design and implementation of an automated electrochemical ... · design and implementation of an...

Design and implementation of an automated electrochemical

flow system coupled with mass spectrometry for investigation

of the dissolution behavior of platinum

Dissertation

zur

Erlangung des Grades

“Doktor der Naturwissenschaften”

an der Fakultät für Chemie und Biochemie

der Ruhr-Universität Bochum

vorgelegt von

Angel Angelov Topalov

aus Ruse, Bulgarien

(geboren in Sankt-Petersburg, Russland)

Bochum 2014

1. Gutachter: Prof. Dr. Martin Stratmann

2. Gutachter: Prof. Dr. Wolfgang Schuhmann

Tag der Einreichnung: 24.01.2014

Tag der Disputation: 04.03.2014

Gewidmet meiner lieben Familie

(�� )

i

Acknowledgement The work described in this thesis was carried out at the Max-Planck-Institute für Eisenforschung

GmbH in Düsseldorf with financial support of Center for Electrochemical Sciences, Bochum.

I would like to express my deepest gratitude to Prof. Dr. Martin Stratmann for supervising

my thesis and his scientific contribution during the annual gatherings of the department. His

critics have strongly motivated me and helped me to find and improve crucial aspects. I

furthermore thank Prof. Dr. Wolfgang Schuhmann for kindly accepting to be second

reviewer and for the time he invested. I appreciate his mentor position during my membership

in CES.

I would like to acknowledge Dr. Karl Mayrhofer, first of all for choosing me as one of the

first PhD students in his starting group and introducing me to the basic concepts of

electrochemistry, the time and patience he spent in several discussions. Especially for the

freedom and trust he provided me over those great three years in the daily work, tinkering in

the labs on the weekends and in many general aspects. I’m grateful for this opportunity, where

he supported me to develop myself as well as in professional and in social fields. Thank you

Karl!

I would like to thank “the core” of the Electrocatalysis Group, Josef Meier, Sebastian

Klemm and Ioannis Katsounaros for a nice working time, their support in building up the

labs of the group in the beginning from ground zero, and for the many nice evenings we spend

together at the Greek restaurant. It is my pleasure to work with you guys!

I would like to thank Andrea Mingers and Jörg Puszcz for taking care of general technical

issues and troubleshooting of ICP-MS, for a smooth daily schedule. I would like to thank also

all members of the Mechanical Workshop for their quick and precise work. A big

acknowledgement is dedicated to all formal and current members of the Electrocatalysis Group

for the nice working atmosphere. Special thanks to Aleksandar Zeradjanin, for helping me to

understand many fundamental issues and being a constructive discussion partner. I really

appreciate the time and ideas he shared with me. X�� !�!

I would like to thanks several people for their help in different experimental issues: Serhij

Cherevko – assisting in the reconfiguration of SFC system for temperature dependent

measurements; Ashokanand Vimalanandan - SKP measurements; Adnan Sarfraz &

Andreas Erbe – in situ Raman and in-situ Elipsometry measurements; Sergiy Borodin & Julia

Klemm – XPS measurements and data evaluation.

ii

Last but not least. My family, I’m proud of having such great parents Angel and Ludmila

Topalovi that provided me wonderful childhood and a lot of possibilities in my life, and

showed me that there are always alternative views. My wife Slaveya Topalova and our little

son Angel, who supported me all the way to the end of this work! Everything that I achieve in

my life wouldn’t be possible without the strong support of my lovely family. Thank you!

Translation of the last paragraph in Bulgarian:

“Последно, но не на последно място. Искам да блогадаря на семейството ми, аз съм

горд да имам такива невероятни родители Ангел и Людмила Топалови, които ми

предоставиха прекрасно детство и много възможности в живота ми и ми показа, че

винаги има алтернативна страна на живота. Същото така искам да благодаря на

съпругарата ми Славея Топалова и нашият син Ангел, които ме подкрепяха през

цялото време до края на тази дисертация! Всичко, което постигнам в живота ми не би

било възможно без силната подкрепа на прекрасното ми семейство. Благодаря ви!”

iv

Abstract The main subject of this thesis is the investigation of the dissolution behavior of

polycrystalline platinum during an electrochemical treatment by utilizing simultaneous online

elemental analysis of the electrolyte. For this purpose, a novel micro-electrochemical scanning

flow cell (SFC) based on the concept of a channel electrode combined with inductively coupled

plasma – mass spectrometer (ICP-MS) has been designed and developed. The developed

system is completely automated to enable high-throughput and combinatorial measurements.

This has been achieved by creating a universal-modular software approach that allows parallel

asynchronous control and acquisition within a single controlling program. A proof of concept

of the new combined system is presented, as well as a long-term performance test showcasing

the reliability of the approach. The created software architecture has been adopted for the

automation of up to now five rotating disc electrode setups and provides a solid backbone

solution for the further development of currently seven SFC systems.

The advantages provided by parallel monitoring of the dissolved species by post analysis

under automated control are used for the systematic investigation of the influence of several

experimental parameters on the dissolution of polycrystalline Pt. The performance of the

coupled SFC/ICP-MS is first optimized to achieve very low effective detection limits below

10 ppt, which enables a precise description of the dissolved amounts in the sub-monolayer

region. Thus for the first time a time resolved Pt concentration profile within a single cyclic

voltammogram is presented. The extensive experimental data explicitly shows that dissolution

appears during the sub-surface oxide formation at ca. 1.1 VRHE as well as during the reduction

of the oxide layer, where irreversible roughening of the surface acts as a necessary precursor to

trigger dissolution. These two different cases of dissolution have been specified as (i) “anodic”

dissolution route with typical material loss in order of ca. 1 ng/cm2 (ca. 0.25% of a single

monolayer of platinum) almost independent of the overpotential for the oxide formation and

the time scale of the experiments, and (ii) “cathodic” dissolution route that represents the main

contribution to the overall dissolution amounts and shows a strong variation with pH, amount

of formed oxide, the experimental time scale and temperature. Within the framework of the

newly gained knowledge, the existing thermodynamic models of Pt dissolution are scrutinized

and new mechanism proposed.

Overall the work summarized in this thesis presents novel methodological developments

and scientific information, and has already been a solid basis for further studies especially of

dissolution processes. So the developed experimental approach has not only been applied for

v

platinum, which is the focus of this work, but has been also applied in several parallel

investigations on stability issues of other noble metals and multi component systems during

this thesis. The major result has been a deep insight into metal dissolution in general, often

even clarifying the mechanism, as well as broad and reliable quantitative datasets of dissolution

rates that provide valuable guidelines for the engineering limits of catalyst materials, such as for

the application of Pt-based materials in fuel cells.

vi

Content Acknowledgement ..................................................................................................... i�Abstract ..................................................................................................................... iv�Glossary ................................................................................................................. viii�1� Motivation – Platinum and its role in the renewable energy concept ................. 1�

1.1� Electrochemical conversion of energy from renewable sources ............................... 2�1.2� Stability – the challenge for Pt catalysts in electrochemical energy conversion ...... 3�

2� Importance, aim and general approach of the study ........................................... 5�2.1� Political-economical aspects and perspectives ............................................................. 5�2.2� Brief overview on platinum dissolution and open questions ..................................... 7�

3� Theoretical background and literature review ..................................................... 9�3.1� Electrode reactions ........................................................................................................... 9�3.2� Anodic corrosion of metals ........................................................................................... 10�3.3� Electrochemistry of platinum ....................................................................................... 12�

3.3.1� Oxide layer ............................................................................................................... 13�3.4� Dissolution of platinum ................................................................................................. 17�

3.4.1� Polycrystalline surfaces ........................................................................................... 17�3.4.2� High surface area Pt catalysts ................................................................................ 20�3.4.3� Structural changes ................................................................................................... 22�3.4.4� Theoretical model ................................................................................................... 23�

4� Technical background ....................................................................................... 25�4.1� Inductively Coupled Plasma - Mass Spectrometry (ICP-MS) .................................. 25�4.2� X-ray Photoelectron Spectroscopy .............................................................................. 27�4.3� Scanning Kelvin Probe microscopy ............................................................................. 27�

5� Automated system development – from design to implementation ................. 29�5.1� Hardware – between commercial and custom solutions .......................................... 30�

5.1.1� Rotating Disc Electrode (RDE) ............................................................................ 30�5.1.2� Scanning Flow Cell system (SFC) ......................................................................... 31�5.1.3� Inductively Coupled Plasma - Mass Spectrometry (ICP-MS) .......................... 34�

5.2� Software development .................................................................................................... 35�5.2.1� Software architecture – modular approach ......................................................... 35�5.2.2� User interface – friendly and efficient design ..................................................... 37�

vii

5.3� Data management ........................................................................................................... 39�5.3.1� Data storage .............................................................................................................. 39�5.3.2� Evaluation – quick data processing ...................................................................... 40�

6� System validation – proof of functionality and limits in application ................ 42�6.1� Automation - proof of concept .................................................................................... 42�6.2� Performance of electrochemical systems – RDE vs. SFC ........................................ 44�6.3� Inductively Coupled Plasma - Mass Spectrometry (ICP-MS) and coupling with the

SFC 46�7� Results and discussion ....................................................................................... 52�

7.1� Electrochemistry of platinum in acidic media ............................................................ 52�7.2� Influence of the overpotential for oxide formation and reduction on platinum

dissolution ........................................................................................................................................ 53�7.3� Steady state dissolution during chronoamperometry ................................................ 56�7.4� Decoupling the influence of the time scale of experiment and the amount of formed

oxide on the dissolution rates ....................................................................................................... 59�7.5� Influence of the concentration of the protons ........................................................... 62�7.6� Effect of the reactive gases on Pt dissolution ............................................................ 65�7.7� Enhancement of Pt dissolution in the presence of chlorides .................................. 68�7.8� Temperature dependence of the dissolution .............................................................. 70�7.9� Ex-situ XPS and SKP investigation of platinum oxide ............................................. 72�

8� Comprehensive discussion ................................................................................ 77�9� Summary and outlook ........................................................................................ 85�References ............................................................................................................... 87�

Appendix .................................................................................................................................. 100�Publications list: .................................................................................................................. 100�Oral presentations: ............................................................................................................. 101�Poster presentations ........................................................................................................... 102�

Curriculum Vitae – Angel A. Topalov ................................................................................. 104�

viii

Glossary Abbreviations

AFC Alkaline fuel cell

ARXPS Angle-resolved X-ray photoelectron spectroscopy

a.u. Arbitrary units

ddl Dynamic link library

ppb Particles per billion

ppm Particles per million

ppt Particles per trillion

ca. Circa (Latin: around)

CAD Computer aided design

CE Counter electrode

CER Chlorine evolution reaction

COM Component object model (Serial port)

CV Cyclic voltammogram

DAQ Digital acquisition

DMFC Direct methanol fuel cell

EEG “Erneuerbare Energie Gesetz” (renewable energy act)

EIS Electrochemical impedance spectroscopy

e.g. exempli gratia (Latin: for example)

et al. et alii (Latin: and others)

etc. et cetera (Latin: so on)

FC Fuel cell

FIFO First-in-first-out

FWHM Full width at half maximum

HFM High field model

i.e. id est (Latin: that is)

ICP-MS Inductively coupled plasma – mass spectroscopy

ICP-OES Inductively coupled plasma – optical emission spectroscopy

LSM Lanthanum Strontium Manganite

LSV Linear sweep voltammetry

MCFC Molten carbonate fuel cell

ix

MO Molecular orbital

n.a. Not available

NASA National Aeronautics and Space Administration

NGM Nucleation and growth mechanism

OCP Open circuit potential

OER Oxygen evolution reaction

ORR Oxygen reduction reaction

PAFC Phosphoric acid fuel cell

PC Personal computer

PDM Point defect model

PEM Place exchange mechanism

PEMFC Proton exchange membrane fuel cell

PGM Platinum group metals

RDE Rotating disc electrode

RHE Reversible hydrogen electrode

RE Reference electrode

RF Radio frequency

RRDE Rotating ring disc electrode

rpm Revolutions per minute

SA Specific activity

SCE Saturated calomel electrode

SECM Scanning electrochemical microscopy

SERS Surface-enhanced Raman spectroscopy

SFC Scanning flow cell

SHE Standard hydrogen electrode

SKP Scanning Kelvin Probe

SOFC Solid oxide fuel cell

UI User interface

UPD Under potential deposition

USB Universal serial bus

VI Virtual instrument

vs. versus

WE Working electrode

XPS X-ray photoelectron spectroscopy

x

YSZ Yttrium stabilized zirconia

Formula abbreviations

Ageo Geometric electrode surface area

Areal Real platinum surface area

C Concentration

D Diffusion coefficient

E Potential

Ea Activation energy

E� Standard potential

Eeq Equilibrium potential

F Faraday constant

G Gibbs energy

im Measured current

idiss Dissolution current

idl Diffusion limited current

ik Kinetic current

j Current density

k Reaction rate

m Mass

M Molar mass

n Number of electrons

QCO Charge determined for CO-stripping

QH Charge determined from HUPD

R Ohmic resistance

t Time

� Scan rate

V Volume velocity

Greek letters

� Charge transfer coefficient

Diffusion layer thickness

� Potential

� Overpotential

xi

� Coverage

� Work function

Chapter 1: Motivation –Platinum and its role in the renewable energy concept

1

1 Motivation – Platinum and its role in the renewable energy concept

The noble metals have fascinated mankind throughout its entire known history. The

ancient cultures like Thracians or/and Mayan were utilizing gold ore for utensils and jewelers

hundreds of years B.C., with the main role to emphasize a social status. The other metals in the

platinum group were rarely present by the ancient cultures [1]. This can be attributed to their

extremely low abundance in terms of amount and geographic distribution and the fact that they

usually appear in nature as alloys. Over the last decades with the development of natural

sciences and industry, the focus of the usage has significantly shifted from esthetic to practical

one, determined predominantly by their physical/chemical properties. In the 17th century, when

platinum was re-discovered by the Spanish conquerors in Colombia, it was considered as a

parasitic product of the gold mining. It acquires the name platinum from the Spanish word

Platina i.e. little silver. The high melting temperature and corrosion resistance make it a very

interesting material for the glass industry. The first real breakthrough in recognizing the

significance of platinum was due to its catalytic properties, e.g. for reforming of crude

petroleum. This was possible due to the separation methodology of platinum ore developed by

the Russian scientist P. G. Sobolevsky, which was later on adapted on large scale [1]. This

scientific/industrial achievement opened a new page in the history of platinum mining and

created the base for the current commercial purification technologies in metallurgy of noble

metals.

Another important historical episode in recognition of the catalytic properties of platinum

was the ability of platinum to promote hydrogen oxidation. This discovery was attributed to the

German scientist J. W. Döbereiner, who showed the spontaneous ignition of gas mixture of

hydrogen and air in presence of platinum [2]. Further investigation of the British scientist

W. R. Grove emphasized the potential of platinum in a so-called gas voltaic battery [3],

nowadays known as a fuel cell. Sir Grove separated the process into two half-cell reactions,

where the electrons are diverted through an external electric circuit. In this way, he performed

energy conversion, i.e. from chemical to electric energy, and showed the basic principles of the

operation of modern hydrogen-based fuel cells.

Chapter 1: Motivation

2

1.1 Electrochemical conversion of energy from renewable sources

The 21th century is marked by the new way of usage of resources especially electrical energy.

Germany as the seventh biggest producer and consumer of electricity in the world (and leader

in the European Union) has a pioneer role in the integration of alternative energy sources [4].

In the year 2000, a new national regulation for renewable energies sources of electricity (EEG)

with the target of replacing/reducing the power supply from the nuclear plants and combustion

of fossil fuels by the year 2050 was approved. This political act led to a rush in the deployment

of renewable energy sources, mainly solar and wind parks that covered more than half of the

overall market of renewable sources by 2012. The industrial upscaling of those technologies

proved their capabilities in supplying hundreds to thousands of megawatts. Installed wind

power sources in Germany by 2012 reached above 31 GW what could provide in theory the

minimum energy consumption [5] (for comparison the biggest German nuclear power plant

Asar-2 has ca. 1.4 GW power). Nevertheless only ca. 8.1 GW of the wind power was effectively

used in 2012 on average [5]. Unfortunately, the integration of the ‘green’ energy sources into

the national power grids and maintaining constant supply turned out to be a serious challenge,

due to the volatility caused by the unavoidable fluctuating meteorological conditions. This is an

issue that can be alleviated only with intermediate buffering of the overproduction and reuse in

case of shortage. Energy storage in chemical bonds is proven to be a concept that can provide

the required capacity on different scales of magnitude. A perspective option is the usage of

hydrogen as an energy carrier with theoretical capacity up to 120 MJ/kg, produced for instance

by water electrolysis [6]. The backwards conversion to electrical energy, can be performed for

instance by separated anodic hydrogen oxidation and cathodic oxygen reaction reactions in

proton exchange membrane fuel cells (PEMFC).

The fuel cell technology is known as a concept from the beginning of the 19th century,

nevertheless the first publically known application is recorded one century later in the space

programs of NASA [3,7]. Records on the usage of FCs in Soviet Union are not publically

available due to their possible military application in submarines. The family of the FCs is

broad with respect to the nature of the charge carrier, operation temperature, the type of

electrolyte etc. In the last decades, the FCs technologies were boosted in several directions.

Table 1.1 represents only some of the commercially available systems. One of the most

Chapter 1: Motivation –Platinum and its role in the renewable energy concept

3

promising technologies is PEMFC. Due to its potential to replace combustion engine in mobile

application, a lot of research was focus in the optimization of the performance.

PEMFC DMFC PAFC AFC MCFC SOFC

Fuel H2 CH3OH H2 H2 H2 H2

Electrolyte

Hydrated Sulfonated

Organic Polymer

Hydrated Sulfonated

Organic Polymer

Phosphoric acid

Potassium Hydroxide

Molten Lithium/ Potassium Carbonate

Yttria- stabilized Zirconia

Charge

carrier H+ H+ H+ OH- CO3

2- O2-

Cathode Pt/FeNxCy PtxRuy Pt NiO NiO LSM

Anode Pt PtxRuy Pt Ni/Precious

metals NiO Ni/YSZ

Operation

temperature 70-100°C 90°C 160-220°C

25-75°C

100-250°C 660°C

800-

1000°C

Table 1.1 Summary of the differences of some of the fuel cells (data adopted from ref. [7])

1.2 Stability – the challenge for Pt catalysts in electrochemical energy conversion

Noble metals, and especially platinum play a key role in the design of electrocatalysts for

energy conversion. The main drawback for the deployment of PEMFC on a large scale is the

high cost of the systems, where one of the main contributors is the price of the catalyst. In

order to compensate the relatively high costs related to catalysis, significant improvements in

performance become necessary. Many research studies have shown ways of improving the

catalyst activity without sufficient consideration of stability, which lead to the development of

many catalyst concepts without industrial relevance. Attempts to replace the noble materials, in

particular platinum, with cheaper alternatives are still ongoing research topics [8–10]. While a

significant improvement has been achieved in tuning the activity of non-noble catalysts, the

long term performance still remains a challenge [11,12]. The state-of-the-art materials used in

commercial PEMFC stacks are therefore platinum-based, which themself suffer from

Chapter 1: Motivation

4

degradation over the extended operation times under applied conditions [13–15]. In order to

render PEMFC technology in electric vehicles viable, an understanding of the underlying

degradation mechanism becomes necessary and these stability issues have to be considered in

the design and development stage of the catalyst.

Recently several groups illustrated the complexity of this problem utilizing in situ electron

microscopy with electrochemical techniques [16–19]. Degradation tests on carbon supported

platinum nano-particles, the typical catalyst for the ORR in low-temperature fuel cells, showed

that different mechanism lead to the loss of active surface area, as summarized in figure 1.1. A

smart design approach has been developed over the last years to inhibit the mechanical

detachment, prevent migration of particles by incorporation in mesoporous structures and also

to reduce the degradation rate of the carbon support by high graphitization [A.10, A.17]. One

of important issues remaining is the dissolution of the catalyst during operation, which appears

to be unavoidable. This leads to the questions: a) what interfacial processes really describe the

Pt dissolution mechanism and b) is it possible to protect Pt during relevant operational

conditions.

Figure 1.1 Illustration of the four major degradation mechanisms of nano-particulated

platinum catalyst (Modified from its original version in ACS Catalysis from ref. [A.5]).

Detachment Support corrosion Dissolution Coalescence

Chapter 2: Importance, aim and general approach of the study

5

2 Importance, aim and general approach of the study

2.1 Political-economical aspects and perspectives

The importance of noble materials is determined mainly by their industrial applications. In

particular, platinum can be found in several fields [20] despite being an extremely rare metal

with crustal abundance of only ca. 3 ppb. Mining sources with sufficient amounts of platinum,

relevant for commercial exploitation, can be found only in two geographic regions in the world,

namely in the border of Zimbabwe/South Africa and several locations in Russia, held

respectively by the Anglo American Holding and Norilsk Nickel [21]. More than 38% of the overall

produced platinum in 2012 was used in the automotive industry in catalytic convertors, with an

average amount of 3-7 g/car [22]. This huge market for platinum was secured by two political

events ‘Clean Air Act’ in the US and the consequent EU derivatives 91/441/EEC &

93/59/EEC for emission standards for vehicles, both triggered officially by environmental

issues. Similar political decisions were made subsequently by a number of countries (Russia,

China, India etc.), which increased the international demand drastically. Moreover platinum

finds also application in several other fields like a) various industrial processes such as the

refinement of petroleum to obtain high octane gasoline or primary feedstocks for polymers,

glass production, hard disc manufacturing, production of nitric acid, fertilizers etc. b) medical

and high tech applications: antitumor drugs, dental restorations sensors etc. c) investment and

jewelry [20]. All this insured the stable market of platinum with the consumption of

approximately 228 tons per year over the last decade. In figure 2.1 are illustrated produced and

consumed amounts of platinum only in 2012, since the demand over the last 6 years is

maintained practically on the same level.


6

Figure 2.1 a) Demand of platinum by application in different industrial branches; b) platinum

supply by producing countries; the data is adapted from Johnson & Matthey annual report

2013 [20].

In contrast to the 19th century, where the economical aspects were dominating

developments, nowadays, sustainability is of crucial importance. It means that symbiosis

between the economical, social and environmental criteria is required to prevent disposable

usage of critical resources and ecological disasters (like Chernobyl, Fukoshima etc.). This trend

can be observed even from figure 2.1, where a significant gap of ca. 63.22 tons per year

between supply and demand is filled by the recycling sector, utilizing old jewelry and car

exhaust catalysts. Figure 2.2 shows the price development of platinum over the last 20 years.

The general tendency in the market indicates a rising price with casual fluctuation, mostly due

to political and economical decisions. Considering the continuously growing price of all PGM

materials, only smart usage can justify their application in future.

Figure 2.2 Chart of the platinum market price over the last 20 years, data adopted from

Johnson and Matthey [20].

�� '&� #�

�� !"�&%�� $�%&��%�#'��

!#�'(�

��&&�"$�

��

%�' �

��$�%&��!#�#"�

��

��

��

��

��

��


7

The unique catalytic properties and the promising application in PEMFCs will have a strong

influence on the overall Pt market. Thus, resolving the issues related with the dissolution

behavior of Pt is of crucial importance for the future of PEMFC technology, system design and

implementation. Moreover, stability limits of platinum will be one of the key parameters in

defining the framework of application also in other technologies. An improvement of the

performance of Pt demands an improved fundamental understanding of the ongoing

degradation phenomena. In particular, the comprehension of the dissolution phenomena on

model system like polycrystalline platinum is an important step in creating an in-depth picture

about the stability of real supported catalysts.

2.2 Brief overview on platinum dissolution and open questions

In 1905 Tafel et al. observed (re)deposition of platinum on the cathode during caffeine

reduction, concluding that platinum is a bad choice for the anode [23]. This was one of the first

reports on the instability of platinum during electrochemical treatment, and in fact implicitly

suggested an anodic route of the dissolution. Later on, in the search for materials for oxygen

evolution reaction during water electrolysis, Kolotkin and Chemodanov developed a

radiotracer method for the investigation of low dissolution rates of PGM elements [24], and

showed for the first time that the dissolution of platinum is determined by the applied anodic

potential. Other studies have shown that dissolution can appear during reduction of the Pt (–

oxide) surface, considering a cathodic route as main contribution for the degradation [25]. Over

the last decades, the opinions on the mechanism diverged, whether Pt dissolution is: a) a

competitive reaction to oxide formation b) intermediate product in the reduction of the oxide,

c) or predominantly a chemical process. Even for the “simplest” system, polycrystalline Pt,

there is no unified point of view on the mechanism.

