Low-temperature formaldehyde oxidation over manganese oxide catalysts: Potassium mediated lattice oxygen mobility

Article abstract of DOI:10.1016/j.mcat.2020.111204

Manganese oxide catalysts with self-modulating K⁺ content and tunable concentration of lattice oxygen and Mn⁴⁺ were synthesized and investigated for HCHO oxidation. The preparation method affects the physicochemical properties and catalytic activity of the catalysts. Herein, the role of K⁺ in enhancing the lattice oxygen mobility of manganese oxide catalysts for enhanced formaldehyde (HCHO) is presented. The presence of K⁺ enhances the redox properties of Mn and promotes catalytic activity by enhancing the mobility of the lattice oxygen and sustaining the availability of surface active oxygen to sustain the reaction. Catalytic activity was observed to improve with increasing K⁺ content and the surface concentration of lattice oxygen and Mn⁴⁺. A drastic reduction in catalytic activity was observed in the acid-treated samples, with low K⁺ concentration. Characterization results indicate that the presence of K⁺ enhances activity and mobility of the lattice oxygen by the weakening the Mn-O bond in manganese oxide and promotes the redox properties of the catalyst. The absence of K⁺ impacted the mobility of the lattice oxygen and the ability of the catalyst to supplement the consumed oxygen species, resulting into reduced catalytic activity and deactivation in the room-temperature (30 °C) activity and stability test.

Full text of DOI:10.1016/j.mcat.2020.111204

Molecular Catalysis 497 (2020) 111204
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.journals.elsevier.com/molecular-catalysis
Low-temperature formaldehyde oxidation over manganese oxide catalysts:
Potassium mediated lattice oxygen mobility
,
,
Abubakar Yusufa, Yong Sunb, Colin Snapec, Jun Hea b,*, Chengjun Wangd **, Yong Rene,
Hongpeng Jiaf
a International Doctoral Innovation Center, University of Nottingham Ningbo China, Ningbo, China
b Department of Chemical and Environmental Engineering, University of Nottingham Ningbo China, Ningbo, China
c Faculty of Engineering, University of Nottingham, University Park, Nottingham, United Kingdom
d College of Resources and Environmental Science, South-Central University for Nationalities, Wuhan, China
e Department of Mechanical, Materials and Manufacturing Engineering, University of Nottingham Ningbo China, Ningbo, China
f Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, China
A R T I C L E I N F O
A B S T R A C T
Manganese oxide catalysts with self-modulating K+ content and tunable concentration of lattice oxygen and
Mn4+ were synthesized and investigated for HCHO oxidation. The preparation method affects the physico-
chemical properties and catalytic activity of the catalysts. Herein, the role of K+ in enhancing the lattice oxygen
mobility of manganese oxide catalysts for enhanced formaldehyde (HCHO) is presented. The presence of K+
enhances the redox properties of Mn and promotes catalytic activity by enhancing the mobility of the lattice
oxygen and sustaining the availability of surface active oxygen to sustain the reaction. Catalytic activity was
observed to improve with increasing K+ content and the surface concentration of lattice oxygen and Mn4+. A
drastic reduction in catalytic activity was observed in the acid-treated samples, with low K+ concentration.
Characterization results indicate that the presence of K+ enhances activity and mobility of the lattice oxygen by
the weakening the Mn-O bond in manganese oxide and promotes the redox properties of the catalyst. The
absence of K+ impacted the mobility of the lattice oxygen and the ability of the catalyst to supplement the
consumed oxygen species, resulting into reduced catalytic activity and deactivation in the room-temperature (30
◦C) activity and stability test.
Keywords:
formaldehyde
manganese oxide catalyst
potassium
catalytic oxidation
lattice oxygen
Introduction
temperature activity for HCHO oxidation. Alkali metals doping, mainly
Na+ and K+, has been demonstrated to be a very effective strategy [5–8].
The removal of formaldehyde (HCHO) – a known and confirmed
carcinogenic substance, from the indoor environment is pertinent to
improving indoor air quality and comfort. Satisfactory pieces of evi-
dence have revealed that exposure to HCHO over a long period, can lead
to nasopharyngeal cancer, sinonasal cancer, and leukaemia [1]. Some of
its major indoor sources include wooden works, furniture and con-
struction materials; others may include indoor combustion processes,
such as cooking, smoking, and heating [2,3]. Amongst the HCHO
removal techniques, catalytic oxidation seems more promising, as the
complete conversion of HCHO into CO2 and H2O can be achieved [4].
Catalysts containing noble metals are promising in this respect, and
various techniques devised have enhanced their low to room
Bai et al.[5] observed a promotional effect of K+ addition on the
activity of Ag/Co3O4 catalyst for HCHO by promoting the generation of
surface OH-. Zhang et al.[8] observed a phenomenal improvement in the
catalytic activity of Pt/TiO2, after its treatment with Na. Total oxidation
of HCHO was achieved at room temperature over the Na modified
Pt/TiO2, in contrast to 150 ◦C for the untreated catalyst. In another
study, Na+ was observed to promote the stabilization and dispersion of
Pd on TiO2 support, and enhanced its activity by facilitating the acti-
vation of chemisorbed oxygen and OH groups to achieve room tem-
perature oxidation of HCHO (compare to 120 ◦C for unmodified
Pd/TiO2) [7]. Similar promotional effects of Na+ were observed by Nie
at al.[6] for HCHO oxidation on Pt supported TiO2 catalysts. However,
* Correspondence author at: International Doctoral Innovation Center, University of Nottingham Ningbo China, Ningbo, China.
** Correspondence author.
E-mail addresses: jun.he@nottingham.edu.cn (J. He), cjwang@scuec.edu.cn (C. Wang).
https://doi.org/10.1016/j.mcat.2020.111204
Received 28 May 2020; Received in revised form 13 August 2020; Accepted 4 September 2020
Available online 20 September 2020
2468-8231/© 2020 Elsevier B.V. All rights reserved.

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
there is a growing interest and focus on transition metals for catalysis
due to their relative abundance and cost-effectiveness [9,10].
0.8) were treated in 100 ml 0.1 M HNO3 solution for 6 hrs under
continuous stirring. The acid-treated samples were then rinsed severally
with deionized water until neutral and then dried at 105 ◦C and denoted
A0.1 Mx. For comparison, some acid-treated samples were subjected to
further calcination at 300 ◦C for 3 hrs and denoted as A0.1 MxH.
Manganese oxide based catalysts are amongst the most promising
transition metal alternatives for low-temperature HCHO oxidation [4].
The effectiveness of metal oxides for catalytic oxidation and other ap-
plications is affected by their propensity to cycle between two valence
states and the mobility of their lattice oxygen [11]. The efficiency of
VOCs oxidation over MnOx catalysts is strongly influenced by the redox
properties of Mn, and the reactivity and mobility of the lattice oxygen
[12], with Mn higher oxidation state (Mn4+) shown to be more active for
the oxidation of o-xylene [13], toluene [14], HCHO [10,15], CO [16],
and methane [17]. Enhancing the redox cycle of Mn (Mn4+↔ Mn3+) by
facilitating the activation of molecular oxygen and enhancing the
mobility of lattice oxygen was shown to be an effective strategy for
improving catalytic oxidation of HCHO [10,18]. Most of the investigated
manganese oxide based catalysts, such as layered birnessite-type MnO2
and octahedral molecular sieves (OMS-2), contain K+. It has been re-
ported that K+ facilitates the dissociation of molecular oxygen and
desorption of water molecules, through an electron transfer mechanism
from potassium to oxygen [19], and also enhances the generation sur-
face active oxygen species such through the creation of oxygen and/or
manganese vacancies [20–22]. It was previously demonstrated that K+
facilitates the migration of the lattice oxygen of 3D CO3O4 for enhanced
catalytic oxidation of HCHO over Ag/CO3O4 [5], it is therefore inter-
esting to understand the effect of K+ on the mobility and activity of the
lattice oxygen of manganese oxide for HCHO oxidation, since K+ is
structurally present in most of the manganese oxide based catalyst.
