Embed Size (px)
Citation preview
www.elsevier.com/locate/apsusc
3 (2006) 2899–2910
Applied Surface Science 25DRIFTS study of surface reactivity to NO2 by zinc nanoparticle
aggregates and zinc hollow nanofibers
Stefano di Stasio a,*, Vladimiro Dal Santo b
a Istituto Motori CNR, Aerosol and Nanostructures Laboratory, Via Marconi, 8-80125 Napoli, Italyb Istituto di Scienze e Tecnologie Molecolari, CNR, Via Golgi, 19-20133 Milano, Italy
Received 21 October 2005; received in revised form 3 April 2006; accepted 5 June 2006
Available online 12 July 2006
Abstract
Zinc nanostructures synthesized with different morphologies from the same evaporation/condensation technique are studied with concern to
surface reactivity to NO2 by Diffuse Reflectance Infrared Fourier Transformed Spectroscopy (DRIFTS). Synthesis of nanopowders is obtained,
according to previous work, by gas flow thermal evaporation at 540 8C of bulk Zn grains. Two types of Zn powders are obtained and studied in
experiments. The first one is collected on the cold walls of the reactor as a deposit produced by thermophoretic effect. It is constituted by grains
(�10 mm) originated by the stratification of smaller aggregates (�200 nm) and isolated primary particles (�50 nm) born in the gas flow. The
second type of powder is grown from the condensation of Zn chemical vapors within the expansion orifice of the quartz reactor after relatively long
time (�1 h) deposition process. It is constituted mainly by hollow Zn nanofibers with external and internal diameter about 100 and 60 nm.
Preliminary characterization of the two types of powders is made by SEM, TEM, XRD. Thereafter, the two types of samples are studied by DRIFTS
at variable temperature (VT). Comparison is made between the home-synthesized nanopowders with respect to commercial Zn standard dust. The
Zn hollow nanofibers when exposed to NO2 are found to exhibit dramatic reactivity, which is not observed at all either in the case of clustered
aggregate zinc or of commercial Zn dust powders. Results indicate that, at increasing temperature from RT to 300 8C, the hollow nanofibers surface
reacts distinctively with adsorbant gas NO2, with contemporary formation of a progressively growing narrow absorption band at 2500 cm�1 and
contemporary depression of a doublet (�1600–1628 cm�1) band. In order to justify this striking spectral feature, we propose the occurring of a
possible polymerization process at nanofibers surface where most probably as a consequence of pre-treatment and exposure to gas NO2 a very thin
film of ZnO is formed. The possible role of huge specific surface of hollow nanofibers as inferred by preliminary SEM, TEM, XRD studies is
discussed.
# 2006 Elsevier B.V. All rights reserved.
PACS: 81.07.�b; 81.20.Rg; 68.43.�h; 78.67.Ch
Keywords: Zn; Nanoparticles; Nanofibers; TEM; SEM; XRD; DRIFTS
1. Introduction
Quasi 1D structures on nanoscale have received recently
increasing attention from the scientific community. Examples
of such 1D objects are nanowires, nanorods, nanowhiskers,
nanocoils, nanobelts and nanotubes, for which investigations
have been presented that illustrate particularly attractive
electrical, optical and thermoelectric properties. A review
about 1D nanostructures has appeared recently [1].
* Corresponding author. Tel.: +39 081 71 77 122; fax: +39 081 239 60 97.
E-mail address: [email protected] (S. di Stasio).
0169-4332/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.apsusc.2006.06.027
The vast majority of such nanostructures are composed of
metal oxide semiconductors such as SnO2, ZnO, etc. [2–4]. For
instance, ZnO is an n-type direct wide band gap semiconductor
(Eg � 3.37 eVat room temperature), which is widely used since
decades in industry of sensors [5,6], electronics [2,3], solar cells
[7], displays [8,9]. Nevertheless, the experimental study
focused on the synthesis of pure metals nanoparticles is
fundamental in many respects, such as for instance to testing the
validity of different nucleation theory models [10]. Moreover, it
is well known from the literature that the electronic
confinement in metal materials does induce a series of
interesting effects [11,12]. Experimental receipts for fabrica-
tion of pure metal 1D nanostructures are reported in the case of
nanowires [13,14], nanorods [15] and nanotubes [11,16].
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–29102900
Between the most popular methods to produce nanomaterials
of different species, included metals (see [17] and references
cited therein) we remind plasma discharge, electrodeposition,
laser ablation and arc discharge within inert gas.
The aerosol routes for fabrication of nanostructures are
attractive in that they allow the good control of the process
parameters, are low-cost with respect to other synthesis
methods (for instance, vacuum techniques) and allow to
control at some degree both the physical (size, concentration,
crystal growth, agglomeration, porosity) and the chemical
(stoichiometry, surface state) properties [17,18].
A few authors report about the simple gas-phase evapora-
tion—condensation scheme (aerosol route) for generation of Zn
particles and 1D nanostructures. This is partially due to the
vapor pressure of such a species, which is relatively large with
respect to other metals even at low temperatures such as about
500 8C. Owing to this fact, it is difficult to limit the size of the
nuclei formed and obtain ultrafine particles below 20 nm.
The generation of very small particles of several metal
species by evaporation at regulated temperature in inert gas at
reduced atmosphere, is reported by Granqvist and Buhrman
[19]. In particular, these authors demonstrate the synthesis of
ultrafine particles with diameters 3–6 nm of Al, Mg, Zn and Sn
at pressure about 0.5–4 Torr of argon atmosphere.
The contributions in the literature specifically devoted to the
synthesis of Zn structures can be considered to start with the
work of Yumoto et al. [20] who first described the fabrication of
Zn crystals from thermal evaporation of Zn metal bars.
The synthesis of zinc nanoparticle powders is studied by
Recknagle et al. [21]. They thermally evaporated bulk Zn in
Argon atmosphere at a pressure 100–200 Torr and a
temperature 900 8C. Particle nucleation was obtained by the
expansion of Zn vapor in a separate chamber at 1 Torr, thus
obtaining nanocrystalline Zn with XRD crystallite size 10–
20 nm.
