Tailoring Catalytic Properties of Copper Manganese Oxide Nanoparticles (Hopcalites-2G) via Flame Spray Pyrolysis

Article abstract of DOI:10.1002/cctc.201800639

Humidity tolerant copper manganese oxide catalysts of the second generation (“Hopcalites-2G”) were produced by flame spray pyrolysis process (FSP) and their catalytic activity for carbon monoxide oxidation under dry and humid conditions was evaluated. The effects of synthesis parameters such as flame specific variables (OFR=oxygen to fuel ratio, sheath gas, reactor) and the precursor specifications (composition, concentration, solvent) were investigated systematically. From the findings of the synthesis variations, it was possible to identify key factors decisive to achieve a high catalytic activity for CO oxidation under humid conditions. The most important characteristics are phase composition, specific surface area and the carbon content to achieve increased long-term stability of the flame made catalysts under humid conditions. A high productivity in the flame-based Hopcalite synthesis was achieved with catalysts performances similar or better than that of the commercial Hopcalite catalyst Carulite 300 (CARUS). The most active catalyst showed 50 % higher CO conversion after one hour on stream at 50 °C and humid conditions compared with the reference. The latter is a crucial step towards an improved humidity tolerance of Hopcalite catalysts in respiratory filter applications.

Full text of DOI:10.1002/cctc.201800639

Accepted Article
Title: Tailoring Catalytic Properties of Copper Manganese Oxide
Nanoparticles (Hopcalites-2G) via Flame Spray Pyrolysis
Authors: Julia Grothe, Karl Wegner, Max Medicus, Elke Schade, and
Stefan Kaskel
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201800639
Link to VoR: http://dx.doi.org/10.1002/cctc.201800639
A Journal of
www.chemcatchem.org

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
Tailoring Catalytic Properties of Copper Manganese Oxide
Nanoparticles (Hopcalites-2G) via Flame Spray Pyrolysis
K. Wegner,[a] M. Medicus,[a] E. Schade,[a] J. Grothe,*[a] and S.Kaskel[a]
Abstract: Humidity tolerant copper manganese oxide catalysts of
the second generation (“Hopcalites-2G”) were produced by flame
spray pyrolysis process (FSP) and their catalytic activity for carbon
monoxide oxidation under dry and humid conditions was evaluated.
The effects of synthesis parameters such as flame specific variables
(OFR = oxygen to fuel ratio, sheath gas, reactor) and the precursor
To overcome these disadvantages, metal and mixed metal
oxides have been insensitively studied as possible catalysts.[5]
As
a well-established and commercially available catalyst
system copper-manganese mixed oxides also known as
“Hopcalites” are in use.[6] This relatively inexpensive, non-toxic
alternative has been intensively studied during the last 100
years and many different approaches such as sol-gel synthesis[7],
specifications
(composition,
concentration,
solvent)
were
investigated systematically. From the findings of the synthesis
variations, it was possible to identify key factors decisive to achieve
a high catalytic activity for CO oxidation under humid conditions. The
most important characteristics are phase composition, specific
surface area and the carbon content to achieve increased long-term
stability of the flame made catalysts under humid conditions. A high
productivity in the flame-based Hopcalite synthesis was achieved
with catalysts performances similar or better than that of the
commercial Hopcalite catalyst Carulite 300 (CARUS). The most
active catalyst showed 50 % higher CO conversion after one hour on
stream at 50 °C and humid conditions compared with the reference.
The latter is a crucial step towards an improved humidity tolerance of
Hopcalite catalysts in respiratory filter applications.
co-precipitations[8],
redox
methods[9],
mechanochemical
synthesis[10] and supercritical antisolvent precipitation[11] were
used to optimize the catalytic activity. However, a major
unresolved shortcoming is still their fast deactivation under
humid conditions which restrict them for an application under
ambient conditions[12]. In order to overcome these shortcomings,
Kunkalekar et al. describe a method of MnO2 doping with copper
and palladium.[13] Liu et al. proposed the addition of 8.5 wt% Sn
to suppressed the water adsorption on Hopcalite surfaces
resulting in improved catalytic activity under moisture.[14] The
modification of cryptomelane by copper was found to improve
the water tolerance during CO oxidation.[15] Njagi et al.
developed a redox method resulting in water stable Hopcalites
up to 15 min on a gas stream with 3 % water at ambient
temperature.[9]
All these investigations show the possibilities for an increase of
the water resistance of catalysts in the CO oxidation. However,
multiple synthesis steps partially involving precious metals are
necessary including substrate production, precipitation, ageing,
filtering, washing, calcination or doping. In addition, the
adjustment of optimal synthesis conditions is often time
consuming and crucial regarding the optimization of each
synthesis step as well as the reproducibility.[16]
Introduction
The development of novel catalytic or adsorption based filter
materials for respiratory filters or indoor air quality improvement
is a crucial task to reduce the risk of exposure to toxic gases.[1,2]
A highly toxic and ubiquitous pollutant, produced by incomplete
industrial, household or engine-based combustion processes is
carbon monoxide. This invisible and odorless gas causes many
damages including nausea, fainting, coma and leads to death on
prolonged exposure even at concentrations below 100 ppm.[3]
Due to the abundant but unspecific exposure sources, several
professional groups such as firefighters, miners and soldiers are
affected. For personal protection the use of respiratory systems
filled with catalyst which decomposes the highly dangerous CO
into the less harmful CO2 is state of the art. In general, the
highest catalytic activity could be achieved with precious metals
supported on metal oxides. Haruta et al. found gold
nanoparticles supported on α-Fe2O3 to be very active even at
temperatures as low as -70°C.[4] However, these systems have
not found industrial applications yet due to the high cost, severe
ageing and poisoning during storage and other ill-defined
deactivation processes.