To investigate the role of K+ in the catalytic oxidation of HCHO over
manganese oxide catalysts, a two way strategy was deployed by syn-
thesizing manganese oxide catalysts with self-modulating and varying
K+ concentration and subjecting the synthesized catalysts to acid-
treatment to remove the K+ from the catalyst’s framework. By varying
the molar ratio of KMnO4 to Mn(NO3)2 in the reaction mixture, not only
is the K+ concentration modulated but also the morphology, lattice
oxygen and Mn4+ content. Increasing the KMnO4 content in the reaction
mixture led to an increase in the K+ content, lattice oxygen and surface
concertation of Mn4+. On the other hand, the acid-treatment of the
pristine catalysts led to a drastic reduction in the K+ content of the
catalysts and the creation of oxygen vacancies. Characterization and
activity evaluation results (in the presence and absence of molecular
oxygen) showed that the presence of K+ enhances the activity and
mobility of the lattice oxygen by weakening the interaction between Mn
and O, and also facilitates the replenishment of consumed oxygen,
thereby enhancing catalytic activity and stability for HCHO oxidation.
The surface reaction mechanisms of both the pristine and acid-treated
catalysts were investigated via in-situ diffuse reflectance infrared
Fourier transform spectroscopy (DRIFTS) measurements.
Characterization
X-ray diffraction patterns of the synthesized catalysts were analysed
on D8 Advanced (Bruker, Germany) instrument operating at 40 mA, 40
mV. Raman spectra of the catalysts were collected on Renishaw inVia
Raman Microscope equipped with an excitation source operating at 532
nm. The spectra were collected between 60 and 1500 cm-1. Hitachi S-
4800 Field Emission Scanning Electron Microscope (FESEM) was used to
observe the morphological structures of the catalysts while elemental
composition and mapping was conducted using energy-dispersive X-ray
spectroscopy (EDS). Prior to the FESEM analysis, the samples were
sputtered with platinum on Hitachi E-1045. Transmission Electron Mi-
croscopy (TEM) and High-Resolution TEM (HRTEM) images were ac-
quired on JEM-2100 200Kv electron microscope (JEOL Ltd., Japan).
Micrometrics ASAP 2020 analyzer (Surface Area and Porosity Ana-
lyser) was used to evaluate the textural properties of the catalysts. The
BET surface area of the catalysts was calculated from the N2 adsorption-
desorption isotherms using the multipoint method, while the pore size
distribution was calculated from the BHJ desorption isotherm. Prior to
the analysis, the samples were outgassed at 200 ◦C for 3 hrs. The
chemical states, binding energy and surface elemental composition of
the catalysts were analyzed using x-ray photoelectron spectroscopy
(XPS) technique. XPS measurements were conducted on Kratos Axis
Ultra DLD, with Al Kα x-ray source with characteristic excitation energy
of 1,486.7 eV. The binding energies of all elements were calibrated
based on the adventitious carbon at 284.8 eV. Inductively Coupled
Plasma Mass Spectrometry (NexION 300, PerkinElmer) was used to
determine the elemental composition of the catalysts.
Hydrogen temperature programmed reduction (H2-TPR) was con-
ducted on Micrometrics Autochem II (chemisorption analyzer). About
50 mg samples were weighed into a quartz tube and outgassed at 150 ◦C,
under a 50 ml min-1 flow of Ar for 1 hr, prior to the analysis. The pre-
treated samples were cooled down to about 30 ◦C. The gas flow was
then switched to 8% H2/Ar mixture (50 ml min-1). The temperature was
slowly ramped at a rate of 5 ◦C min-1 to 600 ◦C, and the H2 consumption
was monitored by a calibrated TCD detector. The surface reaction and
intermediate species of HCHO oxidation were examined via in situ
diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS)
analysis on Thermo fisher, Nicolet 6700 fitted with a reaction cell
(Harrick, Praying Mantis). Purified air was used to inject HCHO (~ 75
ppm) into the cell at a flow of 30 ml min-1, and the spectra recorded
(resolution 4 cm-1 and accumulation of 32 scans). Prior to the DRIFTS
analysis, samples were treated at 200 ◦C in a 20 ml min-1 flow of Ar for
30 minutes.
Experimental
Catalyst synthesis
Catalytic activity test
In a given preparation of MnOx based catalysts, calculated amounts
of KMnO4 (Sinopharm) and Mn(NO3)2 (Sigma Aldrich) equivalent to
0.2, 0.3 and 0.5 molar ratio were dissolved in 100 ml deionized water in
a beaker. KMnO4 solution was gradually added to the Mn(NO3)2 at 50 ◦C
while the pH was gradually adjusted using 2 M KOH until a pH of 10 was
under continuous stirring. The obtained brown precipitates were aged in
the supernatant solution at the same temperature for 2 hrs. The pre-
cipitates were then filtered, rinsed with deionized water and ethanol
until neutral. The samples were then dried in an oven at 105 ◦C for 12
hrs. The dried samples were further calcined at 300 ◦C in air for 3 hrs.
The obtained samples were denoted Mx, where x is the molar ratio of
KMnO4 to Mn(NO3)2 in the precursor solution. For comparison, M0 was
synthesized using the same method and conditions in the absence of
KMnO4. Subsequently, the obtained samples (x = 0.2, 0.3, 0.5, 0.6 and
The catalytic activities of the catalysts were investigated under the
same conditions in a dynamic mode. In a typical experiment, 100 mg of
40-60 mesh catalysts were loaded in a fixed bed quartz reactor with an
internal diameter of 6 mm. About 170 ppm of HCHO gas was generated
by passing air over paraformaldehyde (97% Alfa Aesar) maintained at
30 ◦C, in a U-shaped quartz tube. The relative humidity of the inlet
stream was maintained at 45%, by bubbling air over deionized water.
The concentration of HCHO and relative humidity of the feed stream
was maintained by adjusting the flowrate of the make-up gas. The total
flowrate of the inlet stream was maintained at 100 ml min-1 with the aid
of mass flow controllers, which corresponds to a gas hourly space ve-
locity (GHSV) of 60,000 ml⋅g-1 h-1. The concentration of CO2 and HCHO
from the reactor were monitored online using GC (Agilent 7890B) fitted
2

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
with an Agilent CP-Sil 5 CB column (0.32 × 50 m), Polyarc reactor
(Activated Research Company (ARC), US) and an FID detector. Both
HCHO and CO2 are first converted into CH4 with the aid of the Polyarc
reactor before being detected by the FID. Polyarc reactor, which is a
universal carbon detector with the capacity to convert hydrocarbons
into methane through a 2 step post-column oxidation-methanation
process (HCs → CO2 → CH4) [23]. CO2 was the only detected product of
the oxidation reaction. Prior to the reaction, the reactor was calibrated
by standard gases of HCHO and CO2. The room temperature activity and
stability of the catalysts (M0.5, A0.1 M0.5 and M0) were evaluated in a
continuous flow mode for 72 hrs at a GHSV of 60,000 ml⋅g-1 h-1 at 30 ◦C,
and HCHO concentration of ~10 ppm.
to amorphous in nature. The peaks at 17.9, 28.6, 31.2, 36.4, 37.9, 64.6,
and 67.4◦ are indexed to tetragonal hausmanite structured Mn3O4 (PDF
#24-0734), while the peaks at 36.7, 39.0, 47.5, and 66.2◦ are assigned to
MnO2 [13]. The weak and broad peaks at 36.7, 42.5, and 65.8◦ are
indexed to hexagonal birnessite type MnO2 (JCPDS #18-0802) [21]. M0
is predominantly Mn3O4. Weak features of Mn3O4, in addition to those
of birnessite MnO2, are observed in M0.2. Only the weak peaks attrib-
utable to layered birnessite MnO2 are present in M0.3 and M0.5. In the
presence of permanganate ions, Mn2+ is favourably oxidized to Mn3+
and Mn4+ leading to the formation of birnessite MnO2 [24]. The
observed peak broadening in the acid-treated samples could be attrib-
uted to partial crystal and layer distortion as a result of K+ extraction
[25,26]. Due to the amorphous nature of the catalysts, Raman spec-
troscopy was conducted to further understand their crystal structures
and the effects of K+.
