Adsorption driven formate reforming into hydride and tandem hydrogenation of nitrophenol to amine over PdO: Xcatalysts

Article abstract of DOI:10.1039/d0cy01704c

Due to their high toxicity and non-biodegradability, efficient reduction of nitroarenes to amines is of great practical importance, yet it still remains a significant challenge. Herein, we report PdO/PdO2 nanoparticles uniformly supported on titanate nanotubes (PdOx/TiNTs) for catalyzing the tandem dehydrogenation of sodium formate (SF) and hydrogenation of p-nitrophenol (PNP) to p-aminophenol (PAmP) under mild conditions. Notably, SF adsorption is mainly driven by the hydrogen bonding interactions between the H atom in SF and surface Pd sites, which factually makes the interface of PdOx/TiNT-SF an effective platform for C-H activation. Meanwhile, it is also found that the efficiency of the hydrogenation reaction depends on the reduction rate of the nitro group to nitroso group, and the O atoms adjacent to Pd are considered as the essential sites that facilitate this process. On the basis of the above two effects, the PdOx/TiNT catalyst shows unprecedented catalytic activity (turnover frequency, TOF, is 45.6 h-1) and good selectivity (～100%) during PNP reduction at room temperature. This work deepens our understanding on tandem catalytic (de)hydrogenation systems, and will benefit the design of heterogeneous catalysts for the production of industrially important chemicals.

Full text of DOI:10.1039/d0cy01704c

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Adsorption driven formate reforming into hydride
and tandem hydrogenation of nitrophenol to
amine over PdOx catalysts†
Cite this: DOI: 10.1039/d0cy01704c
Xiaohui Zhu,ab Shipan Liang,b Shuang Chen,b
ab
ab
*
*
Xiangdong Liu
and Renhong Li
Due to their high toxicity and non-biodegradability, efficient reduction of nitroarenes to amines is of great
practical importance, yet it still remains a significant challenge. Herein, we report PdO/PdO2 nanoparticles
uniformly supported on titanate nanotubes (PdOx/TiNTs) for catalyzing the tandem dehydrogenation of
sodium formate (SF) and hydrogenation of p-nitrophenol (PNP) to p-aminophenol (PAmP) under mild
conditions. Notably, SF adsorption is mainly driven by the hydrogen bonding interactions between the H
atom in SF and surface Pd sites, which factually makes the interface of PdOx/TiNT–SF an effective platform
for C–H activation. Meanwhile, it is also found that the efficiency of the hydrogenation reaction depends
on the reduction rate of the nitro group to nitroso group, and the O atoms adjacent to Pd are considered
as the essential sites that facilitate this process. On the basis of the above two effects, the PdOx/TiNT
catalyst shows unprecedented catalytic activity (turnover frequency, TOF, is 45.6 h−1) and good selectivity
(∼100%) during PNP reduction at room temperature. This work deepens our understanding on tandem
catalytic (de)hydrogenation systems, and will benefit the design of heterogeneous catalysts for the
production of industrially important chemicals.
Received 31st August 2020,
Accepted 9th October 2020
DOI: 10.1039/d0cy01704c
rsc.li/catalysis
hydrazine hydrate (N2H4·H2O) are usually required as reducing
Introduction
agents to remove oxygen atoms from nitro groups and
With the world's progress in industrialization, organic
contaminants from a huge amount of industrial effluents have
caused greater concerns over the growing risk of environmental
ultimately give rise to aniline products.8–11
In recent years, a variety of heterogeneous catalytic systems
based on noble metals (e.g. Pd, Pt, Au and Ag) have been
developed for the effective catalytic reduction of PNP to
PAmP.12–16 Lai and co-workers used dispersedly anchored Au
on chitosan functionalized activated coke to improve the
stability of catalysts in the reduction process.17 Jiang et al.
