Photocatalytic H2 production from aqueous solutions of hydrazine and its derivatives in the presence of nitric-acid-activated graphitic carbon nitride

Article abstract of DOI:10.1016/j.cattod.2016.12.024

A treatment of graphitic carbon nitride (GCN) with concentrated nitric acid imparts GCN with the photocatalytic activity in hydrogen evolution from aqueous solutions of hydrazine hydrate, unsymmetric dimethylhydrazine (DMH) and hydrazinium sulfate under illumination with the UV light and in the presence of a Pd/SiO₂ co-catalyst. The rate of photocatalytic water reduction was found to be maximal in the presence of 3–5 M DMH and 12–14 M hydrazine hydrate and comparable to the hydrogen evolution rate in similar systems based on TiO₂ Evonik P25. The X-ray photoelectron spectroscopy of the activated GCN revealed that the acid treatment results in the elimination of some of the nitrogen atoms in the heptazine building blocks of the GCN and formation of carboxylic groups, the modification extent increasing with an increase of both duration of the acid treatment and the GCN synthesis temperature. The activation effect observed after the acid treatment of the GCN is assumed to originate from the participation of the introduced defects in the trapping and interfacial transfer of the photogenerated charge carriers. The maximal photoactivity was observed for the activated GCN produced at 500 °C followed by a 2-h contact with the concentrated HNO₃.

Full text of DOI:10.1016/j.cattod.2016.12.024

G Model
CATTOD-10514; No. of Pages7
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Photocatalytic H production from aqueous solutions of hydrazine
2
and its derivatives in the presence of nitric-acid-activated graphitic
carbon nitride
a
a
a
Alexandra E. Raevskaya , Yaroslav V. Panasiuk , Ganna V. Korzhak ,
Oleksandr L. Stroyuk , Stepan Ya. Kuchmiy , Volodymyr M. Dzhaganb,c,
a,∗
a
Dietrich R.T. Zahnb
a
L. Pysarzhevsky Institute of Physical Chemistry of National Academy of Sciences of Ukraine, Kyiv 03028, Ukraine
Semiconductor Physics, Technische Universität Chemnitz, 09107 Chemnitz, Germany
V. E. Lashkaryov Institute of Semiconductors Physics, National Academy of Sciences of Ukraine, Kyiv 03028, Ukraine
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 June 2016
Received in revised form 12 October 2016
Accepted 16 December 2016
Available online xxx
A treatment of graphitic carbon nitride (GCN) with concentrated nitric acid imparts GCN with the pho-
tocatalytic activity in hydrogen evolution from aqueous solutions of hydrazine hydrate, unsymmetric
dimethylhydrazine (DMH) and hydrazinium sulfate under illumination with the UV light and in the
presence of a Pd/SiO2 co-catalyst. The rate of photocatalytic water reduction was found to be maximal in
the presence of 3–5 M DMH and 12–14 M hydrazine hydrate and comparable to the hydrogen evolution
rate in similar systems based on TiO2 Evonik P25.
The X-ray photoelectron spectroscopy of the activated GCN revealed that the acid treatment results in
the elimination of some of the nitrogen atoms in the heptazine building blocks of the GCN and formation
of carboxylic groups, the modification extent increasing with an increase of both duration of the acid
treatment and the GCN synthesis temperature.
Keywords:
Photocatalysis
Water splitting
Acid treatment
Unsymmetric dimethylhydrazine
X-ray photoelectron spectroscopy
The activation effect observed after the acid treatment of the GCN is assumed to originate from the par-
ticipation of the introduced defects in the trapping and interfacial transfer of the photogenerated charge
◦
carriers. The maximal photoactivity was observed for the activated GCN produced at 500 C followed by
a 2-h contact with the concentrated HNO3.
©
2016 Elsevier B.V. All rights reserved.
1
. Introduction
decomposition of toxic and persistent compounds, such as indus-
trial wastes, disarmament wastes as well as warfare agents and
chemical weapons used as sacrificial donors. Among such toxic
compounds, a special place belongs to hydrazine and its derivatives,
in particular to the unsymmetric 1,1-dimethylhydrazine (DMH)
and hydrazinium sulfate (HS).
The unsymmetric DMH is a widespread component of rocket
fuels broadly utilized in the aerospace program of the former Soviet
Union. The disarmament program launched in the 1980–90ths left
behind large stockpiles of DMH that need to be utilized, amounting
after 1992 to around 5000 tons in Ukraine (the country of origin of
some of the present authors) and more than 80,000 tons in Russia
Photocatalytic splitting of water to hydrogen and oxygen as
well as water reduction with various sacrificial electron donors
by using the inexhaustible solar energy is one of the challenges
faced by the modern photochemistry and material science. A broad
assortment of sacrificial compounds that can be used as a fuel to
reduce water−from sustainable products of agriculture to waste
by-products of various chemical industries [1–8]. A very attrac-
tive way of arranging the photocatalytic hydrogen production is
in combining the water reduction process with the simultaneous
[
9]. The DMH contacting with air oxygen either by a leakage to soil
or during burning procedures can form very carcinogenic and per-
sistent nitroso dimethylamine [9,10] and therefore other “green”
approaches to the DMH utilization are needed.
A good alternative to the thermal oxidation can be the photo-
catalytic oxidation of hydrazine and its derivatives. It was shown
∗
Corresponding author at: Laboratory of Organic Photovoltaics and Electro-
chemistry, L.V. Pysarzhevsky Institute of Physical Chemistry, National Academy of
Sciences of Ukraine, 31 Nauky av 03028, Kyiv, Ukraine.
E-mail addresses: alstroyuk@ukr.net, stroyuk@inphyschem-nas.kiev.ua
(
O.L. Stroyuk).
http://dx.doi.org/10.1016/j.cattod.2016.12.024
920-5861/© 2016 Elsevier B.V. All rights reserved.
