Role of Lattice Oxygen in the Oxygen Evolution Reaction on Co3O4: Isotope Exchange Determined Using a Small-Volume Differential Electrochemical Mass Spectrometry Cell Design

Article abstract of DOI:10.1021/acs.analchem.9b01749

This work demonstrates the role of lattice oxygen of metal oxide catalysts in the oxygen evolution reaction (OER) as evidenced by isotope labeling together with the differential electrochemical mass spectrometry (DEMS) method. Our recent report assessed this role for Co₃O₄ using a flow-through DEMS cell, which requires a large volume of electrolyte. Herein, we extend this procedure to different Co₃O₄ catalyst loadings and particle sizes as well as the mixed Ag + Co₃O₄ catalyst. We introduce, for the first time, a novel small-volume DEMS cell design capable of using disc electrodes and only <0.5 mL of electrolyte. The reliability of the cell is demonstrated by monitoring gas evolution during OER in real time. This cell shows high sensitivity, high collection efficiency, and very short delay time. DEMS results reveal that only the interfacial part (～0.2% of the total loading or 25% of surface atoms) of the catalyst is active for OER. Interestingly, the amount of oxygen exchanged on the mixed Ag + Co₃O₄ catalyst is higher than that on the single Co₃O₄ catalyst, which illustrates the improved electrocatalytic activity previously reported on this mixed catalyst. Furthermore, the real surface area of the catalysts is estimated using different methods (namely, the ball model, double layer capacitance, isotope exchange, and redox peak methods). The surface areas estimated from the Brunauer-Emmett-Teller (BET) and ball models are comparable but roughly three times higher than that of the redox peak method. Our method represents an alternative approach for probing the mechanism and real surface area of catalysts.

Full text of DOI:10.1021/acs.analchem.9b01749

Subscriber access provided by Bibliothèque de l'Université Paris-Sud
Article
3
4
On the role of lattice oxygen in oxygen evolution reaction on CoO: isotope
exchange determined using a new small-volume DEMS cell design
Hatem M.A. Amin, Heinz-Peter Königshoven, Martina Hegemann, and Helmut Baltruschat
Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01749 • Publication Date (Web): 16 Sep 2019
Downloaded from pubs.acs.org on September 17, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 16
Analytical Chemistry
1
2
3
4
On the role of lattice oxygen in oxygen evolution reaction on Co3O4: isotope exchange
determined using a new small-volume DEMS cell design
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Hatem M.A. Amina,b, Peter Königshovenb, Martina Hegemannb, Helmut Baltruschatb,*
a Department of Chemistry, Faculty of Science, Cairo University, 12613 Giza, Egypt
b Institute of Physical and Theoretical Chemistry, University of Bonn, 53117 Bonn, Germany
* Corresponding author: baltruschat@uni-bonn.de
hatem@pc.uni-bonn.de
1
ACS Paragon Plus Environment

Analytical Chemistry
Page 2 of 16
1
2
3
Abstract
4
5
6
7
8
This work demonstrates the role of lattice oxygen of metal oxide catalysts in oxygen evolution
reaction (OER) as evidenced by isotope labeling together with differential electrochemical mass
spectrometry (DEMS) method. Our recent report assessed this role for Co3O4 using a flow-through
DEMS cell, which however requires a large volume of electrolyte. Herein, we extend this
procedure to different Co3O4 catalyst loadings and particle sizes as well as the mixed Ag+Co3O4
catalyst. We introduce, for the first time, a novel small-volume DEMS cell design capable of using
disc electrodes and only < 0.5 mL of electrolyte. The reliability of the cell is demonstrated by
monitoring gas evolution during OER in real time. This cell shows high sensitivity, high collection
efficiency and very short delay time. DEMS results reveal that only the interfacial part (~ 0.2% of
the total loading or 25% of surface atoms) of the catalyst is active for OER. Interestingly, the
amount of oxygen exchanged on the mixed Ag+Co3O4 catalyst is higher than that on the single
Co3O4 catalyst, which illustrates the improved electrocatalytic activity previously reported on this
mixed catalyst. Furthermore, the real surface area of the catalysts is estimated using different
methods (namely ball model, double layer capacitance, isotope exchange and redox peak
methods). The surface area estimated from BET and ball model are comparable, but roughly three
times higher than that of the redox peak method. Our method represents an alternative approach
for probing the mechanism and real surface area of catalysts.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keywords: Water electrolysis; new DEMS cell; cobalt oxide; real surface area; isotope labeling
2
ACS Paragon Plus Environment