Slightly better resolved is the quantification of Pt dissolution, however literature reports are

highly diverging and often not trustful due to missing experimental details. The analysis of the

representative literature shows that the expected dissolution amounts are in the range of some

ng/cm2. A simple order estimation for the expected charge balance, based on the assumption

of a valence of 2 [26] for the dissolved species gives us:

��

��

� ��

��


8

Considering that only the formation charge of a single monolayer of oxide is in the order of

hundreds of �C/cm2, it is obvious that dissolution cannot be resolved purely by measuring

currents at a working electrode during potentiodynamic experiments. Moreover, despite being

highly sensitive, other electrochemical approaches based on double electrodes suffer from

overlapping parasitic processes and are of limited use when multi-element analysis is demanded.

Thus other, complementary techniques with high sensitivity and a capability to analyze various

elements in parallel are required to quantify Pt and Pt-alloy dissolution. Probing amounts as

low as 1012 atoms/cm2 can be achieved for instance with multi-elemental analysis using

inductively coupled plasma mass spectrometer (ICP-MS). As there is no commercially available

system with the required capabilities for time-resolved dissolution monitoring, the first step

within this thesis is the design and development of an electrochemical system that can be

combined online with mass spectrometry. At the second stage of this work, a combinatorial

screening over the experimental parameters should be performed; where the dissolution

behavior of polycrystalline Pt will be revisited in time resolved in situ experiments. The

influence of the experimental parameters in correlation with the surface state will be discussed

so that a tentative dissolution mechanism can be proposed. Additionally, a quantitative

description of the dissolution rates is provided for several operation conditions, which is of

high relevance for engineers working with platinum based materials and can help to predict the

lifetime of catalyst/electrode materials on an industrial scale.

Chapter 3: Theoretical background and literature review

9

3 Theoretical background and literature review

This chapter is targeted to introduce the general aspects of electrochemistry on platinum

and to provide a framework of the existing knowledge on its stability. Platinum has been

intensively studied over the last century, thus the literature review is restricted to the most

relevant publications for this work.

3.1 Electrode reactions Electrochemical processes appearing on the electrode/electrolyte interface with transfer of

charged species can be described as oxidation and reduction reactions as shown in the

schematic energy diagram in figure 3.1. In order to trigger an electrochemical reaction, the

energy level � of the electrons at the electrode has to be higher than the lowest unoccupied

molecular orbital (LUMO) of the electro-active species in the solution for a cathodic reaction,

or to be lower than the highest occupied molecular orbital (HOMO) for anodic one [27].

Figure 3.1 Illustration of energy levels at the electrode/electrolyte interface and changes during

the anodic and cathodic polarizations in red and blue respectively.

This is not always the case for electrode immersed in solution at open circuit condition.

Thus its energy state is typically modified by applied external polarization using a second

electrode, where the potential difference is defined as:

��

That is further related to change of Gibb’s free energy of the system:


10

��

where � is number of electrons, � is Faraday’s constant (96485 As/mol). In this way, different

reactions can be driven to proceed by variation of potential �. Thermodynamically favorable

conditions do not always initiate a reaction due to so called activation barrier of the process.

Arrhenius related experimentally the rate constant k and the activation energy �� as [28]:

� � ��

where � is pre-exponential (frequency) factor, � is temperature. The rate of electrochemical

reaction is proportional to the flowing current that can be directly measured. Applying

overpotential to promote a reaction, affects its rates on different extend.

� � ��

This is attributed to kinetic hindering described by Buttler-Volmer equation [27]:

� � ��

��

The overall current response j is given as a function of the charge transfer coefficient �,

exchange current density �� and overpotential � � � � �� . The described Faradaic processes

can be separated in to two types: (i) continuous and (ii) adsorption processes. In first case, the

reaction can theoretically proceed infinitely, whereas the second one adsorption of the

electrochemically active species causes a blockage of electrode surface and consequent drop of

the reaction rate. The coverage � can be expressed by Langmuir isotherm:

� � ��

or ��

using the Nernst equation, the partial pressure �� of the adsorbed species is related to the

equilibrium potential.

��

Consequently, the coverage can be expressed as:

��

��

��

where �� is the adsorption constant and ��is standard potential.

3.2 Anodic corrosion of metals Dissolution of metals is a result of an electrochemical process, in which the metallic atoms

are released from the lattice of the substrate in form of ions. This phenomenon is part of the

broad field of corrosion. It is commonly accepted that such anodic processes involve


11

intermediate surface state between the metal and the electrolyte. There are several possible

pathways that describes it [29]:

(i) Aquo-ligand mechanism:

� � �� (4.1)

�� (4.2)

(ii) Hydroxo-ligand mechanism:

� � �� (4.3)

�� (4.4)

(iii) Aniono-ligand mechanism

� � �� (4.5)

�� (4.6)

Typically the active anodic corrosion takes a place without significant obstacles due to minor

interaction with passivation film. Thus the dissolution rate is usually determined by the counter

balancing cathodic reaction like hydrogen evolution reaction (HER) or oxygen reduction

reaction (ORR):

�� (4.7)

�� (4.8)

The metal dissolution does not necessary proceed as a continuous process. Formation of

MOHads has a dual role, namely as part of the dissolution reaction pathway (reaction 4.3 and

4.4), but also as a precursor for the passivation of the surface. The MOHads can react further by

de-protonation of the water forming M(OH)2 in a pre-passivation step [29,30]. The following

oxide formation usually gives a dense layer with limited ionic conductivity that significantly

hinders high initial dissolution rates as illustrated in figure 4.2 [31]. Typically the current

response in the passive region is related to the variation of the oxide dimensions and

independent of the applied potential. However, the thickness of the passive film spreads over a

broad range from a few nanometers (e.g. Pt) to several micrometers (e.g. Al). In this region,

metal dissolution is controlled by the chemical dissolution of the oxide film.

Further positive extension of the potential in the region of the OER causes breakdown of

the passivity and the metal dissolution can rise again. This is closely related with the initiation

of the oxygen evolution, especially when the oxide layer is actively participating in the reaction

[32,33].


12

Figure 4.2 Schematic representation of the current density-plot of dissolving metal divided in

to three parts, starting from active dissolution over the passivation and finishing with

transpassive region, where the main current response is result of the OER (image redrawn

from ref. [30]).

3.3 Electrochemistry of platinum Following the thermodynamic predictions for the potential-pH behavior in aqueous

environment, platinum is relatively inert over a broad region as shown in figure 3.1 [34]. The

blue parallelogram indicates the thermodynamic stability window of water, where Pt is mainly

present in metallic form. However, an increase of the potential leads to transition to Pt(II) and

Pt(IV) as illustrated by line 1 and 2 respectively. Platinum in this region is expected in soluble

form (Pt2+) only at highly acidic pH below 0, where the dissolution can proceed over direct

oxidation of the bulk material or chemical dissolution of hydroxide film (Pt(OH)2). For all

other pH values the formation of PtO2 is thermodynamically expected, and dissolution should

only proceed over intermediate oxygenated species of Pt(IV).

��

��

��

��

��

��

��

��

��

��

��

��

��


13

Figure 3.1 Pourbaix diagram of the platinum-water system at 25°C, adopted from ref. [34].

3.3.1 Oxide layer Experimentally, cyclic voltammetry is an excellent approach for an initial study of electrode

surfaces. Platinum is one of the best-known catalysts for HER [35,36], however the potential

region below 0 VRHE is not of interest in this study. A potential excursion starting slightly more

positive than the hydrogen evolution reaction (HER) shows an anodic current response

between 0.05 and ca. 0.4 VRHE (figure 3.2a). Those are characteristic features of the hydrogen

desorption of weakly and strongly bond hydrogen on platinum surfaces with different

crystallographic orientations, known as HUPD (under-potential deposited) region [37]. It can be

used for estimation of the roughness of the electrode based on a ratio between the active

surface area and the geometric one when the double-layer charge is subtracted from the overall

current [38]. Note, that Frumkin et al. showed that a linear correction leads to underestimation

of the real HUPD charge in sulfuric acid. With respect to the nature of the electrolyte, linear

correction can lead also to overestimation like in the concentrated HCl and HBr [39]. Overall

the linear interpolation gives a systematic error in the order of 15-20% that has be taken in to

account.


14

Figure 3.2 a) Linear sweep voltammetry on polycrystalline platinum in 1 N H2SO4 represented

as total capacities (line 1) and true double layer capacities (CDL) (line 2). The dashed line 3 is the

difference between 1 and 2. Line 4 corresponds to formal double layer correction based on

linear interpolation from the intermediate region between hydrogen and oxygen adsorption

regions) [39]. b) Plot of a cyclic voltammogram of Pt in 0.5M H2SO4 including the

corresponding ratio between the anodically consumed charge in the surface oxidation vs. HUPD

charge (Reproduced by permission of Elsevier form ref. [40]).

The flat region between ca. 0.4 and 0.8 VRHE is related to the double layer charging, where no

Faraday reaction is observed in inert media. Further positive increase of the potential above

ca. 0.8 VRHE causes oxidation of the surface. Conway and co-worker described this region as a

formation of 2-dimesional O/OH overlay array on the Pt surface by formation of structures as

Pt4/OH, Pt2/OH and Pt/OH represented by three overlapping anodic peaks OA1, OA2 and OA3

(figure 3.2b) [40]. Around 1.1 VRHE, a full coverage with oxygenated species is reached and the

successive oxide formation continues over interfacial reorganization of the platinum and

absorbed O/OH transforming into a quasi-3D layer. This phenomena is known as place-

exchange and causes irreversible roughening of the surface, as described in the model

introduced by Vetter and Schulze (PEM) [41,42]. Consequent oxide growth follows a

logarithmic relation in time that can be also predicted with various models: PEM, point defect

model (PDM) [43,44], high-field model (HFM) [45,46] or nucleation and growth mechanism

(NGM) [47]. However, only the PDM incorporates a possible dissolution as a parallel reaction

to the oxide formation. Shibata made an important suggestion that the oxide is comprised from

two layers, where a thick oxide is growing between the metal substrate and the outmost


15

monolayer [48,49]. Later works showed that the thin layer is present between the metallic phase

and the outer layer, and was classified into two types [50–52]:

(i) -oxide film (or barrier layer): a monolayer of compact anhydrous metal oxide

containing mainly (e.g. PtO or PtO2).

(ii) �-oxide film (or outer layer): hydrous metal oxide (e.g. PtO(OH)2 or Pt(OH)2.2H2O)

with microscopically porous structure [52,53].

Various experimental techniques have been employed in the characterization of the nature

and the composition of the oxide layer. Studies utilizing XPS analysis showed that platinum is

present mainly in two oxidation states within the oxide layer, namely Pt(II) and Pt(IV), where

different stoichiometry has been postulated like PtO [41,54–57], Pt(OH)2 [58], PtO2 [59–61],

PtO(OH)2 [51] and Pt(OH)4 [59,62–65]. Using ellipsometry, Bockris et al. found that the

thickness of the oxide is the linearly proportional to the applied potential, with scaling factor of

0.95nm/V in the range of 1.0-1.6 VNHE [66]. The authors consider PEM as reasonable to

describe the oxide growth with following reaction path:

�� (3.1)

�� (3.2)

�� (3.3)

Macdonald et al. present diverting values of 2.0-2.5nm/V using electrochemical impedance

spectroscopy (EIS) with same linear relation (figure 3.3b) [67,68]. Despite the discrepancy in

the exact values, it is generally accepted that only a few monolayers of oxide are formed even at

highly positive potentials.


16

Figure 3.3 Dependency of platinum oxide film thickness on the applied potential a) in

0.1N H2SO4 (Reproduced by permission of AIP Publishing LLC from ref. [66]) and b)

0.1M KOH (Reproduced by permission of The Electrochemical Society form ref. [67])

Ross and Wagner concluded lack of crystallographic order in the oxide layer, due to an

absence of reflectivity in low-energy electron diffraction (LEED) experiments on

electrochemically formed film at 1.7 VRHE [69,70]. The amorphous structure can be attributed

to existence of hydroxide outer layer (i.e. �-oxide) that incorporates water molecules [51,53,71].

However, an attenuated LEED reflectivity was observed in the range between 1.2 and 1.7 VRHE,

the authors related this fact to partial oxide coverage. Alternatively, it can also be interpreted as

a growing thicker outer layer over the inner one.

Figure 3.4 Illustration of the ongoing processes that lead to formation of bilayer oxide

structure on metal following the point defect model (PDM) (Reproduced by permission of The

Electrochemical Society form ref. [44]).

Using surface-enhanced Raman spectroscopy (SERS) on thin-film platinum deposited on a

gold substrate, Weaver et al. investigated the oxidation process in ambient environment [72].

The authors envisaged that initial subsurface oxide formation proceed over diffusion of the

oxygen into the platinum lattice via vacancies/defect, rather than “concerted” platinum-oxygen

place-exchange. Macdonald and co-workers also expressed severe doubt about the place-

exchange mechanism, considering that it does not predict a steady-state thickness of the barrier

oxide layer and proposed the point defect model (PDM). In their recent ex-situ investigation


17

utilizing angle-resolved X-ray photoelectron spectroscopy (ARXPS) the bilayer model was

confirmed, by showing that electrochemically formed Pt oxide has a highly defected structure

with an inner part consisting of n-type Pt(II) barrier oxide layer [68,73,74]. Following the PDM,

the authors postulated that formation of the inner oxide (-oxide) is a result of generation and

annihilation of oxygen vacancies at the metal/barrier interface and considered this layer as

passive. The outer oxide is formed by hydrolysis of transmitted cations through the primary

oxide film as illustrated in figure 3.4, where a competition between precipitation of oxygenated

species on the �-oxide/electrolyte interface and possible dissolution occurs.

3.4 Dissolution of platinum 3.4.1 Polycrystalline surfaces

Platinum dissolution was detected for the first time in the beginning of the twentieth

century [23], indirectly as a side reaction occurring on the counter electrode. Furthermore, loses

of platinum electrode material in industrial production of hydrogen peroxide, attracted the

interest of the scientific community towards the dissolution phenomena. Chemodanov and

Kolotkin addressed this issue using Pt foil irradiated in a nuclear reactor, where the dissolution

rates were monitored as a function of the radiation of the electrolyte samples [24]. Figure 3.5

shows a representative measurement performed in 3M HClO4 for different anodic polarization,

where the dissolution current is in a range of ca. 10-5–10-8 A/cm2 (e.g. 10-0.1 ng/cm2s), several

orders lower than the overall current [24,75]. An important finding was the interrelation of the

kinetics between two formally independent processes, namely metal dissolution and the oxygen

evolution reaction (OER). Similar parallelism was observed also by Bockris and co-workers

[76,77], suggesting that both reaction paths proceed over common intermediate surface

oxide(s) [78]. Additionally, the authors reported the appearance of Pt traces during the negative

polarization into the HUPD region with rates of ca. 10-8 A/cm2, which drop after a few minutes

below the detection limit. It was suggested that the oxygenated species on the surface have a

dual impact, namely they are involved in the dissolution process and also act as an inhibitor of

the corrosion at relatively high potentials [79]. It was proposed that the transport of water

molecules, as a main source of oxygen, to the electrode interface has a key role. Moreover, it

was concluded that the direct electrochemical dissolution �� is

thermodynamically very unfavourable. Ota et al. came to a similar conclusion measuring the

platinum losses by a semi-microbalance. After several hundreds of hours of

chronopotentiometry at elevated temperatures, it was shown that the Pt corrosion rate is


18

proportional to the OER current and the activation energy for the dissolution process

remained the same (18 kJ/mol) at 1 and 2 A/cm2 showing the enhanced dissolution with

elevated temperature [80].

Figure 3.5 Comparison of measured dissolution and overall current in relation with the applied

potential at three different temperatures (�)57°C, (�)25°C and (�)-18°C (adopted from ref.

[24]).

Later on Rand and Woods performed electrochemical experiments based on repetitive

cycling between 0.41 and 1.46 VRHE in 1M H2SO4 at 25°C. They reported an average dissolved

amount of 4.8 ng/cm2cycle, where the contribution of Pt(II) vs. Pt(IV) was estimated by post

analysis to be 3:10 [81]. Using a charge imbalance �� between the anodic and

cathodic potential sweep (in the range for oxidation and reduction of the surface) as illustrated

in figure 3.6a, the authors concluded that platinum is dissolving anodically arguing with results

obtained by Chemodanov et al. for the region of OER [24]. This is surprising considering that

none of those published results is performed at OER potentials and the proposed mechanism

responds rather to a chemical process. Moreover, while �� was reported for several upper

potential limits, the dissolution amount was determined only for one positive vertex potential

by post analysis, namely for 1.46 VRHE. The dissolution in this case, corresponds to

��=10.6 �C/cm2, that is smaller by a factor of three in comparison to the charge imbalance

measured electrochemically for the same potential window as shown in figure 3.6b.


19

Figure 3.6 a) Cyclic voltammogram for platinum at potential sweep rate of 40 mV/s; b)

Relation of the charge imbalance �� with the upper potential limit (Reproduced by permission

of Elsevier from ref. [81]).

Johnson et al. made a controversial study using a rotating ring disc electrode (RRDE), where

a reductive current on the gold ring disc was interpreted as redeposition of dissolved platinum

during the oxide reduction [25]. Traces of Pt(II) were found in 0.1M HClO4 with average

amount of 3.3 ng/cm2cycle after potential cycling between 0.2-1.2 VSCE with a scan rate of

0.5 V/min. The increase in the reductive current at the ring was observed around 0.8 VSCE in

the cathodic sweep (ca. 1.05 VSHE), thus the authors considered stripping of the oxide layer as a

descriptor for the dissolution process, given with following reaction:

�� (3.4)

Tsuru and co-workers adopted a similar sample-collector approach using a channel flow

system and confirming the previous results of Johnson et al. [25,82–84]. However, the datasets

should be interpreted cautiously due to the fact that residual oxygen in the electrolyte obviously

causes overlapping reductive currents on the collector electrode. Nevertheless, the post analysis

showed unambiguously the presence of platinum traces in the solution and qualitatively

confirmed the dissolution during the reduction of the oxide layer [85].

A similar observation indicating the cathodic dissolution route during the chlorine evolution

reaction (CER) was made by Roberts et al. utilizing a platinum-coated electrochemical quartz

crystal microbalance (EQCM) [86]. Once the oxide layer is not subjected to reduction, no

significant dissolution can be observed. Massive enhancement in dissolved amounts up to more

than half of a monolayer per cycle (ca. 272.7 ng/cm2cycle) is monitored in 0.1M HCl+0.9M

NaCl for a potential window from -0.2 to 1.05 VSCE. This was expected, when taking into

account that Pt forms a water soluble complex with chloride ions, which can promote the

dissolution by increasing the removal of Pt ions from the vicinity of the electrode and prevent a


20

possible redeposition [87–89]. Several consequent EQCM studies have been performed,

providing large datasets on Pt dissolution predominantly in acidic media, with typical amounts

in the order of few ng/cm2cycle and with a general trend of enhanced dissolution with

increasing acidity and temperature [90–94]. As a consequence, there is no common opinion for

the exact dissolution mechanism, and several reaction pathways have been proposed in

literature:

o Direct electrochemical dissolution:

�� (3.5)

�� (3.6)

o Chemical dissolution:

�� (3.7)

�� (3.8)

�� (3.9)

�� (3.10)

�� (3.11)

The reductive dissolution can be found also as a competition between dissolution and

reduction following an overall reaction [91]:

�� (3.12)

An important unresolved issue is the actual oxidation state of the dissolved species. Recently

Xing et al. addressed this question using ion exchange chromatography, where the platinum

traces were analyzed after continuous cycling in sulfuric acid. Both, the presence of Pt(II) as

well as Pt(IV) ions were reported with a ration of 4:1 for most cases [26].

3.4.2 High surface area Pt catalysts Besides fundamental studies on polycrystalline Pt, the stability of platinum was additionally

investigated on nano-particulate catalysts. However, in this case the long-term stability is

determined not only by dissolution but also other degradation processes [14,95,96] which

constitute overall catalyst degradation: (i) detachment of the catalyst, (ii) coalescence of the

particles, (iii) corrosion of the support and (iv) platinum dissolution. A fifth degradation

process, namely Ostwald ripening, is often mentioned additionally, but can be also considered

as a consequence of primary dissolution. In this case dissolution of the Pt and deposition on

neighboring particles results in a change of the particle size distribution and overall surface


21

morphology [96]. The overlap of several degradation processes makes resolving the dissolution

mechanism on nanoparticles very difficult.

Figure 3.7 a) Measured equilibrium concentration of platinum as a function of the applied

potential in 0.57M HClO4 (Reproduced by permission of The Electrochemical Society form

ref. [97]); b) Extrapolated dissolution rate after polarization for several minutes in 1M HClO4 at

elevated temperature (Reproduced by permission of The Electrochemical Society form

ref. [92]).

Myers et al. showed that the extension of the anodic potential limit during potential cycling

has a significant effect on the degradation of Pt/C catalyst [97]. The trend of increasing

dissolution was confirmed using EQCM for thin film of platinum by Dam and Bruijn after

performing chronoamperometric experiments from 0.45 VRHE to different upper potential

limits (figure 3.7b). An important observation by Myers et al. was that an increase of the

dissolution of polycrystalline Pt in the region between 0.9-1.1 VRHE gives a slope of

92 mV/decade, which is significantly higher than 29.5 mV/decade predicted for two electron

direct electrochemical dissolution as illustrated with reaction 3.2. [34,97]. The maximum

equilibrium concentration was obtained at 1.1 VRHE, where the platinum surface state is

distorted by sub-surface oxide formation [98]. Further positive extension of the applied

potential caused a drop of the dissolved amounts associated to the formation of a passive oxide

layer that prevents the dissolution. Interestingly, the obtained concentrations of dissolved Pt

are very close to the thermodynamic expected equilibrium value of 29 nM for chemical

dissolution governed by reaction 3.10.


22

3.4.3 Structural changes Biegler et al. showed the structural alternation of platinum surface under potential cycling

conditions, concluding the weakening of Pt-Pt interaction for oxygen coverage above 1 with

respect to the Pt atoms [99,100]. The authors also recognized the reduction of the oxide layer

as an important process for the irreversible roughening, where the large amount of adsorbed

oxygen atoms provide the required energy to promote the redistribution of Pt surface atoms

and their possible dissolution. Theoretical investigations of Eikerling et al. confirmed that the

formation of Pt-O leads to a drop of cohesive energy, which is a subject of chemical

dissolution (reaction 3.11) rather than to a direct electrochemical process (reaction 3.5 or 3.6)

[101,102]. STM studies of Itaya and co-workers experimentally proved the increased surface

mobility of platinum atoms even at relatively low potentials, consequently leading to structural

reorganization and formation of diatomic steps [103,104]. In a later work Komanicky et al.

showed the same phenomenon on different crystallographic orientations below the potential of

sub-surface oxide formation [105]. A recent in-situ AFM study, Hoshi related the change of the

surface morphology to the dissolution of low coordinated sites on cubic and tetrahedral Pt

nanoparticles under potential cycling between 0.6 and 1.1 VRHE [106]. It is suggested that

dissolution phenomena, which appear at lower potentials than for polycrystalline surfaces may

be due to the surface defects and their energetics [14].

Figure 3.8 STM images of Pt(111) after electrochemical treatment in 0.1 M HClO4 saturated

with CO between 0.07 and 0.95 VRHE (Reproduced by permission of Journal of the American

Chemical Society from ref. [107]).

Recently Watanabe et al. showed that during CO bulk oxidation under continuous cycling,

the structural reorganization occurs in the form of a movement of adatoms and leads to a

general smoothening of the surface as shown in figure 3.8 [107]. The authors proposed an early


23

stage formation of oxide on the low coordinated sites and their later dissolution in the negative

potential sweep [108]. Promoted mobility of surface Pt atoms is observed also for whole

particles of Pt on carbon support forming circular patterns in the nanometer range [109]. This

was interpreted as indirect effect of the reactive gases on the degradation in general. Indeed

some studies also propose an enhanced dissolution in oxygen saturated solution, and relate this

to early stage formation of subsurface oxide [110,111].