reported Pd nanoparticle-decorated hierarchically porous TiO2
scaffolds with strong synergistic interactions for the reduction
of highly concentrated PNP wastewater. In addition,
pollution.1,2
Among
various
organic
contaminants,
p-nitrophenol (PNP), which has been widely used in the
manufacturing industry, is considered as a priority pollutant
due to its carcinogenic character, high toxicity, and overall
recalcitrance to microbial degradation.3,4 Consequently,
exploring a highly efficient technique to eliminate PNP from
wastewater is becoming an essential environmental concern.5
Importantly, by introducing catalytic hydrogenation reactions,
PNP can be converted into more valuable p-aminophenol
(PAmP), which is a key intermediate in pharmaceuticals,
agrochemicals, dyes and pigments.6,7 To this end, extra
hydrogen resources such as sodium borohydride (NaBH4) or
construction of bimetallic structures is also
a reliable
alternative to establish efficient catalytic systems.18 In this
case, Xiao et al. showed that the catalytic reaction pathway can
be altered by controlling the alloying segment of Cu into Au
nanoparticles, which allows nitroaromatics to be transformed
directly into anilines in a highly selective manner.19 Similarly,
Song and co-workers revealed remarkable Pt–Au synergistic
catalytic effects on PtAu-nanoflowers/rGO with improved
catalytic performances for the reduction reaction.20 So far,
most of the research is limited to the design and preparation of
novel catalysts to enhance the catalytic activity, selectivity, and
stability during PNP hydrogenation reactions, while very few
reports focus on the optimization of the kinetic process and
elucidation of the catalytic mechanism.
a College of Textile Science and Engineering (International Institute of Silk),
Zhejiang Sci-Tech University, Hangzhou, 310018, Zhejiang Province, P. R. China.
E-mail: liuxd2007@yeah.net, lirenhong@zstu.edu.cn
b School of Materials Science & Engineering, Zhejiang Sci-Tech University,
Hangzhou, 310018, Zhejiang Province, P. R. China
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0cy01704c
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
As a water-soluble salt with moderate theoretical H2 weight
efficiency, sodium formate (SF) represents potentially
anhydrous ethanol (>99%), palladium(II) nitrate (Pd(NO3)2),
sodium formate (HCOONa), p-nitrophenol (PNP), p-nitrosophenol
a
interesting reducing agent.21–24 Recently, we reported that
hollow PdO supported on titanate nanotubes (TiNTs) is a
highly active and stable catalyst for the dehydrogenation of SF
solution into molecular hydrogen at room temperature. It is
found that the presence of both Pd centers and their
terminated oxygen atoms strongly enhances the reactivity of
PdO toward the C–H bond as well as water activation.25 Herein,
PdOx/TiNTs are coupled with SF to reduce PNP into PAmP in
an aqueous phase, thus representing an exemplary system for
the investigation of nitro (–NO2) group reduction kinetics. The
operando Raman study and GC-MS results suggest that the
adsorption of SF onto the catalyst surface would drive hydride
in SF to the Pd sites and subsequently construct the novel
Pd⋯H⋯C complex. It is demonstrated that the rate-limiting
step is the reduction of the nitro group to nitroso group, and
the nitrogen atoms in nitro-compound species can be captured
by a high density of surface terminated O atoms, resulting in
the direct reduction of PNP to intermediate species. Therefore,
this unique PdOx/TiNT catalyst shows ∼100% selectivity and a
fast reaction rate (turnover frequency, TOF, is up to 45.6 h−1)
for PNP reduction to PAmP under ambient conditions, which is
one order of magnitude higher than analogous catalytic
systems (Scheme 1). In addition, this reaction is also similar to
our previously reported catalytic hydrogen production from SF
solution using PdO/TiNT catalysts, where only palladium or
palladium oxide is active for the reduction of organic dyes in
the presence of SF.25 This work not only provides insights into
the mechanism of tandem catalytic dehydrogenation and
hydrogenation systems, but also paves a way to improve the
catalytic performance of supported catalysts for the production
of industrially important chemicals.
(PNSP),
p-aminophenol
(PAmP),
p-nitrobenzyl
alcohol,
p-nitrobenzoic acid and distilled water were supplied from local
sources. All the reagents were used without further purification.
Protonated titania nanotubes were prepared the same as
in our previous method.25 Typically, TiO2 nanoparticles (25
nm) were first placed into a beaker (50 mL). Using a syringe,
10 M sodium hydroxide solution (60 mL) was added. The
mixture was left stirring for 12 h in the dark. The resultant
white slurry was transferred to a Teflon vessel which was
then placed into a sealed autoclave and put in an oven at
150 °C for 72 h. After cooling to room temperature, the
precipitated powder was filtered and washed with 0.1 M
aqueous HCl solutions and Milli-Q water, until the filtrate
reached pH = 7. This resulting sample was protonated titania
nanotubes and denoted as TiNTs.