Page 3 of 16
Analytical Chemistry
1
2
3
1. Introduction
4
5
6
7
8
9
The development of a cost-effective and highly active catalyst is of significant importance for
energy storage applications such as metal-air batteries and fuel cells. Oxygen evolution reaction
(OER), in particular, is a central process in these energy technologies. In the last few decades, the
intensive research has resulted in the development of materials with diverse electrocatalytic and
redox properties.1,2 However, OER kinetics is still sluggish and the commercially used catalysts
are based on precious metals.3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
First-row transition metal oxides in particular Co-based materials4-6 are inexpensive alternatives to
7,8
the state-of-the-art IrO2, and have significantly shown high activity for OER.9,10 Additionally,
some bimetallic catalysts such as Ag+Co3O4 mixture4,11,12 have shown bifunctional oxygen
activity.13
In addition to the great efforts that have been exerted to explore highly active catalysts, substantial
effort has been directed to understand the mechanism of OER.9,14,15 As a result, several reaction
mechanisms have been reported15-18 and it is highly complicated to assign a single reaction step to
oxygen electrocatalysis on oxide surfaces. In most of the previously reported mechanisms, changes
in the oxidation states of the metal take place during OER, and thus are mainly responsible for the
enhanced activity observed at cobalt-containing catalysts.4,19,20 Bergmann et al. reported that only
the outermost part of the Co-oxide (1.8% of Co ions) takes part in the redox reactions.20 Switzer et
al. demonstrated the conversion of Co(OH)2 to CoOOH and Co3O4 and compared their OER
activity. CoOOH and Co3O4 showed to have the same OER Tafel slope when the current is
corrected for the electrochemically active surface area, suggesting the importance of assessing the
real surface area and that the active species in both materials is likely to be CoIV.21 Using X-ray
photoelectron spectroscopy, we previously showed a reversible redox switching in Co3O4 when it
is in contact with Ag in a mixture.4
A different mechanism, in which the lattice oxygen of the oxide catalyst is involved in oxygen
evolution and formic acid oxidation, has been proven by mass spectrometry together with isotope
labeling method.22-28 However, this OER mechanism may not be generalized to all metal oxides as
it is dependent on the composition, structure and crystallinity of the material: Although most of the
31
investigated catalysts including Co3O4,22 RuO2,29,30 IrO2,24 Ba0.5Sr0.5Co0.8Fe0.2O3−δ and highly
covalent perovskites32 showed lattice exchange, others including NiFeOxHy,27 less covalent
perovskites32 and oriented RuO2 thin films18 do not. Using operando XAS, Schmidt and co-
workers have recently demonstrated the involvement of lattice oxygen in OER.31 In our previous
report22 we showed using differential electrochemical mass spectrometry (DEMS) technique that
the lattice atoms of Co3O4 participate in OER via an oxygen exchange process. In this study, we
extend our investigations to other Co3O4 catalyst loadings and particle sizes as well as to the mixed
3
ACS Paragon Plus Environment

Analytical Chemistry
Page 4 of 16
1
2
3
4
Ag+Co3O4 catalyst, a catalyst of outstanding activity. This work thus correlates the activity with
the lattice oxygen exchange process in the oxide.
5
6
7
8
In the last two decades, different designs of DEMS cells have been developed to accommodate the
electrode’s geometry and the purpose of the experiment.33-37 For example, in our previous report22
on isotope labeling experiments, we used the thin-layer flow-cell which needs large amount of
electrolyte. The classical cell on the other hand has practical difficulty for employing solid
particles by drop cast and dry method.24,33 Therefore, the high cost of the isotope labeled water
consumed in the flow cell and the need for measurements with particle catalysts boosted us to
develop a new small-volume DEMS cell design, and is presented in this work for the first time (see
Fig. 1). In essence, this is a thin layer cell similar to that of ref 36,38, but the thin layer gap is
variable. In our design, important features are achieved: The use of small disc electrodes (5 mm)
and small electrolyte volume (≤ 0.5 mL) and particularly the very short delay time (≤ 2s), which is
much less than that observed in other DEMS cells (~ 7−9 s)27,30,35,39 and the recent sniffer-chip cell
(~ 3s for OER).40 The proof-of-concept of the new cell is examined for oxygen evolution by
probing the delay time, the ionic current signals and the characteristic features of voltammograms.
In practice, measuring the real surface area under the same conditions of the experiment is
important for catalyst assessment as the experimental conditions may change the state of electrode
surface. Here, we determine the true surface area using several approaches (namely, the ball
model, double-layer capacitance method, redox peak method and isotope exchange method). The
results of these methods are compared to the BET data. Thus, this work highlights the importance
of probing the mechanism and the surface area when assessing the catalyst activity.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2. Experimental
The chemicals, electrode preparation and experimental details can be found in the supporting
information.
New small-volume DEMS cell
A schematic drawing and an actual photo of the small-volume DEMS cell are shown in Fig. 1. A
common feature in most of DEMS cells is the use of a porous Teflon membrane interface which
achieves a shorter delay time33,37 compared to other DEMS setups with a gas carrier.34,40 In our
new design, a porous Teflon membrane interface is used.
For this cell, the working electrode is a GC bulk disc that allows the catalyst ink (e.g. Co3O4
powder in ethylene glycol) to be easily drop casted and dried onto it (In contrast, this is practically
difficult for cells that use a soft and shrinkable porous Teflon membrane as direct support for the
working electrode. Flow-through cells, which also can accommodate such an electrode, have the
disadvantage that the particle catalyst layer may not be stable enough under convective conditions
4
ACS Paragon Plus Environment

