An eco-friendly, highly stable and efficient nanostructured p-type N-doped ZnO photocatalyst for environmentally benign solar hydrogen production

Article abstract of DOI:10.1039/c2gc35519a

We have investigated an economical green route for the synthesis of a p-type N-doped ZnO photocatalyst by a wet chemical method. Significantly, hazardous H₂S waste was converted into eco-friendly hydrogen energy using the p-type N-doped ZnO photocatalyst under solar light, which has previously been unattempted. The as-synthesized p-type N-doped ZnO shows a hexagonal wurtzite structure. The optical study shows a drastic shift in the band gap of the doped ZnO in the visible region (3.19-2.3 eV). The doping of nitrogen into the ZnO lattice is conclusively proved from X-ray photoelectron spectroscopy analysis and Raman scattering. The morphological features of the N-doped ZnO are studied from FESEM, TEM and reveal particle sizes to be in the range of ～4-5 nm. The N-doped ZnO exhibits enhanced photocatalytic hydrogen generation (～3957 μmol h^-1) by photodecomposition of hydrogen sulfide under visible light irradiation, which is much higher as compared to semiconductor metal oxides reported so far. It is noteworthy that a green catalyst is investigated to curtail H₂S pollution along with production of hydrogen (green fuel) using solar light, i.e., a renewable energy source. The green process investigated will have the potential to synthesize other N-doped metal oxides.

Full text of DOI:10.1039/c2gc35519a

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2012, 14, 2790
www.rsc.org/greenchem
PAPER
An eco-friendly, highly stable and efficient nanostructured p-type N-doped
ZnO photocatalyst for environmentally benign solar hydrogen production†
Ashwini P. Bhirud,a Shivaram D. Sathaye,b Rupali P. Waichal,c Latesh K. Nikama and Bharat B. Kale*a
Received 7th April 2012, Accepted 10th July 2012
DOI: 10.1039/c2gc35519a
We have investigated an economical green route for the synthesis of a p-type N-doped ZnO photocatalyst
by a wet chemical method. Significantly, hazardous H2S waste was converted into eco-friendly hydrogen
energy using the p-type N-doped ZnO photocatalyst under solar light, which has previously been
unattempted. The as-synthesized p-type N-doped ZnO shows a hexagonal wurtzite structure. The optical
study shows a drastic shift in the band gap of the doped ZnO in the visible region (3.19–2.3 eV). The
doping of nitrogen into the ZnO lattice is conclusively proved from X-ray photoelectron spectroscopy
analysis and Raman scattering. The morphological features of the N-doped ZnO are studied from
FESEM, TEM and reveal particle sizes to be in the range of ∼4–5 nm. The N-doped ZnO exhibits
enhanced photocatalytic hydrogen generation (∼3957 μmol h−1) by photodecomposition of hydrogen
sulfide under visible light irradiation, which is much higher as compared to semiconductor metal oxides
reported so far. It is noteworthy that a green catalyst is investigated to curtail H2S pollution along with
production of hydrogen (green fuel) using solar light, i.e., a renewable energy source. The green process
investigated will have the potential to synthesize other N-doped metal oxides.
production and H2S decomposition is a highly desirable process.
H2S can be effectively decomposed by photocatalytic processes,
1. Introduction
Major natural energy sources, namely oil, are depleting fast.
Therefore, researchers across the world are focusing their efforts
on renewable alternative sources of energy. Simultaneously,
environmental problems are becoming more and more serious
with the development of industry and economy. Therefore, the
emphasis of using conventional physics and chemistry has
shifted to the use of unconventional green routes in the processes
of tapping various energy sources. Direct splitting of H2O, H2S
or an organic waste using a particulate photocatalyst remains a
promising route to producing H2 that is environmentally clean
and a recyclable form of energy. Water is an ideal hydrogen
source and therefore, photocatalysts that could produce hydrogen
from water splitting under sunlight have been actively sought.1–3
Apart from water, H2S could become an alternative source of
hydrogen as H2S splitting requires less energy compared with
the photocatalytic decomposition of water. H2S is cheaply avail-
able in large quantities as a byproduct in coal and petroleum
industries, natural gas, oil wells and geothermal plants.4,5 In
most cases, the toxic H2S has to be converted into an environ-
mentally friendly end product. Thus, simultaneous hydrogen
preferably under visible light irradiation to produce hydrogen,
the clean fuel. This will not only effectively harness the abun-
dant solar energy, but also clean up the environment. The Claus
process is currently used for the decomposition of H2S.6 The
conventional Claus process comprises the reaction between H2S
and SO2, yielding elemental sulphur and water vapour. Thus, the
Claus process is not a solution to generate H2 from H2S. Also,
the direct thermal decomposition of H2S for the production of
hydrogen and elemental sulphur is energy intensive and econ-
omically more unviable. As a result, there has currently been an
immense emphasis on the development of photocatalysts, which
are active in the visible range for the production of hydrogen.