Wang et al. [22] reported the synthesis of Zn wirelike
nanostructures and nanobelts obtained by thermal evaporation in
Argon flow of a mixture of ZnS and graphite powder followed by
Zn deposition. These authors revealed beltlike structures with
thickness about 20 nm and typical length up to several microns.
They mentioned the difficulty to generate whiskers of 1D
structures from direct evaporation of bulk Zn powders owing to
the fact, we mentioned above, that the production of Zn vapors is
very fast at the increasing of the temperature.
In spite to the conclusion of the cited authors [22] we
demonstrated in our previous work [16] that it is possible to
generate 1D nanostructures, in particular, nanowires and
hollow nanofibers by regulated temperature evaporation in
nitrogen flow of Zn bulk powder at atmospheric pressure
followed by deposition on quartz substrate with evaporation
temperatures relatively low (below about 560 8C).
The aim of the present contribution is to report work about
the influence of different microstructure and morphology with
respect to surface reactivity to gas NO2 by Zn nanostructures
fabricated by the same evaporation–condensation–deposition
route and collected during the same experiments. Nitrogen
oxide (NO) and nitrogen dioxide (NO2) are particularly
harmful gaseous pollutants in urban and industrial areas, owing
to the planetary diffusion of thermal engines and power plants.
Of the two’s, typically at engine exhausts, NO2 is by far the
minority species, the amount depending on the specific
characteristic of the fuel and combustion process, but,
unfortunately, it yields the major damage to vegetation and
living beings owing to the fact that NO2 reacts in the
troposphere and forms acid rains.
We fabricated metallic zinc nanopowders with different
morphologies starting with evaporation of the bulk Zn grains at
the same temperature in nitrogen gas flow at atmospheric
pressure. The following expansion and cooling produce the
nucleation of Zn nanoparticles which stick by thermophoresis
on the inner surface of the quartz reactor where stratified
aggregates are formed. Moreover, the condensation of the Zn
vapors directly inside the expansion quartz nozzle produce a
fluffy deposit that mainly consists of zinc hollow nanofibers.
The evaporation temperature is about Tevap = 540 8C in all the
experiments.
Zinc nanoparticle clusters and zinc hollow nanofibers do
reveal completely different surface properties when exposed to
nitrogen dioxide as inferred by VT-DRIFTS experiments (RT to
300 8C).
We demonstrate a dramatic surface reactivity to NO2 by zinc
hollow nanofibers with respect to both clustered zinc
nanoparticles and commercial �20-mm size zinc powders
considered as standard powder. Different possible surface
mechanisms such as generation of (NxOy)z polymerized
species, formation of nitrites (NO2�) and nitrates (NO3
�) are
proposed, on the basis of available literature, in order to explain
the distinctive nanofiber signatures observed in the infrared
absorbance spectra. Peculiar high reactivity by hollow fibers to
NO2 is discussed and the information about nanofibers specific
surface is retrieved on the basis of preliminary characterization
at SEM, TEM and XRD.
2. Experimental
The synthesis process of Zinc nanoparticles and nanofibers
by an evaporation/condensation scheme is described in detail in
previous papers [16,23–25]. Briefly, the experimental set-up is
constituted by a horizontal quartz reactor, which is heated by a
coaxial furnace and flowing 500 cm3/min Nitrogen (quality
certified bottles) as carrier gas. Nitrogen from bottles is filtered
by hydrophobic EPA filters to remove all the eventual
impurities. Zinc vapors are obtained by thermal evaporation
at 540 8C of commercial Zn granules (Aldrich, 20 mesh, purity
better than 99.8%). Temperature profile inside the furnace is
measured along the reactor axis at different abscissas. The
typical T-profile is parabolic, with a central flat plateau at Tevap
and lower temperatures in correspondence of the oven
extremes. Hereafter, we will refer to Tevap as the evaporation
temperature of the feedstock zinc granules. After heating-up
transient (5–10 min) the temperature regime Tevap = 540 8C is
made inside the reactor and a small quartz boat containing the
feedstock Zn grains is inserted in correspondence of the middle
of the oven. Thereafter, reactor is sealed. Metal vapors under
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–2910 2901
Fig. 1. (a) SEM image of Zn aerosol aggregate collected on 18 mm � 18 mm glass support by thermal precipitator showing the hierarchical structure of the cluster.
Tevap = 540 8C. Bar scale: 1 mm. (b) SEM micrograph of the Zn aggregates collected on the reactor walls. Tevap = 540 8C. Bar scale: 1 mm. (c) TEM images of aerosol
aggregates sampled in the gas flow on carbon coated 3 mm copper grid. Bar scale: 100 nm. (d) Histogram of counts for the smaller aerosol particles synthesized in the
gas flow Tevap = 540 8C.
the furnace are produced with a concentration of the
condensing Zn molecules (monomers) that is proportional to
the saturation (or equilibrium) vapor pressure pSA of the zinc at
the operating temperature. At the output of oven, metal vapors
undergo an expansion through a nozzle and a contemporary
cooling with dilution of a fresh 500 cm3/min nitrogen flux. The
quick cooling down of zinc vapors determines the occurring of
supersaturation condition. Supersaturation starts the homo-
geneous nucleation of Zn particles, namely the formation of
stable, chemically homogeneous molecular Zn clusters.
Several papers [18,26] demonstrate that smaller size of
aerosol particles correspond to lower saturation vapor pressure
pSA, high cooling rate, low system pressure, low molecular
weight of the carrier gas, low surface tension of the species
vaporizing.
Aerosol particles are collected for XRD analysis on high-
efficiency quartz fiber filters. For characterization at SEM and
TEM, Zn nanoparticles are deposited, by a thermal precipitator
and a vacuum impactor, on optical glasses 18 � 18 mm (Menzel-
Glaser) and 3 mm/400 mesh carbon-coated copper grids (Agar
Scientific), respectively. Zinc fibers are collected as a skein of
sponge-like material at the exit of the expansion nozzle after run
times typically of 1 h scanning electron microscope is LEICA
Cambridge 360 Field Emission SEM operated at an acceleration
voltage typically 20 kV and a probe current of 100 pA.
