A platinum promoted Ag/SBA-15 catalyst effective in selective oxidation of methanol-design and surface characterization

Article abstract of DOI:10.1039/d0ra01562h

The aim of this study was better understanding of surface properties of bimetallic (silver-platinum) catalysts and to verify if a very small addition of platinum (ca. 0.05 wt%) to silver (ca. 2.0 wt%) loaded on ordered mesoporous silica, SBA-15, would improve the catalytic properties of bimetallic Ag-Pt materials in selective oxidation of methanol to methyl formate. Ag-Pt catalysts were prepared by one-step and step-by-step procedures and the final Ag/Pt molar ratio in the respective samples was equal to 86 and 63. The catalysts were characterized after calcination and different activation treatments (in Ar and O₂). X-ray diffraction, UV-vis and XP spectroscopy confirmed the lack of Ag-Pt alloy crystallites in the samples and also evidenced a higher resistance of silver oxide species to reduction upon activation in Ar flow in the presence of platinum promoter interacting with silver species. Methanol oxidation over the samples activated in Ar flow and in oxidizing flow (O₂ + Ar) helped identify the role of each component in the bimetallic Ag-Pt catalyst in terms of activity and selectivity in the oxidation of methanol to methyl formate. A highly active bimetallic Pt/Ag/SBA-15 catalyst, selective to methyl formate and stable in methanol oxidation was constructed.

Full text of DOI:10.1039/d0ra01562h

RSC Advances
View Article Online
View Journal | View Issue
PAPER
A platinum promoted Ag/SBA-15 catalyst effective
in selective oxidation of methanol – design and
surface characterization†
Cite this: RSC Adv., 2020, 10, 14570
Joanna Wisniewska, * Izabela Dziedzic and Maria Ziolek
The aim of this study was better understanding of surface properties of bimetallic (silver–platinum) catalysts and
to verify if a very small addition of platinum (ca. 0.05 wt%) to silver (ca. 2.0 wt%) loaded on ordered mesoporous
silica, SBA-15, would improve the catalytic properties of bimetallic Ag–Pt materials in selective oxidation of
methanol to methyl formate. Ag–Pt catalysts were prepared by one-step and step-by-step procedures and
the final Ag/Pt molar ratio in the respective samples was equal to 86 and 63. The catalysts were characterized
2
after calcination and different activation treatments (in Ar and O ). X-ray diffraction, UV-vis and XP
spectroscopy confirmed the lack of Ag–Pt alloy crystallites in the samples and also evidenced a higher
resistance of silver oxide species to reduction upon activation in Ar flow in the presence of platinum
promoter interacting with silver species. Methanol oxidation over the samples activated in Ar flow and in
Received 18th February 2020
Accepted 2nd April 2020
2
oxidizing flow (O + Ar) helped identify the role of each component in the bimetallic Ag–Pt catalyst in terms
DOI: 10.1039/d0ra01562h
of activity and selectivity in the oxidation of methanol to methyl formate. A highly active bimetallic Pt/Ag/SBA-
15 catalyst, selective to methyl formate and stable in methanol oxidation was constructed.
rsc.li/rsc-advances
Methyl formate is a useful organic chemical and can be used
as an intermediate for the preparation of a wide range of other
1
. Introduction
Catalyst surface engineering is important in the construction of chemicals like formic acid, higher carboxylic acids such as
catalysts intended for desired processes. In this work the focus is on acetic acid and propionic acid and their esters. It is also used in
the design and study of surface properties of bimetallic (silver– cereal and tobacco crops as a fumigant, in the cellulose industry
platinum) materials based on ordered mesoporous silica (SBA-15) as a solvent, in foundries in the process of curing resins, in the
proposed for the selective oxidation of methanol to methyl formate. curing of phenol esters. Production of methyl formate from
Catalytic methanol oxidation to formaldehyde is applied in methanol is desirable because methanol is a readily available
1
industry with the use of silver or iron–molybdenum catalysts.
starting material and can be obtained either from natural gas or
4–6
The application of both catalysts requires different reaction from synthesis gas also generated from biomass conversion.
conditions. The rst, silver catalyst works at higher tempera-
7
One of the old patents presented an invention that
tures (823–1023 K) and higher concentration of methanol in the described a process for the preparation of methyl formate from
reagents mixture, whereas the second catalyst works in the methanol in which methanol was contacted at elevated
range 623–723 K in lower methanol concentration. It is well temperature with a platinum group metal catalyst in the
known that methanol oxidation is also an attractive process for absence of an added hydrogen acceptor. These metals were
production of other chemicals like methyl formate or dime- supported on inert solids, among others on silica, having
2
,3
thoxy methane. Both are commercially needed.
hydroxyl groups capable of reacting with the platinum group
compound. The preferable temperature range for this reaction
was 413–453 K. The oxidation of methanol vapor over platinum
8
–11
Faculty of Chemistry, Adam Mickiewicz University in Pozna n´ , Uniwersytetu catalysts was studied by many research groups.
Latest works
Pozna n´ skiego 8, 61-614 Pozna n´ , Poland. E-mail: jc62431@amu.edu.pl; Tel: +48 61
focused on the effects of platinum nanoparticle size and metal/
8
291794
11–13
support interactions. It has been found out
that Pt cluster
†
Electronic supplementary information (ESI) available: S1: functionalization of
diameter inuences the rate and selectivity of the reaction.