To in situ load PdNPs onto TiNTs, a given amount of
Pd(NO3)2·2H2O (palladium(II) nitrate dihydrate) was dissolved
in 50 mL dilute nitric acid solution to form a yellow transparent
solution. 1 mg TiNTs were added into 100 mL deionized water
and ultrasonicated for 10 min to ensure that they were evenly
dispersed in water. Then the yellow palladium precursor
solution was dropped into the aforementioned TiNTs dropwise
under stirring. The mixture was kept stirring overnight, then
the stirrer was switched off and the mixture was set aside for a
certain time till the solid–liquid separation. After abandoning
the supernatant, the precipitant was accumulated and washed
several times with water and ethanol. Finally, the precipitate
was dried overnight in a vacuum box at 80 °C. The PdNPs/TiNTs
were calcined at 300 °C for 3 h in an O2 or N2 atmosphere to
obtain the final catalysts.
Characterization
Experimental
TEM images were recorded on a JEOL JEM-1230 operated at
100 kV. HR-TEM and HAADF-STEM combined with EDS
measurements were performed on an FEI TITAN Cs-corrected
ChemiSTEM equipped with an energy dispersive X-ray (EDX)
spectroscope, operating at 200 kV.
Materials synthesis
Materials. Materials: P-25 TiO2 (Sigma-Aldrich), sodium
hydroxide, hydrochloric acid, nitric acid, ethyl acetate,
X-ray photoelectron spectroscopy (XPS) analysis was
conducted using
a
Thermo Scientific ESCA Lab250
spectrometer, which consists of monochromatic Al Kα as the
X-ray source, a hemispherical analyzer and a sample stage with
multiaxial adjustability to obtain the composition on the
surface of the samples. All the binding energies were calibrated
by the C1s peak of the surface adventitious carbon at 284.6 eV.
To confirm the terminal products of this reaction, gas
chromatography-mass spectrometry (GC-MS) was used. The
reacted solution was centrifuged at 5000 rpm for 5 min to
separate the supernatant and the catalyst solid. Then the
solution was freeze dried to dislodge the water from the
intermediate products or the end-products. Extra ethyl
acetate (EA) was added into the dry products, and then the
organic compounds were dissolved in the EA; some inorganic
salts like NaHCO3 or the unreacted HCOONa were discarded.
Scheme 1 Schematic illustration of tandem catalysis: left hand: room
temperature catalytic reforming of aqueous formate ions yields
bicarbonate and H species on the surface Pd atoms of the catalyst;
right hand: p-nitrophenol adsorbed on O atoms is reduced step by
step in this system by the H species from adjacent Pd atoms.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Catalysis Science & Technology
Paper
In situ Raman spectra were recorded on a Tri Vista Raman Results and discussion
spectrometer. The Raman scattering was induced by
irradiating aqueous solutions of SF and PNP in the presence
Catalyst characterization
PdOx/TiNT and Pd/TiNT catalysts with a Pd loading amount
of 1.0 wt% were synthesized by the incipient wetness
impregnation (IWI) method followed by calcination under an
O2 or N2 atmosphere, respectively (ESI†). Highly dispersed
PdNPs on the surface of nanotubes are observed in the
transmission electron microscopy (TEM) images of PdOx/TiNTs
(Fig. 1a) and Pd/TiNTs (Fig. S1†). The results show that these
two specimens have similar particle sizes, viz. the average sizes
for PdO and Pd are 4.93 and 4.62 nm, respectively. The high-
resolution TEM (HRTEM) image (inset of Fig. S1†) of Pd/TiNTs
presents a typical Pd(111) facet, while the image of PdOx/TiNTs
(inset of Fig. 1a) reveals a lattice spacing of about 0.275 nm for
PdOx/TiNTs, corresponding to the (101) plane of PdO. Energy-
dispersive X-ray spectroscopy (EDS) analysis confirms that Pd,
Ti, and O elements are uniformly distributed (Fig. 1b). The
powder X-ray diffraction (XRD) patterns of the as-prepared
catalysts display diffraction peaks located at 2θ = 11.3°, 24.4°,
29.3°, and 48.4°, which could be attributed to the characteristic
peaks of layered protonated titanates (PDF# 47-0561). Due to
the relatively low loading amount and high dispersity, only
small characteristic XRD peaks for Pd(111), PdO(101) and
PdO2(211) are identified,26 which is consistent with the TEM
results (Fig. 1c).
of solid catalysts with the second harmonic radiation at 633
nm of a Nd: YAG laser. A back scattering geometry was used
to excite sample and collect the Raman scattered light by
reflective optics. The data was recorded with a collection
(integration) time of 100 s and a spectral resolution of 2 cm−1
in the range of 750–1650 cm−1
.