0
Please cite this article in press as: A.E. Raevskaya, et al., Photocatalytic H production from aqueous solutions of hydrazine and its deriva-
2
tives in the presence of nitric-acid-activated graphitic carbon nitride, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.024

G Model
CATTOD-10514; No. of Pages7
ARTICLE IN PRESS
A.E. Raevskaya et al. / Catalysis Today xxx (2016) xxx–xxx
2
that by using TiO /Pt heterostructures as photocatalysts the DMH
the cocatalyst in aqueous solution of a donor (or a pure donor) was
refluxed by a magnetic stirrer and illuminated with a high-pressure
1000 W mercury lamp light in the range of ␭ > 310 nm and an inten-
sity of 50 mW/cm . The reaction mixture was evacuated prior to the
illumination. The composition of the gas phase above the solution
was determined each 30 min by the gas chromatography.
2
can be readily degraded to CO , HNO , water and N avoiding the
2
3
2
formation of the nitroso dimethylamine as a by-product [10]. At
the same time, it was shown as early as in 1982 that hydrazine can
be used as a sacrificial electron donor to photocatalytically reduce
water to H2 [11].
2
Another vital issue that should be addressed in the design of
efficient photocatalytic systems for hydrogen production is the
feasibility to carry out the photoprocess under visible light sup-
plied by the solar illumination. The issue remains a challenge till
the present day as the most active semiconductor photocatalysts
2.3. Instruments
Diffuse reflectance spectra were registered on a Shimadzu
UV-3600 spectrophotometer using BaSO4 as a standard. The dif-
fuse reflectance spectra were converted into absorption spectra
using the Kubelka-Munk function and normalized for the conve-
nience of comparison. X-ray Photoemission Spectroscopy (XPS)
investigations were performed with an ESCALAB
either can absorb only the UV light (TiO , ZnO), or are prone to
2
the photocorrosion emitting toxic metals to environment (CdS), or
have a low conduction band potential that is insufficient for water
reduction to H (WO , Fe O ) [2,3,6]. Recently, the graphitic carbon
TM
250Xi X-ray
2
3
2
3
nitride (GCN) emerged as new visible-light-sensitive photocatalyst
of water splitting and other important light-driven redox processes
Photoelectron Spectrometer Microprobe (Thermo Scientific). The
X-ray photoelectron spectra were acquired under excitation of
a 650 ␮m2 spot with the monochromatized Al K␣ X-ray source
(hꢀ = 1253.6 eV) providing a spectral resolution of 0.5 eV (pass
energy of 20 eV) in high-resolution spectra. The survey XPS spec-
tra were acquired with the pass energy of 200 eV. To prevent
possible charging of the sample it was flooded with low kinetic
energy electrons (Ekin = 0.1 eV, Iem = 0.05 mA). Spectra deconvolu-
tion and quantification were performed using the Avantage Data
System (Thermo Scientific). The number of components used for
each spectral decomposition was kept at the minimum required for
a meaningful representation of the experimental spectra. A Voigt
line shape was used to define the spectral shape of different com-
ponents. Fourier-transform infrared (FTIR) spectra were registered
using a Perkin Elmer Spectrum One spectrometer. The samples
were prepared by pressing 15 mg of the sample with 100 mg KBr at
10 Mpa. X-ray diffraction patterns of the samples were registered
on a Bruker D8 Advance diffractometer using the monochromatized
copper К␣ irradiation (␭ = 0.1540 nm).
[
12–16]. The GCN features an attractive combination of high stabil-
ity and availability, a low toxicity, relatively strong absorption of the
UV and visible light down to 460–470 nm, as well as the conduction
and valence band potentials favorable for both water reduction and
oxidation [12–14]. The GCN coupled to an appropriate cocatalyst
can act as a photocatalyst of water reduction to molecular hydrogen
with a variety of sacrificial donors, such as substituted amines, alco-
hols and inorganic sulfur-containing salts [14]. At the same time,
the feasibility of using hydrazine and its derivatives as electrons
source to reduce water to hydrogen over GCN photocatalysts is yet
to be explored.
In the present paper, we report on the photocatalytic hydro-
gen evolution from aqueous solutions of hydrazine, unsymmetric
dimethylhydrazine and hydrazinium sulfate under illumination
with the UV light. Our results showed that for the GCN to be an
efficient photocatalyst of water reduction it should be preliminary
activated by a treatment with concentrated nitric acid.
2
. Materials and methods
Melamine (1,3,5-triaminotriazine),
3. Results and discussion
hydrazine
hydrate,
Our preliminary results showed that the as-prepared GCN does
not exhibit any photocatalytic activity in molecular hydrogen evo-
lution from aqueous solutions of hydrazine and its derivatives
irrespectively of the temperature of GCN synthesis (Ts). By analogy
with previous reports [19–21], we tried to render the GCN pho-
toactive by treating it with various chemical agents. The aqueous
ammonia, hydrogen peroxide as well as the concentrated sulfuric
and nitric acids were tested and it was found that a positive result
can be achieved by bringing the powdered GCN into contact with
the concentrated HNO3 and such an activation way has not been
reported yet for the photocatalytic systems based on the GCN. The
photocatalytic activity of activated GCN (referred to as a-GCN in
the following discussion) was found to depend both on the dura-
tion of the acid treatment tat and the history of GCN preparation, in
particular, the temperature Ts of melamine condensation.
hydrazinium sulfate, unsymmetric dimethylhydrazine, con-
centrated HNO3 (65 w.%) were purchased from Sigma-Aldrich and
used without further purification. The nanocrystalline TiO2 P25
was supplied by Evonik Corp.
2.1. Preparation of GCN and acid treatment
The GCN was synthesized by the thermal condensation of
melamine in air at a different temperature (further designated
◦
as Ts) in the range of 450–600 C and atmospheric pressure for
1
h in a quartz vessel similarly to [17]. The as-prepared samples
were ground in a mortar and treated with the concentrated nitric
acid. In a typical procedure of the acid treatment, 0.5 g GCN were
immersed in 10 mL HNO3 at room temperature and left with the
magnetic stirring for a certain period of time further designated
as tat (tat = 10 min–24 h). The sample was then washed with copi-
3.1. Kinetic aspects of the photocatalytic hydrogen evolution
ous amount of distilled water till the neutral reaction, washed with
◦
2
-propanol and dried at 60 C.