Larger Pt clusters contain coordinatively more saturated surface
SBA-15 support with 3-mercaptopropyltrimethoxysilane (MPTMS). S2: samples
characterization. Fig. S1: low-angle XRD patterns of SBA-15 (denoted S) and the
support modied with platinum and/or silver. Fig. S2: nitrogen Pt sites than smaller Pt clusters. It results in a weaker binding to
adsorption/desorption isotherms of the samples calcined at 773 K for 4 h.
chemisorbed oxygen and methanol derived intermediates. The
Fig. S3: TEM images of the samples calcined at 773 K for 4 h activated at 673 K
preoxidized platinum catalyst Pt/Al O , in which a larger degree
ꢀ1
2 3
in argon ow (40 mL min ) for 2 h. Fig. S4: UV-vis spectra of Pt/Ag/S catalyst:
ꢀ1
of oxidation of platinum was observed, was more active than the
activated at 673 K in Ar ow (40 mL min ) for 2 h (a) and aer methanol
prereduced catalyst. The differences in reactivity were linked to
oxidation performed at 523 K for 330 min (b). See DOI: 10.1039/d0ra01562h
14570 | RSC Adv., 2020, 10, 14570–14580
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
14
the formation and stabilization of distinct active oxide species to our previous work (S1-ESI†). The functionalized material is
9
during the pretreatment.
marked as SBA-SH.
Inspired by metallic catalysts for the oxidation of methanol
to formaldehyde and methyl formate we have combined both
metals, silver and platinum loaded on silica supports as cata-
2
.2. Modication of functionalized support with platinum
and/or silver
14–16
14
lysts for selective methanol oxidation.
Our previous studies
Bimetallic catalysts containing silver and platinum were
prepared according to two approaches: one-step or step-by-step
modication of functionalized support with metal precursors.
The assumed metals loading in both cases was 2.0 wt% for silver
and 0.1 wt% for platinum.
have been focused on nding the optimal conditions of alloying
silver and platinum loaded on different silicas and studying of
their properties in methanol oxidation performed in the gas
phase. We have been successful in preparation of samples
containing Ag–Pt alloy nanoparticles. However, the alloy species
were not very stable upon activation conditions which always
2.2.1 One-step method – simultaneous incorporation of
two metals. The preparation procedure of the bimetallic Ag–Pt
system was as follows: 25 mL of water solution of silver nitrate
14,17
takes place before oxidation processes.
In the most of
examined bimetallic Ag–Pt samples the activation in argon ow
caused total or partial segregation of platinum and silver on the
surface of silica support. It resulted in a decrease in activity and/
or selectivity in methanol oxidation. From among the studied
samples the only one in which the alloy structure was main-
tained was the catalyst based on mesoporous cellular foam
(AgNO
3
, $99.8%, Sigma-Aldrich) and 25 mL of water solution of
hexachloroplatinate(IV) hydrate (H PtCl , 38.84%, Johnson
2
6
Matthey) were prepared separately and then mixed for 5 min in
one ask. Then the functionalized support was stirred for 4 h
together with the prepared solution at room temperature. The
solid was recovered by ltration and washed with 80 mL of
distilled water. The recovered material was stirred at room
(MCF) containing silver & platinum (Ag/Pt molar ratio ¼ 4.2).
This sample displayed 100% conversion of methanol with 73%
selectivity to methyl formate at a relatively low temperature, 423
K. However, this catalyst not only showed decreased activity
aer a few hours of the reaction, but also changed selectivity
and increased the total oxidation of methanol. The reason was
the following segregation of alloy phase and agglomeration of
single metal particles during the reaction performed under the
applied conditions. Aer segregation of alloy components, both
silver and platinum species acted as separate phases active
temperature with 60 mL of 0.1 M sodium borohydride (NaBH
4
,
>99.8%, Sigma-Aldrich) solution. Aer 40 min, the solid was
recovered by ltration, washed with 80 mL of distilled water.
The material was dried at 333 K for 12 h and calcined in air at
773 K for 4 h. The calcination step was applied for removal of
organosilane (MPTMS) which was used for functionalization of
silica before anchoring of metal ions. The prepared sample was
labelled as AgPt/S, where S stands for SBA-15.
2.2.2 Step-by-step method – rst silver next platinum
2
mainly in total oxidation of methanol to CO . Such instability of
anchoring. The preparation procedure was as follows: the
functionalized material SBA-SH was stirred for 2 h at room
temperature in 25 mL of water solution of silver nitrate. Aer
the bimetallic silver & platinum catalysts provoked us to look for
a different composition of Ag & Pt catalysts. The focus was on
such a composition that would contain signicantly less plat-
inum than silver, because separated platinum species (formed
during methanol oxidation as a result of alloy segregation)
mainly activated the total oxidation of methanol.
The research hypothesis in this work was that a very small
addition of platinum to silver loaded on ordered mesoporous
silica, SBA-15, would be able to improve selectivity of the cata-
lyst into methanol oxidation to methyl formate and that a plat-
inum promoter would provide higher stability of the catalyst
during methanol oxidation and thus allows getting stable
methyl formate production. To check this hypothesis, SBA-15
material was modied with 2.0 wt% of silver and 0.1 wt% of
platinum applying two different procedures, one-step and step-
by-step modication, rst silver and next platinum loading. The
results of methanol oxidation on so-prepared samples appeared
to be promising for selective formation of methyl formate.

ltration and washing with 80 mL of distilled water the recov-
ered solid (R) was divided into two parts, R1 and R2. R1 was
stirred at room temperature with 60 mL of 0.1 M NaBH solu-
4
tion for 40 min to obtain the reduced monometallic sample.
The solid was recovered by ltration, washed with 80 mL of
distilled water. Then it was dried at 373 K for 12 h and calcined
at 773 K for 4 h. In this way the material marked as Ag/S was
obtained, where S stands for SBA-15.
