CuOx-Ag catalysts. All authors read and commented on the manuscript.
59
Chapter 5 Co-feed Study
5.1 Synthesis and characterization
Figure 5.1 Morphological, structural, elemental and physical characterizations of CuOx NPs
synthesized in this study. a) Representative TEM images of CuOx nanoparticles. b) EDX
mapping images of CuOx NPs. c) Size distribution of CuOx NPs. d) Powder X-ray diffraction
(XRD) of the CuOx NPs catalysts with Rietveld refinement. e) Weight fractions of the solid
phases.
For our investigation of the kinetics, product yields and mechanistic pathways of the
electrochemical CO2 reduction reaction (CO2RR) under CO/CO2 co-feeding conditions,
spherical Cu-oxide nanoparticle catalysts were prepared and utilized (See Session 3.11) with
an extra drying process in Frozen dryer. We note that this report is not concerned with the
dynamics of the chemical state of the Cu catalyst under CO/CO2 co-feeding, which is why we
refrained from in-depth operando analytics of the operating catalysts. Yet, to reveal the initial
characteristics of the Cu catalyst employed, Figure 5.1a-c displays transmission electron
microscopy (TEM) images, particle size distributions and scanning transmission electron
microscopy / X-ray energy dispersive spectroscopy (STEM/X-EDS) elemental maps of the
0
10
20
30
40
50
Cu2O1-x
CuO
Cu2O
X / wt%
Cu
2 4 6 8 10 12
Intensity / a.u.
2q @ 78 keV / Degree
Experimental pattern
Fitted pattern
Diff
Cu
Cu2O
CuO
Cu2O1-x
7 8 9 10 11 12
Size (nm)
O K
Cu K Overlap
ab
c
de
60
Chapter 5 Co-feed Study
spherical CuOx nanoparticles (CuOx NPs). The unsupported CuOx NPs with their narrow size
distribution of 9.4±1.1 nm (Figure 5.1c) were synthesized using a liquid-phase, ambient
pressure route under N2 atmosphere, followed by self-oxidation upon exposure to air at room
temperature. The elemental maps of the NPs in Figure 5.1b evidenced an oxygen-rich region
near the surface of the CuOx NPs. Metallic Cu was present and preserved in the center of the
NPs due to the passivating effect of the outer Cu oxide layer. Rietveld refinement-based phase
analyses of synchrotron WAXS patterns confirmed the presence of non-stoichiometry and
structural distortion of the crystal parameters, and allowed the extraction of the exact type,
number, and ratio of the individual Cu oxide phases present in the NP ensemble. Pattern
deconvolution displayed in Figure 5.1d revealed the co-presence of three crystal phases,
namely 18.0±2.9wt% of fcc structured metallic Cu, 33.4±5.3wt% of cubic-Cu2O and
39.2±6.2wt% of monoclinic-CuO. The oxygen occupancies were also refined in the crystal
phases, revealing oxygen occupancy of only 0.821 in Cu2O1-x, indicating an incomplete
diffusion of oxygen atoms inside the unit cells.
5.2 Feeding gases control
5.2.1 In H-cell design
As the working component of H-cell is sealed tightly, the partial pressure of CO2/CO gas can
be adjusted by controlling the individual gas flow rate when the total flow rate is kept at
30 sccm constantly. With the increased amount of CO co-feeding, the feed-gas-ratio is tuned
from 2:1 (CO2 to CO), 1:1 to 1:2. Accordingly, the concentration of CO2 or CO in the
electrolyte will be dependent on the partial pressure of the gases in the atmosphere (298 K,
101.3 KPa).
In pure CO2 condition, c (CO2) = 1/KHenry-CO2 * PCO2 = 33.6 mM
In pure CO condition, c (CO) = 1/KHenry-CO * PCO = 0.95 mM
61
Chapter 5 Co-feed Study
Table 5.1 Gas flow control of various feeds and their relevant ratios of dissolved COx species
acc di g He y’ aw.
CO2 Flow
(sccm)
CO Flow
(sccm)
Volumn
(mL)
Con. CO2
(mmol)
Con. CO
(mmol)
CO2 (aq):CO
(aq) Ratio
30
0
40
1.344
0
-
20
10
40
0.892
0.013
69:1
15
15
40
0.672
0.019
35:1
10
20
40
0.448
0.025
18:1
0
30
40
0
0.038
-
5.2.2 In DEMS flow cell
Instead of controlling the dissolved COx species by flow gas partial pressure in H-cell, it can
be achieved by a home-made mixer (see Session 3.5.3) in DEMS setup as well.
Table 5.2 Electrolyte flow control of various feeds and their relative ratios of dissolved COx
species in DEMS setup.
Saturated Electrolyte Flow
rate CO2: CO
CO2 (aq): CO (aq) Ratio
1:1
31:1
1:3
12:1
5.3 Electrocatalytic rates and product yields under pure and mixed CO2/CO
co-feeds
We now turn to the electrochemical kinetics of the CuOx NPs under pure CO2 feeds, pure CO
feeds and CO2/CO co-feeds at constant total mass flow in buffered neutral (pH=6.8)
electrolytes evaluated in a two-compartment, membrane reactor cells. U i g He y’ aw, he
molar amounts of dissolved CO and CO2 in the electrolyte were calculated for each feed
condition and are plotted in Figure 5.2a. Clearly, the effective total molar amount of dissolved
62
Chapter 5 Co-feed Study
COx (CO and CO2) in the electrolyte decreases with increasing CO partial pressures (Table
5.1). As the NP catalyst loadings were kept constant in all experiments, any effects regarding
surface roughness could be ruled out in our analysis and discussion. To assess the catalytic
activity and selectivity of the CO/CO2 reduction process, we measured the electrochemical
COx reduction rates under the chosen CO2-CO co-feeds under steady-state conditions at
constant potential electrolysis at -1.0 VRHE for multiple hours at room temperature. Figure
5.2b-c displays the time trajectories of absolute production rates of ethylene and methane
during the tests. CH4 production rates strongly depended on feed compositions and followed
the order CO > CO2-CO (co-feed) > CO2. Data revealed that methane production rates were
closely correlated with the experimentally determined H2 production rate (see Figure A2.1),
indicating a possible direct dependence of CH4 formation on the reactant redox state, the
available surface *H coverage, and local proton concentration. The six-electron CO-to-CH4
reaction cascade is expected to proceed faster than the 8-electron pathway starting from CO2.
Further, the dominant hydrogen evolution under pure CO feed conditions can be explained by
higher local pH values under reaction conditions combined with low CO solubilities. Key to
our discussions here, the C2H4 formation rates were significantly promoted under CO2/CO
mixed feeds. Control measurements confirmed that any CO2/CO feed ratio invariably
benefited the ethylene yield (Figure 5.2d). Yields of a number of other relevant liquid
products formed during CO2/CO co-feeding are shown in Figure A2.2 and Figure A2.3. Upon
a 2e- reduction, either formate, HCOO-, or adsorbed CO, denoted henceforth as *CO, was
formed in all feeds containing CO2. Under co-feeding, the partial production rate of HCOO-
was proportional to the CO2 partial pressure, while, obviously, no HCOO- could be formed in
pure CO feeds. Unlike the terminal 2e- reaction pathway to HCOO-, *CO may desorb to
molecular gaseous CO(g) or alternatively, undergo subsequent stepwise proton-coupled
electron transfer pathways to CH4 or alternatively may dimerize in a proton de-coupled
electron transfer pathway to C2H4. The overall production rate of both, CH4 and C2H4
( efe ed he e a he “hyd ca b a e”), is plotted in Figure 5.2d, as well. It shows the
lowest value in pure CO2 feeds, in part because the HCOO- production rate was highest. With
increasing CO in the feed, the HCOO- pathway decreased, while the hydrocarbon yield
sharply increased by about 50%, despite lower solubility of CO and, hence, a lower total COx
63
Chapter 5 Co-feed Study
electrolyte concentration. Thus, COx solubility arguments obviously fail to account for the
observed increase in C2H4 yields under co-feeding. Earlier studies revealed that pure CO
feeds favored proton accessibility to reactive adsorbed intermediates, and, as a result of this,
concluded that C2H4 formed via a hydrogenated dimer (*CO-COH) generated in a
consecutive electron-proton McMurry coupling-type transfer at moderate overpotentials,
which constituted a distinct reaction channel from the CH4 generation pathway. This is why
we are inclined to attribute the observed higher CH4 production rates in CO-containing feeds
to a more favorable proton affinity to *CO, as well. Figure 5.2e displays a volcano-type trend
between ethylene production rate and CO2 partial pressure. The vastly different slopes of the
relation at small and large CO2 partial pressures clearly suggest the existence of an optimum
CO2 to CO co-feed ratio that exhibits maximum C2H4 yields.
Figure 5.2 Catalytic performance detected by on-line GC in H-cell. a) The molar mass of
dissolved CO2 (aq) and CO (aq) in the 40 mL electrolyte with various feed gases according to
He y’ aw a 1 a m 25 °C. The total flow rate of feed gases used in all experiments is 30
sccm. Time-dependent absolute product formation rates on b) C2H4 and c) CH4 at ~ -1.0 VRHE.
Orange: CO2RR in CO2 saturated 0.1 M KHCO3 (pH=6.8); cyan: CORR in 0.1M
K2HPO4/KH2PO4 (pH=6.9); Violet: co-feeds (CO2/CO) reduction reductions in CO2 saturated
0.1 M KHCO3. Catalyst loading: 100 ug cm-2 on glassy carbon. d) C2H4 production rate (left)
and carbon rates on hydrocarbons (CH4 + C2H4) (right) with various feed gas ratios after 4h.
CO2 to CO partial pressure ratios in co-feed vary from 2:1, 1:1 to 1:2, leading to the ratios on
dissolved CO2 (aq) to CO (aq) in the electrolyte from 69:1, 35:1 to 18:1. e) ln(PCO2) vs.
ln(RateC2H4). The error bars are given as standard error of the mean.
050 100 150 200 250 300 350
0.0
0.2
0.4
0.6
0.8
1.0
C2H4
Production Rate / nmol cm-2 s-1
Time / min
CO
CO2/CO
CO2
050 100 150 200 250 300 350
0.1
1
10
CO2
CO2/CO
CO
CH4
Production Rate / nmol cm-2 s-1
Time / min
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0
-0.8
-0.7
-0.6
-0.5
-0.4
Slope 2 = 0.24
Slope 1 = -0.52
ln(PCO2/P0)
ln (RateC2H4)
0.01 0.1 1
CO
CO2
Dissolve COx / mmol
0.8 0.6 0.4 0.2 0.0
C2H4
Production Rate / nmol cm-2 s-1
0.0 0.5 1.0 1.5 2.0
CH4 + C2H4
Production Rate / nmol cm-2 s-1
e
c
b
d
CO2
CO2/CO (2:1)
CO2/CO (1:1)
CO2/CO (1:2)
CO
CO
CO2/CO (1:2)
CO2/CO (1:1)
CO2/CO (2:1)
CO2
a
64
Chapter 5 Co-feed Study
We then focused on the hydrocarbon yield in 1:1 CO2/CO co-feeds as well as pure feeds over
a wider electrode potential range (Figure A2.4). Favorable CH4 production / H2 production
with CO-containing feeds were observed over the entire electrode potential range investigated,
supporting our earlier mechanistic conclusions from Figure 5.2d and Figure A2.1. In CO2/CO
co-feeds, the experimental C2H4 yields first tracked those of pure feeds, until, at potentials
cathodic of -0.93 VRHE, they sharply outperformed them unfolding their kinetic benefits. In
pure CO feeds, the C2H4 yields remained low, suggesting a CO mass-transport limitation
consistent with the low CO solubility and local CO concentration at the interface.
After 2 and 4 h of electrocatalytic COx reduction catalysis at constant applied electrode
potentials, we tracked the changes in the crystalline states and morphology of the initial CuOx
NPs. Data in (Figure A2.5, A2.6, A2.7) evidenced that under pure feeds the resulting Cu
phases and particle morphologies were very similar under either pure feed. The oxidized
copper phases vanished and transformed into a crystalline metallic Cu phase in both reduction
processes, accompanied by a very similar particle agglomeration for all feeds. This led us to
conclude that the observed differences in the catalytic ethylene yields have indeed more likely
kinetic mechanistic roots, and do not arise from catalyst-related chemical factors.
We also investigated the influence of the pH value on the COx electroreduction kinetic in pure
COx feeds (Figure A2.8), and we found that the C2H4 formation, unlike CH4, from pure CO
feeds behaved pH-independent on the SHE scale (Figure A2.8c), suggesting a
proton-decoupled rate determining electron transfer step.74, 75 The C2H4 production exhibits an
uphill trend in CO2RR and slightly downhill trend in CORR, suggesting that the C-C
coupling is limited by CO mass transport.
5.4 Discussions
Recalling the enhanced C2H4 production in co-feed in the more negative potential region,
which is even more obvious than CO2RR, it indicates that the extra CO in the CO2 gas may
help to break the mass-transport-limitation. Unlike the pure feed conditions, C2H4 in co-feed
condition may originate from CO-CO combination, CO2-CO2 combination or crossed
CO-CO2 combination (see Figure 5.3). In order to validate this hypothesis, we tracked and
65
Chapter 5 Co-feed Study
quantified the origin of carbon sources by labelling one of the carbons from co-feed species
by using electrochemical capillary DEMS flow cell.
Figure 5.3 The possible carbon origins in pure feeds and co-feed during our DEMS
experiments. In co-feed condition, carbon origins c u d g h ugh “CO2-CO2” c mbi a i ,
“CO-CO” c mbi a i a d/ c ed “CO2-CO” combination pathway to C2H4.
Quantitative deconvolution of competing mechanistic C2H4 formation pathways
Under CO2/CO co-feed conditions, the detected C2H4 product molecules originate from the
dimerization of two CO species that must be located in atomic proximity. The CO species
may originate from either y CO (he e efe ed a he “CO-CO” a hway), f m y
CO2 (he e efe ed a he “CO2-CO2” a hway) feed m ecu e , e e ha e a mixed igi ,
that is, one *CO derived from CO, while the other *CO was from CO2 (referred to as cross
c u i g “CO2-CO” a hway). The i i ia c ce a i a i f di ed COx (CO and CO2)
in the electrolyte and at the reactive interface are listed in Table 5.2. To obtain a more detailed
mechanistic understanding of the enhancement mechanism of the ethylene production in
co-feeds, we developed and utilized a novel operando DEMS set up (Figure 3.8) equipped
with a newly designed CO2RR capillary cell with a time resolution of ion mass current
dynamics in the milli-second range. This technique enabled us to track and quantify the
atomic origins of the two individual carbon atoms in the ethylene product molecules under the
two pure and mixed CO2/CO co-feeds. We achieve this by using isotope-labelled 13CO in the
feeds. This strategy avoided complications related to mass fragment overlaps.
First, our time-resolved DEMS analysis was used to verify the significantly enhanced C2H4
yields during cyclic voltammetric electrode potential scans, confirming the enhanced
stationary C2H4 yields using online GC analysis discussed above. Figure 5.4a displays
time-resolved DEMS ion mass current scans of C2H4 and CH4 in mixed 12CO2/12CO (1:1,
partial pressures of CO2 and CO 50 kPa each) co-feeding conditions, in comparison to pure
12CO (100 kPa) and 12CO2 (100 kPa) feeds. The mixed feed gave rise to greatly enhanced
C2H4 ion mass currents (detected via the hydrogen-abstracted molecular (M–H)+ fragment
66
Chapter 5 Co-feed Study
throughout) on the cathodic and anodic scan direction. Interestingly, the C2H4 ion mass
currents on the cathodic and anodic sweep branch appeared quite asymmetric, suggesting
local CO depletion and, as a result of this, decreased *CO surface coverage at the most
cathodic applied turning potential of -1.0 VRHE. This depletion is likely due to CO diffusional
mass transport limitations at the large overpotentials.
To account for differences in the individual partial pressures of CO and CO2 under mixed and
pure feeds, we then first compared the ethylene DEMS mass ion sweep under a 12CO2/12CO
(1:3, 25 kPa 12CO2) co-feed (violet sweep curve in Figure 5.4b) with that under a 12CO2/Ar
(1:3, 25 kPa 12CO2) feed (orange curve in Figure 5.4b). We deliberately utilized a somewhat
lower 12CO2 partial pressure to highlight the kinetic effects of CO in the feed. The measured
DEMS data evidenced that the cathodic ethylene production flanks under 12CO2/12CO co-feed
and 12CO2 feed (both at 25 kPa 12CO2) followed each other closely, evidencing that ethylene
largely originates from dimerization of CO2-derived surface *CO in cathodic sweep direction.
A a e y, he “ e f-feedi g” u y f CO by CO2 to CO reduction is sufficient to maintain
the ethylene formation. However, on the anodic sweep, only the 12CO2/12CO co-feed was able
to continue to generate comparable amounts of ethylene (purple curve), proving the
suppression of local CO depletion at the catalytic interface.
Next, keeping the partial pressures of 25 kPa CO2 and 75 kPa CO in the feed constant, we
moved to an investigation of isotope-labelled mixed 12CO2/13CO co-feeds, 12CO2/Ar, and
13CO/Ar feeds at comparable partial pressures to arrive at an accurate deconvolution of the
origin of the ethylene carbon atoms associated with a quantitative assessment of the relative
67
Chapter 5 Co-feed Study
Figure 5.4 Quantitative deconvolution of relative contributions of competing CO-CO
dimerization reaction pathways to ethylene using a novel operando DEMS capillary flow cell
system. Differential electrochemical mass spectrometry (DEMS) sweep data obtained during
COx reduction a CuOx NP catalysts (supported on a flat 0.785 cm2 glassy carbon electrode)
using CO2/CO co-feeds, pure CO2 and pure CO feeds during a cyclic voltammetric scan at
5mV s-1 and electrolyte flow of 5 µL s-1. a) DEMS ion mass current sweeps over time: in the
top, ethylene (molecule-H+, C2H3+ fragment at m/z=27), in the middle methane (CH3+ m/z=15)
measured under 12CO2/12CO (1:1) co-feed (violet), pure 12CO2 (orange) and pure 12CO feed
(cyan), and in the bottom, the concurrent cyclic voltammetric sweep plotted over time.