Illumination of the samples of a-GCN dispersed in evacu-
ated aqueous solutions of hydrazine hydrate (HH), hydrazinium
sulfate N H × H SO (HS) or unsymmetric dimethylhydrazine
2
.2. Photocatalytic experiment
2
4
2
4
H NN(CH ) (DMH) in the presence of a Pd/SiO co-catalyst results
2
3
2
2
The process of photocatalytic hydrogen evolution was studied
in the evolution of molecular hydrogen that proceeds with a con-
stant rate (Electronic Supporting Information (ESI), Fig. S1). At the
same time, no H2 can be detected in a similar system based on the
untreated original GCN irrespectively of the parameters of its syn-
thesis. Also, only traces of hydrogen form in the absence of Pd/SiO2
indicating a crucial role of the palladium for hydrogen evolution
with the participation of a-GCN. No H2 is evolved in the dark and
in glass cylindric thermostated reactors. In a typical experiment, a
sample of GCN or TiO2 (0.02 g) was immersed into 10.0 mL aque-
ous solution of hydrazine or its derivatives and then a co-catalyst
Pd/SiO (0.01 g) was added to the reaction mixture. The co-catalyst
2
(
1 w.% palladium on commercial silica) was prepared by a chemi-
cal reduction method [18]. The suspension of a photocatalyst and
Please cite this article in press as: A.E. Raevskaya, et al., Photocatalytic H production from aqueous solutions of hydrazine and its deriva-
2
tives in the presence of nitric-acid-activated graphitic carbon nitride, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.024

G Model
CATTOD-10514; No. of Pages7
ARTICLE IN PRESS
A.E. Raevskaya et al. / Catalysis Today xxx (2016) xxx–xxx
3
under illumination in the absence of a-GCN and with only Pd/SiO2
present in the system. Also, only traces of hydrogen were detected
in the absence of an electron donor (N H and its derivatives). All
the facts indicate that hydrogen evolution is the result of pho-
tocatalytic water reduction with the participation of a-GCN as a
light-absorbing species, Pd/SiO2 as a catalyst of dark stages and
hydrazine and its derivatives as sacrificial electron donors.
Similar activation effects were also reported in the literature for
the GCN treatment with other chemical agents. In particular, a two-
fold acceleration of the photoelectrochemical hydrogen evolution
was observed in [19] after treatment of the GCN with concentrated
HCl resulting in the protonation of the amine bridging groups con-
necting the heptazine blocks of the GCN. An enhancement of the
photocatalytic water reduction was observed after a hydrothermal
treatment of the GCN in the presence of ammonium thiosulfate
higher co-catalyst content−almost to zero in the presence of 0.04 g
Pd/SiO (Fig. S3b). Both dependencies originate most probably from
2
the light shielding effect. At the photocatalyst content higher than
0.02 g the incoming light can be completely absorbed in a thin near-
surface layer of the reactor while the rest of the reaction mixture
remains unaffected by the light flux. In the case of Pd/SiO2 that
absorbs and scatters quite strongly the UV and visible light, an
excessive loading of the co-catalyst can shield the reaction mixture
from the incoming light thus decreasing considerably the efficiency
of the photocatalytic process.
2
4
3.1.3. The content of electron donors
The a-GCN exhibits a negligible photocatalytic activity in water
reduction in the absence of any electron donors. As the HH or DMH
is introduced into the system the hydrogen evolution begins and
proceeds with a steady rate. The dependence of RH on the concen-
tration of both donors is dome-shaped (Fig. 1b), the highest rate
reached at around 12–14 M donor content in the case of HH (Fig. 1b,
curve 1) and at 3–5 M in the case of DMH (curve 2). For the sake of
comparison the rates of hydrogen evolution in the similar systems
based on TiO2 P25 is presented in Fig. 1b. As can be seen, the rate
of photocatalytic hydrogen evolution is somewhat larger for the
P25 at low donor concentrations. However, at a higher HH/DMH
content the P25 titania rapidly loses its activity, most probably as
a result of a pH increase, and at [HH, DMH] higher than 4 M TiO2 is
completely deactivated (Fig. 1b). In contrast, the a-GCN functions
efficiently in such conditions and tolerates a further increase in the
concentration of both donors retaining a reasonable photocatalytic
activity even in the pure liquid HH and DMH (the cases correspond
to [HH] = 20.4 M and [DMH] = 13.2 M, Fig. 1b). In the case of HS the
[
20]. Etching of the GCN with the ammonia resulted in a consid-
erable increase of its photocatalytic activity in water reduction to
molecular hydrogen [21]. At the same time no reports on using the
concentrated HNO3 to activate the GCN toward the photocatalytic
water reduction could be found in the literature on the moment of
the present paper preparation.
Fig. 1a shows the photoprocess rate achieved with various
donors with a-GCN and TiO2 Evonik P25, the latter used as a
benchmark photocatalyst. The HH was found to be the least active
electron donor and the a-GCN/HH-based system showed roughly
twice lower efficiency of the hydrogen evolution as compared to
the similar TiO /HH-based system. The HS is a more active donor
2
providing almost the same efficiency of hydrogen evolution both
for a-GCN and TiO . The photocatalytic H evolution with the par-
2
2
ticipation of a-GCN and DMH proceeds with the rate that is more
than 4 times higher than that for the unsubstituted hydrazine. In
the DMH-based system, the a-GCN is almost as active as TiO2 P25,
the latter typically showing superior photocatalytic activities in
hydrogen evolution to most of the reported semiconductor pho-
tocatalysts. It should be also noted, that at elevated concentrations
of donors, that is at a high pH, TiO2 rapidly loses its activity while
a-GCN remains stable and active (see discussion below).
We have varied basic parameters of the photocatalytic system
such as a donor type and concentration, temperature, photocatalyst
mass, light intensity as well as Ts and tat to find the optimal con-
ditions and to reach the maximal efficiency of the photocatalytic
process.
rate of photocatalytic H production grows with an increase in [HS],
2
reaches a saturation value at [HS] > 0.1 M and does not change fur-
ther (ESI, Fig. S4), the a-GCN and TiO2 P25 revealing almost the
same efficiency.
