Enhancement of solid base activity for porous boron nitride catalysts by controlling active structure using post treatment

Article abstract of DOI:10.1016/j.apcata.2020.117843

Porous boron nitride (BN) was synthesized using a pyrolysis method in conjunction with varying NH₃ flow rates, followed by washing as a post-treatment. The performance of this material as a solid base catalyst was assessed. It was found that th

Full text of DOI:10.1016/j.apcata.2020.117843

Applied Catalysis A, General 608 (2020) 117843
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Enhancement of solid base activity for porous boron nitride catalysts by
controlling active structure using post treatment
a,b,
c
a,b,c
b
Atsushi Takagaki *, Shohei Nakamura , Motonori Watanabe
, Yoonyoung Kim ,
a,b
d
e
e,f
d
Jun Tae Song , Keiko Jimura , Kanta Yamada , Masaaki Yoshida , Shigenobu Hayashi ,
a,b,c,
Tatsumi Ishihara
*
a
Department of Applied Chemistry, Faculty of Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
b
c
2
International Institute for Carbon-Neutral Energy Research (WPI-I CNER), Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
Department of Automotive Science, Graduate School of Integrated Frontier Sciences, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
Research Institute for Material and Chemical Measurement, National Institute of Advanced Industrial Science and Technology (AIST), Central 5, 1-1-1 Higashi, Tsukuba,
d
Ibaraki 305-8565, Japan
e
f
Graduate School of Sciences and Technology for Innovation, Yamaguchi University, Tokiwadai, Ube, Yamaguchi 755-8611, Japan
Blue Energy Center for SGE Technology (BEST), Yamaguchi University, Tokiwadai, Ube, Yamaguchi 755-8611, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
Porous boron nitride (BN) was synthesized using a pyrolysis method in conjunction with varying NH flow rates,
3
Boron nitride
followed by washing as a post-treatment. The performance of this material as a solid base catalyst was assessed.
It was found that the post-treatment rather than the synthesis conditions significantly improved the activity of
the BN during the nitroaldol reaction. The BN catalyst obtained after washing gave a ten times higher product
yield. Both an increase in the surface areas of the material and the emergence of micropores after washing were
Solid base catalyst
Nitroaldol reaction
1
1
B MAS NMR
B K-edge XAFS
1
1
observed. B solid-state nuclear magnetic resonance spectroscopy demonstrated that boron oxide species were
present in the BN after synthesis but were removed by the washing process. B K-edge X-ray absorption fine
structure analyses indicated the formation of oxygen-substituted trigonal boron sites on the BN surface. Fourier
transform infrared spectroscopy showed that amino groups on the samples functioned as moderately strong basic
sites.
1
. Introduction
[6], and the hydrogenation of crotonaldehyde [5] and 2,5-di-
methylfuran [4]. BN has also been used in heterogeneous catalysis as a
Hexagonal boron nitride (BN) is a layered material that is iso-
blocking layer to stabilize catalytically active metal nanoparticles while
avoiding the agglomeration of such particles that can be caused by
carbon deposition [12,13]. In addition, covering the surfaces of metal
particles with the two-dimensional layer of BN can weaken reactant-
metal interactions, resulting in enhancement of metal-catalyzed reac-
tions such as CO [14] or hydrogen oxidation [15].
structural to graphite, is inert, and exhibits significant chemical and
thermal stability while being physically robust. For these reasons, this
compound is often used as a support in the field of catalysis. The high
thermodynamic stability of BN is beneficial with regard to prolonged
reactions such as the synthesis of NH using Ba-Ru/BN [1] and the
3
oxidation of volatile organic compounds over Pt/BN [2]. In addition,
the chemical inertness and weak metal-support interactions associated
with BN are advantageous because these allow catalytic metal particles
to remain in a reduced state [3] while permitting particle migration [4]
and facilitating the formation of bimetallic clusters of materials such as
PtFe/BN [5] and Rh-Ni/BN [6]. Owing to these properties, BN -sup-
ported metal catalysts have been applied to many reactions, including
CO oxidation [7], hydrocarbon combustion [8,9], selective alcohol
Interestingly, hexagonal BN has recently been reported to act as an
efficient heterogeneous catalyst on its own, based on the utilization of
its unique surface structure [16,17]. BN had been found to exhibit re-
markable catalytic activity during the selective oxidative dehy-
drogenation of light alkanes, including propane [18,19], ethane
[20–22], n-butane and isobutane [23], to the corresponding alkenes
with excellent selectivity. BN has also been shown to effectively pro-
mote methane activation for the production of valuable chemicals [24].
While a variety of active sites have been proposed for BN to date,
oxidation [10], CO methanation [11], the dry reforming of methane
2
⁎
Corresponding authors.
E-mail addresses: atakagak@cstf.kyushu-u.ac.jp (A. Takagaki), ishihara@cstf.kyushu-u.ac.jp (T. Ishihara).
https://doi.org/10.1016/j.apcata.2020.117843
Received 1 July 2020; Received in revised form 18 September 2020; Accepted 20 September 2020
Available online 22 September 2020
0926-860X/ © 2020 Elsevier B.V. All rights reserved.

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
oxidized boron species (BOx) on the BN surface are believed to be re-
sponsible for the oxidation activity of this material [25]. The surface
functionalization of BN also permits the use of this compound as a
photocatalyst [26] or a solid base catalyst [27,28], and theoretical
calculations have indicated that the surface functionalization of BN
sheets with hydroxyl or amino groups decreases its bandgap [29]. This
has been confirmed experimentally and has been shown to increase the
photocatalytic activity of BN during the hydrogen evolution reaction
(BELSORP-max-32 and BELLSORP mini-II, Microtrac-BEL). Each sample
was pretreated by heating under vacuum at 473 K for 2 h prior to
analysis. CO adsorption was also determined at 298 K to assess the
2
micropores in the samples. The morphologies of the materials were
examined using scanning electron microscopy (SEM, JSM-7900F, JEOL)
and transmission electron microscopy (TEM, JEM-ARM200, JEOL). The
elemental analysis of the materials was investigated using energy dis-
N
persive X-ray spectroscopy (EDS, X-Max 80, OXFORD).
[
26]. These functional groups can also be generated on BN surfaces by
The structures of the sample were investigated by solid-state NMR
spectroscopy, XAFS analysis, and X-ray photoelectron spectroscopy.
ball-milling and can serve as active sites for various liquid phase base-
catalyzed reactions [27,28].
1
1
Boron-11 magic-angle spinning ( B MAS) NMR spectra were acquired
with a Bruker Avance III HD 600WB spectrometer operating at a fre-
quency of 192.63 MHz. A Bruker MAS probe head was used with a
zirconia rotor having an outer diameter of 4 mm. The measurement was
The surface functionalization and exposed surface area of BN both
play crucial roles in its utilization as a heterogeneous catalyst. Porous
BN having high surface areas can be synthesized via a pyrolysis method.
A variety of porous BN, including mesoporous BN using amphiphilic
block copolymers, have been reported [30]. In the present study,
porous BN samples were fabricated via a pyrolysis method in con-
1
1
1
performed at room temperature with a spinning rate of 12 kHz. The
spectra were obtained with a single pulse sequence combined with
B
H
high power decoupling during signal acquisition. The flip angle of the
pulse and the recycle delay were π/8 for solution and 10 s, respectively.