Unfortunately, the photocatalysts investigated under the visible
region are still limited in number.7 Recently, a CdS–TiO2 nano-
composite film,8 CdIn2S4 bipyramids/flowers/nanotubes,9,10
a
CdS photocatalyst,11 a β-In2S3 dandelion flower,12 Cu-doped
ZnO nanoparticles,13 and metal oxides,14,15 have been used for
the photodecomposition of H2S. It is also noticed that the photo-
decomposition of H2S by sulfide semiconductor photocatalysts,
especially CdS, is not workable due to photocorrosion of the cat-
alyst during the process. Therefore, there is a need to develop a
stable semiconductor photocatalyst for hydrogen production,
which would be efficient, economical and sustain against photo-
corrosion. ‘Metal oxides’ is the most obvious and popular choice
for consideration, which fulfil the desired properties mentioned
above. Most of the studies have been focused on large band
gaps and stable semiconductor oxides, such as TiO2 and ZnO,
wherein electron–hole pairs are generated under photoexcitation
aCentre for Materials for Electronic Technology, Panchawati, Off
Pashan Road, Pune 411008, India. E-mail: kbbb1@yahoo.com,
bbkale@cmet.gov.in; Fax: +91-20-25898085; Tel: +91-20-25898390
b759/83 Deccan Gymkhana, Pune 411004, India
cNational Chemical Laboratory (NCL), Dr. Homi Bhabha Road,
Pune-411008, India
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c2gc35519a
2790 | Green Chem., 2012, 14, 2790–2798
This journal is © The Royal Society of Chemistry 2012

View Online
by UV light only. In many respects, ZnO is more suited in dye
sensitised solar cell devices than TiO2 due to its excellent elec-
tron mobility and its processability into many shapes and forms
as well as adjusting its band-gap to suit its applications in the
solar spectrum region.
A step forward to extend the spectral response of ZnO to the
visible spectral region is due to impurity doping. An initial
approach to shift the optical response of ZnO from the UV
(3.2 eV) to the visible spectral range has been the doping of tran-
sition metal elements into ZnO.16 However, metal-doping has
several drawbacks. The doped materials have been shown to
suffer from thermal instability and metal centres formed as a
consequence of doping metals, act as electron traps, which is
reducing the photocatalytic efficiency. Recently, it was shown
that the desired photo-response of ZnO can be better achieved by
using anionic dopants rather than metal ions. Suitable group V
elements, such as N, P and As, are used as anionic dopants in
ZnO.17,18 The doping of ZnO has limitations and hence n-type
doping is easily obtained, while it is difficult achieve p-type
materials.19,20 Nitrogen has been considered as the best candidate
for synthesizing p-type ZnO because it has a similar ionic radius
as oxygen and also the smallest ionization energy among the
group V elements.21,22 N-doping into ZnO has been tried by
employing various nitrogen precursors such as NH3, N2O, NO,
N2, Zn3N2 and MMHy.23
Fig. 1 TG-DTA curve of the Zn–urea intermediate.
Table 1 Weight loss stages of the intermediate (Zn–urea complex)
Weight loss stages (°C)
Loss of molecules
% Weight lossa
0–160
Adsorbed H2O
2NH3 + 2NO2
2HCl + 2NO2
4CO2 + NO2
3 (1)
160–260
260–350
350–520
13 (14)
29 (30)
34 (32)
Herein, we report the green synthesis of p-type N-doped ZnO
using a simple wet chemical method. Significantly, nanostruc-
tured p-type N-doped ZnO has been used for the decomposition
of H2S using sunlight for the production of hydrogen which has
previously been unattempted. We have synthesized the photo-
catalyst by a simple chemical method. Characterization of the as-
synthesized N-doped ZnO photocatalyst was done by XRD and
FESEM, revealing its crystal structure and surface morphology.
The incorporation of the ‘N’ impurity into ZnO was confirmed
by Raman spectra and X-ray photoelectron spectroscopy (XPS).
a Figures in parenthesis are predicted values.
would be in close proximity. A further weight loss of 29% is
observed in the range 260–350 °C. This may be attributed to the
loss of two HCl and two NO2 molecules. The corresponding
endothermic DTA curve obtained also confirms the removal of
HCl and NO2. Similarly, the weight loss observed in the range
350–520 °C is around 34%, which is due to the loss of four CO2
and one NO2 molecules. The loss of CO2 and NO2 is due to
complete combustion of the Zn–urea complex and excess biuret.
Accordingly, the exothermic peak is observed at corresponding
temperatures. The total weight loss was observed around 79%,
which is matching with the extent of ZnO formed from
the complex and excess biuret. The weight losses observed
at respective stages is in fairly good agreement with those
predicted. Considering the stepwise weight loss, the following
reaction has been proposed.
2. Results and discussion
In the present investigation, the zinc–urea intermediate complex
was obtained from molten urea and ZnCl2. To understand the
mechanism of formation of ZnO by this route, the intermediate,
most likely a Zn–urea complex, was characterized using
TG-DTA for its thermal decomposition behaviour.
Fig. 1 shows the TG-DTA curves obtained for the Zn–urea
intermediate complex with a dynamic heating program. Its
thermal decomposition occurs in four stages as indicated by the
weight loss in TG (Table 1). The DTA curve shows correspond-
ing heat changes during the course of the reaction.
The TG curve shows that the 3% weight loss is observed at
the first stage of the decomposition of the product, at 160 °C,
which is higher than expected, which is due to the loss of
adsorbed water molecules. This stage is followed by the 13%
weight loss in the 160–260 °C range, which can be attributed to
the loss of two NH3 and one NO2 molecules. The broad
endotherm in the DTA curve shown at these stages also confirm
the removal of H2O, NH3 and NO2. At this stage, the product
would be a Zn–urea complex consisting of Zn–urea decompo-
sition products, wherein Zn and nitrogen from the urea products
ð1Þ
No further loss is observed in thermal studies, indicating the
formation of a stable phase of ZnO, as subsequently revealed in
XRD studies.
Fig. 2 shows the XRD patterns of the as-synthesized products
(S1–S5), prepared under different conditions. It is clear from
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2790–2798 | 2791

View Online
Fig. 3 UV-DRS spectra of samples (a) undoped ZnO (S1) treated at
800 °C for 3 h and N-doped ZnO treated at different reaction temp-
eratures for 3 h at (b) 500 °C (S2), (c) 600 °C (S3), (d) 700 °C (S4)
(e) 800 °C (S5).