Kinetic onset potentials are listed for each feed system as Efeed, referenced to RHE. b)
comparison of DEMS ion current sweeps over time of ethylene-related molecule-H+, (M-H+),
fragments in 12CO2/12CO (1:3) co-feeds (soft purple curve) and the corresponding 12CO2/Ar
68
Chapter 5 Co-feed Study
(1:3) feed maintaining 25 kPa CO2 partial pressure (soft orange curve), and in the bottom, the
concurrent cyclic voltammetric sweep plotted over time. The deconvoluted DEMS ion current
sweeps over time of the three possible M-H+ ethylene fragments resulting under
isotope-labelled 12CO2/13CO co-feeds are shown by the blue, green, and red sweep profiles:
13C12CH3·+ (blue curve) represents the mechanistic pathway involving dimerization of one
13CO-derived and one 12CO2-derived *CO (cross-coupling) according to scheme M1,
12C12CH3·+ (green curve) represents the mechanistic pathway involving dimerization of two
12CO2-derived *CO, shown in scheme M2, and 13C13CH3·+ (red curve) arises from the
mechanistic pathway involving dimerization of two 13CO-derived *CO shown in scheme M3.
The values potential onset are represented for each mechanism, referred to as Emechanism, with
potential referenced to RHE. c) The ratio of reaction between the three mechanisms CO-CO,
CO2-CO2 and CO2-CO, with respective potential range of integration and with steps of 0.05 V
from cathodic till anodic sweep. d) DEMS-derived ethylene mass charges for the three
different feeds during the cathodic and anodic voltammetric sweep (right side), and the
respective net CO2 mass charges (CO2 conversion) on the left side. The purple ethylene
charge bars split into the relative contributions of each ethylene formation pathways (blue,
green, red hashed patterns) from partial Figure 5.4b. e) Comparison of the DEMS – derived
cyclic mass ion current sweep of ethylene under CO2/CO (1:3) co-feeding, plotted in the
potential domain (black line) with the sum of the three DEMS-derived ion mass sweeps of
ethylene isotopes from each mechanistic pathway (blue, green, red colored dash line). The
insert presents the Tafel lines of each partial reactive pathway of ethylene formation with
Tafel slope values in mVRHE dec-1.
contributions of the three competing ethylene formation pathways (see Session 3.5.3). Figure
5.4b compares the three individual ethylene ion mass current sweeps in the time domain (red,
blue, and green curves) corresponding to the three distinct reactive pathways over one cyclic
voltammetric potential scan (red profile at the bottom). The red sweep denotes the sweeping
trajectory of the mass current of the 13C2H3·+ fragment (Im/z 29 (CO-CO)) associated with the
“CO-CO” a hway whe e CO igi a e e i e y f m he i pe labelled 13CO feed. The
green sweep is the C2H3·+ fragment (Im/z 27 (CO2-CO2)) f he “CO2-CO2” pathway, where
adsorbed CO stems entirely from non-labelled CO2. Finally, the blue sweep is the 13CCH3.+
fragment (Im/z 28 (CO2-CO)) related to the cross-coupling “CO2-CO” a hway whe e e hy e e i
formed from CO originating from CO2 and CO. Alongside the mass current sweeps of the
three C2H3·+ ethylene fragments, Figure 5.4b also schematically illustrates the corresponding
three CO dimerization pathways to ethylene in the respective color. Obviously, the
contribution of each mechanistic pathway is a complex function of the applied electrode
potential and scan direction.
69
Chapter 5 Co-feed Study
The milli-second time-resolved measurements of catalytic ethylene sweeps using the DEMS
capillary cell enabled us to accurately evaluate the kinetic onset potentials of ethylene for
each mechanism separately (Figure 5.4b). To the best of our knowledge, there has been no
time-resolved, transient kinetic study to date, where reaction mechanisms with kinetic onset
potentials were reported for, as virtually all previous reports used steady-state or time-delayed
NMR and chromatography-based product analysis techniques. Our analysis revealed that the
C2H4 onset potentials shifted anodically by about 140 mV when operating in pure CO feeds
(cyan sweep, ECO= -0.67 V RHE) compared to pure CO2 feeds (orange sweep, ECO2= -0.81 V
RHE), evidencing a lower kinetic activation barrier of the catalytic CO to C2H4 cascade.
Integration of the three individual DEMS-derived deconvoluted ion mass current sweeps from
Figure 5.4b yielded the total ethylene charges, QMS, of the pathways, which could be
converted into relative charge contributions. The ratio of reaction in Figure 5.4c presents, the
ratio of resulted ethylene charge from intervals of 0.05 V and for each mechanism CO-CO,
CO2-CO2 and CO2-CO. High production ratio over a large potential range is observed for the
CO2-CO mechanism, demonstrating a continuous contribution (Figure A2.9). The right-hand
bars in Figure 5.4d presents, the total ethylene charges (QMS, ethylene) under the three different
feed conditions during the cathodic (top) and anodic (bottom) sweep separately. The length of
the purple bars underlines the enhanced ethylene production under CO2/CO co-feed
conditions compared to the orange and blue bars of the pure CO2 and CO feeds. From the
QMS (ethylene) values, we evaluated the relative contribution of the cross-coupling “CO2-CO”
pathway to the total ethylene generation to as high as 45% over the entire sweep, compared to
33% f he “CO2-CO2” a d a me e 22% f he “CO-CO” a hway in Figure A2.10. A
breakdown of the relative contributions of each pathway during the anodic and cathodic scan
are provided in Table 5.3. In other words, two thirds (67%) of the total ethylene yield can be
kinetically directly attributed to the presence and reactivity of gaseous CO in the reactant feed
during COx electroreduction. Also in Figure 5.4d, the integrated m/z = 44 ion mass charge of
CO2 (QMS (CO2 conversion)) was evaluated for cathodic and anodic scan separately.
Interestingly, CO2 conversion on the cathodic scan is similar for co-feeds and pure CO2 feeds,
while in the anodic scan direction, the conversion of CO2 becomes 3x larger in the presence of
CO, consistent with the larger ethylene formation charge in Figure 5.4c. Evidently, the
70
Chapter 5 Co-feed Study
Table 5.3 Comparison of DEMS-derived kinetic reaction parameters of the three CO
dimerization pathways to ethylene depending on the chemical origin of the CO.
Cross-coupling (CO2-CO) pathway: The dimerizing CO intermediates derive from CO2 and
CO; (CO2-CO2) pathway and (CO-CO) pathway: CO originated only from CO2 or CO,
respectively; iMS (mA cm-2 s-1) is the respective maximum C2H3+ion current , QMS (As) is the
DEMS charge under the deconvoluted C2H3+ion current sweeps under CO2/CO co-feeding in
cathodic and anodic scan direction. “Tafe ” de e he ex e ime a y a a e (i ma
current derived) Tafel slopes (VRHE dec-1), while E (in units of VRHE) is the experimental onset
potential of ethylene formation.
Reactive
CO
dimerizatio
n pathway
Anodic scan direction
Cathodic scan direction
iMS
QMS
iMS
QMS
Tafel
E
CO2-CO
1.5 ±0.1
0.11 ±0.012
2 ±0.2
0.07 ±0.006
89
-0.83
CO2-CO2
1 ±0.4
0.09 ±0.019
1.5 ±0.2
0.05±0.008
65
-0.84
CO-CO
0.63 ±0.05
0.06 ±0. 009
0.77 ±0.05
0.05 ±0.013
215
-0.72
presence of CO is able to maintain higher CO2 reduction rates even at the most cathodic
electrode potentials, where local reactant depletion occurs under the pure CO2 feed. To check
the accuracy of our mechanistic pathway deconvolution, Figure 5.4e compares the
experimentally measured total ethylene ion mass sweep (solid black line) with the sum of the
three individual C2H3·+ ion mass fragment sweeps (dashed curve with color coding for each
mechanism). Clearly, the three contributions reproduce the experimental overall ethylene
sweep curve very well. The inset of Figure 5.4d displays mechanism-specific DEMS-derived
E-log(iMS) Tafel plots and fitted linear Tafel lines with slopes of 81, 125 and 184 mV.dec-1 for
CO2-CO, CO2-CO2 and CO-CO respectively. Thus, the lowest Tafel slope was found for the
cross-coupling CO2-CO pathway, underlining the facile reaction kinetics of the dimerization
of two CO molecules from distinct origins. The CO-CO mechanism, with its lowest onset
potential, showed a significantly less favorable Tafel slope with the most significant voltage
loss per current decade, likely based on the extremely low CO solubility and rapid local
interfacial CO depletion affecting the apparent Tafel slope (see Table 5.3).
71
Chapter 5 Co-feed Study
Co-feed mechanism
Figure 5.5 Mechanistic hypotheses of enhanced ethylene production under CO2/CO reactant
co-feeds. The demonstration of possible dimerization pathways with common intermediates
in CO2 feed, co-feed and CO feed.
Figure 5.5 details the mechanistic implications and conclusions resulting from our
time-resolved kinetic analysis of the electrocatalytic CO2-CO-C2H4 reaction cascade,
sketching plausible reaction pathways and vital intermediates of the three feed conditions.
First and foremost, the DEMS results evidence that the C2H4 production under CO2/CO
co-feeds originates predominantly from dimerization of two adsorbed *CO molecules to form
a neutral OC-CO* or a previously reported negatively charged -OC-CO* adduct,84 which is
further reduced to ethylene. The dominant cross-coupling Langmuir-Hinshelwood-type
mechanism (de ed “1” i Figu e 5.5) with the reactive *CO stemming from both CO2 and
CO is marked as the blue dimerization pathway. The fact that CO in the feed, CO(g), boosts
ethylene yields via a dominant cross-coupling pathway despite a much lower combined COx
concentration at the liquid-solid interface, speaks to non-competing adsorption of CO and
CO2. This suggests the existence of two distinct, non- c amb i g “ eac a - ecific” u face
adsorption sites in atomic proximity on the operating Cu nanocatalyst, one for the adsorption
of dissolved CO2 and another for dissolved interfacial CO. This is a rather unexpected, yet
interesting conclusion with important implications for catalyst and electrolyte design. Also,
a ia y i h m ge e u di ibu i f “ eac a - ecific” *12CO and *13CO sites on the
catalyst surface would consequently result in products enriched in one or the other isotope, as
recently reported by Lum and Ager105.
72
Chapter 5 Co-feed Study
I addi i u “ w - i e” hy he i i i g La gmui -Hinshelwood type pathways, a
Eley-Rideal (ER) type reaction pathway, involving C-C coupling (de ed by “2” i Figu e
5.5) of an adsorbed *CO and a solvated CO(g) at the outer Helmholtz layer, could, in
principle, account for or at least contribute to the enhanced ethylene formation rates
according:
*CO + CO(g) OCCO*.
Following what is known on similar ER mechanisms,74, 82, 120 we cannot fully exclude that
increasing CO(g) concentrations at the double layer may favor ER-type C2H4 production in
parallel to the Langmuir-Hinshelwood pathway. Data in Figure A2.11 showed that increasing
the local interfacial CO(g) concentration under CO-rich CO2/CO (1:3) co-feeds decreased the
net CO formation due to faster C-C coupling and ethylene formation.
Finally, DFT calculations indicated that yet another mechanistic coupling pathway to C2H4,
though energetically somewhat less favorable than the *CO dimerization on Cu(100) facets,82
may open up on non-Cu(100) facets under co-feed conditions with high CO partial pressures.
This pathway involves the coupling of adsorbed *COOH and *CO and becomes energetically
more and more favorable at large *CO coverages and high local CO(g) concentration. This
new C2 pathway of *COOH and *CO/CO(g) may become energetically possible on facets
other than Cu(100) facets and proceeds according to
*COOH + *CO/CO(g) *OCCOOH
in parallel to the existing *OCCO dimerization pathway (see Figure 5.5).
5.5 Extended co-feed mechanism on bifunctional tandem catalyst design
The co-feed mechanism also provides a strategy to design an active catalyst towards C2H4. To
mimic the co-feed condition, a catalyst with high CO production could be used as a
substitution of CO gas feed in CO2RR. In this session, Nickel-nitrogen-functionalized carbon
material (NiNC) and Ag were chosen as the tandem partner for their excellent CO activity, to
increase the local CO concentration on the CuOx NPs interface.
73
Chapter 5 Co-feed Study
Figure 5.6 The realization of enhanced ethylene production using internal CO2/CO “ e f
co-feedi g” u i g non-metallic/metallic tandem catalyst design. a) The tandem catalyst design
is combining NiNC material as a local CO-producer and CuOx NPs on carbon paper electrode. b)
C2H4 production rate with the bifunctional hybrid catalyst with carbon paper as working electrode
for CO2RR at the various component and fixed overpotentials. The experimental C2H4 production
rates at -0.84 VRHE are multiplied by 5 for comparison. No detectable C2H4 formation on pure
CuOx NPs with areal loading of 100 g cm-2 at -0.84 VRHE. No detectable C2H4 formation on pure
NiNC catalyst at both -0.84 VRHE and -0.9 VRHE.
Figure 5.6a displays the first examples of a bifunctional non-metallic/metallic tandem catalyst.
High-surface area Ni-nitrogen-functionalized carbon (NiNC) acts as selective CO-producer47
and supports CuOx NPs with reactant-specific surface sites. The SEM images of the
as-prepared CuOx-NiNC tandem catalyst are shown in Figure A2.12, illustrating the location
of the two distinct components. The neighbouring CuOx NPs and NiNC are among
micrometer scale while partially overlapped with each other. The catalytic C2H4 performance
of CuOx-NiNC tandem catalysts was significantly enhanced compared to supported CuOx
NPs. As shown in Figure 5.6b, the tandem CuOx-NiNC (1:4) catalyst features the same C2H4
production yields at half the mass loading of CuOx NP. Similarly, CuOx-NiNC (1:2) catalysts
yield twice the C2H4 production rate. At overpotentials of -0.84VRHE, CuOx-NiNC tandem
catalysts produce considerably less free gaseous CO (Figure A2.13a) compared to pure NiNC,
evidencing that some of the internally generated CO are immediately consumed by the
tandem catalysts. The CH4 yield increased in the tandem catalysts, as well (Figure A2.13b).
The liquid products analysis for the CuOx-NiNC tandem catalysts is included in Table A2.1.
74
Chapter 5 Co-feed Study
Figure 5.7 The realization of enhanced ethylene/methane production using internal CO2/CO
“ e f c -feedi g” u i g metallic/metallic tandem catalyst design with fixed CuOx areal
particles density of 100 g/cm-2. a) The catalyst is combining Ag material as a local
CO-producer and CuOx NPs on glassy carbon electrode. b) C2H4 and CH4 production rate
with CuOx-NiNC catalyst for CO2RR at the various component and fixed overpotentials.
Error bars are given as standard error of mean.
The metallic/metallic tandem catalyst was also studied. Shown in Figure 5.7a, Ag was
employed as local CO-maker for the CuOx NPs. Figure 5.7b plots the absolute production
rates on various CuOx to Ag component at electrode potentials of -1.0 VRHE. The production
rate of C2H4 was facilitated with all CuOx to Ag ratios compared to pure CuOx, especially at
the ratio of 1 to 3 (CuOx:Ag). The CH4 production rate exhibited a same trend as C2H4, again,
i dica i g ha he “ e f-f med” CO ayed a d mi a i g e i he ca a y ic eac i . To
test the stability of the tandem CuOx-Ag catalyst in the H-cell configuration, 20-hour
electrolysis was carried out using the CuOx: Ag (1:3) catalyst at the chosen electrode potential
(-1.0 VRHE), where the respective maximum C2H4 was observed (Figure A2.14). The tandem
catalysts reported in the literatures28, 121, 122 are also in good accordance with above-mentioned
co-feed system, again, suggesting the ubiquity of our co-feed mechanism.
5.6 Conclusions
The electroreduction of mixed CO2/CO feeds on CuOx NPs and pure CO2 feeds on tandem
Ni/N doped carbon-CuOx catalysts displayed significantly improved catalytic C2H4 yields
over a wide electrode potential range. Using operando DEMS, time-resolved isotope labelling
experiments, carried out in a customized CO2 capillary cell, the possible mechanistic
pathways were quantitatively deconvoluted. Our analysis demonstrated that the enhanced
75
Chapter 5 Co-feed Study
C2H4 production largely originated from a crossed CO2-CO combination pathway and that
over two-thirds of the generated C2H4 involved co-fed CO. The co-fed CO gas does not
compete with CO2 for adsorption sites, which implies the existence of separate,
non-scrambling reactant-specific surface adsorption sites for CO2 and CO. Finally, we
demonstrated that the co-feeding mechanism is the mechanistic basis of internally co-fed
tandem catalyst concepts. In the end, we confirmed our mechanistic conclusions about the
equivalence of external and internal CO co-feeding highlighting the non-metallic/metallic
CuOx-NiNC and metallic/metallic CuOx-Ag tandem catalysts, which involve a (non-)metallic
CO producer coupled to ethylene-producing CuOx. Our mechanistic insights offer a wide
range of practical catalysts design strategies to CO2RR electrocatalyst with improved
ethylene yield, but, in a more general sense, we think our mechanistic concepts are
transferable to other interesting products such as ethanol or propanol, as well.
76
Chapter 5 Co-feed Study
77
Chapter 6 Shaped CuOx Catalysts
Chapter 6 Sheet-like Copper Oxides with Stable and Selective
Ethylene Production for Direct CO2 Electroreduction
Oxidic and oxide-derived nanostructured copper catalysts have frequently been used to
promote the electrochemical reduction of CO2 (CO2RR) to higher-value products such as
ethylene. Here, we present a successful synthesis of free-standing sheet-like copper oxides
nanoparticles, exposing predominantly {001} facets, and investigate CO2RR performance in
an H-cell configuration. The highly active CuOx nanoparticles exhibit ordinary stability
towards ethylene over at least 24 hours. A combination of operando X-ray Absorption
Spectroscopy (XAS) and operando Wide-Angle X-ray Scattering (WAXS) are used to reveal
the changes in structure the catalyst that undergoes at different reaction stages. The onset
potentials of the main products are also determined by operando Differential Electrochemical
Mass Spectrometry (DEMS). The complementary analysis depicts a highly active system, in
which a complete reduction of oxidized Cu progresses from the surface towards bulk layers.
The defect-rich surface introduced by the self-reducing process is proposed to be responsible
for the catalytically stable performance for C2H4 formation.