The shift was expressed with respect to neat boron trifluoride diethyl
junction with various NH flow rates and subsequently examined as
3
solid base catalysts. Analyses of the surface structures, as well as the
bulk structures of these materials, were carried out using a variety of
techniques, including Fourier transform infrared spectroscopy (FTIR), B
3
etherate (CF ·(C
2
H
5
)
2
O) by setting the NaBH signal acquired while
4
spinning at 8 kHz to −42.00 ppm. B K-edge XAFS spectra were ob-
tained at the BL7A beamline of the Photon Factory at the Institute of
Materials Structure Science, High-Energy Accelerator Research
Organization (KEK-IMSS-PF). The spectra were collected using the total
electron yield method with samples mounted on carbon tape, and all
1
1
K-edge X-ray absorption fine structure (XAFS) analysis, and B solid-
state nuclear magnetic resonance (NMR) spectroscopy. A post-treat-
ment involving washing of the material, as opposed to the actual
synthesis conditions, was found to significantly improve the solid base
activity of this BN during the nitroaldol reaction. Specifically, this post-
treatment increased the yield by a factor of ten, even though X-ray
diffraction (XRD) and FTIR analyses did not indicate any changes in the
samples. In contrast, XAFS and solid-state NMR spectra indicated the
removal of BOx species after washing, followed by an increased surface
area and the emergence of micropores. These results clearly demon-
strated that the post-treatment exposed active sites on the BN surface by
removing BOx species, resulting in an enhanced solid base activity. It
should be noted that not BOx species but amino groups were active sites
for the base-catalyzed reaction while BOx species are considered to be
active sites for the oxidation reaction.
spectra were calibrated with reference to the B
2
O peak at 194.0 eV.
3
The oxidation state of elements over the surface was analyzed by XPS
(KRATOS Ultra 2, Shimadzu). The binding energies in each measure-
ment were referenced to the core level of the C1s peak (284.8 eV).
The solid base strength of the sample was determined by acquiring
FTIR spectra following CHCl adsorption, employing a mercury cad-
3
−1
mium telluride detector with a resolution of 4 cm . The sample was
pressed into a disk with a radius of 1.0 cm without KBr and pretreated
in the measurement cell at 473 K under vacuum for 1 h. The sample was
then cooled to 303 K, and a spectrum was recorded. Following this,
gaseous CHCl
3
was introduced into the cell, and the sample was ex-
posed to a saturated CHCl
3
atmosphere for 30 min. Following the re-
2
. Experimental
moval of CHCl from the cell over a span of 30 min, the spectrum of the
3
sample was acquired, and a difference spectrum was obtained by sub-
2.1. Catalyst preparation
tracting the spectrum of the original material.
The BN samples were synthesized using boric acid (Kishida, 99.5 %)
2.3. Nitroaldol reaction
and hexamethylenetetramine (HMTA) (Wako, 99.0 %) according to a
literature procedure [31]. Briefly, 20 mmol of boric acid and 40 mmol
of HMTA were dissolved in 100 mL of water. After evaporating the
solution, the resulting white precursor powder was transferred to an
The catalytic activity of each material as a solid base was de-
termined using the nitroaldol reaction. Reactions were conducted using
50 mg of the catalyst in 2 mL of a toluene solution containing 60.5 μL of
p-methoxybenzaldehyde (0.5 mmol), 68.0 μL of nitromethane
(1.25 mmol) and 19.5 μL of n-decane (0.1 mmol) in a glass reactor
vessel (ACE glass). The reaction was performed at 373 K for 8 h in an oil
bath under stirring. Aliquots of the solution were removed using a
syringe and analyzed by gas chromatography (GC) with a flame ioni-
zation detector (GC-2014, DB-1MS column, Shimadzu). For the quan-
tification of the reactant and products, n-decane was used as an internal
standard, and each calibration curve was made in advance. For the
qualification of products, gas chromatography-mass spectrometry
alumina boat and heated at 1273 K for 3 h in an NH flow (50, 100 or
3
3
−1
1
50 cm (NTP) min ). The brown powder obtained from this proce-
dure was ground using a mortar, then washed by stirring in water
overnight at 333 K. The sample was subsequently collected by cen-
trifugation or filtration, after which it was dried at 353 K in an oven.
The synthesized BN is denoted herein as BNX-Y, where X represents the
3
−1
flow rate of NH (50, 100 or 150 cm (NTP) min ), and Y denotes the
3
method used to collect the sample (no Y: without washing, C: cen-
trifugation and F: filtration).
(
GC–MS, GCMS-QP2010Ultra, Shimadzu) was used. The measurement
2
.2. Characterization
error was within 5%. For example, we tested the activity of BN50-F
three times in which a product yield was 32.0, 31.6, and 32.3 %. This
difference is included in the significant figures. For the recyclability
test, the catalysts were collected by centrifugation at 3000 rpm, washed
with 10 mL of toluene three times, dried overnight, and reused for
further reactions.
The crystal structure of the BN was investigated by XRD (RINT-
2
500HLR+, Rigaku) with Cu Kα radiation (λ = 0.15418 nm), oper-
ating at a voltage of 40 kV and a current of 80 mA. Scans were obtained
−
1
at a rate of 5° min with a step width of 0.05° for 2θ values of 10° to
7
0°. The surface functional groups of the samples were characterized by
FTIR spectroscopy (FT/IR-6600, JASCO) after pressing the materials
into pellets with KBr. The surface areas were evaluated by the
Brunauer-Emmett-Teller (BET) method using nitrogen adsorption
3. Results and discussion
Fig. 1 shows XRD patterns and FTIR spectra obtained from the as-
2

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
Fig. 1. (a) XRD patterns and (b) FTIR spectra obtained from as-synthesized BN samples prepared by heating mixtures of boric acid and hexamethylenetetramine in
NH flow at various flow rates.
3
synthesized BN samples generated from boric acid and HMTA under the
NH atmosphere at various flow rates. All samples produced two broad
3
diffraction peaks at approximately 25° and 42.5° that are attributed to
the (002) and (110) reflections of turbostratic BN [32]. With decreases
3
−1
in the NH flow rate from 150 to 50 cm min , these peaks became
3
narrower. This result indicates that a lower flow rate facilitated the
growth of the BN structure by increasing the contact time between the
precursors and the NH . The FTIR spectra also confirm the formation of
3
−1
BN. Characteristic absorption bands are evident at 1372 and 816 cm
,
ascribed to B–N stretching and B–N–B bending, respectively [31,33,34].
In addition, a broad absorption peak is present between 3400 and
−
1
3
200 cm , indicating the formation of amino and hydroxyl groups on
the surface of the material [35]. The BET surface areas (SBET) of the as-
2
−1
made samples were relatively low at 155, 208, and 171 m g
for
Fig. 2. Results of the nitroaldol reaction using the as-synthesized boron nitride.
Reaction conditions: p-methoxybenzaldehyde (1) (0.5 mmol), nitromethane (2)
BN50, BN100, and BN150, respectively.