Zn–N → Zn–O transformation would be facilitated with the
increase in temperature of reaction. Also, it is known that
undoped ZnO is always ‘n-type’. At the temperature of 800 °C,
even though the Zn–N → Zn–O transformation takes place, it
may not go to completion within the limited period of heating
(3 h). S5 may have some nitrogen left in the lattice but its effect
is overtaken by the ‘defects’ of undoped ZnO formed at high
temperature, which is an ‘n-type’ material.
We observed that there is a one to one correspondence in
these observations, namely, when there is a dominant presence
of nitrogen in the lattice of ZnO, the sample assumes P-type con-
duction as shown by the ‘hot probe’ method. This p-type to
n-type conversion is further supported by the results on the
studies of Raman spectroscopy and XPS results.
The ultraviolet-visible diffuse reflectance spectra (UV-DRS)
of ZnO (S1) and N-doped ZnO (S2–S5) are shown in Fig. 3.
The nature of absorption band with a sharp ‘cut off’ at 388 nm
(band gap = 3.19) of sample S1, is assigned to the phase pure
ZnO formation. The band gap of samples S2–S5 were observed
to be 2.3, 2.4, 2.5 and 2.5 eV, respectively. The samples S2–S5
show drastic red shifts in band gaps, which is due to N-doping.
The N-doped ZnO lattice shall have defects formed in the lattice.
The location of such defects in the band diagram would be
above the valance band. The amount of doped nitrogen shall
determine the width of the defect band. Such a band picture can
explain all the optical absorption observations of the present
study.
Thus, sample S2 shows more absorption in the visible region
than the other samples, indicating maximum doping of nitrogen.
It is well known that the Zn–O bond is more stable than Zn–N.
Therefore, transformation of Zn–N → Zn–O is thermodynami-
cally favoured. As the temperature of the reaction is increased,
this transformation results in lowering of the amount of nitrogen
in the ZnO lattice and consequently lowering of absorption in
visible region.
Fig. 2 XRD spectrum of samples (a) undoped ZnO (S1) treated at
800 °C for 3 h and N-doped ZnO treated at different reaction temp-
eratures for 3 h at (b) 500 °C (S2), (c) 600 °C (S3), (d) 700 °C (S4) and
(e) 800 °C (S5).
Table 2 Type of conduction of ZnO samples
Sample
S1
n
S2
p
S3
p
S4
p
S5
n
Type of conduction
Fig. 2 that the products (S1–S5) show identical XRD pattens.
All the peaks are assigned to the standard hexagonal wurtzite
ZnO crystal structure (JCPDS 36-1451) and no additional peaks
are observed, which infers the formation of phase pure ZnO.
Also, this suggests the possibility of induction of nitrogen into
the ZnO lattice which has not affected the ZnO crystal structure.
The absence of phases such as ZnN, Zn (OH)2, etc., are note-
worthy.24
The type of conductivity of all the end products (S1–S5) is
determined by using a well known ‘hot probe’ method.25,26 The
type of electrical conduction properties of samples S1, S2, S3,
S4 and S5 are indicated in Table 2.
The sample S1 shows n-type conduction and the other
samples, S2, S3 and S4, were changed from n-type conduction
to p-type conduction by annealing at 500, 600 and 700 °C,
respectively. Surprisingly, this p-type conductivity is transformed
to n-type for sample S5 formed by annealing at 800 °C.
Thus, the result of finding the type of conductivity by the hot
probe method shows that ZnO gets converted to ‘p-type material’
when heated in the temperature range 500–700 °C for 3 h in a
partly nitrogenous atmosphere derived by urea products under
the present experimental conditions. This is obviously due to the
replacement of oxygen by nitrogen in the ZnO lattice.18,27 The
absence of a ‘p-type’ nature in sample S5 can be explained as
follows.
The PL study (ESI† (SI-1)) shows emission peaks at 380, 393
and 389 nm for the sample S1, S2 and S5, respectively. There is
2792 | Green Chem., 2012, 14, 2790–2798
This journal is © The Royal Society of Chemistry 2012

View Online
Fig. 5 XPS patterns of N-doped ZnO synthesized at 500 °C for 3 h
(S2) (a) Survey spectrum and high-resolution spectra for (b) Zn2p,
(c) O1s, (d) N1s.
peaks that were assignable to Zn, O, C and N. High resolution
XPS spectra of the Zn2p and O1s lines are shown in Fig. 5b and c.
The core lines of Zn2p3/2 and 2p1/2 are located at 1021.8 eV and
1044.7 eV, respectively. In Fig. 5 the high resolution scanning XPS
spectra of O1s after deconvolution shows that oxygen exists at
least in three forms with the following binding energies: 530.5,
532.0, 533.5 eV. The peak at 530.5 eV is mainly assigned to the
oxygen atoms coordinated with Zn atoms.34 The peak at 532 eV
corresponds to oxygen adsorbed on sample surface probably as (–
OH).35 The peak at 533.5 eV is likely to be due to water molecules
on the surface.36 Fig. 5d show that N1S peaks appeared at 395.7,
397.9, 400 eV which indicate nitrogen incorporation. Three peaks
are assigned to N–Zn, N–C and N–H bonds respectively.37,38
Oxygen to nitrogen atomic ratio of 1 : 0.38 is estimated by XPS.
The XPS results are consistent with the Raman results, which
demonstrate successful N-doping into the ZnO. The XPS of
sample S3, S4 and S5 are given in ESI† (SI-2).
Oxygen to nitrogen atomic ratios for the samples S3 and S4
are 1 : 0.32 and 1 : 0.41, and for the S5 sample there is no nitro-
gen. It is interesting to note that more nitrogen is shown to be
present in sample S4 than in sample S3. This is understandable
as the reaction temperature of S4 is higher than that of S3 and
therefore more nitrogen is present on the surface of the sample
due to the nitrogen out-diffusion process. Such removal of nitro-
gen from N-doped ZnO is reported in the literature.39 XPS is a
surface sensitive technique which efficiently detects these subtle
variations of concentration.