To prepare the bimetallic material, the second part of the
recovered product (R2) was moved into the beaker and stirred
for 2 h at room temperature in 25 mL of water solution of
2 6
hexachloroplatinate(IV) hydrate H PtCl . The next steps were the
same as above described for monometallic sample. Bimetallic
material was labelled as Pt/Ag/S where S stands for SBA-15.
Monometallic platinum catalyst Pt/S was prepared in the
same way as Ag/S with the use of hexachloroplatinate(IV) hydrate
H PtCl as platinum precursor, instead of silver nitrate.
2
6
2. Experimental
2.1. Synthesis and functionalization of SBA-15 support with
2.3. Samples characterization
3-mercaptopropyltrimethoxysilane (MPTMS)
The materials prepared were characterized by using inductively
The SBA-15 support was prepared according to the procedure coupled plasma optical emission spectrometry (ICP-OES), X-ray
described in ref. 18 and 19. It was graed with 3-mercaptopro- diffraction (XRD), N -adsorption/desorption isotherms, Trans-
2
pyltrimethoxysilane (MPTMS) (95%, Sigma-Aldrich) to func- mission Electron Microscopy (TEM), ultraviolet-visible (UV-vis)
tionalize its surface before modication with metals, according spectroscopy, and X-ray photoelectron spectroscopy (XPS). All
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 14570–14580 | 14571

View Article Online
RSC Advances
Paper
these techniques are described in details in the ESI (S2†), of mesoporous silica did not change the mesoporous structure
1
4,15
according to our previous papers.
of the support. Textural parameters calculated from N2
adsorption/desorption isotherms (Fig. S2-ESI†) of all samples
prepared are summarized in Table 1. The surface area reached
2.4. Methanol oxidation
2
ꢀ1
2
ꢀ1
785 m g for pure support and ca. 400–500 m g for the
The catalytic reaction was performed in a xed-bed ow reactor,
as described in our previous works.
granulated catalyst of the size fraction of 0.5 < F < 1 mm was
placed into a tubular reactor (diameter F ¼ 5 mm, length l ¼ 70
mm). The height of the catalyst bed was 6–7 mm. The temper-
ature was controlled by a thermocouple located in the catalyst
support modied with metals. There are no signicant differ-
ences in the average pore diameter and total pore volume of
prepared monometallic and bimetallic catalysts. Metal modi-
cation of silica support rstly functionalized by MPTMS led to
a decrease in surface area, pore diameter and pore volume aer
introduction of modiers. The chemical composition of the
samples was estimated on the basis of ICP-OES and the results
are collected in Table 1. The assumed metals loading was equal
to 2.0 wt% for silver and 0.1 wt% for platinum. The modication
efficiency reached 100% for monometallic silver sample and 87–
1
4,15
A portion of 0.04 g of
ꢀ
1
bed. The samples were activated in argon ow (40 mL min ) or
ꢀ1
ꢀ1
in oxidizing ow (10 mL min
for 2 h (with a ramp of 15 K min ). Then, the temperature was
decreased to that of the reaction. The reactant mixture of Ar/O
2
O + 30 mL min Ar) at 673 K
ꢀ
1
2
ꢀ
/
.
1
CH OH (88/8/4 mol%) was supplied at the rate of 40 mL min
3
95% for silver in bimetallic materials. The efficiency of plat-
The external (inspected by tests with the use of different
volumes of catalysts and constant granular size of catalysts and
contact time) and internal (inspected by tests with the use of
different granular sizes of catalyst and constant volume of
catalyst and contact time) diffusion do not limit the reaction
rate in these conditions. Methanol (Chempur, Poland) was
introduced to the ow reactor by bubbling argon gas through
a glass saturator lled with methanol. The reactor effluent was
analyzed using two online gas chromatographs. One gas chro-
matograph, GC 8000 Top, was equipped with a capillary column
of DB-1, operated at 313 K, and a FID detector applied for
analyses of organic compounds, while the other GC containing
Porapak Q and 5A molecular sieves columns (used for analyses
inum introduction was ca. 50% for all samples. The real Ag/Pt
molar ratio for the sample prepared in the one-step modica-
tion was equal to 86 (AgPt/S), whereas for the sample prepared
by the step-by-step method it was equal to 63 (Pt/Ag/S). These
Ag/Pt molar ratios are much higher than those required for the
20
formation of Ag–Pt alloy nanoparticles. In the one-pot modi-

cation the initial solution contained 20 times lower amount of
platinum ions than silver. It is possible that much higher
+
amount of Ag ions interacting with thiol groups (SH from
ꢀ
MPTMS – functionalization agent) limited the access of PtCl6
to anchoring sites. A kind of competition between Pt and Ag
sources occurred. In contrast, during the step-by-step modi-
+
cation procedure, silver species (Ag ions from silver nitrate)
2 2 2 3
of O , CO , CO, H O, and CH OH) had a TCD detector. The
added rstly were strongly held on SH groups. In such a form
columns in the second chromatograph with TCD were heated
according to the following program: 5 min at 358 K, tempera-
ꢀ
they provided good adsorption sites for anchoring of PtCl
6
. It
can be a reason for a higher amount of platinum anchored in
this procedure.