78
Chapter 6 Shaped CuOx Catalysts
Chapter 6 and its supplementary information section (Appendix A3) were reproduced from a
manuscript under preparation for submission.
Xingli Wang, Katharina Klingan, Jorge Ferreira de Araújo, Henrike Schmies, Isaac Martens,
Fabio Dionigi, Shan Jiang, Holger Dau, and Peter Strasser, “Sheet-like Copper Oxides with
Stable and Selective Ethylene Production for Direct CO2 Electroreduction”
X.W., P.S. conceived and designed the experiments. X.W. performed the experiments and analyzed the
data, K.K. helped the experiment of operando XAS, data analysis and wrote the XAS report; F.J.
helped the DEMS measurements; X.W., H.S., I.M., F.D. contributed to WAXS measurements and data
analysis. S.J. did in situ Raman spectroscopy, X.W. and P.S. wrote the manuscript; all authors
contributed to the discussions.
79
Chapter 6 Shaped CuOx Catalysts
6.1 Synthesis and characterization
The sheet-like CuOX (CuOx NS) catalyst was prepared by solvothermal synthesis through
Cu(OH)2 intermediate in alkaline condition (see Figure A3.1). The wrinkled large-scale thin
Cu(OH)2 layer is then decomposed to CuO and H2O. Figure 6.1a-b presents scanning electron
microscopy (SEM) image and transmission electron microscopy (TEM) image of as-prepared
copper oxides nanoparticles. As shown in the images, the 2-dimension (2D) structure
remained after the thermal decomposition of Cu(OH)2, while the large sheet layer split into
small pieces and the narrow ones merged together. The final products exhibit rectangular
nanosheet-like morphology with serrated edges on the sides of shorter length. Figure A3.2a
shows cross-section image of CuOx NS, indicating a stacked structure in a single free standing
CuOx NS. The thickness of each nanosheet is less than 30 nm. The crystallinity and surface
roughness are highlighted by the high-resolution transmission electron microscopy (HR-TEM)
image and selected area electron diffraction (SAED) image. The HR-TEM image (Figure 6.1c)
of the CuOx NS shows well-defined lattice fringes with an interplanar spacing of 2.8Å,
corresponding to {110} planes of monoclinic CuO. Moreover, from SAED pattern (Figure
6.1d), the rhombus diffraction spots along [001] can be seen with indexed (020), (-110),
(-200), (-1-10), (0-20) and (1-10) planes, indicating that the CuOx NS has {001} exposed
surface. The relative intensity of the diffraction spots is presented in Figure A3.2b. The
crystal structure is also provided in Figure A3.3 with indexed (001), (110) and (11-1) planes.
80
Chapter 6 Shaped CuOx Catalysts
Figure 6.1 Morphological, structural characterizations of sheet-like CuOx nanoparticles
synthesized in this study. a) Large-scale scanning electron microscopy (SEM) image and 3D
structure of CuOx NS (insert, orange). b) Transmission electron microscopy (TEM) image and
c) high resolution TEM images with measured lattice distance and the corresponding fast
Fourier transformation. d) Selected area electron diffraction (SAED) pattern along zone axis
[001]. e) 2D Synchrotron grazing incidence wide-angle X-ray scattering (GI-WAXS) image
demonstrating preferred orientation of as-prepared CuOx NS on glassy carbon. Additional
azimuthally integrated line profiles are shown in Figure A4.4b.
A 2D synchrotron grazing incidence wide-angle X-ray scattering (GI-WAXS) image of
drop-casting CuOx NS on glassy carbon is shown in Figure 6.1e. The diffraction patterns are
mainly contributed by the CuO phase, while a small reflection due to an additional Cu2O
phase is also observed with low angle. The higher order Cu (I) peaks appear absent. It can be
seen that the (002) and (11-1) reflections of Cu (II) exhibit a stronger intensity in the
meridional direction, demonstrating that the CuOx nanosheets are stacked along the <00l>
direction and that this one is perpendicular to the incident beam. A partial contour plot is
shown in Figure A4.4a, indicating the coexistence of both the Cu2O phase and the CuO phase.
The distribution of scattering intensity was obtained by integration of the 2D detector
GI-WAXS image numerically. Azimuthal integrations of CuOx NS line profiles are included
in Figure A4.4b.
81
Chapter 6 Shaped CuOx Catalysts
6.2 Electrochemical CO2RR over sheet-like CuOx catalysts
The catalytic CO2RR activity of as-prepared CuOx NS was tested in a two-compartment
membrane-separated H-cell configuration within the potential range of -0.75V to -1.0V versus
RHE in CO2 saturated 0.1 M KHCO3. We deposited the catalysts onto glassy carbon electrode
via drop coasting of a catalyst ink to keep the catalyst loading of 100 g cm-2 constant for
each measurement. The detailed absolute production rates for all products after 1h are listed
in Figure 6.2a. C2H4 exhibited earlier formation compared to CH4. Thus, a region with only
C2H4 as the unique hydrocarbon can be observed at relatively low overpotentials. In contrast,
CH4 formation experienced a rapid growth after -0.9 VRHE and exceeded C2H4 at -0.97 VRHE.
The corresponding partial densities are shown in Figure 6.2b, demonstrating an outstanding
selectivity toward hydrocarbon. The geometric current density of C2H4 reached up to 6.2 mA
cm-2 at -0.97 VRHE, and for CO remained at low level during the entire given overpotential
range. The production rate-derived Faradic efficacies (FEs) of all detectable products over
selected electrode potentials are shown in Figure A3.5. The total FEs for each measurement
reaches 90%, with clear suppression of HER and enhancement of C2H4 with higher
overpotentials. Since the catalytical electroreduction process on copper oxides always
involves a self-reducing period with a mixed matrix of metallic and oxidized copper,
long-term electrolysis up to ~20 h has been carried out at -0.76 VRHE, -0.84 VRHE and -1.0
VRHE, respectively. Figure 6.2c-f shows the production rates (and FEs) as a function of
reaction time. The production rates towards the main carbon-based products (CO, CH4, and
C2H4) displayed a similar activation process in the initial stage of CO2 electrolysis, while with
higher overpotentials, a larger slope was observed (Figure 6.2c-e), indicating a faster
transition time. After 120 min, the production rates reached to the peak value. At all selected
overpotentials, C2H4 formation remains constant after peak values, while CH4 only formed at
-1.0 VRHE and declined after the highest point. It is noteworthy that C2H4 is the only
hydrocarbon formed at -0.84 VRHE and C2H4 production rate is even higher (1.6 nmol cm-2 s-1)
than conventional electropolished Cu foil at -1.0 VRHE.123 The testing time was then prolonged
to 1200 min. As we can see in Figure 6.2f, the production rate and FE towards C2H4 displayed
minor changes over the reaction time. The morphology change of the as-prepared CuOx NS
82
Chapter 6 Shaped CuOx Catalysts
was determined by SEM after 1200-min electrolysis, where dramatically change was found
displayed as particle agglomeration (see Figure A3.6). The same phenomena were observed
after long-term reaction at -0.76 VRHE (Figure A3.7).
Figure 6.2 Electrochemical Performance of CO2RR with as-prepared CuOx NS in H-cell
configuration. a) Absolute product formation rates of major gaseous products as a function of
applied electrode potentials during CO2RR in CO2-saturated 0.1M KHCO3 at 60 min. b)
Partial current density of major gaseous products. Time-dependent absolute product formation
rates on c) CO, d) CH4 and e) C2H4 with various overpotentials. f) Chronoamperometric
performance stability of the CO2 reduction reaction on CuOX NS in CO2-saturated 0.1M
KHCO3 at -0.84 VRHE.
83
Chapter 6 Shaped CuOx Catalysts
6.3 Discussions
The above-mentioned observations of activity/selectivity evolution at fixed overpotential
suggest that the catalytic performance is mainly dependent on either structural-morphological
evolution and/or crystalline phase transition. Taking the long-stability test at -0.84 VRHE as an
example, which showed a considerable C2H4 formation but no CH4, the products formation
was unaffected after reaching their peak point. Now, the sintering and coalescence of CuOx
nanosheets could be observed as early as 1-h after start of the electrolysis experiment (see
Figure A3.8). In addition, the highest C2H4 selectivity was achieved after the original shape
was lost. This implies that the catalyst morphology and its evolution do not appear to be the
main factor or determinant for the experimental high catalytic performance. Considering
relatively large negative overpotentials used during the measurements, copper oxides are
likely quantitatively reduced to metallic Cu. To determine the prevalent local structure motifs,
operando XAS was employed to examine the local atomic structure evolution and in situ
GI-WAXS for crystalline phase change of as-prepared CuOx electrocatalyst under catalytic
operating condition.
84
Chapter 6 Shaped CuOx Catalysts
Figure 6.3 Catalyst evolution during CO2RR characterized by operando XAS. a) XANES at
the Cu K-edge of the CuOx NS film at OCP in 0.1 M KHCO3 at pH 6.8 and during 130 min
CO2RR at -0.84 VRHE. b) Linear combinations (with the weighting factors as fit parameters)
of Cu foil and the CuOx NS film at OCP were fitted to experimental data of the CuOx NS film
during 130 min CO2RR at -0.84 VRHE. Colored lines represent the experimental data and
black lines the linear combinations. c) Amount of Cu metal in the CuOx nanoparticle film
during 130 min CO2RR at -0.84 VRHE. Values are obtained by linear combinations (with the
weighting factors as fit parameters) of Cu foil and the CuOx film at OCP, which were fitted to
experimental data. d) FT of k3 weighted EXAFS at the Cu K-edge of the CuOx film at OCP
and during 130 min CO2RR at -0.84 VRHE in 0.1 M KHCO3 at pH 6.8. Colored lines represent
the experimental data and black lines the simulations. The distance on the x-axis is reduced
by 0.35 Å relative to the real distance. e) EXAFS (k3 weighted) at the Cu K-edge of the CuOx
film at OCP and during 130 min CO2RR at -0.84 VRHE in 0.1 M KHCO3 at pH 6.8. Colored
lines represent the experimental data and black lines the simulations. Coordination numbers
of the first Cu-O (f) coordination sphere and the first intermetallic Cu-Cu (g) shell. Catalyst
loading: 100 g cm-2.
The CuOx NS electrode was prepared and characterized by operando XAS in fluorescence
geometry. X-ray absorption near edge spectra (XANES) shown in Figure A3.9 compares the
oxidation state of CuOx NS film with reference samples, which is identical in its dry state and
at OCP in 0.1 M KHCO3 at pH 6.8 with single CuO being the main present component. The
small portion of Cu2O which was detected by WAXS (Figure 6.1e) is attributed to the grazing
85
Chapter 6 Shaped CuOx Catalysts
incidence mode, which offers a much larger surface sampled by the X-ray beam. The catalyst
structure at resting condition is a highly ordered CuO structure. The nanoparticle film is even
more ordered than the reference powder (see higher Debye Waller parameters for the CuO
powder reference in the simulation table). However, CuO reference powder and CuO
nanoparticle film have the same atomic structure (see Figure A3.10 and simulation results in
Table A3.1). Contributions from Cu(OH)2 and Cu2O in the relaxed state of the CuO NP film
can be excluded.
Applying a constant electrode potential of -0.84 VRHE, complete reduction to metallic copper
is observed after 130 min. The transformation from CuO to Cu was followed by the intensity
decrease of the pre-edge feature at 8996 eV. The reduction from cupric oxide, CuO, to
metallic Cu occurs directly without any detectable intermediate formation of Cu2O (Figure
6.3a). Linear combinations of Cu foil and the CuO NS film at OCP were calculated and fitted
with the weighting factors as fit parameters to the experimental spectra. Directly after the
LSV scan, 35% of the bulk volume has been reduced to Cu metal (see Figure 6.3b-c). At the
end of the experiment, all Cu species are present as Cu metal. (The linear combination gives
99%, the remaining 1% is within the uncertainty of the data quality). The gradual change
from CuO to Cu metal is also indicated by the intensity decrease of the Cu-O bond (1.94 Å)
and the intensity increase of the Cu-Cu (2.53 Å) bond (Figure 6.3d-e). Coordination numbers
(CN) of Cu-O decreased with the reduction process, while CN of Cu-Cu increased, and
stabilized at 10 after 120 min. There is no change in the Cu-O distance (first coordination
sphere) and no change in the Cu-Cu bond length. The obtained CN of self-reduced Cu is
smaller than that expected for the face centered cubic (FCC) structure of metallic copper,
indicating a defective structure formed during catalytic reduction. The resulting
undercoordinated sites with strong binding ability continuously contribute to the outstanding
catalytic performance, which is also proposed and discussed by previous study.94, 124, 125
86
Chapter 6 Shaped CuOx Catalysts
Figure 6.4 Catalyst shape evolution detected by in situ WAXS. a, b) in situ WAXS patterns
of CuOx NS electrode with catalyst loading of 20 ug cm-2 during 130 min CO2RR at -0.84
VRHE.
The reduction of CuOx NS to metallic copper during operation at catalytic potential was
further confirmed by in situ WAXS measured in grazing incidence mode (see Figure 6.4a).
The dry powder showed clear reflections of a single Cu(II) oxide phase with the whole pixel
integration on the 2D detector. The majority of this CuO phase was reduced within the first 30
min, while a small portion of a residual oxide phase lasted for almost 120 min. The vanishing
of distinct CuO reflections and growth of metallic copper reflections are shown in Figure
6.4b.
The evolution of the dominant catalytic reaction products during the change in the catalyst
structure owing to the reduction of CuOx NS are shown in Figure 6.5. The time resolved
transient ion currents of each product was recorded over applied electrode potential and time
using a custom-designed millisecond-resolved differential electrochemical mass spectrometry
(DEMS) flow cell. During repeated cyclic voltammetric scans, the product ion current -
potential curves showed increasing ion currents in cathodic scan directions, while the ion
currents decreased in the anodic scan direction (as shown in Figure 6.5).
87
Chapter 6 Shaped CuOx Catalysts
Figure 6.5 Products formation and onset potential shift along the electroreduction of CuOx
NS catalyst. a) operando Differential electrochemical mass spectrometry (DEMS) sweep data
obtained during CO2RR on CuOx NS catalyst (supported on a flat 0.785 cm2 glassy carbon
electrode) using CO2-saturated 0.1M KHCO3 by continuous cyclic voltammetric scan at 5mV
s-1. b) DEMS-derived mass charges for various products formed during the cathodic and
anodic voltammetric sweep. c) Spider plot shows the variations in the onset electrode
potential of key products during CO2RR by continuous cyclic voltammetric scan at 5mV s-1.
Product molecules considered are: m/z =28 CO, m/z = 15 corresponding to methane, m/z =26
corresponding to ethylene, m/z =31 corresponding to ethanol.
Figure 6.5a displays and compares the characteristic ion currents of CH4, C2H4 and ethanol
(EtOH) over the voltammetric cycle number. C2H4 and ethanol (EtOH) exhibited very similar
trend, where the ion currents peaked after the 10th cycle, while CH4 increased monotonically.
The mass ion charges of CO, CH4, C2H4 and EtOH (ethanol) were considered and
deconvoluted at m/z values of 28, 15, 26, 31, respectively. These m/z values yield quantitative
measures of the product formation rates (Yields) over each cycle (see Figure 6.5b). All
product yield showed growing trends as the metallic character of the Cu catalysts increased in
the course of the reduction process; this is in contrast to CO, the yield of which appeared to
declined after 5th cycle. The onset potentials of the key catalytic reduction products and their
changes with cycle number and with catalyst structure are included in Figure 6.5c,
88
Chapter 6 Shaped CuOx Catalysts
demonstrating a significant influence on C2H4 and CO. After 40 min, the onset potentials of
C2H4 and CO moved to more and more negative values, where a large portion of the initial
oxidized copper was reduced. Thus, in the beginning of CO2RR (within 10 cycles), the
catalyst with a large portion of cupric oxide phase seems inactivate towards CO2RR, leading
to relatively lower ion currents toward the major products. This may also indicate the
co-presence of a carbonate passivation layer, which suppresses the products evolution. While
it shows a positive effect on onset potentials towards CO and ethylene, while not affecting the
CH4 and EtOH onset potential.
6.4 Conclusions
We have developed a facile synthesis route towards sheet-like CuOx nanoparticles with
predominantly {001} facet. Catalytic tests in an H-Cell showed highly active CO2RR
performance and excellent stability. As illuminated by ex-situ SEM, the loss of initial
morphology did not cause catalyst deactivation. Using the combination of operando XAS and
in situ WAXS, we were able to trace the reconstruction process that CuOx NS underwent
during CO2RR, indicating a conversion from the initial copper oxide phase to catalytically
active metallic copper phase. The sustained presence of ionic Cu species below our detection
limit cannot be dismissed, nor confirmed as the origin of sustained catalytic activity.
Meanwhile, under-coordinated surface sites were generated by self-reducing catalysts, which
was held responsible for the outstanding experimental performance. Our operando DEMS
results suggested that the initial oxidized copper phase and/or a carbonate passivation layer
shifted the onset potentials on CO and C2H4 more anodically, but left those of CH4 and EtOH
unaffected. Together the materials and kinetic catalytic study presented here on previously
unexplored sheet-like CuOx nanoparticles have established valuable
structure-activity-selectivity correlations for future catalyst designs.
89
Chapter 7 Flow Cell Study
Chapter 7 Achieving high C2H4 Evolution at industrial current
densities on CuOx nanosheet derived gas diffusion electrode
The electrochemical conversion of CO2 (CO2RR) to valuable carbon-based products at an
industrial equivalent scale is posed to assess the potential economic and technical feasibility
of the technology. Delivered by the work presented in Chapter 6, the sheet-like CuOx
nanoparticles exhibit the promise to selectively convert CO2 into ethylene during CO2RR. We
thus employ such CuOx nanosheets (CuOx NS) as a catalyst candidate for a type of Gas
Diffusion Electrode (GDE) combined Micro-Flow-Cell (MFC) measurement. The catalytic
performance of CuOx NS exhibits an outstanding ethylene production under industry
equivalent neutral condition, demonstrating C2H4 partial current density of 229 mA cm-2 and
stable faradic C2H4 efficiencies around 30 % for up to 24 hours (at 300 mA cm-2). The partial
current towards C2+ products reaches ~410 mA cm-.2 The morphological and structural
evolution of the reactive CuOx NS catalysts is also investigated in combination with an
electrochemical liquid TEM micro reactor. Our quasi in situ TEM monitoring tracked a
remarkable amount of Cu particle evolution, which is in good agreement with the
ex situ-TEM images of catalysts measured under GDE-flow-cell conditions. The present study
corroborates the power of in situ electrochemical liquid TEM studies for the understanding of
activity-selectivity-morphology relations under catalytic CO2RR operating conditions at the
industry level.