Since CO2 was the only detected product, the conversion of HCHO
(%) over the catalysts was calculated as thus:
CO2
HCHO
Conversion (%) =
× 100
(1)
The Raman spectra of the catalysts are presented in Fig. 2. As seen in
Fig. 2a, the peaks at 650, 480, 393, 304 and 266 cm-1 are attributed to
Mn3O4 [27,28]. The Raman peaks located at 500-510 cm-1
(ν1), 569-582
where CO2 and HCHO stand for the concentration (ppm) of carbon di-
oxide in the outlet stream and that of HCHO in the inlet stream,
respectively. The conversion efficiency was monitored between the
temperature ranges of 30 – 120 ◦C.
cm1 (ν2) and 626-634 cm-1
(ν3) are characteristics of layered birnessite
MnO2 [29]. They are respectively assigned to Mn-O-Mn stretching (ν1),
Mn-O stretching (ν2) of the basal plane of MnO6 and the symmetric
stretching of MnO6 (ν3) [21,29]. A significant weakening in the Mn3O4
features (650 cm-1) was observed while the peaks corresponding to
birnessite MnO2 (582 and 626 cm-1) became more prominent in M0.2, in
agreement with the XRD results. Birnessite MnO2 is mainly composed of
layers of edge-sharing [MnO6] octahedra with K+ and water molecules
at the interlayer spaces to compensate for the electric charge of the
layers [30,31]. Only the three peaks characteristics of birnessite-type
layered MnO2 were observed in M0.3 and M0.5 (Fig. 2a) and in M0.6
and M0.8 (Fig. S1), in agreement with the XRD results. The relative
The specific rate was calculated as follows [22]:
ꢀ
)
[HCHO] × f ×
η
-1
m ol m -2m in
-5
Specific rate
μ
=
× 10
(2)
22.4 × SBET × m
where [HCHO] represents the inlet concentration of HCHO (ppm), f is
the total flowrate (ml. min-1), m is the mass of catalyst (g),
η is the
conversion (%) and SBET stands for the measured BET surface area of the
catalysts.
Lattice oxygen test was conducted on M0.5 and A0.1 M0.5 samples
according to the modified procedure reported elsewhere [15]. Prior to
the test, the catalysts were treated in a flow of pure air (50 ml min-1) for
30 min. at 150 ◦C, the treated catalysts were then purged with pure N2
(50 ml min-1) for 30 min. at the same temperature, to remove molecular
oxygen and weakly adsorbed species. A flow of N2 (99.99%) containing
~ 170 ppm of HCHO (GHSV 60,000 ml⋅g-1 h-1) was passed over the
catalysts at 60 ◦C and the reaction was monitored online for 600 min..
The conversion of HCHO was computed according to Eq. 1, while the
evolution of CO2 and HCHO were monitored using GC-Polyarc reactor.
Results and Discussion
Phase structure and morphology
The XRD patterns of the investigated catalysts are depicted in Fig. 1.
The weak peaks of the catalysts showed that they are poorly crystalline
Fig. 1. XRD patterns of as pristine and acid-treated catalysts.
Fig. 2. Raman spectra of (a) synthesized catalysts, (b) acid-treated catalysts.
3

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
increase in the intensity of ratio of ν2 to ν3 (Fig. 2a) could be ascribed to
the increase in K+ content in the interlayer structure of the birnessite
MnO2 [32], in agreement with the elemental analysis (Table 1). On the
other hand, a weakening in the intensity ratio of ν2 to ν3 was observed
after the acid treatment (Fig. 2b). This is due to the polarizability vari-
ations between the two vibrational modes attributable to the removal of
K+ ions from the interlayer structure of birnessite [25,26,32].
shown by the ICP-MS results, the concentration of K increases with
increasing amount of KMnO4 in the precursor solutions. Increasing the
molar ratio beyond 0.5 (0.6 and 0.8), does not result in any increase in
the K+ content of M0.6 and M0.8, as shown in Table S1, likely due to the
presence of excess amount of KMnO4 in the reaction mixtures (Fig. S4
and Supplementary Note 1). As such the optimum K+ was achieved in
M0.5. The acid treatment succeeded in drastically reducing the content of
Morphological evolution due to variation in the KMnO4 concentra-
tion in the reaction mixture were observed, as shown in Fig. 3. In the
absence of KMnO4, M0 is composed of somewhat circular nanostructures
with diameter ranging from 83 to149 nm and thickness from 25 to 40
nm. After the introduction of KMnO4 (M0.2), the morphology completely
transformed into an open flower-like structure composed of long inter-
weaved and packed nanorods, with an average thickness of 29 nm,
which further compacted into loose and irregular spheres, composed of
nanosheets with an average size of 25 nm (M0.3). Interestingly, at a
molar ratio of 0.5 (M0.5), there was a complete morphological trans-
formation into dense and aggregated nanospheres, which did not change
after the acid treatment (A0.1 M0.5). No further changes in morphology
was observed when the molar ratio was increased to 0.6 (M0.6) and 0.8
(M0.8), as shown in Fig. S2. However, a slight reduction in the size of the
nanospheres was observed after the acid treatment, as shown in the TEM
images (Fig. 4a and c). This could be attributed to the abstraction of K+
from the interlayers of the birnessite [26], and the possible distortion of
the layered structure during the acid treatment. The HRTEM images of
M0.5 (Fig. 4b) and A0.1 M0.5 (Fig. 4d) showed both the pristine and
acid-treated catalysts have exposed (001) and (100) lattice planes,
which correspond to ~ 0.7 and 0.24 nm interplanar spacing of birnessite
MnO2, respectively.
K+ in the treated samples (A0.1 Mx), in agreement with the Raman re-
sults. As for the elemental mapping of M0.5 and A0.1 M0.5 (see Fig. S5
and S6), both Mn and O are well dispersed on both samples, while the
intensity and dispersion of K significantly reduced after the acid treat-
ment, further confirming the leaching of K+ from the acid-treated
samples. The trend in the K+ content of the catalysts is consistent with
the EDS and XPS results as presented in Table 1 and Fig. S7 (XPS K 2p
spectrum).
The XPS spectra of Mn 2p, O 1s and Mn 3s of the analyzed catalysts
are presented in Fig. 5a, 5b, and 5c, respectively. The Mn 2p and 3 s high
resolution spectra may provide insights about the chemical environment
of Mn. The deconvolution of the Mn 2p3/2 peak indicated the presence of
two overlapped peaks with binding energies within the range of 642 –
642.3 eV and 642.9 – 643.6 eV corresponding to Mn4+and Mn3+
respectively [33]. No peaks related to Mn2+ were observed. The surface
content of Mn3+ and Mn4+ were estimated based on their respective
peak areas in the deconvoluted Mn 2p3/2 peak and presented in Table 2.
Higher content of Mn4+ may suggest higher oxidation, which increases
with increasing K+ concentration. The Mn 3 s doublet splitting (ΔE,
Fig. 5c) can be used to calculate the average oxidation state (AOS) of the
catalysts based on the equation AOS = 8.956 – (1.126*ΔE) [34]. The
AOS values of the catalysts are presented in Table 2 and ranges from 3.4
to 3.83, suggesting the presence of mixed states (Mn4+ and Mn3+), as
observed in the Mn 2p3/2 deconvolution results (Table 2 and Fig. 5a).
The increase in the Mn4+ and AOS with K+ concentration, indicates that
the synthesis route and K+ concentration may have an has impact on the
oxidation of Mn in the catalyst. It was suggested that higher ratio of
Mn4+/Mn3+ facilitates the redox cycle of Mn and enhances catalytic
oxidation of HCHO and other VOCs [10,14,17], by enhancing the cyclic
redox reaction of Mn (Mn4+ ↔ Mn3+) during the catalytic oxidation
process [13,35]. Interestingly, a significant reduction in both the AOS
and Mn4+ content of the acid-treated M0.5 (A0.1 M0.5) was observed, as
shown in Table 2, in agreement with previous reports [20,26]. The
presence of Mn3+ is associated with the creation of oxygen vacancies in
MnO2 to maintain electrostatic balance due to Mn4+ vacancies [36]. The
surface concentration of Mn3+ can be used as a relative measure for
determining oxygen vacancies. Therefore, the acid treatment does not
only facilitate the removal of K+ but also lead to the creation of va-
cancies, as reported elsewhere [26].
Physicochemical properties
The BET surface area, pore volume, and pore diameters of the cata-
lysts are presented in Table 1. The adsorption-desorption isotherms of all
the catalysts correspond to Type IV adsorption isotherm (Fig. S3a),
which is typical of mesoporous materials. This is further confirmed by
the pore size distribution (see Fig. S3b), while the average pore size of all
the catalysts (both pristine and acid-treated) falls within the range of 7-
21 nm. As seen from Table 1, the surface area initially increases with the
addition of KMnO4, but subsequently reduces as the amount of KMnO4
increases. The observed surface area trend conforms to the morpho-
logical transformation and particle size of the catalysts. As the particle
size decreases, an increase in surface area was observed until the for-
mation of compact and dense nanospheres. It is understood that the
aggregation of the particles into compacted nanospheres led to a loss in
surface area. However, an increase in surface area was observed in all
the acid treated catalysts. This might be related to decrease in particle
size, K+ leaching and the disruption of the layered structure during acid
treatment [25,26], as observed in the Raman and TEM results.