Flame spray pyrolysis (FSP) is a scalable and continuous
production method for metal and mixed metal oxide
nanoparticles suitable for large scale industrial production.[17–19]
Characteristic for this process is the formation of a combustible
aerosol consisting of volatile metal precursors dissolved in
organic solvents. The ignition of the formed aerosol droplets and
the subsequent combustion of the precursor results in the
formation of uniform shaped nanoparticles with narrow size
distribution.[20] The possibility of producing almost any mixed
metal oxide makes the FSP ideal for
a wide range of
applications including catalysis, sensors, phosphors and electro
ceramics.[21] Especially the high specific surface area of
nanoparticles is of great importance for catalytic applications. In
our previous work we reported the synthesis of Hopcalite
nanoparticles via FSP for the first time.[22] In a single-stage
process, an improved moisture stability of the catalyst in CO
oxidation could be achieved. We found out that the increased
water stability in comparison to commercial Hopcalite (Carulite
300, CARUS) is caused by residual carbonaceous coating on
the catalyst particle surface acting as a protective hydrophobic
membrane. In order to achieve a better understanding of the
[a]
K. Wegner, M. Medicus, E. Schade, Dr. J. Grothe, Prof. S. Kaskel
Department of Inorganic Chemistry
University of Technology Dresden, Neubau Chemische Institute
Bergstrasse 66, 01069 Dresden
E-mail: Julia.Grothe@tu-dresden.de
Supporting information for this article is given via a link at the end
of the document.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
membrane origin and the process variables affecting the
catalytic CO oxidation performance under humid conditions we
present here a systematic investigation the synthesis variables
in the FSP process and its effect on the structural and textural
Hopcalite properties. In particular the solvent, the precursor and
the oxygen to fuel ratio (OFR) were investigated. Moreover, a
tubular quartz tube reactor was tested to form a thicker carbon
membrane enwrapping the catalyst particles following a coating
methodology published by Ernst et al. for Pt@C.[23] Furthermore,
scale up experiments were carried out in order to increase the
production rate while retaining the improved catalytic activity.
For an in-depth insight into structure and textural characteristics
the Hopcalites synthesized with different synthesis conditions
were characterised by means of X-ray powder diffraction,
transmission electron microscopy and nitrogen physisorption.
Elemental analyses revealed the carbon content and with
subsequent water adsorption isotherms we deduce structure
property relationships in particular focussed on the
hydrophobicity of the materials and the consequences for
achieving an increased catalytic activity in the presence of water
vapor.
(
)
풓풆풂풍
풎풐풍 푶 / 풎풐풍 풇풖풆풍
ퟐ
푶푭푹 =
(1)
(
)
풔풕풐풊풄풉
풎풐풍 푶 / 풎풐풍 풇풖풆풍
ퟐ
During the syntheses an increasing flame length could be
observed with decreasing OFR. The reason for this is the
compensation of the oxygen deficiency with oxygen from
ambient air. In all cases the Cu1.5Mn1.5O4 (ICSD [01-070-0262])
mixed spinel was identified as the main crystalline phase by
means of powder X-ray diffraction. Due to the excess of
manganese in the precursor solution, which was chosen
because of the best catalytic activities reported in earlier works,
an additional phase consisting of Mn3O4 (ICSD [01-089-4837])
appears (Figure S1). Peak narrowing of the reflexes indicates
the crystal size of the mixed spinel is slightly increasing with
decreasing OFR in both cases, without but also with additional
nitrogen sheath gas surrounding the flame. The obtained
increase in flame length leads to a longer residence time of the
generated particles in the high temperature regime and therefore
to bigger crystals. The OFR is also influencing the specific
surface area as shown in Figure 1. The longer residence time of
the particles in the flame not only results in higher crystallinity
but also in the formation of larger aggregates and agglomerates.
The elemental analysis for all samples revealed carbon contents
below 0.3 wt% (Table 1). Regarding these results, it must be
assumed that the variation of the OFR as well as additional
nitrogen sheath gas cannot provide synthesis conditions which
result in higher carbon contents of the nanoparticles. The
catalytic studies showed only low CO conversion rates at
temperatures below 100 °C for all samples, which is mainly due
to their low specific surface areas and the associated poor
desorption of the resulting CO2. The obtained T50 and T90 values
for the samples with an OFR of 1.0 are shown in Table S1.
Results and Discussion
OFR and sheath gas influence
The oxygen to fuel ratio plays a key role in controlling the
incomplete combustion and thus the formation of carbon
deposits responsible for the hydrophobic functionalization of the
Hopcalite nanoparticles. Variation of the oxygen to fuel ratio
(eq. 1) from 1.0 to 0.4 as well as additional nitrogen sheath gas
during FSP were investigated regarding their influence on
structural particle properties. Incomplete precursor combustion
was expected to achieve higher carbon contents on the
nanoparticles resulting in higher catalytic activity during carbon
monoxide oxidation under humid conditions. The textural
properties and compositions of the obtained particles are listed
in Table 1.
100
90
80
Table 1. Textural properties of FSP derived Hopcalites synthesized with
different OFR.
70
[a]
[b]
[c]
Sample
OFR
N2S
[slm]
SSABET
[m2/g]
dBET
dXRD
C[d]
[wt.%]
60
[nm]
[nm]
0 slm N2S
OFR-1 co
OFR-2 co
OFR-3 co
OFR-4 co
OFR-5 co
OFR-6 co
OFR-7 co
OFR-8 co
OFR-9 co
OFR-10 co
OFR-11 co
OFR-12 co
1.0
0.8
0.6
0.4
1.0
0.8
0.6
0.4
1.0
0.8
0.6
0.4
0
0
0
0
10
10
10
10
20
20
20
20
95
77
56
51
80
76
57
46
81
69
76
52
12
14
20
22
14
15
19
24
14
16
15
21
10
12
14
14
11
11
12
16
11
12
10
13
0.21
0.13
n.d.
10 slm N2S
50
20 slm N2S
n.d.
0.11
0.11
0.14
0.09
0.17
0.15
0.14
0.11
1.0
0.8
0.6
0.4
OFR
Figure 1. Specific surface areas from Hopcalites synthesized varying OFRs
and sheath gas flows.