3.1.4. The synthesis temperature of original GCN and duration of
the acid treatment
It was found that the temperature of the synthesis of original
GCN that was used for the preparation of photocatalysts influences
quite strongly the photocatalytic activity of the resulting a-GCN.
The rate of H2 evolution increases to some extent as a result of
◦
the Ts elevation from 475 to 500 C and decreases drastically at a
◦
higher synthesis temperature reaching at Ts = 600 C only around
10% of the RH observed at Ts = 500 C (Fig. 2a). The dependences
◦
3.1.1. The light intensity and temperature
The rate of photocatalytic hydrogen production increases with
of the same shape were obtained also for the HH and DMH (not
shown).
an increase of the exciting light intensity in an almost linear manner
throughout the studied range of up to 50 mW/cm2 (ESI, Fig. S2a).
The hydrogen evolution accelerates by a factor of around four with
The dependences of RH on the duration of post-synthesis GCN
acid treatment are also similar for all three donors and reveal quite a
sharp maximum at tat = 2 h (Fig. 2b). A more prolonged contact with
the concentrated HNO3 results in a decrease of the photocatalytic
activity of a-GCN. Such a dome-shaped dependence and a peak at
tat = 2 h were observed for the GCN samples synthesized at different
temperatures.
an increase of the temperature of the reaction mixture from 20 to
◦
5
0 C (ESI, Fig. S2b, curve 1). The temperature range was dictated
by the leakage of vacuum grease at a higher T and the moisture
condensation on the outer reactor walls at a lower T. An estima-
tion of the activation energy basing on a linearization of the RH(T)
dependence as ln(RH) vs 1000/T (ESI, Fig. S2b, curve 2) showed it
to be as small as 25−30 kJ/mole. Such activation energy is typical
for the photocatalytic H2 evolution systems where limiting stages
are formation and desorption of H2 molecules from the co-catalyst
surface [18].
3.2. The characterization of a-GCN samples and origins of the
activation effect
The acid treatment of the GCN results in a “blue” shift of the
fundamental absorption band edge, however, the shift magnitude
does not correlate with the duration of the treatment. In the case of
3
.1.2. The photocatalyst and co-catalyst content
The dependences of RH both on the a-GCN and Pd/SiO2 con-
◦
Ts = 500 C the band edge of original GCN can be observed at 440 nm
tent are dome-shaped (ESI, Fig. S3). The rate of hydrogen evolution
grows with an increase in the photocatalyst content, reaches a
peak value at 0.02 g (per a 10-mL reactor) and then decreases (Fig.
S3a). Similarly, the RH reaches a saturation value at an increase of
the Pd/SiO2 content to 0.0075–0.010 g and drops drastically at a
(ESI, Fig. S5a, curve 1) corresponding to the band gap Eg of 2.82 eV.
A 30-min contact with HNO3 results in a band edge shift to 415 nm
(Fig. S5a, curve 2, Eg = 2.99 eV) and further treatment for 24 h only
shifts the band edge to 410 nm (Fig. S5a, curve 4, Eg = 3.03 eV). It
can be concluded that the contact with HNO3 results in a partial
Please cite this article in press as: A.E. Raevskaya, et al., Photocatalytic H production from aqueous solutions of hydrazine and its deriva-
2
tives in the presence of nitric-acid-activated graphitic carbon nitride, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.024

G Model
CATTOD-10514; No. of Pages7
ARTICLE IN PRESS
A.E. Raevskaya et al. / Catalysis Today xxx (2016) xxx–xxx
4
Fig. 1. (a) Rate RH of photocatalytic hydrogen formation in different systems. (b) Dependence of RH on HH (curve 1) and DMH concentration (curve 2). Conditions: 0.02 g
◦
◦
a-GCN, 0.01 g Pd/SiO2, T = 20 C. Photocatalyst: Ts = 500 C, tat = 2 h.
◦
Fig. 2. RH as a function of temperature of the GCN synthesis Ts (a) and duration of the acid treatment tat (b). Conditions: 0.02 g a-GCN, 0.01 g Pd/SiO2, [HS] = 0.2 M, T = 20 C.
◦
Photocatalyst: in (a) tat = 2 h, in (b) Ts = 500 C.
destruction of the aromatic heptazine framework and a shrinking of
the conjugation area with a corresponding increase of the interband
electron transition energy [19,22,23]. However, the depth of such
destruction for different tat cannot be estimated from the optical
absorption spectra.
be discerned. It should also be mentioned that estimations of the
thickness of the GCN and a-GCN particles made from the spectral
width of the main XRD peak using the Scherrer equation showed
that both values reside in a nanometer range changing from 6.5 nm
for the original GCN to 4.8 nm for the a-GCN produced at tat = 2 h.
Scanning electron microscopy (SEM) images (ESI, Fig. S7)
The XRD pattern of starting GCN (ESI, Fig. S5b, curve 1) reveals
two main peaks characteristic for the graphitic carbon nitride. The
◦
showed that the a-GCN (Ts = 500 C, tat = 2 h) consists of microscopic
◦
first peak at 27.3 reflects a layer-to-layer periodicity in the GCN
formations with a considerable portion of platelet-shaped parti-
cles. Many particles visible on the images reveal a lamellar periodic
structure that most probably originates from the inherently lay-
ered character of a-GCN. The similar lamellar structure was also
with an interlayer distance of 3.26 Å, while the second one at around
◦
1
3
corresponds to a periodicity between regular heptazine units
within the single layers of carbon nitride (one of the most plausi-
ble structures of GCN single layer is provided in ESI, Fig. S6) with a
period of 6.8 Å. The acid treatment of the GCN results in a consider-
able decrease of the intensity of both peaks (ESI, Fig. S5b, curve 2)
and some shrinking of the interlayer distance to 3.22 Å. The changes
indicate that the contact with HNO3 results in a disruption of both
periodic alteration of the single layers and the regularity within the
GCN single layers [19,20,24,25]. However, as in the above case of
optical absorption experiments, no clear dependence of the mag-
nitude of such changes and the duration of the acid treatment can
◦
reported in [26] for the GCN produced at 500 C via the conden-
sation of melamine and cyanuric chloride. As suggested by XRD,
the microscopic formations visible on the SEM images consist of
smaller and much thinner platelets with the thickness of around
5 nm that cannot be resolved by SEM.