Solid base catalytic activity was evaluated using the nitroaldol re-
action as a test reaction. In the present study, the reaction of p-meth-
oxybenzaldehyde (1 in Scheme 1) with nitromethane (2) gave p-
methoxy-β-nitrostyrene (3). The subsequent Michael addition reaction
of 3 with nitromethane then afforded the dinitro compound, 1-
methoxy-4-[2-nitro-1-(nitromethyl)ethyl]benzene (4). The main pro-
duct of this reaction is a nitroalkene, and these compounds have sig-
nificant synthetic potential in organic chemistry [36,37]. Because nitro
groups are strongly electron-withdrawing, nitroalkenes have applica-
tions as reagents in Michael addition [38], Friedel–Crafts alkylation
(
1.25 mmol), catalyst (50 mg), toluene (2 mL), 373 K, 8 h. Products were p-
methoxy-β-nitrostyrene (3) and 1-methoxy-4-[2-nitro-1-(nitromethyl)ethyl]
benzene (4).
suggesting that the as-synthesized samples contained surface im-
purities.
In a previous study, as-synthesized BN was washed with hot water
to remove possible boron oxide byproducts [31]. Consequently, in the
present work, the BN was washed by stirring overnight in water, fol-
lowed by recovery via centrifugation or filtration. Fig. 3 provides the
XRD patterns and FTIR spectra of the BN50-C, BN100-C, and BN150-C
samples, all of which were recovered by centrifugation. In addition to
[
39], and Morita–Baylis–Hillman reactions [40]. Nitroalkenes are also
important precursors for the production of pharmaceuticals [41], an-
tifungal drugs [42], and potential anticancer drugs [43].
the broad peaks related to BN, the samples prepared using higher NH
3
Fig. 2 shows the results of the nitroaldol reaction using the as-syn-
thesized BN, along with data for a commercial hexagonal BN and two
conventional solid bases (MgO and hydrotalcite) as comparisons. While
the highly crystalized commercial BN, which had a low surface area of
flow rates (the BN100-C and BN150-C) generated numerous new peaks
that are attributed to ammonium pentaborate tetrahydrate ((NH
·4H O, PDF #00-031-0043) [44]. The FTIR spectra of the BN100-C
and BN150-C also exhibit three absorption peaks between 1200 and
4
)
5
B O
8
2
2
−1
3
m g , was inactive, the as-synthesized BN50, BN100, and BN150 all
−
1
9
00 cm
that are ascribed to ammonium borate (Fig. 3(b)). After the
afforded the corresponding nitroalkene (3). The order of the activity of
these materials was BN100 > BN150 > BN50, which is the same as
that of the surface areas. In contrast, the MgO and hydrotalcite afforded
not only 3 but also 4 because of their strong basicity. Although the as-
synthesized BN showed comparable activity to those of the conven-
tional solid bases, the product yield was insufficient. This reduced ac-
tivity could be related to the low surface area of the material,
recovery of these samples by centrifugation, each sample was trans-
ferred to a dish and dried in an oven. However, it was necessary to add
a small amount of water when collecting the sample from a centrifuge
tube, and this water may have hydrolyzed the B–N bonds of the
amorphous BN to generate the ammonium borate. It is reported that
porous boron nitride [45,46] and boron carbon nitride [47] are slowly
Scheme 1. The nitroaldol reaction of p-meth-
oxybenzaldehyde (1) and nitromethane (2) to
p-methoxy-β-nitrostyrene (3). The subsequent
Michael addition of 3 with nitromethane (2)
affords 1-methoxy-4-[2-nitro-1-(nitromethyl)
ethyl]benzene (4).
3

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
Fig. 3. (a) XRD patterns and (b) FTIR spectra obtained from BN samples after washing and recovery by centrifugation. Filled circles indicate the formation of
NH ·4H O.
4
B
5
O
8
2
hydrolyzed by moisture in the air at room temperature for a long time.
Addition of water hydrolyzed boron nitride to form boric acid and
ammonia by the following reactions. Then, the condensation of boric
acid produced pentaborate.
and hydroxyl groups, suggesting the exposure of these active sites. The
FTIR spectra acquired from the BN100-F and BN150-F also exhibit
−
1
additional broad absorption bands ranging from 1200 to 900 cm that
−1
are attributed to B–O bending (1100 cm ) and B–N–O bending
−1
2
BN + 3H
2
O → B
O → 2H
+ NH → (NH
2
O
3
+ 2NH
3
(925 cm ) and are most likely due to the formation of a small amount
of ammonium borate. The BET surface areas of the samples recovered
B
2
H
O
3
BO
+ 3H
2
3
BO
3
)B
2
−1
5
3
3
3
4
5
O
8
·4H
2
O + 3H
2
O
by filtration were also increased, to 307, 414, and 274 m g
for the
The extent to which this hydrolysis proceeded would also have been
affected by the degree of crystallinity of the BN. As shown in Fig. 1,
BN50-F, BN100-F, and BN150-F, respectively. These values are ap-
proximately twice those of the as-synthesized samples. Comparing the
surface areas of BN50 and BN100 before washing, the BN100 had
higher surface areas which could be due to the lower crystallinity than
the BN50. The highest surface areas of BN100-F among all samples
could be derived from the BN100 having the highest surface areas
among the as-made samples.
with increases in the NH flow rate, the crystallinity of the boron nitride
3
decreased. Thus, the BN150 (which had the lowest crystallinity of the
samples) was easily hydrolyzed and produced the most ammonium
borate. The formation of ammonium borate significantly lowered the
surface area of the material. The BET surface areas of the samples re-
2
−1
covered by centrifugation were 3 m g
for both the BN100-C and
Fig. 5 compares the results of the nitroaldol reaction using BN be-
fore and after the washing treatment, applying the same reaction con-
ditions as detailed in the caption to Fig. 2. It is obvious that the washing
process drastically altered the activity of these materials. Specifically,
the yield of 3 obtained from the BN50 was significantly increased from
3% to 16 % by washing followed by centrifugation (BN50-C) and to 32
% by washing followed by filtration (BN50-F). However, the BN100-C
and BN150-C still showed relatively low activity after washing and
centrifugation. As noted above, the as-synthesized BN100 gave the
highest yield among the as-synthesized samples and showed compar-
able activity to other solid bases (Fig. 2). The yield of 3 provided by the
BN100 was also increased to 29 % after washing, followed by filtration
(BN100-F). All the samples provided improved yields following
washing and recovery by filtration. Table 1 summarizes the results of
the nitroaldol reaction. These data indicate that the BN50-F and BN150-
F afforded the highest yields of 3 along with good selectivity (78 and 59
BN150-C. The BN50-C, which did not contain an ammonium borate
2
−1
phase, had a surface area of 345 m g that exceeded the value for the
as-synthesized BN50.
Fig. 4 shows the XRD patterns and FTIR spectra obtained from the
BN samples recovered by filtration (the BN50-F, BN100-F, and BN150-
F). In contrast to the results for the materials recovered using cen-
trifugation, the XRD patterns show that all samples consisted of a tur-
bostratic BN phase (Fig. 4(a)) and no ammonium borate was present.
The difference between the outcomes obtained from centrifugation and
filtration is ascribed to the presence of water during the drying process.
Specifically, after filtration, no additional water was added to the
samples before drying, which limited hydrolysis to form ammonium
−
1
borate. The FTIR spectra of all samples show a clear peak at 816 cm
attributed to the B–N–B bending mode of BN (Fig.4(b)). Compared with
the BN50, the BN50-F generated distinct absorption peaks due to amino
Fig. 4. (a) XRD patterns and (b) FTIR spectra obtained from BN samples after washing and recovery by filtration.