Attenuated total reflectance (ATR-FTIR) was used to
confirm the real time reaction process. The range of
wavenumbers was set from 400 to 4000 cm−1. 1–2 drops of
0.01 M PNP solution were dropped into the center of the
crystal stage, and then a small amount of catalysts was
added. The spectral changes during the reaction were
measured at regular intervals. Considering the difficulty in
determining the amount of catalyst, this method could just
reflect the catalytic activity in the qualitative aspect.
Catalytic activity
The catalytic reduction of PNP to PAmP was carried out as
follows: first, 1 M SF and 0.1 mM PNP were selected to
determine the activity of the catalysts. Second, in separate
standard quartz cuvettes, 2 mL probe solution was added,
followed by adding 1 mg catalysts without stirring. After the
solid catalysts sank into the bottom of the cuvette in several
seconds, the upper solution was recorded by using a UV-2550
spectrometer (Shimadzu Inc.). The parameters were set as
follows: spectral scanning range from 200 to 500 nm, and
cyclic scan of 60 cycles with an interval of 60 s.
A detailed investigation of the electronic states of the
surface metal species is carried out by X-ray photoelectron
spectroscopy (XPS) analysis (Fig. 1d and S2†). To determine the
oxidation states of palladium, the main peak of the Pd3d5/2
For the recycling experiment, 10 mg catalysts were added into
20 mL PNP/SF reaction solution in a 50 mL beaker; after 30 min of
reaction, the solution was centrifuged, and the upper solution was
detected using the UV-vis spectrometer to confirm the reduction
degree and the solid catalyst was collected for the next cycle.
Analysis of gases in the reaction system was carried out with
1 mg of catalysts suspended in a 10 mL solution containing 1 M
SF and 0.01 M PNP in a 55 mL Pyrex test tube under stirring
(400 10 rpm). A water bath was used to maintain the reaction
temperature at room temperature (25 0.5 °C). Nitrogen gas
was used to eliminate the oxygen in the test tubes, and the
tubes were finally sealed with silica gel stoppers. Gas volumes
of 400 μL were extracted from the test tubes using a microliter
syringe at regular intervals, and GC-TCD was employed for
evaluating the gas evolution amount, including H2, O2, CO2
and CO. Triplicate reactions were studied from which
statistical analysis of the reactivity could be extracted.
The TOF reported here is an apparent TOF value based on
the surface number of Pd atoms in the catalyst, which is
calculated by the following equation: TOF = (nH /nPd × t) ×
2
(d/1.3), where nH and nPd represent the molar amounts of
2
Fig. 1 (a) TEM image of PdOx/TiNTs (inset shows the HRTEM image);
(b) STEM image of PdOx/TiNTs and EDS maps of O, Pd, and Ti,
respectively; (c) XRD patterns of 1.0 wt% Pd/TiNTs calcined under N2
or O2; (d) high resolution Pd3d XPS spectra of 1.0 wt% Pd/TiNTs
calcined under N2 or O2.
evolved H2 within t h and Pd catalyst, respectively, and d is
the average diameter of PdNPs; 1.3/d is the empirical surface
atom dispersion degree for nanoscale particles. The TOF
values in references are also in similar calculations.
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
core-level spectra is fitted with two components at 335.11 and
336.04 eV corresponding to Pd0 and surface Pd2+ species,
respectively (Fig. 1d). A large fraction of palladium is present in
the Pd0 metallic state in the Pd/TiNT samples. Additionally, the
absence of any shake-up satellite feature in the spectra at
binding energies above 338.0 eV further indicates the lower
palladium oxidation state. The PdOx/TiNT catalyst shows that
the position of the palladium main peak shifts slightly toward
a higher binding energy, which is strongly related to the
modification of the electron density on Pd species after
calcination under different atmospheres. Moreover, it appears
that the fraction of Pd2+ on the surface, as estimated from the
ratio of PdO/(Pd + Ti + O) peaks tends to increase within PdOx/
TiNTs. More critically, a new peak at around 340.3 eV is
observed, indicating that the oxidation state of a fraction of Pd
species becomes +4 after calcination in an O2 atmosphere.27–29
However, only negligible shifts of O1s as well as Ti2p are
observed in the PdOx/TiNT and Pd/TiNT samples (Fig. S3†),
suggesting that the incorporated nanoparticles do not cause
obvious modification in the TiNT structure.