ꢀ
1
ture increase to 408 K (heating rate 5 K min ), 4 min at 408 K,
cooling to 358 K (for the automatic injection onto the column
with 5A), 10 min at 358 K, temperature increase to 408 K
ꢀ
1
(
heating rate 10 K min ), and 11 min at 408 K. Argon was 3.2. State of silver and platinum
applied as a carrier gas. The outlet stream line from the reactor
to the gas chromatograph was heated at about 373 K to avoid
3
.2.1 Catalysts aer calcination. The types of active species
in calcined samples were identied rstly from UV-vis spectra
Fig. 1). According to literature the surface plasmon resonance
SPR) band of spherical metallic silver nanoparticles appears at
condensation of the reaction products. The selectivity, S
i
, was
(
(
calculated as the molar concentration of the indicated product
(
i) divided by the sum of the concentrations of all products
21–23
ca. 400 nm,
whereas the band at ca. 230–260 nm comes from
P
detected, S
i
¼ c
i
/
c ꢁ 100%. The following products detected
10
9 1
+
4
d /4d s transition characteristic of well-dispersed Ag
by FID and TCD detectors were analyzed: methanol, formalde-
hyde, methyl formate, dimethoxymethane, dimethyl ether, and
carbon dioxide. The product distribution was illustrated by
selectivities.
d+
24–26
cations and/or small Ag
n
clusters.
The spectra of Ag/S
display two broad bands: at ca. 245 nm and ca. 370 nm. The
rst one seems to have two components at ca. 235 nm and ca.
54 nm, both assigned to silver cations in silver oxide. The

2
second one is characteristic of metallic silver. All of them are
slightly blue-shied in comparison with literature data but it
can be caused by many factors such as interaction with the
support, or size and shape of particles. Although the spectrum
3
. Results and discussion
3.1. Composition and structure of catalysts
To conrm typical structure of SBA-15, pure support and silica of monometallic platinum sample is shown additionally in
modied with metals were examined by X-ray diffraction in expanded scale (because of a very low content of platinum
small-angles range. Fig. S1-ESI† shows the XRD patterns of the modier) (Fig. 1) it still does not show any UV-vis bands. The
prepared support and metallic samples. Three distinct peaks spectra of bimetallic samples AgPt/S and Pt/Ag/S display a band
+
24–26
assigned to the (100), (110) and (200) reections were observed at 229 nm from Ag cations.
There is also a broad band with
ꢂ
ꢂ
in the 2q ranges between 0.5 and 2 , which are characteristic of the center at ca. 330 nm. It can cover three bands: from metallic
SBA-15 with p6mm 2D-hexagonal symmetry. Thus, modication silver (ca. 400 nm), from silver Ag
d+
n
clusters (ca. 260 nm) and
14572 | RSC Adv., 2020, 10, 14570–14580
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Table 1 The chemical composition and texture parameters of studied samples
Ag/Pt molar ratio
-assumed/real
BET surface area
Total pore volume,
Average pore diameter,
DFT (nm)
a
2
ꢀ1
3
ꢀ1
Catalyst
Pt [wt%]
Ag [wt%]
(m g
)
DFT (cm g
)
S
Pt/S
Ag/S
AgPt/S
Pt/Ag/S
—
0.05
—
0.04
0.05
0.50
—
—
785
532
433
457
416
—
0.94
0.68
0.65
0.69
0.67
—
10.1
9.4
9.8
9.4
9.3
—
2.00
1.90
1.74
1.1
—
36.0/85.9
36.0/63.0
7.2/4.2
b
0
.5Pt2Ag/M
a
b
8
S stands for SBA-15 support. From our previous work, M stands for MCF support.
4
+
2+
from cationic platinum species Pt (ca. 270 nm) or Pt (ca. 380
and 417 nm). Thus, on the basis of UV-vis spectroscopy only it is
not possible to unambiguously identify the metal species
present in bimetallic AgPt/S and Pt/Ag/S samples.
Another technique which allows getting information on the
metals phases present in the samples is X-ray diffraction. XRD
patterns in the wide-angle range were examined to observe the
presence of metal and/or metal oxide crystallites of modiers.
As shown in Fig. 2 the only reections visible in XRD patterns
ꢂ
ꢂ
ꢂ
are located at 2q z 31.7 , 2q z 33.0 and 2q z 33.7 . They come
27
from Ag O (111), (120) and (031) planes, respectively. The
2
intensity of these reections is higher for bimetallic samples,
which suggests larger particles in these samples. It is worth
mentioning that X-ray diffraction has some limitations – it Fig. 2 Wide-angle XRD patterns of the samples calcined at 773 K for
permits detecting particles with diameters bigger than 1.5 nm 4 h.
and if their concentration is sufficiently high. Thus the lack of
any reection in XRD pattern of Pt/S can be caused by high
dispersion of platinum particles or rather by too low content of aer calcination but mainly aer activation which took place
this metal (0.05 wt%).
before methanol oxidation. The UV-vis spectra of the studied
3
.2.2 Catalysts aer activation at 673 K in argon ow (40 catalysts activated in argon ow at 673 K are collected in Fig. 3.
ꢀ
1
17
mL min at 673 K). In our previous studies it has been shown All spectra display a well separated band located at ca. 400 nm
0
23
that characterization of catalysts aer activation (673 K in argon coming from metallic silver Ag . As expected, the activation in
ow) is extremely important because the samples are activated inert gas caused the reduction of silver oxide species to metallic

before a catalytic reaction and their surface properties can be form. The reduction of silver cationic species upon activation
changed. High mobility of silver species as well as its suscep- was not total, as the broad band characteristic of these species
tibility to changes in oxidation state was the reason for segre- did not disappear completely. This band consists of two peaks,
gation of Pt–Ag alloy reported in ref. 14. Thus, this time we also one at ca. 229 nm and the other at ca. 263 nm, both assigned to
24–26
turned our attention to the states of metal modiers not only silver cations.