The morphology and particle size of the nanostructured com-
pounds were investigated by FESEM and TEM. Fig. 6 shows the
FESEM images of samples S1–S5, indicating the morphologies
of undoped and all N-doped ZnO samples. The FESEM of
sample S1 (Fig. 6a, b) shows the agglomerated hexagons of
sizes ∼3–6 μm. The FESEM of the N-doped ZnO sample S2
(Fig. 6c, d) shows a spongy structure, whereas samples S3
(Fig. 6e, f) and S4 (Fig. 6g, h) show agglomerated nanoparticles
with sizes of ∼50–60 nm and ∼50–100 nm, respectively.
However, sample S5 (Fig. 6i, j) reveals distorted hexagons of
sizes ∼4–5 μm.
Fig. 4 Raman spectrum of samples (a) undoped ZnO (S1) treated at
800 °C for 3 h and N-doped ZnO treated at different reaction tempera-
ture for 3 h at (b) 500 °C (S2), (c) 600 °C (S3), (d) 700 °C (S4),
(e) 800 °C (S5).
a red shift observed due to the replacement of oxygen by nitro-
gen in ZnO. Apparently, the red shift has been influenced by the
N-doping as well as the creation of defects during the thermal
process.28
Raman scattering was performed for the undoped and
N-doped ZnO, as illustrated in Fig. 4. In all the spectra, a
common Raman line is located at 437 cm−1 (Fig. 4a–c), which
corresponds to the E2 (high) vibrational mode which implies
the existence of the wurtzite structure for both undoped and the
N-doped ZnO. The peak located at 333 cm−1 corresponds to the
second order (E2 (high) − E2 (low)).29 The other modes at A1
(275–277 cm−1), A3 (509 cm−1) and A1 (LO) (580 cm−1) can be
assigned to local vibration modes (LVMs) of nitrogen in ZnO.30
Raman modes at 275 cm−1 was attributed to the localized
vibration of Zn atoms, where parts of their first nearest neighbour
O atoms are replaced by N atoms in the ZnO lattice.31 Also, the
Raman peaks observed at 509 and 580 cm−1 is related to the
presence of nitrogen.32,33 It may be noted that the nitrogen
related peaks, namely (275–277 cm−1), (509 cm−1) and (LO)
(580 cm−1) are absent in samples S1 and S5. The Raman
spectroscopic observations, thus, confirms N-doping in the
lattice and concomitantly shows that sample S1 and S5 do not
have an effective presence of nitrogen.
The presence of nitrogen and its concentration in N-doped
ZnO was analyzed by XPS. During the thermal treatment, the
complex decomposes into ZnO but traces of ‘Cl’ get adsorbed
on the surface during cooling of the sample. The traces of ‘Cl’
were removed by washing with copious amounts of hot water
and its complete removal is confirmed by XPS (Fig. 5a). The
full XPS spectrum of the N-doped ZnO (Fig. 5a) showed only
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2790–2798 | 2793

View Online
Fig. 7 TEM of N-doped ZnO treated at 500 °C for 3 h (S2).
Fig. 6 FESEM of samples (a, b) undoped ZnO (S1) treated at 800 °C
and N-doped ZnO treated at different reaction temperature for 3 h at
(c, d) 500 °C (S2), (e, f) 600 °C (S3), (g, h) 700 °C (S4), (i, j) 800 °C
(S5).
Scheme 1 Schematic representation of formation mechanism of
undoped and N-doped ZnO.
phase of the N-doped ZnO, which is consistent with the XRD
results. Fig. 7d shows a high-resolution TEM (HRTEM) image
of the nanocubes. The lattice fringes show that the nanocubes
are well crystalline and the observed d-spacing corresponds to
the (101) plane of N-doped ZnO.
After careful evaluation of the above results, the over-all
mechanism of N-doped ZnO formation by the ‘urea route’ can
be visualized as below.
The formation mechanism of undoped and N-doped ZnO is
represented in Scheme 1. When ZnCl2 and urea are mixed in
alcoholic medium and heated at 150 °C, a ZnCl2 based Zn–urea
complex is formed (Scheme 1). However, the product remains
Since the morphology of sample S2 was not clear in FESEM
(spongy structure), we have characterized the sample with TEM
(Fig. 7a–d). In the TEM images (Fig. 7a, b) we observed
∼50–100 nm sized nanocubes as the primary growth. Interest-
ingly, growth of secondary nanoparticles of size ∼4–5 nm over
the nanocubes was observed. The secondary growth is perhaps
seen as a spongy structure in the FESEM. Such a spongy struc-
ture with highly crystalline nanoparticles (Fig. 7b) is likely to
have a high surface area.
The electron diffraction (Fig. 7c) dots can be indexed as the
(100), (101), (102) and (201) planes of the hexagonal wurtzite
2794 | Green Chem., 2012, 14, 2790–2798
This journal is © The Royal Society of Chemistry 2012

View Online
hygroscopic when exposed to the atmosphere. We have per-
formed the reaction using excess urea so as to have a larger
number of nitrogen species available for doping. Hence, along
with the complex some unreacted biuret is also formed.40
Further, the complex along with biuret (as per eqn (1)) is
annealed at 500, 600, 700 and 800 °C on the basis of the precise
thermal study by TG-DTA. The thermal study shows a four-stage
weight loss to obtain ZnO. Our main aim is to dope nitrogen
through urea, of which we have achieved at 500 °C.
In the case of sample S5, it was expected to achieve the same
morphology, i.e., basic hexagons as obtained for sample S1.
However, due to the small number of defects, their influence in
the distorted hexagons obtained in sample S5 is well understood.