Fig. 2 Concentration changes of PNP and kinetic curves at 400 nm:
(a) PdOx/TiNTs and (b) Pd/TiNTs. Reaction conditions: PNP solution
(10 μM) + HCOONa (1 M) (insets are the images of the visual color
change as a function of reaction time).
The reduction of PNP is chosen as an ideal probe reaction
to assess the catalytic activity of as-prepared PdOx/TiNTs. The
UV-vis spectra of standard PNP solutions with or without SF
are presented in Fig. S4.† Two control experiments indicate
that PNP will not be reduced in the absence of SF, yet PNP
shows better adsorbability on the PdOx/TiNT surface than on
pristine TiNTs (Fig. S5†). The PNP reduction reaction over
PdOx/TiNTs and Pd/TiNTs obeys the quasi-first-order kinetics,
and the fitting linear dependency with slopes of 0.15 and
0.07 corresponding to PdOx/TiNTs and Pd/TiNTs respectively
is obtained from logarithmic plots of c0/c vs. reaction time,
indicating that surface oxygen species most likely participate
in this reaction (Fig. 2). Two effects result in the different
catalytic activities of PdOx/TiNTs and Pd/TiNTs. First,
oxidized palladium performs far superior to metallic Pd-
based counterpart catalysts in the activation of the C–H bond
in SF, as discussed in a previous report.25 Second, the oxygen
atoms in the vicinity of Pd supply more appropriate sites for
the capture of nitrogen atoms in nitro-compound species
benefited by the high density of surface terminated O atoms
and the conjugation effect of the nitro group in the
para-position of nitrophenol. The reaction efficiency of PdOx/
TiNTs is nearly one order of magnitude higher than those of
other analogous PNP catalytic reduction reactions. We also
compared the activity and selectivity among different catalytic
systems with similar reaction conditions, as listed in Table
S1.† To assess the reusability and stability of PdOx/TiNTs,
10 consecutive utilization cycles are performed. Similar
conversion and selectivity are determined as compared with
the fresh catalyst, indicating the excellent reusability of the
PdOx/TiNT catalyst (Fig. S6†).
1113 and 1170 cm−1 can be attributed to the C–H in-plane
bending vibration of the benzene ring, while the bands at
1284 and 1350 cm−1 are assigned to the stretching vibration
of –NO2 groups.30 Additionally, the SF stretching vibration of
CO also centers at around 1350 cm−1, which is interfered
by the characteristic peak of PNP. The intensity of PNP
vibrational peaks decreases steadily as
a function of
monitoring time, and the final disappearance indicates its
successful reduction.
Fig.
3 Real time Raman vibration spectrum change: (a) SF/PNP
solution reacted with PdOx/TiNTs, (b) SF solution with or without
PdOx/TiNTs, and bare PdOx/TiNTs (black line). Notes: stretching
vibration: ν, asymmetric stretching vibration: νs, bending vibration: β,
flexion: δ. (c) GC-TCD detected the change of gas composition in the
system of PNP/SF. (d) Real-time ATR-FTIR vibration of CO2 in solution.
0.01 M PNP + 1 M SF + PdOx/TiNTs, 25 °C.
To further confirm the degradation process of PNP over
the PdOx/TiNT catalyst in the presence of SF, in situ Raman
spectroscopy is conducted. As shown in Fig. 3a, the
characteristic vibration peaks of PNP groups are observed at
1113, 1170, 1284, 1350, and 1595 cm−1.The former peaks at
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Catalysis Science & Technology
Paper
Notably, neither the relative intensity nor the peak
position of all the PNP vibration modes changes when the
PdOx-TiNT catalyst is initially introduced into the system
involving both PNP and SF, which is in sharp contrast to the
influence of the PdOx-TiNT catalyst on the vibration modes
of bare SF. As shown in Fig. 3b, the bare SF solution (blue
line) shows four characteristic peaks at 1212, 1297, 1348 and
the vibration intensity of the CO2 doublet peaks increases
immediately and then decreases until vanishing. The variation
tendency can also be observed more intuitively in the inset
figure, which shows the U-like shape evolution of CO2
absorption intensity at 2360 cm−1. Such a trend demonstrates
that CO2 is more likely to generate around the surface of
the PdOx/TiNT catalyst, and desorbs from the solution.