Page 5 of 16
Analytical Chemistry
1
2
3
4
5
6
7
8
9
in addition to the higher electrolyte consumption). The disc electrode is then carefully installed and
pressed into a Kel-F shaft so that both disc and shaft forefronts lie on the same planar level. The
disc has a diameter of 5 mm (0.196 cm2). The shaft is an analogue to a RDE tip. It has a conical
shape at the apex, as shown in the drawing, which results in good electrochemical performance.
This shape is found to be important, after some trials with a pure cylindrical tip, as it allows a
shorter distance between the disc and the reference electrode (ca. 3 mm). The tip containing the
working electrode disc is inserted into a PTFE lid, and screwed into an upper Teflon mounting nut.
The nut acts as a manipulator to control the distance of the working electrode from the Teflon
membrane (range is from ~20 µm to ~3 mm). The Teflon lid tightly fits into the glass housing,
which is made for a small-volume electrolyte. The glass cell is mounted on a stainless steel holder,
which contains 5 mm diameter steel frit in the centre and is directly connected to a differentially
pumped quadruple mass spectrometer MS (QMG-422, Pfeiffer). As an interface between the
electrolyte and MS, a Teflon membrane is mechanically supported on the steel frit.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
b
Fig. 1. a) Sketch and cross section of the new small-volume DEMS cell with its components, where
1) copper contact pin, 2) Teflon mounting nut, 3) reference electrode connection, 4) Teflon lid, 5)
Kel-F shaft, 6) glass bead, 7) disc working electrode, 8) steel frit, 9) stainless steel connection to
MS, 10) mounting ring, 11) steel screw, 12) electrolyte, 13) glass housing, 14) counter electrode,
15) electrolyte inlet and outlet, 16) gas outlet. b) Actual photo of the cell.
The working electrode was exactly facing the steel frit to allow efficient transfer of volatile
products to MS. The working electrode was connected from the back side via a copper wire and a
contact pin to get electrical contact with the potentiostat. Few tens of micrometers (~50 µm) is
found to be the optimal distance between the working electrode and the membrane for our setup.
5
ACS Paragon Plus Environment

Analytical Chemistry
Page 6 of 16
1
2
3
4
5
6
7
8
9
A Pt wire, as a counter electrode, was inserted through the PTFE lid to the electrolyte. The
reference electrode was a reversible hydrogen electrode (RHE). The reference electrode was
electrically connected to the working solution via a salt bridge made of a Teflon tube filled with
the same alkaline solution, and ended with a glass bead. All potentials in this work are referred to
RHE. Ar-purging was achieved by a Teflon tube, which was immersed into the solution at certain
distance from the membrane. The electrolyte was saturated with Ar for about 10 minutes prior to
and during the experiment. All experiments were carried out at room temperature (25 ±1 oC).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3. Results
18
3.1. Oxygen exchange in Co3O4 (50 nm) catalyst in 20% marked H2 O solution
DEMS is a versatile technique that monitors the gaseous reaction products and intermediates of an
electrochemical reaction spectrometrically. To test the reliability of the new DEMS cell, OER at
Co3O4/GC was monitored. In this experiment, Fig. 2, we used a higher catalyst loading (400 µg
cm-2) and isotope concentration (20%) compared to our previous measurements with the flow thin
layer cell.22 In fact, the catalyst loading has an effect on the OER currents, where higher loadings
led to higher OER currents, see Fig. S1. In a similar procedure to that in our recently published
article,22 two series of experiments were conducted in the new cell: In series A, we performed the
18
18
experiments in marked H2 O solution to label the oxygen atoms of the Co3O4 with O atoms;
16
while in series B the pre-labeled oxide was treated in unlabeled H2 O solution to evolve O2. By
16
18
monitoring the signals of different isotopic O2 (16O2, O18O and O2) we get information on the
extent of oxygen exchange in the oxide. From these experiments we prove that the oxygen of the
Co3O4 catalyst is involved in the OER mechanism. In series A of experiments, the cell was filled
from the electrolyte reservoir using a Teflon tube inlet with the marked solution: 0.8 M LiOH
18
solution containing 20% (w/w) of marked H2 O. CVs and MSCVs were recorded between 0.05 V
and 1.6 V vs. RHE for 6 cycles at a scan rate of 10 mV s-1. Afterwards, the potential was stopped
at 0.4 V for electrolyte exchange for preparing for series B. Fig. 2a shows the respective
16
16
18
voltammograms for the first 3 cycles. The ionic currents corresponding to O2, O18O and O2
gases are clearly observable at potentials of OER (above 1.5 V), proving the successful operation
of the cell. In MSCV, the ionic currents correspond to the pure OER process, while in CV the
faradaic currents involve oxidation of Co3O4 at onset potentials of OER. Importantly, there is no
significant hysteresis between anodic and cathodic scans in the MSCV, as compared to the thin
layer cell33 or the Kel-F cell used by Krtil et al.30 The delay time in this cell is 1-2 s, as can be
noticed from the CVs and MSCVs. This proves the efficient mass transport to the mass
spectrometer in this cell.
6
ACS Paragon Plus Environment

Page 7 of 16
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 2. CVs and MSCVs for OER (first 3 cycles) obtained at Co3O4 (400 µg/cm2 of 50 nm
size)/GC electrode in the new small-volume DEMS cell in (a) Ar-saturated 0.8M LiOH solution-
18
16
containing 20% (w/w) marked H2 O, and the upper potential limit is 1.6 V, (b) 0.8M LiOH/H2 O
18
after the electrode has been pre-treated in 0.8M LiOH/H2 O. The upper panel displays the
faradaic current, while the lower ones represent the ionic currents for m/z 32, 34 and 36. Scan rate
= 10 mV/s.
In addition to the validation of the quality of the cell, we aimed to investigate the contribution of
the oxide in OER and its mechanism. After evaluation of the signals, the mole fraction of 16O2 and
16O18O was calculated and displayed in Fig. 3a. The results show an increase of 16O18O
concentration with a concomitant decrease in concentration of 16O2 with each cycle, reaching
steady-state values at the 4th cycle, see Fig. 3a. This indicates that the oxide indeed participates to
16
some extent in OER. The assumption that the increase of O18O content in the evolved O2 could
16
18
be due to the exchange of an OH by OH can be ruled out since such OH groups would more
probably exchanged during electrolyte exchange. Thus, they would not show up during OER. The
plotted points in the figure represent the average of three data points due to the uncertainty in the
integration of the peaks.
In series B, the marked solution was withdrawn from the cell using a syringe under constant
potential of 0.4 V. The electrolyte was then replaced gradually while keeping a little of solution
over the membrane; this process was repeated several times to ascertain the substitution of all the
16
marked solution with H2 O solution. The pre-labeled oxide was then scanned in the fresh 0.8M
16
18
LiOH/H2 O solution for 7 cycles. Fig. 2b shows the first three cycles. The signal for O2 is not
7
ACS Paragon Plus Environment