90
Chapter 7 Flow Cell Study
Chapter 7 and its supplementary information section (Appendix A4) were reproduced from a
manuscript under preparation for submission.
Xi g i Wa g, Tim Mö e , Ma e K i ge h f, S efa ie Küh a d Pe e S a e , “Achie i g
high C2H4 Evolution at industrial current densities on CuOx nanosheet derived gas diffusion
e ec de”
P.S., X.W. conceived and designed the experiments. X.W. M. K. and S. K. performed the microscopy
experiments, X.W analyzed the data, T.M. helped with the flow cell measurement and data analysis;
X.W. and P.S. wrote the manuscript; all authors contributed to the discussions
91
Chapter 7 Flow Cell Study
7.1 CO2RR electrolysis using GDE combined MFC
To assess the technological potential of CuOx nanosheets catalysts (synthesis details in
Chapter 3.12) for industrial CO2 co-electrolysis, we employed a single three-electrode
electrolyzer with multi-chambers cell set-ups for pressurizing the gas flow and circulated
electrolyte flows in the cathodic chamber (see Figure 7.1a). In the electrolyzer, a catalyst
suspension containing a known amount of catalyst, is usually spray-coated onto a carbon
paper substrate as a working electrode. The hydrophobic microporous layer on carbon paper
helps to form a uniform gas-liquid interface between the substrate and catalyst during the
reaction. The microporous layer also helps to lower the ohmic contact resistances between
porous layer and catalyst layer. The solubility of CO2 in liquid aqueous electrolytes limits its
transport rates to the reactive interface. This is why current density in CO2-saturated liquid
electrolytess49 remains low. In contrast, a Gas Diffusion Electrode (GDE) set-up with a
three-phase interface of electrolyte, catalyst and pressurized CO2 inside the GDE offers a
solution. With a catalyst loading of 1 mg cm-2 on an active geometric surface area of 3 cm2
GDE, constant current densities between 50 and 700 mA cm-2 were applied galvanostatically
to the cell in 1 M KHCO3. Each current is held for 2 h before moving on to the next current
step.
92
Chapter 7 Flow Cell Study
Figure 7.1 Electrochemical characterization in micro flow-cell configuration. a) Schematic
representation of the Flow-Cell. b) FEs as a function of applied geometric current density for
CuOx NS with a catalyst loading of 1 mg/cm2 in 1 M KHCO3. c) Partial current density of
C2H4 vs the total applied current density. d) Plot of ethylene partial current density in a
flow-cell system (compared with references).
Figure 7.1b shows experimentally measured FEs of major CO2RR products on CuOx
electrocatalyst as a function of the applied current densities. It shows a high selectivity
towards CO2RR exceeding 58 % and HER selectivity is only 13% at 700 mA cm-2. The FE
towards C2H4 exhibits an uphill trend with increase of applied current densities. Maintaining a
constant FE of C2H4 as 33% at high current densities, the partial current of C2H4 is reported as
229 mA cm-2 (at 700 mA cm-2) in mild condition (see figure 7.1c). The geometric partial
current density towards C2+ products also reaches ~ 410 mA cm-2 (at 700 mA cm-2) (see figure
A4.1) CH4 formation is suppressed at all current densities (0.58% at 700 mA cm-2), leading to
unique C2H4 formation as single hydrocarbon product. Figure 7.1d compares the C2H4 partial
current density in a flow-cell system reported in the literatures. Our catalyst is among the best
93
Chapter 7 Flow Cell Study
catalysts for CO2RR in neutral condition with flow cell configuration. Note that we used
nanoparticles for the GDE, which also brings the flexibility for the GDE design with tunable
catalyst loading. While most studies focused on the electro-deposition system.
Figure 7.2 SEM image of CuOx NS GDE after CO2 electrolysis.
The electroreduction of CO2 on a catalyst is almost always accompanied by a morphology
evolution of the catalytic active material. SEM was employed to determine the morphology
change of the GDE after the catalytic reaction. As shown in Figure 7.2, particle agglomeration
and dendrite growth can be observed. Even though the CuOx NS strongly deshaped during the
CO2 electrolysis process, the rough surfaces remained after the current-screening, which was
considered to contribute continuously to the high activity and outstanding selectivity.
7.2 Discussions
The CuOx NS catalyst reached a stable faradaic efficiency of ethylene, FEC2H4 , of nearly 30%
when the applied current density reached 300 mA cm-2. The catalyst and its GDE were able to
maintain this FE efficiency even at higher current densities. A stability test was subsequently
performed at 300 mA cm-2 for 24 hours. The FEC2H4 reached 31% already after 4-hour test.
Only a negligible decrease in C2H4 efficiency was observable after as long as 20 hours.
Particle agglomeration and dendrite growth were observed by SEM after the long-term
electrolysis. A control experiment was taken in H-cell configuration at similar overpotential
(-1 VRHE) for 24 h (see Figure A4.2), while no evident dendrites were observed in the SEM
images, the observed aggregation was similar (see Figure A4.3).
94
Chapter 7 Flow Cell Study
Figure 7.3 a) Gaseous products efficiencies with 300 mA cm-2 in CO2 electrolyzer flow cell.
All tests were performed with CO2 saturated 1 M KHCO3 and 1 mg cm-2 catalyst loading. b)
SEM images after flow cell test.
The morphological restructuring of copper oxides catalysts during catalytic operation has
been suggested to play an important role during the reaction. However, direct imaging of the
morphological and structural evolution of reactive copper oxide catalysts has remained
elusive. In situ microscopy technique offers us an exciting opportunity to trace the
morphological evolution during the reaction. Figure 7.4a shows the interaction between the
electron beam and CuOx electrocatalyst in a customized microscopic electrochemical liquid
cell setup. The cross-section scheme of the liquid cell holder is shown with the electrodes is
given on the top chip. The quasi in situ TEM experiments were realized in a Protochips
Poseidon holder, in which a three-component flow cell chip was inserted and equipped with a
silicon nitride electron transparent window. Several control experiments were carried out first,
because the typical K+ -containing KHCO3 electrolyte is known to be unsuitable for the
silicon nitride window, risking of the membrane break (according to the user manual). Then
CO2-saturated 0.1 M KHCO3 was changed to 0.1M Na2HPO4/NaH2PO4 (pH=6.9) to maintain
a similar pH value. We performed 15-min CO2 electrolysis in the H-cell with 0.1 M
Na2HPO4/NaH2PO4. As we can see in Figure A4.4, the morphological evolution of the CuOx
NS catalyst does not depend on the interaction of CO2 with the catalyst surface. The shape
loss of CuOx NS started already after 15 min seen by sheet cracking. Thus, 0.1 M
Na2HPO4/NaH2PO4, without CO2 feed is used for the following quasi in situ TEM
experiments. Another reason to leave the CO2 feed in the first experiment is that the gas
bubbles in electrochemical liquid TEM micro reactor always cause short-circuit.
95
Chapter 7 Flow Cell Study
Figure 7.4 in situ Observation of dendrite growth after electrochemical reduction of CuOx NS
catalysts inside an electrochemical liquid cell TEM holder. a) Cross-section of the chip with
the corresponding bottom chip and the electrolyte layer between the chips. b) Dimension of
the electrochemistry chip (upper left), photography of top E chip (upper right) and ex situ test
cell. c) Chronoamperometric performance in 0.1 M Na2HPO4/NaH2PO4 (pH=6.9) for 25 min.
STEM image of CuOx NS on electrochemistry chip d) before and e) after reduction. f) TEM
image of CuOx NS on electrochemistry chip after reduction.
Figure 7.4d represents the STEM image of CuOx NS attached on the carbon working
electrode of the top window of the electrochemical liquid TEM cell. After TEM imaging
(Figure 7.4d), the whole chips, containing the top electrochemical chip, bottom chip and as
well as a gasket, were transferred into the ex situ cell with a similar operating system for
electrochemistry test. Chronoamperometric performance was then measured for 25 min in an
ex situ electrochemical liquid TEM cell. A very similar CA curve was thereby obtained
(Figure 7.4c) as that obtained in the H-cell. After rinsing, the ex situ cell was carefully
unmounted and the same chip (“E-chi ”) configuration was mounted into the electrochemical
liquid TEM cell again for imaging. Figure 7.4 e,f show the direct images after
electrochemistry of the CuOx NS catalyst. Again, the sheet-like catalyst deshaped and the
particles agglomerated along the edge of the glassy carbon electrode. Moreover, the formation
of clear dendrite was observed on the top E-chip. The quasi in situ data suggest particle
evolution in liquid TEM cell agreed better with practical flow cell condition. We attribute this
to the harsh electroreduction condition in both flow cell electrolyzer and liquid TEM micro
reactor, where the current densities are quite high, comparing to the relative catalyst loading.
96
Chapter 7 Flow Cell Study
7.3 Conclusions
We have investigated the electrocatalytic reduction of CO2 to selective C2H4 streams on CuOx
NS catalysts, with exposing predominantly {001} facets, in a Micro-Flow-Cell (MFC)
configuration equipped with a Gas Diffusion Electrode (GDE). The catalyst poses a faradaic
efficiency exceeding 33% ethylene at high current (700 mA cm-2) with suppressed HER. The
partial current of C2H4 and C2+ reaches 229 and ~410 mA cm-2, respectively. We then
demonstrated the morphological transformations in an electrochemical liquid TEM cell set up.
Dendrites growth was observed in both flow cell and liquid TEM cell due to the similar harsh
electroreduction condition, which differs from the H-cell. Our quasi in situ experiment
verified the feasibility of in situ microscopy technology for future investigation on
catalytically morphology evolution under CO2RR operating conditions at industry relevant
current levels.
97
Chapter 8 Summary and Outlook
Chapter 8 Summary and Outlook
The present Dissertation set out to establish a new in-depth understanding of CO2
electroreduction reactivity on known and new catalysts with high C2+ product selectivity. By
adjusting the local environment of our studied CuOx nanoparticle model catalysts, including
tuning the particle assembles and introducing co-reactant-gas-feeding system, we were able to
uncover the Critical role of local interfacial CO concentrations and, using advanced
electrochemical mass spectrometry, succeeded in deducing previously unavailable details of
the reaction mechanism and the relative contribution of reactive pathways. Beyond this
fundamental investigation, novel tandem catalysts, consisting of the C2 product selective
CuOx nanoparticles in combination with a selective CO producer (such as Ag or a novel
non-metallic NiNC carbon catalyst), have delivered significant enhancements in ethylene
production and yield in comparison to CuOx and NiNC control catalysts. More importantly, a
new type of sheet-shaped CuOx nanoparticles was designed, developed and systematically
investigated based on the knowledge found in this work, and its impressive performance for
ethylene formation was further demonstrated in both liquid H-cell configuration and
industry-relevant micro flow cell. Moreover, a combination of operando/(quasi) in situ
techniques were used to discovery the morphology-structure-phase-activity-selectivity
relationship.
8.1 Manipulating *CO behavior by local condition
Recalling the importance of *CO formed after 2e- transfer process during CO2 feed, which
can be followed by a protonation pathway to CH4 or dimerization pathway to C2H4 (as
illustrated in Figure 8.1).
98
Chapter 8 Summary and Outlook
Figure 8.1 Graphical summary of topics and issues discussed and investigated in the first part
of work, including areal density control, mechanistic co-feed study, and tandem catalyst.
(Reprinted with permission from The John Wiley and Sons and Springer Nature)126,127
We started our mechanistic study on particle assembles, aiming to control re-adsorption
behavior of *CO (see Chapter 4). In this contribution, tunable product distribution of ethylene
and methane during CO2 electroreduction is observed by simply adjusting the areal particle
density. As CO2RR is a multiple proton and electron reaction, the desorption, diffusional
transport and possible re-adsorption of chemical reaction intermediates or products on closely
spaced particles may result in variations in product selectivity. By increasing the areal NP
density, the mean interparticle distances decrease and the real surface-area increases as
deduced from geometric voltammetric current scans. A shift in Faradic efficiency towards
C2H4 over CH4 in higher areal density is observed. This observation is attributed to higher
dimerization rates of adsorbed *CO at smaller interparticle distances and more surface active
sites, which enhanced CO(g) re-adsorption on NP in close proximity. The following catalytic
performance tests over extended reaction times corroborate the role of interparticle distances.
To further increase the local concentration of CO (g), a CO2 with CO co-feeding system was
he i e iga ed ( ee Cha e 5). I hi a , he e ec chemica educ i f “modified gas
feed” (CO2 with CO co-feeding) is systematically investigated on uniform and catalytically
active Cu oxide NPs as a promising method to improve C2H4 formation over a wide electrode
99
Chapter 8 Summary and Outlook
potential range. Using a custom-designed DEMS flow cell, carbon origins of C2H4 in the
co-feed system were successfully identified by isotope labelling of one of the carbon species
from the co-feed (12C and 13C). Our analysis demonstrated for the first time that the enhanced
C2H4 production mainly originated from a kinetically favored CO2-CO crossed combination
pathway. Our study highlights the role of extra CO on the ethylene product, indicating
additional channels for ethylene production opened up without affecting the existing
pathways. The co-feed mechanism inspired a new fabrication of catalysts with a CO-maker
catalyst and copper-based material, yielding high selectivity to hydrocarbons with a facile
preparation method.
8.2 Shaped CuOx nanoparticles for CO2RR
Our focus moved to new catalyst development in the second part of this work. A new type of
sheet-like copper oxide nanoparticles with preferred {001} orientation was successfully
synthesized and first applied for CO2RR in an H-cell configuration (see Chapter 6). The
highly active CuOx nanoparticles exhibit outstanding stability and selectivity towards ethylene.
A combination of (in situ/operando) WAXS and XAS were used to examine the phase/local
atomic structure evolution of as-prepared CuOx electrocatalyst under operating CO2RR
electrolysis condition. Millisecond-resolved operando DEMS was employed to determine the
onset potential shift of the products with gradually self-electroreduced CuOx electrocatalyst to
a purely metallic Cu phase. Ex situ SEM measurements showed that the initial sheet-like
morphology de-shaped during the reduction reaction and progressing particle aggregation in
the H-cell, while unaffecting the catalytic performance. These results suggest that the
improved catalytic performance for oxide-derived copper catalyst arise from a combination of
the increased content of metallic copper induced under-coordinated sites, and the stable, high
local pH created through nanostructured morphology.
100
Chapter 8 Summary and Outlook
Figure 8.2 Graphical summary of topics and issues discussed and investigated in the second
part of this work, including shaped CuOx nanoparticles synthesis, characterization, H-cell
tests, a combination of operando studies, and flow cell test at industrial relevant level.
To reach the industrial level CO2 electrolysis, performance exhibited in common H-type cell
is far from sufficient, mostly due to mass transfer limitations of the CO2 reactant. Thus, micro
flow cell experiments were done in a following-up work (see Chapter 7). The catalyst
exhibited a stable faradaic efficiency towards ethylene exceeding 33 %. The partial current of
C2H4 and C2+ reached 229 and 356 mA cm-2, respectively, obtaining as an excellent catalyst
for ethylene under neutral condition. Our study on morphological transformations of
free-standing CuOx nanosheets was done in electrochemical liquid TEM micro reactor.
Dendrite growth caused by harsh electroreduction conditions was observed in both flow cell
and liquid TEM cell experiments, which differs from observations in the H-cell. Our data
suggest the particle evolution in liquid TEM cell agreed better with the micro flow cell due to
similar practical flow conditions. The present study corroborates the power of in situ
electrochemical liquid TEM studies for the understanding of morphology-activity-selectivity
relations under catalytic CO2RR operating conditions at the industry level.
101
Chapter 8 Summary and Outlook
8.3 Outlook
An unprecedented contribution towards understanding the mechanism in pure CO2 reduction
reactions was provided in this work, offering new guidelines for obtaining active and stable
electrocatalysts. However, it still remains a challenge to selectively form a single product with
extended catalyst stability on Cu-based electrocatalysts. Currently, most of the catalysts are
screened in H-cell devices. The mass transfer limitation shown in the H-cell is often
considered as a barrier for continuously large scale operation of CO2 electrolysis. To solve
this problem, the translation to flow cell has proven to be a highly effective method for the
development of industrial scale electrochemical reactors. The continuously circulated gas and
liquid flow in the flow cell reactor, moving reactants and products to and away from the
electrodes, helps to break the mass transport limitations. Therefore, further investigations in
flow cell reactors with Cu-based catalysts, together with new catalyst development, provide
an opportunity for the practical application from lab scale to industrial scale.51
102
Chapter 8 Summary and Outlook
103
Reference
Reference
1. Li F, MacFarlane DR, Zhang J. Recent advances in the nanoengineering of
electrocatalysts for CO2 reduction. Nanoscale 2018, 10(14): 6235-6260.
2. Shi J, Jiang Y, Jiang Z, Wang X, Wang X, Zhang S, et al. Enzymatic conversion of
carbon dioxide. Chemical Society Reviews 2015, 44(17): 5981-6000.
3. Qiao J, Liu Y, Hong F, Zhang J. A review of catalysts for the electroreduction of
carbon dioxide to produce low-carbon fuels. Chemical Society Reviews 2014, 43(2):
631-675.
4. Costentin C, Robert M, Saveant J-M. Catalysis of the electrochemical reduction of
carbon dioxide. Chemical Society Reviews 2013, 42(6): 2423-2436.
5. Spinner NS, Vega JA, Mustain WE. Recent progress in the electrochemical
conversion and utilization of CO2. Catalysis Science & Technology 2012, 2(1):
19-28.
6. Jitaru M, Lowy DA, Toma M, Toma BC, Oniciu L. Electrochemical reduction of
carbon dioxide on flat metallic cathodes. Journal of Applied Electrochemistry 1997,
27(8): 875-889.
7. Annual Energy Outlook 2019. US Energy Information Administration
Office of Energy Analysis 2019.
8. Renewable capacity statistics 2019. International Renewable Energy Agency
(IRENA), Abu Dhabi.