The peaks at 529.6–530 eV, 531.2–531.6 eV and 533.4-533.8 eV
were respectively allocated to lattice oxygen (Olat), surface adsorbed
oxygen species (Osuf) and surface adsorbed water (Owat) [13,35]. The
relative amounts of the different surfaces oxygen species of the catalysts
were determined from their respective areas and presented in Table 2.
The catalysts possess surface defective sites and unsaturated oxygen
belonging to defective oxides and hydroxyl groups (OH-, O-, O2-) [22,
35]. The amount of lattice oxygen Olat increases monotonously with
increasing K+ and Mn4+ content (Table 2), in accordance with the ob-
servations of Li et al.[14]. This shows that the K+ content, the oxidation
state of manganese, and the lattice oxygen content of the catalysts can be
systematically controlled by tuning the concentration of KMnO4 in the
precursor solution. The increase in the electron density around Olat due
to presence of K+ leads to a slight decrease in the binding energy of Olat
[19,37], in agreement with the observed weakening of the Mn-O bonds
with increasing K+ content in the Raman results (see next two para-
graphs). The surface concentration of Osuf significantly increased after
the acid treatment from 25% in M0.5 to 33% in A0.1 M0.5, indicating the
formation of oxygen vacancies, in agreement with the observed decrease
in the Mn4+ content after the acid treatment. A similar observation was
The elemental composition of the catalysts is presented in Table 1. As
Table 1
Textural properties and K contents of pristine and acid-treated catalysts
Potassium content
Pore
BET
Pore volume
(cm3/g)
Catalyst
size
(m2/g)
ICP
EDS
XPS
(%)
(nm)
(wt.%)
(wt.%)
M0
44.6
20.8
11.9
10.8
0.23
0.30
0.40
0
0
-
M0.2
A0.1
M0.2
M0.3
A0.1
M0.3
M0.5
A0.1
M0.5
115.6
160.2
2.28
0.35
4.57
-
2.79
-
51.8
12.2
7.1
0.15
0.37
9.98
0.86
8.92
-
6.68
-
223.4
22.4
43.7
12.5
8.4
0.05
0.09
13.72
1.63
12.71
2.46
14.89
1.5
4

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
Fig. 3. SEM images of (a) M0, (b) M0.2, (c) M0.3, (d) M0.5 and (e) A0.1 M0.5
.
Fig. 4. TEM and HR-TEM images of (a and b) M0.5, (b and d) A0.1 M0.5
.
5

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
Fig. 5. XPS Spectra of catalysts (a) Mn 2p, (b) O 1s, and (c) Mn 3s.
the activation energy for desorption and activation of water and oxygen
molecules to replenish the consumed oxygen species during HCHO
oxidation [19].
Table 2
XPS surface content and average oxidation state.
O content (%)
lat Osuf
Mn4+ content (%)
The Raman results (Fig. 2) were used to obtain more information on
the local structure of the catalysts. The increase in the intensity of the
Sample
O
AOS
ν2) peak (Fig. 2a), which is mainly due to Mn4+ in MnO6 basal plane
M0.2
39.23
44.72
86.52
36.86
55.89
65.98
70.00
62.30
40.24
30.51
25.14
33.23
3.42
3.55
3.83
3.40
(
[29], with increasing K+ content is related to the increase in the con-
centration of Mn4+, in agreement with the XPS results. The red shift of
the peak from 582.6 cm-1 in M0.2, to 575.5 and 569 cm-1 in M0.3 and M0.5
respectively, could be attributed to the resulting weakening of the Mn-O
bond strength of MnO2, due to its interaction with K+. The Raman red
shift of the Mn-O stretching is due to the strong interaction between K+
and O, which results in the weakening of the bond strength, and en-
hances the mobility of the lattice oxygen [21]. Interestingly, no further
red shift was observed in M0.6 and M0.8 (Fig. S1), perhaps because they
contain similar amount of K+ as M0.5. As such, it is expected that the
presence of K+ enhances the activity of the lattice oxygen due to the
weakened interaction between Mn and O, in agreement with the XPS
results. On the other hand, the observed peaks broadening and weak-
ening of the peak intensity of (ν2) in the acid treated catalyst (Fig. 2b) is
related to the decrease in the oxidation state of Mn/amount of Mn4+ and
local disorder due to vacancies in the MnO6 sheets [21,29], in agreement
M0.3
M0.5
A0.1 M0.5
reported elsewhere for acid-treated birnessite MnO2 [26,38].
However, the binding energy of the Olat in A0.1 M0.5 increased
significantly after the acid treatment (530.0 eV) compared to the pris-
tine M0.5 (529.6 eV). This indicates a reduction in the electron density
around the lattice oxygen due to the creation of manganese vacancies as
a result of K+ removal [19], in agreement with the observed decrease in
the content of Mn4+ in A0.1 M0.5. A similar shift to the higher binding
energy of the lattice oxide of birnessite MnO2 after the abstraction of K+
was observed elsewhere [19,38]. As such, more energy will be needed to
mobilize the lattice oxygen of the acid-treated catalysts, compared to its
pristine counterpart. It is therefore expected that K+ plays a vital role in
enhancing the mobility of lattice oxygen. K+ is also reported to decrease
6

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
with the XPS results. Therefore the acid treatment led to the removal of
K+, creation of oxygen vacancies, and reduction in the Mn4+ and lattice
oxygen content of the catalysts.
M0.5 and A0.1 M0.5 for 72 hrs continuous test. The temperatures for 50 %
(T50) and 90 % (T90) conversion were used to compare the catalysts. As
shown in Table 3, the T50 values for the pristine catalysts are 36 ◦C, 56
◦C, 64 ◦C and 107 ◦C, while T90 values are 64 ◦C, 68 ◦C, 77 ◦C and 118
◦C, for M0.5, M0.3, M0.2 and M0, respectively. Therefore the order of
increasing activity is as follows: M0.5 > M0.3 > M0.2 > M0. It is interesting
to observe that both M0.5 and M0.3 attained 100% at 70 ◦C, however, at
lower temperatures, M0.5 showed superior catalytic activity. The trend
of increasing catalytic is in agreement with the K+ content of the cata-
lysts and its effect in enhancing the activity of the lattice oxygen of the
catalysts, as observed in the Raman, XPS and H2 TPR results. It should
The surface reducibility and oxygen mobility of the catalysts were
investigated using H2-TPR, and presented in Fig. 6. M0.3 and M0.5 exhibit
similar reduction behavior with a broad reduction peaks which can be
attributed to the sequential reduction of Mn (IV) to Mn (II): MnO2
→
Mn2O3 → Mn3O4 → MnO [22,33]. Conversely, M0, M0.2 and A0.1 M0.5
showed two reduction peaks which are related to the reduction of MnO2
to Mn3O4 (227-252 ◦C) and Mn3O4 to MnO (294-371 ◦C) [13,20]. As
shown in Fig. 6, the reduction behavior of the catalysts is greatly
influenced by their K+ and Mn4+ content. Compared to M0.3, both the
main reduction peak and the lower temperature peak (reduction of
surface adsorbed oxygen), are shifted to lower reduction temperatures.
Similarly, all the reduction peaks of M0.2 are shifted to lower reduction
temperatures compared to M0. This indicates that the presence of K+
influences the reduction of Mn and enhance the activity and mobility of
the lattice oxygen [18]. The low-temperature reducibility could be
attributed to the weakening of the Mn-O bond and enhanced oxygen
mobility as in accordance with the XPS and Raman results. This shows
that the lattice oxygen is easily mobilized and activated to react with H2
in the presence of K+, as previously reported [5,19,37]. The presence of
also be noted that similar catalytic activity was observed over M0.5, M0.6
,
and M0.8, as shown in Fig. S8, possibly due to their similar content of K+,
as no increase in K+ concentration was observed beyond M0.5 Table S1.