3.4.4 Theoretical model The oxide formation causes a structural alternation of the surface state. Based on the

assumption that a partial coverage of oxygenated species on the platinum surface leads to

dissolution, Darling and Meyer model this phenomenon for spherical nanoparticles using

kinetic analysis of three possible states Pt, Pt2+ and PtO [112]. The authors neglect the chemical

dissolution of PtO justifying this with a rather low, but unknown rate constant, although

several experimental studies on extended platinum surfaces suggested rather an equilibrium

chemical reaction as the key process. Moreover, a certain potential shift is introduced to adjust

the missing initial step of OH adsorption in the oxide formation. Figure 3.9 shows the

predicted amounts of dissolved species with respect to the potential scale, where the change of

coverage starts around 0.8 VSHE and the surface is fully covered at ca. 1.25 VSHE. Following this

theoretical model, the dissolution should vanish with reaching a full surface coverage and

should not be enhanced with increase of the potential above ca. 1.25 VSHE. Certainly, this is

contradictory to the trend observed in their experimental study as depicted in figure 3.9b.

Furthermore the dissolution is described as a direct electrochemical process (reaction 3.5) that

appears in parallel to the surface passivation due to oxide formation, however experimentally

the first dissolution traces (figure 3.9a) are present 300 mV more negative than the

thermodynamically expected for Pt/Pt2+ reaction.


24

Figure 3.9 a) Simulated oxide coverage in correlation with soluble platinum concentration

during potential cycling with 10 mV/s; b) Predicted dissolution amount versus the upper

potential limit for potential sweep experiments (Reproduced by permission of The

Electrochemical Society form ref. [112]).

In summary, major discrepancies on the topic of Pt dissolution exist in the literature, and

even for polycrystalline platinum the exact dissolution mechanism has not been resolved to

date.

Chapter 4: Theoretical background

25

4 Technical background The aim of this section is to provide a brief introduction into the technical details of the

experimental methodology and specification of the used equipment.

4.1 Inductively Coupled Plasma - Mass Spectrometry (ICP-MS)

Inductively coupled plasma - mass spectrometry is a multi-elemental analytical technique

that allows qualitative and quantitative description of species in a concentration range from

ppm to ppt. Small portions of a liquid sample are pumped into the introduction system that is

comprised of a nebulizer and a cyclone chamber [113]. The resulting aerosol is introduced

through an injector tube into the ICP torch, where it undergoes number of physical changes.

Initially, the finely dispersed droplets are losing the solvation shell forming small solid particles.

The consequent propagation in the plasma leads to sublimation and transformation in to

ground state atoms. The final conversion to ions takes place in the so-called normal analytic

zone at ca. 6000°K, as illustrated in figure 4.1. The available energy is ca. 15.8 eV, which is

sufficient to cover only the first ionization energy of most of the elements of the Mendeleev

table [114]. This insures a transfer of predominantly single charged ions in to vacuum chamber.

Figure 4.1 Schematic representation of the ICP torch consisting of three concentric quartz

tubes. The finely dispersed aerosol is introduced over an injector tube in the plasma, where it

undergoes several “dissociation” steps as illustrated on top. The resulting charged ion stream is

directed toward a triplet interface directly into the vacuum chamber.


26

After the ions pass the interface, the stream is focused by ion optics and forwarded to the

mass analyzer. Nowadays most commercially used systems are based on quadrupole mass

spectrometers, as is the ICP-MS adopted in this study (NexION 300x, Perkin Elmer, US).

Their principle of operation is briefly illustrated in figure 4.2. Using a combination of oscillating

and constant magnetic field, conditions at which only ions with a certain mass-to-charge ratio

have a stable flight path are created. In this way, the isotopes of interest are filtered and arrive

to the detector. The residual portion of the initial ion stream that passes the selection criteria of

the mass filter hits the analyzer. In order to cover six to seven orders of magnitude in

concentrations, there is a long list of technical solutions. NexION 300x is equipped with an

electron multiplier with dual operation mode (digital and analog) that allows measurements in a

broad range by reducing the sensitivity at high ion load. The resulting information is presented

as intensity in arbitrary units (i.e. count) that can be quantified using an initial calibration of the

device (described in more details in chapter 6).

Figure 4.2 a) Schematic illustration of flight path of ions with different masses; b) Simplified

Mathieu stability diagram of a quadrupole, showing the stability conditions for two isotopes

and the influence of modifying the DC and RF components [114].

It is worth to mention that the performance of the mass filter is a compromise between the

mass selectivity and element sensitivity. Namely, increasing the precision in mass determination

leads to a decrease of the effective detection limit and vise versa. Using an ICP-MS several

elements of the periodic table can be analyzed. While real-time parallel measurements are not

possible because only one mass is acquired at a time, the short dwell time during switching

between masses (50-100 ms) allows for efficient sequential investigations.


27

4.2 X-ray Photoelectron Spectroscopy XPS is an experimental method based on the photoelectric effect utilizing energy dispersive

analysis of the emitted photoelectrons, which enables the determination of sample composition

and the electronic state in the vicinity of the surface. An X-ray beam is generated in an

excitation source (aluminum anode, K=1486.6 eV, PHI2000) with a typical energy range of

200-1200 eV. The X-ray can penetrate several micrometers of a solid, nevertheless XPS is still a

surface sensitive technique. This is determined by the fact that the maximum depth of escaping

electrons is only some nanometers, for example in case of platinum the mean free path is

ca. 11 nm [115]. The electrons ejected from the sample with a specific energy are filtered by a

hemispherical analyzer. The binding energy is estimated from the analysis of the kinetic energy

by the following relation:

�� where � is Planck’s constant (6.626�10-24 J/s), � is the frequency of the excitation X-rays,

�� is the measured kinetic energy of photoelectrons, and�� is the work function of the

spectrometer [116]. The binding energy spectrum is processed with a library adjusted to C1s

Carbon peak position for PHI2000 at 284.1 eV using CasaXPS software.

4.3 Scanning Kelvin Probe microscopy The Kelvin probe technique was introduced to measure the work function of different

materials by Lord Kelvin [117]. Utilizing a lateral oscillation of the probe, a current flow is

induced between the tip and the sample. It is compensated by applying an external bias that

results in a shift of the Fermi energy of the tip. The substrate is usually grounded to prevent

change of the energy level. This so called compensation method enables an estimation of the

work function by measurement of the applied potential. The exact value is calculated using

following relation [27]:

��

In its modern version, the sample is placed on translational stage with a backside contact to

a reference tip, thus enabling the scanning of a certain area of the sample.


28

Figure 4.5 a) Schematic setup of the SKP system; b) Energy levels of the sample and tip

before approaching the surface; c) Illustration of equilibration of the energy levels between the

tip and the sample by applying a potential difference of �U.

Chapter 5: Automated system development – from design to implementation

29

5 Automated system development – from design to implementation

The discovery of new fundamental phenomena and development of improved materials is

often achieved by adopting smart design of advanced experimental methods and combination

of different, complementary techniques. Also modern electrochemistry requires adequate

approaches to increase the information depth on electrode reactions and in particular handle

the large amount of parameters that come into play. Several electrochemical and analytical

methods with high-throughput screening capabilities have therefore been developed during the

previous decades, as for instance the scanning micropipette contact method (SMCM) [118],

flow type-scanning droplet cell (f-SDC) [119,120], scanning electrochemical microscopy

(SECM) [121–123], multi working electrode array [124] etc., using the benefits of computational

control over the hardware. All these systems have something in common, namely the self-made

components running under a single software application.

The investigation of the stability during electrochemical treatment requires an approach

based on the coupling of electrochemical methods online with a spectrometer. As many of the

functionality criteria were not strictly defined (or not defined at all) at the beginning of the

project, the design of the following experimental equipment was done following the spiral

model of software/system development, namely to code the development with maximum

flexibility with respect to the changing or unknown, upcoming requirements [125]. The

electrochemical experiments are usually performed using a potentiostat that allows current

and/or potential control. In this project the Reference 600 from Gamry Intruments was chosen.

Its compatibility with several programming languages is very beneficial for the implementation

of custom measuring procedures. The final source code was implemented in graphical

programming environment LabVIEW using embedded ActiveX object. The controlling

programs were distributed in a compiled form as executable on the laboratory computers.


30

5.1 Hardware – between commercial and custom solutions

Appropriate choice of the equipment is crucial for the accurate and efficient high

throughput system. An important criterion is the interface of the hardware, which has to allow

external control. All devices have digital and/or analog input. The communication with

hardware components was integrated with self-made routines directly in LabVIEW, or in some

cases indirectly over dynamic link libraries (ddl).

5.1.1 Rotating Disc Electrode (RDE) The rotating disc electrode is a well-established technique for electrochemical investigations.

It founds often applications in the fundamental studies mainly due to the precisely defined

mass transport regime [126,127], nevertheless no commercial vendor provides a completely

automated solution for usage of RDE, due to the number of components that have to work in

parallel and different requirements in laboratories. The developed RDE system within this

work can be separated into three main sections:

(i) Electrochemical part – The potentiostat (Reference 600, Gamry) is connected to the

working, counter and reference electrodes placed into in-house made Teflon® three

compartment cell.

(ii) Rotator part – The working electrode is fixed on a motor (EDI101), where the

rotation rate is controlled by Speed Controlling Unit (CVT101 from Radiometer

Analytical). The digital access is made by a DAQ card (NI USB 6009), where 1 mV

of output voltage corresponds to approximately 1 rpm. Each RDE system has its

own correction factor for the precise speed control. The communication is

performed without feedback, due to limitations of the Speed Controlling Unit.

(iii) Gas system – The self-made two channels gas system is assembled from Swagelok

parts as illustrated in figure 5.1. The gases can be exchanged between argon,

oxygen, hydrogen, carbon monoxide and carbon dioxide using six magnet valves

(6011, Bu�rkert) controlled by a customized unit containing ICP-DAS i-7520-CR and

i-7055D-CR cards. Additionally, the argon stream can be mixed with one of the

listed gases in precisely defined ratio. The flow rates are adjusted by two

commercial mass flow controllers (Bronkhorst, Series:EL-FLOW) in the range of 0-

200 ml/min connected via a FlowBUS network and accessible over a converter


31

unit.

Important to notice is the separate interface of the hardware components in each section.

This makes possible their asynchronous operation without causing race condition and allows

quick replacement in case of defect or general change of the hardware. Figure 5.1 illustrates a

sketch of the hardware. For simplicity, several physical connections are removed (like the

cables between the potentiostat and the electrodes, power supplies, valve connections,

FlowBUS network etc.).

Figure 5.1 Schematic representations of the hardware components in the rotating disc

electrode setup and their communication interface.

5.1.2 Scanning Flow Cell system (SFC) The SFC system is based on the concept of a channel electrode, where the electrolyte is

flowing over the working electrode. In contrast to its classical version, the sample (i.e. working

electrode) is not integrated in to the walls, but it is externally introduced on a three-dimensional

translational stage. For this purpose a completely new design of the flow cell was made in this

work. Figure 5.2a-b depicts only two of the developed CAD models of electrochemical cells in

the SFC system, which were in-house manufactured using a milling machine (CAM 4-02

IMPRESSION, VHF camfacture AG) from polyacrylate or polycarbonate. The control over the

cell geometry during fabrication with a �m-precision insures the reproducibility of the

measurements with different cells. The contact area with the sample is restricted by silicon

gasket (RTV 118Q, Momentive), so that the cells with 0.4 and 1 mm channels have defined


32

contact areas of 0.25 and 1.1 mm2, respectively. The illustrated silicone sealing is not gas tight,

thus to prevent possible diffusion of ambient gases (especially oxygen) into the electrolyte, two

outer purging channels with argon are employed (figure 5.2). In this manner, the spacing

between the sample and the main body of the SFC is continuously filled with inert gas that

shields the gasket from air.

Figure 5.2 a) CAD model of the SFC cell with 1mm electrolyte channels and external purging

modification, b) alternative SFC cell geometry with 0.4 mm inner diameter of the channels;

Magnified image of SFC tip: c) in non-contact mode and d) in contact with the sample. [A.11]

The SFC cell can be positioned at various points in the XY-plane allowing investigation of

local properties of the working electrode. The counter electrode is introduced in the electrolyte-

supplying channel. In this way a possible redeposition of the dissolved species from the sample

on the counter electrode is prevented. The reference electrode is positioned in the opposite

channel to the counter electrode (i.e. in the outlet), in order to reduce the IR drop by acting like

classical Luggin capillary and also to prevent possible chloride contaminations from leaking of

the commercial electrode (Ag/AgCl/3M KCl, Metrohm). This electrode configuration is used

for the measurements in the following chapters. Note, that the cell design is only one of the

bottlenecks in the development of the whole SFC. The system itself is made out of a number

of different devices that increases the complexity significantly in comparison to the RDE

system, as illustrated in figure 5.3.

��

��

��

� � � ��

��

"��"��

!��

��

��

��

��

��

��

��

!�� #��

��

��


33

Figure 5.3 Simplified schematic representation of the hardware components in the scanning

flow cell (SFC) with their communication interface and connections. [A.18]

Similar to the RDE system, the hardware can be classified in three main sections:

(i) Electrochemical part – The potentiostat (Reference 600, Gamry) is connected to the

electrodes over a shielded extension cable box. The reference and counter are

placed stationary with respect to the SFC tip. The working electrode (WE) is

contacted by a tungsten needle and can be moved over three axes; Additionally the

WE is placed on a thermal conductive platform that is incorporated in the

temperature controlled water circuit of the main electrolyte vessel maintained by a

bath circulator (Fisherbrand FBC 735, Fisher Scientific);

(ii) Positioning part – The sample is fixed on a platform on top of a three-dimensional

translational stage (M403.6DG motors, step precision of 0.1 �m) with independent

axis control over a PCI card (C843.4, Physik Instrumente) integrated in the controlling

laboratory PC. Optional optical adjustment of the position can be done using two


34

USB cameras with 10x and 70x magnification (SMX-M83C, Sumix). The contact

force between SFC cell and the sample is monitored with bending force sensor

(KD45-2N, ME-Meßsysteme) over a self-made acquisition module (ICP-DAS i-7520-

CR and i-7016-CR cards). Typical values in contact mode of the SFC are between

50 and 500 mN;

(iii) Gas system – Similar configuration as described in chapter 5.1.1-(iii), with minor

modification to the outlet. The channel 1 is used for outer purging of the SFC tip

and the second one is used for the saturation of the electrolyte in the main vessel,

as shown on figure 5.3;

The electrolyte flow is driven through the cell by an integrated MP2 peristaltic pump,

incorporated directly in the ICP-MS. In order to decrease the time delay and improve the

lateral velocity between the SFC and the ICP-MS, tubing with reduced dimensions with respect

to the supply channels and the one in the SFC tip is used. In non-contact mode a continuous

electrolyte flow is sustained (figure 5.2c). Several technical modifications are incorporated to

provide stable system performance in non-contact mode like precise adjustment of the

curvature of the meniscus. Maintaining constant electrolyte level in the main vessel is achieved

by continuous supply by a second peristaltic pump that compensates the outflowing amount.

5.1.3 Inductively Coupled Plasma - Mass

Spectrometry (ICP-MS) In conventional applications the ICP-MS is used for multi elemental analysis of species with

constant concentrations in the region down to ppb/ppt. The samples are usually prepared in

acidic media and mixed addition of a reference element, later on used as an internal standard.

In case of the coupling with SFC, to prevent possible interference of the internal standard with

the electrochemical measurements, both solutions are mixed after the electrochemical cell and

shortly before the introduction system of the ICP-MS, as shown on figure 5.4. Using tubing

with different inner diameter (typically 0.38 or 1 mm) allows the user to vary the mixing ratios

(4:1, 1:1 and 1:4). This is used additionally to increase/decrease the effective sensitivity in some

cases. Afterwards, the mixed solution is dispersed in to the spray chamber by the nebulizer

(Meinhard) with argon stream of ca. 0.9 L/min. The operation range of the nebulizer is the

determining component for the electrolyte flow in the combined system and thus has major

impact on the system design. For example, a SFC cell configuration with 1 mm channel


35

diameter operates with an average volume velocity of ca. 0.2 mL/min (through the cell) at a 1:1

mixing ratio. The aerosol is introduced in the ICP torch, where it is decomposed, dissociated

and ionized. After that the ion stream passes through the triple cone interface in to the vacuum

chamber, where the isotopes of interest are ‘filtered’. As mentioned before, the quadrupole

provides a stable ‘flight’ path only for a single charged isotope over a short period of time (i.e.

dwell time of 50-100 ms). Thus a set of isotopes is measured in sequential order (i.e. method)

by several repetitions (i.e. sweeps) in accumulative mode, which gives the final acquisition value

of one method. In case of platinum, a method covering the two most abundant isotopes (194Pt

and 195Pt) and the internal standard of 187Re gives an effective acquisition rate of ca. 1 Hz for

5 sweeps/reading.

Figure 5.4 Schematic representation of the modified introduction system of the ICP-MS and

the hardware components within the device.

5.2 Software development 5.2.1 Software architecture – modular approach

The hardware-controlling program represents the core of the development. It has to

combine different devices with several kinds of communication in terms of interface and

timing under the roof of single software. This requires maximum flexibility in the software

architecture in the run-time procedure. Nevertheless, the general structure follows a sequential


36

logic as illustrated in the flowchart on figure 5.5. First, the input parameters for the header of

the data are requested, and then the main program is initialized. In the next step, the

communication with the hardware components is validated. After successful start-up of the

equipment, the program goes in stand-by for the run-time mode, which can be terminated and

followed by safety shut down of the hardware in case of closing the software. As a final step,

the main program closes with all running processes in the background and creates an event log

for the operation time in a single ASCII file.

Figure 5.5 Flowchart of the programming logic of the main controlling application; Steps 1

and 2 are performed in parallel during real application, in order to reduce the boot time. Step 3

initialize and check the hardware with possible transition to step 5 in case of error in the

hardware functionality. Step 4 is described in details in figure 5.6;

The challenging part is to create the source code architecture for step (4) of the flowchart

from figure 5.5. It requires several parallel functionalities like: visualization, data storage,

communication (acquisition/control) etc., where all those processes have different execution

timing. The first step is to create a straightforward separation of the programming logic and

user inputs from single hardware functionalities, or to generate a so-called abstraction level

[125,128]. That will allow at a later stage an integration of smart algorithms in between, like

automatic execution under predefined conditions [129]. This is achieved by extending the

classical consumer/producer design pattern [125]. In its original version, it has one producer


37

loop usually for the user interface (in pink in figure 5.6) and only one programming loop for

the operations (in blue in figure 5.6). That limits the execution frequency to a single time cycle

of the consumer loop, which can be overcome by introducing a separate structure for each

hardware component. The operation commands to control the different processes are

generated centralized in the “producer loop” and distributed over the “System Queue” to the

consumer. Here appears a problem in distinguishing the destinations (e.g. consumer loop) of a

single task. In addition, many continuously running parallel processes will cause overuse of the

laboratory computer resources (most likely crash) even with integrated timeout. Those issues

are resolved by the development of universal algorithm that can be inserted and configured

individually for each consumer loop. This enables continuous, standby or task-oriented

execution, just by a simple modification of the input constants. The block diagram of the

source code is illustrated in the zoom in of the “CheckVI” in figure 5.6.

Figure 5.6 Schematic illustration of the developed modular design pattern in LabVIEW for the

run-time procedures. The pink represents the UI loop with event handler and the blue is used

for the single hardware components or individual programming functionalities (e.g. data save or

graphical visualization). (Reproduced by permission of AIP Publishing LLC from ref. [A.3])

It is crucial to emphasize that data acquisition is separated from data processing procedures

on purpose. The information is buffered on the principle of first-in-first-out (FIFO) in the

“Data Queue” (fig.5.6). In this way, no data is lost, even in case of slow performance of the

laboratory computer or redistribution of the resources on other tasks.

5.2.2 User interface – friendly and efficient design The programming functionality is the heart of the software, and the user interface (UI) is its

face. The UI has to be intuitive and restrictive for the operator in order to be used efficiently.

This is achieved by following the general guidelines for UI, by separating it into three main


38

fields: status indicators, controlling parameters and visualization [130]. Figure 5.7 depicts the

UIs of the RDE and SFC controlling programs. Their similarity enables operators of one

system to become familiar with the other in a short period of time. The UI has on top the first

field (in red box) containing status indicators for running processes and the predefined worklist

for the automatic mode. The second field includes controlling parameters for all hardware,

which can be changed online for the manual mode or can be used in creating/modifying the

worklist for the automatic one. In the last part (in blue box), a simultaneous graphical data

representation is made for the last running method in different forms (like current vs. potential,

or potential/current vs. time, etc.). Furthermore, different functions are activated or disabled

with respect to the operation mode or the status of the program. It can be easily recognized

that, for example, the mode cannot be changed during a running measurement.

Figure 5.7 User interface of the controlling software application of the (a) SFC and (b) RDE

system. The Front Panel is divided into three fields. (i) Status indicators and worklist in the red

box; (ii) Controlling parameters and settings in the green and (iii) online data display in the

blue;


39

5.3 Data management 5.3.1 Data storage

An automated approach implies also automated storage of a large amount of data. As shown

in figure 5.6, saving and visualization of the data is separated from the acquisition process,

which allows different data treatments to be performed independent on the acquisition

procedure. All folders and data files are created in the background without any interaction of

the user, reducing the operation effort and possible errors. The data storage is made in defined

structures in separate folders for each worklist in the automatic mode or in a temp folder for

the manual execution as shown on figure 5.8.

Figure 5.8 Schematic tree structure of the storage folders and the data files; Example of

consumed HDD space and amount of data files after a single degradation test on commercial

catalyst Pt/C over ca. 10h;

An additional security algorithm is implemented to prevent loss of information in case of

sudden system collapse or failure of the main power supply. The acquired values for the

running experiment are buffered in the file last-measurement.txt in real time. Even in case of

shortage of the main power supply of the equipment, the dataset is saved automatically. An

��

!� �� "��

! ��"��

��

��

��

��

��

��

��

��

��

��


40

important aspect is the reverse data tracking, meaning that not only the general structure is

strictly defined, but also the file formats. All files are saved in ASCII format (figure 5.9), which

it is not that space-saving as binary format, but on the other hand doesn’t restrict the evaluation

procedure only to the developed programs in LabVIEW. The data file can thus be accessed

with any text editing application.

Figure 5.9 File format of the descriptor file in a) and a selected single measurement file in b).

Both are separated in three section: (i) Information about used hardware including user

comments, (ii) Header of the content and (iii) Data or descriptive table.

5.3.2 Evaluation – quick data processing Treatment of a large amount of data as produced by automated experimental equipment is

typically a time consuming procedure. Loading and visualization of data in most of the cases,

follows several rather routine steps, and analysis of the experimental results by the user is thus

often a rather demanding process. The data overflow is unavoidable during high throughput

experiments, but the efficiency of the evaluation methodology can be improved. For that

purpose, a set of supporting programs has been created. Figure 5.10 depicts two of them: a)

processing of electrochemical data and in b) combined evaluation of ICP-MS data with the

electrochemical one from the SFC. Operations used on a daily basis (like determination of

active surface area from CO-stripping/Hupd region, calculation of number of exchanged


41

electrons, Levich-Koutecky plot, Tafel etc.) are completed in a few mouse clicks. The evaluation

protocol can be saved and loaded in a binary TDMS format. The architecture of the software

follows the so-called state machine design pattern with integrated event handler. That allows

simple extension without affecting the existing algorithms, just by introducing new cases in the

existing source code and accessing the data from a shift register.

Figure 5.10 User interface of the data evaluation programs: a) for the electrochemical dataset;

b) for a combined evaluation of the datasets from the ICP-MS and SFC system;

Chapter 6: System validation – proof of functionality and limits in application

42

6 System validation – proof of functionality and limits in application

As described in the previous chapter, a variety of different hardware and software

algorithms have been combined in order to achieve fully automated experimental setups. The

increasing complexity of the system inevitably leads to a higher probability of hidden errors.

The programming logic of individual components was proven in several separate tests. Here

will be shown only the reliability of the computer control during real operation conditions and

verified of the full automation of the coupling between the electrochemical and spectrometric

systems.

6.1 Automation - proof of concept The system test is performed using a degradation study of a 3 nm commercial carbon-

support platinum catalyst (TKK, Japan) with one of the RDE setups. Different glassy carbon

RDE tips are prepared with the same loading of 20 �gPt/cm2 dried under nitrogen atmosphere.