The position of surface plasmon resonance
Fig. 1 UV-vis spectra of the samples calcined at 773 K for 4 h.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 14570–14580 | 14573

View Article Online
RSC Advances
Paper
ꢀ1
Fig. 3 UV-vis spectra of the samples activated at 673 K in argon flow (40 mL min ) for 2 h.
band of metallic silver is worth of discussion. Monometallic about the oxidation states of metals on the silica surface, the
silver sample exhibits this band at 428 nm, whereas a blue shi XPS spectra of the samples aer activation were recorded. In the
(to 419 nm for AgPt/S–Ar and 408 nm for Pt/Ag/S–Ar) is visible spectrum of Ag/S–Ar (Fig. 6A), two well dened bands assigned
for bimetallic materials. It is clear that addition of platinum to to metallic Ag 3d5/2 (binding energy (BE) 368.4 eV) and Ag 3d3/2
silver changes the band position indicating a synergistic effect (BE 374.4 eV) indicated that silver species on the surface
0
31–33
between metallic silver and platinum species. This effect was occurred as Ag species.
The XPS spectra of the bimetallic
larger for bimetallic catalysts prepared by the step-by-step catalysts display the Ag 3d bands in exactly the same positions
method because platinum introduced aer silver loading on as those of the monometallic sample. On the basis of literature
3
1–33
14
the support could partially covered silver species. The observed data
and our own studies, it has been concluded that the
0
blue shi of SPR band could originate from a decrease in metal ranges of binding energies of Ag 3d5/2 reported for Ag , Ag
2
O
particle size. However, the analysis of TEM images discussed and AgO overlapped. Thus, to check if there are some metal
below excludes such a possibility. Thus, the blue shi comes oxides on silica surface the XP spectra in O 1s region were
from interactions between silver phase and platinum promoter. deconvoluted and examined (Fig. 6B). All materials display an
Careful look at the spectra of the platinum sample (Pt/S–Ar) intense band corresponding to Si–O–Si bonds in silica structure
reveals some changes in the spectra aer activation of the (binding energy 532.8 eV). In the spectra of Ag/S–Ar and AgPt/S–
sample in argon ow. The low-intensive broad band with the Ar the other band centered at ca. 530.2 eV is also visible. It
34
maximum at ca. 380 nm appeared, indicating the presence of comes from the oxygen in Ag–O–Ag bonds. The band at
2
+
28
partially reduced Pt species. It suggests that upon activation 531.9 eV visible in Pt/Ag/S–Ar spectrum can be assigned to the
4
+
2+
35
Pt cations accept electrons and Pt is formed. It may limits oxygen from surface hydroxyl groups or the subsurface oxygen
34
reduction of silver oxide species. Transmission electron in silver species. It is worth to note that the band coming from
microscopy images shown in Fig. S3-ESI† allow determination Ag–O–Ag bonds is much larger in bimetallic catalysts compared
of metal particle size distribution in the samples activated in to Ag/S–Ar catalyst. These results are in accordance with XRD
argon ow (Fig. 4). It is well visible that in bimetallic samples studies of activated samples and conrm that the reduction of
the average particle size of metal species signicantly increased silver upon activation was inhibited if platinum was added to
in comparison with the size of silver nanoparticles (NPs) in the sample.
monometallic sample. It is also important to stress that in
To sum up: on the basis of X-ray diffraction, UV-visible and
monometallic silver containing catalyst the contribution of very XP spectroscopy of catalysts activated in argon ow we can
small Ag NPs is almost 50%, whereas the share of these NPs in conclude that silver in all samples was present in both forms:
0
bimetallic samples is signicantly lower. The XRD patterns of Ag and Ag
2
O, however the reduction of silver oxide was deeper
activated samples (Fig. 5) conrmed the results from UV-vis in monometallic silver catalysts Ag/S. It is not possible to say
spectroscopy and TEM analysis. Two additional reections at explicitly that platinum was in the metallic or oxide form
ꢂ ꢂ
2
q z 38.1 and 2q z 44.1 conrmed the presence of metallic because of very low concentration of platinum in materials
29,30
silver crystallites.
No reections which could be assigned to (0.05 wt%), which did not allow their detection by the used
14
bimetallic Ag–Pt alloy-like crystallites appeared in the XRD techniques. The only form which was detected by UV-vis spec-
patterns. The appearance of reections from Ag is correlated troscopy was Pt species. In the activated samples there was no
0
2+
ꢂ
with reduced intensity of the reections at 2q z 33.7 and silver–platinum nanoalloy, but the blue shi in the UV-vis
ꢂ
2
q z 31.7 coming from silver oxide. In the XRD pattern of the spectra proves interactions between platinum promoter and
monometallic sample Ag/S–Ar the reections from oxide species silver phase.
disappeared totally, which suggests that the reduction of silver 3.2.3 Catalysts aer activation at 673 K in oxidizing
ꢀ
1
ꢀ1
oxide was deeper in this sample. It means that platinum species mixture (10 mL min of O + 30 mL min of Ar). For deeper
2
partially stabilized silver oxide species. To get more information understanding of cationic silver/silver oxide role in methanol
14574 | RSC Adv., 2020, 10, 14570–14580
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
Fig. 5 Wide-angle XRD patterns of the samples activated at 673 K in
ꢀ1
argon flow (40 mL min ) for 2 h.