Furthermore, GC-MS is used to confirm the intermediates and
final products of the reaction solution. From Fig. S7,† it can be
seen that before the reaction, only PNP is detected at 18.3 min
(black line). Upon 5 min of the reaction, the intensity of PNP
obviously decreases, whereas one additional characteristic peak
of PAmP appears at 14.9 min. The substances in both positions
are further confirmed subsequently using a connected MS
detector (Fig. S8†). In addition, the reaction for a longer period
(30 min) gives the same and exclusive product of reduced PNP.
Therefore, based on the GC-MS results, we reasonably suggest
that PAmP is the final product after the whole reduction
reaction. Moreover, this reaction system can be also applied
to other organic contaminants, e.g. nitrobenzene derivative
p-nitrobenzyl alcohol, and the results are shown in Fig. S9–S11.†
The C–H bond in SF can be activated and cleaved to produce
active hydride (*H) species on the PdOx/TiNT surface, as
indicated by the Raman results. Meanwhile, the formate
residue after dehydrogenation can also be hydrolyzed to form
1378 cm−1, assigned respectively to the β(C–H)in-plane
,
β(C–H)out-plane, ν(CO), and ν(C–H) vibrations. The stretching
vibration is the main mode for SF, and the C–H vibration
intensity is apparently much stronger than that of CO.31,32
However, once SF and PdOx/TiNTs co-exist, two bending
vibrations of C–H bonds vanish (red line). Furthermore, the
intensities of stretching vibration of C–H and CO are also
exchanged while no relative shift is observed. In analogy to
the size-dependent behavior of the adsorption modes of
formate species on PVP–Ru nanoparticles, the ν(CO)
around 1348 cm−1 most likely belongs to the asymmetric
mode of O–C–O stretching in a monodentate mode and no
bidentate formate species is observed.33 The obvious decrease
in the intensity of ν(C–H) vibration coupled with the absence
of bands characteristic of β(C–H) upon introduction of the
PdOx/TiNTs suggests that the coordination environment of
SF over the PdOx/TiNT surface has been changed. As Pd
possesses favorable hydrogen affinity, it is reasonable to
propose that the H atom from SF would combine with Pd to
form a Pd⋯H⋯C complex on the catalyst surface, which
restricts and thus decreases the vibration of the C–H bond.
In addition, the enhanced ν(CO) vibration mode can be
interpreted in terms of a delicate balance between the force
of the C–H bond to the metal and the back force to the CO
bond. This transformation factually makes the interface of
PdOx/TiNTs and SF an effective platform for activating the H
atom and subsequently driving it away from the SF molecule,
suggesting great potential applications for subsequent –NO2
reduction and/or H2 generation.
−
bicarbonate (HCO3 ). According to the Henderson–Hasselbalch
equation (pH = pKa + log([base]/[acid])) and the conjugate acid–
base equilibrium theory, it is estimated from Table S2† that
both SF and NaHCO3 are alkaline and NaHCO3 has a relatively
stronger alkalinity. PNP is alkalescent (acid dissociation
constant pKa = 7.15) and PAmP is a weak acid (pKa = 5.48),
thus in principle, the catalytic products PAmP and NaHCO3
will continue to complete the so-called acid and alkali
neutralization reaction in aqueous solution. Consequently, the
carbonate was easily decomposed into water and CO2, which
explains the origin of CO2 detected in GC-TCD and ATR-IR.
After all the PNP molecules were transformed into PAmP, the
redundant *H species generated by SF activation will combine
with each other to produce H2 gas.
The composition of the evolved gas is examined in real
time by GC-TCD (Fig. 3c). It is observed that the amount of
evolved carbon dioxide (CO2) increases gradually in the initial
2
hour reaction, and the reaction solution also turns
More importantly, the fact that no aqueous by-products or
intermediates are detected by GC-MS shows the high catalytic
selectivity of PdOx/TiNTs. Within the catalytic process from
PNP to PAmP, nitroso, imine and hydroxylamine are
considered as the possible intermediate products. However,
both imine and hydroxylamine are unstable and facile to
reduce or oxidize, and one of the nitroso intermediate
products p-nitrosophenol (PNSP) with relative better stability
is selected and used for a series of designed experiments to
determine the rate-limiting step (Fig. 4). It should be noted
that PNP and PNSP cannot be distinguished from the peak
position as both of them exhibit the major peaks at 400 nm
in the UV-vis spectra (blue and pink lines in Fig. S4†).