The long-term performance is validated under automatic mode of operation using an

accelerated degradation protocol simulating harsh start-stop conditions of operation [131]. The

experimental procedure is illustrated in the flowchart in figure 6.1a. Initially, the potential of the

reference electrode versus RHE is evaluated in hydrogen atmosphere by measuring the open

circuit potential (OCP), and then a short pretreatment is applied to insure the adhesion of the

catalyst layer to the RDE tip. In the next step, the specific activity (SA) for the oxygen

reduction reaction (ORR) is determined at four rotation rates (400, 900, 1600 and 2500 rpm)

using the equation [132]:

��

��

� ��

��

� ��

where�� is the measured current at 0.9 VRHE, �� is the diffusion-limited current at the same

rotation rate, �� is the oxidation charge from the first CO-stripping procedure in the region

between 0.7 and 1 VRHE (figure 6.1b) and � correspond to the charge density of polycrystalline

platinum (ca. 0.195 mC/cm2) [133]. The variation of the SA value is based on experimental


43

results from five different samples. The following degradation procedure leads to a decrease in

the active surface area over time, as measured at different stages in between the degradation

protocol by CO stripping method. After 7200 degradation cycles, more than 60% of the active

surface area is lost and/or not accessible (figure 6.1c). The maximum relative standard

deviation in the determined area at the end of the degradation protocol is less than 8% as

determined from five different samples, and is attributed to the precision of the preparation

methodology of thin films on the RDE tips.

Figure 6.1 a) Schematic illustration of the applied experimental sequence; b) Time evolution of

the active surface area during the degradation procedure, measured by CO stripping (inset).

The error bar is based on the deviation in the results of five samples on five measuring days. c)

The arrows indicate the evolution of the cyclic voltammograms between 0.4 and 1.4 VRHE at a

scan rate of 1 V/s (black shows the forward scan and in red the backward one) (Reproduced by

permission of AIP Publishing LLC from ref. [A.3]).

��

��

��


44

The automated degradation test showed long term operation of the developed system over

several hours without any user interaction. As the software architecture of the RDE and SFC

follows similar algorithms, no separate validation for the SFC is performed. Moreover the

reliability of the SFC system(s) can be observed in the many experimental series shown in the

next chapter.

6.2 Performance of electrochemical systems – RDE vs. SFC

The electrochemical response of both systems is evaluated using the same polycrystalline

platinum sample in freshly prepared 0.1 M perchloric acid. Figure 6.2 shows the cyclic

voltammograms (CVs) in argon-saturated electrolyte for the RDE and SFC setups in a) and b),

respectively. The CVs in both systems are comparable and reproducible. The HUPD region in

case of the SFC system is slightly shifted downward to negative currents. This is an indicator

for residual oxygen in the solution that can be reduced at these potentials (i.e. causes a reductive

current response). The partial leaking of oxygen under the silicon gasket remains an issue. It

has to be mentioned that despite the fact that the SFC configuration has its drawbacks, it still

provides a powerful tool for comparative analysis, online combinations of electrochemistry

with complementary techniques and in comparison with RDE it has negligible IR-drop due to

the typically low working currents. The electrochemical measurements with the SFC are limited

in the region where there is no massive gas evolution on the counter and/or working electrode,

thus the potential windows are chosen precisely to prevent blocking of the channels by bubbles.

In the next step, the electrolyte is saturated with oxygen and the specific activity of platinum

for the ORR is determined. The kinetic current is extracted from the overall current (figure

6.2.c-d), which is corrected for diffusional effects in the positive-going potential sweep

following the Levich-Koutecky equation. The Tafel plot has been acquired using the evaluation

software from chapter 5. The experimental results from the RDE system show coinciding

curves for the different rotation rates, which is not the case for the SFC. An increase in the

flow velocity in the channels leads to a positive shift in the kinetic current response. This can

be attributed to the change of the flow profile distribution over the surface with different flow

velocities, e.g. increase of the velocity leading to shrinking of the hydrodynamic layer, Later is

linearly proportional to the diffusion layer thickness and leads in combination with an


45

inhomogeneous flow profile to an inhomogeneous current density distribution on the surface

in the kinetically controlled region.

Figure 6.2 Electrochemical measurements in 0.1M HClO4 on polycrystalline platinum: Cyclic

voltammograms made in argon are shown in a) for the SFC and in b) for the RDE; ORR

measurements and the corresponding Tafel plots in c,e) and d,f) for the SFC and RDE system

respectively; The insets in c) and d) shows the change in the limiting current with square root

of the angular velocity for the RDE and third root of the volume velocity for the SFC;


46

The non-uniform distribution of the flow profile is critical and does not allow a direct

analytical solution of the diffusion limited current with respect to the flow rate at this statge.

Nevertheless the extracted mass transport limited current follows the third root relation

predicted for the channels electrode [134]. Furthermore, the order of the diffusion layer

thickness can be estimated using the general relation for the measured currents [27]:

� � ��

where � is the number of exchanged electrodes (for ORR on platinum is assumed to be 4), � is

the Faraday constant (96485 As/mol), C is the bulk oxygen concentration in 0.1M HClO4

(1.2 mM) and k is the mass transfer coefficient that can be further expressed in case of a

diffusion limited reaction as the ratio between the diffusion coefficient of oxygen (D=1.67�10-

5 cm2/s) and the diffusion layer thcikness. From the limited current in region within 0.2-

0.5 VRHE (Fig. 6.2d), the average thickness of the diffusion layer can be estimated to be in the

range of ca. 78-157 �m. A single precise value cannot be considered for certain flow velocity

due to the inhomogeneous distribution of the flow profile as illustrated in figure 6.3. Exact

evaluation of the SFC system will require numerical simulation in future.

Figure 6.3 Illustration of the distribution of the flow profile lines for: a) scanning flow cell, b)

channel electrode and c) rotating disc electrode;

6.3 Inductively Coupled Plasma - Mass Spectrometry (ICP-MS) and coupling with the SFC

The ICP-MS is particularly interesting for the investigation of low dissolution rates and

provides its results in relative units (i.e. counts) for the isotope(s) of interest. The quantification

of the signal is made on the basis of a calibration made before the actual measurements. For

that purpose, electrolytes with known concentration are prepared and measured. Typically, the

intensities of the investigated isotopes are normalized to the relative constant intensity of the

��

��

��


47

internal standard. Figure 6.4 depicts an example of calibration curves for some of the noble

metals in sulfuric acid.

Figure 6.4 Calibration curves for some of the platinum group metals in concentration range of

0.5-5 �g/L, performed by the method with two groups for 193Ir, 195Pt, 197Au measured against

5 �/L of an internal 187Re standard, and for 102Ru, 103Rh, 106Pd versus 89Y as an internal standard

(5 �g/L). The equation for the linear fitting is presented for each isotope in the legend. Mixing

ratio of the internal standard with the analyte is 1:1.

For the purpose of conventional quantitative analysis, the solution of interest is

continuously introduced into the mass spectrometer until a constant response is achieved,

which is further measured several times (i.e. replicates). The final concentration is calculated

from the normalized intensity using a calibration curve (figure 6.4) for the corresponding

electrolyte matrix. In the case of coupled SFC and ICP-MS, the content of the species in the

electrolyte is changing with time. The transient signals are recorded based on the same principle

as the stationary one, however, can be measured only once (replicate(s)=1). In order to evaluate

the performance of quantification in this mode and the correlation between the spectrometric

and electrochemical signal, on the dissolution of Cu from a bulk copper sample is investigated

by applying a chronopotentiometric sequence. Figure 6.5 shows the resulting spectrometric

signal calculated as “dissolution current” using Faraday’s law:

��

� � � ��

��

��

�


48

where � is the valence of the ions, � is the calculated concentration using the corresponding

calibration curve, � is the volume velocity, � is Faraday’s constant and � is the molar mass.

Initially, the ICP-MS signal is recorded with the SFC not in contact with Cu sample, which

provides the background signal as shown in the first seconds of the inset in figure 6.5. Once

the contact with the sample is achieved, the native layer of oxide is dissolved in a chemical

process and results in the initial peak in the Cu signal. Applying anodic currents causes an

enhancement in the dissolution rate and thus in the measured spectrometric signal. A good

correlation between the applied current and the dissolution rate from the ICP-MS is achieved.

Note, however, with the increase in the applied currents a substantial mismatch in the

magnitude of the measured spectrometric signal appears. This is related to the fact that the

dissolution current is calculated for monovalent copper ions based on the thermodynamic

prediction of the Pourbaix diagram [34]. Under the chosen experimental conditions, copper

can however also be dissolved partially in a bivalent state, which tends to be dependent on the

current density. This phenomenon has been previously reported in other studies [135,136].

Figure 6.5 On top the experimental procedure for the measurements is illustrated. The

measurement started in non-contact mode, the Cu working electrode surface is then


49

approached and the sequence with modulated current is applied once contact has been

established. Finally the cell is lifted up (i.e. away from the sample). Below the spectrometric

signal (Idiss) and the applied current (Iapplied) are plotted. The inset shows a magnification of the

initial dissolution peak after the contact with the sample. Electrolyte: 1 mM HCl. [A.4]

A straightforward correlation of the electrochemical and spectrometric signals is achieved by

modifying the experimental conditions. Increasing the acidity of the electrolyte leads to

dissolution of copper practically only in Cu+ form. Figure 6.6 shows the nice overlap of both

datasets, where the ICP-MS signal closely follows the electrochemistry.

Figure 6.6 Correlation between the spectrometric and electrochemical signal in 10 mM HCl

electrolyte. The applied experimental sequence is similar to the one used in figure 6.5 [A.4].

Synchronization between electrochemical and spectrometric signal is another important

issue that has to be addressed. Figure 6.7a illustrates an experimental sequence of applied

anodic current on a copper sample, while the corrosion rate is recorded in parallel. Two main

parameters are extracted from the results, namely the time shift between SFC and ICP-MS

response and the peak broadening with respect to the flow rate of the electrolyte. Once the

volume velocity through the SFC is above 50 �L/min only minor changes are detected in both

parameters, as shown in figure 6.7c. The optimal flow rate for the following experiments is

determined also from a combination of two further criteria: (i) stable electrolyte flow in the


50

SFC during non-contact mode and (ii) optimal nebulizer performance of the ICP-MS, where

the second one is dominating in order to achieve high sensitivity. The experimental results

reveal a volume flow of ca. 51 and ca. 193 �L/min as optimal for the 0.4 and 1 mm SFC cell

configurations, respectively, with mixing ratios of 1:4 and 1:1 between the electrolyte and the

internal standard solution.

Figure 6.7 Schematic representations of the applied current signal of 200 nA on a copper

sample in a) and illustration of the measured signal by the ICP-MS in b); The analyte and the

internal standard are mixed in ratio 1:4. The full width at half maximum (FWHM) and time

delay are graphically shown in c) for different pump velocities in 0.1M H2SO4. for the cell with

0.4 mm diameter channels.

The time delay has two contributing parts, as shown in figure 6.8: (i) time required to escape

from the diffusion layer and (ii) time for the lateral transition of the dissolved species to the

detector of the ICP-MS.

��

��

��

��

��

��

��

��

�

��

��

��

��

��

� ��

�!�

�!�

$��

#""��

"��

��"�"%�� #�

�!�


51

The main impact originates from the convectional part, although the diffusion time can also

cause certain variations with the thickness of the diffusion layer, as plotted in figure 6.7b. The

shift in the time scale is determined experimentally on each measuring day, and lies typically in

the range of 16-20 seconds. The peak spreading is an unavoidable aspect of the experimental

system. It originates from different lateral velocities in convective flow in the tubing (figure6.3)

and the inhomogeneous profile of the diffusion layer. Species dissolution at the same moment

will thus be introduced at different time in the convective stream and the transport in the

tubing to the ICP-MS is accompanied with additional variation of the time delay before

reaching the detector. In the case of copper, five seconds of active dissolution result in peaks

with a FWHM of three to seven times broader than the electrochemical signal (figure 6.7c).

Figure 6.8 a) Schematic representation of the factors influencing the overall delay time; b)

theoretical prediction for the required time to diffuse through certain distance for diffusion

coefficient of 5�10-6cm2/s.

Chapter 7: Results and discussion

52

7 Results and discussion 7.1 Electrochemistry of platinum in acidic media

Polycrystalline platinum is a well-studied catalytic material, often used as a model system.

Figure 7.1 depicts exemplary the cyclic voltammogram of the polycrystalline surface measured

with SFC cell (1 mm diameter of the channels). In the potential window of 0-0.4 VRHE, the

typical current response for hydrogen adsorption and desorption can be observed (so called

HUPD region). It is often used for the “active” surface determination based on the works of

Conway and co-workers under the assumption of 210 �C/cm2 (1e-/Pt atom) for a single

monolayer formation [40,59]. Unfortunately minor variation of the lower potential limit cause a

significant change in the integrated value and can lead to random errors and several studies are

questioning the reliability of this method [132,137]. Additionally, the different time scale

contributes also to the variation of the HUPD charge, as shown in figure 1b. It has to be

mentioned that the correction of the double layer charging is performed with a simplified linear

interpolation, which introduces an additional uncertainty as already discussed by Frumkin and

co-workers [138]. An alternative method of active surface area determination is CO-stripping,

which however could affect the dissolution behavior of platinum. As the roughness can change

during these experiments and its influence on the dissolution is unknown, in a first approach all

measured values in the further work are represented as current as well as dissolution rate

normalized to the geometric area of the electrode. The second region between 0.4 and 0.8 VRHE

is related to the double layer charging, where the platinum shows rather chemically inert

behavior. Above ca. 0.8 VRHE an increase in the current is observed. The region between 0.8 and

1.0 VRHE has different interpretations in the literature. Conway and co-worker suggested an

initial water de-protonation and OH adsorbed on the surface [139]. An alternative statement is

made by Jerkiwiecz et al. who assume the surface oxidation to proceed directly with

chemisorbed oxygen [140]. Nevertheless most of the authors agree on the formation of

subsurface oxide starting around 1.1 VRHE [42,98]. Further extension of the anodic potential

promotes the oxide growth and above ca. 1.6 VRHE the oxygen evolution reaction (OER) begins.

The negative going sweep contains a reductive peak (1.0-0.5 VRHE) that can be used for

quantification of the amount of oxide formed in the anodic scan.


53

Figure 7.1 a) Cyclic voltammogram measured with SFC-1mm between 0.05 and 1.5 VRHE with

a scan rate of 0.2 V/s in 0.1M HClO4 electrolyte continuously purged with an argon stream.

The volume flow rate is 194 �L/min that corresponds to average lateral velocities of ca. 6 and

1 mm/s for 0.4 and 1 mm channels, respectively; b) Integrated charge in region between 0.05

and 0.4 VRHE with subtraction of the capacitive response for seven different scan rates.

7.2 Influence of the overpotential for oxide formation and reduction on platinum dissolution

As it becomes clear from Chapter 3, it is not resolved in literature what exactly triggers Pt

dissolution and how much Pt dissolves under different conditions. Most of the representative

studies assume that the oxide plays a key role in the process of dissolution [25,81,112].

Figure 7.2 shows an experimental sequence of applied cyclic voltammograms, where the

positive potential limit is extended stepwise from 1.0 to 1.85 VRHE, thus modifying the

overpotential for the oxide formation. The recorded online ICP-MS signal shows no detectable

dissolution as long as the potential cycling is performed within the HUPD and double-layer

region, even up to initial Pt oxidation at 1.0 VRHE. Note that the typical detection limit of the

ICP-MS is less than 10 ppt (equal to ca. 0.7 �g/cm2L or ca. 3 pg/cm2s or ca.

9.3�109 atoms/cm2s), based on three times the standard deviation of the blank signal [113,141].

For further simplicity in the following text is used the statement “no dissolution” is used,

which should be considered correctly as no significant dissolution for the given conditions with

respect to the detection limit for our analysis. Once the upper potential is increased above

ca. 1.1 VRHE first traces of platinum are observed. With further extension of the anodic limit, the

amounts of dissolved platinum per cycle become more pronounced, and the two peaks can be


54

distinguished during one CV as shown in figure 7.2c. A relatively small peak is observed in all

positive going sweeps, which on a first view tends to be constant with respect to the upper

vertex potential. The second one appears in the cathodic scan around 1.0 VRHE and is

significantly larger. This finding is in good agreement with the results obtained by Johnson et al.

who consider the cathodic route of dissolution as the main one [25].

Figure 7.2 a) The applied experimental sequence in 0.1 M HClO4/Ar consists of 2 cyclic

voltammograms with a scan rate of 0.01 V/s for each potential window; always starting from

0.1 VRHE to an upper potential limit between 1.0 and 1.8 VRHE and raised in steps of 0.05 VRHE

b) corresponding time-resolved dissolution profile of Pt presented on the same time axis as a).

c) The inset provides a magnification of the region around 6000 s. d) Amount of dissolved Pt

normalized per cycle plotted versus the upper vertex potential for cyclic voltammograms from

a) and additional measurement with SFC-0.4mm, including the three representative literature

references from Johnson et al. [25], Darling et al. [112] and Woods et al. [81]. [A.11]

The dissolution profile in Figure 7.2b can be furthermore quantified by numerical

integration of the peaks with a certain time region, using the following relation (all data is

processed by the developed software from chapter 5.3.2):

� � � � ��

� ��

��

where � is the volume velocity of the electrolyte, � is the concentration profile measured

within the time period between �� and �� . The dissolution amounts in each cycle are

interpreted versus the upper potential values in figure 7.2d. The experimental sequence is

performed with two different configurations of the SFC systems with 0.4 and 1 mm channels.

The results from both experiments are nicely overlapping, what shows the negligible influence


55

of the cell geometry on the platinum dissolution. The integrated values change from “bellow

the detection limit” (for potentials less than 1.1 VRHE) to almost 10 ng/cm2cycle (at ca. 1.8 VRHE).

Considering an average surface density of 1.3�1015 atom per cm2 for a polycrystalline platinum

sample (equal to ca. 420 ng/cm2) [139], the dissolution amount can be represented in

percentages of the single monolayer. From figure 7.2d it becomes obvious that the amounts do

not exceed 2.4% of a monolayer per cycle in this series and addition a rough comparison with

the few existing values of the three most representative studies is given. Despite the slightly

different experimental conditions (scan rate, pH of electrode etc.), the normalized dissolution

amounts are quite similar for the corresponding upper potential limit.

The platinum dissolution is strongly influenced by the applied potential profile, especially in

the region of oxide formation/reduction as shown in figure 7.2c. In order to investigate the

effect of the overpotential for the oxide reduction on Pt dissolution, a sequence of cyclic

voltammograms is used, with fixed upper vertex limit at 1.55 VRHE and variable lower one as

plotted in figure 7.3a. Significant dissolution is detected when the operational potentials spread

over the region that allows oxidation and follow-up reduction. As soon as the potential window

is narrowed to potentials that are not sufficient to reduce the surface, the dissolution amount

drops beneath the detection limit. Thus, if the lower vertex potential stays above ca. 1.05 VRHE,

the spectrometric signal remains below the detection limit. This can be again easily recognized

from the integrated amounts in figure 7.3c. In the region of oxide formation below the onset of

the oxygen evolution reaction platinum surface is passivated. Once the lower potential limit is

shifted back to reductive values dissolution is observed again. Thus the transition between the

oxidized surface and reduced state is a necessary condition for the significant platinum

dissolution at room temperature and potentials not exceeding 1.55 VRHE. Two “stability”

windows can be defined for the electrochemical treatment in 0.1M HClO4 (i) between 0.05 and

ca. 1.05 VRHE and (ii) between ca. 1.05 and 1.55 VRHE, where platinum dissolution can also be

neglected.


56

Figure 7.3 a) Applied potential sequence in 0.1M HClO4/Ar at room temperature cyclic

voltammograms with a scan rate of 0.01 V/s and an upper potential limit of 1.55 VRHE; the

lower potential limit is gradually changed from 0.05 VRHE until 1.05 VRHE and down again in

steps of 0.05 V; b) Corresponding platinum dissolution profile plotted on the same time axis as

in (a), the dashed horizontal line shows the detection limit for the measurement; c) Integrated

amount of dissolved Pt for each single cycle to different lower potential limits of the cyclic

voltammograms from (b). (Reproduced from ref. [A.15] with permission from The Royal

Society of Chemistry)

7.3 Steady state dissolution during chronoamperometry

From the experiments above it is clear that a perturbation of the potential signal in time is

affecting the dissolution and can accelerate or suppress its rate. Maintaining constant potential

at relatively high anodic value was also already reported as a promoter of the dissolution [24,79]

in contrast to the dissolution behavior within the stability window of the water that is rather

unclear. Figure 7.4ab illustrates one representative experimental sequence where initially the

sample is polarized at 0.15 VRHE for 200 s, than oxidized at 1.6 VRHE for 300 s and further

reduced in the consequent steps. For each variation of the potential the dissolution profile

increases sharply and afterwards declines over the time of ca. 200 s. Again an increase of the

platinum signal is detected only in a case of change in the surface state. Modulation of the

applied potential within one of the emphasized “stability” windows does not cause dissolution.

A further, more refined chronoamperometric series shows again that transition in the region

around 1.1 VRHE triggering the dissolution during the oxide formation (figure 7.4cd) as well as

during the reduction, while a clear drop in the dissolution signal occurs during extended


57

polarization times. Thus it can be concluded that steady-state dissolution within the stability

window of water plays a minor role compared to the transient dissolution

Figure 7.4 Representative chronoamperometric experimental sequences that demonstrates

typical behavior of Pt dissolution under steady-state conditions in 0.1M HClO4/Ar. The graph

a) shows the applied potential sequence (potential holds at 0.15, 1.6, 1.0 and 0.15 VRHE) with

the corresponding dissolution profile in b). [A.11] c) Applied potential profile, starting from

OCP followed by 30 sec potential hold at 0.12 VRHE, then increasing the potential value from

0.17 VRHE up to 1.47 VRHE with steps of 0.1 V and backwards where in seven of the

chronoamperometric steps the duration time was extended to 300 sec as indicated. The vertical

red dashed line illustrates the transition from 1.07 VRHE to 1.17 VRHE step during the oxidation.

d) corresponding dissolution profile to c) with integrated amounts for the peaks. The dotted

line indicates the detection limit as determined from the daily calibration.

In order to investigate platinum dissolution in the region of oxygen evolution, an additional

experimental series is conducted. As long-term chronoamperometric measurements at high

anodic potentials e with the SFC are critical due to technical limitations related with blockage

of the tubing by gas bubbles, a linear sweep voltammetry with a low scan rate of 2 mV/s is

applied. This allows short excursions to relatively high anodic potentials for a reasonable period

of time and avoids extensive gas evolution in the SFC system. Note, that an alternative

approach of using a conventional three electrode cell and performing post-analysis is

accompanied with several pitfalls like possible deposition of the dissolved species on the

counter electrode and a large volume of electrolyte that leads to even lower effective

concentration to analyze. Again the onset of the dissolution can be clearly observed around

1.1 VRHE (figure 7.5). Considering the data acquired in the region of oxide formation (figure 7.4)

it is expected that the dissolution rate should drop after certain period of time (ca. 200 s).


58

Nevertheless, exceeding 1.6 VRHE leads to a rather small, but still detectable dissolution rate

during the stationary polarization in the potential region of oxygen evolution reaction (OER).

A similar phenomena was observed during electrolysis in chloride containing solutions after

tens to hundreds of hours of operation using platinum as an anode [79,142]. Nevertheless, Pt

seems to be highly corrosion resistant in comparison with other noble materials at such anodic

potentials [77]. This can be attributed to the formation of a chemically stable Pt-oxide layer,

and to the so-called electrolyte route [50,76] for the oxygen evolution reaction on platinum,

which does not involve the oxide in the reaction pathway.

Figure 7.5 a) Platinum dissolution profile during anodic polarization from 0.05 to 1.85 VRHE

with scan rate of 2 mV/s plotted versus the applied potential in 0.01M, 0.1M, 1M and 5M

HClO4 saturated with argon solutions. The dashed line indicates the detection limit for the

measurement; b) Tafel plot of the dissolution current (calculated with valence 2) for the anodic

and cathodic polarization in 0.1M HClO4 from a).