To further investigate the effect of K+, the acid-treated catalysts were
evaluated under the same conditions as the pristine catalysts. As seen in
Fig. 7a (dashed lines), the overall catalytic activities of all the acid-
treated catalysts drastically reduced after the acid treatment. The T50
values for A0.1 M0.3 and A0.1 M0.2 are 71 ◦C and 74 ◦C respectively,
while the corresponding T90 values for A0.1 M0.5, A0.1 M0.3 and A0.1
M
0.2 are 87 ◦C, 86 ◦C and 94 ◦C. It is, however, important to mention that
despite the observed decrease in the overall catalytic activity, an initial
increase in activity was observed on all the acid-treated catalysts. This
could be ascribed to the enhanced reducibility of the surface adsorbed
oxygen and the generation of oxygen vacancies after acid treatment, as
seen in the XPS and H2-TPR results. This observation further buttresses
the role of surface adsorbed oxygen species in enhancing the catalytic
oxidation of HCHO, as previously reported [20,35].
K
+ in MnO2 catalysts weakens the Mn-O bond and increases the electron
density around the lattice oxygen, which increases their electron
acceptance potential, reducibility and activity [13,19,39]. It is however
important to note that, compared to M0.5, A0.1 M0.5 showed a lower
reduction temperature for the surface adsorbed oxygen, which could be
ascribed to the created oxygen vacancies due to the acid treatment [26],
in accordance with the XPS and Raman result. Nonetheless, it exhibited
higher temperature reduction profile, further demonstrating the role of
K+ in enhancing the activity and mobility of lattice oxygen.
Nevertheless, a slowdown in the activity of the acid-treated catalysts
is observed after prolonged exposure to HCHO. This could be related to
Catalytic activity for HCHO oxidation
Oxidation reaction (O2 rich environment)
The results of the oxidation reaction (60,000 ml⋅g-1 h-1, ~170 ppm of
HCHO, ~ 45% relative humidity) are presented in Fig. 7a, while Fig. 7b
presents the results of the room temperature oxidation and stability tests
(10 ppm HCHO, 60,000 ml⋅g-1 h-1, ~ 45% relative humidity, 30 ◦C) of
Fig. 7. Catalytic activity and stability tests (a) catalytic conversion of HCHO
over pristine and acid-treated catalysts, (b) room-temperature and stability tests
over M0.5, A0.1 M0.5 and M0 at 30 ◦C.
Fig. 6. H2 TPR analysis of catalysts.
7

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
Mn-La oxides for methane oxidation [17]. Tang et al.[10] also reported
that higher oxidation state Mn (Mn4+) and rich lattice oxygen, are
responsible for the high catalytic activity of MnO2-CeO2 composites for
HCHO oxidation. A similar phenomenon is observed here. As such, it is
rational to deduce that abundant lattice oxygen and higher Mn4+ con-
tent are favorable for HCHO oxidation, and that K+ facilitates the ac-
tivity of the lattice oxygen. The consumed lattice oxygen are
subsequently replenished by the activation of molecular oxygen and
water molecules through an electron transfer from K+ to oxygen [17,
19].
Table 3
Catalyst activity and specific rate.
Catalyst
Conversion Temperatures
Rate at 60 ◦
C
Specific rate at 60 ◦C
(◦C)
(μ
mol. min-
(μ
mol. m-2. min-1
)
1
T50 T90
)
M0
107 118
64 77
71 94
56 67
74 86
36 64
- 87
0.046
0.319
0.289
0.454
0.311
0.635
0.548
0.010
0.028
0.018
0.088
0.014
0.283
0.126
M0.2
A0.1 M0.2
M0.3
A0.1 M0.3
M0.5
A0.1 M0.5
It is also noteworthy to mention that despite the significant
improvement in the surface area of the acid-treated catalysts, no cor-
relation was observed between the surface area and catalytic activity. In
effect, a negative relationship was observed between surface area and
activity in the pristine catalysts and their acid-treated counterparts. The
effects of surface area and physical adsorption were further evaluated by
determining the specific reaction rates [22,35]. As observed in Table 3,
the specific rates of all the catalysts decreased drastically after the acid
treatment. From the above observation, it is obvious that the catalytic
activity is independent of the catalysts’ surface area.
the depletion of surface adsorbed oxygen, and the reduced ability of the
acid-treated catalysts to mobilize lattice oxygen and/or activate water
and oxygen molecules into oxygen species due to the absence of K+. The
observed lower reducibility and increase in the binding energy of the
lattice oxygen of A0.1 M0.5 could be a plausible explanation. The
removal of K+ results into reduction in the electron density around the
lattice oxygen and the reduction in the content of Mn4+. At lower re-
action temperature (< 50 ◦C) and exposure time, the higher concen-
tration of surface active oxygen species of the acid-treated catalysts
enhanced their catalytic activities. However, after the consumption of
the surface active oxygen species (< 50 ◦C), the replenishment of the
consumed species depends on the migration and activation of the lattice
oxygen [5], which is facilitated by the presence of K+ and the abundance
of the lattice oxygen.
Furthermore, as depicted in the room temperature HCHO oxidation
at 30 ◦C and stability studies (Fig. 7b), A0.1 M0.5 exhibited higher cat-
alytic activity in the first 15 hrs of the test, after which a sharp decline in
activity was observed. The initial higher activity is attributed to its
higher content of surface active oxygen species while the sharp decline
in activity is related to the depletion of surface active oxygen species and
its reduced ability to mobilize and activate lattice oxygen, as earlier
noted. Conversely, a relatively stable activity, an average of 68% con-
version, was maintained on M0.5 over the 72 hrs of continuous test. To
further probe the effect of K+, stability test was also conducted on M0
sample at 30 ◦C and the results are shown in Fig. 7b. Like the A0.1 M0.5
sample, a catalytic deactivation was also observed. It should however be
noted that its lower activity could be related be due to it being pre-
dominantly composed of Mn3O4 as shown by the XRD and Raman results
(Fig. 1 and 2a). MnO2 was previously shown to be more active than
Mn3O4 for HCHO oxidation [3].
HCHO oxidation, like many other VOCs, follows the Mars-van
Krevlin (MvK) reaction mechanism [40,41], as such, the redox proper-
ties of the metal oxide (Mn in this case) and the lattice oxygen mobility
play a very vital role in the reaction. The MvK mechanism involves the
consumption of lattice oxygen by the reactant (catalyst reduction) and
the replenishment of the consumed oxygen by molecular oxygen (cata-
lyst re-oxidation) [41,42]. High content Mn4+ enhances the redox cycle
of Mn and the transfer of lattice oxygen to promote the catalytic
oxidation of HCHO and other VOCs, by facilitating the cyclic redox re-
action (Mn4+ ↔ Mn3+) to supplement consumed lattice and surface
active oxygen during the catalytic oxidation process [10,13,14,17,35].
The activation and transfer of lattice oxygen to Mn is critical to
completing the redox cycle and enhances catalytic oxidation process
[10]. Zhang et al.[15] noted that the abundance of lattice oxygen in
δ-MnO2 is critical for HCHO oxidation. Wu et al.[13] also illustrated the
prime role of lattice oxygen and its mobility for the oxidation of o-xylene
In addition, the possible effect of surface contamination from resid-
ual acid on the deactivation and reduced activity was investigated by
subjecting the acid-treated samples to further calcination (A0.1 MxH) at
300 ◦C for 3 hrs and investigated under the similar conditions. Com-
parison of the activity of the non-calcined acid-treated samples (A0.1
Mx) and the calcined acid-treated samples (A0.1 MxH) (Fig. S9) and the
stability test (Fig. S10) showed similar performance trends in both cases.
These results further suggest that the observed decline in activity results
from the removal of K+.
over
α-MnO2. Furthermore, the catalytic activity of cryptomelane
octahedral molecular sieve MnO2 for the oxidation of benzene was
attributed to enhanced mobility of lattice oxygen [37]. As such, the
lattice oxygen activity, which is influenced by K+ content and abun-
dance of lattice oxygen, to complete the redox cycle of Mn, plays a
central role in HCHO oxidation.
Lattice oxygen test (O2 lean environment)
To further underscore the role of K+ in the mobility of lattice oxygen
for HCHO oxidation, an oxidation reaction was conducted in the absence
of molecular oxygen. The results are depicted in Fig. 8. To rule out the
contribution of external sources of O2 and CO2, pure N2 (99.99%) was
used, and blank tests were conducted with no evidence of CO2 in the
outlet gas stream. It is clearly evident that even in the absence of mo-
lecular O2, deep oxidation of HCHO into CO2 took place, with M0.5
attaining an average conversion of 59% in the first 60 min., while an
average of 21.5% was recorded on A0.1 M0.5. The fast and sharp decline
in the activity of A0.1 M0.5 is attributed to the depletion of surface active
oxygen, as observed earlier in the room temperature and stability test.