Such very thin a-GCN platelets can be observed by TEM (ESI, Fig.
S8a,b) on the edges of large agglomerates that cannot be penetrated
by the electron beam. The HRTEM images of separate platelets (ESI,
Fig. S8c,d) showed the a-GCN platelets to be 5–6 nm-thick, in accor-
Please cite this article in press as: A.E. Raevskaya, et al., Photocatalytic H production from aqueous solutions of hydrazine and its deriva-
2
tives in the presence of nitric-acid-activated graphitic carbon nitride, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.024

G Model
CATTOD-10514; No. of Pages7
ARTICLE IN PRESS
A.E. Raevskaya et al. / Catalysis Today xxx (2016) xxx–xxx
5
dance with the XRD results, and exhibiting a layered structure with
an interlayer distance of around 0.4 nm typical for the GCN.
It should also be noted that the TEM/HRTEM images of a mixture
of a-GCN and Pd/SiO2 presented in Fig. S8 (ESI) were taken after
the completion of the photocatalytic test. No additional higher-
contrast (darker) spots can be observed on separate a-GCN platelets
Ts and tat. The assignment of the components of C1s, N1s, and O1 s
bands was carried out using the reported XPS data pools [28,29]
as well as the reports on the GCN photocatalysts activated by the
thermal treatment [23,24,27] and chemical agents [19–21,25,30].
The N1 s band in the X-ray photoelectron spectra of the
untreated GCN samples (Fig. 3a) can be in all cases approximated
by a combination of four components with maxima at 397.2 eV,
398.2 eV, 399.8 eV, and 403.3 eV. The first peak at 397.2 eV can be
assigned to the binding energy of 1s electrons in the nitrogen atoms
(
Fig. S8a,b) that can be assigned to Pd nanoparticles indicating that
no migration of palladium from the microcrystalline Pd/SiO2 to
a-GCN took place during the photoreaction as a result of possi-
ble Pd dissolution/photo-redeposition. Also, the examination of the
mixure with SEM/EDX (ESI, Fig. S9) showed that palladium is only
present on the high-contrast areas assigned to the Pd/SiO2 beads
and is not observed on other spots of the sample, thus confirming
that no migration of Pd from the co-catalyst to the photocatalyst
occurred during the photocatalytic process.
of C
N C fragments of the heptazine cycles (pyridinic-N, high-
lighted by the blue color in Fig. S6, ESI). The feature at 398.2 eV is
typically ascribed to the nitrogen atoms bound with three C atoms
in the heptazine heterocycles (NC fragment). In a broadly accepted
3
model of a GCN single layer such atoms can be found in the cen-
ter of heptazine heterocycles (highlighted by the green color in
Fig. S6, ESI). The peak at 399.8 eV can be assigned to the nitrogen
atoms in –NH– and –NH2 fragments of GCN (highlighted by the
red color in Fig. S6, ESI). A broad band at 402–407 eV is typically
assigned to plasmon excitation of the aromatic system of heptazine
heterocycles.
◦
The original GCN produced at Ts = 500 C has a comparatively
2
low specific surface area of 10 m /g as revealed by an analysis of
the nitrogen adsorption/desorption isotherms (ESI, Fig. S10). The
acid treatment of the GCN results in an increase of the surface area
2
to 23, 36, and 39 m /g for the 2-h, 5-h, and 24-h treatment, respec-
tively (Fig. S10). The a-GCN samples are, therefore, more disordered
than the original GCN, in accordance with the XRD data, but have
approximately the same specific surface area. The isotherms reveal
a hysteresis indicating the presence of mesopores. The size of meso-
pores varies in the range of 20–30 nm indicating that the porosity of
a-GCN samples originates from the inter-platelet voids rather than
from the presence of real mesopores.
The C1s band of the GCN samples can be deconvoluted into four
separate components peaked at 284.5 eV, 287.8 eV, 288.8 eV, and
293 eV (Fig. 3b). The most intense peak at 287.8 eV is characteristic
for aromatic carbon in the heptazine heterocycles (N
C N frag-
ments). The peak at 284.5 eV corresponds to the 1s electron binding
3
energy in the carbon atoms of single C C bonds (sp -C) originating
from aliphatic residuals and adventitious carbon (absorbed by the
samples from the environment). The low-intensity peak at 288.8 eV
is typical for the carbon atoms in carboxylic groups. Similarly to the
N1s series, the broad band peaked at 293 eV is ascribed to plasmon
excitation of the heptazine heterocycles.
Each GCN sample revealed also a low-intensity O1s band that
combines two components with maxima at 530.3 eV and 531.4 eV
(Fig. 3a, insert) typical for oxygen atoms in C O and C O groups,
respectively.
Relative contributions of the above-discussed components into
the N1s and C1s bands were calculated as a ratio of the area of the
selected components to the integral area of the N1s, C1s or O1s
bands for a given sample (ESI, Table S3). The major part of nitrogen
It was also found that the original GCN and the product of a 2-h
acid treatment are characterized by almost identical FTIR spectra
(
ESI, Fig. S5c), which is also the typical situation for the thermally-
treated [24,27] and chemically-treated GCN [19–21]. Taking into
account the XRD, SEM and FTIR results we can only conclude
that the structural changes in GCN introduced by the acid treat-
ment occur mostly in the surface layer of the GCN particles and
the concentration of the introduced defects (new functionalities)
is low with respect to the overall GCN volume. By this reason,
we subjected the a-GCN samples produced in different conditions
to a study with the X-ray photoelectron spectroscopy which is a
surface-sensitive method allowing to probe predominantly a thin
surface layer of the GCN and a-GCN particles.