Note on this occasion, that due to the small currents in the system, the experiments are

conducted without IR compensation. The typical operation range of the potentiostat is

between 6 and 60 �A and the measured resistance is around 100-200 Ohm. This results in an

IR drop in the range of 0.6 to 12 mV. The current exceeds to 6 �A usually only above 1.6 VRHE,

which keeps the uncertainty of the potential axis less than 1.2 mV for the low potential values.

The dissolution rate can be interpreted as a current response and presented in Tafel plot

(figure 7.5b). The extrapolated slopes for the anodic and cathodic potential sweeps are ca. 120

and ca. 305 mV/decade, the values are significantly higher than the previously reported

92 mV/decade by Myer et al. [97]. This can be attributed to the fact that the literature values are

extracted from transient measurements and are based on the assumption that the overall

dissolution amount per cycle is directly proportional to reaction rate without separating the

“anodic” and “cathodic” contribution. Interestingly, the dissolution current in the cathodic


59

potential sweep practically remains with a constant slope over more than 600 mV, this rather

indicates a reaction independent on the potential like possible chemical dissolution of some

intermediate oxide.

7.4 Decoupling the influence of the time scale of experiment and the amount of formed oxide on the dissolution rates

It is important to have in mind that a variation of the potential window has the additional

effect of extending/reducing the effective time of reactions(s) in certain potential regions, like

for example in the region of oxide formation. Thus it is also important to investigate the

influence of the scan rate during transient experiments. Figure 7.6ab shows an experimental

series in a fixed potential window with an increasing scan rate from 0.005 to 0.5 V/s and the

online detected platinum, respectively. The corresponding dissolution amounts are decreasing

from ca. 5.9 to 1 ng/cm2cycle with increasing scan rate. A detailed investigation of the

electrochemical signal reveals that the amount of formed oxide also varies significantly with the

scan rate, as evaluated by the integration of the oxide reduction peak in the region from 1 to

0.5 VRHE with subtraction of the capacitive response. The reductive charge is calculated as

following from the cyclic voltammograms:

� � � ��

��

��

Assuming that reduction charge of a single monolayer (ML) oxide is equal to 420 �C/cm2

[92], it is observed that no more than 1.5 ML are formed for a calculated roughness of the

electrode of ca. 1.3. The correlation of the oxide reduction charge with the dissolution amount

is plotted in figure 7.6c. It is important to mention again that the dissolution charge is 2-3

orders of magnitude smaller than the calculated charges from the electrochemical signal, as

predicted in the chapter 2 (1 ng/cm2~0.99 �C/cm2).


60

Figure 7.6 a) Applied experimental series in 0.1M HClO4/Ar; a cyclic voltammogram between

0.05 and 1.5 VRHE with a scan rate of 0.005, 0.01, 0.025, 0.05, 0.1, 0.2 and 0.5 V/s; b)

Corresponding platinum dissolution profile plotted on the same time axis as in (a), the dashed

line shows the detection limit for the measurement; c) The amount of dissolved Pt and

dissolution rate for each cycle versus the integrated charge for Pt oxide reduction recorded

during the same cycle whereas on top is indicated the corresponding scan rate. The dissolution

charge is calculated for an assumed valence of two. (Reproduced from ref. [A.15] with

permission from The Royal Society of Chemistry)

The results suggest that (figure 7.6c) the dissolution amounts are directly proportional to the

amount of oxide formed, which is itself inversely related to the scan rate. Unfortunately, the

contribution of anodic and cathodic dissolution can only be explicitly separated for scan rates

lower than 25 mV/s; at higher scan rates the two peaks overlap. Moreover, the influence of the

scan rate cannot be resolved unambiguously, as it changes both, the amount of formed oxide

and the time for its reduction. Due to the fact that steady-state dissolution within 0.0-1.6 VRHE

diminishes during constant polarization (figure 7.4), an advanced experimental protocol was

applied to split the influence of the timescale of the experiment from the influence of the

amount of formed oxide on dissolved Pt amount, while also resolving the anodic and cathodic

dissolution. An exemplary potential profile is presented in figure 7.7ab, where the dissolution is

clearly separated for the positive and negative going potential sweep. The effect of

overpotential of oxide formation is also investigated by changing the upper potential limit from

1.25 to 1.65 VRHE, while always using the same extended amount of time during constant

polarization. Identical experimental series are performed with six different scan rates of the

LSV steps. The data for the cathodically and anodically dissolved amount determined from


61

these experiments is plotted in figure 7.7c (in green color) and figure 7.7d (in red color),

respectively.

Figure 7.7 a) Applied experimental sequence in 0.1M HClO4/Ar, potential hold at 0.05 VRHE

followed by a sweep with a scan rate of 0.01 V/s to the corresponding upper potential limit,

hold for 300 seconds, and then a sweep back to 0.05 VRHE and successive hold; b) Platinum

dissolution profile plotted on the same time axis as in (a); Amount of dissolved platinum per

sweep for the same experiments as on (a-b) for scan rates of 0.01, 0.2, 0.5, 1, 2 and 4 V/s for

the positive sweep in (d) and negative in (c); (e) cathodically dissolved amount plotted versus

the reductive charge in the negative sweep. (Reproduced from ref. [A.15] with permission from

The Royal Society of Chemistry)

An increase of the upper potential limit has only a minor influence on the anodic dissolution

that coincides with results from figure 7.2b where similar reproducible peaks are observed

above ca. 1.3 VRHE (in the positive potential sweep). Surprisingly, a variation of the scan rate

from 0.01 to 4 V/s has also no distinguishable effect. It appears to be practically constant value

around 0.8 ��0.3 ng/cm2cycle (i.e. amount within ca. 0.2% of a single monolayer). Neither the

amount of the formed oxide nor the experimental time affects the anodic dissolution. It seems

that the anodic dissolution is triggered by the subsurface oxide formation and is rather related


62

to the initial state of platinum surface or some unstable intermediate species like a “transition”

oxide.

A different behavior is observed during the reductive scan (figure 7.7c). An increase of the

overpotential for the oxide formation leads to an enhancement of dissolved amounts for all

scan rates. From the work of Bockris et al., it is known that in the region between 1 and

1.6 VSHE an increase of the overpotential leads to a proportional increase in the amount of

formed oxide [66]. This relation is considered to be valid for thickness of less than a monolayer.

Similar to experiments in figure 7.6c, the reductive charge is evaluated from the negative going

potential sweep and compared with the cathodically dissolved amount (figure 7.7e). Thus the

previously found lower total platinum dissolution can be understood from the similar behavior

of the dominating cathodic dissolution. Interestingly, while the qualitative behavior that

dissolved platinum scales almost linearly with the oxide reductive charge is true for each scan

rate, the quantitative results are clearly diverting. Moreover, similar reductive charges do not

lead to identical dissolution amounts. In other words, it is not only important how much oxide

is formed, but also on what timescale it is reduced. The amount of dissolved platinum is

decreasing with high scan rates, becoming comparable with the anodic case and almost

independent from the amount of formed oxide when the surface is nearly instantaneously

reduced. It seems that additionally diffusion of the dissolved species from the surface into the

bulk electrolyte plays a role and has to be considered in the discussion of the mechanism at a

later stage.

7.5 Influence of the concentration of the protons

The electrolyte composition and especially the proton concentration are expected to have a

significant contribution to the dissolution of platinum. Several reports showed the

enhancement in overall dissolution in case of platinum-based nano-catalysts with higher acidity

[83,143]. Figure 7.8 shows a summary of dissolution amounts derived from cyclic

voltammograms in sulfuric and perchloric acids for several potential regions and different

electrolyte concentrations. Independent on the electrolyte acidity, the onset of platinum

dissolution coincides for all measurements around 1.1 VRHE. This suggests a proton dependent

reaction to be the rate-determining step. The onset in a case of sulfuric acid is thereby slightly

shifted positive. This might be related to higher adsorption strength of sulfates/bisulfates in

comparison with perchlorates that results in a shift in the onset for platinum oxidation [139].

Furthermore, the dissolution amounts can be investigated above ca. 1.3 VRHE separately for the


63

anodic and cathodic branches. Similar to the previous findings, the platinum dissolved in the

anodic sweep shows a minor increase with extending the upper potential limit. The values are

rather scattered around 1 ± 0.3 ng/cm2 cycle, showing the minor increase (almost independent)

with the pH change on the anodic dissolution for concentrations up to 1M. A clear

enhancement is detected only in the extreme case of 5M HClO4. Nevertheless, it still stays with

factor lower to amount in the cathodic scan that remains the main contributor to the overall

dissolution for all electrolytes. The amounts in the cathodic case, scale with the increased

acidity of the electrolyte. This is expected considering that the oxide reduction proceeds via

protonation. Surprisingly, the quantities for the dissolved platinum are different for almost

identical conditions in perchloric and sulfuric acids. It can be assumed that both acids have

similar pH for the same concentration, since the second dissociation constant of sulfuric acid is

1010 times smaller than the first one [144]. It has to be mentioned that the variation of the

concentration changes the pH value as well as the amount of the spectator species present in

the electrolyte. The nature of the electrolyte can therefore additionally affect the quantitative

behavior of the dissolution, while the qualitative behavior follows same trends in both acids.

Figure 7.8 The amount of dissolved platinum during potential cycling at a scan rate of

0.01 V/s between 0.05 VRHE and the corresponding upper potential in a) HClO4 and b) H2SO4

with different concentrations. The red and green color codes show the dissolved amount

during the positive and negative sweep for the corresponding concentration (�) 5M; (�) 1M;

(�) 0.1M and (�) 0.01M. All electrolytes are continuously purged with argon. (Reproduced

from ref. [A.15] with permission from The Royal Society of Chemistry)

Taking into account that protons are influencing the dissolution, a logical question appears

about the behavior in basic solutions. Figure 7.9 shows the dissolution amounts in 1 mM

sodium hydroxide, which is at the limit of the technical capabilities for the ICP-MS regarding


64

alkaline pH. Again the dissolution is obtained in the anodic and cathodic sweeps around

potentials for formation and reduction of the oxide respectively. The onset of the dissolution is

similar as in the case of acidic electrolytes at ca. 1.1 VRHE. The integrated amounts follow the

general trend of decreasing dissolution amount with increasing pH. The quantitative

description has to be considered criticality, however, since the dissolution profile is only slightly

above the quantification limit. Nevertheless the overall behavior is qualitatively not different

from the acidic conditions, the predominant dissolution occurs during the reduction of the

oxide. The amounts detected in the anodic sweep are again almost identical for different upper

potential limits. Another aspect that can affect the dissolution amounts is the solubility of the

platinum in electrolytes with different pH value. It is known that transport of Pt ions in acidic

media proceeds over complexing with anions and forming water-soluble product (e.g. [PtCl6]2-),

where in basic solutions hydroxylated complexes (e.g. PtOH+ and Pt(OH)2(aq)) play a major

role in transport within the electrolyte [26,89,145]. The solubility of the products in the acid

environment is significantly higher than in the basic one. However, the maximum values in the

dissolution profile remian three orders of magnitude lower than the solubility of Pt2+ in 1mM

NaOH (i.e. ca. 152 �g/L) [146]. This shows that the obtained results are not an artifact of the

experiment, but a real effect of change of the pH.

Figure 7.9 Overall dissolution amount during potential cycling from 0.05 VRHE to the

corresponding upper potential limit with scan rate of 0.01 V/s in 1mM NaOH/Ar. The black

color represent the amount for a single cycle, where the red and green color present the

amounts detected in the positive and negative potential sweep respectively.


65

7.6 Effect of the reactive gases on Pt dissolution

Platinum is typically used as a catalytic material and should ideally be stable during certain

operational conditions. Especially, the stability issues in the presence of reactive gases like

oxygen, hydrogen and carbon monoxide are of high relevance for the application in fuel cells.

A recent study of Kongkanand et al. suggested a certain influence of oxygen as a promoter for

the dissolution of platinum-based nanocatalyst during potential treatment using post analysis

[111]. This phenomenon is addressed in this work by performing identical experiments in

perchloric acid saturated with different gases. The comparison of the dissolution profiles for Ar,

O2, H2 and CO-saturated electrolyte is presented in figure 7.10. No quantitative or qualitative

difference can be observed in the dissolution profiles using electrolyte saturated with oxygen,

hydrogen and argon (i.e. ca. 7 and 8 ng/cm2cycle for the corresponding CVs up to 1.55 and

1.65 VRHE). Further variation of the potential window in oxygen-saturated solution showed the

same results as found in the case of argon (not shown). Therefore it can be concluded that

previously reported results in literature on Pt/C degradation are rather due to structural

changes than to dissolution of active material. Distinctively different is the dissolution behavior

during bulk CO oxidation, where a massive increase in the platinum amount is detected (ca. 25

and 37 ng/cm2cycle). Surprisingly, almost no anodic dissolution peak is observed and the

cathodic one appears ca. 100 mV more negative than in case of argon.

Figure 7.10 a) Applied experimental sequence of two cyclic voltammograms from 0.1 up to

1.55 and 1.65 VRHE with a scan rate of 0.01 V/s followed by open circuit potential

measurement; b) The corresponding dissolution profiles in 0.1M HClO4 saturated and

continuously purged with Ar, H2, O2 or CO on the same time scale. (Reproduced by

permission of The Electrochemical Society form ref. [A.16])


66

Further, the influence of the overpotential for the bulk CO oxidation is correlated with the

platinum dissolution signal. Figure 7.11 depicts a series of electrochemical CO oxidation, where

the upper potential limit is varied from 0.87 VRHE upwards. Once the potential window is

extended above ca. 1.1 VRHE platinum starts dissolving considerably, similar to the dissolution in

inert argon atmosphere, which indicates the important contribution of sub-surface oxide to the

overall dissolution also in this case. Again almost no dissolution signal is observed in the

positive going sweep for all potential ranges similar to the results in figure 7.10. This is

surprising considering that the CO is smoothening the platinum surface even at low potentials

below 0.8 VRHE by removing adislands [132,147]. Further increase of the overpotential causes

the increase in the dissolution amounts. This can be related to the fact that the anodic

extension of the potential window, leads to formation of more Pt-oxides, which can be partially

dissolved during subsequent reductive scan. The exact amount of oxide(s) cannot be extracted

from the electrochemical signal, due to the overlapping with the CO oxidation response. For all

potential windows, the onset of the dissolution in the cathodic potential sweep is observed at

ca. 0.85 VRHE what is very similar to the value for the initial CO oxidation.


67

Figure 7.11 a) Cyclic voltammograms from 0.15 VRHE to different vertex potential with a scan

rate of 0.01 V/s; b) The dissolution amounts per cycle vs. the corresponding upper potential; c)

Dissolution profile in the negative going potential sweep. (Reproduced by permission of The

Electrochemical Society form ref. [A.16])

CO oxidation is often used for the estimation of active surface area of platinum-based

catalyst, a method known as CO Stripping [132]. Figure 7.12 shows the expected platinum

losses for potential treatment during change of the used gases. Initially, the electrochemical

treatment in presence of argon shows identical behavior as previously discussed. A change of

the used gas to CO under potential control does not trigger any dissolution. A reference cyclic

voltammogram for CO bulk oxidation results in the same dissolved amount as in figure 7.11.

Consequently, the gas supply was switched back to argon and the electrolyte is purged for an

hour to remove the CO from the solution. The following CVs showed smaller dissolved


68

amounts than expected. In the first cycle only a cathodic dissolution peak is detected, mainly

due to the remaining, pre-adsorbed CO on the surface that is already removed in the negative

potential sweep. The second cycle has an increased dissolution and the typical dissolution

features reappear as for example in figure 7.2c. A close look at the electrochemical signal

reveals the origin for the reduced amount, namely residual CO is present in the electrolyte. A

complete removal of the CO from the electrolyte remains an issue for this SFC experimental

setup. Thus, a different experiment is performed, where the platinum sample is pretreated in a

bulk electrochemical cell with CO-saturated solution. The surface is then covered with CO by

holding the potential at 0.05 VRHE for 5 minutes. Afterwards the sample is taken out and

contacted with the SFC under potential control and three cyclic voltammograms are performed

as shown on figure 7.12cd. Surprisingly, the dissolution during the CO stripping procedure is

significantly lower than expected for CO bulk oxidation or under Ar atmosphere and in a good

agreement with the results in figure 7.12ab.

Figure 7.12 a) and c) Illustrates the applied potential with the applied gas purging sequence; b)

and d) Corresponding dissolution profile and integrated amounts for the detected peaks for an

investigation of a polycrystalline Pt sample pre-treated in a separate, CO-purged cell with

potential hold for 5 min at 0,05 VRHE, before investigation in the SFC under Ar. (Reproduced

by permission of The Electrochemical Society form ref. [A.16])

7.7 Enhancement of Pt dissolution in the presence of chlorides

Usage of platinum in chloride-containing electrolytes introduces additional complexity to

the process of dissolution. A few literature reports showed the quantification of the detected

dissolution rates under chronoamperometric conditions after several hour of operation [24,142].


69

It is known that chlorides promote platinum dissolution, but the mechanism of promotion is

still unclear. A series of experiments were performed with the SFC in perchloric acid with the

addition of different concentrations of chlorides to address this question. It is important to

mention that despite all experiments are conducted with Suprapure® chemicals from Merck, the

produced HClO4 is always accompanied with minimal amounts of Cl-. The original 70%

(i.e. 11.6 M) solution has less than 1 ppm Cl- by quality certificate, which means a “pure”

0.1M HClO4 as used before has no more than ca. 0.4 �M of Cl-. Figure 7.13 shows the

dissolution profile for various potential windows and the integrated amounts of dissolved Pt.

The general behavior seems to be unchanged with addition of Cl-. The dissolution is detected

again above ca. 1.1 VRHE in the positive and negative potential sweep direction, where the peaks

become only more emphasized with increasing Cl- content. This suggests that the chlorides

don’t change the dissolution pathway. As mentioned above, platinum forms a water-soluble

complex with Cl- ions, which can facilitate the removal of platinum ions from the vicinity of

the electrode. In this way a possible redeposition could be partially inhibited. Quantitatively, the

addition of small amounts of HCl in the order of a few 10-6 moles do not induce any

alternation from the dissolution in 0.1M HClO4, only for concentrations above 10 �M the

difference becomes significant. Note that the dissolution indicates a drop in the overall

amounts after a certain potential. This can be considered as an artifact of the measurement

possibly related to blocking of the surface. All dissolution amounts are normalized to the

geometric area, not to the effective/accessible one. For example in the case of addition of

0.1mM HCl, chlorine evolution reaction is observed in the cyclic voltammograms above

1.45 VRHE that can result in bubble formation and therefore decrease the accessible surface area.

Figure 7.13 a) Dissolution amounts of the platinum are plotted for cyclic voltammograms

from 0.05 VRHE to the various corresponding upper potential limits. 0.1M HClO4 was used with

addition of HCl and the electrolyte was continuously purged with argon. The black color


70

corresponds to the total dissolution in a single cycle, red color for the anodic sweep and the

green for the cathodic one.

7.8 Temperature dependence of the dissolution

Large-scale industrial applications of platinum as a catalyst are usually not at room

temperature. For example, a PEMFC stack or low-temperature electrolyzers are operated at

approximately 85°C [148]. All experiments performed in the previous sections are made in

temperature-controlled laboratories (at ca. 20°C). Figure 7.14 shows cyclo-voltammetric

measurements made at five different temperatures within a fixed potential window using a scan

rate of 0.01 V/s. The increase in the temperature is achieved using a circulator bath, which

heats up the electrolyte vessel as well as the sample itself within the same water circuit. An

intermediate period of 5 min is used to achieve a thermal equilibrium of the system before each

CV is recorded using sweep rate of 0.01 V/s. Surprisingly, the overall dissolved amounts are

decreasing with increasing temperature. While the dissolution peak during the oxide formation

is increasing with temperature and appears to be for all five cycles exactly at 1.1 VRHE as shown

in figure 7.14ab, an opposite behavior is observed for the negative potential sweep. Variation of

the temperature can affect not only the dissolution, but also the oxide formation. Thus the

amount of formed oxide is estimated from the voltammograms. A correlation between the

reductive charges from the negative going potential sweeps with the dissolved amount is

presented in figure 7.15.

Figure 7.14 a) Applied potential profile consisting of cycles between 0.05 and 1.6 VRHE in

0.1M H2SO4 for 25, 35, 45, 55 and 65 degree Celsius with intermediate potential hold at

0.5 VRHE for 300 sec to equilibrate the temperature; b) Corresponding spectrometric profile


71

with indicated dissolution onset; c) dissolution amounts for the positive and negative potential

sweeps, and the overall one within a single cycle plotted for different temperatures. [A.18]

For the same potential window, an increase in temperature results in an increased oxide

formation, but for a decrease in the overall dissolution per cycle. The distinction originates

from the unusual temperature dependence of dissolution in the cathodic sweep. In contrast to

the previous results, formation of more oxide at elevated temperatures leads to diminished

amount of dissolve platinum in the reductive scan. The difference in dissolution behavior

during positive and negative potential sweeps also explains the trend observed by several

groups for which during the potential treatment an increase in the dissolution amount was

detected whilst raising the temperature, suggesting the predominant contribution of the anodic

dissolution at high scan rates [91,92,149].

Figure 7.15 a) Relation between the reduction charge and the dissolution amount for different

temperatures (data from figure 7.14); b) Logarithm of the dissolution amounts from the anodic

and cathodic potential sweep plotted versus one over the temperature.

Furthermore, from the reaction rate law and Faraday’s law stems that the dissolution

amount should be linearly proportional to the rate e.g. the slope of the log(m(Pt)) vs. 1/T

should be the same. Namely, the rate constant could be expressed as the product of the

dissolved amount and proportionality factor that will affect only the offset in figure 7.15.

However, this is only under the assumption of presence of steady-state dissolution. In the case

of “anodically” dissolved platinum, an Arrhenius behavior can be observed and the activation

energy of the process can be estimated from the slope using following relation:

� � � �� or ��

��


72

where � is the rate constant of the process, �� is the activation energy of the process, � is the

universal gas constant (8.314 J/(K.mol)), � is the temperature in Kelvin and � is pre-

exponential factor. The activation energy for the dissolution in the anodic sweep is calculated

ca. 13 kJ/mol. This value is very close to the calculated 18 kJ/mol for Pt dissolution during

OER at high anodic potential by Ota et al. [80].

In the cathodic case is the observed non-Arrhenius behavior, which does not allow a direct

extraction of Ea (“negative” activation energy) and rather indicates a more complex pathway.

The proposed approach of Eyring–Polanyi can be helpful in treating complex reactions, where

the reaction rate is defined as [150]:

� � ��

where �� is the Boltzmann constant (1.38�10-23 J/K), � is the Planck’s constant (6.63�10-34

J/s) and �� is the Gibbs energy. The rate law can be further modified using the definition of

�� as:

� � ��

� �� or ��

� � ��

��

From figure 7.15b, the dissolution in the reductive scan has a positive slope that means the

partial derivative of the �� with respect ��should be bigger than zero:

� � � ��

��

��

This results that enthalpy of the reaction �� , where at room temperature

�� . The negative sign is indicating for the “cathodic” dissolution to be an

exothermic process.

Note, that the following evaluation is made under the assumption that dissolution proceed

for the “anodic” and “cathodic” route within the same time frame under steady-state condition,

where the detected amount should be linearly propositional to the rate. However the previous

results indicate rather a transient nature of the dissolution. Thus the value of the calculated

activation energy and the trend Arrhenius/non-Arrhenius behavior of the dissolution processes,

have to be considered critical.

7.9 Ex-situ XPS and SKP investigation of platinum oxide

Understanding the nature of the formed platinum oxides has an essential role for resolving

the dissolution mechanism. Although platinum is an extensively studied material, there is still a


73

lot of uncertainty concerning the chemical structure of the electrochemically formed oxide layer.

The opinions spread from hydroxide species to different combinations of hydrous oxides

containing various oxidation states [139]. Pure electrochemical investigations have their

limitations in clarifying the oxide and thus require complementary techniques. For this purpose,

an oxide layer has been electrochemically prepared by first cleaning a Pt foil in 0.1M HClO4 by

several cycles between 0.05 and 1.6 VRHE. Afterwards, it was polarized at 0.27 VRHE for 10 min.

Under potential control, the foil is partially pulled out of the solution and then the potential is

changed to 1.27 VRHE for 10 min. The same procedure is done once again with polarization at

1.97 VRHE for 10 min. Finally, the sample is taken completely out of the electrolyte, dried in a

nitrogen stream and introduced into the translation chamber of the X-ray photoelectron

spectrometer (XPS).