For M0.5 to achieve high conversion in the absence of molecular oxygen,
the supply of such of the consumed oxygen (see Fig. S11) could only
have come from its lattice oxygen, as evident by its higher concentration
of lattice oxygen in the XPS results (Table 2). This is indicative of its
superior ability to mobilize higher amount of lattice oxygen, which is
attributed to the effect of K+ in weakening the M-O bonds of MnO2,
thereby making the lattice oxygen more active. The participation of the
lattice oxygen in HCHO oxidation indicates that the reaction follows
Here, we observed a strong correlation between catalytic activity and
surface concentration of lattice oxygen, whose activity is enhanced by
K+. Potassium activates molecular oxygen and water to replenish the
supply of consumed oxygen species through an electron transfer to ox-
ygen [19]. The lattice oxygen mobility of OMS-2 significantly increases
with a corresponding increase in K+ concentration through electronic
delocalization effect [37].The presence of K+ weakens the Mn-O bonds
of manganese oxide thereby enhancing reducibility and mobility of the
lattice oxygen as shown by the Raman, XPS and H2-TPR results, and the
observed catalytic activity trend with increasing K+ concentration.
High oxidation state of Mn (Mn4+) favors the catalytic oxidation of
HCHO and other VOCs [10,13,14]. Zhang et al.[15] observed that
amongst the various crystal phases of MnO2, the catalyst with the
highest ratio of Mn4+/Mn3+ showed the best activity for HCHO oxida-
tion. Similar observations were reported for MnO2 based catalysts with
varying content of Mn4+ for the o-xylene oxidation [13], PdOx supported
manganese oxide catalysts for CO oxidation [16], and Ag modified
8

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
molecular HCHO were observed on both samples, indicating that HCHO
is instantaneously converted into its intermediate species upon adsorp-
tion onto the surface of both catalysts. The strong absorption bands in
the range of 3500 – 2900 cm-1 are due to surface strongly bonded hy-
droxyl groups [44]. The bands at 1570 – 1594, and 1352 cm-1 are
respectively assigned to the asymmetric stretching
ν (OCO)as and sym-
metric stretching vibration ν (OCO)s of formate species, while the peaks
at 2874 and 1374 cm-1 are attributed to
ν(CH) and δ(CH) of formate
species [44–46].[21,22] The peaks at 2695 and 2790 cm-1 are related to
the symmetric (νs) and asymmetric (νas) CH stretch of dioxymethylene
(DOM), 1415 cm-1 for δ(CH) modes, while the peaks 1161 and 1047 cm-1
are associated with the
ν (CO) stretch of DOM species [35,46,47]. The
peaks at 2326 cm-1 arise from the CO2 stretching vibration, while the
peaks at 1484-1493 and 1654 cm-1 are ascribed adsorbed carbonates
[20,22,48]. As reaction time increases, from 0 to 60 min., the intensity of
both the symmetric and asymmetric vibrations of formate species in-
creases, indicating that they are the dominant species through which
HCHO is converted into CO2.
As seen from Fig. 9a, HCHO is immediately converted into DOM
intermediates as it adsorbs onto the surface of M0.5, which is then pro-
gressively converted into formates as evidenced by the initial decrease in
the intensity of the peaks at 1493 cm-1 and the weakening of 1415 cm-1
after 10 min., followed by the rapid increase in the absorption of the
formate peak at 1594 cm-1. Mn4+ serves as active site for the activation
of HCHO and the generation of DOM species, from the reaction of HCHO
and surface lattice oxygen. The generated DOM species are further
converted into formates through oxidative dehydrogenation [3,49]. A
further reaction of the formates with surface OH- generates CO2 and
H2O. Surface OH- could be generated from the reaction of surface lattice
oxygen and H+ from the oxidative dehydrogenation of DOM [5], or from
Fig. 8. Lattice oxygen test (a) HCHO conversion over M0.5 and A0.1 M0.5 (b)
CO2 and HCHO gas evolution with reaction time.
MvK reaction mechanism as previously reported [40,41].
As such, a simple oxygen mobility and transfer mechanism is
envisaged. As the surface oxygen species are consumed in the reaction,
surface oxygen vacancies are created. Sub-surface or bulk lattice oxygen
species are mobilized into the created vacancies to replenish the
consumed surface oxygen and to sustain the reaction [15,43]. The
higher catalytic activity of M0.5 is an indication of the role of K+ in
enhancing the mobility of lattice oxygen and the ease of its re-oxidation
after reduction by HCHO. The decline in the activity of M0.5, even in the
presence of K+, is a result of the increasing difficulty to mobilize inner
lattice oxygen. Therefore, the role of K+ is not only limited to the
mobility of lattice oxygen but also facilitates the activation of molecular
oxygen to replace the consumed oxygen species [19], as evidenced by
the higher activity and stability of the K+ rich catalysts in the presence of
molecular oxygen (Fig. 7a and b). The XPS result of the spent catalysts
(Fig. S12) shows a drastic drop in the surface concentration of lattice
oxygen species compared to the fresh catalyst samples (Table 2). A drop
from 70 % to 52.9 % was observed on M0.5, while a corresponding
decrease from 62.3 % to 56.2 % was also observed on the spent A0.1
M
0.5. This further confirms the consumption of lattice oxygen in the
lattice oxygen test with more lattice oxygen consumed on M0.5 as seen in
Fig. 8.
Surface reaction pathway
The reaction pathway and the surface intermediate species formed
during HCHO oxidation on M0.5 and A0.1 M0.5 were investigated using
in-situ DRIFTS measurements (170 ppm, 30 ml min-1), at room tem-
perature for 60 min.. The DRIFTS results are depicted in Fig. 9a and 9b
for M0.5 and A0.1 M0.5 respectively. No peaks corresponding to
Fig. 9. In situ DRIFTS of (a) M0.5 (b) A0.1 M0.5
.
9

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
the reaction of water molecules with surface oxygen at the oxygen va-
cancy sites [22,35]. However, while DOM species were clearly observed
on the surface of M0.5, only a very weak feature related to the
ν (CO)
stretch (1047 cm-1) of DOM was observed on A0.1 M0.5 (Fig. 9b). The
initial higher activity of A0.1 M0.5 (high and room-temperature tests)
could be attributed to the direct formation of formates from the nucle-
ophilic attack of HCHO by surface active oxygen [50], or the faster rate
of DOM conversion into formates then to carbonate due its higher
density of surface adsorbed oxygen [22,35], as shown by the XPS results.
During the HCHO oxidation, surface oxygen species are continuously
consumed and replenished [5,48,51]. After the depletion of surface
oxygen, their replenishment largely depends on the activation and
transfer of lattice oxygen [5]. The Raman, XPS and H2-TPR results
suggest that the presence of K+ enhances the mobility of the lattice
oxygen. As a result of the weakened interaction between Mn-O, due to
the presence of K+, lattice oxygen is easily mobilized to re-oxidize the
reduced Mn (Mn4+ ↔ Mn3+), thus replenishing the consumed surface
oxygen. High content of Mn4+ also helps in enhancing the cyclic redox
behaviour of Mn [10,18]. Through an electron transfer from K+, mo-
lecular oxygen is activated into lattice oxygen in the created oxygen
vacancies [5,19], thus sustaining the supply of oxygen species for the
conversion of HCHO into its intermediate species, as depicted in Fig. 10.
Therefore the catalytic oxidation ability, to a large, depends on the ac-
tivity and mobility of the lattice oxygen to replenish the consumed
surface oxygen and re-oxidize the reduced Mn. As such, the observed
progressive deactivation of the A0.1 Mn (0.1) catalyst (Fig. 7b), can be
attributed to the continuous depletion of the surface oxygen and its
inability to sustain continuous supply, as opposed to M0.5, which even
though exhibited lower initial conversion, sustained an average
room-temperature of 68% over 72 hrs of continuous test, and achieved
complete conversion of high concentration of HCHO (170 ppm) at lower
temperature.
Fig. 10. Surface reaction of HCHO over K-rich M0.5
.
active oxygen, was attributed to the depletion of the surface oxygen
species, and lower activity and mobility of the lattice oxygen to sustain
the replenishment of active oxygen species, due to low K+ content.