24,25
species.
display a less intense band at ca. 370 nm from small metallic
silver clusters Ag . The spectrum of Pt/Ag/S–O synthesized by
2 2
Two samples (Ag/S–O and AgPt/S–O ) additionally
n
2
the step-by-step method differs from the other ones because it
shows only one band characteristic of cationic silver. Interest-
ingly, the spectrum of Pt/S–O2 is similar to that of Pt/S–Ar
(
Fig. 7) and shows a broad band with a maximum at 363 nm
2
+
suggesting the presence of cationic platinum species Pt ,
probably from PtO particles. The presence of silver oxide/
cationic silver clusters in activated samples was conrmed by
X-ray Diffraction (Fig. 8). From XRD patterns it is clearly seen
that silver oxide particles are bigger in bimetallic catalysts (very
intense reections) than in Ag/S–O
addition of very small amounts of platinum (0.05 wt%) to silver
2.0 wt%) catalyst promotes the formation of larger silver oxide
species.
2
sample. It indicates that
(
3
.3. Methanol oxidation
The results of methanol oxidation in the ow system at 423 K,
73 K and 523 K are displayed in Fig. 9. Despite a very low
4
content of platinum, Pt/S–Ar catalyst was very active even
below 423 K and led to total oxidation of methanol to carbon
dioxide. The other materials were not very active (below 10% of
methanol conversion) at 423 K but their activity rapidly
increased at 473 K. With increasing methanol conversion over
Ag/S–Ar a drastic decrease in selectivity of the reaction was
observed and the total oxidation of methanol was the only
reaction path at 523 K. Very high activity of Ag/S–Ar sample can
be correlated with a large amount of very small Ag NPs (ca. 50%
of NPs were 1–3 nm) responsible for chemisorption of oxygen
and its activity for participation in total oxidation of methanol.
Interestingly, although bimetallic materials also ensure high
conversion of methanol at 473 K, they appeared to be selective
to methyl formate (61% selectivity on AgPt/S–Ar and 69% on Pt/
Ag/S–Ar). What is more, the increase in the reaction tempera-
ture to 523 K led to 98% conversion of methanol and 75%
selectivity to methyl formate when Pt/Ag/S sample was used.
Fig. 4 Metal particle sizes distribution on the support after activation
at 673 K in Ar flow (40 mL min ) for 2 h (basing on TEM images and
measuring of 250 crystallites in each sample).
ꢀ1
oxidation (results shown in the next section) the materials were
studied aer treatment in oxidizing mixture. As seen in Fig. 7
the UV-vis spectra of activated Ag/S–O , AgPt/S–O and Pt/Ag/S–
2
2
2
O samples display the main band, which covers both peaks at
ca. 230 nm and ca. 260 nm coming from cationic silver
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 14570–14580 | 14575

View Article Online
RSC Advances
Paper
ꢀ1
Fig. 6 XPS spectra (Ag 3d (A) and O 1s regions (B)) for Ag/S, AgPt/S and Pt/Ag/S catalysts activated at 673 K in argon flow (40 mL min ) for 2 h.
These data show that the composition of active species present species. In the sample prepared by modication with silver
on the surface of the catalyst prepared by the step-by-step and platinum using the one-step method the interaction
modication favors selective oxidation of methanol to methyl between both metals was weaker, as evidenced by a smaller
formate. This composition consists of silver oxide, metallic blue shi in the UV-vis spectra and it resulted in a lower
silver and platinum promoter which interacts with silver selectivity to methyl formate. It seems that addition of small
14576 | RSC Adv., 2020, 10, 14570–14580
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
ꢀ1
ꢀ1
Fig. 7 UV-vis spectra of the catalysts activated at 673 K in oxidative mixture flow (10 mL min
2
O + 30 mL min Ar) for 2 h.
above mentioned hypothesis, methanol oxidation was per-
formed on the catalysts activated in oxidizing mixture. Such
activation enhanced the amount of silver oxide species as
shown in Fig. 7 and 8 (described above). As revealed in Fig. 10,
the activation of platinum catalyst in oxygen ow did not
change the activity and selectivity of this catalyst. However,
this activation treatment of Ag/S sample signicantly
decreased its activity and increased its selectivity (100% and
9
5% selectivity to methyl formate at 473 and 523 K, respec-
tively) in methanol oxidation. Thus, evidently silver oxide is
responsible for activation of the reaction pathway towards
methyl formate generation, but it is not responsible for high
activity. For higher activity metallic silver and/or platinum are
required. Thus, platinum not only acts as the promoter, as
mentioned above, but also can provide active centers for acti-
Fig. 8 Wide-angle XRD patterns of the samples activated at 673 K in vation of adsorbed oxygen. In the absence of silver in the
ꢀ1
ꢀ1
oxidative mixture flow (10 mL min
2
O + 30 mL min Ar) for 2 h.
catalyst, platinum catalyzed total oxidation of methanol.
Basing on presented results some general conclusion and
hypothesis of what is happening in catalysts can be proposed.
amounts of platinum (as the promoter) to silver catalyst results In combination with silver species, platinum activated the
d+
in higher resistance of cationic silver (Agn , Ag O) to reduction chemisorbed oxygen for further reaction with intermediates
2
to metallic species. Unreduced silver species are responsible chemisorbed on silver oxide. Silver oxide NPs were active
for high selectivity of bimetallic Ag & Pt catalysts. To check the centers for chemisorption of formaldehyde (the rst product of
ꢀ
1
Fig. 9 Results of methanol oxidation at 423 K, 473 K and 523 K over the catalysts activated at 673 K in Ar flow (40 mL min ) for 2 h; FA –
formaldehyde, MF – methyl formate.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 14570–14580 | 14577

View Article Online
RSC Advances
Paper
ꢀ1
Fig. 10 Results of methanol oxidation at 473 K and 523 K over the catalysts activated at 673 K in oxidative mixture flow (10 mL min
mL min ) for 2 h; DIE – dimethyl ether, DMM – dimethoxymethane, FA – formaldehyde, MF – methyl formate.