The reference measurement performed on a polycrystalline platinum surface gives the

typical double peak response at 71.1 eV and 74.4 eV for Pt4f as shown in figure 7.16a, which is

in good agreement with literature values [151]. The second area polarized at relatively low

overpotential for the oxide formation (ca. 1.27 VRHE), shows the characteristic fingerprint of

PtO2 at 75.1 eV and 78.4 eV, in line with previous reports [56–58,152]. The formation of a

Pt(IV) state at this “early” stage of oxidation is not directly intuitive, but thermodynamically

possible [34]. This can be correlated to observed passivity in section 7.2, where potential

cycling within the stability window of water does not lead to significant dissolution. In the third

case, the sample is anodically polarized in the region of intensive oxygen evolution reaction. No

clear stoichiometry for PtO or PtO2 can be derived from the XPS spectra. The fitted data

indicates a mixed oxide layer that consists of both oxidation states, +2 and +4. Unexpectedly,

the intermediate oxidation state of PtO appears in the region of OER, where transpassive

dissolution starts to appear again [24,79].


74

Figure 7.16 Pt-4f X-ray photoelectron spectra measured at location initially polarized for

10 min at: a) 0.27 VRHE; b) 1.27 VRHE and c) 1.97 VRHE; The red curve represents the measured

data, other color represent the fittings generated by CASA software.

Probing the surface of the Pt/PtxOy substrate is of crucial importance for recognizing

surface states involved in the dissolution mechanism proceeding on the electrode-electrolyte

interface. XPS spectra provide information about the average composition over the top layers.

Note, however, that these results have to be considered with caution for the discussion, as the

oxidation states might change firstly upon drying after the polarization, and secondly upon

introduction into the vacuum chamber for analysis.


75

In addition to XPS, Scanning Kelvin Probe (SKP) was utilized to investigate the change of

the average electron energy after different polarizations (see Figure 7.17). The work function

can be evaluated in different ways [153], in this work is used the following relation:

� ��

that was extracted from the absolute potential scale introduced in Bard et al. [27] (pp.54). The

initial surface scan over the reduced area gives almost an identical value of 5.51 eV that is

relatively close to the previously reported 5.4 eV by Trasatti [153]. A slight variation in the

comparison with literature is predominantly related to the use of different equations.

Polarization at 1.1 VRHE leads to a significant drop of the work function related to the change

of the dipole orientation of Pt-O on the surface during the place-exchange mechanism. Further

positive extension of the applied potential again increases the value of �, which is expected for

oxide growth. The same reduced platinum sample was introduced under CO stream and

consequently measured by SKP, a significant drop is observed to �� in

comparison to the “free” surface (5.51 eV). It is known from measurements in UHV systems

that the adsorbed CO can lead to a decrease up to ca. 1.1 eV [154].

Figure 7.17 Scanning Kelvin Probe measurements made on a polycrystalline platinum surface

pretreated in 0.1M HClO4/Ar at indicated potential values for 10 min in dried nitrogen

atmosphere.

While the electrochemically formed oxides are typically formed quite fast, all the SKP

measurements are made ex-situ after long-term polarization. Thus a one-to-one correlation to

the dissolution results in the previous chapters has to be considered critically. Moreover, the

tests performed in humidified oxygen atmosphere in the SKP showed a fluctuating response,

which are expected considering the adjustment of the oxide layer to the different environment.


76

Finally, after a certain time the work function approaches the one corresponding to pure

platinum, related to instability of the oxide layer. This remains as an issue of ex-situ

measurements.

Chapter 8: Comprehensive discussion

77

8 Comprehensive discussion Platinum dissolution is closely related to the electrochemical formation and reduction of the

oxide layer as shown in several experimental series in chapter 7. It can be described as a

transient process taking place during a potential change with time around ca. 1.1 VRHE, the

critical potential for subsurface oxide formation. The results furthermore indicate that

dissolution is detected during positive and negative polarization, where the second one prevails

mainly with respect to the overall amount per cycle. While the “cathodic” dissolution indicated

a strong influence on the applied potential window, time scale, pH and anion concentration,

the “anodic” case follows a rather moderate dependence on the experimental parameters. To

grasp the meaning beyond the presented phenomenological relations, first of all, the most

important finding are summarized in table 8.1.

PPaarraammeetteerrss ““AAnnooddii cc”” ddii ssssoo lluutt iioonn ““CCaatthhooddii cc”” ddii ssssoo lluutt iioonn

Time dependence transient process transient process

Anodic limit minor influence significant increase with

more positive anodic limit

Scan rate minor influence significant increase with

decreasing scan rate

Cathodic limit minor influence significant increase with

more negative cathodic limit

pH increase with drop in pH significant increase with

decreasing in pH

Temperature increase with higher

temperatures

decrease with higher

temperatures

Table 8.1. Distinction behavior between “anodic” and “cathodic” dissolution behavior in

relation to the experimental parameters.

Due to the possible interrelation between oxide layer formation and the dissolution process,

it is important to revisit some of the existing concepts for Pt oxidation. During anodic

polarization, initial oxidation of the Pt surface in the region between 0.8-0.9 VRHE is described

as OH adsorption or alternatively as direct adsorption of O2-. Based on the analysis of Conway


78

and co-workers, it can be assumed that a full coverage of one with oxygenated species is

achieved at about 1.1 VRHE [40]. However, Bockris et al. considered a maximal surface coverage

with oxygen containing species to be around 0.25 of a full monolayer at equilibrium conditions

[155]. Below the critical coverage, the enthalpy of formation of the chemisorbed phase is

higher than the enthalpy of formation of the oxide. As the oxygen coverage approaches the

critical value, the repulsive interactions between the oxygenated species gradually reduce the

enthalpy of chemisorption until, at the critical coverage, it becomes equal to the enthalpy of

formation of the oxide film. Beyond the critical coverage, repulsive interactions in the densely

packed electronegative O adlayer induce surface reconstruction into a more energetically

favorable configuration by occupation of the sub-surface sites, known as place exchange

[41,98,156]. Of essential importance is to notice that the transition between the O chemisorbed

phase and the appearance of an oxide film is a thermodynamically, and not a kinetically, driven

process. Taking into consideration the observations of Conway et al. that around 1.05-1.1 VRHE

the amount of adsorbed species corresponds to the full monolayer and that a critical coverage

according to Bockris et al. is around 0.25 of the full monolayer, it seems that place exchange

starts after reaching the critical coverage, while additional adsorption until a full monolayer

occurs via sub-surface oxide formation [139,155]. It is important to mention, that independent

of the definition of full coverage, it is generally accepted that formation of sub-surface oxide

initiates at ca. 1.1 VRHE on polycrystalline platinum.

During the online trace analysis, no dissolved platinum was detected between 0.8-0.9 VRHE,

in the region in which the adsorption of oxygenated species on the surface begins. The

determined onset potential for platinum dissolution was ca. 1.1 VRHE. This means that

dissolution is triggered in the moment when the critical coverage is reached. Above 1.1 VRHE,

three processes are possible, adsorption of oxygen containing species, oxide growth and Pt

dissolution. Anodic polarization causes a downshift of the Fermi level and induces an increase

in the work function. To equilibrate, the system tends to reduce the work function by injecting

of electrons from adsorbed oxygen containing species towards the electrode and by inversing

the direction of the Pt-O dipoles by further place exchange between Pt atoms (Ptex) in the

lattice and oxygen containing species on top, as shown experimentally in figure 7.17. As a

consequence further oxidation causes an irreversible roughening of the surface. Platinum atoms

being pushed up/away from their original position in the lattice structure and O/OH moving

below the platinum atoms results in Pt atoms weakly bond to the bulk platinum lattice [157],

which makes them susceptible to dissolution. The existence of an equilibrium between the

process of adsorption of oxygen containing species and the process of oxide growth is a


79

potential explanation why the typical feature of the cyclic voltammograms of Pt in acidic media,

between 1.05 and 1.55 VRHE is a plateau with clear absence of an increase of a faradaic current

(figure 7.1). Thermodynamically the oxidation of the platinum surface can be expressed by the

following equations:

�� (8.1)

�� (8.2)

�� (8.3)

The experimentally applied anodic potentials in this work are predominantly below 1.8 VRHE

thus the formation of �� is disregarded. Two states are expected, namely �� and ��,

where the second one should be present over a broader potential range. However the

experimental investigations of the oxide layer have shown a complex bilayer structure, as

discussed in chapter 3. In a similar potential range direct electrochemical dissolution is expected

to proceed in parallel to the chemical dissolution:

�� (8.4)

�� (8.5)

A list of models has been proposed in the past to describe the oxide growth on platinum,

while only a few of them incorporate the dissolution phenomena. One of the most famous

models for the platinum dissolution is proposed by Darling and Meyer [112]. It considers the

dissolution as competition between dissolution (reaction 8.4) and formation of chemically

stable oxide (reaction 8.1). Namely, as long as the surface is not fully covered with passive

oxide species, platinum can be dissolved into the electrolyte by direct oxidation. Figure 8.3

illustrates schematically the change of the oxide coverage during cyclic voltammetry as

suggested by this partial oxide coverage model. During partial coverage the dissolution should

reach a maximum, once full coverage is achieved, no dissolution should be observed anymore.

There are, however, several incoherencies with the experimental results. First, experimentally

dissolution is only observed after full oxide coverage has been reached. Secondly, an increase of

the upper potential limit (beyond 1.1 VRHE) leads to a higher dissolved amount, which cannot

be explained from this model as shown schematically in figure 8.1. Furthermore, variation of

the scan rate should affect the dissolved amount in the negative as well as in the positive

potential sweep, which is however different from the experimental fact (table 8.1) that only the

“cathodic” route is influenced.


80

Figure 8.1 Illustration of change of oxide coverage (limit by 1 with respected to the model of

Darling and Meyer [112]) during potential cycling within different potential windows. The

green rectangles represent the regions of dissolution.

Platinum forms only a few monolayers of oxide film consisting of two different kinds of

composition: a thin inner one consists of platinum in the second oxidation state (also called -

oxide) and a thicker outer one (�-oxide) (figure 8.2). An alternative approach describing the

oxide growth in the framework of a possible dissolution is the point defect model (PDM)

suggested by Macdonald et al. [43,68]. It proceeds over initial formation of a compact passive

film (-oxide on figure 8.2), where the consequent oxide formation undergoes over

precipitation of hydroxide species on the outer oxide/electrolyte interface. A disturbance of the

oxide layer from its equilibrium state leads to a depassivation of the surface. Following the

point defect model, this will induce movement of metal ions across the oxide layer, and cause a

partial dissolution. Within the PDM, a positive linear polarization will cause more or less

continuous dissolution once the outer oxide layer has been formed. However, the

chronoamperometric series performed in the region of oxide formation did not lead to

dissolution, which speaks against the beliefs of the PDM. Also a variation of the time scale

should influence the “anodically” dissolved amount, which is contradicted by the results in this

work.

��

��

��

��

��

��

��

��

��

��

��

��

��


81

Figure 8.2 Illustration of change of the surface state during anodic polarization 1-4 and the

cathodic one 4-6, including the intermediate steps during the oxide formation in 3-4.

The experimental results indicate 1.1 VRHE as an initial potential for the dissolution for all

used acidic solutions as well as for 1 mM NaOH, suggesting a proton dependent reaction to be

involved. The direct electrochemical dissolution (8.4) should proceed 88 mV more positive, so

that ca. 1.04 mM Pt2+ (equal to ca. 202.8 mg/L) are required in the vicinity of the electrode to

shift the equilibrium potential to 1.1. The effectively measured concentrations are several

orders of magnitude smaller, typically in the lower range of �g/L. Nevertheless, the equilibrium

potential of reaction 8.4 could be different for lower coordinated sites. Alternatively, a chemical

dissolution of stoichiometric �� could cause the platinum loss, as shown in reaction 8.5.

Nevertheless, no constant dissolution could be explicitly detected in region 1.1 to 1.2 VRHE or

above, as the spectrometric signal decreases for chronoamperometric measurements below the

detection limit.

Considering all the results it is clear that the underlying oxide formation and reduction is

determining the dissolution processes of Pt. However, thermodynamic considerations fail to

explain the mechanism in detail. Thus a kinetic model becomes necessary to describe both, the

anodic and cathodic dissolution. During anodic polarization a change in the electrode potential

causes a disturbance of the equilibrium between oxide formation and adsorption of O/OH

species. The downshift of the Fermi level (i.e. increase of the work function) will be manifested

instantly by an increase of the coverage and a re-establishment of equilibrium by new place

exchange. At these positive potentials, the Ptex atoms/ions formed due to place-exchange can

be easily passivated by adsorption of additional O/OH species. However, minor amounts of

the formed Ptex sites are also prone to dissolution. In particular, it can be assumed that Pt sites

that are already low-coordinated before oxidation are more susceptible than Pt sites from facets.

��

��

� ��

��

�"��

��

�#��

��

�$��!��

��

��

��

��

��

��

��

��

��

��

��


82

This could explain on the one hand that the amount of Pt dissolved during the anodic

dissolution is rather constant and independent from pH, potential limits and scan rate, as the

amount of low-coordinated sites in the initial state does not vary extensively. Of course, more

extended single-crystal studies will be necessary to confirm this in future. Moreover, the drop

of the dissolution rates to values below the detection limit during constant polarization can be

explained based on this. Once the Ptex sites from the outer surface of the Pt-oxide are removed

and a quasi steady-state is reached, no further place-exchange takes place and the remaining Ptex

are stabilized in the surface oxide. It could be expressed as chemical reaction:

�� (8.6)

In contrast to the “anodic” dissolution, the dominating “cathodic” dissolution is

qualitatively different. It appears below ca. 1.05 VRHE during the reduction of the surface after

potential excursion above 1.1 VRHE, and strongly depends on several experimental parameters.

An almost linear relation was obtained between the dissolved Pt and the amount of formed

oxide, with a slope depending on the time scale of the experiment. The surface state at

potentials above 1.3 VRHE can be assumed with simplified �� stoichiometry, taking into

account the XPS data from figure 7.16. A negative sweep will lead to protonation of the oxide

layer and changes in the passivity of the layer. The processes can be expressed with following

reactions:

�� (8.7)

�� (8.8)

�� (8.9)

Chemical dissolution of Pt/Pt4+ is unlikely, considering that no evidence was found in acidic

media and also most of the literature reports Pt2+ as the soluble product in the electrolyte [26].

The electrochemical reductive dissolution (8.8) is expected thermodynamically to proceed at:

�� An order estimation of the effective measured Pt concentration gives 10-8 molar (ca. 2 �g/L),

thus the equilibrium potential can be written as:

��

A shift of the potential with 118 mV with each pH unit will give a much lower value than

experimentally observed (ca. 1.05 VRHE), but is certainly thermodynamically possible in very

acidic environments (above 1M). Moreover, the reaction 8.8 requires two electrons and four

protons or alternatively it can be split into two consecutive steps. Initial reduction of �� to

�� (8.2) followed by chemical dissolution of �� (8.5). A possible dissolution pathway in this

case is illustrated in figure 8.3.


83

Figure 8.3 Illustration of reductive reaction pathway/s including solvable intermediate.

Taking into consideration that there is no significant cathodic dissolution before reaching

1.05 VRHE it can be assumed that the predominant form of surface oxide above 1.05 VRHE is

�� (before region of intense OER). It seems that �� (or ��) is a crucial intermediate

and the chemical dissolution is an unavoidable part of the overall picture. It has to be

mentioned that reaction 8.5 (or 8.6) is based on protonation of �� and is independent on the

electrode potential. Comparing the oxide reduction and the dissolution charges (figure 7.7e),

the predominant fraction of the formed �� is reduced to ��, while only a much smaller

portion is chemically dissolved. The low amount of dissolved platinum can be related to

possible re-deposition of the dissolved species in the vicinity of the electrode, which is

reasonable considering the low potentials during the reduction process. Moreover, the

influence of re-deposition on the cathodic dissolution can also explain the decreasing amount

of Pt with increasing scan rate, as it depends on the effective diffusion from the surface into

the bulk electrolyte.

Important additional hints to the dissolution mechanism are provided by the CO stripping

and CO bulk oxidation experiments. It is known that CO smoothens the platinum surface by

increasing the surface diffusion and therefore reducing the amount of defect sites and adatoms

[107]. A novel finding is that anodic dissolution is vanishing during CO bulk oxidation and

almost completely disappears for CO stripping. This confirms the influence of low coordinated

sites or their lack in case of anodic dissolution (like Ptex). In addition, the adsorbed CO can also

physically block the surface from possible redeposition in the cathodic sweep and promote in

this way also the dissolution amount during CO bulk oxidation, whereas the absence of CO in

the solution give smaller dissolved amount for the stripping experiment. Once the surface is

smoothened, the cathodically dissolved Pt diminishes even less than in case of an Ar saturated

solution for the same conditions (figure 7.12d). Further potential cycling increases the

roughness of the surface already in the following cycle, and the expected dissolution is soon

reestablished.

PtO2� PtO� Pt�

Pt4+�

+2H++2e-�

Pt2+�

+2H++2e-�

12

3

4

5

6


84

In summary, the oxide formation leads to lattice distortion of the platinum interface due to

the place-exchange mechanism, creating higher amounts of Ptex that can chemically dissolve. At

anodic polarization above 1.1 VRHE, the high driving force for the adsorption of O/OH leads

to a quick passivation of the surface followed by a transition to a higher oxidation state. During

this transition some limited amount of dissolved platinum originating from low coordinated

sites occurs, i.e. anodic dissolution, which is almost independent from the experimental

parameters. A more or less opposite picture appears during the reduction of the oxide layer

where the transition between different oxidation states and low overpotential for O/OH

adsorption leads to de-passivation of the surface, where anodically formed �� can be

dissolved before a full reduction of the surface. However the exact surface state of the

platinum oxide layer during electrochemical treatments remains under discussion and will be

the key to completely resolve the picture on the dissolution mechanism.

Chapter 10: References

85

9 Summary and outlook This work presented describes a comprehensive study on platinum dissolution mostly in

acidic media: from design and implementation of new experimental equipment through

extended investigation of the dissolution behavior to finally a tentative model on the

mechanism. In the initial stage a universal modular software approach based on an extended

consumer/producer design pattern with asynchronous control and acquisition over different

hardware components was developed. The software architecture incorporates smart algorithms

for automatic execution of electrochemical high-throughput and combinatorial experimental

series. This technical achievement was adopted in the development of five automated Rotating

Disc Electrode systems (as black box solution) and seven Scanning Flow Cell backbones (as a

basic platform for further development). The developed software with its modular approach

provides a solid base for further modifications and extensions of the functionality by including

new procedures and/or incorporation of other hardware equipment without affecting the main

programming logic or architecture. The source code and the implemented logic are

documented and can be reused by the successors in follow-up studies.

The innovative design and functionality of the Scanning Flow Cell enables straightforward

integration of analytical methods for online-monitoring of the electrolyte constituents during

electrochemical measurements. Especially the coupling with inductively coupled plasma mass

spectrometer (ICP-MS) and the further optimization of the system performance allow online

detection of dissolved species in the sub-monolayer region. The reliability of the developed

software/system was verified in several individual steps and the long-term performance was

evaluated over 5 days of degradation test on Pt/C catalysts. Additionally, a proof-of-concept

was achieved for the combined SFC/ICP-MS technique on a copper bulk material, used as a

model system, where the direct correlation between the applied galvanostatic sequence and

measured spectrometric signal was shown. Thus, not only a qualitative but also a quantitative

description can be presented for an estimation of corrosion rates.

In the second stage of this project, a targeted screening of the experimental parameters was

performed to determine which of them have a crucial influence on the platinum dissolution

and to which extent. It was found that a transition between the reduced and oxidized surface

state around the potential of sub-surface oxide formation (e.g. 1.1 VRHE) is triggering significant

platinum loss. Clear separation of the dissolution during the anodic and cathodic polarization

was demonstrated in correlation with the amount of formed oxide, time scale of the


86

experiment, pH and temperature. Two potential windows of electrochemical treatment have

been emphasized within which the dissolution rate is negligible. The possible dissolution

pathways were revisited and discussed intensively in the previous chapter 8 in relation to the

new findings. The effect of the reactive gases on the dissolution of polycrystalline Pt was

presented for the first time. Improved understanding of the dissolution phenomena of

platinum is of high importance from the fundamental point of view as well as from applied

perspectives. The gained knowledge about stability issues of Pt is determining also its

application limits, and provides guidelines for optimization of operation conditions of

platinum-based materials. The extended quantitative description of the dissolution rate can be

further used for engineering purposes to predict Pt losses during various experimental

conditions investigated in this work.

For the complete resolving of the dissolution mechanism of platinum, however, a missing

puzzle part remains the surface state during the electrochemical treatment. Only with in-situ

experiments in a comparable electrochemical environment, it will be possible to answer this

question in future works.

During this work, also several parallel studies on noble metals and bimetallic compositions

were performed and published that explicitly show broad range of application of the developed

methodology. The novel electrochemical scanning flow system (SFC) can be further extended

with alternative analytical tool. One example is an ongoing project on the investigation of the

activity and selectivity of electrocatalytic materials with a combined SFC and DEMS system.

Overall, the ability of parallel monitoring over several parameters like stability, activity and

selectivity provides a powerful tool for systematic investigation of real systems like fuel cell

catalyst and electrode materials.


87

References [1] D. McDonald, A history of platinum and its allied metals, Johnson Matthey :

Distributed by Europa Publications, London, 1982.

[2] P.M.D. Collins, The Pivotal Role of Platinum in the Discovery of Catalysis, Platinum

Metals Rev. 30 (1986) 141–146.

[3] W.R. Grove, On a Gaseous Voltaic Battery, Phil. Mag. 21 (1942) 417–420.

[4] J. Renn, R. Schlo�gl, H.-P. Zenner, Herausforderung Energie ausgewählte Vorträge der

126. Versammlung der Gesellschaft Deutscher Naturforscher und Ärzte e.V., Ed. Open Access,

Berlin, 2011.

[5] EWEA, Wind power 2012 European statistics, The European Wind Energy

Association (EWEA), 2013.

[6] B. Sørensen, Renewable energy conversion, transmission, and storage,

Elsevier/Academic Press, Amsterdam ; Boston, 2007.

[7] J.H. Hirschenhofer, D.B. Stauffer, R.R. Engleman, M.G. Klett, Fuel Cell Handbook,

7th ed., Parsons Corporation, 2004.

[8] M. Lefevre, E. Proietti, F. Jaouen, J.-P. Dodelet, Iron-Based Catalysts with Improved

Oxygen Reduction Activity in Polymer Electrolyte Fuel Cells, Science. 324 (2009) 71–74.

[9] E. Proietti, F. Jaouen, M. Lefèvre, N. Larouche, J. Tian, J. Herranz, et al., Iron-based

cathode catalyst with enhanced power density in polymer electrolyte membrane fuel cells, Nat.

Commun. 2 (2011) 416.

[10] G. Wu, K.L. More, C.M. Johnston, P. Zelenay, High-Performance Electrocatalysts for

Oxygen Reduction Derived from Polyaniline, Iron, and Cobalt, Science. 332 (2011) 443–447.

[11] F. Jaouen, E. Proietti, M. Lefèvre, R. Chenitz, J.-P. Dodelet, G. Wu, et al., Recent

advances in non-precious metal catalysis for oxygen-reduction reaction in polymer electrolyte

fuel cells, Energy Environ. Sci. 4 (2011) 114.

[12] Z. Chen, D. Higgins, A. Yu, L. Zhang, J. Zhang, A review on non-precious metal

electrocatalysts for PEM fuel cells, Energy Environ. Sci. 4 (2011) 3167.

[13] Handbook of fuel cells: fundamentals, technology, and applications, Wiley, Chichester,

England ; Hoboken, N.J, 2003.

[14] Y. Shao-Horn, W.C. Sheng, S. Chen, P.J. Ferreira, E.F. Holby, D. Morgan, Instability

of Supported Platinum Nanoparticles in Low-Temperature Fuel Cells, Top. Catal. 46 (2007)

285–305.


88

[15] A.B. Yaroslavtsev, Y.A. Dobrovolsky, N.S. Shaglaeva, L.A. Frolova, E.V. Gerasimova,

E.A. Sanginov, Nanostructured materials for low-temperature fuel cells, Russ. Chem. Rev. 81

(2012) 191–220.