CRediT authorship contribution statement
Abubakar Yusuf: Conceptualization, Investigation, Writing - orig-
inal draft. Yong Sun: Data curation, Formal analysis. Colin Snape:
Supervision, Writing - review & editing. Jun He : Supervision, Re-
sources, Project administration, Funding acquisition, Writing - review &
editing. Chengjun Wang: Resources, Conceptualization, Writing - re-
view & editing. Yong Ren: Investigation. Hongpeng Jia: Investigation,
Resources.
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgements
Conclusion
The morphological structure, and the content of K+, Mn4+, and lat-
tice oxygen of MnOx catalysts can be methodically controlled by
adjusting the ratio of KMnO4 and Mn(NO3)2 in the reaction mixture.
High surface content of Mn4+ enhances the cyclic redox behavior of the
catalyst, while K+ enhances the lattice oxygen activity and helps in the
activation of molecular oxygen and to replace the consumed lattice
oxygen, thereby enhancing activity for HCHO oxidation. On the other
hand, the abstraction of K+, via acid treatment, led to the creation of
vacancies and an increase in concentration surface active oxygen spe-
cies, resulting into an initial enhancement of catalytic activity. However,
due to the absence of or at low K+ concentration, lattice oxygen mobility
and the activation of molecular and/or water molecules, to replace the
consumed surface active oxygen, is impacted, leading to reduced cata-
lytic activity. In addition, lattice oxygen test shows that the catalysts
lattice oxygen is consumed during the HCHO oxidation process. Despite
the increase in surface area of the acids-treated catalysts, and surface
area variation of the pristine catalysts, no correlation was observed
between surface area and catalytic activity.
This work was carried out at the International Doctoral Innovation
Centre (IDIC), University of Nottingham Ningbo, China. The author
acknowledges the financial support from IDIC, Ningbo Education Bu-
reau, Ningbo Science and Technology Bureau, and the University of
Nottingham. This work was also partially supported by Ningbo Science
and Technology Bureau under the Innovation Team Project
(2017C510001) and UK Engineering and Physical Sciences Research
Council (EP/G037345/1 and EP/L016362/1).
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.mcat.2020.111204.
References
[1] International Agency for Research on Cancer, World Health Organization,
Formaldehyde, 2-butoxyethanol and 1-tert-butoxypropan-2-ol., Vol. 88, 2004,
pp. 39–325. Lyon, France.
Catalytic activity increases with increasing K+ content, and the
surface concentration of lattice oxygen and Mn4+. M0.5 with the highest
concentration of K+, Mn4+, and lattice oxygen exhibited the best cata-
lytic activity, attaining complete conversion of HCHO at 70 ◦C (170 ppm
GHSV of 60,000 ml/g⋅h), while similar conversion was attained by A0.1
M0.5 at 100 ◦C. Room temperature tests conducted at 30 ◦C under a
continuous flow of ~10 ppm HCHO showed that M0.5 is stable for up to
72 hrs with an average conversion of 68%, while fast deactivation was
observed on A0.1 M0.5. The stability and ability of the K+ rich catalyst
(M0.5) to sustain room temperature oxidation of HCHO are attributed to
the lattice oxygen mobility and oxygen/water activation effect of K+ to
replace the consumed lattice oxygen. Conversely, the fast deactivation of
the acid-treated catalyst, despite possessing higher content surface
[2] T. Salthammer, S. Mentese, R. Marutzky, Formaldehyde in the indoor environment,
Chem. Rev. 110 (4) (2010) 2536–2572.
[3] Y. Sekine, Oxidative decomposition of formaldehyde by metal oxides at room
temperature, Atmos. Environ. 36 (35) (2002) 5543–5547.
[4] A. Yusuf, C. Snape, J. He, H. Xu, C. Liu, M. Zhao, G.Z. Chen, B. Tang, C. Wang,
J. Wang, S.N. Behera, Advances on transition metal oxides catalysts for
formaldehyde oxidation: A review, Catal. Rev.: Sci. Eng. 59 (3) (2017) 189–233.
[5] B. Bai, J. Li, Positive effects of K+ ions on three-dimensional mesoporous Ag/Co3O4
catalyst for HCHO oxidation, ACS Catal. 4 (8) (2014) 2753–2762.
[6] L. Nie, J. Yu, X. Li, B. Cheng, G. Liu, M. Jaroniec, Enhanced performance of NaOH-
Modified Pt/TiO2 toward room temperature selective oxidation of formaldehyde,
Environ. Sci. Technol. 47 (6) (2013) 2777–2783.
[7] C. Zhang, Y. Li, Y. Wang, H. He, Sodium-Promoted Pd/TiO2 for catalytic oxidation
of formaldehyde at ambient temperature, Environ. Sci. Technol. 48 (10) (2014)
5816–5822.
[8] C. Zhang, F. Liu, Y. Zhai, H. Ariga, N. Yi, Y. Liu, K. Asakura, M. Flytzani-
Stephanopoulos, H. He, Alkali-metal-promoted Pt/TiO2 opens a more efficient
10

A. Yusuf et al.
Molecular Catalysis 497 (2020) 111204
pathway to formaldehyde oxidation at ambient temperatures, Angew. Chem., Int.
[30] N.P. Arias, M.E. Becerra, O. Giraldo, Structural and electrical studies for birnessite-
type materials synthesized by solid-state reactions, Nanomater. 9 (8) (2019) 1156.
[31] H. Yin, K.D. Kwon, J.-Y. Lee, Y. Shen, H. Zhao, X. Wang, F. Liu, J. Zhang, X. Feng,
Distinct effects of Al3+ doping on the structure and properties of hexagonal
turbostratic birnessite: A comparison with Fe3+ doping, Geochim. Cosmochim.
Acta. 208 (2017) 268–284.
Ed. Engl. 51 (38) (2012) 9628–9632.
[9] J. Quiroz, J.-M. Giraudon, A. Gervasini, C. Dujardin, C. Lancelot, M. Trentesaux, J.-
F. Lamonier, Total oxidation of formaldehyde over MnOx-CeO2 catalysts: The effect
of acid treatment, ACS Catal. 5 (4) (2015) 2260–2269.
[10] X. Tang, Y. Li, H. Xiumin, Y. Xu, H. Zhu, J. Wang, W. Shen, MnOx–CeO2 mixed
oxide catalysts for complete oxidation of formaldehyde: Effect of preparation
method and calcination temperature, Appl. Catal., B 62 (2006) 265–273.
[11] Y.-G. Yin, W.-Q. Xu, S.L. Suib, C.L. O’Young, Lattice oxygen mobility and structural
stability of Ni and Cu octahedral molecular sieves having the cryptomelane
structure, Inorg. Chem. 34 (16) (1995) 4187–4193.
[32] D. Chen, D. Ding, X. Li, G.H. Waller, X. Xiong, M.A. El-Sayed, M. Liu, Probing the
charge storage mechanism of a pseudocapacitive MnO2 electrode using in
operando Raman Spectroscopy, Chem. Mater. 27 (19) (2015) 6608–6619.
[33] X. Zhang, Z.a. Ma, Z. Song, H. Zhao, W. Liu, M. Zhao, J. Zhao, Role of cryptomelane
in surface-adsorbed oxygen and Mn chemical valence in MnOx during the catalytic
oxidation of Toluene, J. Phys. Chem. C 123 (28) (2019) 17255–17264.
´
˜
[12] V. Santos, M. Pereira, J. Orfao, J. Figueiredo, The role of lattice oxygen on the
activity of manganese oxides towards the oxidation of volatile organic compounds,
Appl. Catal., B 99 (1-2) (2010) 353–363.
´
˜
[34] V. Santos, O. Soares, J. Bakker, M. Pereira, J. Orfao, J. Gascon, F. Kapteijn,
J. Figueiredo, Structural and chemical disorder of cryptomelane promoted by alkali
doping: Influence on catalytic properties, J. Catal. 293 (2012) 165–174.
[35] J. Wang, G. Zhang, P. Zhang, Layered birnessite-type MnO2 with surface pits for
enhanced catalytic formaldehyde oxidation activity, J. Mater. Chem. A 5 (12)
(2017) 5719–5725.
[13] Y. Wu, Y. Lu, C. Song, Z. Ma, S. Xing, Y. Gao, A novel redox-precipitation method
for the preparation of
α
-MnO2 with a high surface Mn4+ concentration and its
activity toward complete catalytic oxidation of o-xylene, Catal. Today 201 (2013)
32–39.