[a] Derived from nitrogen physisorption at 77 K with multipoint BET analysis at
p/p0= 0.05–0.20.
[b] Calculated from Eq.(5) .
[c] Calculated from Eq.(4).
[d] Determined by combustion method with CHNS-Analyzer.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
Solvent variation
80
70
60
50
40
30
20
10
0
-4500
-4000
-3500
-3000
-2500
-2000
-1500
-1000
-500
The solvent affects the flame temperature due to variations in
combustion enthalpies but also may partially play a role as
carbon precursor in the formation of the hydrophobic membrane.
Therefore, experiments with different solvents were carried out
and the materials were investigated focusing on structural
particle properties. The textural properties and compositions of
the obtained particles are listed in Table 2.
Table 2. Textural properties of FSP derived Hopcalites using different
solvents.
0
Figure 2. Specific surface areas of Hopcalites synthesized from different
solvents compared with their combustion enthalpy.
[a]
[b]
[c]
Sample
Solvent
SSABET
[m2/g]
dBET
dXRD
C[d]
[nm]
[nm]
[wt.%]
MA-1 co
MFA-1 co
Et-1 co
M:A = 9
M:FA = 99
E
THF
H
79
78
73
60
47
63
14
14
15
18
23
17
12
11
13
18
18
16
0.15
0.19
0.14
0.07
0.16
0.13
Quartz tube adaption
THF-1 co
H-1 co
Following the method proposed by Ernst et al., a quartz tube
was connected to the FSP nozzle to investigate whether an
inert-like atmosphere could be generated in which the desired
OFR was obtained to foster the formation of carbon deposits
ideally as a carbon coating of the Hopcalite nanoparticles.[23]
Xylene was used as the solvent in the syntheses because, with
more than 90 % by weight, it has the largest carbon content of
the solvents used and it is known to form a soot-burning flame.
The synthesis conditions were selected in such a way that the
influence of the OFR, the solvent flow and the precursor
concentration could be investigated. The textural properties and
compositions of the obtained particles are listed in Table 3.
X-1 co
X
[a] Derived from nitrogen physisorption at 77 K with multipoint BET analysis at
p/p0= 0.05–0.20.
[b] Calculated from Eq. (5).
[c] Calculated from Eq. (4).
[d] Determined by combustion method with CHNS-Analyzer.
To elucidate correlations between particle properties and
solvents, the spraying conditions were kept constant having an
OFR of 1.0.
In Figure 2 It can be seen that higher combustion enthalpies of
the solvent lead to smaller BET surface areas. This trend can be
explained with a higher flame temperature and thus increased
probability for sintering processes as well as aggregate
formation. The associated increase in the particle size leads to
smaller specific surface areas. Using xylene as a solvent, a
relatively high surface area with respect to the combustion
enthalpy can be achieved. This deviation might be associated
with the fact, that xylene is the only aromatic solvent and
therefore shows a different decomposition behavior during
pyrolysis. In addition, due to the higher carbon content of xylene,
also a higher dispersion oxygen stream is required in order to
achieve the same OFR value as compared to the other solvents
(20 % more than hexane), which in turn has an influence on the
flame length and - temperature as well as the residence time of
the particles in the flame. All samples were investigated by
means of powder X-ray diffraction resulting in the same crystal
structures (Figure S2). The solvent does not have an influence
on the crystal phase but on the crystal size which increases with
increasing combustion enthalpy. This can be explained by the
higher energy input promoting the crystal growth. The elemental
analysis revealed carbon contents below 0.2 wt%. Regarding
these results, it must be assumed that the influence of the
solvent on the production of carbon species forming a carbon
coating on the catalyst particles is too low. The combustion
temperature of the solvent is much lower compared to that of the
precursor. Consequently, the particle formation is independent
of solvent evaporation and combustion. The catalytic
performance is listed in Table S2.
Table 3. Textural properties of FSP derived Hopcalites with quartz tube.
[a]
[b]
[c]
Sample
OFR
SSABET
[m2/g]
dBET
dXRD
C[d]
[wt.%]
HI75[e]
[nm]
[nm]
Q-1
0.8
1.0
1.0
1.0
1.1
1.0
0.9
0.8
340
51
35
27
25
26
28
98
5
-
94.87
2.39
1.12
0.10
0.09
0.07
0.10
16.53
0.17
0.33
0.38
0.40
0.39
0.43
0.40
0.25
Q-2 co
22
32
40
44
42
39
11
28
22
16
26
28
28
15
Q-3 sm-s
Q-4 sm-s
Q-5 sm-s
Q-6 sm-s
Q-7 sm-s
Q-8 sm-s
[a] Derived from nitrogen physisorption at 77 K with multipoint BET analysis at
p/p0= 0.05–0.20.
[b] Calculated from Eq. (5).
[c] Calculated from Eq. (4).
[d] Determined by combustion method with CHNS-Analyzer.
[e] Calculated from Eq. (2).
Samples Q-2 to Q-4 show the influence of precursor flow and
concentration on the Hopcalite particles. From the XRD
measurements an increase of the Mn3O4 phase in comparison to
the particles synthesized without quartz tube could be obtained.
After the synthesis, the inside wall of the quartz tube showed a
copper coating which explains the lack of copper in the product.
In addition, the synthesis conditions with quartz tube (high
temperature, small temperature gradient) promote the formation
of Mn3O4. These conditions yield also a higher crystallinity and
increased crystallite sizes.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
100
90
80
70
60
50
40
30
20
10
0
20
19
18
17
16
15
14
3
2
1
0
Figure 4. Transmission electron mircograph of sample Q-8.
Q-5
Q-6
Q-4
Q-7
Q-3
Q-2
Q-8
Figure 3. Specific surface area and carbon content of samples synthesized
with quartz tube.
The hydrophobicity index, HI75 is a valuable number to estimate
the hydrophobicity of the particles.[22] The HI75 value is defined
by equation (2).