Transmission electron microscope is Philips EM 208 S operated
at acceleration voltage 100 kV with a probe current 20 mA.
X-ray diffraction (XRD) measurements are carried out on a
PW1710 Philips rotating anode X-ray diffractometer at Cu Ka1
(l = 1.54056 A) and Cu Ka2 (l = 1.54439 A) radiation with
generator operating at 40 kVand 20 mA. The X-ray wavelength
for the measurement is l = 1.54056 A. A scanning step of
0.0208 is applied to record the patterns in the 2u range of 20–
858. To obtain stable analytic results, step-scan is set at 2.0 s per
step. Reduction of raw data includes Ka2-stripping, smoothing
and corrections for background intensity. Peaks position is
determined by means of a quadratic differential curve of the
measurement data.
Commercial Zinc dust powder used for comparison in XRD
analysis is obtained by Fluka Chemical and used as received.
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–29102902
DRIFT spectra are recorded on a Digilab FTS-60 equipped
with a KBr beamsplitter and a MCT detector operating
between 400 and 4000 cm�1, using an Harrick reaction
chamber with KBr windows, and an Harrick DRA-2C1 diffuse
reflectance accessory. The test cell is connected to vacuum and
gas lines for sample treatment, and its temperature is controlled
by a programmer from RT to 450 8C. The volatile decom-
position products are detected downstream within the carrier
gas by an on-line quadrupole mass-spectrometer (Leda-Mass
Satellite, 0–200 amu), interfaced with the DRIFT cell by a
heated stainless steel line. In each DRIFTS experiment a small
amount of sample (�0.1 mg) is located in a sample holder of
5 mm diameter on a Boron Nitride (BN) bed. Blank
experiments with BN beds showed almost complete inertia
towards O2 and NO2 in the temperature range studied. The
sample layer is 2–3 mm thick. Spectra are measured in
Kubelka–Munk (K–M) units. Backgrounds are recorded by
using dry KBr powder.
All the samples are treated according to the following
procedure: (a) Calcination under O2 (30%)/He flow (10 mL/
min) at 200 8C for 1 h (heating rate 5 8C/min). (b) Cooling
down to RT in He flow (10 mL/min). (c) Replacement of He
Fig. 2. (a) and (b) SEM pictures of a sponge of nanofibers obtained by the deposition
feedstock Zn grains is the same as in Fig. 1, Tevap = 540 8C. (c) and (d) TEM microg
fibers are mostly hollow inside.
flow with a NO2 (1000 ppm vol. ca.)/He flow (10 mL/min). (d)
Heating-up to 300 8C at 10 8C/min rate, holding the
temperature for 6 min at room temperature (RT), 50, 100,
150, 200, 250, 300 8C. DRIFT spectra are recorded 2 min after
the reaching of the plateau temperature. (e) Cooling-down in
NO2/He flow. After cooling at RT, a comparative spectra are
recorded for each sample to check with the spectra at the
beginning of the heating treatment.
3. Results and discussion
Fig. 1a is a typical SEM image a Zn aggregate collected by
thermal precipitator on a 18 mm � 18 mm glass support when
furnace is operated at Tevap = 540 8C. We observe that the
process of aggregation is hierarchical in nature in that smaller
particles stick and partially sinter together to form larger
globular units of some hundreds nanometers. These last are the
constituting bricks of clusters of some microns and these, again,
can form by sticking even larger ultimate aggregates. Fig. 1b
shows a fragment of Zn stratified aggregates, generated in the
carrier gas (nitrogen) at the same Tevap = 540 8C and collected
directly in correspondence of the outlet section of furnace on
of the Zn metallic vapors. Bar scale: 5 mm. The evaporation temperature for the
raphs of samples relative to the same experiments, which illustrate the fact that
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–2910 2903
Fig. 3. XRD patterns of zinc aerosol aggregates (a: top), zinc hollow nanofibers
(b: middle) and commercial Zinc dust (c: bottom). Nanoparticle aggregates and
nanofibers are synthesized at Tevap = 540 8C. All the three samples are indexed
as hexagonal zinc structure. The (h k l) lattice reflections are indicated.
the cold surface of the quartz reactor, where nanoparticles
aggregates stick as effect of thermophoresis. This powder,
which is constituted by stratification of Zn clusters typically
sized below 10 mm, is the first kind of samples that are object of
the present study. In order to observe the smaller particles we
collect some samples directly from the gas flow on carbon
coated 3 mm grids and observe them by TEM. Fig. 1c is an
example of isolated aggregate composed by three main larger
sub-units sized some hundred nanometers. The evaporation
temperature is the same above Tevap = 540 8C. At the periphery
of each of this globes constituting the cluster, smaller
nanoparticles are visible. They are disposed as a necklace in
the image section and coating the entire surface. These very
small particles are partially sintered, with dimensions
distributed according to a log-normal size distribution with
modal peak about 20 nm, average size 51 nm and standard
deviation 60 nm. Fig. 1d is the associated histogram of counts.
On the basis of previous observations, we comment that the
aggregation process starts directly in the gas flow as effect of
nanoparticle reciprocal collisions. It is possible to evaluate a
characteristic coalescence time tcoal [18,26], i.e., the time
requested to two particles to coalesce together (sintering) and
form a new larger spheroidal particle. If the collision
characteristic time tcollis [18,26] is larger or smaller with
respect to tcoal, the net result of the process is the formation of
spheroid or chain-shaped aggregates, respectively.
Fig. 2a and b are examples of SEM micrographs of nanofiber
bundles obtained by the condensation–deposition and growth
process occurring within the expansion nozzle of reactor. The
evaporation temperature of bulk Zn grains feedstock is still
Tevap = 540 8C. This is the second type of nanopowders object
of the present study. It is possible to distinguish both thin fibers
(external diameter about 85–200 nm) and more thick nanorods
(diameter about 400–700 nm). Previous work [16] has shown
that most of these fibers are hollow inside, thus resulting in Zn
crystalline nanotubes, about 100 nm width with wall thickness
variable within 10–20 nm. Fig. 2c and d show a group of Zn
nanotubes with an average diameter of about 85 nm, as inferred
by histogram of counts (not shown here) obtained from
processing all the available TEM pictures relatively to the
fibrous samples corresponding to Tevap = 540 8C.