The band gap of N-doped ZnO is red shifted drastically as com-
pared to ZnO, which is due to N-doping, as N-doping produces
external strain as well as lattice strain.42,43 This strain also affects
the total morphology of the N–ZnO. Hence, different mor-
phologies are obtained for the N-doped samples.
On the basis of the critical assignment of weight loss and cor-
responding exo/endothermal changes in DTA observed during
thermal decomposition of the Zn–urea complex we have pro-
posed the reaction, eqn (1). According to the TG-DTA, further
thermal treatment from 160–260 °C shows gradual decompo-
sition of the complex and biuret in the form of NO2 and NH3, as
explained in the TG-DTA section. After 260 °C the gradual
weight loss of the complex and biuret is continuous and there
may be formation of ZnO1−xNx at 500 °C along with HCl, NO2
and CO2.
Photocatalytic activity
The photocatalytic activity of the undoped and N-doped ZnO
samples for hydrogen generation via photodecomposition of H2S
was performed. The freshly prepared undoped and N-doped ZnO
powders were employed for examining the photocatalytic
hydrogen generation activities.
Table 3 shows the hydrogen evolution rate from the photode-
composition of H2S using undoped and N-doped samples, i.e.,
S1–S5, respectively. The photocatalytic activity of sample S1
(undoped ZnO) showed a moderate hydrogen evolution rate, i.e.,
around 1874 μmol h−1, while the N-doped ZnO samples S2, S3,
Urea and its decomposed products, such as a biuret, are in
intimate contact with the complex and hence the species contain-
ing nitrogen in these decomposed products can act as a nitrogen
source for doping in ZnO, which is being formed at high temp-
eratures in an air atmosphere.41 Ammonia that is evolved at the
second stage (160–260 °C) is partially adsorbed by means of
dissociative chemisorption on the surface of the intermediate
(Zn–urea complex) formed at that stage. During chemisorption,
the lone pair of electrons of the nitrogen atom is condensed on
the surface, due to which an equal number of hydroxyl groups
and amide ions are formed on the surface.26 At high temperature
(500 °C) N-doping takes place during the formation of ZnO. The
formation of ZnO1−xNx at 500 °C is already confirmed by
Raman and XPS. The decomposition of ZnO1−xNx starts after
500 °C and at 800 °C pure phase ZnO has been formed
(Scheme 1), which is also confirmed by Raman and XPS. The
schematic representation (Scheme 1) also shows the morphologi-
cal transformation of N-doped ZnO at different temperatures.
The FESEM image of sample S1 shows well developed hexa-
gons. It is not surprising because it does not contain intentionally
added impurities in the matrix. The addition of impurities/
dopants hampers the growth of the basic structures of the crys-
tals. Sample S2 shows the secondary growth of tiny nanoparti-
cles of size ∼4–5 nm on ∼50–100 nm size nanocubes (Fig. 7) of
the N-doped ZnO, as presented in Scheme 1. Sample S3 shows
the disappearance of tiny nanoparticles (spongy structure demon-
strated by FESEM Fig. 6c, d) and the formation of bigger nano-
particles as compared to sample S2. This is consistent with the
earlier proposed mechanism, that as the temperature increases
the ZnN → ZnO transformation occurs leading to a decrease in
N-doping. Consequently, the numbers of defects are also
lowered and thus the bigger particles are formed due to high
temperature. At high temperatures, the formation of bigger par-
ticles via aggregation of smaller particles due to sintering is
quite obvious. The same trend continues for S4 and S5 samples.
FESEM and Raman of sample S5 is similar to sample S1, and
implies minimization of N-doping. The S5 sample is expected to
have a negligible nitrogen content, as suggested earlier because
of its out-diffusion process, and therefore is likely to have fewer
defects, compared to the defects in samples S2–S4.
S4 and S5 showed 3957, 3480, 2875 and 2448 μmol h−1
respectively. The maximum H2 evolution, i.e., 3957 μmol h−1
,
,
was obtained for sample S2. The hydrogen evolution versus
irradiation time plots for catalysts S1–S5 are shown in Fig. 8.
The linearity of the graphs clearly show the steady evolution rate
of the undoped and N-doped ZnO samples. However, overall the
Table 3 Photocatalytic activity of undoped and N-doped ZnO samples
for hydrogen evolutiona
Sample
S1
S2
S3
S4
S5
H2 (μmol h−1
)
1874
3957
3480
2875
2448
a Catalyst, 0.2 g; 200 ml (0.25 M aqueous KOH solution); 300 W Xe
lamp (oriel).
Fig. 8 Time versus volume of H2 (μmol) evolution of the as-syn-
thesized (a) undoped ZnO (S1) and (b–e) N-doped ZnO (S5–S2)
samples.
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2790–2798 | 2795

View Online
N-doped ZnO sample S2 shows higher activity as compared to
undoped ZnO (S1) and the other metal oxides.14,15 Significantly,
the photocatalytic activity of N-doped ZnO was observed to be
much higher as compared to sulfide semiconductors, such as
CdS, CdIn2S4, In2S3, etc.,9–12 reported so far.
The higher hydrogen evolution rate for sample S2 is observed,
which is due to the higher doping concentration of nitrogen than
the other samples and also due to the secondary growth of nano-
particles (∼4–5 nm) over the primary growth of nanocubes. Due
to N-doping, the band gap of the N-doped ZnO has been shifted
completely in the visible region and hence, maximum visible
light absorption is favoured. Thus, the maximum evolution rate
obtained is justified. Accordingly, it is observed that the evol-
ution rate decreases in the case of samples S3–S5, which is due
to lowered levels of N-doping.