atoms (∼70%) in the original GCNs belong to the C
N C groups. A
Survey X-ray photoelectron spectra of original GCN samples
revealed two signals originating from carbon and nitrogen, and a
low-intensity peak indicating the presence of oxygen admixture
minor portion of the nitrogen atoms (∼20%) occupies central posi-
tions in the heptazine heterocycles binding to three neighboring
carbon atoms. Around 10% of the nitrogen atoms belong to the
(
ESI, Fig. S11). An atomic carbon-to-nitrogen ratio C/N was found
bridging C NH groups that connect heptazine cycles and form the
2
to be 0.68 for all the untreated GCN samples (ESI, Table S1). The
value is somewhat lower than that expected for the stoichiometric
C N , C/N = 0.75, attesting to the presence of an excess of nitrogen
infinite poly heptazine chains (ESI, Fig. S6). A major part of the
carbon atoms in the GCN sample (∼93%) belongs to the heptazine
heterocycles. Besides, the sample reveals the presence of adventi-
tious carbon (∼5%) and a very small amount of carboxylic/carbonyl
groups that can be regarded as point defects in the poly heptazine
structure.
3
4
and, therefore, to an incomplete character of the melamine con-
densation. The GCN samples reveal the presence of a small amount
of oxygen, around 1%, incorporated into the carboxylic groups (see
discussion below). The acid treatment of GCN in all cases results in
an increase of the C/N ratio to 0.70–0.73, depending on Ts. Simul-
taneously, the oxygen content increases from 1 to 3% after the acid
treatment (Table S1). The observations indicate that the samples
lose some nitrogen atoms and gain additional carboxylic groups
It was found that even a relatively short, 0.5 h, contact of the
GCN with the concentrated HNO results in considerable changes in
3
the sample structure (ESI, Table S3) that become more pronounced
at a longer acid treatment (Fig. 3c,d). In particular, a decrease of
the relative contribution of pyridinic-N from 70% to 56% and a
corresponding increase of the relative content of nitrogen in NC3
fragments from 20% to 34% can be observed. A more prolonged
acid treatment results in a further decrease of the relative content
of the pyridinic nitrogen – down to 48% for the 24-h contact with
the nitric acid. These observations are in accordance with the anal-
ysis of the C1s series indicating that the acid treatment induces a
decrease of the relative content of the aromatic carbon from 93% in
the untreated sample to 81% for the a-GCN sample contacting with
the acid for 24 h. Simultaneously, an increase in the relative content
of carbon atoms bound in the carboxylic groups is observed- from
2% for the original GCN to 15% for the a-GCN sample produced by the
as a result of the contact with HNO . For each particular GCN the
3
atomic C/N ratio increases in the first hours of the acid treatment,
however, returns to the starting value after a more prolonged con-
tact between the GCN and concentrated HNO3 (see in ESI, Table S2
◦
for the sample produced at Ts = 500 C). The oxygen content shows
a steady increase as the acid treatment proceeds and reaches 4%
after the 24-h acid treatment.
Examination of high-resolution X-ray photoelectron spectra in
the characteristic range of N1s, C1s, and O1 s electron binding ener-
gies allowed to shed light on changes in the GCN structure produced
by the contact with the concentrated nitric acid both as a function of
Please cite this article in press as: A.E. Raevskaya, et al., Photocatalytic H production from aqueous solutions of hydrazine and its deriva-
2
tives in the presence of nitric-acid-activated graphitic carbon nitride, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.024

G Model
CATTOD-10514; No. of Pages7
ARTICLE IN PRESS
A.E. Raevskaya et al. / Catalysis Today xxx (2016) xxx–xxx
6
Fig. 3. High-resolution X-ray photoelectron spectra of GCN (sections a, b) and a-GCN (sections c, d) in the range of N1s (a, c) and C1s (b, d) electron binding energy. Insert in
◦
(
b): O1s XPS band of a-GCN sample. Preparation conditions: Ts = 500 C, tat = 2 h.
2
4-h acid treatment. At that, the relative contribution of the adven-
result in similar changes, that is, in a partial destruction of the
heptazine heterocycles with the elimination of the C C frag-
ments and the formation of various structural defects, in particular,
nitrogen vacancies and carboxylic groups. The post-synthesis acid
treatment of such defect-enriched samples results in the formation
of additional defects in the structure of GCN samples.
3
titious carbon (sp -C) remains low and almost steady, around 5%.
Therefore, the XPS results show that interaction of the GCNs with
nitric acid results in a partial destruction of the heptazine heterocy-
cles of the GCN single layers in the near-surface area of the samples
with a decrease of the nitrogen content and formation of structural
defects. One of the possible defects detected by the XPS is the car-
boxylic group. Similar effects were observed in [24] to be produced
N
Summarizing, the XPS results show that the photoactivity of
a-GCN originates most probably from new structural defects intro-
duced by the acid treatment. The chemical nature of the defects can
be versatile and a particular sort of defects that can be observed by
XPS is the carboxylic groups, but a contribution of other species
cannot be ruled out. The defects can apparently serve as traps for
the photogenerated charge carriers slowing down their recombina-
tion and allowing for the more efficient interfacial charge transfer
to the reactants to occur. Most probably, as is the case for inor-
ganic semiconductor photocatalysts, the structural defects can also
participate in the recombination processes. Therefore, an optimal
defect density exists yielding the maximal photocatalytic activity,
while at a higher concentration of defects the recombination pro-
cesses prevail and the photoactivity of a photocatalyst deteriorates.
In the case of GCN, which is characterized by a complex structure
(molecular polyheptazine units constituting infinite sheets where
a periodic potential for the photogenerated charges is created) and
a versatility of possible defects, the introduction of a new defect can
either quench PL or, in some cases, result in the PL enhancement.