Zn aggregates and hollow nanofiber powders as synthesized
above with Tevap = 540 8C are analyzed by XRD. The results are
shown in Fig. 3a and b, respectively. Fig. 3c is the diffraction
pattern of commercial Zn dust here taken for comparison.
Clusters of aerosol particles (Fig. 3a) can be indexed as zinc
hexagonal (wurtzite) phase with lattice constants a = 2.649 A
and c = 4.924 A, as evaluated for the two Zn lattice reflections
(1 0 1) and (1 0 0) corresponding to d = 2.080 and 2.295 A,
respectively. The inferred lattice constants are consistent with
the standard values for bulk Zn, a = 2.665 A, c = 4.947 A
(JCPDS Card no 4-0831). However, the (1 0 1) peak is stronger
with respect to the (1 0 0) peak, being the ratio about 3.53,
contrarily to the bulk metallic Zn (JCPDS Card no 4-0831) for
which the intensity ratio of (1 0 1) with respect to (1 0 0) is
about 2.50. Moreover, the ratio between the observed (1 0 0) to
(1 0 2) is 2.6 with respect to an expected ratio (1 0 0) to (1 0 2)
equal 1.4 for bulk zinc. Thus, the peak (1 0 2) appears to be
strongly depressed with respect to standard metallic zinc. The
same is observed for the (1 0 3), (1 1 0) and (1 1 2) lattice
reflections.
From XRD pattern of hollow nanofibers (Fig. 3b) it is
possible, also in this case, to index the sample as hexagonal
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–29102904
Table 2
Correspondence between chemical species and infrared absorption band loca-
tions from the previous literature
Species Band (cm�1)
NO (gas) 1876
zinc. Following the same procedure above, from the two Zn
lattice reflections (1 0 1) and (1 0 0) we obtain lattice constants
a = 2.662 A, c = 4.954 A and d = 2.090 and 2.306 A, respec-
tively. By comparison of the measured peak intensity with
respect to the expected values from bulk zinc (JCPDS Card no
4-0831) it is possible to comment that, in the case of nanofibers,
the peak (1 0 0) of nanofibers is strongly depressed. As
discussed elsewhere [16], this can be interpreted as an indicator
of the fact that the nanofibers grow preferentially along the
(1 0 0) plane. Nanofibers grown at Tevap = 540 8C will be the
second type of nanostructured powders here object of study.
The diffraction pattern of Zn commercial dust (Fig. 3c) is
indexed as hexagonal zinc with lattice constants a = 2.660 A
and c = 4.947 A evaluated as described above for the two Zn
lattice reflections (1 0 1) and (1 0 0) corresponding to d = 2.088
and 2.303 A, respectively.
We observe that in the case of cluster formed by aerosol
particles the c-value is slightly smaller. This accord SEM
observations (not reported here) that reveal hexagonal prism
crystals with an aspect ratio H/W lower than unity where H, W
are the height and the width of the hexagonal prism structure.
Scherrer’s analysis of XRD peaks is used to infer the
crystallite size for the three classes of samples. In particular, the
line broadening of the (1 0 1) Zn diffraction peak of 2u patterns
is considered. The instrumental broadening line of the
diffractometer is measured by a NIST standard (LaB6). The
grain size hei(1 0 1) is evaluated by the Scherrer’s equation
hei = (0.9 l)/(b cos u), where l is the radiation wavelength
(1.54056 A), b the FWHM-width of the diffraction peak
corrected for the instrumental broadening and 2u the diffraction
angle corresponding to the considered peak.
Table 1 reports in column 2 and 3 the summary of the
previous results.
The values of hei(1 0 1) are reported in Table 1, column 4. We
note that the apparent size is smaller in the case of the aerosol
particles which produce the cluster deposit on cold wall reactor
and considerably larger in the case of standard Zn.
The surface interactions with NO2 by the two kinds of
nanostructures, namely, aggregate with respect to fiber
powders, are extensively investigated by DRIFT spectroscopy
experiments and compared to standard Zn dust powder.
It is well known from the literature that NO2 behaves
according to a complex scheme on the metallic zinc as
recognized by previous authors [27,28]. After adsorption of
nitrogen dioxide, the detection of N, O, NO, NO2 and NO3 on
the metal surface have been reported [28] on the basis of
Table 1
Columns 2 and 3: lattice parameters a, c of the Zn hexagonal phase evaluated
from (1 0 1) and (1 0 0) XRD peaks for the three Zn structures object of this
study and column 4: apparent crystallite size evaluated from the line broadening
of the (1 0 1) peak for each of the considered structure
a (A) c (A) hei(1 0 1) (A)
Cluster of AP 2.649 4.924 662
Nanofibers 2.662 4.954 876
Standard Zn 2.660 4.947 1111
All units in Angstrom.
synchrotron photoemission and absorption near-edge spectro-
scopy (XANES) measurements. NO2 is recognized by those
authors to be a very good oxidizing agent for prepare zinc oxide
from metallic Zn.
Concerning to our work, we stress that as a consequence of
pre-treatment and subsequent exposure to NO2 it is presumable
the formation of a thin film of ZnO at the surface of Zn
nanofibers and nanoparticle aggregates.
The preliminary experiments by DRIFTS demonstrate that
zinc aggregates and zinc nanofibers are strongly absorbing with
respect to infrared radiation. The baseline shift occurring
during our experiments is a typical effect encountered in VT-
DRIFT spectra [29] enhanced by the non-ideality of our
samples. In fact, a similar behavior is recorded also during the
oxidation pretreatment (spectra not shown). Moreover, the
comparative spectra recorded at RT after cooling-down, step (e)
of experimental procedure, do not show any significant change
of baseline shape with respect to the first spectra recorded at the
beginning of heating-up phase, step (d). Only for stratified
aggregates there is a change, but of minor entity, of the whole
spectrum after the NO2 adsorption. Any contribution of boron
nitride spectra to the spectra reported below is excluded from
the fact that separate experiments performed on pure boron
nitride, which replicate the same treatments of zinc samples,
show complete inertness of BN towards oxygen and NO2.