It is worth mentioning that crystallite size increases with
decreasing concentration. This can be explained by the fact that
more energy from the combustion is available for less crystal
nuclei for growing. In contrast to that a decrease in precursor
flow results in smaller crystallites according to XRD, since less
energy from combustion is available for crystal growth at the
same precursor concentration. In comparison, the particle
diameter calculated from the specific surface area is decreasing
for lower precursor concentrations. This apparent contradiction
can be explained by the increased carbon content not
considered in the calculations (Figure 3). The carbon is lighter
than the oxide, and therefore the mass-related surface area
increases. From the elemental analysis, an increase in the
carbon content can be observed when the metal concentration is
reduced. This can be explained by the fact that fewer metal
oxide nuclei are present per mole of solvent and thus more
carbon can be formed per product particle. In the experiments
carried out, a doubling of the carbon content could be observed
at half of the concentration. Beside the precursor concentration
and –flow, also the OFR was investigated on its influence on
particle properties. The resulted XRDs can be found in
Figure S3. Between an OFR of 1.1 to 0.9 no significant
differences in crystal phases as well as crystallite diameter could
be observed. At an OFR of 0.8 which corresponds to 80 % of the
required oxygen for a complete combustion, a peak broadening
is observed. The crystallite diameters calculated from the
Scherrer equation are only half as large as for the other OFR.
This might relate to the large amount of carbon of about 16 wt%.
A reduction in crystallite growth could be accompanied by the
formation of carbon around the particles. A reason could be that
more energy is consumed by carbon formation. In addition, the
growth could be restricted by the condensation of carbon on the
metal oxide particles.[23] The high content of carbon also
explains the high specific surface area of around 100 m2g-1
(Figure 3). From TEM investigations the carbon forms a thin
layer of 2-5 nm thickness around the oxide particles (Figure 4).
푽
풑풐풓풆,ퟎ.ퟕퟓ (푯ퟐ푶)
푯푰ퟕퟓ
=
(2)
푽
풑풐풓풆,ퟎ.ퟕퟓ (푵ퟐ)
Here, Vpore,0.75(ads) is specific pore volume accessible for the
adsorptive (ads) at the respective p/p0 = 0.75. An example for
the evaluation from a water as well as nitrogen isotherm can be
found in the Figure S4. During the investigation of all quartz tube
samples an exponential relationship between carbon content
and HI75 could be found (Figure 5). For this correlation, a pure
carbon sample (Q-1) without metal precursor was synthesized,
too. The analysis of this sample can be found in the supporting
information (Figure S5-6). However, this relationship is not
universal, since the HI75 also depends on the metal oxide
system. Despite the high hydrophobicity achieved the materials
with high carbon content showed low catalytic activity in CO
oxidation, due to the low surface areas as well as the lack of
copper in the particles, (Table S3).
100
90
80
70
60
50
40
30
20
10
0
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
HI75
Figure 5. Correlation between carbon content of the Hopcalites synthesized
with quartz tube and HI75.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
Precursor concentration and productivity
The FSP experiments with a quartz tube have shown a
possibility to significantly increase the carbon content but also
lead to catalysts with low catalytic activity due to extreme
influences on the particle properties e.g. crystal structure and
particle size. As no significant influence of the solvent on the
carbon content of the obtained particles could be observed, an
investigation of the influence of the precursor seemed to be
appropriate. In previous work the best catalytic performance
under humidity could be achieved with manganese to copper
molar ratio of 2.4 and therefore this ratio was chosen. The
textural properties and compositions of the obtained particles
are listed in Table 4.
a higher surface area and hence a higher carbon content. At
concentrations above 0.2 M no more significant decrease in the
specific surface area is observed. To get a better understanding
of the role of the carbon on the particles in the catalytic oxidation
reaction, the catalyst particles were tested under humid
conditions. The tests were carried out at 50 °C oven temperature.
For a better comparison, the CO conversion was converted into
reaction rates using equation (3).
풏
̇
푪푶 ∙ 푿
푪푶
풓푪푶
=
(3)
푺푺푨 ∙풎
풄풂풕
In Figure 8 the reaction rates for the FSP derived as well as for a
commercial Hopcalite (Carulite 300, CARUS) are shown after 10
and 20 min on stream, respectively. After 10 min only 3 samples
have higher reaction rates than the Carulite 300 while after
20 min all 6 samples surpass the reference. The catalyst with
the highest surface area shows an outstanding reaction rate
more than twice as high as the Carulite 300 after 10 min and
almost 4.5 times higher after 20 min. All samples investigated
show much lower deactivation behavior between 10 and 20 min
in comparison to the reference. While the reaction rate of the
reference decreases by 50 %, the reaction rate of sample SU-1
only decreases by ca. 7 %. The reaction rate under humid
conditions correlates well with the carbon content of the samples
(Figure 8). Obviously the carbon deposits mainly influence the
long term stability under humid conditions while the initial
catalytic activity is determined by the specific surface area. The
sample with the highest reaction rate was chosen for further
tests at different temperatures. The CO oxidation was carried
out at 25, 40 and 50 °C oven temperature. In comparison to the
commercial reference the FSP made Hopcalite shows higher
activity and longtime stability at any temperature investigated. At
50 °C the SU-1 yields 50 % more CO conversion compared to
the reference even after one hour on stream (Figure 7).
Table 4. Textural properties and compositions of FSP derived Hopcalites
synthesized with different concentrations and a commercial reference.
a
b
c
c
tspray
[min]
SSABET
[m2/g]
dBET
dXRD
Cd
[wt.%]
Sample
[mol/l]
[nm]
[nm]
SU-1
SU-2
SU-3
SU-4
SU-5
SU-6
Carulite 300
0.05
0.10
0.14
0.19
0.27
0.325
-
130
66
45
34
24
20
-
193
176
150
137
137
133
310
5.7
6.3
7.4
8.0
8.0
8.3
3.6
5.2
4.8
4.7
4.6
5.0
5.7
-
0.46
0.40
0.29
0.27
0.31
0.20
0.12
[a] Derived from nitrogen physisorption at 77 K with multipoint BET analysis at
p/p0= 0.05–0.20.
[b] Calculated from Eq. (5).