Therefore, the kinetics of concentration change is carried
out. By comparing the kinetic curves in Fig. 4 and 2, kPNSP is
more than three times higher than kPNP during the reduction
process, and kPNP/PNSP is in between them. These results
colorless. Further prolonging the reaction time leads to a
slight increase of the CO2 amount, with a maximum amount
of around 80 μmol during the entire reaction. Because partial
CO2 is dissolved in the reaction solution, the free CO2 gas
detected by GC-TCD cannot reach the theoretical content
(∼100 μmol). However, an opposite reaction trend is
observed in the yield of H2 gas: almost no H2 is produced in
the first 2 h of reaction, whereas the H2 evolution sharply
increases after this critical range of reaction time. This result
indicates that, at the PNP fading stage, only CO2 gas is
generated in the reaction system.
ATR-IR and GC-MS are conducted to explore the reduction
pathway of PNP. As shown in Fig. 3d, in the absence of catalysts,
no vibration is observed within 2250–2450 cm−1 (the CO2
vibration peaks appear at around 2330 and 2360 cm−1).34 After
the PdOx/TiNT catalyst is introduced into the catalytic system,
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
PdOx/TiNTs and SF an effective platform for C–H activation.
The efficiency of the hydrogenation reaction depends on the
reduction rate of the nitro group to nitroso group, and the
terminated O atoms in the PdOx species are considered as the
essential active sites to this specific reaction. These two effects
result in the unprecedented catalytic activity and excellent
selectivity (∼100%) over the PdOx/TiNT catalyst during PNP
reduction to PAmP at room temperature. The present
work deepens our understanding on tandem catalytic
dehydrogenation and hydrogenation systems, and will benefit
the design of heterogeneous catalysts for the production of
industrially important chemicals.
Conflicts of interest
The authors declare that they have no conflict of interest.
Fig. 4 Concentration changes of control experiments: (a) PNP/PNSP
and (b) PNSP with corresponding kinetic curves monitored at 400 nm.
Reaction conditions: PNP and/or PNSP solution (10 μM) + HCOONa
(1 M) + PdOx/TiNTs.
Acknowledgements
We are grateful for financial support from the National
Science Foundation of China (21872123) and the Zhejiang
Provincial
Natural
Science
Foundation
of
China
(LY18B030007).
mean that imine is reduced to hydroxylamine and amine
rapidly, while reduction of nitro to nitroso is relatively
difficult, which is consistent with the GC-MS results.
Consequently, the rate-limiting step is most likely the
reduction process of nitro to nitroso.
Notes and references
1 S. C. Jensen, S. Bettis Homan and E. A. Weiss, J. Am. Chem.
Soc., 2016, 138, 1591–1600.
The whole reaction is shown in Fig. S12:† SF molecules
are adsorbed on PdOx/TiNTs through the formation of a
H–Pd intermediate, and the C–H bond is cleaved to form
HCO3– and *H species. PNP molecules are also adsorbed
simultaneously. Since the –NO2 group in PNP is characteristic
of its electron affinity, the N atom is prone to be adsorbed on
the terminated O atoms in PdOx due to the conjugation
effect.28,35 Thus, the –NO2 group is easily reduced into –NH2
by free *H species without the production of H2 unless the
reduction process is completely finished. Because the –NH2
group in PAmP shows electron donating ability, PAmP
molecules are readily desorbed from the catalyst surface,
leading to the high catalytic selectivity.36 Meanwhile, as a
by-product, NaHCO3 will continue to react with PAmP to
produce CO2 and sodium salt. Therefore, the present work
2 W. Shi, B. Zhang and Y. Lin, et al., ACS Catal., 2016, 6,
7844–7854.
3 M. Misra, S. R. Chowdhury and T. I. Lee, Appl. Catal., B,
2020, 272, 118991.
4 Z. Wu, X. Yuan and H. Zhong, et al., Sci. Rep., 2016, 6, 25638.
5 W. Zhang, G. Li and W. Wang, et al., Appl. Catal., B,
2018, 232, 11–18.
6 P. P. Ghimire, L. Zhang and U. A. Kinga, et al., J. Mater.
Chem. A, 2019, 7, 9618–9628.