Analytical Chemistry
Page 8 of 16
1
2
3
4
5
6
7
8
observable since its concentration is below limit of detection. The amount of 16O18O decreases
with cycling reaching the final concentration of 18O in solution, while the amount of 16O2
increases, see Fig. 3b. This directly proves the oxygen exchange mechanism and participation of
oxide atoms in OER. The steady-state values were reached after the 4th cycle, and the isotope
content at/in the oxide is similar to that in the solution, thus no further exchange takes place. The
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16
isotopic composition of the formed O2 correspond to a concentration of ~16.5 w% of H2 O. The
difference from the as-prepared solution (20 w%) can be due to some losses on the walls of the
syringe during transfer and thus dilution of the marked solution by the rest of water on the walls
and membrane of the cell before start of measurement.
Fig. 3. Plot of the isotopic content as a function of scan number for OER obtained at Co3O4 (400
18
16
µg/cm2)/GC with scan rate of 10 mV/s in (a) 0.8M LiOH/(20 w%) H2 O, (b) 0.8M LiOH/H2 O,
18
the electrode has been pre-treated in 0.8M LiOH/H2 O. Data were taken from the experiment
shown in Fig. 2. The uncertainty in X34 is ~ 2% and in X32 is ~ 0.3%. The plotted data points
represent the average of three independent integrations.
To determine quantitatively the extent of oxygen exchange process, the number of moles of
exchanged oxygen atoms of the oxide were calculated and compared to the total atoms and also
surface atoms of the catalyst, following the same procedure and equations reported in our previous
study.22 For this catalyst loading, a total of 13.5 nmol cm-2 oxygen atoms are exchanged, thus the
amount of oxygen exchanged of the total oxygen atoms in Co3O4 loading is rex~ 0.2%. When the
amount of exchanged oxygen is compared to the surface atoms obtained from the charge of the
redox reaction Co3+/Co4+, we get the amount of exchanged atoms to the surface atoms yex~ 25%
which are active and participate in the oxygen exchange process, see Table 1. For comparison,
previous results on IrO2/Ti showed an isotope exchange of about 1-2% (few monolayers) of the
total loading of the oxide.24,27
Moreover, the real surface area of the catalyst is determined using different approaches: i) The ball
model in which the particles were treated as spheres with homogenous radius, ii) Double-layer DL
capacitance approach in which a DL capacitance of the oxide of 60 µF cm-2 is considered9,41
8
ACS Paragon Plus Environment

Page 9 of 16
Analytical Chemistry
1
2
3
4
5
6
7
8
9
although this value has a high degree of uncertainty. However, there are no well-accepted values
for the DL capacitance for oxides including Co3O4 as they exhibit significant pseudo capacitance.
iii) Redox peak method in which the moles of the surface atoms is calculated using the charge of
the redox couple Co3+/Co4+ (peak at about 1.45 V), iv) Isotope exchange method based on the
number of moles of atoms exchanged and v) BET analysis. Details of the procedure and the
equations used for each method are described in our previous report.22 The results are summarized
in Table 2 and compared in the discussion section.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
Oxygen exchange in Co3O4 (50 nm) in 50% marked H2 O solution
18
We also studied the behavior of the same catalyst but in a higher H2 O concentration (50 w%) in
solution and using higher upper potential limit (1.65 V), Fig. S2-S4. The results are displayed in
Table 1 and revealed that the dependence of the isotope exchange ratio on the potential limit of
scanning and the loading amount is not significant, see SI.
Table 1. Summary of results of oxygen exchange experiments at different catalysts and loadings.
nt(O2-) /
rsurf /
%
rex / %
yex /%
rex /%#
yex /%#
Catalyst
nmol cm-2
Co3O4 (40 µm) 100 µg/cm2 in 0.5 M KOH
with 2% isotope till 1.62 V 22
1660
3320
6640
6640
9960
664
0.23
1.78
0.79
0.63
--
0.08
34
0.03
13
Co3O4 (50 nm) 200 µg/cm2 in 0.1 M LiOH
0.22
0.20
12
25
0.1
8.5
6.4
with 10% isotope till 1.8 V22
Co3O4 (50 nm) 400 µg/cm2 in 0.8 M LiOH
with 20% isotope till 1.6 V
Co3O4 (50 nm) 400 µg/cm2 in 0.5 M LiOH
with 50% isotope till 1.65 V(1.6 V)
Co3O4 (10 µm) 600 µg/cm2 in 0.8 M LiOH
with 20% isotope till 1.61 V
Ag+Co3O4 (50 nm) 10 w%, 400 µg/cm2 in
0.5 M LiOH with 50% isotope till 1.65 V
0.05
0.15(0.2) 23(22)
0.07(0.1) 11(3)
0.03
0.9
--
--
--
0.02
0.3
--
--
--
--
--
Ag+Co3O4 (50 nm) 10 w%, 600 µg/cm2 in
0.8M LiOH with 20% isotope till 1.65 V
664
--
0.7
rsurf = nsurf / nt is the ratio of moles of surface atoms (from surface redox peak charge) to the total
number of moles in catalyst loading. (Not applicable to the catalysts mixed with Ag, because the
corresponding peak is not separable from that belonging to Ag)
rex: Exchanged oxygen atoms to the total oxygen atoms in the catalyst loading
yex: Exchanged oxygen atoms to the oxygen atoms on the surface obtained from redox peak
18
# Isotope exchange in H2 O-free solution after Co3O4 labeling (i.e. in series B)
18
3.2. Oxygen exchange in Co3O4 (10 µm) in 20% marked H2 O solution
It is widely recognized that the particle size of the catalyst influences its activity.28 Therefore, we
investigated this effect using RRDE and isotope labeling measurements, see Fig. S5 and S6. The
results (Tables 1 and 2) proved the isotope exchange mechanism also for the large particles.
According to the ball model, this catalyst should have a surface area of only about one hundredth
of that of the 50 nm catalyst layer. The BET data, however, show that the real surface area is much
9
ACS Paragon Plus Environment