4

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
shows the reusability of the BN50-F and hydrotalcite. These catalysts
were centrifuged, washed with toluene, dried overnight, and recycled
for further reactions. Compared to hydrotalcite, which showed a sig-
nificant decrease in the activity (Fig. 6(b)), the BN50-F catalyst could
show better reusability (Fig. 6(a)).
Because the BN50-F exhibited high activity and significant changes
in activity were also identified in tests using the BN50, BN50-C, and
BN50-F, the morphologies and bulk and surface structures of these
materials were investigated. Fig. 7 provides SEM images of the BN50,
BN50-C, and BN50-F showing the presence of small particles less than
3
0 nm in size that have aggregated in all samples. These results are in
good agreement with the XRD patterns, which show broad diffraction
peaks attributable to BN nanoparticles. A reduction in particle size was
observed after the washing treatment. The average particle size was
2
9.2, 26.8, and 12.1 nm for BN50, BN50-C, and BN50-F, respectively.
Fig. 5. Yields of p-methoxy-β-nitrostyrene (3) from the nitroaldol reaction
using boron nitride. Reaction conditions: p-methoxybenzaldehyde (1)
Fig. 8 shows TEM image of the BN50-F. As observed for porous BN
materials in previous study [31,50], BN layers were arranged in a
randomly collapsed manner, indicating the turbostratic structure with
low crystallinity. Fig. 9 shows the EDS results of the BN50 and BN50-F
samples. These samples mainly composed of boron and nitrogen, but
they also contain oxygen and small amount of carbon. No significant
change in the composition was observed before and after washing.
As noted above, the BET surface area (SBET) of the BN50 was in-
(
0.5 mmol), nitromethane (2) (1.25 mmol), catalyst (50 mg), toluene (2 mL),
3
73 K, 8 h.
Table 1
Results obtained from the nitroaldol reaction of p-methoxybenzaldehyde and
a
nitromethane using various boron nitride samples .
2
−1
3 yield/%b
4 yield/%c
2
−1
Entry
Catalyst
S
BET /m
g
1 conv./%
creased from 155 to 336 and 309 m g
by washing. Fig. 10 presents
the nitrogen sorption isotherms and pore size distributions of the BN50,
BN50-C, and BN50-F and shows that each material generated a typical
type I isotherm, indicating a microporous structure. The micropore
surface areas (Smicro) were determined using the t-plot method, and
1
2
3
4
5
6
7
8
9
BN50
155
336
309
208
3
10
19
3
0
0
0
0
0
0
0
0
0
0
8
11
BN50-C
16
d
d
BN50-F
41, 70
32, 40
BN100
41
41
35
23
26
54
9
11
9
BN100-C
BN100-F
BN150
2 −1
values of 152, 328, and 297 m g were obtained for the BN50, BN50-
C, and BN50-F, respectively. Thus, the increases in the BET surface
areas after washing are primarily attributed to greater micropore sur-
face areas. The pore size distributions obtained from Barret-Joyner-
Halenda (BJH) plots demonstrate increases in the number of micropores
414
171
3
29
6
BN150-C
BN150-F
Commercial BN
Calcined MgO
Hydrotalcite
7
274
3
32
0
1
1
1
0
1
2
20
13
16
5
after washing, and CO
2
adsorption data further verified the appearance
molecule
has a high vapor
109
5
of a greater number of narrow micropores. Because the CO
2
a
has a kinetic diameter of 0.33 nm [51] and CO
Reactionconditions: p-methoxybenzaldehyde (1) (0.5 mmol), nitromethane (2)
2
(
1.25 mmol), catalyst (50 mg), toluene (2 mL), 373 K, 8 h.
pressure at moderate temperatures, this compound can be used for the
b
c
p-Methoxy-β-nitrostyrene.
analysis of ultramicropores as small as 0.4 nm [52]. Fig. 11 shows the
1
-Methoxy-4-[2-nitro-1-(nitromethyl)ethyl]benzene.
CO adsorption isotherms obtained from the BN50 and BN50-F at
2
d
2
0 h.
298 K. A clear difference can be seen between the two samples. The
BN50 adsorbed very little CO , indicating the absence of ultra-
micropores, while the BN50-F adsorbed 2.9 mL g at 101 kPa, corre-
2
−1
%
, respectively) (entries 3 and 9). When using the BN50-F, prolonging
−1
the reaction time to 20 h increased the yield of 3 to 40 % with a 1
conversion of 70 %. It is reported that amino groups-tethered meso-
porous silica catalysts exhibited a high yield to nitroalkene and could be
reused for five cycles without decrease of the activity [48,49]. Fig. 6
sponding to 0.12 mmol g
.
1
1
The bulk structures of these materials were characterized by
B
solid-state NMR spectroscopy. Boron has a quadrupolar nucleus with a
nuclear spin of 3/2 and is affected by nuclear quadrupole interactions.
Fig. 6. Recyclability of BN50-F and hydrotalcite catalysts. Reaction conditions: p-methoxybenzaldehyde (1) (0.5 mmol), nitromethane (2) (1.25 mmol), catalyst
50 mg), toluene (2 mL), 373 K, 8 h.
(
5

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
flow of 50 cm min−1. These samples were produced
3
Fig. 7. SEM images of BN prepared by heating a mixture of boric acid and hexamethylenetetramine in an NH
3
(
a) without washing (as-synthesized), (b) with washing and recovery by centrifugation, and (c) with washing and recovery by filtration.
Fig. 9. EDS results of BN50 and BN50-F.
Thus, unlike protons and carbon nuclei, second-order nuclear quadru-
pole interactions remain unaveraged even by the magic angle rotation
1
1
of the sample, resulting in a unique line shape. Fig. 12 shows the
B
MAS NMR spectra of the BN50, BN50-C, and BN50-F along with si-
mulated spectra generated using the Bruker TopSpin 3.6 software
package. Table 2 summarizes the NMR parameters used for the simu-
Fig. 8. TEM image of BN prepared by heating a mixture of boric acid and
11
lation, including the B isotropic chemical shift (δiso), quadrupole
3
−1
hexamethylenetetramine in an NH flow of 50 cm min . The sample was
3
coupling constant (C
Q
), and asymmetry factor (η ). The BN50 gener-
Q
produced with washing and recovery by filtration (BN50-F).
ated three signals. The first broad signal with a δiso of 28.5 ppm (shown
in yellow) is ascribed to trigonal boron bonded with nitrogen, namely
6

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
Fig. 10. Nitrogen adsorption isotherms and pore size distributions for BN50, BN50-C, and BN50-F.
Table 2
Parameters employed to simulate 11B MAS NMR spectra.
b
c
Sample
Site
δ
iso
ppm
/
C
Q
/MHz
η
Q
Relative peak
proportion/%
a
BN50
Trigonal B–N
Trigonal B–O
Tetrahedral
28.5
18.5
1.4
2.8
2.5
0.1
0
56
0.30 33
0
0
11
81
d
B–O
BN50-C
BN50-F
Trigonal B–N
Trigonal B–O
Tetrahedral
29.0
17.7
1.5
2.8
2.2
1.0
0.30 12
0
7
d
B–O
Trigonal B–N
Trigonal B–O
Tetrahedral
28.8
17.8
2.8
2.8
1.4
1.3
0
98
1
0.30
0
1
d
Fig. 11. CO
2
adsorption isotherms for BN50 and BN50-F as acquired at 298 K.