[14] K. Li, C. Chen, H. Zhang, X. Hu, T. Sun, J. Jia, Effects of phase structure of MnO2
and morphology of δ-MnO2 on toluene catalytic oxidation, Appl. Surf. Sci. 496
(2019), 143662.
[36] J. Hou, Y. Li, L. Liu, L. Ren, X. Zhao, Effect of giant oxygen vacancy defects on the
catalytic oxidation of OMS-2 nanorods, J. Mater. Chem. A 1 (23) (2013)
6736–6741.
[15] J. Zhang, Y. Li, L. Wang, C. Zhang, H. He, Catalytic oxidation of formaldehyde over
manganese oxides with different crystal structures, Catal. Sci. Technol. 5 (4)
(2015) 2305–2313.
[37] J. Hou, L. Liu, Y. Li, M. Mao, H. Lv, X. Zhao, Tuning the K+ concentration in the
tunnel of OMS-2 nanorods leads to a significant enhancement of the catalytic
activity for benzene oxidation, Environ. Sci. Technol. 47 (23) (2013)
13730–13736.
[16] J.S. Park, D.S. Doh, K.Y. Lee, High catalytic activity of PdOx/MnO2 for CO
oxidation and importance of oxidation state of Mn, Top. Catal. 10 (1) (2000)
127–131.
[38] T. Gopi, G. Swetha, S. Chandra Shekar, C. Ramakrishna, B. Saini, R. Krishna, P.V.
L. Rao, Catalytic decomposition of ozone on nanostructured potassium and proton
containing δ-MnO2 catalysts, Catal. Commun. 92 (2017) 51–55.
[39] Y. Chen, Z. Huang, M. Zhou, Z. Ma, J. Chen, X. Tang, Single silver adatoms on
nanostructured manganese oxide surfaces: Boosting oxygen activation for benzene
abatement, Environ. Sci. Technol. 51 (4) (2017) 2304–2311.
[40] M. He, J. Ji, B. Liu, H. Huang, Reduced TiO2 with tunable oxygen vacancies for
catalytic oxidation of formaldehyde at room temperature, Appl. Surf. Sci. 473
(2019) 934–942.
[17] A. Machocki, T. Ioannides, B. Stasinska, W. Gac, G. Avgouropoulos, D. Delimaris,
w. Grzegorczyk, S. Pasieczna-Patkowska, Manganese–lanthanum oxides modified
with silver for the catalytic combustion of methane, J. Catal. 227 (2004) 282–296.
[18] X. Tang, J. Chen, X. Huang, Y. Xu, W. Shen, Pt/MnOx–CeO2 catalysts for the
complete oxidation of formaldehyde at ambient temperature, Appl. Catal., B 81 (1-
2) (2008) 115–121.
[19] J. Wang, J. Li, P. Zhang, G. Zhang, Understanding the “seesaw effect” of
interlayered K+ with different structure in manganese oxides for the enhanced
formaldehyde oxidation, Appl. Catal., B 224 (2018) 863–870.
[20] R. Fang, Q. Feng, H. Huang, J. Ji, M. He, Y. Zhan, B. Liu, D.Y.C. Leung, Effect of K+
ions on efficient room-temperature degradation of formaldehyde over MnO2
catalysts, Catal. Today 327 (2019) 154–160.
[41] P. Liu, H. He, G. Wei, X. Liang, F. Qi, F. Tan, W. Tan, J. Zhu, R. Zhu, Effect of Mn
substitution on the promoted formaldehyde oxidation over spinel ferrite: Catalyst
characterization, performance and reaction mechanism, Appl. Catal., B 182 (2016)
476–484.
[42] I.E. Wachs, Catalysis science of supported vanadium oxide catalysts, Dalton Trans.
42 (33) (2013) 11762–11769.
[21] S. Rong, K. Li, P. Zhang, F. Liu, J. Zhang, Potassium associated manganese vacancy
in birnessite-type manganese dioxide for airborne formaldehyde oxidation, Catal.
Sci. Technol. 8 (7) (2018) 1799–1812.
[43] E. Aneggi, C. de Leitenburg, A. Trovarelli, On the role of lattice/surface oxygen in
ceria–zirconia catalysts for diesel soot combustion, Catal. Today 181 (1) (2012)
108–115.
[22] J. Wang, J. Li, C. Jiang, P. Zhou, P. Zhang, J. Yu, The effect of manganese vacancy
in birnessite-type MnO2 on room-temperature oxidation of formaldehyde in air,
Appl. Catal., B 204 (2017) 147–155.
[44] C. Li, K. Domen, K.-i. Maruya, T. Onishi, Spectroscopic identification of adsorbed
species derived from adsorption and decomposition of formic acid, methanol, and
formaldehyde on cerium oxide, J. Catal. 125 (2) (1990) 445–455.
[45] J.P. Durand, S.D. Senanayake, S.L. Suib, D.R. Mullins, Reaction of formic acid over
amorphous manganese oxide catalytic systems: An in situ study, J. Phys. Chem. C
114 (47) (2010) 20000–20006.
[23] C.S. Spanjers, C.A. Beach, A.J. Jones, P.J. Dauenhauer, Increasing flame ionization
detector (FID) sensitivity using post-column oxidation–methanation, Anal.
Methods 9 (12) (2017) 1928–1934.
[24] J.T. Sampanthar, J. Dou, G.G. Joo, E. Widjaja, L.Q.H. Eunice, Template-free low
temperature hydrothermal synthesis and characterization of rod-shaped
manganese oxyhydroxides and manganese oxides, Nanotechnol. 18 (2) (2006),
025601.
[46] Z. Fan, Z. Zhang, W. Fang, X. Yao, G. Zou, W. Shangguan, Low-temperature
catalytic oxidation of formaldehyde over Co3O4 catalysts prepared using various
precipitants, Chinese J. Catal. 37 (6) (2016) 947–954.
[25] V.B.R. Boppana, S. Yusuf, G.S. Hutchings, F. Jiao, Nanostructured alkaline-cation-
containing δ-MnO2 for photocatalytic water oxidation, Adv. Funct. Mater. 23 (7)
(2013) 878–884.
[47] C.F. Mao, M.A. Vannice, Formaldehyde Oxidation over Ag Catalysts, J. Catal. 154
(2) (1995) 230–244.
[48] J. Wang, P. Zhang, J. Li, C. Jiang, R. Yunus, J. Kim, Room-Temperature Oxidation
of Formaldehyde by Layered Manganese Oxide: Effect of Water, Environ. Sci.
Technol. 49 (20) (2015) 12372–12379.
[26] Y. Liu, W. Yang, P. Zhang, J. Zhang, Nitric acid-treated birnessite-type MnO2: An
efficient and hydrophobic material for humid ozone decomposition, Appl. Surf. Sci.
442 (2018) 640–649.
[49] J.T. Dickinson, R.J. Madix, The mechanism of methanol formation and other
reactions following formaldehyde adsorption on Ni (110), Int. J. Chem. Kinet. 10
(8) (1978) 871–882.
[27] M.R. Awual, S.A. El-Safty, A. Jyo, Removal of trace arsenic (V) and phosphate from
water by a highly selective ligand exchange adsorbent, J. Environ. Sci. 23 (12)
(2011) 1947–1954.
[50] C. Ma, D. Wang, W. Xue, B. Dou, H. Wang, Z. Hao, Investigation of formaldehyde
oxidation over Co3O4ꢀ CeO2 and Au/Co3O4ꢀ CeO2 catalysts at room temperature:
effective removal and determination of reaction mechanism, Environ. Sci. Technol.
45 (8) (2011) 3628–3634.
[28] F. Buciuman, F. Patcas, R. Craciun, D.R.T. Zahn, Vibrational spectroscopy of bulk
and supported manganese oxides, Phys. Chem. Chem. Phys. 1 (1) (1999) 185–190.
[29] C. Julien, M. Massot, R. Baddour-Hadjean, S. Franger, S. Bach, J. Pereira-Ramos,
Raman spectra of birnessite manganese dioxides, Solid State Ionics 159 (3-4)
(2003) 345–356.
[51] L. Zhu, J. Wang, S. Rong, H. Wang, P. Zhang, Cerium modified birnessite-type
MnO2 for gaseous formaldehyde oxidation at low temperature, Appl. Catal., B 211
(2017) 212–221.
11