Difference spectra are taken in each DRIFTS experiment,
namely, the spectrum when NO2/He mixture (1000 ppm NO2)
minus that at RT when pure He gas is flowing.
Table 2 reports the infrared absorption bands relative to
nitrogen oxides and neutral/charged adsorbed species from the
analysis of literature [30–32].
It is well known [28] that the interaction of NO2 on the zinc
surface, even at room temperature, is expected to generate NO
species and zinc surface oxidation thus generating a ZnO
coating above the metal surface according to the following
reaction:
ZnðsolidÞ þ NO2ðgasÞ ! ZnOðsolidÞ þ NOðgasÞ (1)
Such interaction is found to be favorite at higher temperatures
[28]. Thus, as discussed above, beside pre-treatment of samples
NO+ (ads) 2390–2102
NO (ads) 1966–1710
NO2 (gas) 1612
NO2 (ads) 1642–1605
Species n3 n1
NO3� (monodentate) 1530–1480, 1290–1250 1035–970
NO3� (bidentate) 1565–1500, 1300–1260 1040–1010
NO3� (bridging) 1650–1600, 1225–1170 1030–1000
NO2� (bridging nitrite) 1220–1205
NO2� (nitro) 1440–1335 1350–1315
NO2� (monodentate) 1470–1450 1065–1050
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–2910 2905
Fig. 4. Samples of stratified Zn aggregates reported in Fig. 1 are studied by DRIFTexperiments. Difference spectra (spectra under NO2/He minus spectra before NO2/
He introduction) at 860–2640 cm�1 region at increasing temperature under NO2/He flow (1000 ppm vol. ca.). (a) RT; (b) 50 8C; (c) 100 8C; (d) 150 8C; (e) 200 8C; (f)
250 8C; (g) 300 8C.
in O2/He flow, a ZnO film is expected to be formed on the surface
of aggregates or nanofibers as a consequence of reaction with test
gas NO2.
NO (gas) is detected in our experiments by DRIFTS and
quadrupole mass spectrometry measurements after admittance
of NO2/He mixture in the reaction chamber.
Fig. 4 shows the DRIFTS spectra of nanoparticle aggregate
powder, synthesized at Tevap = 540 8C (see Fig. 1), exposed in
the test chamber to the NO2/He flow at different test
temperatures (RT to 300 8C). A doublet located at 1628 and
1600 cm�1 is visible in all the temperature range investigated,
thus decreasing at the increase of temperature, which could be
ascribed to R and P branches of NO2 free (gas).
A broad band located at 1300 cm�1 is evident in low
(<250 8C) temperature spectra and decreases raising the
temperature disappearing almost completely at 300 8C. Around
1500 cm�1 another band very broad and feeble seems to be
present. The analysis of bands evolution of Fig. 4 is
complicated by the fact that, at increasing the temperature, a
negative feature is observable between 1550 and 1300 cm�1.
Thus, the two peaks at 1300 and 1500 cm�1 could be ascribed
to two vibrational modes of mono- or bidentate NO3� species
[31] adsorbed on the ZnO thin film surface formed on the Zn
surface, even if for a definitive attribution the presence of
eventual absorption bands in the 970–1040 cm�1 region (not
observable in our experiments) should be checked.
At the increase of temperature the band at 1300 cm�1 first
grows till 150 8C. Thereafter, it becomes less pronounced and it
almost disappears at 300 8C. The band is restored by cooling
down the sample at RT in NO2/He flow.
Fig. 5 are the DRIFT spectra at the same working
temperatures above recorded in the case the hollow nanofibers
powder grown by the deposition within the quartz reactor
orifice of metallic zinc vapors obtained at Tevap = 540 8C.
Preliminarily, we note that the differences of amplitude (about
one order magnitude) of the IR absorption bands between Zn
aggregates (Fig. 4) and Zn nanofibers (Fig. 5) could be due to
the different amount of diffused IR radiation emerging from the
sample resulting from different packing of sample cup and/or
different nanostructure (micro-crystalline aggregates versus
mono-crystalline rods [16]).
The direct comparison of zinc fiber versus zinc aggregates
DRIFT spectra at the same operated temperature reveals
dramatic differences between them.
Concerning analysis of nanofiber spectra recorded under
NO2/He flow in Fig. 5, we note that at room temperature no
bands attributable to adsorbed species are evident. Only the
doublet located at 1600–1628 cm�1, due to NO2 (gas) free is
present. The presence of NO2 feebly adsorbed on Zn2+ sites can
not be excluded since the IR adsorption of such species fall in
the 1642–1605 cm�1 range that is masked by gas phase NO2
adsorption. Upon increasing the temperature from RT to
300 8C, the NO2 doublet disappears and a broader shoulder
between about 1900 and 1500 cm�1 grows up. At the mean time
also a relatively narrow peak located at 2500 cm�1 appears.
Contemporarily a new, complex spectral feature does appear
between 1400 and 1200 cm�1.
Fig. 6 shows the DRIFT spectra obtained for commercial
zinc dust (Fluka Chemicals) in the same variable temperature
experiments of Figs. 4 and 5.
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–29102906
Fig. 5. Samples of Zn hollow nanofibers reported in Fig. 2 are studied by DRIFT. Difference spectra (spectra under NO2/He minus spectra before NO2/He
introduction) at 860–2640 cm�1 region at increasing temperature under NO2/He flow (1000 ppm vol. ca.). (a) RT; (b) 50 8C; (c) 100 8C; (d) 150 8C; (e) 200 8C; (f)
250 8C; (g) 300 8C.