B–O
Ammonium
borate
Trigonal B–O
Tetrahedral
18.6
1.5
2.5
0.4
0.30 78
22
0
d
trigonal-planar BN
The second broad signal with a δiso of 18.5 ppm (shown in pink) is at-
tributed to trigonal boron with oxygen: trigonal-planar BO [54]. The
third narrow signal with a δiso of 1.4 pm (shown in orange) corresponds
to tetrahedral boron. Since the formation of cubic BN having BN sites
is unlikely in the present work because the synthesis of this compound
typically requires high pressures, this signal can be assigned to BO
55]. The relative peak proportions were determined to be 56, 33, and
1 % for the BN , BO , and BO sites, respectively. Fig. 9(b) presents the
B spectrum of ammonium borate, (NH )B ·4H O, which contains
three- and four-coordinate boron sites at a 4:1 ratio [56,57]. The
3
. Hexagonal BN contains only this component [53].
B–O
a
b
c
Isotropic chemical shift.
quadrupole coupling constant.
asymmetry factor.
3
4
d
η
Q
is assumed to be 0. C
Q
and η could have large errors because the line
Q
width is narrow.\.
4
[
spectrum exhibits two signals; one with a δiso of 18.6 ppm attributed to
trigonal boron with oxygen (BO ) and the other with a δiso of 1.5 ppm
ascribed to tetrahedral boron with oxygen (BO ). The BO : BO ratio in
the ammonium borate was 78:22, which is in good agreement with the
1
3
3
4
1
1
3
4
5
O
8
2
4
3
4
Fig. 12. 11B MAS NMR spectra obtained from (a) BN50, BN50-C, and BN50-F and (b) ammonium pentaborate tetrahydrate ((NH
)B
O
8
·4H O).
4
5
2
7

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
Fig. 13. B K-edge XAFS spectra obtained from BN50, BN50-C, and BN50-F.
value expected from the crystal structure. This ratio is also close to that
in the BN50 (75:25), suggesting that this material contained borates
Fig. 15. Difference FTIR spectrum of BN50-F following CHCl3 adsorption.
1
1
such as ammonium pentaborate tetrahydrate. After washing, the
B
11
→
B cross-polarization (CP) MAS NMR spectrum could be deconvo-
spectrum of the BN50 was significantly changed, in that the BO
BO proportions were decreased in both the BN50-C and BN50-F. The
relative peak proportions in the BN50-C spectrum were 81, 12, and 7%
for the BN , BO , and BO sites, respectively, while those in the BN50-F
3
and
luted into five peaks in this same region by assuming that the two
4
parameters C
Q
and η were constant based on the ideal powder pattern
Q
from the material [60]. However, it should be taken into consideration
that the powder lineshapes become distorted by second-order nuclear
quadrupole interactions when the CP method is applied to quadrupolar
3
3
4
spectrum were 98, 1, and 1%. These decreases in BO
3
and BO sites
4
indicate the removal of borate species by washing, which in turn in-
creased the surface areas of the materials and exposed active sites on
the surfaces.
nuclei [61,62]. Thus, we simply assigned this region to trigonal B–N
11
(
BN
3
) sites in the B NMR spectra, even though BN
2
O and BNO sites
2
were clearly observed in the B XAFS data. Another reason for the sole
Further characterization was carried out by soft X-ray absorption
spectroscopy. Fig. 13 shows the B K-edge XAFS spectra obtained from
the BN50, BN50-C, and BN50-F. Four sharp peaks can be seen from 191
to 195 eV that correspond to the B1s‒π* transition. The first peak at
assignment of this region to BN sites in the NMR study is the difference
3
in surface sensitivity between methods. The B XAFS spectra were re-
corded in the surface-sensitive total electron yield mode, which probes
11
to a depth of about 0.6 nm from the surface [63,64]. In contrast, the
B
1
92.0 eV is ascribed to the excitation of B1s electrons to the π* state in
NMR spectra were acquired using a single pulse sequence combined
BN
3
, which is related to trigonal B–N [58,59]. The second and third
1
with H high power decoupling to show the bulk structure and enable
peaks at 192.7 and 193.3 eV can be attributed to the excitation of B1s
1
11
quantitative analysis. This is in contrast to the H → B CP method,
electrons to π*(BN
2
O) and π* (BNO
2
) states, respectively [58,31].
1
11
which transfers polarization from the H spin and detects the B spin
signal. The differences between the surface-sensitive XAFS spectra and
These peaks are characteristic of oxygen-doped boron nitrides in which
three coordinate nitrogen atoms are partially replaced with oxygen.
Note that the presence of hydrogen does not influence the spectra. As an
the NMR spectra in this study indicate that such BN
were preferentially located on the surface.
2
O and BNO
2
sites
example, the BN
2
O sites are considered to actually comprise N-B(N)-OH
The fourth peak at 194.0 eV is ascribed to the B1s-π* (BO
3
) transi-
and NH
2
-B(N)-OH groups. Thus, these peaks are related to the forma-
tion, indicating the presence of trigonal B–O. This peak was much in-
tense in the BN50 spectrum than in those produced by the BN50-C and
BN50-F, which is in good agreement with the NMR results. The broad
peak from 197 to 201 eV involves the transition of B1s electrons to the
tion of amino and hydroxyl groups on the surfaces, as also indicated by
1
1
the FTIR data. These results suggest that the B NMR spectra acquired
from the BN50, BN50-C, and BN50-F may reflect the presence of tri-
gonal BN
2
O and BNO
2
in the region from 30 to 20 ppm. In a recent
σ* state for three-coordinate boron (BN ) in addition to four coordinate
3
1
study of exfoliated BN nanosheets, which also have BN
x
O
y
sites, the H
Fig. 14. XPS spectra obtained from BN50, BN50-C, and BN50-F. (a) B1s, (b) N1s.
8

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
Fig. 16. Proposed reaction sequence of the nitroaldol reaction on the BN catalyst.
). It has also been reported that the B1s‒σ* (BN ) transition 4. Conclusions
boron (BO
4
3
produces a broad signal with two peaks resulting from resonance
splitting, centered at 198.9 eV [25,31,58] and that the B1s‒σ* (BO
transition gives a single peak centered at 198.0–198.3 eV [64,63].
4
)
The work reported herein shows that a washing post-treatment ra-
ther than the synthesis conditions greatly affected the level of activity
during the nitroaldol reaction over BN solid base catalysts prepared
from boric acid and hexamethylenetetramine via pyrolysis under an
The surface structure was further investigated by using XPS. Fig. 14
shows the XPS spectra of BN50, BN50-C, and BN50-F samples.
Fig. 14(a) displays the B1s spectra. The BN50 had three peaks centered
at 190.4, 191.9, and 193.1 eV. The peak position of 190.4 eV is attrib-
uted to B–N [65,66]. The peak at 191.9 eV is assigned to oxidized boron
which is related to the formation of B–OH groups [67,68]. The peak
position of 193.1 eV is close to that of ammonium borate (192.8 eV)
NH flow. The as-synthesized BN contained boron oxide species such as
3
ammonium borate, had a low surface area and exhibited poor catalytic
activity for this reaction. After removing these boron oxide species by
washing, the surface area was greatly increased with the simultaneous
formation of micropores. As a result, amino and hydroxyl groups were
exposed on the surface, leading to much higher activity compared with
[
69], thus is attributable to oxidized boron B–O in ammonium borate.