9. Benson EE, Kubiak CP, Sathrum AJ, Smieja JM. Electrocatalytic and homogeneous
approaches to conversion of CO2 to liquid fuels. Chemical Society Reviews 2009,
38(1): 89-99.
10. Inglis JL, MacLean BJ, Pryce MT, Vos JG. Electrocatalytic pathways towards
sustainable fuel production from water and CO2. Coordination Chemistry Reviews
2012, 256(21-22): 2571-2600.
11. Goeppert A, Czaun M, Jones J-P, Surya Prakash GK, Olah GA. Recycling of carbon
dioxide to methanol and derived products - closing the loop. Chemical Society
Reviews 2014, 43(23): 7995-8048.
12. Albo J, Alvarez-Guerra M, Castano P, Irabien A. Towards the electrochemical
conversion of carbon dioxide into methanol. Green Chemistry 2015, 17(4):
2304-2324.
13. Li Y, Sun Q. Recent Advances in Breaking Scaling Relations for Effective
Electrochemical Conversion of CO2. Advanced Energy Materials 2016, 6(17):
1600463.
14. Elgrishi N, Chambers MB, Wang X, Fontecave M. Molecular polypyridine-based
metal complexes as catalysts for the reduction of CO2. Chemical Society Reviews
2017, 46(3): 761-796.
15. Handoko AD, Wei F, Jenndy, Yeo BS, Seh ZW. Understanding heterogeneous
electrocatalytic carbon dioxide reduction through operando techniques. Nature
Catalysis 2018, 1(12): 922-934.
104
Reference
16. Mota FM, Kim DH. From CO2 methanation to ambitious long-chain hydrocarbons:
alternative fuels paving the path to sustainability. Chemical Society Reviews 2019,
48(1): 205-259.
17. Jouny M, Luc W, Jiao F. General Techno-Economic Analysis of CO2 Electrolysis
Systems. Industrial & Engineering Chemistry Research 2018, 57(6): 2165-2177.
18. Higgins D, Hahn C, Xiang C, Jaramillo TF, Weber AZ. Gas-Diffusion Electrodes for
Carbon Dioxide Reduction: A New Paradigm. ACS Energy Letters 2019, 4(1):
317-324.
19. Zheng T, Jiang K, Ta N, Hu Y, Zeng J, Liu J, et al. Large-Scale and Highly Selective
CO2 Electrocatalytic Reduction on Nickel Single-Atom Catalyst. Joule 2019, 3(1):
265-278.
20. Hori Y. Electrochemical CO2 Reduction on Metal Electrodes. vol. 42, 2008, pp
89-189.
21. Kuhl KP, Hatsukade T, Cave ER, Abram DN, Kibsgaard J, Jaramillo TF.
Electrocatalytic Conversion of Carbon Dioxide to Methane and Methanol on
Transition Metal Surfaces. Journal of the American Chemical Society 2014, 136(40):
14107-14113.
22. Zhang Y-J, Sethuraman V, Michalsky R, Peterson AA. Competition between CO2
Reduction and H2 Evolution on Transition-Metal Electrocatalysts. Acs Catalysis 2014,
4(10): 3742-3748.
23. Kim D, Resasco J, Yu Y, Asiri AM, Yang P. Synergistic geometric and electronic
effects for electrochemical reduction of carbon dioxide using gold-copper bimetallic
nanoparticles. Nature Communications 2014, 5: 4948.
24. Lum Y, Kwon Y, Lobaccaro P, Chen L, Clark EL, Bell AT, et al. Trace Levels of
Copper in Carbon Materials Show Significant Electrochemical CO2 Reduction
Activity. Acs Catalysis 2016, 6(1): 202-209.
25. Singh MR, Kwon Y, Lum Y, Ager JW, III, Bell AT. Hydrolysis of Electrolyte
Cations Enhances the Electrochemical Reduction of CO2 over Ag and Cu. Journal of
the American Chemical Society 2016, 138(39): 13006-13012.
26. Clark EL, Hahn C, Jaramillo TF, Bell AT. Electrochemical CO2 Reduction over
Compressively Strained CuAg Surface Alloys with Enhanced Multi-Carbon
Oxygenate Selectivity. Journal of the American Chemical Society 2017, 139(44):
15848-15857.
27. Higgins D, Landersp AT, Ji Y, Nitopi S, Morales-Guio CG, Wang L, et al. Guiding
Electrochemical Carbon Dioxide Reduction toward Carbonyls Using Copper Silver
Thin Films with Interphase Miscibility. Acs Energy Letters 2018, 3(12): 2947-2955.
28. Morales-Guio CG, Cave ER, Nitopi SA, Feaster JT, Wang L, Kuhl KP, et al.
Improved CO2 reduction activity towards C2+ alcohols on a tandem gold on copper
electrocatalyst. Nature Catalysis 2018, 1(10): 764-771.
29. Zeng J, Bejtka K, Ju W, Castellino M, Chiodoni A, Sacco A, et al. Advanced Cu-Sn
foam for selectively converting CO2 to CO in aqueous solution. Applied Catalysis
B-Environmental 2018, 236: 475-482.
105
Reference
30. Zhu W, Zhang L, Yang P, Hu C, Dong H, Zhao Z-J, et al. Formation of Enriched
Vacancies for Enhanced CO2 Electrocatalytic Reduction over AuCu Alloys. Acs
Energy Letters 2018, 3(9): 2144-2149.
31. Huang J, Mensi M, Oveisi E, Mantella V, Buonsanti R. Structural Sensitivities in
Bimetallic Catalysts for Electrochemical CO2 Reduction Revealed by Ag-Cu
Nanodimers. Journal of the American Chemical Society 2019, 141(6): 2490-2499.
32. Chen Y, Li CW, Kanan MW. Aqueous CO2 Reduction at Very Low Overpotential on
Oxide-Derived Au Nanoparticles. Journal of the American Chemical Society 2012,
134(49): 19969-19972.
33. Zhu W, Michalsky R, Metin O, Lv H, Guo S, Wright CJ, et al. Monodisperse Au
Nanoparticles for Selective Electrocatalytic Reduction of CO2 to CO. Journal of the
American Chemical Society 2013, 135(45): 16833-16836.
34. Mistry H, Reske R, Zeng Z, Zhao Z-J, Greeley J, Strasser P, et al. Exceptional
Size-Dependent Activity Enhancement in the Electroreduction of CO2 over Au
Nanoparticles. Journal of the American Chemical Society 2014, 136(47):
16473-16476.
35. Zhu W, Zhang Y-J, Zhang H, Lv H, Li Q, Michalsky R, et al. Active and Selective
Conversion of CO2 to CO on Ultrathin Au Nanowires. Journal of the American
Chemical Society 2014, 136(46): 16132-16135.
36. Kim C, Jeon HS, Eom T, Jee MS, Kim H, Friend CM, et al. Achieving Selective and
Efficient Electrocatalytic Activity for CO2 Reduction Using Immobilized Silver
Nanoparticles. Journal of the American Chemical Society 2015, 137(43):
13844-13850.
37. Kumar B, Asadi M, Pisasale D, Sinha-Ray S, Rosen BA, Haasch R, et al. Renewable
and metal-free carbon nanofibre catalysts for carbon dioxide reduction. Nature
Communications 2013, 4: 2819.
38. Mao X, Hatton TA. Recent Advances in Electrocatalytic Reduction of Carbon
Dioxide Using Metal-Free Catalysts. Industrial & Engineering Chemistry Research
2015, 54(16): 4033-4042.
39. Wu J, Ma S, Sun J, Gold JI, Tiwary C, Kim B, et al. A Metal-free electrocatalyst for
Carbon Dioxide Reduction to Multi-carbon Hydrocarbons and Oxygenates. Nature
Communications 2016, 7: 13869.
40. Xie J, Zhao X, Wu M, Li Q, Wang Y, Yao J. Metal‐Free Fluorine‐Doped Carbon
Electrocatalyst for CO2 Reduction Outcompeting Hydrogen Evolution. Angewandte
Chemie International Edition 2018, 57(31): 9640-9644.
41. Zhang S, Kang P, Ubnoske S, Brennaman MK, Song N, House RL, et al.
Polyethylenimine-Enhanced Electrocatalytic Reduction of CO2 to Formate at
Nitrogen-Doped Carbon Nanomaterials. Journal of the American Chemical Society
2014, 136(22): 7845-7848.
42. Wu J, Yadav RM, Liu M, Sharma PP, Tiwary CS, Ma L, et al. Achieving Highly
Efficient, Selective, and Stable CO2 Reduction on Nitrogen-Doped Carbon Nanotubes.
Acs Nano 2015, 9(5): 5364-5371.
43. Zheng T, Jiang K, Wang H. Recent Advances in Electrochemical CO2-to-CO
Conversion on Heterogeneous Catalysts. Advanced Materials 2018, 30(48): 1802066.
106
Reference
44. Siahrostami S. Designing Carbon-Based Materials for Efficient Electrochemical
Reduction of CO2. Industrial & Engineering Chemistry Research 2019, 58(2):
879-885.
45. Zhu DD, Liu JL, Qiao SZ. Recent Advances in Inorganic Heterogeneous
Electrocatalysts for Reduction of Carbon Dioxide. Advanced Materials 2016, 28(18):
3423-3452.
46. Liu J, Ma Q, Huang Z, Liu G, Zhang H. Recent Progress in Graphene-Based
Noble-Metal Nanocomposites for Electrocatalytic Applications. Advanced Materials
2019, 31(9): 1800696.
47. Ju W, Bagger A, Hao G-P, Varela AS, Sinev I, Bon V, et al. Understanding Activity
and Selectivity of Metal-nitrogen-doped Carbon Catalysts for Electrochemical
Reduction of CO2. Nature Communications 2017, 8(1): 944.
48. Huan TN, Ranjbar N, Rousse G, Sougrati M, Zitolo A, Mougel V, et al.
Electrochemical Reduction of CO2 Catalyzed by Fe-N-C Materials: A
Structure-Selectivity Study. Acs Catalysis 2017, 7(3): 1520-1525.
49. Moeller T, Ju W, Bagger A, Wang X, Luo F, Trung Ngo T, et al. Efficient CO2 to CO
Electrolysis on Solid Ni-N-C Catalysts at Industrial Current Densities. Energy &
Environmental Science 2019, 12(2): 640-647.
50. Hori Y, Wakebe H, Tsukamoto T, Koga O. Electrocatalytic Process of CO Selectivity
in Electrochemical Reduction of CO2 at Metal Electrodes in Aqueous Media.
Electrochimica Acta 1994, 39(11-12): 1833-1839.
51. Kim C, Dionigi F, Beermann V, Wang X, Möller T, Strasser P. Alloy Nanocatalysts
for the Electrochemical Oxygen Reduction (ORR) and the Direct Electrochemical
Carbon Dioxide Reduction Reaction (CO2RR). Advanced Materials 2018: 1805617.
52. Yoshio H, Katsuhei K, Akira M, Shin S. Production of Methane and Ethylene in
Electrochemical Reduction of Carbon Dioxide at Copper Electrode in Aqueous
Hydrogencsrbonate Solution. Chemistry Letters 1986, 15(6): 897-898.
53. 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 1991, 64(1): 123-127.
54. Hori Y, Takahashi R, Yoshinami Y, Murata A. Electrochemical Reduction of CO at a
Copper Electrode. Journal of Physical Chemistry B 1997, 101(36): 7075-7081.
55. 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 2003, 199(1–2): 39-47.
56. Tang W, Peterson AA, Varela AS, Jovanov ZP, Bech L, Durand WJ, et al. The
Importance of Surface Morphology in Controlling the Selectivity of Polycrystalline
Copper for CO2 Electroreduction. Physical Chemistry Chemical Physics 2012, 14(1):
76-81.
57. Guo H, Chen Y, Cortie MB, Liu X, Xie Q, Wang X, et al. Shape-Selective Formation
of Monodisperse Copper Nanospheres and Nanocubes via Disproportionation
Reaction Route and Their Optical Properties. The Journal of Physical Chemistry C
2014, 118(18): 9801-9808.
107
Reference
58. Schouten KJP, Gallent EP, Koper MT. The Influence of pH on the Reduction of CO
and CO2 to Hydrocarbons on Copper Electrodes. J Electroanal Chem 2014, 716:
53-57.
59. Sen S, Liu D, Palmore GTR. Electrochemical Reduction of CO2 at Copper
Nanofoams. Acs Catalysis 2014, 4(9): 3091-3095.
60. Ka R, K e e R, Yı maz H, K e MTM, Mu G. Ma i u a i g he Hyd ca b
Selectivity of Copper Nanoparticles in CO2 Electroreduction by Process Conditions.
ChemElectroChem 2015, 2(3): 354-358.
61. Figueiredo MC, Ledezma-Yanez I, Koper MTM. In situ Spectroscopic Study of CO2
Electroreduction at Copper Electrodes in Acetonitrile. Acs Catalysis 2016, 6(4):
2382-2392.
62. Ledezma-Yanez I, Gallent EP, Koper MTM, Calle-Vallejo F. Structure-sensitive
Electroreduction of Acetaldehyde to Ethanol on Copper and Its Mechanistic
Implications for CO and CO2 Reduction. Catalysis Today 2016, 262: 90-94.
63. Varela AS, Kroschel M, Reier T, Strasser P. Controlling the Selectivity of CO2
Electroreduction on Copper: The Effect of the Electrolyte Concentration and the
Importance of the Local pH. Catalysis Today 2016, 260: 8-13.
64. Wang Z, Yang G, Zhang Z, Jin M, Yin Y. Selectivity on Etching: Creation of
High-Energy Facets on Copper Nanocrystals for CO2 Electrochemical Reduction.
ACS Nano 2016, 10(4): 4559-4564.
65. Kim D, Kley CS, Li Y, Yang P. Copper Nanoparticle Ensembles for Selective
Electroreduction of CO2 to C2-C3 products. Proceedings of the National Academy of
Sciences of the United States of America 2017, 114(40): 10560-10565.
66. Li Y, Cui F, Ross MB, Kim D, Sun Y, Yang P. Structure-Sensitive CO2
Electroreduction to Hydrocarbons on Ultrathin 5-fold Twinned Copper Nanowires.
Nano Letters 2017, 17(2): 1312-1317.
67. Bertheussen E, Hogg TV, Abghoui Y, Engstfeld AK, Chorkendorff I, Stephens IEL.
Electroreduction of CO on Polycrystalline Copper at Low Overpotentials. Acs Energy
Letters 2018, 3(3): 634-640.
68. Reske R, Mistry H, Behafarid F, Roldan Cuenya B, Strasser P. Particle Size Effects in
the Catalytic Electroreduction of CO2 on Cu Nanoparticles. Journal of the American
Chemical Society 2014, 136(19): 6978-6986.
69. Mistry H, Varela AS, Bonifacio CS, Zegkinoglou I, Sinev I, Choi Y-W, et al. Highly
Selective Plasma-activated Copper Catalysts for Carbon Dioxide Reduction to
Ethylene. Nature Communications 2016, 7: 12123.
70. Varela AS, Ju W, Reier T, Strasser P. Tuning the Catalytic Activity and Selectivity of
Cu for CO2 Electroreduction in the Presence of Halides. ACS Catalysis 2016, 6(4):
2136-2144.
71. Somorjai GA. Catalysis on the Atomic Scale (Emmett Award Lecture of 1977).
Catalysis Reviews 1978, 18(2): 173-197.
72. Peterson AA, Abild-Pedersen F, Studt F, Rossmeisl J, Norskov JK. How Copper
Catalyzes the Electroreduction of Carbon Dioxide into Hydrocarbon Fuels. Energy &
Environmental Science 2010, 3(9): 1311-1315.
108
Reference
73. Peterson AA, Nørskov JK. Activity Descriptors for CO2 Electroreduction to Methane
on Transition-Metal Catalysts. The Journal of Physical Chemistry Letters 2012, 3(2):
251-258.
74. Kortlever R, Shen J, Schouten KJP, Calle-Vallejo F, Koper MTM. Catalysts and
Reaction Pathways for the Electrochemical Reduction of Carbon Dioxide. Journal of
Physical Chemistry Letters 2015, 6(20): 4073-4082.
75. Huang Y, Handoko AD, Hirunsit P, Yeo BS. Electrochemical Reduction of CO2
Using Copper Single-Crystal Surfaces: Effects of CO* Coverage on the Selective
Formation of Ethylene. ACS Catalysis 2017, 7(3): 1749-1756.
76. Cheng T, Xiao H, Goddard WA, III. Full Atomistic Reaction Mechanism with
Kinetics for CO Reduction on Cu(100) From ab initio Molecular Dynamics
Free-energy Calculations at 298 K. Proceedings of the National Academy of Sciences
of the United States of America 2017, 114(8): 1795-1800.
77. Goodpaster JD, Bell AT, Head-Gordon M. Identification of Possible Pathways for
C-C Bond Formation during Electrochemical Reduction of CO2: New Theoretical
Insights from an Improved Electrochemical Model. Journal of Physical Chemistry
Letters 2016, 7(8): 1471-1477.
78. Garza AJ, Bell AT, Head-Gordon M. Mechanism of CO2 Reduction at Copper
Surfaces: Pathways to C2 Products. Acs Catalysis 2018, 8(2): 1490-1499.
79. Cheng T, Xiao H, Goddard WA, III. Reaction Mechanisms for the Electrochemical
Reduction of CO2 to CO and Formate on the Cu(100) Surface at 298 K from
Quantum Mechanics Free Energy Calculations with Explicit Water. Journal of the
American Chemical Society 2016, 138(42): 13802-13805.
80. Xiao H, Cheng T, Goddard WA, III. Atomistic Mechanisms Underlying Selectivities
in C1 and C2 Products from Electrochemical Reduction of CO on Cu(111). Journal of
the American Chemical Society 2017, 139(1): 130-136.
81. Montoya JH, Peterson AA, Nørskov JK. Insights into C-C Coupling in CO2
Electroreduction on Copper Electrodes. ChemCatChem 2013, 5(3): 737-742.
82. Schouten KJP, Kwon Y, van der Ham CJM, Qin Z, Koper MTM. A New Mechanism
for the Selectivity to C1 and C2 Species in the Electrochemical Reduction of Carbon
Dioxide on Copper Electrodes. Chemical Science 2011, 2(10): 1902-1909.
83. Kuhl KP, Cave ER, Abram DN, Jaramillo TF. New Insights into the Electrochemical
Reduction of Carbon Dioxide on Metallic Copper Surfaces. Energy & Environmental
Science 2012, 5(5): 7050-7059.