At RT the spectra of standard Zn powder are markedly
different with respect to those of Zn aggregates (Fig. 4) and
hollow fibers (Fig. 5) even if the 1600–1628 cm�1 doublet of
NO2 gas absorption is observed also in this case. In particular,
three intense bands located respectively at 1510, 1315 and
1023 cm�1 are visible, which are increasing with the time of
contact with NO2. These bands can be attributed to adsorbed
bidentate nitrates (Table 2).
At increasing the temperature from RT to 300 8C, these
1510, 1315, 1023 cm�1 absorption bands exhibit an initial
Fig. 6. DRIFT difference spectra (spectra under NO2/He minus spectra before N
2640 cm�1 region at increasing temperature under NO2/He flow (1000 ppm vol. ca.
growth till 100 8C, followed by a progressive depression, which
can be interpreted as the process of thermal decomposition of
adsorbed nitrates. In the meanwhile a negative feature located
at 1650–1760 cm�1 develops. After cooling down to RT the
sample in NO2/He flow all the bands are restored. This can be
interpreted as the fact that the cooling down at RT again yields
adsorption of NO2 free gas on the ZnO thin film surface
previously formed on samples with formation of fresh nitrates.
In fact, nitrites and nitrates are rather unstable at 300 8C in NO2/
He atmosphere and most probably they decompose. After
O2/He introduction) of the standard Zn dust available in commerce at 860–
). (a) RT; (b) 50 8C; (c) 100 8C; (d) 150 8C; (e) 200 8C; (f) 250 8C; (g) 300 8C.
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–2910 2907
cooling down at RT they can be formed again and result stable
under the pure He flow. The alternative process of nitrites NO2�
and nitrates NO3� destruction/creation can be iterated by
periodic thermal cycles of heating/cooling and, in this sense, it
can be retained reversible.
The following interpretative scheme for surface reactions
yielding nitrites and nitrates has been reported in the literature
for SnO2 powders [33]:
NO2ðgasÞ þ e�ðconductive bandÞ ? NO2�ðadsÞ (2)
NO2ðgasÞ þ VOþ? NO2
�ðadsÞ þ VO2þ (3)
2NO2ðgasÞ þ O2�ðsolidÞ þ e�ðconductive bandÞ
? 2NO3�ðadsÞ (4)
2NO2ðgasÞ þ O2�ðsolidÞ þ VO
þ? 2NO3
�ðadsÞ þ VO2þ
(5)
NO2ðgasÞ þ O�ðsolidÞ ? NO3�ðadsÞ (6)
In the above Eqs. (3)–(5) VO+, VO
2+ are oxygen vacancies at
ZnO film surface over Zn samples.
Our previous experimental work on current response
measurements [34] of Zn nanopowders (nanofibers and
nanoparticle aggregates) when exposed at NO2 sub-ppm levels
revealed relative abundance of conductive electrons in the case of
nanofibers at room temperature. Also in that case samples were
pre-treated, specifically by exposure to humid synthetic air for
24 h at 200 8C. In particular, the current response of nanofibers at
RT, 50%RH and 0.4 ppm NO2 resulted almost 1 and 4 order of
magnitudes higher with respect to nanoparticle aggregate and
standard Zn powders, respectively. The essential point is that,
even at low temperature, a thin film of ZnO is probably formed
soon on the sample powders. The current response of the
nanofibers samples decreases at the increasing temperature from
RT to 150 8C. This previous experimental evidence is in contrast
Fig. 7. Detail of difference spectra (spectra under NO2/He minus spectra before NO2/
under NO2/He (1000 ppm vol. ca.). Left: Zn stratified aggregates; Right: Zn nanofi
200 8C; (f) 250 8C; (g) 300 8C. Each curve is arbitrarily shifted up to discern clos
with the previous scheme above of nitrites and nitrates formation,
in that all the surface reactions (2) to (5) above, except Eq. (6),
lead to surface trapping of electrons at increasing temperature,
i.e., when decomposition of nitrites and nitrates occur.
A supplementary mechanism can be advocated to justify the
previous experimental results [34]. Kase et al. [35] proposed a
particular process of NO chemisorption by the formation of the
so-called pseudo-oxygen defects. Again, the mechanism is cited
according to the presumable formation of a thin film of ZnO on
the surface of nanofiber and nanoparticle aggregates samples. It
may describe the observed changes of electrical conductivity
for an n-type semiconductor, which undergo contemporary NO
chemisorption and is written as the following:
O2�ðsÞ þ NO ! PVO� þNO2
�ðadsÞ (7)
That results also in an increase of conductive electrons
through the following:
PVO� $ PVO
� þ e� (8)
In the above PVO� and the asterisk denote an electronvacancy
at pseudo-oxygen defect and the neutral state, respectively; (s)
and (ads) indicating the surface ion and chemisorbed species.
Eq. (7) means that an oxygen defect having a quasi-free electron
is produced on the oxidation of NO using the surface oxygen; the
quasi-free electron is released from the surface oxygen defect
through Eq. (8), leading to the electrical conductivity increase.
In our case we propose at RT a similar mechanism occurring
for NO2 at ZnO film interface of nanopowders described by the
following reaction:
O2�ðsÞ þ NO2ðgasÞ ! PVO� þNO3
�ðadsÞ (9)
leading to adsorbed NO3� and releasing of conductive electrons
from the surface oxygen defects. This process, according to the
cited paper [35] is most probable to occur at RT and thwarted at
higher temperatures at which nitrates are not stable. The
He introduction) relatively to 2440–2620 cm�1 region at increasing temperature
bers. Legend key: (a) room temperature; (b) 50 8C; (c) 100 8C; (d) 150 8C; (e)
e plots, scale on the left is a bar scale for each separate curve.
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–29102908
Fig. 8. Detail of difference spectra (spectra under NO2/He minus spectra before NO2/He introduction) relatively to 1500–1800 cm�1 region at increasing temperature
under NO2/He (1000 ppm vol.ca.). Left: Zn stratified aggregates; right: Zn nanofibers. Legend key: as in Fig. 7.
decrease of bands attributed to adsorbed nitrites and nitrates at
the increasing of the temperature should be related to their
thermal decomposition.