[c] Calculated from Eq. (4).
[d] Determined by combustion method with CHNS-Analyzer.
The synthesis time could be reduced by a factor of 7 for the
production of the same amount of catalyst, only by increasing
the precursor concentration. The crystallographic analysis of the
obtained particles (Figure S7) showed neither big differences in
crystallite size nor undesired crystal phases. In contrast, the
surface area as well as the carbon content decrease with
increasing precursor concentration (Figure 6).
100
90
0.5
190
80
0.4
180
65 % (1h)
70
170
0.3
60
50
40
30
160
0.2
150
140
0.1
15 % (1h)
20
SU-1
Carulite300
10
130
0.00
0.0
0
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0
10
20
30
40
50
60
c / mol l-1
TOS / min
Figure 6. Specific surface area and carbon content in dependence of
precursor concentration.
Figure 7. Results of catalytic CO oxidation for the sample SU-1 in
comparison to the commercial reference material Carulite 300 under humid
conditions (according to 75% humidity at 20 °C) at 50 °C oven temperature.
This correlation leads to the assumption that the carbon is
mainly deposited on the particle surface. Smaller particles have
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
1.2
1.0
0.8
0.6
0.4
0.2
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
5 min on stream
20 min on stream
10 min on stream
20 min on stream
0.0
Caru300 SU-1
SU-2
(176)
SU-3
(150)
SU-4
(137)
SU-5
(137)
SU-6
(133)
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
(310)
(193)
WC / wt%
Figure 8. Results of catalytic CO oxidation for the samples SU-1 to SU-6 in comparison to the commercial reference material Carulite 300 (left) and reaction rate
of SU-1 to SU-6 in dependence of carbon content (right), under humid conditions (according to 75% humidity at 20 °C) at 50 °C oven temperature.
The T50 as well as T90 values of Hopcalites synthesized with
different precursor concentrations under dry conditions can be
found in Table S4.
stream. After 20 min, the stearate sample shows a higher
conversion and a lower deactivation behavior compared to the
2-EH sample. As the same Mn:Cu ratio is detected by EDX, the
differences in catalytic activity are due to different surface areas
and carbon contents (Figure 9). Since the stearate and 2-EH
samples have the same surface area and nearly the same
particle size, it can be concluded that the carbon content from
the precursor is responsible for the long-term stability of the
catalyst under humid conditions. Unfortunately, there was no
carbon layers or residuals detectable in the TEM images
(Figure S8).
Precursor comparison
Regarding the investigated synthesis parameter, it could be
shown that the influence of the precursor is more important than
the solvent in terms of catalytic activity. From these findings we
investigated the possibility to increase the carbon content under
synthesis conditions which lead to active particles under dry
conditions and humidity-tolerant particles. Therefore, carbon rich
precursors e.g. stearates were used. The textural properties and
composition of the obtained particles are listed in Table 5.
10 min on stream´
20
20 min on stream
Table 5. Textural properties and compositions of FSP derived Hopcalite
synthesized from different precursors.
nitrate
2-EH
stearate
15
Wc (precursor) / wt%
Wc (product)[a] / wt%
Mn:Cu[b] / mol mol-1
SSABET[c] / m2 g-1
dXRD[d] / nm
[a] Determined by combustion method with CHNS-Analyzer.
[b] Detected by EDX.
[c] Derived from nitrogen physisorption at 77 K with multipoint BET analysis at
0
55.8
0.31
1.93
160
7.9
69.2
1.16
1.94
160
6.3
0.11
1.95
114
10.9
10
5
p/p0= 0.05–0.20.
[d] Calculated from Eq. (4).
The comparison of the catalysts synthesized from three different
precursor systems under comparable FSP conditions shows an
increased carbon content in the product particles with increasing
carbon content of the precursor. This leads to the assumption
that there is a precursor-dependent combustion behavior in the
FSP whereby the carbon content in the catalyst can be adjusted.
From the catalytic investigations under humid conditions, it could
clearly be seen that the reference as well as the particles made
from nitrates show no considerable CO conversion. The other
two catalysts show nearly the same conversion after 10 min on
0
Carulite 300
nitrate
2-EH
stearate
Figure 9. Results of catalytic CO oxidation of Hopcalite nanoparticles
synthesized out of nitrates in inverse microemulsions (sample nitrate[26]) , 2-
ethylhexanoates (sample 2-EH[22]), stearates (S-1) in comparison to the
commercial reference material Carulite 300 (Carus) under humid conditions
(according to 75% humidity at 20 °C) at ambient temperature.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
Conclusions
Experimental Section
The origins of a hydrophobic carbon membrane responsible for
the humidity resistance of flame-derived Hopcalite nanoparticles
under catalytic conditions were studied by systematically varying
the synthesis parameters in the flame spray pyrolysis and the
consequences for structural, textural, compositional and catalytic
particle properties. The influence of the flame generation (OFR,
sheath gas, quartz tube) and the precursor (composition,
concentration, solvent) were investigated. Regarding the carbon
content as well as the moisture stability neither the OFR nor the
solvent have a significant influence. However, the particle size
and crystallinity can be adjusted by these parameters.
Encapsulating the FSP using a quartz tube results in higher
carbon contents but leads to Hopcalite nanoparticles with
considerably reduced catalytic activity in CO oxidation. The
nature of this carbon is soot like and amorphous and consists of
disordered and graphitic regimes, investigated by Raman and
TEM. Without a quartz tube, the productivity of the process was
increased using higher precursor concentrations. However, as a
consequence, a decrease in surface area and catalytic activity of
the catalysts was observed. Nevertheless, all samples
investigated showed an increased long-term stability under
humid CO oxidation compared with a commercial reference
Hopcalite (Carulite 300). Samples produced under optimized
conditions (SU-1) showed an outstanding performance with a
loss of only 7 % of catalytic activity from 10 to 20 min on reaction
gas stream at 50 °C while the decrease of the reference catalyst
was 50 %. The comparison of different precursors showed that
carbon rich Cu-Mn-precursors lead to higher carbon content in
the product particles indicating the precursor carboxylate ligand
to be the main source of the protective carbon membrane. At
constant surface area and Cu:Mn ratio, catalysts with higher
carbon content tend to show improved long-term stability in CO
oxidation under humid conditions as it is relevant for respiratory
applications. The carbon produced by incomplete precursor
combustion leads to a reduced water affinity but does not form a
thick layer around the particles. A further analysis of these
carbon using XPS and other surface techniques will be the next
steps to understand its effect on the particle properties.