1
1
The BN50-F showed the absence of the peak around 193 eV, indicating
that the oxidized boron species were successfully removed. This is in a
the as-synthesized sample as well as conventional solid bases. FTIR,
B
MAS NMR, and B K-edge XAFS spectroscopy were all confirmed to be
powerful techniques for the characterization of the BN structure. In
addition to trigonal B–N sites, partially oxygen-substituted trigonal
1
1
good agreement with the results of B MAS NMR. Fig. 14(b) shows the
N1s spectra. Again, the BN50 had three peaks centered at 398.1, 399.2,
and 400.4 eV. The first peak at 398.1 eV is ascribed to NeB. The second
peak at 399.2 eV is likely attributable to amino groups [68]. The last
peak at 400.4 eV is assigned to ammonium ion in ammonium borate
boron sites (BN
2
O and BNO ) related to amino and hydroxyl groups
2
were determined to be present in these samples. The post-treated BN
was shown to contain moderately strong basic sites, believed to com-
prise surface amino groups.
[
69]. From the XPS spectra, the removal of ammonium borate was
observed. In addition, the presence of hydroxyl and amino groups on
the surface of BN50-F was indicated.
CRediT authorship contribution statement
The solid base strength of the BN50-F was investigated by acquiring
FTIR spectra following the adsorption of CHCl
3
. Fig. 15 shows the
Atsushi Takagaki: Conceptualization, Methodology, Investigation,
Writing - original draft, Supervision, Funding acquisition. Shohei
Nakamura: Conceptualization, Investigation. Motonori Watanabe:
Methodology, Validation. Yoonyoung Kim: Investigation. Jun Tae
Song: Methodology, Validation. Keiko Jimura: Investigation, Formal
analysis. Kanta Yamada: Investigation. Masaaki Yoshida:
Investigation, Formal analysis, Writing - review & editing. Shigenobu
Hayashi: Investigation, Formal analysis, Writing - review & editing.
Tatsumi Ishihara: Validation, Writing - review & editing, Supervision.
difference spectrum for the BN50-F sample. The C–H stretching band of
−
1
the adsorbed CHCl
of gaseous CHCl at 3019 cm [70]. The shift indicates that the acidic
proton in the CHCl molecule interacted with the basic sites of the BN
via hydrogen bonding. A simultaneous change was observed in the
3
appears at 2984 cm and is red-shifted from that
−1
3
3
−
1
range from 3300 to 3100 cm . As noted above, the pure BN50-F
−
1
generated a broad peak between 3400 and 3200 cm
cribed to surface amino and hydroxyl groups (Fig. 4(b)). The negative
−1
that was as-
−
1
peak centered at 3212 cm along with the positive peak at 3160 cm
in the difference spectrum indicate that the amino groups on the BN
surface interacted with the CHCl , and so functioned as basic sites. The
shift in the position of the C–H stretching band of the adsorbed CHCl is
Declaration of Competing Interest
3
3
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
a useful means of assessing the base strength of these catalysts [71].
−
1
−1
This peak is reported to appear at 3035 cm on SiO
2
[72], 3025 cm
−
1
−1
on γ-Al
2
O
3
[73], 3015 cm
on LiX zeolite [74], 2995 cm
−1
on NaX
on CsX
−
1
zeolite [74], 2966 cm
on KX zeolite [74], and 2925 cm
Acknowledgments
zeolite [74]. Thus, the BN50-F evidently had moderately basic sites
stronger than those on SiO
2
, γ-Al
2
O , LiX, and NaX zeolites, but weaker
3
This work was supported by a Grant-in-Aid for Scientific Research
those on than KX and CsX zeolites.
(
B) (No. 18H01785) from JSPS, Japan. A part of this work was per-
The chloroform-adsorbed FTIR measurement indicated that the
amino groups on the BN surface functioned as moderate basic sites.
Similar to aminosilica catalysts [75,76], we propose the imine forma-
tion mechanism for the nitroaldol reaction. Fig. 16 shows the proposed
reaction sequence of the nitroaldol reaction. First, a proton from ni-
tromethane was abstracted by an amino group of the BN catalyst. Then,
an imine intermediate was formed by the nucleophilic attack of an
amino group of the catalyst on the aldehyde group of benzaldehyde.
Finally, β-nitrostyrene was formed by the nucleophilic attack of de-
protonated nitromethane.
formed at the KEK-IMSS-PF (2017G529 and 2019G674). It was sup-
ported by the AIST Nanocharacterization Facility (ANCF) platform as a
program of the “Nanotechnology Platform” of the MEXT, Japan, Grant
Number JPMXP09A18AT0056.
References
[
1] C.J.H. Jacobsen, J. Catal. 200 (2001) 1–3, https://doi.org/10.1006/jcat.2001.3200.
[2] J.C.-S. Wu, Z.-A. Lin, J.-W. Pan, M.-H. Rei, Appl. Catal. A Gen. 219 (2001) 117–124,
https://doi.org/10.1016/S0926-860X(01)00673-1.
[
[
3] C.-A. Lin, J.C.S. Wu, J.-W. Pan, C.-T. Yeh, J. Catal. 210 (2002) 39–45, https://doi.
org/10.1006/jcat.2002.3638.
4] H. Goto, A. Takagaki, R. Kikuchi, S.T. Oyama, Appl. Catal. A Gen. 548 (2017)
1
22–127, https://doi.org/10.1016/j.apcata.2017.06.041.
9

A. Takagaki, et al.
Applied Catalysis A, General 608 (2020) 117843
[
[
[
[
5] J.C.S. Wu, T.-S. Cheng, C.-L. Lai, Appl. Catal. A Gen. 314 (2006) 233–239, https://
doi.org/10.1016/j.apcata.2006.08.022.
I.N.N. Namboothiri, Chem. Commun. 50 (2014) 6973–6976, https://doi.org/10.
1039/C4CC02279C.
6] J.C.S. Wu, H.-C. Chou, Chem. Eng. J. 148 (2009) 539–545, https://doi.org/10.
[41] L. Biewenga, T. Saravanan, A. Kunzendorf, J.-Y. van der Meer, T. Pijning,
P.G. Tepper, R. van Merkerk, S.J. Charnock, A.-M.W.H. Thunnissen,
G.J. Poelarends, ACS Catal. 9 (2019) 1503–1513, https://doi.org/10.1021/acscatal.
8b04299.
1
016/j.cej.2009.01.011.
7] L. Lin, Z. Li, Y. Zheng, K. Wei, J. Am. Ceram. Soc. 92 (2009) 1347–1349, https://
doi.org/10.1111/j.1551-2916.2009.02964.x.
8] G. Postole, B. Bonnetot, A. Gervasini, C. Guimon, A. Auroux, N.I. Ionescu,
M. Caldararu, Appl. Catal. A Gen. 316 (2007) 250–258, https://doi.org/10.1016/j.
apcata.2006.09.026.
[42] K.K. Kimmel, M.T. Robak, J.A. Ellman, J. Am. Chem. Soc. 131 (2009) 8754–8755,
https://doi.org/10.1021/ja903351a.
[43] R. Mohan, N. Rastogi, I.N.N. Namboothiri, S.M. Mobin, D. Panda, Bioorg. Med.
Chem. 14 (2006) 8073–8085, https://doi.org/10.1016/j.bmc.2006.07.035.