Figs. 7 and 8 show a direct comparison of the DRIFT
difference spectra of cluster versus nanofiber samples on
enlarged wavenumber scales, 2440–2630 cm�1 and 1500–
1800 cm�1, respectively. As already quoted above, at increas-
ing temperatures, the nanofibers spectra exhibit the rising of a
narrow peak at 2500 cm�1 and, contemporarily, the progressive
smoothing and disappearing of the doublet at 1600–1628 cm�1,
thus indicating probable consumption due to further oxidation
of Zn to ZnO and/or desorption of NO2 at increasing
temperatures from RT to 200 8C.
The distinctive feature retrieved in the case of nanofibers, is
analyzed in Fig. 9 by plotting the intensity (height) evolution of
1628 cm�1 branch of the NO2 doublet (squares) and 2500 cm�1
(diamonds). This furnishes the evidence that a correlation does
exist between the two mechanisms occurring at hollow
Fig. 9. Comparative plot of DRIFT bands intensity (height) for DRIFT spectra
of Zn hollow nanofibers at 1628 vs. 2500 cm�1 spectral region as a function of
temperature under NO2/He flux (1000 ppm vol. ca.). Symbols: squares
for1628 cm�1; diamonds for 2500 cm�1.
nanofibers surface during interaction with NO2/He. The
increasing rate of 1628 cm�1 band intensity at increasing
temperatures is linear such as the decreasing rate of 2500 cm�1
band and the two slopes are of reversed sign each with respect
to the other. It should be stressed that the band height at
1628 cm�1 (squares) does not vary between 250 and 300 8Cand the relative values are not strictly zero owing to the fact that
spectra ‘f’ and ‘g’ are shifted upwards in order to hold a unique
method for band intensity evaluation. Nevertheless, this point
does not affect our conclusions since the relative peak height
variations are investigated and not the absolute values.
The assignment of 2500 cm�1 band in Fig. 7-right
(nanofibers) is not easy. According to the literature (see [30]
and references cited therein), we found out that bands about
2500 cm�1 may be attributed to the presence of adsorbed
polymeric species of type (NxOy)z, but the absence of the band
typical of NO2 dimers located at 1748–40 cm�1 does not
support previous hypothesis, thus the nature of this band
remains rather unclear. Therefore, further work is necessary in
order to clarify this point.
A structural peculiarity of the hollow fibers, which should be
accounted for in the discussion of surface reactivity effects, is
the huge surface area with respect to the other samples. For
instance, 1 g of Zn hollow fibers with 100 and 60 nm as external
and internal diameters, respectively, coherently with the TEM
results, at a temperature 25 8C (density of Zn [36] equal to
7.14 g/cm3), can be associated to a surface about
140 � 103 cm2. The same weight of 20 mm sized Zn grains
(commercial standard sample) is characterized by a surface
about 0.42 � 103 cm2.
The evaluation of the surface of clustered Zn aggregates
deserves a careful consideration. We have observed by TEM
that the clusters formed on the reactor wall are about 5–10 mm
size and that at the surface a layer of much smaller particles
(20–50 nm) are observed such as a necklace all around (see
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–2910 2909
Fig. 1b and c). If we consider a size of 10 mm (or 5 mm) and a
layer of 50 nm smaller particles coating the surface, a weight of
1 g of such powder will correspond to a total surface about
3.43 � 103 cm2 (or 0.87 � 103 cm2).
Thus, the ratio between the specific surface (cm2/g) of
hollow nanofiber with respect to Zn standard powder will be
about 330 and the ratio between the specific surface (cm2/g) of
hollow nanofibers with respect to stratified aggregates powder
will be about 40 and 160 for clusters of 10 and 5 mm,
respectively. Therefore, for any mechanism occurring at the
surface of the sample, the hollow fibers are expected to possess
exceptional sensitivity with respect to the other two kinds of
samples. On the other hand, the layer of smaller 50 nm
particles observed on the surface of the clusters collected on
reactor walls (Fig. 1b), has a negligible effect on the total
volume, but a dramatic (300% enhancement) on the total
surface of the resulting structures. Thus, on a geometrical
basis, Zn clusters are expected to possess a surface reactivity
intermediate with respect to the one of Zn nanofibers and
standard Zn powder. It should be noted that the above
reasoning is based exclusively of size effects, that is, it neglects
eventual role of the impurities and surface defects associated to
the surface of the Zn samples, which, on the other end, are
found all indexed as hexagonal lattice (wurtzite) on the basis of
XRD analysis
4. Conclusions
Zinc powders constituted by nanoparticles aggregates and
hollow nanofibers have been synthesized via thermal evapora-
tion process at the same Tevap = 540 8C followed by condensa-
tion of the metallic vapors. Preliminary characterization of
samples is made by electron microscopy analysis (SEM, TEM)
and X-ray diffraction (XRD). Thereafter, the two kinds of
samples have been studied with respect to NO2 reactivity by
DRIFTS and compared with standard Zn dust. The surface
interaction chemistry of the two types of zinc nanostructures is
found to be very different, depending markedly on their shape,
scale and geometry. DRIFT spectra of zinc aggregates and
standard dust show the formation of nitrates (NO3�) and nitrites
(NO2�) adsorbed species, that are stable at RT, decompose as
the temperature increases, and can be again formed at RT after
their decomposition.
On the contrary, DRIFT spectra of nanofibers exhibit a
peculiar reactivity with respect to test gas NO2. A specific
absorption band at 2500 cm�1 is recorded for nanofibers, which
is not observed at all for irregularly shaped zinc aggregates of
nanoparticle primary particles and for standard commercial Zn
dust powder. We found out that the spectra of zinc hollow
nanofibers, external, internal diameters about 100, 60 nm, when
exposed to NO2/He flux, are characterized by two distinctive
absorption bands at 2500 cm�1 and 1600–1628 cm�1, the last
one representing the NO2 free gas.