[16] C.E. Carlton, S. Chen, P.J. Ferreira, L.F. Allard, Y. Shao-Horn, Sub-Nanometer-

Resolution Elemental Mapping of “Pt3Co” Nanoparticle Catalyst Degradation in Proton-

Exchange Membrane Fuel Cells, J. Phys. Chem. Lett. 3 (2012) 161–166.

[17] K. Schlo�gl, M. Hanzlik, M. Arenz, Comparative IL-TEM Study Concerning the

Degradation of Carbon Supported Pt-Based Electrocatalysts, J. Electrochem. Soc. 159 (2012)

B677.

[18] J.C. Meier, I. Katsounaros, C. Galeano, H. Bongard, A.A. Topalov, A. Kostka, et al.,

Stability investigations of electrocatalysts on the nanoscale, Energy Environ. Sci. (2012).

[19] N. Hodnik, M. Zorko, M. Bele, S. Ho�evar, M. Gaber��ek, Identical Location

Scanning Electron Microscopy: A Case Study of Electrochemical Degradation of PtNi

Nanoparticles Using a New Nondestructive Method, J. Phys. Chem. C. 116 (2012) 21326–

21333.

[20] Johnson Matthey PLC, Platinum 2013, Johnson Matthey PLC, 2013.

[21] C. Hagelüken, Markets for the catalyst metals platinum, palladium and rhodium, Metall.

60 (2006) 31–42.

[22] H. Abe, Current Status and Future of the Car Exhaust Catalyst, QUARTERLY

REVIEW. 39 (2011) 21–31.

[23] j. Tafel, B. Emmert, The Cause of Spontaneous Depression of the Cathode Potential

During Electrolysis of Diluted Sulfuric Acid, Z Phys Chem. 52 (1905) 349–373.

[24] A.N. Chemodanov, Y.M. Kolotyrkin, M.A. Dembrovski, Isledovanie procesa

raztvorenia platiny w kislotnyh elektrolitah pri razlichnyh polyarizacizah, Elektrohimia. 6 (1970)

460–468.

[25] D.C. Johnson, D.T. Napp, S. Bruckenstein, A ring-disk electrode study of the

current/potential behaviour of platinum in 1.0 M sulphuric and 0.1 M perchloric acids,

Electrochimica Acta. 15 (1970) 1493–1509.

[26] L. Xing, G. Jerkiewicz, D. Beauchemin, Ion exchange chromatography coupled to

inductively coupled plasma mass spectrometry for the study of Pt electro-dissolution, Anal.

Chim. Acta. 785 (2013) 16–21.

[27] A.J. Bard, Electrochemical methods: fundamentals and applications, 2nd ed, Wiley,

New York, 2001.

[28] P.W. (Peter W. Atkins, Physical chemistry, W H Freeman, New York, 2006.


89

[29] Corrosion mechanisms in theory and practice, 2nd ed., rev. and expanded, Marcel

Dekker, New York, 2002.

[30] H. Kaesche, Die Korrosion der Metalle: physikalisch-chemische Prinzipien und

aktuelle Probleme, Springer, Heidelberg [u.a., 2011.

[31] Corrosion and oxide films, Wiley-VCH, Weinheim, 2003.

[32] M. Wohlfahrt-Mehrens, J. Heitbaum, Oxygen evolution on Ru and RuO2 electrodes

studied using isotope labelling and on-line mass spectrometry, J. Electroanal. Chem. Interfacial

Electrochem. 237 (1987) 251–260.

[33] O. Diaz-Morales, F. Calle-Vallejo, C. de Munck, M.T.M. Koper, Electrochemical water

splitting by gold: evidence for an oxide decomposition mechanism, Chem. Sci. 4 (2013) 2334.

[34] M. Pourbaix, Atlas of electrochemical equilibria in aqueous solutions., Nat’L Assoc. Of

Corrosion, [S.l.], 1974.

[35] B.E. Conway, G. Jerkiewicz, Relation of energies and coverages of underpotential and

overpotential deposited H at Pt and other metals to the “volcano curve” for cathodic H2

evolution kinetics, Electrochimica Acta. 45 (2000) 4075–4083.

[36] J. Greeley, T.F. Jaramillo, J. Bonde, I. Chorkendorff, J.K. Nørskov, Computational

high-throughput screening of electrocatalytic materials for hydrogen evolution, Nat. Mater. 5

(2006) 909–913.

[37] F.G. Will, Hydrogen Adsorption on Platinum Single Crystal Electrodes, J.

Electrochem. Soc. 112 (1965) 451.

[38] M. Hayes, A.T. Kuhn, Determination of platinum catalyst surface area with

potentiodynamic techniques — Effect of experimental parameters, Appl. Surf. Sci. 6 (1980) 1–

14.

[39] A.N. Frumkin, O.A. Petrii, Topics in pure and applied electrochemistry, Society for the

Advancement of Electrochemical Science and Technology , Karaikudi, India, 1975.

[40] H. Angerstein-Kozlowska, B.E. Conway, W.B.A. Sharp, The real condition of

electrochemically oxidized platinum surfaces, J. Electroanal. Chem. Interfacial Electrochem. 43

(1973) 9–36.

[41] K.J. Vetter, J.W. Schultze, The kinetics of the electrochemical formation and reduction

of monomolecular oxide layers on platinum in 0.5 M H2SO4: Part II. Galvanostatic pulse

measurements and the model of oxide, J. Electroanal. Chem. Interfacial Electrochem. 34

(1972) 141–158.


90

[42] K.J. Vetter, J.W. Schultze, The kinetics of the electrochemical formation and reduction

of monomolecular oxide layers on platinum in 0.5 M H2SO4: Part I. Potentiostatic pulse

measurements, J. Electroanal. Chem. Interfacial Electrochem. 34 (1972) 131–139.

[43] C.Y. Chao, A Point Defect Model for Anodic Passive Films, J. Electrochem. Soc. 128

(1981) 1187.

[44] D.D. Macdonald, The Point Defect Model for the Passive State, J. Electrochem. Soc.

139 (1992) 3434.

[45] N. Cabrera, N.F. Mott, Theory of the oxidation of metals, Reports Prog. Phys. 12

(1949) 163–184.

[46] A. Damjanovic, Temperature Study of Oxide Film Growth at Platinum Anodes in

H[sub 2]SO[sub 4] Solutions, J. Electrochem. Soc. 127 (1980) 874.

[47] D. Gilroy, Oxide growth at platinum electrodes in H2SO4 at potentials below 1.7 V, J.

Electroanal. Chem. Interfacial Electrochem. 71 (1976) 257–277.

[48] S. Shibata, Conductance measurement of thin oxide films on a platinum anode,

Electrochimica Acta. 22 (1977) 175–179.

[49] S. Shibata, Kinetics and mechanism of electrochemical reduction of multilayer oxides

on a smooth platinum electrode surface in acidic electrolyte, J. Electroanal. Chem. Interfacial

Electrochem. 89 (1978) 37–58.

[50] V.I. Birss, A. Damjanovic, A Study of the Anomalous pH Dependence of the Oxygen

Evolution Reaction at Platinum Electrodes in Acid Solutions, J. Electrochem. Soc. 130 (1983)

1694.

[51] V.I. Birss, M. Chang, J. Segal, Platinum oxide film formation—reduction: an in-situ

mass measurement study, J. Electroanal. Chem. 355 (1993) 181–191.

[52] B.E. Conway, G. Tremiliosi-Filho, G. Jerkiewicz, Independence of formation and

reduction of monolayer surface oxide on Pt from presence of thicker phase-oxide layers, J.

Electroanal. Chem. Interfacial Electrochem. 297 (1991) 435–443.

[53] S.J. Xia, V.I. Birss, Hydrous oxide film growth on Pt. I. Type I �-oxide formation in

0.1M H2SO4, Electrochimica Acta. 44 (1998) 467–482.

[54] A. Damjanovic, Oxide Growth at Platinum Anodes with Emphasis on the pH

Dependence of Growth, J. Electrochem. Soc. 126 (1979) 555.

[55] D. Gilroy, B.E. Conway, Surface oxidation and reduction of platinum electrodes:

Coverage, kinetic and hysteresis studies, Can. J. Chem. 46 (1968) 875–890.


91

[56] J.S. Hammond, N. Winograd, XPS spectroscopic study of potentiostatic and

galvanostatic oxidation of Pt electrodes in H2SO4 and HClO4, J. Electroanal. Chem.

Interfacial Electrochem. 78 (1977) 55–69.

[57] K.S. Kim, N. Winograd, R.E. Davis, Electron spectroscopy of platinum-oxygen

surfaces and application to electrochemical studies, J. Am. Chem. Soc. 93 (1971) 6296–6297.

[58] G.C. Allen, P.M. Tucker, A. Capon, R. Parsons, X-ray photoelectron spectroscopy of

adsorbed oxygen and carbonaceous species on platinum electrodes, J. Electroanal. Chem.


[59] B.V. Tilak, B.E. Conway, H. Angerstein-Kozlowska, The real condition of oxidized pt

electrodes, J. Electroanal. Chem. Interfacial Electrochem. 48 (1973) 1–23.

[60] A. Damjanovic, Growth of Oxide Films at Pt Anodes at Constant Current, Density in

H[sub 2]SO[sub 4], J. Electrochem. Soc. 122 (1975) 471.

[61] L.B. Harris, Initial Anodic Growth of Oxide Film on Platinum in 2N H[sub 2]SO[sub

4] under Galvanostatic, Potentiostatic, and Potentiodynamic Conditions: The Question of

Mechanism, J. Electrochem. Soc. 122 (1975) 593.

[62] H. Angerstein-Kozlowska, B.E. Conway, Evaluation of rate constants and reversibility

parameters for surface reactions by the potential-sweep method, J. Electroanal. Chem.


[63] M. Peuckert, H.P. Bonzel, Characterization of oxidized platinum surfaces by X-ray

photoelectron spectroscopy, Surf. Sci. 145 (1984) 239–259.

[64] L. Kövér, C. Ujhelyi, D. Berényi, D. Varga, I. Kádár, á. Kövér, et al., X-ray

photoelectron spectroscopic investigation of electrochemically oxidised and reduced platinum

surfaces, J. Electron Spectrosc. Relat. Phenom. 14 (1978) 201–214.

[65] T. Dickinson, A.F. Povey, P.M.A. Sherwood, X-ray photoelectron spectroscopic

studies of oxide films on platinum and gold electrodes, J. Chem. Soc. Faraday Trans. 1 Phys.

Chem. Condens. Phases. 71 (1975) 298.

[66] A.K.N. Reddy, M.A. Genshaw, J.O. Bockris, Ellipsometric Study of Oxygen-

Containing Films on Platinum Anodes, J. Chem. Phys. 48 (1968) 671.

[67] J. Bao, D.D. Macdonald, Electrocatalysis: proceedings of the international symposium,

Electrochemical Society, Pennington, NJ, 2005.

[68] A. Sun, J. Franc, D.D. Macdonald, Growth and Properties of Oxide Films on Platinum,

J. Electrochem. Soc. 153 (2006) B260.

[69] F.T. Wagner, P.N. Ross, Leed analysis of electrode surfaces, J. Electroanal. Chem.



92

[70] F.T. Wagner, P.N. Ross, LEED spot profile analysis of the structure of

electrochemically treated Pt(100) and Pt(111) surfaces, Surf. Sci. 160 (1985) 305–330.

[71] V.I. Birss, M. Goledzinowski, The unusual reduction behavior of thin, hydrous

platinum oxide films, J. Electroanal. Chem. 351 (1993) 227–243.

[72] H. Luo, S. Park, Y.H. Chan, M.J. Weaver, Surface Oxidation of Platinum-Group

Transition Metals in Ambient Gaseous Environments: Role of Electrochemical versus

Chemical Pathways, J. Phys. Chem. B. 104 (2000).

[73] A. Sun, D.D. Macdonald, Growth and Properties of Oxide Films on Platinum II. pH

Dependence in Alkaline Solutions, in: ECS, 2007: pp. 1–10.

[74] J. Bao, D.D. Macdonald, Growth Kinetics of the Anodic Oxide Film on Platinum

under Potentiodynamic Polarization Conditions, Z. Für Phys. Chem. 227 (2013) 541–559.

[75] A.N. Chemodanov, I.M. Djulai, Corrosion-electrochemical behaviour of noble metals -

platinum, iridium, gold, Zashchita Met. 27 (1991) 658–666.

[76] A. Damjanovic, A. Dey, J.O. Bockris, Kinetics of oxygen evolution and dissolution on

platinum electrodes, Electrochimica Acta. 11 (1966) 791–814.

[77] A. Damjanovic, A. Dey, J.O. Bockris, Electrode Kinetics of Oxygen Evolution and

Dissolution on Rh, Ir, and Pt-Rh Alloy Electrodes, J. Electrochem. Soc. 113 (1966) 739.

[78] Y.M. Kolotyrkin, Use of radioactive indicator and electrochemical methods for

determining low corrosion rates, Electrochimica Acta. 18 (1973) 593–606.

[79] Y.M. Kolotyrkin, V.V. Losev, A.N. Chemodanov, Relationship between corrosion

processes and oxygen evolution on anodes made from noble metals and related metal oxide

anodes, Mater. Chem. Phys. 19 (1988) 1–95.

[80] K.-I. Ota, S. Nishigori, N. Kamiya, Dissolution of platinum anodes in sulfuric acid

solution, J. Electroanal. Chem. Interfacial Electrochem. 257 (1988) 205–215.

[81] D.A.J. Rand, R. Woods, A study of the dissolution of platinum, palladium, rhodium

and gold electrodes in 1 m sulphuric acid by cyclic voltammetry, J. Electroanal. Chem.


[82] A.P. Yadav, A. Nishikata, T. Tsuru, Channel-Flow Double-Electrode Study on the

Dissolution and Deposition Potentials of Platinum under Potential Cycles, J. Electrochem. Soc.

156 (2009) C253.

[83] A.P. Yadav, T. Okayasu, Y. Sugawara, A. Nishikata, T. Tsuru, Effects of pH on

Dissolution and Surface Area Loss of Platinum Due to Potential Cycling, J. Electrochem. Soc.

159 (2012) C190.


93

[84] Y. Sugawara, T. Okayasu, A.P. Yadav, A. Nishikata, T. Tsuru, Dissolution Mechanism

of Platinum in Sulfuric Acid Solution, J. Electrochem. Soc. 159 (2012) F779–F786.

[85] S. Kim, J.P. Meyers, The influence of hydrogen- and cation-underpotential deposition

on oxide-mediated Pt dissolution in proton-exchange membrane fuel cells, Electrochimica Acta.

56 (2011) 8387–8393.

[86] F.-B. Li, A. Robert Hillman, S.D. Lubetkin, D.J. Roberts, Electrochemical quartz

crystal microbalance studies of potentiodynamic electrolysis of aqueous chloride solution:

surface processes and evolution of H2 and Cl2 gas bubbles, J. Electroanal. Chem. 335 (1992)

345–362.

[87] B.V. Ershler, About the passivity of platinum, Reports of National Academy of

Science USSR. XXXVII (1942) 258–261.

[88] D.M. Novak, B.E. Conway, Competitive adsorption and state of charge of halide ions

in monolayer oxide film growth processes at Pt anodes, J. Chem. Soc. Faraday Trans. 1. 77

(1981) 2341.

[89] M. Azaroual, B. Romand, P. Freyssinet, J.-R. Disnar, Solubility of platinum in aqueous

solutions at 25°C and pHs 4 to 10 under oxidizing conditions, Geochim. Cosmochim. Acta. 65

(2001) 4453–4466.

[90] M. �ukaszewski, A. Czerwi�ski, Dissolution of noble metals and their alloys studied by

electrochemical quartz crystal microbalance, J. Electroanal. Chem. 589 (2006) 38–45.

[91] G. Inzelt, B. Berkes, Á. Kriston, Temperature dependence of two types of dissolution

of platinum in acid media. An electrochemical nanogravimetric study, Electrochimica Acta. 55

(2010) 4742–4749.

[92] V.A.T. Dam, F.A. de Bruijn, The Stability of PEMFC Electrodes, J. Electrochem. Soc.

154 (2007) B494.

[93] G. Inzelt, B.B. Berkes, Á. Kriston, A. Székely, Electrochemical nanogravimetric studies

of platinum in acid media, J. Solid State Electrochem. 15 (2010) 901–915.

[94] M. Umeda, Y. Kuwahara, A. Nakazawa, M. Inoue, Pt Degradation Mechanism in

Concentrated Sulfuric Acid Studied Using Rotating Ring�Disk Electrode and Electrochemical

Quartz Crystal Microbalance, J. Phys. Chem. C. 113 (2009) 15707–15713.

[95] K.J.J. Mayrhofer, S.J. Ashton, J.C. Meier, G.K.H. Wiberg, M. Hanzlik, M. Arenz, Non-

destructive transmission electron microscopy study of catalyst degradation under

electrochemical treatment, J. Power Sources. 185 (2008) 734–739.

[96] K.J.J. Mayrhofer, J.C. Meier, S.J. Ashton, G.K.H. Wiberg, F. Kraus, M. Hanzlik, et al.,

Fuel cell catalyst degradation on the nanoscale, Electrochem. Commun. 10 (2008) 1144–1147.


94

[97] X. Wang, R. Kumar, D.J. Myers, Effect of Voltage on Platinum Dissolution,

Electrochem. Solid-State Lett. 9 (2006) A225.

[98] Z. Nagy, H. You, Applications of surface X-ray scattering to electrochemistry

problems, Electrochimica Acta. 47 (2002) 3037–3055.

[99] T. Biegler, An Electrochemical and Electron Microscopic Study of Activation and

Roughening of Platinum Electrodes, J. Electrochem. Soc. 116 (1969) 1131.

[100] T. Biegler, R. Woods, Limiting oxygen coverage on smooth platinum anodes in

acid solution, J. Electroanal. Chem. Interfacial Electrochem. 20 (1969) 73–78.

[101] S.G. Rinaldo, J. Stumper, M. Eikerling, Physical Theory of Platinum

Nanoparticle Dissolution in Polymer Electrolyte Fuel Cells, J. Phys. Chem. C. 114 (2010)

5773–5785.

[102] S.G. Rinaldo, W. Lee, J. Stumper, M. Eikerling, Model- and Theory-Based

Evaluation of Pt Dissolution for Supported Pt Nanoparticle Distributions under Potential

Cycling, Electrochem. Solid-State Lett. 14 (2011) B47.

[103] K. Itaya, In situ scanning tunneling microscopy of platinum (111) surface with

the observation of monatomic steps, J. Vac. Sci. Technol. Vac. Surfaces Films. 8 (1990) 515.

[104] K. Sashikata, N. Furuya, K. Itaya, In situ electrochemical scanning tunneling

microscopy of single-crystal surfaces of Pt(111), Rh(111), and Pd(111) in aqueous sulfuric acid

solution, J. Vac. Sci. Technol. B. 9 (1991) 457.

[105] V. Komanicky, K.C. Chang, A. Menzel, N.M. Markovic, H. You, X. Wang, et

al., Stability and Dissolution of Platinum Surfaces in Perchloric Acid, J. Electrochem. Soc. 153

(2006) B446.

[106] Y. Onochi, M. Nakamura, N. Hoshi, Atomic Force Microscopy of the

Dissolution of Cubic and Tetrahedral Pt Nanoparticles in Electrochemical Environments, J.

Phys. Chem. C. 116 (2012) 15134–15140.

[107] J. Inukai, D.A. Tryk, T. Abe, M. Wakisaka, H. Uchida, M. Watanabe, Direct

STM Elucidation of the Effects of Atomic-Level Structure on Pt(111) Electrodes for Dissolved

CO Oxidation, J. Am. Chem. Soc. 135 (2013) 1476–1490.

[108] M. Wakisaka, Y. Udagawa, H. Suzuki, H. Uchida, M. Watanabe, Structural

effects on the surface oxidation processes at Pt single-crystal electrodes studied by X-ray

photoelectron spectroscopy, Energy Environ. Sci. 4 (2011) 1662.

[109] K. Hartl, M. Nesselberger, K.J.J. Mayrhofer, S. Kunz, F.F. Schweinberger, G.

Kwon, et al., Electrochemically induced nanocluster migration, Electrochimica Acta. 56 (2010)

810–816.


95

[110] M. Matsumoto, T. Miyazaki, H. Imai, Oxygen-Enhanced Dissolution of

Platinum in Acidic Electrochemical Environments, J. Phys. Chem. C. 115 (2011) 11163–11169.

[111] A. Kongkanand, J.M. Ziegelbauer, Surface Platinum Electrooxidation in the

Presence of Oxygen, J. Phys. Chem. C. 116 (2012) 3684–3693.

[112] R.M. Darling, J.P. Meyers, Kinetic Model of Platinum Dissolution in PEMFCs,

J. Electrochem. Soc. 150 (2003) A1523.

[113] Inductively coupled plasma mass spectrometry handbook, Blackwell ; CRC

Press, Oxford : Boca Raton, FL, 2005.

[114] H.E. Taylor, Inductively coupled plasma-mass spectrometry: practices and

techniques, Academic Press, San Diego, 2001.

[115] G. Fischer, J. Vancea, Mean free path and density of conductance electrons in

platinum determined by the size effect in extremely thin films, Phys. Rev. B. 22 (1980) 6065–

6073.

[116] J.W. Niemantsverdriet, Spectroscopy in catalysis: an introduction, 3rd

completely rev. and enl. ed, Wiley-VCH ; John Wiley, distributor], Weinheim : [Chichester,

2007.

[117] Analytical methods in corrosion science and engineering, CRC Press, Boca

Raton, 2006.

[118] C.G. Williams, M.A. Edwards, A.L. Colley, J.V. Macpherson, P.R. Unwin,

Scanning Micropipet Contact Method for High-Resolution Imaging of Electrode Surface

Redox Activity, Anal. Chem. 81 (2009) 2486–2495.

[119] M.. Lohrengel, C. Rosenkranz, I. Klüppel, A. Moehring, H. Bettermann, B.V.

den Bossche, et al., A new microcell or microreactor for material surface investigations at large

current densities, Electrochimica Acta. 49 (2004) 2863–2870.

[120] A.W. Hassel, M.M. Lohrengel, The scanning droplet cell and its application to

structured nanometer oxide films on aluminium, Electrochimica Acta. 42 (1997) 3327–3333.

[121] A.J. Bard, F.R.F. Fan, J. Kwak, O. Lev, Scanning electrochemical microscopy.

Introduction and principles, Anal. Chem. 61 (1989) 132–138.

[122] G. Wittstock, W. Schuhmann, Formation and Imaging of Microscopic

Enzymatically Active Spots on an Alkanethiolate-Covered Gold Electrode by Scanning

Electrochemical Microscopy, Anal. Chem. 69 (1997) 5059–5066.

[123] A. Hengstenberg, C. Kranz, W. Schuhmann, Facilitated tip-positioning and

applications of non-electrode tips in scanning electrochemical microscopy using a shear force

based constant-distance mode, Chem. Weinh. Bergstr. Ger. 6 (2000) 1547–1554.


96

[124] S. Guerin, B.E. Hayden, C.E. Lee, C. Mormiche, J.R. Owen, A.E. Russell, et al.,

Combinatorial Electrochemical Screening of Fuel Cell Electrocatalysts, J. Comb. Chem. 6

(2004) 149–158.

[125] R. Bitter, LabView advanced programming techniques, 2nd ed, CRC

Press/Taylor & Francis Group, Boca Raton, FL, 2007.

[126] J. Nikolic, E. Expósito, J. Iniesta, J. González-Garcia, V. Montiel, Theoretical

Concepts and Applications of a Rotating Disk Electrode, J. Chem. Educ. 77 (2000) 1191.

[127] T.J. Schmidt, Characterization of High-Surface-Area Electrocatalysts Using a

Rotating Disk Electrode Configuration, J. Electrochem. Soc. 145 (1998) 2354.

[128] S. Sumathi, LabVIEW based advanced instrumentation systems, Springer,

Berlin ; New York, 2007.

[129] R.W. Larsen, LabVIEW for engineers., Pearson, Boston, Mass ; London, 2010.

[130] P.A. Blume, The LabVIEW style book, Prentice Hall, Upper Saddle River, NJ,

2007.