Analytical Chemistry
Page 10 of 16
1
2
3
4
5
6
7
8
9
larger due to the intrinsic roughness of these larger particles. The real surface area that is obtained
from the isotope exchange method in Series A at the 10 µm catalyst is roughly half of that
obtained at the 50 nm catalyst suggesting an even larger roughness effect than the BET data.
Correspondingly, the percentage of active surface sites as determined from the isotope exchange is
twice as large as for the 50 nm particles (24% vs. 12 %).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
3.3. Oxygen exchange at Ag+ Co3O4 (10 w%) mixed catalyst in 50% marked H2 O solution
Recent literature showed that the mixed Ag 311+ Co3O4 catalyst has a marked catalytic activity for
OER and ORR in alkaline media.4,11,12 Ag 311 stands for Ag particles from Ferro GmbH (and not
for a crystallographic orientation). The enhanced bifunctionality was attributed possibly to a redox
switching in Co3O4 (i.e. transitions between Co3O4 and Co(OH)2).4 We thus investigated OER on
the mixed catalyst based on Ag 311 (1 µm) + Co3O4 (50 nm) 10 w% in 0.5M LiOH solution
18
containing 50% (w/w) of marked H2 O, as shown in Fig. 4a. The redox peaks observed in the CV
are assigned to oxidation and reduction of Ag and cobalt and their oxides. Ag is oxidized to AgI2O
and then further oxidized to AgIAgIIIO2.11
In series A of experiments on the mixed catalyst, an increase of 16O18O concentration with a
16
concomitant decrease of O2 concentration with consecutive scans is observable, Fig. S7a. While
in series B of experiments, a decrease of concentration of O18O with a concomitant increase in
16
16O2 concentration with each cycle is noticed, see Fig. S7b. This behavior is similar to that
observed on pure Co3O4 catalyst. Again, this supports the contribution of the oxide, and also could
be the Ag-oxide since it is oxidized at higher potentials, in the oxygen exchange process. The total
amount of 18O atoms exchanged during the successive scans corresponds to ~1% of the total
loading with respect to only Co3O4 mass since Ag is mostly inactive for OER. The surface redox
peak of cobalt oxide in the mixture is not separable from that belonging to Ag, and thus calculation
of the rsurf and yex values is not applicable for the mixed catalyst. In series B, 0.3% is obtained,
which is less than that in series A. This might be due to 18O interdiffusion or loss upon rinsing.
The above results show that the amount of surface atoms in the mixed Ag+ Co3O4 catalyst taking
part in the oxygen exchange mechanism is higher than that observed on Co3O4 catalyst (~1% for
the mixed catalyst compared to 0.2% for the Co3O4 catalyst). Roughly 50% of the surface atoms
are active (for isotope exchange) as compared to 12 % for the pure 50 nm Co3O4 catalyst.
Therefore, the surface of the mixed catalyst is more active than that of Co3O4. This suggests that
OER takes place at the interphase between Co3O4 and Ag particles is important for OER in the
mixture, but also that a spillover of oxygen atoms seems to be involved. This is important for the
interpretation of the improved catalytic activity observed on the mixed catalyst.4 Thus, the real
surface areas obtained from BET, ball model and isotope exchange model revealed good
10
ACS Paragon Plus Environment

Page 11 of 16
Analytical Chemistry
1
2
3
4
5
6
7
8
qualitative agreement, illustrating the suitability of the used methods for the characterization of our
catalysts, as displayed in Table 2. However, the results based on the DL capacitance method
deviate from the other methods which might be because part of the exposed surface is oxide and
the rest is metallic Ag.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 4. DEMS results for 6 cycles OER at Ag + Co3O4 (10 w%) (400µg/cm2)/GC electrode in the
new small volume DEMS cell in (a) Ar-saturated 0.5M LiOH solution containing 50% (w/w) of
18
16
marked H2 O, scan rate=10 mV/s. (b) in 0.5M LiOH/H2 O after the electrode has been pre-
18
treated in 0.5M LiOH/H2 O. The upper panel displays the faradaic current, while the lower ones
represent the ionic currents for m/z 32, 34 and 36.
4. Discussion
A comparison of the results of different approaches to estimate the true surface area including the
amount of exchanged oxygen (n) is given in Table 2. For the small particle size catalyst (50 nm),
the ball model, redox peak model and the BET data revealed relatively good consistency, which
indicates their smooth surface. However, for the larger particles (40 µm), the surface area
estimated from the redox peak method is much larger than that obtained from the ball model,
which seems to be due to their roughness. The results also show that the DL capacitance method
gives higher true surface area than the BET method as the BET-determined surface area has not
direct correlation to the electrochemically active area. Moreover, as can be noticed from the Table,
the true surface area (roughness factor) obtained from the redox peak method for Co3O4 (50 nm) is
roughly one-third to half of the BET value, indicating that only a part of the surface is active for
surface redox reactions. Exception is the experiment in 0.1M LiOH in the thin layer cell, where the
11
ACS Paragon Plus Environment