84. Calle-Vallejo F, Koper MTM. Theoretical Considerations on the Electroreduction of
CO to C2 Species on Cu(100) Electrodes. Angewandte Chemie-International Edition
2013, 52(28): 7282-7285.
85. Nie X, Esopi MR, Janik MJ, Asthagiri A. Selectivity of CO2 Reduction on Copper
Electrodes: The Role of the Kinetics of Elementary Steps. Angewandte
Chemie-International Edition 2013, 52(9): 2459-2462.
86. Montoya JH, Shi C, Chan K, Norskov JK. Theoretical Insights into a CO
Dimerization Mechanism in CO2 Electroreduction. Journal of Physical Chemistry
Letters 2015, 6(11): 2032-2037.
109
Reference
87. Durand WJ, Peterson AA, Studt F, Abild-Pedersen F, Nørskov JK. Structure Effects
on the Energetics of the Electrochemical Reduction of CO2 by Copper Surfaces.
Surface Science 2011, 605(15): 1354-1359.
88. Nie X, Luo W, Janik MJ, Asthagiri A. Reaction Mechanisms of CO2 Electrochemical
Reduction on Cu(111) Determined with Density Functional Theory. Journal of
Catalysis 2014, 312: 108-122.
89. Luo W, Nie X, Janik MJ, Asthagiri A. Facet Dependence of CO2 Reduction Paths on
Cu Electrodes. Acs Catalysis 2016, 6(1): 219-229.
90. Huang Y, Ong CW, Yeo BS. Effects of Electrolyte Anions on the Reduction of
Carbon Dioxide to Ethylene and Ethanol on Copper (100) and (111) Surfaces.
Chemsuschem 2018, 11(18): 3299-3306.
91. Tsang CF, Javier AC, Kim Y-G, Baricuatro JH, Cummins KD, Kim J, et al.
Potential-Dependent Adsorption of CO and Its Low-Overpotential Reduction to
CH3CH2OH on Cu(511) Surface Reconstructed from Cu(pc): Operando Studies by
Seriatim STM-EQCN-DEMS. Journal of the Electrochemical Society 2018, 165(15):
J3350-J3354.
92. Li CW, Kanan MW. CO2 Reduction at Low Overpotential on Cu Electrodes
Resulting from the Reduction of Thick Cu2O Films. Journal of the American
Chemical Society 2012, 134(17): 7231-7234.
93. Li CW, Ciston J, Kanan MW. Electroreduction of Carbon Monoxide to Liquid Fuel
on Oxide-derived Nanocrystalline Copper. Nature 2014, 508: 504.
94. Verdaguer-Casadevall A, Li CW, Johansson TP, Scott SB, McKeown JT, Kumar M,
et al. Probing the Active Surface Sites for CO Reduction on Oxide-Derived Copper
Electrocatalysts. Journal of the American Chemical Society 2015, 137(31):
9808-9811.
95. Manthiram K, Beberwyck BJ, Aivisatos AP. Enhanced Electrochemical Methanation
of Carbon Dioxide with a Dispersible Nanoscale Copper Catalyst. Journal of the
American Chemical Society 2014, 136(38): 13319-13325.
96. Chen CS, Handoko AD, Wan JH, Ma L, Ren D, Yeo BS. Stable and Selective
Electrochemical Reduction of Carbon Dioxide to Ethylene on Copper Mesocrystals.
Catalysis Science & Technology 2015, 5(1): 161-168.
97. Dutta A, Rahaman M, Luedi NC, Mohos M, Broekmann P. Morphology Matters:
Tuning the Product Distribution of CO2 Electroreduction on Oxide-Derived Cu Foam
Catalysts. ACS Catalysis 2016, 6(6): 3804-3814.
98. Handoko AD, Ong CW, Huang Y, Lee ZG, Lin L, Panetti GB, et al. Mechanistic
Insights into the Selective Electroreduction of Carbon Dioxide to Ethylene on
Cu2O-derived Copper Catalysts. The Journal of Physical Chemistry C 2016, 120(36):
20058-20067.
99. Ren D, Ang BS-H, Yeo BS. Tuning the Selectivity of Carbon Dioxide
Electroreduction toward Ethanol on Oxide-Derived CuxZn Catalysts. Acs Catalysis
2016, 6(12): 8239-8247.
100. Eilert A, Cavalca F, Roberts FS, Osterwalder J, Liu C, Favaro M, et al. Subsurface
Oxygen in Oxide-Derived Copper Electrocatalysts for Carbon Dioxide Reduction.
The Journal of Physical Chemistry Letters 2017, 8(1): 285-290.
110
Reference
101. Liu C, Lourenço MP, Hedström S, Cavalca F, Diaz-Morales O, Duarte HA, et al.
Stability and Effects of Subsurface Oxygen in Oxide-Derived Cu Catalyst for CO2
Reduction. The Journal of Physical Chemistry C 2017, 121(45): 25010-25017.
102. Lum Y, Yue B, Lobaccaro P, Bell AT, Ager JW. Optimizing C–C Coupling on
Oxide-Derived Copper Catalysts for Electrochemical CO2 Reduction. The Journal of
Physical Chemistry C 2017, 121(26): 14191-14203.
103. Lum Y, Ager JW. Stability of Residual Oxides in Oxide-Derived Copper Catalysts
for Electrochemical CO2 Reduction Investigated with O18 Labeling. Angewandte
Chemie-International Edition 2018, 57(2): 551-554.
104. Pander JE, III, Ren D, Huang Y, Loo NWX, Hong SHL, Yeo BS. Understanding the
Heterogeneous Electrocatalytic Reduction of Carbon Dioxide on Oxide-Derived
Catalysts. Chemelectrochem 2018, 5(2): 219-237.
105. Lum Y, Ager JW. Evidence for Product-specific Active Sites on Oxide-derived Cu
Catalysts for Electrochemical CO2 reduction. Nature Catalysis 2019, 2(1): 86-93.
106. Hahn C, Hatsukade T, Kim Y-G, Vailionis A, Baricuatro JH, Higgins DC, et al.
Engineering Cu Surfaces for the Electrocatalytic Conversion of CO2: Controlling
Selectivity toward Oxygenates and Hydrocarbons. Proceedings of the National
Academy of Sciences of the United States of America 2017, 114(23): 5918-5923.
107. Roberts FS, Kuhl KP, Nilsson A. High Selectivity for Ethylene from Carbon Dioxide
Reduction over Copper Nanocube Electrocatalysts. Angewandte
Chemie-International Edition 2015, 54(17): 5179-5182.
108. Loiudice A, Lobaccaro P, Kamali EA, Thao T, Huang BH, Ager JW, et al. Tailoring
Copper Nanocrystals towards C2 Products in Electrochemical CO2 Reduction.
Angewandte Chemie-International Edition 2016, 55(19): 5789-5792.
109. Xiao H, Goddard WA, Cheng T, Liu Y. Cu metal Embedded in Oxidized Matrix
Catalyst to Promote CO2 Activation and CO Dimerization for Electrochemical
Reduction of CO2. Proceedings of the National Academy of Sciences 2017, 114(26):
6685-6688.
110. Gao D, Zegkinoglou I, Divins NJ, Scholten F, Sinev I, Grosse P, et al.
Plasma-Activated Copper Nanocube Catalysts for Efficient Carbon Dioxide
Electroreduction to Hydrocarbons and Alcohols. Acs Nano 2017, 11(5): 4825-4831.
111. Favaro M, Xiao H, Cheng T, Goddard WA, Yano J, Crumlin EJ. Subsurface Oxide
Plays a Critical Role in CO2 Activation by Cu (111) Surfaces to Form Chemisorbed
CO2, the First Step in Reduction of CO2. Proceedings of the National Academy of
Sciences 2017, 114(26): 6706-6711.
112. Cavalca F, Ferragut R, Aghion S, Eilert A, Diaz-Morales O, Liu C, et al. Nature and
Distribution of Stable Subsurface Oxygen in Copper Electrodes During
Electrochemical CO2 Reduction. The Journal of Physical Chemistry C 2017, 121(45):
25003-25009.
113. Garza AJ, Bell AT, Head-Gordon M. Is Subsurface Oxygen Necessary for the
Electrochemical Reduction of CO2 on Copper? The Journal of Physical Chemistry
Letters 2018, 9(3): 601-606.
111
Reference
114. Liu C, Hedström S, Stenlid JH, Pettersson LGM. Amorphous, Periodic Model of a
Copper Electrocatalyst with Subsurface Oxygen for Enhanced CO Coverage and
Dimerization. The Journal of Physical Chemistry C 2019, 123(8): 4961-4968.
115. Rietveld H. A Profile Refinement Method for Nuclear and Magnetic Structures.
Journal of applied Crystallography 1969, 2(2): 65-71.
116. Fassel VA, Kniseley RN. Inductively Coupled Plasma. Optical Emission
Spectroscopy. Analytical Chemistry 1974, 46(13): 1110A-1120A.
117. Risch M, Klingan K, Heidkamp J, Ehrenberg D, Chernev P, Zaharieva I, et al.
Nickel-oxido Structure of a Water-oxidizing Catalyst Film. Chemical
Communications 2011, 47(43): 11912-11914.
118. Risch M, Ringleb F, Kohlhoff M, Bogdanoff P, Chernev P, Zaharieva I, et al. Water
Oxidation By Amorphous Cobalt-based Oxides: in situ Tracking of Redox
Transitions and Mode of Catalysis. Energy & Environmental Science 2015, 8(2):
661-674.
119. Hansen TW, DeLaRiva AT, Challa SR, Datye AK. Sintering of Catalytic
Nanoparticles: Particle Migration or Ostwald Ripening? Acc Chem Res 2013, 46(8):
1720-1730.
120. Schreier M, Yoon Y, Jackson MN, Surendranath Y. Competition Between H and CO
for Active Sites Governs Copper-mediated Electrosynthesis of Hydrocarbon Fuels.
Angewandte Chemie International Edition 2018, 57(32): 10221-10225.
121. Huang J, Mensi M, Oveisi E, Mantella V, Buonsanti R. Structural Sensitivities in
Bimetallic Catalysts for Electrochemical CO2 Reduction Revealed by Ag-Cu
Nanodimers. J Am Chem Soc 2019, 141(6): 2490-2499.
122. Lum Y, Ager JW. Sequential Catalysis Controls Selectivity in Electrochemical CO2
Reduction on Cu. Energy & Environmental Science 2018, 11(10): 2935-2944.
123. Jovanove Z, Araújo J, Li S, Strasser P. Catalyst Preoxidation and EDTA Electrolyte
Additive Remedy Activity and Selectivity Declines During Electrochemical CO2
Reduction. Journal of Physical Chemistry C 2019, 123(4): 2165-2174.
124. Feng X, Jiang K, Fan S, Kanan MW. A Direct Grain-Boundary-Activity Correlation
for CO Electroreduction on Cu Nanoparticles. ACS Central Science 2016, 2(3):
169-174.
125. Liu X, Xiao J, Peng H, Hong X, Chan K, Nørskov JK. Understanding Trends in
Electrochemical Carbon Dioxide Reduction Rates. Nature Communications 2017, 8:
15438
126. Wang X, Varela AS, Bergmann A, Kuehl S, Strasser P. Catalyst Particle Density
Controls Hydrocarbon Product Selectivity in CO2 Electroreduction on CuOx.
ChemSusChem, 2017, 10: 4642–4649.
127. Wang X, Araújo J, Ju W, Bagger A, Schmies H, Kuehl S, et al. Mechanistic reaction
pathways of enhanced ethylene yields during electroreduction of CO2–CO co-feeds
on Cu and Cu-tandem electrocatalysts. Nature Nanotechnology, 2019, 14: 1063-1070.
112
Reference
113
Appendix
Appendix
A1. Supplementary Information to Chapter 4
Catalyst particle density controls hydrocarbon product selectivity in CO2 electroreduction
on CuOx
Reproduced with permission from ChemSusChem, 2017, 10, 4642-4649. Copyright of The
John Wiley and Sons (2017).126
Figure A1.1 a) b) Representative TEM images of CuOx nanoparticles after 7 days in hexane.
The inset shows the size distribution of CuOx nanoparticles.
Figure A1.2 a) Diffraction pattern of as-prepared CuOx NPs. Also indicated are diffraction
patterns of metallic Cu (red, PDF#98-000-0172), CuO (blue, PDF#01-078-0428), and Cu2O
(green, PDF#98-000-0186).
114
Appendix
Figure A1.3 The crystal structures of a) metallic Cu, b) cubic-Cu2O (cuprite), c)
monoclinic-CuO (Tenorite). Color legend: blue spheres, Cu; red sphere, O.
Figure A1.4 The total Faradaic efficiencies of gaseous products as a function of areal density
at various applied overpotentials. Faradaic efficiencies of CO2RR and HER over a) 4 µg cm-2,
b) 15 µg cm-2, c) 31 µg cm-2.
115
Appendix
Figure A1.5 The Faradaic efficiencies for all detected products on CuOx NPs after 400 min
with the areal particle density of 31 µg cm-2 at -1.0VRHE.
Figure A1.6 a) Total Faradaic efficiency for gaseous products as a function of areal particle
density. b) Trends in C2H4/CH4 production ratio at varying catalyst areal densities at -1.01
VRHE applied electrode potential. Potentials are IR corrected.
116
Appendix
Figure A1.7 Electrochemical CO2 reduction activity (left axis of ordinates) and selectivity
(right axis of ordinates) of glassy carbon as blank at various overpotentials. The amount of H2
at -0.66 VRHE is not detectable (under TCD detecting limitation).
117
Appendix
A2. Supplementary Information to Chapter 5
Mechanistic reaction pathways of enhanced ethylene yields during electroreduction of
CO2-CO co-feeds on Cu and Cu-tandem electrocatalysts
Reproduced in part with permission from Nat. Nanotechnol., 2019, 14, 1063-1070. Copyright
of Springer Nature.127
Figure A2.1 Time-dependent absolute product formation rates on H2. Orange: CO2RR in
CO2 saturated 0.1 M KHCO3 (pH=6.8); cyan: CORR in 0.1 M K2HPO4/KH2PO4 (pH=6.9);
Violet: co-feeds (CO2/CO) reduction reductions in CO2 saturated 0.1 M KHCO3. Catalyst
loading: 100 ug/cm-2.
118
Appendix
Figure A2.2 Liquid products analysis after 4-hour measurements in various feeds conditions.
Absolute product formation rates of detected liquid products on CuOx NPs with various
CO2/CO ratios.
Figure A2.3 Formate analysis after 4-hour measurements in various feeds conditions. a)
Absolute product formation rate of formate on CuOx NPs with various CO2/CO ratios. b) The
pathway from CO2 to formate. It shows that the formate only forms in CO2-involving feeds.
119
Appendix
Figure A2.4 The potential-dependent tests of CORR, CO2RR and co-feed reduction
(CO2/CO 1:1). Absolute product formation rates of a) CH4, b) C2H4 and c) H2 on CuOx
nanoparticles.
-1.00 -0.95 -0.90 -0.85
0.0
0.5
1.0
1.5
2.0
C2H4
Potential / VRHE
Production Rate / nmol cm-2 s-1
CO2
CO2RR
CO2/CO (1:1)
CORR
-1.00 -0.95 -0.90 -0.85
1
10
100
Potential / VRHE
Production Rate / nmol cm-2 s-1
H2
-1.00 -0.95 -0.90 -0.85
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
CH4
Potential / VRHE
Production Rate / nmol cm-2 s-1
a
b
c
120
Appendix
Figure A2.5 XRD profiles of as-prepared working electrodes tested before and after
electrochemical reductions. Grazing incidence X-ray Diffraction (GI-XRD) analysis of thin
film catalysts coated on glassy carbon as working electrodes. The bottom lines show pdf
reference patterns of metallic copper (Cu, blue), pdf # 98-000-0172; cuprite (Cu2O, green),
pdf # 98-000-0186, and of tenorite, (CuO, pink), pdf # 01-078-0428.
121
Appendix
Figure A2.6 SEM images of as-prepared working electrodes after electrochemical reductions.
Morphology evolution of CuOx NPs during a-c) CO2RR in CO2 saturated 0.1 M KHCO3
(pH=6.8); d-e) CORR in 0.1 M K2HPO4/KH2PO4 (pH=6.9); g-i) co-deed reduction (CO2: CO
1:1) in CO2 saturated 0.1 M KHCO3 (pH=6.8).
Figure A2.7 High angle annular dark field (HADDF) images and EDX (Energy dispersive
X-ray spectroscopy) line scan of the CuOx NPs after co-feed (1:1) reduction. EDX line
analysis of the CuOx NPs was carried out by detection of X-ray band derived from O K, F K,
S K, and Cu K with two distinct locations.
122
Appendix
Figure A2.8 Investigation of pure feed reductions of CORR and CO2RR. a) Absolute product
formation rates of major gaseous products (black: H2; blue: CH4; pink: C2H4) as a function of
applied electrode potentials during CORR (solid plots) in 0.1 M K2HPO4/KH2PO4 (pH=6.9)
and CO2RR (hollow plots) in CO2-saturated 0.1M KHCO3. b, c) Faradaic efficiencies of
major gaseous products (blue: CH4; red: C2H4) as a function of applied electrode potentials
during CORR in 0.1 M K2HPO4/KH2PO4 (pH=6.9) and 0.1 M KOH (pH=13). Potentials are
IR corrected.
123
Appendix
Figure A2.9 The ratios of reaction between the three mechanisms. The ratio of reaction
between the three mechanisms on a) CO-CO, CO2-CO2, and b) CO2-CO, with respective
potential range for integration at 0.05 V from cathodic to anodic sweep of potentials.
Figure A2.10 Decomposition of the three pathways calculated by the co-feed system of a
mixture of carbon isotopes. The charge under deconvoluted curves in each correspondent
pathway CO-CO (red area), CO2-CO (blue area) and CO2-CO2 (green area) in fragment M-H+.
Analysis were carried out during cathodic, and anodic direction and the sum with
combination direction (total charge)
124
Appendix
Figure A2.11 Differential electrochemical mass spectrometry data obtained on 0.785 cm2
large glassy carbon electrode drop-coated with CuOx NPs catalysts and using pure CO, pure
CO2, and CO2-CO co-feed. The graph above presents the ion mass current over time with
products deconvoluted CO (m/z 28) measured in pure CO, pure CO2 and CO2-CO co-feed
system in a mixed ratio of 1:1 or in the diluted gas ratio (1:3). The bar graph in the figure
illustrates the ionic mass charge for CO evolution (taken by m/z 28) for each feeds measured.