7 I. Ibrahim, I. O. Ali and T. M. Salama, et al., Appl. Catal., B,
2016, 181, 389–402.
8 D. Bhattacharjee, K. Mandal and S. Dasgupta, RSC Adv.,
2016, 6, 64364–64373.
9 S. M. Ansar and C. L. Kitchens, ACS Catal., 2016, 6, 5553–5560.
10 F. Nawaz, Y. Xie and H. Cao, et al., Catal. Today, 2015, 258,
595–601.
provides
a new tandem catalytic system that couples
hydrogen production with organic contaminant degradation
in a “one-pot” reaction vessel.
11 B. M. Mogudi, P. Ncube and N. Bingwa, et al., Appl. Catal., B,
2017, 218, 240–248.
12 D. Hu, S. Jin and Y. Shi, et al., Nanoscale, 2017, 9,
3629–3636.
13 M. M. Mohamed and M. S. Al-Sharif, Appl. Catal., B,
2013, 142–143, 432–441.
14 C. Li, P. Wang and Y. Tian, et al., ACS Catal., 2017, 7,
5391–5398.
15 J. A. Johnson, J. J. Makis and K. A. Marvin, et al., J. Phys.
Chem. C, 2013, 117, 22644–22651.
Conclusions
In summary, PdOx/TiNTs are prepared by the IWI method and
subsequent annealing in an oxygen atmosphere, and show
high selectivity and catalytic activity for the tandem
dehydrogenation of SF and hydrogenation of PNP to PAmP
under mild conditions. Notably, SF adsorption is mainly driven
by the hydrogen bonding interactions between the H atom in
SF and surface Pd sites, which factually makes the interface of
16 Y.-Y. Shen, Y. Sun and L.-N. Zhou, et al., J. Mater. Chem. A,
2014, 2, 2977.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Catalysis Science & Technology
Paper
17 Y. Fu, L. Qin and D. Huang, et al., Appl. Catal., B, 2019, 255,
117740.
27 O. Lupan, V. Postica and M. Hoppe, et al., Nanoscale,
2018, 10, 14107–14127.
18 T. Liu, Y. Sun and B. Jiang, et al., ACS Appl. Mater. Interfaces,
2020, 12, 28100–28109.
28 Y. Zheng, L. Kovarik and M. H. Engelhard, et al., J. Phys.
Chem. C, 2017, 121, 15793–15803.
19 Q. Xiao, S. Sarina and E. R. Waclawik, et al., ACS Catal.,
2016, 6, 1744–1753.
29 A. N. Hoffman, Y. Gu and L. Liang, et al., npj 2D Mater.
Appl., 2019, 3, 50.
20 P. Song, L.-L. He and A.-J. Wang, et al., J. Mater. Chem. A,
2015, 3, 5321–5327.
30 S. Vijayalakshmi, S. Kalyanaraman and T. R. Ravindran,
Phys. Scr., 2014, 89, 095501.
21 A. Zoran, Y. Sasson and J. Blum, J. Mol. Catal., 1984, 26,
321–326.
22 A. Chen, Y. Li and J. Chen, et al., ChemPlusChem, 2013, 78,
1370–1378.
23 W. Peng, F. Zhang and G. Zhang, et al., Catal. Commun.,
2011, 12, 568–572.
24 X. Xi, T. Chen and J.-S. Zhang, et al., Chem. Commun.,
2018, 54, 1521–1524.
31 E. Spinner, J. Chem. Soc. B, 1967, 879–885.
32 I. Tajima, H. Takahashi and K. Machida, Spectrochim. Acta,
Part A, 1981, 37, 905–910.
33 K. Tedsree, A. T. S. Kong and S. C. Tsang, Angew. Chem., Int.
Ed., 2009, 121, 1471–1474.
34 C. Ehnes, M. Lucas and P. Claus, Chem. Ing. Tech., 2016, 88,
1474–1479.
35 O. Exner and S. Böhm, J. Phys. Org. Chem., 2006, 19,
1–9.
25 X. Zhu, L. Du and Z. Guo, et al., Catal. Sci. Technol., 2018, 8,
6217–6227.
26 P. Fu, P. Zhang and J. Li, Appl. Catal., B, 2011, 105, 220–228.
36 I. S. Han, C. K. Kim and C. K. Kim, et al., J. Phys. Chem. A,
2002, 106, 2554–2560.
This journal is © The Royal Society of Chemistry 2020
Catal. Sci. Technol.