[131] N. Garland, T. Benjamin, J. Kopasz, DOE Fuel Cell Program: Durability

Technical Targets and Testing Protocols, in: ECS, 2007: pp. 923–931.

[132] K.J.J. Mayrhofer, D. Strmcnik, B.B. Blizanac, V. Stamenkovic, M. Arenz, N.M.

Markovic, Measurement of oxygen reduction activities via the rotating disc electrode method:

From Pt model surfaces to carbon-supported high surface area catalysts, Electrochimica Acta.

53 (2008) 3181–3188.

[133] Electrocatalysis, Wiley-VCH, New York, 1998.

[134] J.A. Cooper, R.G. Compton, Channel Electrodes — A Review, Electroanalysis.

10 (1998) 141–155.

[135] A. Jardy, A. Legal Lasalle-Molin, M. Keddam, H. Takenouti, Copper

dissolution in acidic sulphate media studied by QCM and rrde under ac signal, Electrochimica

Acta. 37 (1992) 2195–2201.

[136] B. Yuan, C. Wang, L. Li, S. Chen, Real time observation of the anodic

dissolution of copper in NaCl solution with the digital holography, Electrochem. Commun. 11

(2009) 1373–1376.

[137] D.F. van der Vliet, C. Wang, D. Li, A.P. Paulikas, J. Greeley, R.B. Rankin, et al.,

Unique Electrochemical Adsorption Properties of Pt-Skin Surfaces, Angew. Chem. Int. Ed. 51

(2012) 3139–3142.


97

[138] G.A. Tsirlina, 2 Surface Thermodynamics of Metal/Solution Interface: the

Untapped Resources, in: C.G. Vayenas (Ed.), Interfacial Phenom. Electrocatalysis, Springer

New York, New York, NY, 2011: pp. 107–158.

[139] B.E. Conway, Electrochemical oxide film formation at noble metals as a

surface-chemical process, Prog. Surf. Sci. 49 (1995) 331–452.

[140] G. Jerkiewicz, G. Vatankhah, J. Lessard, M.P. Soriaga, Y.-S. Park, Surface-

oxide growth at platinum electrodes in aqueous H2SO4, Electrochimica Acta. 49 (2004) 1451–

1459.

[141] J.-L. Todoli, Liquid sample introduction in ICP spectrometry: a practical guide,

1st ed, Elsevier, Amsterdam ; Boston, 2008.

[142] G. Benke, W. Gnot, The electrochemical dissolution of platinum,

Hydrometallurgy. 64 (2002) 205–218.

[143] S. Mitsushima, Y. Koizumi, S. Uzuka, K.-I. Ota, Dissolution of platinum in

acidic media, Electrochimica Acta. 54 (2008) 455–460.

[144] W.L. Marschall, E.V. Jones, Second Dissociation Constant of Sulfuric Acid

from 25 250°C Evaluated from Solubilities of Calcium Sulfate in Sulfuric Acid Soultions, J

Phys Chem. 70 (1966) 4028–4040.

[145] S. Mitsushima, Y. Koizumi, K.-I. Ota, N. Kamiya, Solubility of Platinum in

Acidic Media (I) - in sulfuric Acid, Electrochemistry. 75 (2006) 155–158.

[146] S.A. Wood, Experimental determination of the hydrolysis constants of Pt2+

and Pd2+ at 25°C from the solubility of Pt and Pd in aqueous hydroxide solutions, Geochim.

Cosmochim. Acta. 55 (1991) 1759–1767.

[147] D.S. Strmcnik, D.V. Tripkovic, D. van der Vliet, K.-C. Chang, V. Komanicky,

H. You, et al., Unique Activity of Platinum Adislands in the CO Electrooxidation Reaction, J.

Am. Chem. Soc. 130 (2008) 15332–15339.

[148] A. Rabis, P. Rodriguez, T.J. Schmidt, Electrocatalysis for Polymer Electrolyte

Fuel Cells: Recent Achievements and Future Challenges, ACS Catal. 2 (2012) 864–890.

[149] S. Kawahara, S. Mitsushima, K.-I. Ota, N. Kamiya, Consumption of Pt Catalyst

under Electrolysis and Fuel Cell Operation, in: ECS, 2006: pp. 85–100.

[150] K.J. Laidler, M.C. King, Development of transition-state theory, J. Phys. Chem.

87 (1983) 2657–2664.

[151] J.F. Moulder, J. Chastain, R.C. King, Handbook of x-ray photoelectron

spectroscopy: a reference book of standard spectra for identification and interpretation of XPS

data, Physical Electronics, Eden Prairie, Minn., 1995.


98

[152] J. Knecht, G. Stork, Roentgenphotoelektronen - spektroskopische

Untersuchung der Thallium-Oxidelektrode, Fresenius Z. Fuer Anal. Chem. 289 (1978) 206–206.

[153] S. Trasatti, Electronegativity, work function, and heat of adsorption of

hydrogen on metals, J. Chem. Soc. Faraday Trans. 1. 68 (1972) 229.

[154] S. Trasatti, Work function, electronegativity, and electrochemical behaviour of

metals, J. Electroanal. Chem. Interfacial Electrochem. 33 (1971) 351–378.

[155] M.L.B. Rao, A. Damjanovic, J.O. Bockris, OXYGEN ADSORPTION

RELATED TO THE UNPAIRED d-ELECTRONS IN TRANSITION METALS, J. Phys.

Chem. 67 (1963) 2508–2509.

[156] C.. Parkinson, M. Walker, C.. McConville, Reaction of atomic oxygen with a

Pt() surface: chemical and structural determination using XPS, CAICISS and LEED, Surf. Sci.

545 (2003) 19–33.

[157] L.D. Burke, J.F. O’sullivan, The stability of hydrous oxide films on platinum, J.

Appl. Electrochem. 21 (1991) 151–157.


99

Appendix

100

Appendix Publications list: 20. Cherevko S.; Zeradjanin, A.R.; Topalov, A. A.; Kulik, N.; Katsounaros, I.; Mayrhofer, K.J.J.; Stability-Activity Interrelation for the Electrocatalysis of Acidic Oxygen Evolution on Noble Metals, (2014) --- under revision --- 19. Zeradjanin, A.R.; Topalov, A. A.; Overmeere, Q.v.; Cherevko, S.; Chen X.X.; Ventosa, E.; Schuhmann W.; Mayrhofer, K.J.J.; Rational design of the electrode morphology for oxygen evolution – enhancing the performance for catalytic water oxidation, (2014) ---in press--- 18. Cherevko S.; Topalov, A. A.; Zeradjanin, A.R.; Keeley, G.P.; Mayrhofer, K.J.J.; Temperature-Dependent Dissolution of Platinum in Acidic Media, (2014) Electrocatalysis --- in press --- 17. Meier, J.C.; Galeano, C.; Katsounaros, I.; Witte, J.; Bongard, H.J.; Topalov A. A.; Baldizzone, C.; Mezzavilla, S.; Schüth, F.; Mayrhofer, K.J.J.; Design criteria for active and stable Pt/C fuel cell catalysts, 5 (2014) 44-67 Beilstein Journal of Nanotechnology, doi: 10.3762/bjnano.5.5 16. Topalov, A. A.; Zeradjanin, A.R.; Cherevko, S.; Mayrhofer, K.J.J.; The impact of dissolved reactive gases on platinum dissolution in acidic media, 40 (2014) 49-53 Electrochemistry Communications, doi: 10.1016/j.elecom.2013.12.021 15. Topalov, A. A.; Cherevko, S.; Zeradjanin, A. R.; Meier, J. C.; Katsounaros, I.; Mayrhofer, K. J.J.; Towards a comprehensive understanding of platinum dissolution in acidic media, 5 (2), (2014) 631-638 Chemical Science, doi: 10.1039/C3SC52411F 14. Cherevko, S.; Topalov, A. A.; Zeradjanin, A. R.; Katsounaros, I.; Mayrhofer, K.J.J.; Gold Dissolution: Towards Understanding of Noble Metal Corrosion, 3 (37), (2013) 16516 - 16527 RSC Advances, doi: 10.1039/C3RA42684J 13. Scuppert, A. K.; Topalov, A. A.; Savan, A.; Ludwig, A.; Mayrhofer, K.J.J.; Composition-Dependent Oxygen Reduction Activity and Stability of Pt–Cu Thin Films, 1 (2013) ChemElectroChem, doi: 10.1002/celc.201300078 12. Cherevko, S.; Topalov, A. A.; Katsounaros, I.; Mayrhofer, K.J.J.; Electrochemical dissolution of gold in acidic medium, 28 (2013) 44-46 Electrochemistry Communications, doi: 10.1016/j.elecom.2012.11.040 11.a) Topalov, A. A.; Katsounaros, I; Auinger, M.; Cherevko, S.; Meier, J.C.; Klemm, S.O.; Mayrhofer, K.J.J.; Dissolution of platinum – limits for the deployment of electrochemical energy conversion? , 51 (50), (2012) 12613-12615 Angewandte Chemie International Edition VIP Status, doi: 10.1002/anie.201207256 11.b) Topalov, A. A.; Katsounaros, I; Auinger, M.; Cherevko, S.; Meier, J.C.; Klemm, S.O.; Mayrhofer, K.J.J.; Die Auflösung von Platin – Grenzen für den Einsatz zur elektrochemischen Energieumwandlung? , 124 (50), (2012) 12782-12785 Angewandte Chemie VIP Status, doi: 10.1002/ange.201207256 10. Galeano, C.; Meier, J.C.; Peinecke, V.; Bongard, H.J.; Katsounaros, I.; Topalov, A. A.; Lu, A.H.; Mayrhofer, K.J.J.; Schueth, F.; Towards Highly Stable Electrocatalysts via Nanoparticle Pore Confinement 134 (50), 20457-20465, JACS (2012) doi: 10.1021/ja308570c 9. Ankah, G. N.; Pareek, A.; Cherevko S.; Topalov, A. A.; Rohwerder, M.; Renner, F. U.; The influence of halides on the initial selective dissolution of Cu3Au (111), Electrochimica Acta, 85 (2012) 384-392 doi:10.1016/j.electacta.2012.08.059 8. Meier, J.C.; Katsounaros, I; Galeano, C.; Bongard, H.; Topalov, A.A.; Kostka, A.; Karschin, A.; Stability investigations of electrocatalysts on the nanoscale, Energy & Environment Science 5 (11), (2012), 9319-9330, doi:10.1039/C2EE22550F 7. Schuppert, A.K.; Topalov, A. A.; Katsounaros, I.; Klemm, S. O.; Mayrhofer, K. J. J., A Scanning Flow Cell System for Fully Automated Screening of Electrocatalyst Materials, Journal of The Electrochemical Society 11 (2012) 159 (2012) doi: 10.1149/2.009211jes

Appendix

101

6. Klemm, S. O.; Karschin, A.; Topalov, A. A.; Mingers, A. M.; Katsounaros, I.; Mayrhofer, K. J. J., Time and potential resolved dissolution analysis of rhodium using a microelectrochemical flow cell coupled to an ICP-MS, Journal of Electroanalytical Chemistry 677–680 (2012), 50–55, doi:10.1016/j.jelechem.2012.05.006 5. Meier, J. C.; Galeano, C.; Katsounaros, I.; Topalov, A. A.; Kostka, A.; Schu �th, F.; Mayrhofer, K. J. J., Degradation Mechanisms of Pt/C Fuel Cell Catalysts under Simulated Start-Stop Conditions, ACS Catal. 2 (2012), 832-843, doi:10.1021/cs300024h 4. Klemm, S. O.; Topalov, A. A.; Laska, C. A.; Mayrhofer, K. J. J., Coupling of a high throughput microelectrochemical cell with online multielemental trace analysis by ICP-MS, Electrochemistry Communications 13 (2011), 1533–1535, doi:10.1016/j.elecom.2011.10.017 3. Topalov, A. A.; Katsounaros, I.; Meier, J. C.; Klemm, S. O.; Mayrhofer, K. J. J., Development and integration of a LabVIEW-based modular architecture for automated execution of electrochemical catalyst testing, Review of Scientific Instruments 82 (2011), no. 11, 114103-1-114103-5,doi:10.1063/1.3660814 2. Auinger, M.; Katsounaros, I.; Meier, J. C.; Klemm, S. O.; Biedermann, P. U.; Topalov, A. A.; Rohwerder, M.; Mayrhofer, K. J. J., Near-surface ion distribution and buffer effects during electrochemical reactions, Phys. Chem. Chem. Phys. 13 (2011), 16384–16394, doi:10.1039/c1cp21717h 1. Katsounaros, I.; Meier, J. C.; Klemm, S. O.; Topalov, A. A.; Biedermann, P. U.; Auinger, M.; Mayrhofer, K. J. J., The effective surface pH during reactions at the solid–liquid interface, Electrochemistry Communications 13 (2011), no. 6, 634-637, doi:10.1016/j.elecom.2011.03.032

Oral presentations: 13. Schuppert, A. K.; Topalov, A. A.; Savan, A.; Ludwig, A.; Mayrhofer, K. J. J., Pt-Cu Alloys as Catalysts for the Oxygen Reduction Reaction – A Thin-Film Study of Activity and Stability, 2013. 224th ECS Meeting, San Francisco, CA, USA, 2013-10-27 to 2013-11-01

12. Topalov, A. A.; Zeradjanin, A. R.; Mayrhofer, K. J. J., (Elektro-)chemische Energieumwandlung - essentieller Baustein der Energiewende, 2013. 26. Spektrometertagung, Friedrichshafen, Germany, 2013-09-10 to 2013-09-11 (invited talk)

11. Topalov, A. A.; Cherevko, S.; Zeradjanin, A. R.; Meier, J. C.; Katsounaros, I.; Mayrhofer, K. J. J., Dissolution of Pt during oxygen reduction in acidic media, 2013. GDCh-Wissenschaftsforum Chemie 2013, Darmstadt, Germany, 2013-09-01 to 2013-09-04 (Förderpreis der GDCh-Fachgruppe Angewandte Elektrochemie 2013)

10. Topalov, A. A.; Cherevko, S.; Zeradjanin, A. R.; Mayrhofer, K. J. J., Stability of Electrocatalyst Materials – a Limiting Factor for the Deployment of Electrochemical Energy Conversion?, 2013. Third Russian-German Seminar on Catalysis “Bridging the Gap between Model and Real Catalysis. Energy-Related Catalysis”, Burduguz, Lake Baikal, Russia, 2013-06-24 to 2013-06-27, (Award for an excellent oral presentation of young scientist)

9. Zeradjanin, A. R.; Topalov, A. A.; Cherevko, S.; Schuhmann, W.; Mayrhofer, K. J. J., Rational design of morphological pattern for efficient electrocatalytic gas evolution, 2013. 4th Regional Symposium on Electrochemistry South-East Europe, Ljubljana, Slovenia, 2013-05-26 to 2013-05-30

8. Topalov, A. A.; Zeradjanin, A. R.; Cherevko, S.; Mayrhofer, K. J. J., Investigation of platinum stability by in-situ mass spectrometry, 2013. 4th Regional Symposium on Electrochemistry South-East Europe, Ljubljana, Slovenia, 2013-05-26 to 2013-05-30

7. Topalov, A. A., Wissenschaftler-/in, was ist das?, 2012. Berufs- und Studieninformationstag, Lore-Lorentz Schule, Düsseldorf, Germany, 2012-12-04

Appendix

102

6. Meier, J. C.; Galeano, C.; Katsounaros, I.; Topalov, A. A.; Schüth, F.; Mayrhofer, K. J. J., Electrode Materials for Electrochemical Energy Conversion, 2012. Electrochemistry 2012, Fundamental and Engineering Needs for Sustainable Development, Munich, Germany, 2012-09-17 to 2012-09-19

5. Topalov, A. A.; Mayrhofer, K. J. J., Pt dissolution monitored by ICP-MS, 2012. Workshop "Challenges in Energy Research", MPI-BAC, Mülheim, Germany, 2012-03-12 to 2012-03-12

4. Cherevko., S; Topalov, A. A.; Mayrhofer, K. J. J., Effect of Cathodic Polarization on the Electrochemistry of Gold Surfaces, 2012. 63rd Annual Meeting of the International Society of Electrochemistry, Prague, Czech Republic, 2012-09-19 to 2012-09-24

3. Topalov, A. A.; Mayrhofer, K. J. J., Kopplung ICP-MS mit Elektrochemie: Online Untersuchung von Materialkorrosion sowie Stabilität von Brennstoffzellenkatalysatoren, 2012. Anorganica 2012, Hilden, Germany, 2012-09-13 to 2012-09-13 (invited talk)

2. Meier, J. C.; Galeano, C.; Katsounaros, I.; Topalov, A. A.; Schüth, F.; Mayrhofer, K. J. J., Role of Carbon Support for Activity and Stability of Fuel Cell Catalysts, 2011. 15th Annual Green Chemistry & Engineering Conference, Washingtion D.C., USA, 2011-06-21 to 2011-06-23

1. Katsounaros, I.; Meier, J. C.; Topalov, A. A.; Klemm, S. O.; Hodnik, N.; Mayrhofer, K. J. J., Oxygen reduction reaction: Surface coverage vs. activity, 2011. 8th Greek Conference of Chemical Engineering, Thessaloniki, Greece, 2011-05-26 to 2011-05-28

Poster presentations 11. Topalov, A. A.; Zeradjanin, A. R.; Cherevko, S.; Mayrhofer, K. J. J., Investigation of electro-chemical dissolution of platinum under the influence of reactive gases by in-situ mass spectrometry, /2013. São Paulo School of Advanced Sciences on Electrochemistry, Energy Conversion and Storage - SPASECS, São Paulo, Brazil, 2013-12-07 to 2013-12-14 (Poster Award) 10. Schuppert, A. K.; Topalov, A. A.; Savan, A.; Klemm, S. O.; Ludwig, A.; Mayrhofer, K. J. J., Fast Screening of PEMFC-Catalysts with a Scanning Flow Cell System, /2012. Electrochemistry 2012, Munich, Germany, 2012-09-17 to 2012-09-19 9. Topalov, A. A.; Klemm, S. O.; Meier, J. C.; Katsounaros, I.; Mayrhofer, K. J. J., Investigation of the anodic and cathodic dissolution of platinum in acidic media, /2012. GDCh meeting – Electrochemistry 2012, Munich, Germany, 2010-09-17 to 2010-09-19 8. Mayrhofer, K. J. J.; Meier, J. C.; Galeano, C.; Katsounaros, I.; Topalov, A. A.; Schüth, F., Activity and stability of Pt/HGS catalysts for application in fuel cells, /2012. GDCh meeting – Electrochemistry 2012, Munich, Germany, 2010-09-17 to 2010-09-19 7. Klemm, S. O.; Karschin, A.; Schuppert, A. K.; Topalov, A. A.; Katsounaros, I.; Mayrhofer, K. J. J., Rhodium electrode dissolution in sulfuric acid during electrochemical treatment investigated with a scanning flow cell coupled to an ICP-MS, /2012. 63rd Annual Meeting of the International Society of Electrochemistry, Prague, Czech Republic, 2012-09-19 to 2012-09-24 6. Schuppert, A. K.; Topalov, A. A.; Savan, A.; Klemm, S. O.; Ludwig, A.; Mayrhofer, K. J. J., Fast Screening of PEMFC-Catalysts with a Scanning Flow Cell System, /2012. 63rd Annual Meeting of the International Society of Electrochemistry, Prague, Czech Republic, 2012-09-19 to 2012-09-24 5. Topalov, A. A.; Meier, J. C.; Katsounaros, I.; Klemm, S. O.; Mayrhofer, K. J. J., Online Monitoring of the Dissolution of Platinum during Electrochemical Experiments by Coupled ICP-MS, /2012. 63rd Annual Meeting of the International Society of Electrochemistry, Prague, Czech Republic, 2012-09-19 to 2012-09-24

Appendix

103

4. Auinger, M.; Katsounaros, I.; Meier, J. C.; Biedermann, P. U.; Topalov, A. A.; Klemm, S. O.; Rohwerder, M.; Mayrhofer, K. J. J., The Effective Surface pH during Reactions at the Solid/Liquid Interface , /2012. 63rd Annual Meeting of the International Society of Electrochemistry, Prague, Czech Republic, 2012-08-19 to 2012-08-24 3. Meier, J. C.; Galeano, C.; Katsounaros, I.; Topalov, A. A.; Schüth, F.; Mayrhofer, K. J. J.;, IL-TEM and IL-Tomography Stability Investigations of Fuel Cell Catalysts, /2012. 63rd Annual Meeting of the International Society of Electrochemistry, Prague, Czech Republic, 2012-09-19 to 2012-09-24, (Poster Award) 2. Auinger, M.; Katsounaros, I.; Meier, J. C.; Klemm, S. O.; Biedermann, P. U.; Topalov, A. A.; Rohwerder, M.; Mayrhofer, K. J. J., Near Surface Ion Distribution and Buffer Effects during Electrochemical Reactions, /2011. 14th Austrian Chemistry Days, Linz, Austria, 2011-09-26 to 2011-09-29 1. Katsounaros, I.; Topalov, A. A.; Mayrhofer, K. J. J., Electrochemical reduction of CO2 to fuels: directions and perspectives, /2010. Electrochemistry 2010: From Microscopic Understanding to Global Impact, Bochum, Germany, 2010-09-13 to 2010-09-15

Appendix

104

Curriculum Vitae – Angel A. Topalov PERSONAL INFORMATION

� Birthday: 11.02.1985, Sankt-Petersburg � Nationality: Bulgarian, German � Martial status: married � e-mail: [email protected]

PHD STUDIES

05/2010 - present Max-Planck-Institut für Eisenforschung, Düsseldorf

� PhD student in the Electrocatalysis Group of Dr. Karl J.J. Mayrhofer in the Department for Interface Chemistry and Surface Engineering of Prof. Martin Stratmann in close cooperation with Prof. Wolfgang Schuhmann

� PhD thesis about “Design and implementation of automated system based on coupling of electrochemical flow cell with mass spectrometry for investigation of dissolution behavior of platinum”

� PhD courses at the Ruhr-Universität Bochum (RUB): - Intellectual Property Rights - Advanced Electroanalytical Methods

� Soft skill courses of the International Max-Planck Research School: (Presentation Skills, Scientific Writing & Publishing, Leadership Skills, Project & Self Management)

� Certified LabVIEW Developer (6/2011-5/2013) � Training and supervision of new PhD students and interns � Supervisor in DAAD RISE program 2011 � Building up the laboratories of the Mayrhofer group

- System design of Scanning Flow Cell (SFC) - Software development for RDE and SFC setups - Coupling of ICP-MS with Scanning Flow Cell

UNIVERSITY

2004 - 2009 University of Leipzig, Physics Department

� M.Sc. degree in physics, International Physics Study Program 2003 - 2004 Sofioter University “Sv. Kliment Ohridski”, Physics Department

AWARDS

12/2013 São Paulo School of Advanced Sciences on Electrochemistry, Energy

Conversion and Storage - SPASECS 2013, Sao Paulo, Brazil

� Energy & Environmental Science Poster Prize

09/2013 GDCh-Wissenschaftsforum Chemie 2013, Darmstadt, Germany

� Award of German Chemical Society (GDCh) for Applied Electrochemistry 2013

Appendix

105

06/2012 3rd Russian-German Seminar on Catalysis, Burduguz, Russia

� Best oral presentation of young scientist

SCIENTIFIC RECORD

Peer-reviewed articles published till Januar 2014 18 Total number of citations (according to google scholar) 180 H-index (according to google scholar) 8

WORK EXPERIENCE

10/2012 – present Max-Planck-Institut für Eisenforschung GmbH (TVöD E13)

05/2010 – 09/2012 Center for Electrochemical Sciences, RUB (TV-L E13)

Multiple student jobs (e.g.):

11/2008 – 03/2010 CAD support, Ingenieurbüro für Bauabrechnung – Reißmüller

09/2009 – 12/2009 Technical support, Malberg EDV-Systemberatung GmbH

04/2008 – 05/2008 HIWI, University of Leipzig, Physics Department

SCHOOL

1998 - 2003 Secondary Mathematical School “Baba Tonka”, Ruse, Bulgaria

1991 - 1998 Elementary School “Paisii Hilendaski”, Ruse, Bulgaria

LANGUAGE SKILLS

English fluent

German fluent

Bulgarian Mother tongue

Russian Mother tongue

design and implementation of an automated electrochemical ... · design and implementation of an...

Documents