Precursor Synthesis:
Cu(II)-2-ethylhexanoate: In a typical synthesis 30 g (0.12 mol)
copper sulfate pentahydrate (Sigma Aldrich, 99 %) as well as
40.75 g (0.245 mol) sodium-2-ethylhexanoate (Sigma Aldrich,
97 %) were dissolved in each 150 ml distilled water resulting in a
blue (Cu) and colorless (hexanoate) clear solution. Under
vigorous stirring the copper sulfate solution was added dropwise
to the sodium-2-ethylhexanoate solution. The resulting green
precipitate was vacuum filtered with a Büchner funnel, washed
with water several times and dried in a vacuum oven at 80 °C to
constant weight. Yield: quantitative.
Mn(II)-2-ethylhexanoate: In a typical synthesis 30 g (0.12 mol)
manganese acetate tertahydrate (Sigma Aldrich, >99 %) were
suspended in 74.4 g 2-ethylhexanoic acid (0.516 mol) (Sigma
Aldrich, 99 %). The mixture was refluxed at 140 °C for 7 h giving
a clear yellow solution which turns into dark orange by cooling to
room temperature. The produced acetic acid was removed
means rotary evaporator to purify the manganese(II)-2-
ethylhexanoate solution. The manganese content was analyzed
by TGA to be 9 %.
Cu/Mn(II)-stearate: In a typical synthesis 47.5 g (0.159 mol)
stearic acid (Sigma Aldrich, >95 %) were melted at 90 °C in
300 ml distilled water. To this solution, 635 ml (0.159 mol) of a
0.25 M sodium hydroxide solution were added dropwise over 1 h.
To the resulting clear solution, 500 ml of 0.2 M (0.1 mol) copper
chloride dehydrate (Sigma Aldrich, 99 %) were added dropwise
over 1 h. The resulting blue-green precipitate was vacuum
filtrated with a Büchner funnel washed with 1 l hot water and
dried in a vacuum oven at 80 °C to constant weight. Yield:
quantitative
For the synthesis of manganese(II)-stearate manganese
chloride dihydrate (Sigma Aldrich, 99 %) was used instead of the
copper salt resulting in a voluminous white powder. Yield:
quantitative.
Cu/Mn(II)-2-ethylhexanoate mixed precursors: In general self-
made or commercial copper(II)-2-ethylhexanoate (Sigma Aldrich
98%) and self-made manganese(II)-2 ethylhexanoate were
dissolved in various solvents resulting in overall metal
concentrations of 0.025 M to 0.325 M with Mn:Cu molar ratios
ranging from 2.0 to 2.4. Details on the samples produced from 2-
ethylhexanoates are given in Table 6.
From these findings we deduce the correct phase composition to
be the most important characteristic followed by the specific
surface area. The carbon content critically determines the
catalytic activity in CO oxidation under humid conditions, but the
nature of the selective carbon membrane is significantly affected
by the precursor. An unspecific thick coating despite higher
carbon content, does not improve the catalyst performance. The
laboratory experiments described are a valuable basis for
upscaling and continuous industrial production of Hopcalite
catalysts with significantly increased humidity tolerance as
compared to Hopcalites commercially available today. Such
humidity tolerant second generation Hopcalites (Hopcalites-2G)
are valuable system components for innovative respiratory and
indoor air filters.
Cu/Mn(II)-stearate mixed precursors: Copper(II) and
manganese(II)-stearate were dissolved in
mixture of 2-ethylhexanoic acid:methanol:xylole resulting in
overall metal concentrations of 0.05 M with a Mn:Cu molar ratio
of 2.0 (Table 6).
a 2:2:1 (V:V:V)
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
Table 6. Sample codes and synthesis conditions.
Mn:Cu
Cmetal
[mol/l]
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
SFc
[ml/min]
5.0
O2Dd
[slm]
4.13
3.30
2.48
1.65
4.13
3.30
2.48
1.65
4.13
3.30
2.48
N2De
O2Sf
N2Sg
[slm]
0
Sample
solventb
OFR
[mol:mol]
2.0
[slm]
[slm]
OFR-1co
OFR-2co
OFR-3co
OFR-4co
OFR-5co
OFR-6co
OFR-7co
OFR-8co
OFR-9co
OFR-10co
OFR-11co
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
1.0
0.8
0.6
0.4
1.0
0.8
0.6
0.4
1.0
0.8
0.6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2.0
5.0
0
2.0
5.0
0
2.0
5.0
0
2.0
5.0
10
10
10
10
20
20
20
2.0
5.0
2.0
5.0
2.0
5.0
2.0
5.0
2.0
5.0
2.0
5.0
OFR-12co
M:A=9
2.0
0.05
5.0
0.4
1.65
0
0
20
Etco
THFco
H co
E
2.0
2.0
2.0
2.0
2.0
2.0
-
0.05
0.05
0.05
0.05
0.05
0.05
0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
2.5
5.0
5.0
5.0
5.0
2.5
2.5
2.5
2.5
2.5
2.5
1.0
1.0
1.0
1.0
1.0
1.0
0.81
1.0
1.0
1.0
1.1
1.0
0.9
0.8
3.6
3.6
3.6
3.6
3.6
3.6
5.76
7.60
8.08
9.64
4.13
4.12
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
T
0
0
H
0
0
Xco
X
0
0
MAco
MFAco
Q-1
M:A=9
M:FA=99
X
0
0
0
0
5
7.84
9.64
9.64
4.83
10.60
9.64
8.68
7.71
0
Q-2co
Q-3ih
Q-4ih
Q-5ih
Q-6ih
Q-7ih
Q-8ih
SU-1ih
SU-2ih
SU-3ih
SU-4ih
SU-5ih
SU-6ih
X
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.4
2.4
2.4
2.4
2.4
2.4
0.025
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.1
0
5
X
0
5
X
0
5
X
0
8.77
8.77
8.77
8.77
0
X
0
X
0
X
0
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
M:A=9
EH:M:X
=2:2:1
7.5
7.5
7.5
7.5
7.5
7.5
0
0
0.14
0.19
0.27
0.325
0
0
0
0
0
0
0
0
S-1ih
2.0
0.05
2.5
2.4
8.0
0
0
0
co commercial copper-2-ethylhexanoate was used.