During the photodecomposition reaction, in 0.25 M KOH
solution (pH of 12–13), the weak diprotic acid H2S (two pKa
values are 7.0 and 11.96) dissociates and maintains equilibrium
with hydrosulphide HS− ions. The oxide semiconductor absorbs
light and generates electron–hole pairs. The photogenerated
valence band hole (h+VB) upon band gap excitation of the
powder samples of N-doped ZnO oxidizes the HS− ions to
sulphur (2S), liberating protons. The conduction band electron
(e−CB) from the N-doped ZnO photocatalyst reduces protons to
produce molecular hydrogen.
stirring until a homogeneous solution was achieved. This solu-
tion was then kept in oven at 150 °C to obtain a white product,
which was found to be hygroscopic. The hygroscopic product
was preserved in a desiccator, and then further subdivided for
heat treatment at 500 °C, 600 °C, 700 °C and 800 °C for 3 h to
obtain brown products, which are no longer hygroscopic. For
comparison, we also synthesized the undoped (S1) sample. It
was synthesized by heat treating a 0.5 M ZnCl2·3H2O solution
(without any additives like urea) at 150 °C to obtain the product,
which was further annealed at 800 °C for 3 h. The other products
with dopants were annealed at 500 °C, 600 °C; 700 °C and
800 °C for 3 h. After annealing at the different temperatures, all
samples (including S1) were washed with copious amounts of
hot distilled water until all chlorine is removed, and were labeled
as S2, S3, S4 and S5 respectively. These products were then sub-
jected to UV-DRS, XRD, Raman, XPS, FESEM, TEM analysis
for their characterization.
3.2. Characterization
Thermogravimetric (TG) and differential thermal analysis
(DTA) of the intermediate was carried out at a heating rate of
10 °C min−1 in air (SETARAM-16/18). The optical properties of
the powder samples were studied using an UV-visible-near infra-
red spectrometer (UV-VIS-NIR, Perkin Elmer Lambda-950).
The crystalline phases and the crystallite size of the photocatalyst
was identified using an X-ray powder diffraction (XRD) tech-
nique (XRD-D8, Advance, Bruker-AXS) with Cu Kα radiation.
Room temperature micro Raman scattering (RS) was performed
using a HR 800-Raman Spectroscopy, Horiba Jobin Yvon,
France, with an excitation at 632.81 nm by a coherent He–Ne
ion laser and a liquid nitrogen cooled CCD camera to collect and
process the scattered data. The chemical bonds of these N-doped
ZnO samples were examined using X-ray photoelectron
spectroscopy (XPS, ESCA-3000, VG Scientific Ltd, England)
with a base pressure greater than 1.0 × 10−9 Pa and Mg Kα radi-
ation (1253.6 eV) was used as an X-ray source operated at
150 W. The morphologies of the N-doped ZnO nanostructures
were characterized by field emission scanning electron
microscopy (FESEM, Hitachi, S-4800) and high resolution trans-
mission electron microscopy (HRTEM, JEOL, 2010F). For
HRTEM studies, samples were prepared by dispersing the
powder in acetone, followed by sonication in an ultrasonic bath
for 2 min and then drop-casting the dispersion on a carbon-
coated copper grid and by subsequent drying in a vacuum.
H2S þ OHꢀ $ HSꢀ þ H2O
ð2Þ
ð3Þ
ð4Þ
ð5Þ
Semiconductor N–ZnO ! hþVB þ eꢀ
CB
Oxidation reaction : 2HSꢀ þ 2hþ ! 2S þ 2Hþ
VB
Reduction reaction : 2Hþ þ 2eꢀ ! H2
CB
The p-type N-doped ZnO shows higher photocatalytic activity,
which may be because of the availability of a larger number of
holes (due to p-type doping), that ultimately accelerates the pro-
duction of protons and increases molecular hydrogen generation
(eqn (4) and (5)). However, the detailed mechanism requires
more experimentation which is in progress.
It is noteworthy that this is the first time that we have reported
enhanced hydrogen generation from waste H2S using an eco-
friendly, stable N-doped ZnO as a photocatalyst under solar
light. The stability of the photocatalyst has also been examined
by reusing the photocatalyst samples (S2) and we obtained the
same results (SI-3†). The XRD of the reused catalyst did not
show any changes in the phase purity of the N-doped ZnO
(SI-4†), which implies good stability of the catalyst.
3.3. Photodecomposition of H2S
The cylindrical quartz photochemical reactor was filled with
200 ml 0.25 M aqueous KOH and purged with argon for 1 h.
Hydrogen sulphide (H2S) was bubbled through the solution for
about 1 h at the rate of 2.5 ml min−1 at 299 K. N-doped ZnO
(0.2 g) was introduced into the reactor and irradiated with a
visible light source Xe-lamp (300 W, LOT ORIEL GRUPPE,
EUROPA, LSH302, see ESI† SI-5 for Xe lamp spectrum) with
constant stirring. The escaped hydrogen sulfide was trapped in
a NaOH solution. The amount of hydrogen gas evolved was
collected in a graduated cylinder and measured. The evolved
3. Experimental section
3.1 Synthesis of N-doped ZnO
The synthesis of the N-doped ZnO was performed by using a
wet chemistry technique. Analytical grade zinc chloride (ZnCl2),
urea and absolute ethanol were purchased from the local chemi-
cal manufacturer (Qauligene chemicals) and were used as
received. In a typical synthesis, to 0.5 M ZnCl2 solution in absol-
ute ethanol, urea (1 : 2 w/w) was added slowly with constant
2796 | Green Chem., 2012, 14, 2790–2798
This journal is © The Royal Society of Chemistry 2012

View Online
8 W. So, K. Kim and S. Moon, Int. J. Hydrogen Energy, 2004,
29, 229.