Bands at 2500 cm�1 and 1600–1628 cm�1 are observed to
grow and fall off as temperature rises from RT to 300 8C with
the increasing rate of the former band related to the decreasing
rate of the latter band. This feature is not at all observed in the
case of the other samples and, therefore, may represent a
possible fingerprint of Zn nanofibers structure with respect to
species formed at their surface when fibers are exposed to gas
NO2. This peculiarity can be explained as nanoscale effect
related to the much higher specific surface (cm2/g) and/or to a
confinement effect of hollow nanofibers with respect to
commercial Zn standard. Zn aggregates collected on reactor
walls are estimated to possess intermediate specific surface
with respect to hollow nanofibers and standard dust.
Acknowledgements
Kind assistance of Mr. Giuseppe Mocci from Materials and
Production Eng. Dept. (DIMP) University of Napoli ‘‘Federico
II’’ during XRD measurements is sincerely acknowledged.
Work supported by the FIRB-MIUR grant # RBAU01K749.
References
[1] Y. Xia, P. Yang, Y. Sun, Y. Wu, B. Mayers, B. Gates, Y. Yin, F. Kim, H.
Yan, Adv. Mater. 15 (2003) 353.
[2] Z.W. Pan, Z.R. Dai, Z.L. Wang, Science 291 (2001) 1947.
[3] Z.R. Dai, Z.W. Pan, Z.L. Wang, Adv. Funct. Mater. 13 (2003) 9.
[4] Z.L. Wang, Z.W. Pan, Int. J. Nanosci. 1 (2002) 41.
[5] X.W. Sun, H.S. Kwok, J. Appl. Phys. 86 (1999) 408.
[6] J. Xu, Q. Pan, Y. Shun, Z. Tian, Sens. Actuators B 66 (2000) 277.
[7] K.L. Chopra, S.R. Das (Eds.), Thin Film Solar Cells, Plenum, New York,
1983.
[8] C.T. Troy, Photonics Spectra 31 (1997) 34.
[9] K. Vanheusden, C.H. Seager, W.L. Warren, D.R. Tallant, J. Caruso, J.
Hampden-Smith, T.T. Kodas, J. Lumin. 75 (1997) 11.
[10] A.A. Onischuk, P.A. Purtov, A.M. Baklanov, V.V. Karasev, S.V. Vosel, J.
Chem. Phys. 124 (2006) 14506.
[11] M. Nishizawa, V.P. Menon, C.R. Martin, Science 268 (1995) 700.
[12] T.M. Whitney, J.S. Jing, P.C. Searson, C.L. Chien, Science 261 (1993)
1316.
[13] W.K. Hsu, J. Li, H. Terrones, N. Grobert, Y.Q. Zhu, S. Trasobares, J.P.
Hare, C.J. Pickett, H.W. Kroto, D.R.W. Watton, Chem. Phys. Lett. 301
(1999) 159.
[14] E. Braun, Y. Eichen, U. Sivan, G. Ben-Yoseph, Nature 391 (1998)
775.
[15] B. Nikoobakht, Z.L. Wang, M.A. El-Sayed, J. Phys. Chem. B 104 (2000)
8635.
[16] S. di Stasio, Chem. Phys. Lett. 393 (2004) 498.
[17] F.E. Kruis, H. Fissan, A. Peled, J. Aerosol Sci. 29 (1998) 511.
[18] Y. Singh, J.R.N. Javier, S.H. Ehrman, M.H. Magnusson, K. Deppert, J.
Aerosol Sci 33 (2002) 1309.
[19] C.G. Granqvist, R.A. Buhrman, J. Appl. Phys. 47 (1976) 2200.
[20] H. Yumoto, R.R. Hasiguti, T. Kaneko, J. Cryst. Growth 75 (1986)
289.
[21] K. Recknagle, Q. Xia, J.N. Chung, C.T. Crowe, H. Hamilton, G.S. Collins,
NanStruct. Mater. 4 (1994) 103.
[22] Y. Wang, L. Zhang, G. Meng, C. Liang, G. Wang, S. Sun, Chem. Commun.
(2001) 2632.
[23] S. di Stasio, A. Onischuk, A. Baklanov, in: Proceedings of the Interna-
tional Aerosol Conference IAC 2002, Taipei, Taiwan, (2002), p. 343.
[24] S. di Stasio, V. Dal Santo, in: Proceedings of the European Aerosol
Conference, Madrid, (2003), p. 563.
[25] S. di Stasio, A. Onischuk, C. Baratto, G. Sberveglieri, in: Proceedings of
the European Aerosol Conference, Madrid, (2003), p. 565.
[26] S. Panda, S.E. Pratsinis, NanoStruct. Mater. 5 (1995) 755.
[27] J.A. Rodriguez, J. Hrbek, J. Vac. Sci. Technol. A 12 (1994) 2140.
[28] J.A. Rodriguez, T. Jirsak, J. Dvorak, S. Sambasivan, D. Fischer, J. Phys.
Chem. B 104 (2000) 319.
S. di Stasio, V. Dal Santo / Applied Surface Science 253 (2006) 2899–29102910
[29] R.L. White, Anal. Chem. 64 (17) (1992) 2010.
[30] K. Hadjiivanov, Catal. Lett. 68 (2000) 157.
[31] K. Hadjiivanov, Catal. Rev Sci. Eng. 42 (2000) 71.
[32] T. Eickhoff, P. Grosse, W. Theiss, Vibr. Spectrosc. 1 (1990) 229.
[33] A. Chiorino, G. Ghiotti, F. Prinetto, M.C. Carotta, D. Gnani, G. Martinelli,
Sens. Actuators B 58 (1999) 338.
[34] C. Baratto, G. Sberveglieri, A. Onischuk, B. Caruso, S. di Stasio, Sens.
Actuators B 100 (2004) 261.
[35] K. Kase, M. Yamaguchi, T. Suzuki, K. Kaneko, J. Phys. Chem. 99 (1995)
13307.
[36] B.J. Aylett, in: J.C. Bailar, A.F. Trotman-Dickenson (Eds.), Comprehen-
sive Inorganic Chemistry, Pergamon Press, Oxford, 1973.