To deconvolute of CO signal from mass fragmentation signal of CO2 (CO+ m/z 28), we used
the relation of molecular mass signal m/z 44 and fragment signal m/z 28. Even though the
signal intensity of molecular mass ethylene (m/z 28) is relatively small (~3%) compared to
CO signal, it was subtracted from the CO molecule mass signal.
125
Appendix
Figure A2.12 SEM images of as-prepared the bifunctional hybrid catalyst. SEM images of
CuOx NPs (50 ug cm-2)-NiNC (1:4) tandem catalyst with carbon paper as working electrode
(carbon fiber), a) SEI mode; b) COMPO mode.
Figure A2.13 Catalytic performance of the tandem CuOx-NiNC catalysts detected by on-line
GC in H-cell. Absolute product formation rates on a) CO and b) CH4 with different catalyst
loading variation at -0.84 VRHE and -0.9 VRHE. No detectable CH4 formation on pure NiNC
and pure CuOx NPs with areal loading of 100 g cm-2at -0.84 VRHE.
126
Appendix
0 4 8 12 16 20 24
0
5
10
15
20
25
Time / h
H2
CO
CH4
C2H4
Production Rate / nmon cm
-2
s
-1
Figure A2.14 The long stability test of tandem CuOx-Ag (1:3) catalyst at a constant electrode
potential of -1.0 VRHE in CO2- a u a ed 0.1 M KHCO3.
Table A2.1 Liquid products of tandem CuOx-NiNC catalysts during CO2RR after 90 min at
fixed overpotential.
Tandem Catalyst
Acetaldehyde
(nmol cm-2 s-1)
Propionaldehyde
(nmol cm-2 s-1)
EtOH
(nmol cm-2 s-1)
PrOH
(nmol cm-2 s-1)
AllylOH
(nmol cm-2 s-1)
CuOx NP-100 ug
cm-2 : NiNC (1:2)
@ -0.9 VRHE
1.96E-4 ±
2.6E-5
1.16E-4 ± 9.4E-6
0.00111 ± 2.8E-5
5.01E-4 ± 3.1E-5
2.12E-4 ±
3.9E-6
127
Appendix
A3. Supplementary Information to Chapter 6
Sheet-like Copper Oxides with Stable and Selective Ethylene Production for Direct CO2
Electroreduction
Figure A3.1 SEM images of Cu(OH)2 intermediate formed during synthesis process.
Figure A3.2 a) Cross-section TEM images of as-prepared CuOx NS and b) corresponding
intensity of SAED intensity.
128
Appendix
Figure A3.3 Crystal structure of monoclinic CuO with indexing plane.
Figure A3.4 a) Partial contour plot of CuOx NS. b) Azimuthally integrated line profiles of
as-prepared CuOx NS.
129
Appendix
Figure A3.5 Faradaic efficiencies of main products distribution after CO2RR in H-cell
design.
Figure A3.6 a) Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NS in CO2-saturated 0.1M KHCO3 at -0.84 VRHE. SEM image of as-prepared CuOx
catalyst on GC electrode (100 g cm-2) before b) and c) after 20-hour electrolysis.
-0.76 -0.84 -0.88 -0.99
0
20
40
60
80
100
HCOOH
AllylOH
PrOH
EtOH
Aceton
C2H4
CH4
CO
H2
FE / %
E / VRHE
130
Appendix
Figure A3.7 a) Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NS in CO2-saturated 0.1M KHCO3 at -0.76 VRHE. Insert: SEM image of as-prepared
CuOx catalyst on GC electrode (100 g cm-2) before and after 20-h electrolysis. Production
rates b) and Faradaic efficiencies c) of main gaseous products over 1200 min.
Figure A3.8 SEM image of 1-h Chronoamperometric performance of the CO2 reduction
reaction on CuOx NS in CO2-saturated 0.1M KHCO3 at -0.84 VRHE.
131
Appendix
Figure A3.9 Comparison of XANES at the Cu K-edge of references powders CuO, Cu(OH)2
and the CuOx NS film in the dry state, and at OCP in 0.1 M KHCO3 at pH 6.8.
Figure A3.10 a) FT of k3 weighted EXAFS a) and b) EXAFS at the Cu K-edge of Cu metal
foil, Cu(OH)2 powder, CuO powder, and the CuOx NS film at OCP in 0.1 M KHCO3 at pH
132
Appendix
6.8. Colored lines represent the experimental data and black lines in the simulations. The
distance on the x-axis in a) is reduced by 0.35 Å relative to the real distance.
Figure A3.11 a) FT of k3 weighted EXAFS at the Cu K-edge of the CuOx NS film at OCP in
0.1 M KHCO3 at pH 6.8 and during 130 min CO2RR at -0.84 VRHE. The distance on the
x-axis is reduced by 0.35 Å relative to the real distance. b) EXAFS (k3 weighted) at the Cu
K-edge of the CuOx NS film at OCP in 0.1 M KHCO3 at pH 6.8 and during 130 min CO2RR
at -0.84 VRHE.
133
Appendix
Figure A3.12 EXAFS distances of the first Cu-O coordination sphere and the first
intermetallic Cu-Cu shell.
020 40 60 80
1.91
1.92
1.93
1.94
1.95
1.96
1.97
1.98
020 40 60 80 100 120 140
2.50
2.51
2.52
2.53
2.54
2.55
RCu-O (Å)
a
Time (min)
Time (min)
OCP
b
RCu-Cu (Å)
OCP
134
Appendix
Table A3.1 Cu K-edge simulation parameters of the FT-EXAFS oscillations of as-prepared
CuOx NS film. Data range k= 3-11 Å-1, Amplitude reduction factor
=0.8.
CuO NS film as-prepared
Shell
R /Å
N
σ
Rf
OCP
Cu-O
Cu-Cu
1.94 ± 0.01
2.69 ± 0.02
4.2 ± 0.13
2.2 ± 0.22
0.0431*
0.0415*
11.39
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
2.88 ± 0.02
3.03 ± 0.01
3.14 ± 0.01
3.30 ± 0.01
4.1 ± 0.14
4.1 ± 0.32
2.3 ± 0.32
1.8 ± 0.13
0.0629*
0.0536*
0.0318*
0.0997*
0 min
Cu-O
Cu-Cu
1.93 ± 0.01
2.53 ± 0.02
2.82 ± 0.21
3.9 ± 1
0.0431*
0.0853*
10.19
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
2.74 ± 0.03
2.93 ± 0.03
3.11 ± 0.03
4.43 ± 0.02
5.37 ± 1.9
7.17 ± 2.1
3.46 ± 0.8
4.51 ± 1
0.0629*
0.0536*
0.0318*
0.0885*
10 min
Cu-O
Cu-Cu
1.93 ± 0.01
2.51 ± 0.01
2.25 ± 0.22
4.93 ± 0.32
0.0431*
0.0853*
11.9
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
3.63 ± 0.11
4.41 ± 0.03
4.92 ± 0.02
3.48 ± 0.07
1.84 ± 2.4
6.4 ± 1.3
2.81 ± 0.66
4.12 ± 2.4
0.0906*
0.0885*
0.0344*
0.0997*
30 min
Cu-O
Cu-Cu
1.94 ± 0.01
2.53 ± 0.002
1.56 ± 0.23
6.95 ± 0.26
0.0431*
0.0853*
9.62
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
3.63 ± 0.02
4.44 ± 0.008
4.95 ± 0.01
3.47 ± 0.02
5.74 ± 1.6
9.91 ± 1.3
3.7 ± 0.64
5.61 ± 1.7
0.0906*
0.0885*
0.0344*
0.0997*
135
Appendix
Numbers marked with * are fixed according to the information in the cif file.
CuO NS film as-prepared
Shell
R /Å
N
σ
Rf
50 min
Cu-O
Cu-Cu
1.94 ± 0.02
2.53 ± 0.01
0.975 ± 0.22
8.23 ± 0.26
0.0431*
0.0853*
9.01
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
3.61 ± 0.02
4.44 ± 0.01
4.95 ± 0.01
3.45 ± 0.02
7.72 ± 1.6
11.9 ± 1.3
4.93 ± 0.65
6.98 ± 1.8
0.0906*
0.0885*
0.0344*
0.0997*
80 min
Cu-O
Cu-Cu
1.95 ± 0.03
2.52 ± 0.004
0.656 ± 0.23
9.23 ± 0.26
0.0431*
0.0853*
8.438
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
3.64 ± 0.02
4.44 ± 0.01
4.96 ± 0.01
3.46 ± 0.02
8.44 ± 1.6
14.4 ± 1.3
5.63 ± 0.64
6.88 ± 1.7
0.0906*
0.0885*
0.0344*
0.0997*
90 min
Cu-Cu
2.523 ± 0.004
9.74 ± 0.25
0.0853*
12.63
Cu-Cu
Cu-Cu
Cu-Cu
3.58 ± 0.02
4.44 ± 0.008
4.96 ± 0.009
2.8 ± 0.7
14.9 ± 1.2
5.95 ± 0.62
0.0906*
0.0885*
0.0344*
120 min
Cu-Cu
2.53 ± 0.004
10.1 ± 0.25
0.0853*
12.1
Cu-Cu
Cu-Cu
Cu-Cu
3.61 ± 0.01
4.44 ±0.01
4.95 ± 0.01
3.91 ± 0.27
14.8 ± 1.2
6.54 ± 0.62
0.0906*
0.0885*
0.0344*
Only when the Cu metal content is ≤90% it is reasonable to include CuO distances (marked red)
136
Appendix
Reference materials
Shell
R /Å
N
σ
Rf
CuO
powder
Cu-O
Cu-O
1.95 ± 0.01
2.74 ± 0.01
4*
2*
0.0614
0.0997
9.07
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
Cu-Cu
2.88 ± 0.02
3.02 ± 0.01
3.16 ± 0.02
3.38 ± 0.03
3.88 ± 0.03
4.79 ± 0.01
4.97 ± 0.01
4*
4*
2*
2*
2*
5*
4*
0.081
0.0717
0.0317
0.99
0.0998
0.0747
0.0333
Cu(OH)2
powder
Cu-O
Cu-O
1.97 ± 0.01
1.89 ± 0.02
2.2 ± 0.18
2.2 ± 0.17
0.0317*
0.0596*
10.2
Cu-O
Cu-Cu
Cu-Cu
2.89 ± 0.02
2.93 ± 0.03
3.33 ± 0.03
2.11 ± 0.27
2.08 ± 0.15
3.68 ± 0.16
0.0987*
0.0316*
0.0987*
Cu metal
Cu-Cu
Cu-Cu
2.53 ± 0.001
3.60 ± 0.01
12 ± 0.24
5.57 ± 0.21
0.0853*
0.0906*
10.87
Cu-Cu
Cu-Cu
4.44 ± 0.004
4.96 ± 0.004
21 ± 1.2
8.65 ± 0.59
0.0885*
0.0344*
Simulation approach: first I fixed the coordination number N to the expected value (cif file)
and calculated the distances R for each shell, and the Debeye Waller. Then, the Debye Waller
factor was fixed to obtained values above and N was calculated, highly similar, but slightly
lower Rf factor (better fit quality).
137
Appendix
A4. Supplementary Information to Chapter 7
Achieving high C2H4 Evolution at industrial current densities on CuOx nanosheet derived
gas diffusion electrode
Figure A4.1 The partial current density of C2+ products as a function of applied geometric
current density for CuOx NS in 1 M KHCO3.
138
Appendix
0 4 8 12 16 20 24
0
20
40
60
80
Time / h
Partial Current Density-C2H4 / mA mg-1
0 4 8 12 16 20 24
0.0
0.5
1.0
1.5
2.0
Time / h
Production Rate Ratio: C2H4:CH4
0 4 8 12 16 20 24
0
10
20
30
40
50 H2
CO
CH4
C2H4
Time / h
FE / %
0 4 8 12 16 20 24
1
10
100 H2
CO
CH4
C2H4
Time / h
Production Rate / nmol cm-2 s-1
a)b)
c)d)
Figure A4.2 Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NS in CO2-saturated 0.1 M KHCO3 at -1.0 VRHE, performed in an H-call configuration.
a) Absolute product formation rates of major gaseous products over 24 h. b) Faradaic
efficiencies of major gaseous products. c) The experimental results of C2H4/CH4 product
selectivity over 24 h. d) Partial current of C2H4 with time.
Figure A4.3 SEM image of as-prepared CuOx catalyst after H-cell test.
139
Appendix
Figure A4.4 Morphology evolution of CuOx NS. a) SEM image of as-prepared CuOx
NS/glassy carbon electrode. Insert: TEM image of CuOx NS. b) SEM image of CuOx NS in
CO2-saturated 0.1 M KHCO3 after 1-hour reaction at -1.0 VRHE. c) TEM image of CuOx NS in
N2-saturated 0.1 M Na2HPO4/NaH2PO4 (pH=6.9) after 15 min at -1.0 VRHE.
140
Appendix
Table of figures and schemes
Figure 1.1 Total global renewable energy capacity from Hydropower and marine (orange),
Wind (green), Solar (violet), Bioenergy (yellow), Geothermal (blue) sources. Data taken from
ref 8. Copyright International Renewable Energy Agency…………………………………….1
Figure 1.2 General scheme of a typical electrolyzer for CO2RR……………………………...3
Figure 1.3 Classification of various metal depending on formation of major products in
electrochemical CO2 reduction. a) Periodic table and faradaic efficiency of major product
from experimental data by Hori. H2 (red), CO (blue), formate (yellow), hydrocarbon (green) b)
The experimental product classification of H2, CO, a d HCOOH by he ΔEH*-ΔECOOH* and the
ΔEH* c) Faradaic efficiency of CO2 educ i eac i by ΔEH*. a-c) Modified and reproduced
from ref 51 with permission. Copyright 2018, John Wiley and Sons…………………………..4
Figure 1.4 Possible C1 species for C-C coupling……………………………………………...5
Figure 1.5 a) Tuning the product distribution of CO2 electroreduction on OD-Cu foam
catalysts. b) Optimizing C-C coupling on oxide-derived copper. c) Effects of CO* coverage
on the selective formation of ethylene. d) Nature and distribution of stable subsurface oxygen
in OD-Cu. a) to d) are adapted from ref 97, 102, 75 and 112, respectively, with the permission,
copyright American Chemical Society. …………………………………….…………………7
Figure 2.1 Schematic overview of scientific questions and goals dealt with in the course of
this work. ………………………………………………………………………………………9
Figure 3.1 lists the interaction between electrons and samples used for different
techniques…………………………………………………………………………………….17
Figure 3.2 Schematic illustration of the Bragg equation……………………………………..18
Figure 3.3 Bohr model of an atom. When energy is absorbed by an atom, electron jumps from
its ground state orbital to an orbital with higher energy level. For the stabilizing, the atom
may decay back to a lower energy state by emitting a photon, . …………………………21
Figure 3.4 Schematic (a) and photograph (b) of H-type two compartments cell divided by a
polymer membrane. ………………………………………………………………………….22
Figure 3.5 Geometry defining of the incident beam and scattered beam with scattering vector
. ……………………………………………………………………………………………..27
Figure 3.6. a) Schematic illustration of the in situ WAXS setup under electrochemical
condition using at a synchrotron facility, showing the incident and scattered X-rays, the 2D
detector and the grazing incidence cell. The side view (b) and the top view (c) of the test
station at the European Synchrotron Radiation Facility (ESRF). …………...……………….29
Figure 3.7 a) Photograph and dimension of the electrochemistry chips with enlarged big
E-chip, in comparison to 1 euro cent. b) Chip design and three-electrode system applied on
141
Appendix
the big E-chip, showing the glassy carbon working electrode in the center, surrounded by the
Pt reference electrode (grey) and the Pt counter electrode (light grey). c) Assembling of the
liquid E Chem holder with gasket, big/small E-chips and lid. d) Side view of the chip with the
corresponding bottom chip and the electrolyte layer between the chips. e) ex situ cell with
same chip assemble, flow mode and electrochemical connection. f) Gamry potentiostat for the
liquid E Chem holder and ex situ ce . Image a e a y ad ed f m “W kf w & T ai i g
P eid Se ec Ve i 1.2; C y igh 2017, P chi , I c.” wi h e mi i f m
Protochips. …………………………………………………………………………………...30
Figure 3.8 Schematic DEMS setup - custom-made TU Berlin electrochemical capillary
DEMS flow cell. ……………………………………………………………………………..33
Figure 3.9 Schematic view and photography of feed gas control system at DEMS setup -
saturation of electrolyte in the flow stream. ………………………………………………….33
Figure 3.10 A typical mass spectrum of ethylene gas directly dissolved in 0.1 M KHCO3 and
measured at DEMS setup. ……………………………………………………………………35
Figure 3.11 Photograph of operando XAS test station at Bessy. ……………………………40
Figure 4.1 Physiochemical characterization of as-prepared CuOx NPs. ……………………..45
Figure 4.2 Catalytic activity represented by liner sweep voltammograms taken at 5 mV/s in
CO2-saturated 0.1 M KHCO3 for 4 µg cm-2, 15 µg cm-2, 31 µg cm-2 CuOx NPs areal densities,
catalyst layers were supported on glassy carbon. ……………………………………………46
Figure 4.3 Catalyst mass-normalized absolute product formation rates of major gaseous
products as a function of applied electrode potentials during CO2 electroreduction in
CO2-saturated 0.1 M KHCO3 at CuOx nanoparticle catalyst areal densities. ……………..…48
Figure 4.4 Faradaic efficiency for each product as a function of areal density at various
applied overpotentials. Faradaic efficiency of CO2 electrochemical reduction over a) 4 µg
cm-2, b) 15 µg cm-2, c) 31 µg cm-2. d) The Faradaic efficiency ratios of C2H4/CH4 versus
overpotential for the different areal densities. ……………………………………………….49
Figure 4.5 a) Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NPs in CO2-saturated 0.1 M KHCO3 at -1.0 VRHE. b) Partial current densities of CO,
CH4, C2H4. c) Absolute product formation rates of gaseous products over time. d) The
consumption ratio of *CO for CO, CH4, C2H4, respectively. Catalyst areal density: 31 µg
cm-2……….....................................................................................................………………..50
Figure 4.6 a) TEM image of CuOx nanoparticles after the chronoamperometric stability test at
-1.0 VRHE from Figure 4.5. b) The mean particle size after 400 min significantly increased
compared to Figure 4.1a. Catalyst areal density: 31 µg cm-2. ……………………………….53
Scheme 4.1 The favored pathway for *CO in low/high areal NP densities condition. In low
areal particle density, *CO tend to leave as CO(g); while in high areal NP density, the
142
Appendix
desorbed CO(g) molecules are more likely to re-adsorb on the surface of nearby Cu
particles……………………………………………………………………………….………52
Figure 5.1 Morphological, structural, elemental and physical characterizations of CuOx NPs
synthesized in this study. …………………………………………………………………….59
Figure 5.2 Catalytic performance detected by on-line GC in H-cell. ………………………..63
Figure 5.3 The possible carbon origins in pure feeds and co-feed during our DEMS
experiments. In co-feed condition, carbon origins c u d g h ugh “CO2-CO2” c mbi a i ,
“CO-CO” c mbi a i a d/ c ed “CO2-CO”c mbi a i a hway C2H4. ………….65
Figure 5.4 Quantitative deconvolution of relative contributions of competing CO-CO
dimerization reaction pathways to ethylene using a novel operando DEMS capillary flow cell
system. Differential electrochemical mass spectrometry (DEMS) sweep data obtained during
COx reduction a CuOx NP catalysts (supported on a flat 0.785 cm2 glassy carbon electrode)
using CO2/CO co-feeds, pure CO2 and pure CO feeds during a cyclic voltammetric scan at
5mV s-1 and electrolyte flow of 5 µL s-1. …………………………………………………….67
Figure 5.5 Mechanistic hypotheses of enhanced ethylene production under CO2/CO reactant
co-feeds. ……………………………………………………………………………………..71
Figure 5.6 The realization of enhanced ethylene production using internal CO2/CO “ e f
co-feedi g” u i g non-metallic/metallic tandem catalyst design. …………………………..73
Figure 5.7 The realization of enhanced ethylene/methane production using internal CO2/CO
“ e f c -feedi g” u i g metallic/metallic tandem catalyst design. a) The catalyst is combining
Ag material as a local CO-producer and CuOx NPs on glassy carbon electrode. b) C2H4 and
CH4 production rate with CuOx-NiNC catalyst for CO2RR at the various component and fixed
overpotentials. Error bars are given as standard error of mean. …… ……………………….74
Figure 6.1 Morphological, structural characterizations of sheet-like CuOx nanoparticles
synthesized in this study. …………………………………………………………………….80
Figure 6.2 Electrochemical Performance of CO2RR with as-Prepared CuOx NS in H-cell
configuration. ………………………………………………………………………………...82
Figure 6.3 Catalyst evolution during CO2RR characterized by operando XAS. ……………84
Figure 6.4 Catalyst shape evolution detected by in situ WAXS. …………………………….86
Figure 6.5 Products formation and onset potential shift along electroreduction of CuOx NS
catalyst. ………………………………………………………………………………………87
Figure 7.1 Electrochemical characterization in micro Flow-Cell configuration. ……………92
Figure 7.2 SEM image of CuOx NS GDE after CO2 electrolysis. …………………………...93
Figure 7.3 a) Gaseous products efficiencies with 300 mA cm-2 in CO2 electrolyzer flow
cell. …………………………………………………………………………………………..94
143
Appendix
Figure 7.4 Quasi in situ demonstration of dendrites growth after reduction of CuOx NS
catalyst in E Chem holder…………………………………………………………………… 95
Figure 8.1 Graphical summary of topics and issues discussed and investigated in first part of
work, including areal density control, mechanistic co-feed study, and tandem catalyst.