9 A. P. Bhirud, N. S. Chaudhari, L. K. Nikam, R. S. Sonawane, K. R. Patil,
B. J. Ook and B. B. Kale, Int. J. Hydrogen Energy, 2011, 36, 11628.
10 B. B. Kale, J. O. Baeg, S. M. Lee, H. Chang, S. J. Moon and C. W. Lee,
Adv. Funct. Mater., 2006, 16, 1349.
hydrogen was then analyzed for its purity using a gas chromato-
graph (Model Shimadzu GC-14B, MS-5 Å column, TCD, Ar
carrier). All the samples (S1–S5) were tested for their catalytic
activity under identical conditions.
11 S. k. Apte, S. N. Garje, G. P. Mane, A. Vinu, A. D. Naik,
D. P. Amalnerkar and B. B. Kale, Small, 2011, 7, 957.
12 S. D. Naik, T. C. Jagadale, S. K. Apte, R. S. Sonawane, M. V. Kulkarni,
S. I. Patil, S. B. Ogale and B. B. Kale, Chem. Phys. Lett., 2008, 452,
301.
13 K. G. Kanade, B. B. Kale, J. -O. Baeg, S. M. Lee, C. W. Lee, S. -
J. Moon and H. Chang, Mater. Chem. Phys., 2007, 102, 98.
14 E. Subramanian, J. O. Baeg, B. B. Kale, S. M. Lee, S. J. Moon and
K. J. Kong, Bull. Korean Chem. Soc., 2007, 28, 2089.
15 B. B. Kale, J. O. Baeg, K. J. Kong, S. J. Moo, S. M. Lee and W. W. So,
Int. J. Energy Res., 2010, 34, 404.
16 K. C. Barick, S. Singh, M. Aslam and D. Bahadur, Microporous Meso-
porous Mater., 2010, 134, 195; D. Jung, Solid State Sci., 2010, 12, 466;
M. C. Zlatko, D. T. L. Alexander, S. Andrzej, M. C. Marijana, L. Forró
and M. Arnaud, Cryst. Growth Des., 2010, 10, 4437; J. Zhao, H. Liang,
J. Sun, Q. Feng, S. Li, J. Bian, L. Hu, G. Du, J. Ren and J. Liu, Phys.
Status Solidi A, 2011, 208, 825; Y. Liu, T. Wang, X. Sun, Q. Fang, Q. Lv,
X. Song and Z. Sun, Appl. Surf. Sci., 2011, 257, 6540; C. Wu, L. Shen,
Y. C. Zhang and Q. Huang, Mater. Lett., 2011, 65, 1794; T. Xia, Y. Zhao,
T. Sager, S. George, S. Pokhrel, N. Li, D. Schoenfeld, H. Meng, S. Lin,
X. Wang, M. Wang, Z. Ji, Z. J. I. Zink, L. Madler, V. Castranova, S. Lin
and A. E. Nel, ACS Nano, 2011, 5, 1223.
17 D. Fan, R. Zhang and Y. Li, Solid State Commun., 2010, 150, 1911;
H. von Wenckstern, H. Schmidt, M. Brandt, A. Lajna, R. Pickenhain,
M. Lorenz, M. Grundmann, D. M. Hofmann, A. Polity, B. K. Meyer,
H. Saal, M. Binnewies, A. Börger, K.-D. Becker, V. A. Tikhomirov and
K. Jug, Prog. Solid State Chem., 2009, 37, 153; S. P. Wang, C. X. Shan,
B. H. Li, J. Y. Zhang, B. Yaoa, D. Z. Shen and X. W. Fan, J. Cryst.
Growth, 2009, 311, 3577.
4. Conclusions
In nutshell, the synthesis of a p-type N-doped ZnO catalyst was
demonstrated by a simple wet chemical method. Significantly,
the efficient and stable production of eco-friendly hydrogen
using hazardous H2S waste with the help of abundantly available
solar light was demonstrated for the first time using the p-type
N-doped ZnO catalyst. The optical study clearly shows a shift in
the band gap to the visible region (3.19–2.3 eV) due to
N-doping, which is in good agreement with the Raman and XPS
studies. XPS confirms the bond formation of nitrogen with zinc
in the ZnO lattice with an oxygen-to-nitrogen atomic ratio of
1 : 0.38. The lowering of the amount of N-doping with increas-
ing temperature agrees well with the XPS results. In the case of
the N-doped ZnO catalyst obtained at 500 °C (S2), nanocubes of
the size ∼50–100 nm were obtained as the primary growth,
while the secondary growth was nanoparticles of size ∼4–5 nm
over the nanocubes. The naked N-doped ZnO sample prepared at
500 °C showed excellent photocatalytic activity (3957 μmol h−1
)
for the hydrogen production under visible light irradiation,
which is much higher as compared to the oxide semiconductor
catalysts reported so far. Significantly, the catalyst is highly
stable and can be used repeatedly. The green route investigated
herewith, will have potential applications in the production of
other N-doped metal oxides.
18 X. Yang, A. Wolcott, G. Wang, A. Soba, R. C. Fitzmorris, F. Qian,
J. Z. Zhang and Y. Li, Nano Lett., 2009, 9, 2331.
19 S. B. Zhang, S.-H. Wei and A. Zunger, Phys. Rev. Lett., 2000, 84, 1232.
20 S. B. Zhang, S. H. Wei and A. Zunger, Phys. Rev. B, 2001, 63, 075205.
21 C. H. Park, S. B. Zhang and S. H. Wei, Phys. Rev. B, 2002, 66, 073202.
22 K. Iwata, P. Fons, A. Yamada, K. Matsubara and S. Niki, J. Cryst.
Growth, 2000, 209, 526.