2
O + 30
ꢀ1
Table 2 The selectivity and yield of methyl formate (MF) in methanol the reaction temperature caused easier desorption of formal-
8
oxidation over Pt/Ag/S and 0.5Pt2AgPt/M (from previous studies ).
dehyde and therefore the selectivity to formaldehyde increased
ꢀ
1
Both samples were activated at 673 K in Ar flow (40 mL min ), for 2 h
before the reaction
and that to methyl formate decreased at 523 K. It was especially
pronounced for bimetallic samples (AgPt/S–O and Pt/Ag/S–
2
O ). The question is if doping of silver with a small amount of
2
Methanol conv., Selectivity Yield of
Catalyst
Temp., K
423
%
to MF
MF
platinum (shown in this work) would give as efficient catalyst
as Ag–Pt alloy containing bimetallic catalysts from our
Pt/Ag/S–Ar
6
100
69
75
59
73
17
6.0
54
74
59
73
13
1
4,17
previous studies.
For comparison purposes the sample
4
5
73
23
78
98
100
100
74
which displayed the best catalytic properties in our previous
studies, labelled as 0.5Pt2Ag/M, was taken into account. This
sample contained 0.5 wt% of Pt and 1.6 wt% of Ag (Table 1.)
which formed bimetallic alloy crystallites loaded on meso-
porous silica, MCF type. The yield of methyl formate and the
stability of the samples Pt/Ag/S–Ar and 0.5Pt2Ag/M–Ar were
considered. As seen in Table 2. Pt/Ag/S–Ar displayed 74% yield
of methyl formate at 523 K, whereas the sample from our
previous work reached a similar yield at 423 K. Thus, one can
conclude that Ag–Pt alloy particles are more active in methyl
a
0
.5Pt2Ag/M–Ar
373
4
4
23
73
a
8
Sample from previous work.
methanol oxidation chemisorbed on Lewis acid sites as evi-
denced in ref. 3) and its further reaction with the second
methanol molecule towards methyl formate. The increase in
8
Fig. 11 Results of methanol oxidation over 0.5Pt2Ag/M at 423 K (from paper ) (A) and Pt/Ag/S at 523 K (B); MF – methyl formate, FA – form-
ꢀ1
aldehyde. Both samples were activated before the reaction for 2 h in ar flow (40 mL min ) at 673 K.
14578 | RSC Adv., 2020, 10, 14570–14580
This journal is © The Royal Society of Chemistry 2020

View Article Online
Paper
RSC Advances
formate production because the same yield of this product was
obtained at a lower temperature. Bearing in mind the easy
Acknowledgements
segregation of Ag–Pt alloy structure, the reaction at 423 K over National Science Centre in Poland (Grant No. UMO-2018/29/B/
.5Pt2Ag/M–Ar was performed for a longer time and the results ST5/00137) is acknowledged for the partial nancial support
0
are shown in Fig. 11A. Aer 3 hours of the reaction, a decrease in of this work.
both, methanol conversion and methyl formate selectivity were
noted. Especially signicant changes were in the selectivity to
methyl formate (decrease from 73% to 40%). The increase in the
References
reaction temperature considerably decreased the methyl formate
yield (Table 2). The same test of stability but at 523 K was per-
formed over Pt/Ag/S–Ar catalyst. As shown in Fig. 11B methanol
conversion did not change during the reaction time and reached
the value of ca. 100%. The selectivity to methyl formate was in the
range 63–75% and there were no signicant changes during the
reaction time. The UV-vis spectra of Pt/Ag/S–Ar aer stability test,
shown in Fig. S4-ESI† conrm the stability of the material and the
lack of changes in silver active species. Thus, the Pt/Ag/S catalyst
prepared within this work, containing silver doped with platinum
species, displays even better catalytic properties than the mate-
rials containing Pt & Ag alloy-like nanoparticles. On the one
hand, higher temperature of methanol oxidation on Pt/Ag/S–Ar is
required (523 K instead of 423 K) for sufficient methyl formate
selectivity but on the other hand, very good catalytic properties
are preserved for a longer time. What is more, in this study the
1 C. H. Bartholomew and R. J. Farrauto, Fundamentals of
Industrial Catalytic Processes, John Wiley & Sons, 2nd edn,
2010, vol. 8, pp. 560–634.
2 J. M. Tatibouet, Appl. Catal., A, 1997, 148, 213–252.
3 G. Busca, Catal. Today, 1996, 27, 457–496.
4 B. N. Pattanaik, Int. J. Chem. Petrochem. Technol., 2013, 3, 55–
70.
5 J. S. Lee, J. C. Kim and Y. G. Kim, Appl. Catal., 1990, 57, 1–30.
6 J.-P. Lange, Catal. Today, 1980, 64, 3–8.
7 R. P. Aplin, Process for the production of methyl formate,
Eur. Pat., EP0189283A1, 1985.
8 J. G. Firth, Trans. Faraday Soc., 1970, 67, 212–215.
9 L. R. Merte, M. Ahmadi, F. Behafarid, L. K. Ono, E. Lira,
J. Matos, L. Li, J. C. Yang and B. R. Cuenya, ACS Catal.,
2013, 3, 1460–1468.
amount of platinum in catalyst was 10 times lower, which 10 M. P. Zum Mallen and L. D. Schmidt, J. Catal., 1996, 161,
signicantly reduced the cost of the material.