ih In house made copper-2-ethylhexanoate was used
a OFR = oxygen to fuel ratio; Et, THF, H, X, MA, MFA = different solvents; Q = quartz tube adaption; SU = precursor concentration (scale up); S = stearate
b M = methanol, A = acetic acid, E = ethanol, T = THF, H = n-hexane, X = xylene, FA = formic acid, EH = 2-ethylhexanoic acid
c SF = solvent flow; d O2D = oxygen dispersion gas flow; e N2D = nitrogen dispersion gas flow; f O2S = oxygen sheath gas flow; g N2S = nitrogen sheath gas flow
Flame Spray Pyrolysis: The synthesis of Cu-Mn-oxide
nanoparticles was carried out using a flame spray reactor
described in detail elsewhere.[18,20] In short, the precursor
solutions were fed through a commercial nozzle (NPS10, Tethis)
with liquid feed rates of 2.5 and 5.0 mL min−1, ignited by a
surrounding supporting-flame consisting of 1.50 slm CH4 and
3.00 slm O2. The liquid precursor solution was delivered by a
micro annular gear pump (mzr-2905, HNP Mikrosysteme GmbH)
and dispersed by 1.65 to 9.64 slm O2 or 5.00 to 8.77 slm N2 with
a dispersion gas pressure drop accounting to 2.0 bar. For some
synthesis, additional O2 or N2 sheath gas was supplied through
a sintered metal plate ring surrounding the flame. In addition, the
samples named with Q were synthesized with a quartz tube
(l: 400 mm, AD: 50 mm, ID: 46 mm) enclosing the flame to avoid
complete precursor combustion. All gas flows were controlled by
thermal mass flow controllers (Bronkhorst). The product particles
were collected by a binder-free glass fiber filter (GF/A Whatman)
with the help of a rotary vane pump (Vacuubrand RZ 9). All
gases had a purity of 99.999 % and were purchased from Air
Liquide. The specific synthesis conditions for each sample are
listed in table S1.
Characterization methods: X-ray powder diffraction patterns
were collected with a PANalytical X’Pert Pro diffractometer in
reflection mode using Cu Kα1 (1.5406 Å) radiation. A step size
of 0.013° 2θ was applied. The obtained diffractograms were
compared with reference data from the ICSD powder diffraction
data base, which are referenced by the respective powder
diffraction file (PDF) numbers. Applying the SCHERRER-
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
equation[24] (eq. 4), the crystallite diameters (dXRD
diffraction peaks of Cu1.5Mn1.5O4 in the range of 10 to 80° 2θ
)
for all
stream of 86 mL min-1 and switching to O2 (16.7 mL min-1) after
10 min, holding the temperature for additional 50 min. CO and
CO2 signals were recorded in the exhaust gas during the
activation procedure. For testing under humid conditions, the
gas stream was saturated with water vapor by flowing through a
reservoir containing deionized water at 15 °C. A humidity of
75 % RH at 20 °C was simulated in this way.
In order to guarantee sufficient gas permeability through the
catalyst bed, the as received voluminous Hopcalite powders
were dispersed in water and densified by centrifugation.
Subsequent drying, grinding and sieving, gave grain sizes of
125–250 μm that were used for the catalytic tests.
were calculated.
퐾∙휆
푑푋푅퐷
=
(4)
a
(
)
퐹푊퐻푀∙cos 휃
Nitrogen physisorption was conducted at -196 °C with
Quantachrome NOVA 4000 after activating the samples at
150 °C for 12 h in dynamic vacuum. The specific surface area
(SSABET) was determined by applying the multi-point BET-
method in the range of p/p0 = 0.05–0.20. The BET diameters
(dBET) were calculated with equation (5) from the measured
specific surface area (SSA) and the density of Cu1.5Mn1.5O4
(ρ=5.44 g/cm3)[25]
.
Acknowledgements
6
푑퐵퐸푇
=
(5)
a
푆푆퐴∙ρ
The authors would like to thank Dr. Danny Haubold for TEM
measurements and Dr. Stefan Klosz for Raman investigations.
Water physisorption isotherms were recorded with
Quantachrome HYDROSORB 1000 at 25 °C after activating the
samples in the abovementioned way.
Keywords: CO oxidation, flame spray pyrolysis, Hopcalite,
humidity tolerance, respiratory filter
Scanning electron micrographs were obtained using a SU8020
(HITACHI) operating at 3 kV. Prior to the measurements the
samples were fixed on an adhesive carbon pad and sputtered
with gold to provide the necessary electrical conductivity.
EDX measurements were obtained at a X-MAX Silicon Drift
Detector (OXFORD INSTRUMENTS) while the acceleration
voltage was set to 20 kV.
Transmission electron micrographs were obtained using a
TECNAI F30 (PHILIPS) operating at 300 kV using a field emitter
as electron source. Therefore, the samples were dispersed in
acetone, dropped on a copper grid (S160, plano) and dried
overnight prior to use.
[1]
[2]
[3]
H. Huang, Y. Xu, Q. Feng, D. Y. C. Leung, Catal. Sci. Technol. 2015,
5, 2649–2669.
D. G. Hannan, C. R. E. Merkle, Am. Ind. Hyg. Assoc. J. 1968, 29,
136–139.