(Reprinted with permission from The John Wiley and Sons). . ……………………………..98
Figure 8.2 Graphical summary of topics and issues discussed and investigated in second part
of this work, including shaped CuOx nanoparticles synthesis, characterization, H-cell tests,
combination of operando studies, and flow cell test at industrial level. ……………………100
Figure A1.1 a) b) Representative TEM images of CuOx nanoparticles after 7 days in hexane.
The inset shows the size distribution of CuOx nanoparticles. ………………………………113
Figure A1.2 a) Diffraction pattern of as-prepared CuOx NPs. Also indicated are diffraction
patterns of metallic Cu (red, PDF#98-000-0172), CuO (blue, PDF#01-078-0428), and Cu2O
(green, PDF#98-000-0186). ………………………………………………………………..113
Figure A1.3 The crystal structures of a) metallic Cu, b) cubic-Cu2O (cuprite), c)
monoclinic-CuO (Tenorite). Color legend: blue spheres, Cu; red sphere, O. ……………...114
Figure A1.4 The total Faradaic efficiencies of gaseous products as a function of areal density
at various applied overpotentials. Faradaic efficiencies of CO2RR and HER over a) 4 µg cm-2,
b) 15 µg cm-2, c) 31 µg cm-2. ……………………………………………………………….114
Figure A1.5 The Faradaic efficiencies for all detected products on CuOx NPs after 400 min
with the particle areal density of 31 µg cm-2 at -1.0VRHE. ………………………………….115
Figure A1.6 a) Total Faradaic efficiency for gaseous products as a function of areal particle
density. b) Trends in C2H4/CH4 production ratio at varying catalyst areal densities at -1.01
VRHE applied electrode potential. Potentials are IR corrected. ……………………………..115
Figure A1.7 Electrochemical CO2 reduction activity (left axis of ordinates) and selectivity
(right axis of ordinates) of glassy carbon as blank at various overpotentials. The amount of H2
at -0.66 VRHE is not detectable (under TCD detecting limitation). …………………………116
Figure A2.1 Time-dependent absolute product formation rates on H2. …………………….117
Figure A2.2 Liquid products analysis after 4-hour measurements in various feeds conditions.
Absolute product formation rates of detected liquid products on CuOx NPs with various
CO2/CO ratios. ……………………………………………………………………………...118
Figure A2.3 Formate analysis after 4-hour measurements in various feeds conditions. …...118
Figure A2.4 The potential-dependent tests of CORR, CO2RR and co-deed reduction (CO2:
CO 1:1). …………..…………………………………………………………………………119
Figure A2.5 XRD profiles of as-prepared working electrodes tested before and after
electrochemical reductions. …………………………………………………………………120
144
Appendix
Figure A2.6 SEM images of as-prepared working electrodes after electrochemical
reductions. ………………………………………………………………………………….121
Figure A2.7 High angle annular dark field (HADDF) images and EDX (Energy dispersive
X-ray spectroscopy) line scan of the CuOx NPs after co-feed (1:1)
educ i ...………………….……………………………………………………………….121
Figure A2.8 Investigation of pure feed reductions of CORR and CO2RR. ………………..122
Figure A2.9 The ratios of reaction between the three mechanisms. The ratio of reaction
between the three mechanisms on a) CO-CO, CO2-CO2, and b) CO2-CO, with respective
potential range for integration at 0.05 V from cathodic to anodic sweep of
e ia .. ………………….……………………………………………………………….123
Figure A2.10 Decomposition of the three pathways calculated by the co-feed system of a
mixture of carbon isotopes….……………………………………………………………….123
Figure A2.11 Differential electrochemical mass spectrometry data obtained on 0.785 cm2
large glassy carbon electrode drop-coated with CuOx NPs catalysts and using pure CO, pure
CO2, and CO2-CO co-feed. ………………………………………………………………….124
Figure A2.12 SEM images of as-prepared the bifunctional hybrid catalyst. SEM images of
CuOx NPs (50 𝝁g cm-2)-NiNC (1:4) tandem catalyst with carbon paper as working electrode
a) SEI mode; b) COMPO mode. ………..……………………………………………..….125
Figure A2.13 Catalytic performance of the tandem CuOx-NiNC catalysts detected by on-line
GC in H-cell. Absolute product formation rates on a) CO and b) CH4 with different catalyst
loading variation at -0.84 VRHE and -0.9 VRHE. ……………………………………………..125
Figure A2.14 The long stability test of tandem CuOx-Ag (1:3) catalyst at constant electrode
potential of -1.0 VRHE in CO2- a u a ed 0.1 M KHCO3. ……………………………………126
Figure A3.1 SEM images of Cu(OH)2 intermediate formed during synthesis process. ……127
Figure A3.2 a) Cross-section TEM images of as-prepared CuOx NS and b) corresponding
intensity of SAED intensity. ………………………………………………………………..127
Figure A3.3 Crystal structure of monoclinic CuO with indexing plane. …………………...128
Figure A3.4 a) Partial contour plot of CuOx NS. b) Azimuthally integrated line profiles of
as-prepared CuOx NS. ………………………………………………………………………128
Figure A3.5 Faradaic efficiencies of main products distribution after CO2RR in H-cell
design. ………………………………………………………………………………………129
Figure A3.6 a) Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NS in CO2-saturated 0.1 M KHCO3 at -0.84 VRHE. SEM image of as-prepared CuOx
catalyst on GC electrode (100 g cm-2) before b) and c) after 20-hour
electrolysis. …………………………………………………..……………………...……...129
145
Appendix
Figure A3.7 a) Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NS in CO2-saturated 0.1 M KHCO3 at -0.76 VRHE. Production rates b) and Faradaic
efficiencies c) of main gaseous products over 1200 min.…………………………………..130
Figure A3.8 SEM image of 1-h Chronoamperometric performance of the CO2 reduction
reaction on CuOx NS in CO2-saturated 0.1M KHCO3 at -0.84 VRHE. ………………………130
Figure A3.9 Comparison of XANES at the Cu K-edge of references powders CuO, Cu(OH)2
and the CuOx NS film in the dry state, and at OCP in 0.1 M KHCO3 at pH 6.8……………131
Figure A3.10 a) FT of k3 weighted EXAFS a) and b) EXAFS at the Cu K-edge of Cu metal
foil, Cu(OH)2 powder, CuO powder, and the CuOx NS film at OCP in 0.1 M KHCO3 at pH
6.8. …………………………………………………………………………………………..131
Figure A3.11. a) FT of k3 weighted EXAFS at the Cu K-edge of the CuOx NS film at OCP in
0.1 M KHCO3 at pH 6.8 and during 130 min CO2RR at -0.84 VRHE. b) EXAFS (k3 weighted)
at the Cu K-edge of the CuOx NS film at OCP in 0.1 M KHCO3 at pH 6.8 and during 130 min
CO2RR at -0.84 VRHE. ………………………………………………………………….…132
Figure A3.12 EXAFS distances of the first Cu-O coordination sphere and the first
intermetallic Cu-Cu shell. …………………………………………………………………..133
Figure A4.1 The partial current density of C2+ products as a function of applied geometric
current density for CuOx NS in in 1 M KHCO3. ……………………………………………137
Figure A4.2 Chronoamperometric performance stability of the CO2 reduction reaction on
CuOx NS in CO2-saturated 0.1M KHCO3 at -1.0 VRHE, performed in an H-call
configuration. ……………………………………………………………………………….138
Figure A4.3 SEM image of as-prepared CuOx catalyst after H-cell test. …………………..138
Figure A4.4 Morphology evolution of CuOx NS……………………………..……………..139
146
Appendix
Table of tables
Table 3.1 Overview of different materials used in this work with the corresponding chapters
in which their characterization is discussed. …………………………………………………14
Table 3.2 Overview of Methods for physicochemical characterization used in this thesis with
reference to the sections in which they are described……………………………………….. 16
Table 3.3 Summary of the detectable products in the H2O/CO2RR electrolysis and the
corresponding detection methods. …………………………………………………………...24
Table 3.4 All main possible fragments of ethylene including CO and CO2 species. ………..36
Table 3.5 Resume of the main ethylene fragments used for the DEMS based deconvolution of
the contribution of each reaction pathways with relative intensity (r.i.) and fragment
assignments (Assgn.) of the three *CO dimerization mechanisms resulting in ethylene
formation (C2H4, 13CH2CH2 and 13CH213CH2). The three mechanisms are denoted as
“CO2-CO2” if b h ca b a m f e hy e e de i e f m 12CO2, ”CO2-CO” i ca e f mixed
origin and “CO-CO” if b h ca b a m de i e f m 13CO. ………………………………37
Table 5.1 Gas flow control of various feeds and their relevant ratios of dissolved COx species
acc di g He y’ aw. ……………………………………………………………………61
Table 5.2 Electrolyte flow control of various feeds and their relevant ratios of dissolved COx
species in DEMS setup. ……………………………………………………………………...61
Table 5.3 Comparison of DEMS-derived kinetic reaction parameters of the three CO
dimerization pathways to ethylene depending on the chemical origin of the CO. Cross
coupling (CO2-CO) pathway: The dimerizing CO intermediates derive from CO2 and CO;
(CO2-CO2) pathway and (CO-CO) pathway: CO originated only from CO2 or CO,
respectively; iMS (mA cm-2 s-1) is the respective maximum C2H3+ion current , QMS (As) is the
DEMS charge under the deconvoluted C2H3+ion current sweeps under CO2/CO co-feeding in
cathodic and anodic scan direction. “Tafe ” de e he ex e ime a y a a e (i ma
current derived) Tafel slopes (VRHE dec-1), while E (in units of VRHE) is the experimental onset
potential of ethylene formation. ……………………………………………………………...70
Table A2.1 Liquid products of tandem CuOx-NiNC catalysts during CO2RR after 90 min at
fixed overpotential…………………………………..………………………………………126
Table A3.1 Cu K-edge simulation parameters of the FT-EXAFS oscillations of as-prepared
CuOx NS film. ………………………………………………………………………………134
147
Appendix
List of Abbreviations
CO2RR
(Electrochemical) CO2 reduction reaction
HER
Hydrogen redution reaction
atm
atmospheres
XRD
X-ray diffraction
M-N-C
Metal-Nitrogen-Carbon
OER
Oxygen evolution reaction
OD-Cu
Oxide-derived copper
ICP-OES
Inductively Coupled Plasma
Optical Emission Spectrometry
CE
Counter electrode
WE
Working electrode
RE
Reference electrode
NHE
Normal hydrogen electrode
RHE
Reversible hydrogen electrode
LSV
Linear sweep voltammertry
CA
Chronoamperometry
GDE / L
Gas diffusion electrode / layer
DFT
Density functional theory
MEA
Membrane electrode assembly
h
hours
min
minutes
DEMS
Differential electrochemical mass
spectrometry
WAXS
Wide-angle X-ray scattering
AW
Acid washing
HT
Heat treating
EDX
Energy dispersive X-ray spectroscopy
PEIS
Potentiostatic electrochemical
impedance spectroscopy
TEM
Transmission electron microscopy
HRTEM
High resolution TEM
PTFE
Polytetrafluoroethene
XAS
X-ray absorption spectroscopy
OCP
Open circuit potential
EXAFS
Extended X-ray absorption fine
structure
GC
Gas chromatograph / Glassy carbon
HAADF
High-angle annular dark field
PCET
Proton coupled electron transfer
HE-XRD
High-energy XRD
HPLC
High performance liquid chromatograph
MFC
Micro flow cell
SEM
Scanning electron microscopy
RT
Room temperature
iR
Current times resistance (ohmic drop)
FE
Faradaic efficiency
TCD
Thermal conductivity detector
FID
Flame ionization detector
148
Appendix
ER
Eley-Rideal
SAED
Selected area electron
diffraction
XANES
X-ray absorption near edge spectra
GI-WAXS
Grazing incidence wide-angle
X-ray scattering
NPs
Nanoparticles
NS
Nanosheet
GDE
Micro-flow-cell
PTFE
Polytetrafluorethylen
Echem
Electrochemistry
149
Appendix
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.
Ketjen EC 600JD
-
-
AzkoNobel
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
H2
99.999 %
Air Liquide
Carbon monoxide
CO
99.997 %
Air Liquide
Nitrogen
N2
99.999 %
Air Liquide
150
Appendix
Oxygen
O2
99.998 %
Air Liquide
Argon
Ar
99.999 %
Air Liquide
Helium
He
99.999 %
Air Liquide
Trioctylphosphine oxide
TOPO
90%
Sigma Aldrich
Borane tert-butylamine complex
TBAB
97%
Sigma Aldrich
Oleylamine
OA
90%
Sigma Aldrich
N, N-dimethylformamide
DMF
99%
Sigma Aldrich
Cuprous Bromide
CuBr
99%
Sigma Aldrich
Cupric Acetate
Cu(Ac)2
99%
Sigma Aldrich
151
Appendix
List of Publications during Ph.D. Study
Catalyst Particle Density Controls Hydrocarbon Product Selectivity in CO2
Electroreduction on CuOx. X. Wang, A. S. Varela, A. Bergmann, S. Kuehl, P.
Strasser. ChemSusChem, 2017, 10, 4642–4649.
Mechanistic reaction pathways of enhanced ethylene yields during electroreduction of
CO2–CO co-feeds on Cu and Cu-tandem electrocatalysts X. Wang, J. Araújo, W. Ju,
A. Bagger, H. Schmies, S. Kuehl, J. Rossmeisl, P. Strasser. Nat. Nanotechnol., 2019,
14, 1063-1070.
Sheet-like Copper Oxides with Stable and Selective Ethylene Production for Direct
CO2 Electroreduction from H-cell to Flow Cell. X. Wang, K. Klingan, T. Möller, J.
Araújo, S. Jiang, H. Dau, P. Strasser. In Preparation for submission.
Tracking Copper Oxide Dendrite Catalyst Growth during CO2 Electroreduction in a
Liquid Electrochemical TEM Cell. X. Wang, T. Möller, M. Klingenhof, S. Kuehl, P.
Strasser. In Preparation for submission.
Efficient CO2 to CO Electrolysis on Solid Ni-N-C Catalysts at Industrial Current
Densities, T. Moeller, W. Ju, A. Bagger, X. Wang, F. Luo, T. N. Thanh, A. Varela, J.
Rossmeisl, P. Strasser, Energy and Environmental Science, 2019, 12,640-647.
Alloy Nanocatalysts for the Electrochemical Oxygen Reduction (ORR) and the Direct
Electrochemical Carbon Dioxide Reduction Reaction (CO2RR), C. Kim, F. Dionigi,
V. Beermann, X. Wang, T. Möller, P. Strasser, Adv. Mater. 2018, 1805617.