Analytical Chemistry
Page 12 of 16
1
2
3
4
5
6
7
8
ohmic resistance was too high so that the peaks are not well separated, thus the uncertainty in peak
charge could be high.22 In an independent experiment at Co3O4 (50 nm) with 1 mg cm-2 and 50 mV
s-1 in RRDE cell, the area calculated using the redox peak method is roughly one-third of the BET
value. This ratio is consistent with the results in our new cell.
9
Moreover, the true surface area obtained from isotope exchange method is a fraction (roughly a
quarter) of the surface area obtained from the redox peak method or corresponds roughly to 12 %
of the BET surface, suggesting that only a part of the surface atoms is exchanged. In particular
there is no indication of an exchange of subsurface oxygen atoms. No significant effect of catalyst
loading on the amount of exchanged oxygen of the total catalyst is observed. However, larger
surface area is obtained as the particle size decreases: the average true surface areas (obtained from
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
isotope exchange experiments) are 0.007, 0.006 and 0.02 cm2 /g for the 40 µm, 10 µm and 50 nm
tr
Co3O4 catalysts, respectively. Additionally, the ‘active’ surface area (calculated from isotope
experiments) for the 50 nm catalyst increases as the loading of the catalyst increases from 200
µg/cm2 to 400 µg/cm2, but the surface area obtained from DL capacitance and redox peak methods
does not increase accordingly, which could be due to the limited conductivity within the layer, see
Table 2. The fraction of oxygen atoms of the catalyst participating in oxygen exchange mechanism
approaches 0.2% of the total loading, independent of the loading. The true surface area of different
sizes of Co3O4 was in the order: 50 nm>10 µm>40 µm which is reasonable, as depicted in Table 2.
Therefore, the transition from higher oxidation state oxide to lower oxidation state oxide, and thus
releasing oxygen, is possibly involved during OER at Co3O4. This can be via an oxygen exchange
mechanism, in which surface lattice oxygen atoms of the oxide are exchanged with oxygen atoms
from the solution. A scheme correlating oxygen evolution reaction and the redox transitions in
Co3O4 catalyst is proposed in Fig. 5. In this scheme, adsorption of OH− species via one electron
transfer is the first step, and then followed by the formation of O−O bond in the OOH group. The
terminal step involves the disproportionation of higher valent oxide (CoO2) to lower valent oxide
(CoO), and thus releasing oxygen.
From the results of our work and the literature, isotope labeling experiments together with mass
spectrometry studies have revealed that a lattice oxygen exchange mechanism is dependent on
structure, composition, metal-oxygen covalency and crystallinity of the catalyst.18,23,27,32 For
example, Shao-Horn et al. showed the lack of oxygen exchange on polycrystalline film and
nanocrystalline particles RuO2 in contrast to the results of Wohlfahrt-Mehrens and Krtil and their
coworkers who demonstrated the involvement of lattice oxygen during OER. This discrepancy
might be due to the different surface crystallinity or the presence of defect active sites that are
formed during the different synthesis procedures of polycrystalline RuO2.18,23 Another parameter
that influences the involvement of lattice oxygen in OER is the metal-oxygen covalency as
12
ACS Paragon Plus Environment

Page 13 of 16
Analytical Chemistry
1
2
3
4
demonstrated for perovskites by Shao-Horn et al.32 This difference in behavior emphasizes the
critical role of the preparation conditions on the mechanism.
5
6
7
8
Further application of the isotope labeling procedure to other catalysts and systematic investigation
of the effect of potential window could be valuable for mechanistic investigations. Moreover, the
new cell design, first, allows the use of small volume of electrolyte, second, offers the use of
particle layers and in particular those which are not stable enough for working under flow-through
conditions, and third, provides a very short delay time. The exploitation of these features could add
new opportunities for future studies: characterization of submonolayer amounts of organic
adsorbates on electrode surfaces because of its high sensitivity; batteries and fuel cells research as
well as evaluation of enzyme activities.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Summary of results of estimation of true surface area of different catalysts and loadings
using different methods.
Surface
Method
Catalyst
Isotope
exchange
exp. A
atoms
exchanged peak
(per BET
area)
Redox
Isotope
exchange
exp. B#
Ball
model
DL-cap.
method
BET
method
Co3O4(40µm) 100
µg/cm2 in 0.5M KOH
with 2% isotope till
1.62V 22
Co3O4(50nm) 200
µg/cm2 in 0.1M LiOH
with 10% isotope till
1.8V 22
Co3O4 (50 nm) 400
µg/cm2 in 0.8 M LiOH
with 20% isotope till
1.6V
Co3O4 (50 nm) 400
µg/cm2in 0.5 M LiOH
with 50% isotope till
1.65 V (1.60 V)*
Atr
tr
0.016
0.03
39
--
--
12
22
0.72
1.34
3.9
2.1
3.9
32
0.28
0.51
1.8
(cm2 /cm2
)
geo
--
12%
12%
8(13)%
24%
50%
40%
n (nmol/cm2
)
)
)
)
)
)
)
geo
Atr
33
61
66
122
66
122
14
25
6.6
12
9.9
18
140
255
68
(cm2 /cm2
)
geo
tr
n (nmol/cm2
72
7.3
59
3.4
geo
Atr
79
7.3
29
1.8
(cm2 /cm2
)
geo
tr
n (nmol/cm2
145
79
126
13.5
5.2(8)
9.7(15)
3.4
53
3.4
geo
Atr
105(89)
23(22)
42(41)
--
2.4(0.7)
4.5(1.3)
1.3
(cm2 /cm2
)
geo
tr
n (nmol/cm2
145
0.6
1.0
8
194(165)
25
geo
Co3O4 (10 µm) 600
µg/cm2 in 0.8 M LiOH
with 20% isotope till
1.61 V
Atr
(cm2 /cm2
)
geo
tr
n (nmol/cm2
45
6.2
--
2.4
geo
£Ag+Co3O4 (50 nm) 10
Atr
77
3.2
--
0.9
w%, 400 µg/cm2
(cm2 /cm2
)
geo
tr
in 0.5 M LiOH with
50% isotope till 1.65 V
£Ag+Co3O4 (50 nm) 10
w%, 600 µg/cm2
in 0.8M LiOH with
20% isotope till 1.65 V
n (nmol/cm2
15
142
44
5.9
--
1.6
geo
Atr
12
3.8
--
0.1
(cm2 /cm2
)
geo
tr
n (nmol/cm2
22
81
7.1
--
0.2
geo
18
# Isotope exchange in H2 O-free solution after Co3O4 labeling (i.e. in series B)
£ Calculations were done with respect to only the mass of Co3O4 (50 nm) in the mixture.
* Repeated experiments with different upper potential limit.
13
ACS Paragon Plus Environment