E. L. Pulster, J. V. Hillman, in Hamilt. Hardy’s Ind. Toxicol., John
Wiley & Sons, Inc., Hoboken, New Jersey, 2015, pp. 309–316.
M. Haruta, J. Catal. 1989, 115, 301–309.
S. Royer, D. Duprez, ChemCatChem 2011, 3, 24–65.
B. B. Arthur Lamb, W. C. Bray, J. C. W Frazer, J. Ind. Eng. Chem.
1920, 12, 213.
[4]
[5]
[6]
[7]
M. Krämer, T. Schmidt, K. Stöwe, W. F. Maier, Appl. Catal. A Gen.
2006, 302, 257–263.
[8]
G. J. Hutchings, A. A. Mirzaei, R. W. Joyner, M. Rafiq, H. Siddiqui,
S. H. Taylor, Catal. Letters 1996, 42, 21–24.
[9]
E. C. Njagi, C.-H. Chen, H. Genuino, H. Galindo, H. Huang, S. L.
Suib, Appl. Catal. B Environ. 2010, 99, 103–110.
T. J. Clarke, T. E. Davies, S. A. Kondrat, S. H. Taylor, Appl. Catal. B
Environ. 2015, 165, 222–231.
Z.-R. Tang, S. A. Kondrat, C. Dickinson, J. K. Bartley, A. F. Carley,
S. H. Taylor, T. E. Davies, M. Allix, M. J. Rosseinsky, J. B. Claridge,
et al., Catal. Sci. Technol. 2011, 1, 740.
C, H,
N and S contents were determined via a triple
measurement, using a vario MICRO cube (Elementar).
[10]
[11]
TG and DTG measurements were performed using an STA
409C/CD (Netzsch). The samples were heated up from room
temperature to 1000 °C with a heating rate of 5 K/min in an
Al2O3 crucible. TG measurements were carried out in synthetic
air, DTG measurements were carried out in oxygen. TG-MS
measurements were performed at a TG/STA-QMS 403 C Aëolos
(Netzsch). Identification of CO2 was done by analyzing the signal
at m/z=44.
[12]
[13]
C. Chen, E. Njagi, S. Sun, H. Genuino, Chem. 2010.
R. K. Kunkalekar, A. V. Salker, React. Kinet. Mech. Catal. 2012, 108,
173–182.
[14]
[15]
Y. Liu, Y. Guo, H. Peng, X. Xu, Y. Wu, C. Peng, N. Zhang, X. Wang,
Appl. Catal. A Gen. 2016, 525, 204–214.
W. Y. Hernández, M. A. Centeno, S. Ivanova, P. Eloy, E. M.
Gaigneaux, J. A. Odriozola, Appl. Catal. B Environ. 2012, 123–124,
27–35.
FT-IR spectra were obtained using a Bruker Vertex 70 (Bruker
Optik GmbH) with 32 scans between 650-4000 cm-1 and a scan
rate of 2 cm-1.
[16]
G. J. Hutchings, A. A. Mirzaei, R. W. Joyner, M. R. H. Siddiqui, S. H.
Taylor, Appl. Catal. A Gen. 1998, 166, 143–152.
W. Y. Teoh, R. Amal, L. Mädler, Nanoscale 2010, 2, 1324–1347.
L. Mädler, H. K. Kammler, R. Mueller, S. E. Pratsinis, J. Aerosol Sci.
2002, 33, 369–389.
[17]
[18]
Raman spectra were obtained using a Raman spectrometer RM-
2000 (Renishaw). The wavelength of the used laser was 532 nm.
As described in earlier work [22,26], the catalytic tests were
performed in a fixed bed tubular reactor (din = 6 mm) equipped
with two NDIR-sensors smartModulPREMIUM (Pewatron) for the
simultaneous determination of CO and CO2 concentrations in
the exhaust gas stream. The temperature was monitored inside
the catalyst bed by a thermocouple inserted in a glass capillary.
In a typical setup, 100 mg of catalyst were exposed to a gas
mixture consisting of 0.72 vol.% CO, 71.43 vol.% N2, and 27.85
vol.% O2 (59.6 mL min-1) for at least 20 min at each temperature.
Catalyst activation was conducted by heating to 300 °C in a N2
[19]
[20]
R. Koirala, S. E. Pratsinis, A. Baiker, Chem. Soc. Rev. 2016, 45,
3053–3068.
H. K. Kammler, L. Mädler, S. E. Pratsinis, Chem. Eng. Technol.
2001, 24, 583–596.
[21]
[22]
R. Strobel, S. E. Pratsinis, J. Mater. Chem. 2007, 17, 4743–4756.
T. Biemelt, K. Wegner, J. Teichert, M. R. Lohe, J. Martin, J. Grothe,
S. Kaskel, Appl. Catal. B Environ. 2016, 184, 208–215.
F. O. Ernst, R. Büchel, R. Strobel, S. E. Pratsinis, Chem. Mater.
2008, 20, 2117–2123.
[23]
[24]
[25]
[26]
A. R. Lang, Acta Metall. 1956, 4, 102.
G. Blasse, J. Phys. Chem. Solids 1966, 27, 383–389.
T. Biemelt, K. Wegner, J. Teichert, S. Kaskel, Chem. Commun.
(Camb). 2015, 51, 5872–5.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800639
ChemCatChem
FULL PAPER
Hopcalites-2G: Humidity tolerant
copper manganese oxide catalysts
were produced by a scalable one
step flame spray pyrolysis process
(FSP) and their catalytic activity for
carbon monoxide oxidation under dry
and humid conditions was evaluated
Karl Wegner, Maximilian Medicus, Elke
Schade, Julia Grothe*, Stefan Kaskel
Page No. – Page No.
Tailoring Catalytic Properties of
Copper
Manganese
Oxide
systematically.
Generating
a
Nanoparticles (Hopcalites-2G) via
Flame Spray Pyrolysis
protecting carbon membrane leads
to nanoparticles with outstanding
catalytic performance under humid
conditions
compared
with
a
commercial reference.
This article is protected by copyright. All rights reserved.