[44] G. Moussa, R. Moury, U.B. Demirci, P. Miele, Int. J. Hydrogen Energy 38 (2013)
7888–7895, https://doi.org/10.1016/j.ijhydene.2013.04.121.
[45] F. Cao, K. Liu, Z. Fang, S. Wang, J. Mater. Sci. Technol. 28 (2012) 956–960, https://
doi.org/10.1016/S1005-0302(12)60158-3.
[
9] Y.-R. Liu, X. Li, W.-M. Liao, A.-P. Jia, Y.-J. Wang, M.-F. Luo, J.-Q. Lu, ACS Catal. 9
(
2019) 1472–1481, https://doi.org/10.1021/acscatal.8b03666.
[
[
10] Y. Wang, L. Shi, W. Lu, Q. Sun, Z. Wang, C. Zhi, A.-H. Lu, ChemCatChem 9 (2017)
1
363–1367, https://doi.org/10.1002/cctc.201700002.
11] M. Fan, J.D. Jimenez, S.N. Shirodkar, J. Wu, S. Chen, L. Song, M.M. Royko,
J. Zhang, H. Guo, J. Cui, K. Zuo, W. Wang, C. Zhang, F. Yuan, R. Vajtai, J. Qian,
J. Yang, B.I. Yakobson, J.M. Tour, J. Lauterbach, D. Sun, P.M. Ajayan, ACS Catal. 9
[46] A.N. Streletskii, D.G. Permenov, B.B. Bokhonov, A.V. Leonov, S.N. Mudretsova,
Colloid J. 72 (2010) 553–558, https://doi.org/10.1134/S1061933X10040174.
[47] M. Florent, T.J. Bandosz, J. Mater. Chem. A 6 (2018) 3510–3521, https://doi.org/
10.1039/C7TA11359E.
(
2019) 10077–10086, https://doi.org/10.1021/acscatal.9b02197.
[
[
[
[
[
[
[
12] A.C. Chien, J.A. van Bokhoven, Catal. Sci. Technol. 5 (2015) 3518–3524, https://
doi.org/10.1039/C5CY00021A.
[48] S.W. Jun, M. Shokouhimehr, D.J. Lee, Y. Jang, J. Park, T. Hyeon, Chem. Commun.
49 (2013) 7821–7823, https://doi.org/10.1039/C3CC43568G.
[49] M. Shokouhimehr, M.S. Asl, B. Mazinani, Res. Chem. Intermed. 44 (2018)
1617–1626, https://doi.org/10.1007/s11164-017-3188-9.
13] L. Gao, G. Fu, M. Wei, Y. Zhu, Q. Liu, E. Crumlin, Z. Liu, X. Bao, ACS Catal. 6 (2016)
6
814–6822, https://doi.org/10.1021/acscatal.6b02188.
14] Y. Zhang, X. Weng, H. Li, H. Li, M. Wei, J. Xiao, Z. Liu, M. Chen, Q. Fu, X. Bao, Nano
Lett. 15 (2015) 3616–3623, https://doi.org/10.1021/acs.nanolett.5b01205.
[50] R. Shankar, M. Sachs, L. Francàs, D. Lubert-Pequel, G. Kerherve, A. Regoutz,
C. Petit, J. Mater. Chem. A 7 (2019) 23931–23940, https://doi.org/10.1039/
C9TA02793A.
15] L. Gao, Y. Wang, H. Li, Q. Li, N. Ta, L. Zhuang, Q. Fu, X. Bao, Chem. Sci. 8 (2017)
5
728–5734, https://doi.org/10.1039/C7SC01615H.
16] J.M. Venegas, W.P. McDermott, I. Hermans, Acc. Chem. Res. 51 (2018) 2556–2564,
https://doi.org/10.1021/acs.accounts.8b00330.
[51] O.P. Mahajan, Carbon 29 (1991) 735–742, https://doi.org/10.1016/0008-
6223(91)90010-G.
17] L. Shi, Y. Wang, B. Yan, W. Song, D. Shao, A.-H. Lu, Chem. Commun. 54 (2018)
[52] M. Thommes, K. Kaneko, A.V. Neimark, J.P. Olivier, F. Rodriguez-Reinoso,
J. Rouquerol, K.S.W. Sing, Pure Appl. Chem. 87 (2015) 1051–1069, https://doi.
org/10.1515/pac-2014-1117.
1
0936–10946, https://doi.org/10.1039/C8CC04604B.
18] J.T. Grant, C.A. Carrero, F. Goeltl, J.Venegas P. Mueller, S.P. Burt, S.E. Specht,
W.P. McDermott, A. Chieregato, I. Hermans, Science 354 (2016) 1570–1573,
https://doi.org/10.1126/science.aaf7885.
[53] P.S. Marchetti, D. Kwon, W.R. Shmidt, L.V. Interrante, G.E. Maciel, Chem. Mater. 3
(1991) 482–486, https://doi.org/10.1021/cm00015a023.
[
[
[
19] L. Shi, D. Wang, W. Song, D. Shao, W.-P. Zhang, A.-H. Lu, ChemCatChem 9 (2017)
[54] D. Zielniok, C. Cramer, H. Eckert, Chem. Mater. 19 (2007) 3162–3170, https://doi.
org/10.1021/cm0628092.
1
788–1793, https://doi.org/10.1002/cctc.201700745.
20] L. Shi, B. Yan, D. Shao, F. Jiang, D. Wang, A.-H. Lu, Chin. J. Catal. 38 (2017)
[55] G.L. Turner, K.A. Smith, R.J. Kirkpatrick, E. Oldfield, J. Magn. Reson. 67 (1986)
544–550, https://doi.org/10.1016/0022-2364(86)90391-4.
3
89–395, https://doi.org/10.1016/S1872-2067(17)62786-4.
21] R. Huang, B. Zhang, J. Wang, K.-H. Wu, W. Shi, Y. Zhang, Y. Liu, A. Zheng,
R. Schlögl, D.S. Su, ChemCatChem 9 (2017) 3293–3297, https://doi.org/10.1002/
cctc.201700725.
[56] A.R. Ploszajski, M. Billing, N.T. Skipper, J.K. Cockcroft, Chem. Commun. 55 (2019)
8290–8292, https://doi.org/10.1039/C9CC02720C.
[57] Y. Song, J. Wang, L. Chen, P. Yang, J. Lumin. 225 (2020) 117325, , https://doi.org/
10.1016/j.jlumin.2020.117325.
[
[
22] Y. Honda, A. Takagaki, R. Kikuchi, S.T. Oyama, Chem. Lett. 47 (2018) 1090–1093,
https://doi.org/10.1246/cl.180510.
[58] M. Petravic, R. Peter, I. Kavre, L.H. Li, Y. Chen, L.-J. Fan, Y.-W. Yang, Phys. Chem.
Chem. Phys. 12 (2010) 15349–15353, https://doi.org/10.1039/C0CP00984A.
[59] S.P. Huber, E. Gullikson, R.W.E. van de Kruijs, F. Bijkerk, D. Prendergast, Phys. Rev.
B 92 (2015) 245310, , https://doi.org/10.1103/PhysRevB.92.245310.
[60] R.W. Dorn, M.J. Ryan, T.-H. Kim, T.W. Goh, A. Venkatesh, P.M. Heintz, L. Zhou,
W. Huang, A.J. Rossini, Chem. Mater. 32 (2020) 3109–3121, https://doi.org/10.
1021/acs.chemmater.0c00104.