Analytical Chemistry
Page 14 of 16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 5. Proposed scheme for oxygen evolution reaction on cobalt-containing spinel catalyst in
alkaline media. After one OER cycle, some of O-atoms in the oxide, Co2+O, will be exchanged
with 18O atoms (red).
5. Conclusions
In this study, we extend our investigations of oxygen exchange process to different particle sizes
of Co3O4 as well as to the mixed Ag+Co3O4 catalyst. For the first time, we present a new small-
volume DEMS cell design, which allows measurements on nanoparticle-modified disc electrodes
and uses only sub-milliliters of electrolyte during measurement. The feasibility of the new cell is
achieved by monitoring OER reaction. This cell shows a good performance and a very short delay
18
16
time (1−2 s). We found that in H2 O-marked solution, the concentration of O18O increases in
16
parallel to a decrease in O2 concentration with each cycle before reaching a steady-state value.
Thus suggests that the oxide takes part in OER. Only the outermost parts of the catalyst are active:
~ 0.2 % of the total loading. Interestingly, the amount of oxygen exchanged on the mixed Ag +
Co3O4 catalyst is higher than that at the single Co3O4, which can be a possible origin of the
improved electrocatalytic activity previously observed on the mixed catalyst. The real surface area
of the catalysts is estimated using different methods and the results are compared. The
electroanalytical approach presented here can lead to more in-depth studies that promote advances
in electrocatalysis and electroanalysis.
Supporting Information
Experimental details; RRDE measurements; Additional isotope exchange results on Co3O4 (50
nm), Co3O4 (10 µm) and the mixed Ag+Co3O4; SEM of the mixed catalyst.
Acknowledgements
14
ACS Paragon Plus Environment