23 Z. Z. Ye, J. G. Lu, H. H. Chen, Y. Z. Zhang, L. Wang, B. H. Zhao and
J. Y. Huang, J. Cryst. Growth, 2003, 253, 258; X. L. Guo, H. Tabata and
T. Kawati, J. Cryst. Growth, 2002, 544, 237; C. Wang, Z. Ji, K. Liu,
Y. Xiang and Z. Ye, J. Cryst. Growth, 2003, 259, 279.
24 C. W. Zou, X. D. Yan, J. Han, R. Q. Chen, W. Gao and J. Metson, Appl.
Phys. Lett., 2009, 94, 171903.
25 G. Golan, A. Axelevitch, B. Gorenstein and V. Manevych, Microelectron.
J., 2006, 37, 910.
Acknowledgements
The authors would like to thank the Department of Information
and Technology, Govt of India, New Delhi for financial support
and Dr D. P. Amalnerkar, Executive Director, C-MET for their
kind support. The authors are very grateful and also wish to
express their gratitude to the nanocrystalline materials group
C-MET, Pune for their support. The authors would also acknowl-
edge to Dr K. R. Patil for XPS analysis and Dr V. H. Rane for
the GC analysis.
26 B. V. Zeghbroeck, Principles of Semiconductor Devices, 2004, ch. 2,
pp. 161–162.
27 T. Morimoto, H. Yanal and M. Nagao, J. Phys. Chem., 1976, 80, 471.
28 C. Persson, C. Platzer-Bjorkman, J. Malmstron, T. Torndahl and
M. Edoff, Phys. Rev. Lett., 2006, 97, 146403.
29 J. Serrano, A. H. Romero, F. J. Manjon, R. Lauck, M. Cardona and
A. Rubio, Phys. Rev. B, 2004, 69, 094306.
Notes and references
30 K. Saito, Y. Hosokai, K. Nagayama, K. Ishida, K. Takahashi, M. Konagai
and B. P. Zhang, J. Cryst. Growth, 2004, 272, 805.
31 X. Zhu, H. Z. Wu, D. J. Qiu, Z. Yuan, G. Jin, J. Kong and W. Shen, Opt.
Commun., 2010, 283, 2695.
32 B. Sieber, H. Liu, G. Piret, J. Laureyns, P. Roussel, B. Gelloz,
S. Szunerits and R. Boukherroub, J. Phys. Chem. C, 2009, 113, 13643.
33 G. T. Du, Y. Ma, Y. T. Zhang and T. P. Yang, Appl. Phys. Lett., 2005, 87,
213103.
34 S. Anandana, A. Vinu, K. L. P. Sheeja Lovely, N. Gokulakrishnan,
P. Srinivasu, T. Mori, V. Murugesan, V. Sivamurugan and K. Ariga,
J. Mol. Catal. A: Chem., 2007, 266, 149.
35 S. F. Chen, S. J. Zhang, W. Liu and W. Zhao, J. Hazard. Mater., 2008,
155, 320.
36 S. F. Chen and L. Chen, Mater. Chem. Phys., 2006, 98, 116.
37 H. Wang, H. P. Ho, K. C. Lo and K. W. Cheah, J. Phys. D: Appl. Phys.,
2007, 40, 4682.
1 P. Xu, T. Xu, J. Lu, S. Gao, N. S. Hosmane, B. Huang, Y. Dai and
Y. Wang, Energy Environ. Sci., 2010, 3, 1128; F. Dionigi,
P. C. K. Vesborg, T. Pedersen, O. Hansen, S. Dahl, A. Xiong, K. Maeda,
K. Domen and I. B. Chorkendorff, Energy Environ. Sci., 2011, 4, 2937;
J. Shi, H. Cui, Z. Liang, X. Lu, Y. Tong, C. Su and H. Liu, Energy
Environ. Sci., 2011, 4, 466; K. Maeda, D. Lu, K. Teramura and
K. Domen, Energy Environ. Sci., 2010, 3, 471; M. Liu, L. Wang, G. Lu,
X. Yao and L. Guo, Energy Environ. Sci., 2011, 4, 1372.
2 H. G. Kim, D. W. Hwang and J. S. Lee, J. Am. Chem. Soc., 2004, 126,
8912.
3 H. G. Kim, P. H. Borse, W. Choi and J. S. Lee, Angew. Chem., Int. Ed.,
2005, 44, 4585.
4 M. Barbeni, E. Pelizzetti, E. Borgarello, N. Serpone, M. Gratzel and
L. Balducci, Int. J. Hydrogen Energy, 1985, 10, 249.
5 N. Buhler, K. Meier and J. P. Reber, J. Phys. Chem., 1984, 88, 3261.
6 M. P. Elsner, M. Menge, C. Muller and D. W. Agar, Catal. Today, 2003,
79–80, 487.
38 C. L. Perkins, S. H. Lee, X. Li, S. E. Asher and T. J. Coutts, J. Appl.
Phys., 2005, 97, 034907.
7 H. Kato and A. Kudo, J. Phys. Chem., 2004, 106, 5029.
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 2790–2798 | 2797

View Online
39 A. Tsukazaki, A. Ohtomo, T. Onuma, M. Ohtani, T. Makino, M. Sumiya,
K. Ohtani, S. F. Chichibu, S. Fuke, Y. Segawa, H. Ohno, H. Koinuma
and M. Kawasaki, Nat. Mater., 2005, 4, 42.
40 Q. F. Ana, P. Marina, V. Aurea, M. del Francisco, G. H. Mar and
M. G. C. Jose, Chem. Mater., 2010, 22, 6529.
41 C. Giordano, C. Erpen, W. Yao, B. Milke and M. Antonietti, Chem.
Mater., 2009, 21, 5136.
42 A. M. Smith, A. M. Mohs and S. Nie, Nat. Nanotechnol., 2009,
4, 63.
43 M. Onuki and S. Kubo, J. Phys. Soc. Jpn., 1962, 17, 1516.
2798 | Green Chem., 2012, 14, 2790–2798
This journal is © The Royal Society of Chemistry 2012