23] J.M. Venegas, J.T. Grant, W.P. McDemott, S.P. Burt, J. Micka, C.A. Carrero,
I. Hermans, ChemCatChem 9 (2017) 2118–2127, https://doi.org/10.1002/cctc.
2
01601686.
[
[
24] Y. Wang, L. Zhao, L. Shi, J. Sheng, W. Zhang, X.-M. Cao, P. Hu, A.-H. Lu, Catal. Sci.
Technol. 8 (2018) 2051–2055, https://doi.org/10.1039/C8CY00163D.
25] A.M. Love, B. Thomas, S.E. Specht, M.P. Haanrahan, J.M. Venegas, S.P. Burt,
J.T. Grant, M.C. Cendejas, W.P. McDermott, A.J. Rossini, I. Hermans, J. Am. Chem.
Soc. 141 (2019) 182–190, https://doi.org/10.1021/jacs.8b08165.
[61] S. Hayashi, K. Hayamizu, Chem. Phys. Lett. 203 (1993) 319–324, https://doi.org/
10.1016/0009-2614(93)85575-9.
[
[
[
[
[
[
26] Q. Liu, C. Chen, M. Du, Y. Wu, C. Ren, K. Ding, M. Song, C. Huang, ACS Appl. Nano
Mater. 1 (2018) 4566–4575, https://doi.org/10.1021/acsanm.8b00867.
27] S. Torii, K. Jimura, S. Hayashi, R. Kikuchi, A. Takagaki, J. Catal. 355 (2017)
[62] S. Hayashi, Solid State Nucl. Magn. Reson. 3 (1994) 93–101, https://doi.org/10.
1016/0926-2040(94)90027-2.
[63] M. Kasrai, M.E. Fleet, S. Muthupari, D. Li, G.M. Bancroft, Phys. Chem. Minerals 25
(1998) 268–272, https://doi.org/10.1007/s002690050113.
1
76–184, https://doi.org/10.1016/j.jcat.2017.09.013.
28] S. Namba, A. Takagaki, K. Jimura, S. Hayashi, R. Kikuchi, S.T. Oyama, Catal. Sci.
Technol. 9 (2019) 302–309, https://doi.org/10.1039/C8CY00940F.
29] A. Bhattacharya, S. Bharracharya, G.P. Das, Phys. Rev. B 85 (2012) 035415, ,
https://doi.org/10.1103/PhysRevB.85.035415.
[64] D. Li, G.M. Bancroft, M.E. Fleet, P.C. Hess, Z.F. Yin, Am. Mineral. 80 (1995)
873–877, https://doi.org/10.2138/am-1995-9-1001.
[65] C.-Y. Su, Z.-Y. Juang, K.-F. Chen, B.-M. Cheng, F.-R. Chen, K.-C. Leou, C.-H. Tsai, J.
Phys. Chem. C 113 (2009) 14681–14688, https://doi.org/10.1021/jp904402h.
[66] Z. Wu, Y. Chen, W. Li, J.Li.H. Yu, L. Liu, G. Wu, T. Yang, L. Luo, RSC Adv. 8 (2018)
17944–17949, https://doi.org/10.1039/C8RA02017E.
30] M. Maleki, A. Beitollahi, J. Lee, M. Shokouhimehr, J. Havadpour, E.J. Park, J. Chun,
J. Hwang, RSC Adv. 5 (2015) 6528–6535, https://doi.org/10.1039/C4RA11431K.
31] Q. Weng, D.G. Kvashnin, X. Wang, O. Cretu, Y. Yang, M. Zhou, C. Zhang, D.-
M. Tang, P.B. Sorokin, Y. Bando, D. Golberg, Adv. Mater. 29 (2017) 1700695, ,
https://doi.org/10.1002/adma.201700695.
[67] L. Fu, G. Chen, N. Jiang, J. Yu, C.-T. Lin, A. Yu, J. Mater. Chem. A 4 (2016)
19107–19115, https://doi.org/10.1039/C6TA06409D.
[68] A. Takagaki, Catal. Today 352 (2020) 279–286, https://doi.org/10.1016/j.cattod.
2019.11.007.
[
[
[
[
[
[
[
[
[
32] A. Nag, K. Raidongia, K.P.S.S. Hembram, R. Datta, U.V. Waghmare, C.N.R. Rao, ACS
Nano 4 (2010) 1539–1544, https://doi.org/10.1021/nn9018762.
33] M.I. Baraton, T. Merle, P. Quintard, V. Lorenzelli, Langmuir 9 (1993) 1486–1491,
https://doi.org/10.1021/la00030a011.
[69] V. Cholet, L. Vandenbulcke, J.P. Rouan, P. Baillif, R. Erre, J. Mater. Sci. 29 (1994)
1417–1435, https://doi.org/10.1007/BF00368905.
[70] S. Huber, H. Knözinger, J. Mol. Catal. A Chem. 141 (1999) 117–127, https://doi.
org/10.1016/S1381-1169(98)00255-6.
34] J. Li, L. Gao, J. Mater. Chem. 13 (2003) 628–630, https://doi.org/10.1039/
B210385K.
[71] J.C. Lavalley, Catal. Today 27 (1996) 377–401, https://doi.org/10.1016/0920-
5861(95)00161-1.
35] T. Sainsbury, A. Matti, P. May, Z. Wang, I. McGovern, Y.K. Gun’ko, J. Coleman, J.
Am. Chem. Soc. 134 (2012) 18758–18771, https://doi.org/10.1021/ja3080665.
36] M.L. Kantam, P. Sreekanth, Catal. Lett. 57 (1999) 227–231, https://doi.org/10.
[72] E.A. Paukshtis, N.S. Kotsarenko, L.G. Karakchiev, React. Kinet. Catal. Lett. 12
(1979) 315–319, https://doi.org/10.1007/BF02064262.
1
023/A:1019012019131.
[73] T.A. Gordymova, A.A. Davydov, React. Kinet. Catal. Lett. 23 (1983) 233–238,
https://doi.org/10.1007/BF02065566.
37] H. Yu, J. Xie, Y. Zhong, F. Zhang, W. Zhu, Catal. Commun. 29 (2012) 101–104,
https://doi.org/10.1016/j.catcom.2012.09.032.
[74] J. Xie, M. Huang, S. Kaliaguine, React Kinet. Catal. Lett. 58 (1996) 217–227,
https://doi.org/10.1007/BF02067026.
38] T. Okino, Y. Hoashi, Y. Takemoto, J. Am. Chem. Soc. 125 (2003) 12672–12673,
https://doi.org/10.1021/ja036972z.
[75] K.K. Sharma, A.V. Biradar, T. Asefa, ChemCatChem 2 (2010) 61–66, https://doi.
org/10.1002/cctc.200900259.
39] N. Takenaka, J. Chen, B. Captain, R.S. Sarangthem, A. Chandrakumar, J. Am. Chem.
Soc. 132 (2010) 4536–4537, https://doi.org/10.1021/ja100539c.
40] D.K. Nair, R.F.S. Menna-Barreto, E.N. da Silva Júnior, S.M. Mobin,
[76] J.D. Bass, A. Solovyov, A.J. Pascall, A. Katz, J. Am. Chem. Soc. 128 (2006)
3737–3747, https://doi.org/10.1021/ja057395c.
10