230–246.
11 H. Wang, Y. Wang, Z. Zhu, A. Sapi, K. An, G. Kennedy,
W. D. Michalak and G. A. Somorjai, Nano Lett., 2013, 13,
4
. Conclusions
2
976–2979.
1
2 H. Wang, K. An, A. Sapi, F. Liu and G. A. Somorjai, Catal.
Lett., 2014, 144, 1930–1938.
3 W. Tu and Y. H. Chin, J. Catal., 2014, 313, 55–69.
4 J. Wisniewska and M. Ziolek, RSC Adv., 2017, 7, 9534–9544.
5 J. Wisniewska, M. Ziolek, N. Artioli and M. Daturi, J. Catal.,
The results presented in this work not only allow construction
of effective bimetallic (Ag & Pt) catalyst for production of
methyl formate from methanol but also contributed to better
understanding of the role of each component in this bimetallic
catalyst in activity and selectivity in methanol oxidation.
1
1
1
2
016, 336, 58–74.
6 J. Czaplinska, P. Decyk and M. Ziolek, Appl. Catal., A, 2015,
04, 361–372.
7 J. Wisniewska, C. M. Yang and M. Ziolek, Catal. Today, 2019,
33, 89–96.
8 D. Zhao, J. Sun, Q. Li and G. D. Stucky, Chem. Mater., 2000,
2, 275–279.
9 J. Czaplinska, I. Sobczak and M. Ziolek, J. Phys. Chem. C,
014, 118, 12796–12810.
0
Metallic silver Ag was found to be responsible for chemi-
1
1
1
1
2
2
2
2
2
sorption of oxygen, while the platinum active phase took part
in activation of adsorbed oxygen, too. Platinum species also
acted as an electronic promoter and preserved total reduction
of silver oxide upon activation and reaction conditions. This
role was crucial because silver oxide was responsible for
selective formation of methyl formate (via chemisorption of
formaldehyde on Lewis acid sites). Thus, addition of small
5
3
1
2
amount of platinum (as the promoter) to silver catalyst resulted
in a higher resistance of cationic silver (Ag
0 J. Wisniewska, H. Guesmi, M. Ziolek and F. Tielens, J. Alloys
Compd., 2019, 770, 934–941.
1 T. Benk ´o , A. Beck, K. Frey, D. F. Srank ´o , O. Geszti, G. S ´a fr ´a n,
B. Mar ´o ti and Z. Schay, Appl. Catal., A, 2014, 479, 103–111.
2 D. Vargas and E. Guilleminot, Appl. Catal., B, 2017, 216, 1–
d+
n 2
, Ag O) to reduc-
tion and increased the activity and selectivity in oxidation of
methanol to methyl formate. Interaction between platinum
promoter and silver species ensured the best catalytic proper-
ties (activity, selectivity, stability) and such properties were
observed for the sample prepared by the step-by-step method
10.
3 L. Suber, I. Sondi, E. Matijevi ´c and D. V. Goia, J. Colloid
Interface Sci., 2005, 288, 489–495.
4 P. M. More, D. L. Nguyen, M. K. Dongare, S. B. Umbarkar,
N. Nuns, J. S. Girardon, C. Dujardin, C. Lancelot,
A. S. Mamede and P. Granger, Appl. Catal., B, 2015, 162,
(Pt/Ag/S). The catalyst obtained in this method displayed 74%
yield of methyl formate production at 523 K and it was stable in
the reaction time.
1
1–20.
Conflicts of interest
2
5 S. Miao, Y. Wang, D. Ma, Q. Zhu, S. Zhou, L. Su, D. Tan and
There are no conicts to declare.
X. Bao, J. Phys. Chem. B, 2004, 108, 17866–17871.
This journal is © The Royal Society of Chemistry 2020
RSC Adv., 2020, 10, 14570–14580 | 14579

View Article Online
RSC Advances
Paper
26 D. Chen, Z. Qu, S. Shen, X. Li, Y. Shi, Y. Wang, Q. Fu and 31 A. M. Ferraria, A. P. Carapeto and A. M. Botelho Do Rego,
J. Wu, Catal. Today, 2011, 175, 338–345. Vacuum, 2012, 86, 1988–1991.
2
7 G. Valverde-Aguilar, J. Garc ´ı a-Macedo and V. Renter ´ı a-Tapia, 32 G. B. Hound, Z. F. Hazos and G. N. Salaita, Phys. Rev. B,
J. Nano Res., 2008, 3, 103–114. 2000, 62, 126–133.
28 C. S. S. R. Kumar, UV-VIS and photoluminescence spectroscopy 33 V. K. Kaushik, J. Electron Spectrosc. Relat. Phenom., 1991, 56,
for nanomaterials characterization, Springer Berlin
Heidelberg, Berlin, 2013.
9 H. Granbohm, J. Larismaa, S. Ali, L.-S. Johansson and
S.-P. Hannula, Materials, 2018, 11, 1–17.
273–277.
34 N. J. Firet, M. A. Blommaert, T. Burdyny, A. Venugopal,
D. Bohra, A. Longo and W. A. Smith, J. Mater. Chem. A,
2019, 7, 2597–2607.
2
30 X. Huang, M. Yang, G. Wang and X. Zhang, Microporous 35 L. Wolski, J. E. Whitten, I. Sobczak and M. Ziolek, Mater. Res.
Mesoporous Mater., 2011, 144, 171–175. Bull., 2016, 85, 35–46.
14580 | RSC Adv., 2020, 10, 14570–14580
This journal is © The Royal Society of Chemistry 2020