For example, we have recently reported on the PL properties of GCN
produced at different condensation temperature [31] and found
that the higher is the synthesis T, the lower is the PL intensity of GCN
as a result of the formation of additional defects during the thermal
treatment. However, after the exfoliation of the GCN to single-layer
and a-few-layer sheets this dependence has reversed – the exfoli-
ation products of GCN produced at a higher T revealed a higher PL
efficiency reflecting a versatility of defects that can participate in
radiative processes and a strong influence of the environment of the
photophysical processes in GCN. Also, it is widely reported for the
graphene quantum dots and carbon nanoparticles that the intro-
duction of additional hydroxyl/carboxyl groups to their basal plane
or surface, respectively, results in an enhancement of fluorescence
[32]. So, the COH/COOH groups can act in these cases not only as
◦
by a thermal treatment of the GCN at 660 C. The formation of struc-
tural defects was supposed to be the reason for an increase of the
photocatalytic activity of the treated GCN in water reduction and
Rhodamine B decomposition [24] as well as N2 reduction [27].
A relative content of different structural units for a series of the
GCN samples produced at a different Ts – from 475 C to 600 C
as well as for the corresponding a-GCNs (tat = 2 h) is provided in
ESI (Table S4). The untreated GCN samples are characterized by a
different oxygen content and by some differences in the content of
aromatic carbon and the two-coordinated nitrogen. In particular,
the fraction of pyridinic nitrogen increases from 53% to 70% at an
elevation of the Ts from 475 to 500 C and decreases at a higher
temperature – down to 51% for Ts = 600 C. In a similar manner,
the maximal relative content of the aromatic carbon in the studied
series, 93%, is observed for the GCN sample produced at 500 C. At
the same time, an increase of Ts from 475 C to 600 C results in
an increase of the relative content of the carbon atoms bound to
the oxygen atoms- from 2% for Ts = 475–500 C to 8% for the GCN
◦
◦
◦
◦
◦
◦
◦
◦
◦
sample synthesized at 600 C.
The acid treatment of all the samples results in changes in their
structure that show a similar trend. In particular, we observed a
decrease of the relative contribution of the pyridinic nitrogen and
an increase in the relative content of the NC3 fragments, as well as
a decrease in the relative contribution of the aromatic carbon and
an increase of the population of the carbon atoms bound to oxygen
(
ESI, Table S4). The latter trend is most vividly pronounced for the
◦
GCN sample produced at 500 C, where the acid treatment results
in an increase of the relative carbon content in the C OH/C O frag-
ments from 2 to 7%. We can conclude, therefore, that an increase
in the temperature of GCN synthesis from 500 C to 600 C and the
post-synthesis acid treatment of the sample produced at Ts = 500 C
◦
◦
◦
Please cite this article in press as: A.E. Raevskaya, et al., Photocatalytic H production from aqueous solutions of hydrazine and its deriva-
2
tives in the presence of nitric-acid-activated graphitic carbon nitride, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.024

G Model
CATTOD-10514; No. of Pages7
ARTICLE IN PRESS
A.E. Raevskaya et al. / Catalysis Today xxx (2016) xxx–xxx
7
point defects capable of charge trapping but also as additional sites
of radiative recombination.
photocatalytic materials for the degradation of contaminants of
environmental concern. The work was supported by the com-
plex program of the fundamental research of National Academy
of Sciences of Ukraine “Fundamental problems of new nanoma-
terials and nanotechnologies” (project #32/15-Н). The authors
acknowledge Dr. A.V. Shvets (L.V. Pysarzhevsky Institute of Physical
Chemistry of Nat. Acad. Sci. of Ukraine) for the data of XRD and FTIR
spectroscopy and Dr. M. Skoryk (LLC NanoMedTech, Kyiv, Ukraine)
for SEM data.
A direct evidence of the increased charge trapping by shal-
low traps of the a-GCN can be found in photoluminescence (PL)
measurements (ESI, Fig. S12). The acid treatment results in a con-
siderable growth of the PL intensity even at early stages of the acid
treatment (the first 10 min). However, at a more prolonged acid
treatment, the PL intensity remains steady and almost does not
vary with the treatment time. At the same time, at a certain treat-
ment duration, around 2 h and longer, the photocatalytic activity
also reaches a saturation level and further generation of the defects
only enhances the radiationless recombination thus resulting in a
decrease of the photocatalytic activity.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.cattod.2016.12.
024.
A similar situation is most probably realized by the variation
of the synthesis temperature. Formation of structural defects by
annealing of the GCN both in the air and in an inert atmosphere
is a well-reported way of influencing the photocatalytic properties
of GCN [24,26,27]. In particular, thermal activation of the GCN as
a photocatalyst of hydrogen evolution observed in [22,23,26] was
mainly ascribed to the generation of structural defects.
References
[
[
[
1] A. Kudo, Pure Appl. Chem. 79 (2007) 1917–1927.
2] A. Kudo, Y. Miseki, Chem. Soc. Rev. 38 (2009) 253–278.
3] X. Chen, S. Shen, L. Guo, S.S. Mao, Chem. Rev. 110 (2010) 6503–6570.
◦
With an increase of Ts in the range of 500–650 C, the density
[4] D. Jing, L. Guo, L. Zhao, X. Zhang, H. Liu, M. Li, S. Shen, G. Liu, X. Hu, X. Zhang, K.
Zhang, L. Ma, P. Guo, Inter. J. Hydrogen Energy 35 (2010) 7087–7097.
of structural defects in the GCN is expected to increase as well. We
can, therefore, anticipate that the introduction of the defects makes
the GCN more susceptible to the attack of the nitric acid and the fol-
lowing acid treatment should produce more additional defects for
the samples synthesized at a higher temperature which is in accor-
dance with the XPS data. As a result, we observe the dome-shaped
dependencies of the hydrogen evolution rate both on the acid treat-
ment duration and the synthesis temperature of the starting GCN
[
[
5] K. Maeda, K. Domen, J. Phys. Chem. Lett. 18 (2010) 2655–2661.
6] O.L. Stroyuk, A.I. Kryukov, S. Ya Kuchmiy, V.D. Pokhodenko, Theor. Exp. Chem.
4
5 (2009) 199–222.
[7] J. Ran, J. Zhang, J. Yu, M. Jaroniec, S.Z. Qiao, Chem. Soc. Rev. 43 (2014)
787–7812.