Page 15 of 16
Analytical Chemistry
1
2
3
4
5
6
7
8
This work was supported by the Federal Ministry of Education and Research of Germany (BMBF)
through the project "Strom aus Luft und Li" (03X4624A). H.M. Amin acknowledges the financial
support from MoHE (Egypt) and DAAD. The authors thank S. K. Eswara and U. Kaiser for
performing SEM and D. Wittmaier and K. A. Friedrich for providing the BET data.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6. References
(1) McCrory, C. C. L.; Jung, S.; Ferrer, I. M.; Chatman, S. M.; Peters, J. C.; Jaramillo, T. F. J. Am. Chem. Soc. 2015, 137, 4347-4357.
(2) Trotochaud, L.; Young, S. L.; Ranney, J. K.; Boettcher, S. W. J. Am. Chem. Soc. 2014, 136, 6744-6753.
(3) Reier, T.; Oezaslan, M.; Strasser, P. ACS Catal. 2012, 2, 1765-1772.
(4) Amin, H.; J. Bondue, C.; Eswara, S.; Kaiser, U.; Baltruschat, H. Electrocatalysis 2017, 8, 540-553.
(5) Eßmann, V.; Barwe, S.; Masa, J.; Schuhmann, W. Anal. Chem. 2016, 88, 8835-8840.
(6) Koza, J. A.; He, Z.; Miller, A. S.; Switzer, J. A. Chem. Mater. 2012, 24, 3567-3573.
(7) Lee, Y.; Suntivich, J.; May, K. J.; Perry, E. E.; Shao-Horn, Y. The Journal of Physical Chemistry Letters 2012, 3, 399-404.
(8) Trasatti, S. J. Electroanal. Chem. 1980, 111, 125-131.
(9) Bockris, J. O. M.; Otagawa, T. J. Electrochem. Soc. 1984, 131, 290-302.
(10) Hoang, T. T. H.; Gewirth, A. A. ACS Catalysis 2016, 6, 1159-1164.
(11) Amin, H. M. A.; Baltruschat, H.; Wittmaier, D.; Friedrich, K. A. Electrochim. Acta 2015, 151, 332-339.
(12) Wittmaier, D.; Wagner, N.; Friedrich, K. A.; Amin, H. M. A.; Baltruschat, H. Journal of Power Sources 2014, 265, 299-308.
(13) Jiang, Z. Q.; Jiang, Z. J.; Maiyalagan, T.; Manthiram, A. J. Mater. Chem. A 2016, 4, 5877.
(14) Rossmeisl, J.; Qu, Z. W.; Zhu, H.; Kroes, G. J.; Nørskov, J. K. Journal of Electroanalytical Chemistry 2007, 607, 83-89.
(15) Suntivich, J.; May, K. J.; Gasteiger, H. A.; Goodenough, J. B.; Shao-Horn, Y. Science 2011, 334, 1383-1385.
(16) Plaisance, C. P.; van Santen, R. A. J. Am. Chem. Soc. 2015, 137, 14660-14672.
(17) Hamdani, M.; Singh, R. N.; Chartier, P. International Journal of Electrochemical Science 2010, 5, 556-577.
(18) Stoerzinger, K. A.; Diaz-Morales, O.; Kolb, M.; Rao, R. R.; Frydendal, R.; Qiao, L.; Wang, X. R.; Halck, N. B.; Rossmeisl, J.; Hansen, H. A.;
Vegge, T.; Stephens, I. E. L.; Koper, M. T. M.; Shao-Horn, Y. ACS Energy Letters 2017, 2, 876-881.
(19) Wittmaier, D.; Cañas, N. A.; Biswas, I.; Friedrich, K. A. Advanced Energy Materials 2015, 5, 1500763.
(20) Bergmann, A.; Martinez-Moreno, E.; Teschner, D.; Chernev, P.; Gliech, M.; de Araújo, J. F.; Reier, T.; Dau, H.; Strasser, P. Nature
Communications 2015, 6, 8625.
(21) Liu, Y.-C.; Koza, J. A.; Switzer, J. A. Electrochim. Acta 2014, 140, 359-365.
(22) Amin, H. M. A.; Baltruschat, H. Phys. Chem. Chem. Phys. 2017, 19, 25527-25536.
(23) Wohlfahrt-Mehrens, M.; Heitbaum, J. J. Electroanal. Chem. Interfacial Electrochem. 1987, 237, 251-260.
(24) Fierro, S.; Nagel, T.; Baltruschat, H.; Comninellis, C. Electrochem. Commun. 2007, 9, 1969-1974.
(25) Fierro, S.; Nagel, T.; Baltruschat, H.; Comninellis, C. Electrochem. Solid State Lett. 2008, 11, E20-E23.
(26) Fierro, S.; Ouattara, L.; Calderon, E. H.; Passas-Lagos, E.; Baltruschat, H.; Comninellis, C. Electrochim. Acta 2009, 54, 2053-2061.
(27) Roy, C.; Sebok, B.; Scott, S. B.; Fiordaliso, E. M.; Sørensen, J. E.; Bodin, A.; Trimarco, D. B.; Damsgaard, C. D.; Vesborg, P. C. K.; Hansen,
O.; Stephens, I. E. L.; Kibsgaard, J.; Chorkendorff, I. Nature Catalysis 2018, 1, 820-829.
(28) Jirkovský, J.; Makarova, M.; Krtil, P. Electrochem. Commun. 2006, 8, 1417-1422.
(29) Wohlfahrt-Mehrens, M.; Heitbaum, J. Journal of Electroanalytical Chemistry 1987, 237, 251-260.
(30) Macounova, K.; Makarova, M.; Krtil, P. Electrochemistry Communications 2009, 11, 1865-1868.
(31) Fabbri, E.; Nachtegaal, M.; Binninger, T.; Cheng, X.; Kim, B.-J.; Durst, J.; Bozza, F.; Graule, T.; Schäublin, R.; Wiles, L.; Pertoso, M.;
Danilovic, N.; Ayers, K. E.; Schmidt, T. J. Nature Materials 2017, 16, 925.
(32) Grimaud, A.; Diaz-Morales, O.; Han, B.; Hong, W. T.; Lee, Y.-L.; Giordano, L.; Stoerzinger, K. A.; Koper, M. T. M.; Shao-Horn, Y. Nature
Chemistry 2017, 9, 457.
(33) Baltruschat, H. Journal of the American Society for Mass Spectrometry 2004, 15, 1693-1706.
(34) Trimarco, D. B.; Pedersen, T.; Hansen, O.; Chorkendorff, I.; Vesborg, P. C. K. Review of Scientific Instruments 2015, 86, 075006.
(35) Wang, H.; Rus, E.; Sakuraba, T.; Kikuchi, J.; Kiya, Y.; Abruña, H. D. Anal. Chem. 2014, 86, 6197-6201.
(36) Hartung, T.; Baltruschat, H. Langmuir 1990, 6, 953-957.
(37) Heinen, M.; Chen, Y. X.; Jusys, Z.; Behm, R. J. Electrochim. Acta 2007, 52, 5634-5643.
(38) Bawol, P. P.; Reinsberg, P. H.; Baltruschat, H. Anal. Chem. 2018, 90, 14145-14149.
15
ACS Paragon Plus Environment

Analytical Chemistry
Page 16 of 16
1
2
(39) Bani Hashemi, A.; La Mantia, F. Anal. Chem. 2016, 88, 7916-7920.
3
4
5
6
(40) Trimarco, D. B.; Scott, S. B.; Thilsted, A. H.; Pan, J. Y.; Pedersen, T.; Hansen, O.; Chorkendorff, I.; Vesborg, P. C. K. Electrochim. Acta 2018,
268, 520-530.
(41) Levine, S.; Smith, A. L. Discussions of the Faraday Society 1971, 52, 290-301.
7
8
9
TOC:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16
ACS Paragon Plus Environment