7
[
[
8] M.R. Gholipour, C.-T. Dinh, F. Beland, T.-O. Do, Nanoscale 7 (2015) 8187–8208.
9] V.G. Voloshchuk, Y.L. Yagupolskyi, Y.V. Tanchuk, V.M. Boyko, R.V. Bubnov,
EPMA J. 5 (2014) A92.
[
10] P.A. Kolinko, D.V. Kozlov, A.V. Vorontsov, S.V. Preis, Catal. Today 122 (2007)
78–185.
1
◦
samples with the peaks at tat = 2 h and Ts = 500 C.
[
[
11] Y. Oosawa, J. Chem. Soc. Chem. Commun. (1982) 221–222.
12] A. Thomas, A. Fischer, F. Goettmann, M. Antonietti, J.-O. Müller, R. Schlögl, J.M.
Carlsson, J. Mater. Chem. 18 (2008) 4893–4908.
4
. Conclusion
[13] X. Wang, S. Blechert, M. Antonietti, ACS Catal. 2 (2012) 1596–1606.
[
[
14] S. Cao, J. Yu, J. Phys. Chem. Lett. 5 (2014) 2101–2107.
15] F. Fresno, R. Portela, S. Suarez, J.M. Coronado, J. Mater. Chem. A 2 (2014)
The graphitic carbon nitride treated with the concentrated nitric
2863–2884.
acid revealed a photocatalytic activity in hydrogen evolution from
aqueous solutions of hydrazine, unsymmetric dimethylhydrazine
and hydrazinium sulfate in the presence of a Pd/SiO2 co-catalyst.
The rate of photocatalytic water reduction was found to be maxi-
mal in the presence of 3–5 M DMH and 12–14 M hydrazine hydrate
and comparable to the hydrogen evolution rate in similar systems
[16] Y. Zheng, L. Lin, B. Wang, X. Wang, Angew. Chem. Int. Ed. 54 (2015)
12868–12884.
[
17] Ya.V. Panasiuk, A.E. Raevskaya, O.L. Stroyuk, P.M. Lytvyn, S. Ya Kuchmiy, RSC
Adv. 5 (2015) 46843–46849.
[
18] A.V. Korzhak, N.I. Ermokhina, O.L. Stroyuk, V.K. Bukhtiyarov, A.E. Raevskaya,
V.I. Litvin, S. Ya Kuchmiy, V.G. Ilyin, P.A. Manorik, J. Photochem. Photobiol. A
198 (2008) 126–134.
[
19] Y. Zhang, A. Thomas, M. Antonietti, X. Wang, J. Am. Chem. Soc. 131 (2009)
50–51.
based on titania Evonik P25. However, in contrast to TiO , the acti-
2
vated GCN retains photocatalytic activity at higher concentrations
of electron donors and can operate even in the pure liquid hydrazine
hydrate and DMH while the TiO2 P25 is completely deactivated in
such alkaline media.
[20] Z. Hong, B. Shen, Y. Chen, B. Lin, B. Gao, J. Mater. Chem. A 1 (2013)
1754–11761.
21] P. Yang, J. Zhao, W. Qiao, L. Li, Z. Zhu, Nanoscale 7 (2015) 18887–18890.
[22] H. Zhang, A. Yu, J. Phys. Chem. C 118 (2014) 11628–11635.
[23] D. Zheng, C. Huang, X. Wang, Nanoscale 7 (2015) 465–470.
1
[
[
[
24] P. Niu, G. Liu, H.-M. Cheng, J. Phys. Chem. C 116 (2012) 11013–11018.
25] W. Wu, J. Zhang, W. Fan, Z. Li, L. Wang, X. Li, Y. Wang, R. Wang, J. Zheng, M.
Wu, H. Zeng, ACS Catal. 6 (2016) 3365–3371.
The rate of photocatalytic hydrogen evolution from aqueous
solutions of hydrazine and its derivatives was found to depend on
the duration of the treatment with the nitric acid tat as well as on
the GCN synthesis temperature Ts. Dependences of the photopro-
cess rate on tat and Ts are dome-shaped with maxima at tat = 2 h
[26] Y. Cao, Z. Zhang, J. Long, J. Liang, H. Lin, H. Lin, X. Wang, J. Mater. Chem. A 2
2014) 17797–17807.
(
[
[
27] G. Dong, W. Ho, C. Wang, J. Mater. Chem. A 3 (2015) 23435–23441.
28] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, Handbook of X-Ray
Photoelectron Spectroscopy: A Reference Book of Standard Spectra for
Identification and Interpretation of Xps Data, Physical Electronics, 1995.
29] A.V. Naumkin, A. Kraut-Vass, S.W. Gaarenstroom, NIST X-ray Photoelectron
Spectroscopy Database, NIST, 2012, Stand. Ref. Database 20, Version 4.1.
30] Y. Li, H. Xu, S. Ouyang, D. Lu, X. Wang, D. Wang, J. Ye, J. Mater. Chem. A 4
(2016) 2943–2950.
◦
and Ts = 500 C.
An X-ray photoelectron spectroscopic study revealed that the
GCN acid treatment results in the elimination of some of the nitro-
gen atoms in the heptazine building blocks of GCN and formation of
carboxylic groups. The amount of introduced defects and oxygen-
containing groups increases with an increase of both tat and Ts. The
GCN activation effect induced by the acid treatment was supposed
to originate from the participation of the introduced defects/COOH
groups in the trapping and interfacial transfer of the photogener-
ated charge carriers.
[
[
[
[
31] A.L. Stroyuk, Ya.V. Panasiuk, A.E. Raevskaya, S.Ya. Kuchmiy, Theor. Exp. Chem.
51 (2015) 243–251.
32] S.Y. Lim, W. Shen, Z. Gao, Chem. Soc. Rev. 44 (2015) 362–381.
Acknowledgements
The authors wish to express their deepest and sincerest recog-
nition of Prof. András Dombi as a key figure in the topic of
Please cite this article in press as: A.E. Raevskaya, et al., Photocatalytic H production from aqueous solutions of hydrazine and its deriva-
2
tives in the presence of nitric-acid-activated graphitic carbon nitride, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.024