Copyright © 2007 American Scientific Publishers
All rights reserved
Printed in the United States of America
Journal of
Nanoscience and Nanotechnology
Vol. 7, 1–9, 2007
Homogeneous Precipitation of Dy2O3
Nanoparticles—Effects of Synthesis Parameters
Happy1 , A. I. Y Tok1 ∗ , L. T. Su1 , F. Y. C. Boey1 , and S. H. Ng2
1
2
School of Materials Science and Engineering, Nanyang Technological University, Singapore 639798
AMR International Corp—Singapore Representative Office, 61 Science Park Road, #01-19 The Galen
Singapore Science Park III, Singapore 117525
The experimental parameters that control the size and size distribution of dysprosium oxide nanoparticles synthesized by homogeneous precipitation technique have been systematically investigated. The particles were characterized with respect to their size, shape, and thermal decomposition
behavior. It was found that the precipitated particles were spherical, uniform in size, and amorphous,
which upon heating in air, decomposed into the oxide form with no change in morphology. The size
and size distribution of the particles showed strong dependence on the metal cation concentration
([Dy3+ ]) and weak dependence on urea concentration and aging time. In addition, the presence of
chlorine ions (Cl− ) was found to have significant effect on the growth and agglomeration of the particles. Aggregation mechanism as the growth mechanism is offered to explain the effects of these
synthesis parameters on the morphology, size, and size distribution of dysprosium oxide particles.
Keywords: Nanoparticles, Dysprosium Oxide, Homogeneous Precipitation.
Nano-particles are usually referred to as condensed phase
particles in the size range of 1–100 nm in diameter. In this
size range, the particles have a high proportion of atoms
located at its surface as compared to bulk materials, giving
rise to unique physical and chemical properties that are
totally different from their bulk counterparts.1 The oxides
of rare earth elements such as yttrium, neodymium, samarium, europium, dysprosium, and lutetium have found many
important applications as catalysts, high efficiency phosphors, and magnetic to dielectric formulations for multilayer ceramic capacitors.2–5 In the preceding years, some
successful efforts have been made to synthesize the nanoparticles of rare earth oxides.6–10 Among them, homogenous precipitation by urea has emerged to be the most
flexible and promising technique as it is relatively simple,
reproducible, and economically feasible for potential large
scale production. The precipitates are dense, can be readily
filtered and are of considerably high purity. Furthermore,
they decompose to the oxide form at relatively low calcinations temperature without change in morphology.
Although some studies on the synthesis of rare earth
oxide particles using homogeneous precipitation by urea
have been reported, a proper understanding on the effect
∗
Author to whom correspondence should be addressed.
J. Nanosci. Nanotechnol. 2007, Vol. 7, No. 3
of synthesis parameters on the growth of the particles
during precipitation, which in turn affects size and size
distribution of the particles, has not been addressed. For
example, a few investigations on the effect of aging time
on the particle formation by homogeneous precipitation
using urea have been reported, however with contradictory results.10–12 14 Matijević11 reported that the particle
grew as the aging time increasing, yet the growth of the
particle discontinued after reaching a stable size. On the
contrary, Gao10 observed that the particle size decreased
and increased again with aging time increasing. Interestingly, Taş14 found out that gallium particles have a size
of 200–500 nm in length with “zeppelin” shape right after
collected at the precipitation start point, which grew to
1–2 m after aging for 95 minutes. Holding the solution for 24 hours, morphology change from “zeppelin” to
spherical particles with average particle size of 50–60 nm
was observed. Thus, there is no clear conclusion could be
drawn on how the particles grow during the precipitation
process.
In previous articles,11 15–16 the short nucleation time
model of LaMer17 has been used to describe particle
growth in a urea precipitation process. In this model,
nuclei are formed homogeneously from solution when the
reactive species in the solution reaches the supersaturation
point. These nuclei will grow further to become larger particles by capturing the solute species in solution through
1533-4880/2007/7/001/009
doi:10.1166/jnn.2007.203
1
RESEARCH ARTICLE
1. INTRODUCTION
RESEARCH ARTICLE
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
molecular diffusion process. However, in recent years,
aggregation mechanism has been described as a more relevant growth mechanism in homogeneous precipitation process by urea.14 18–20 For example, Hu21 showed that the
zirconia particles produced using urea precipitation consisted of much smaller crystalline sub-units.
The growth process in the aggregation model starts by
nuclei capturing solutes species through molecular addition to form small (primary) particles. Once these small
particles have been formed, they will aggregate with one
another forming larger, secondary (final) particles. The
aggregation process could be facilitated by Brownian
motion, Van der Waals forces, or changes in the chemical environment of the solution. For example, a change in
the pH or increase in ionic strength in the solution can
reduce the electrostatic barrier surrounding the particles,
thus promoting aggregation. This aggregation process will
stop when a stable size of secondary particles has been
reached. In the field of colloid stability theory, the barrier could be attributed to steric and/or electrostatic forces.
Such growth mechanism has been observed and described
for the preparation of mono-dispersed silica22 and gold
particles.23
Therefore, it is the aim of this report to verify on the
growth mechanism of rare earth oxide particles using
homogeneous precipitation by urea by collecting precipitates at different aging time. In addition, it is also important to understand the effect of other various synthesis
parameters on the size and size distribution of the precipitated nano-particles such as metal cation concentration, urea concentration, and chlorine ions, with the aim
of achieving narrow size distribution of rare earth oxide
particles.
2. EXPERIMENTAL DETAILS
In this study, dysprosium (III) oxide (99.99%), hydrochloric acid, and urea (98%) were used as raw materials. The
dysprosium oxide was obtained from AMR International
Corp2 , while the hydrochloric acid and urea were procured
from Sigma Aldrich.
Aqueous dysprosium chloride solution was prepared by
dissolving dysprosium oxide in an appropriate amount of
hydrochloric acid. In a typical synthesis, aqueous urea
solution was prepared by dissolving urea in 1500 mL
of deionized water (DI H2 O), mixed with an appropriate
amount of dysprosium chloride and stirred homogeneously
in a glass beaker. The amount of urea and dysprosium
chloride in the solution was adjusted depending on the synthesis conditions. This mixture was then heated at 90 C
in a water bath, stirred throughout the process. The aging
time in our experiment was defined after the first onset of
turbidity was observed during experiment (e.g., a bluish
tint). The variation of synthesis parameters in our study are
summarized in Table I. After the precipitation process, the
solution was cooled and the precipitates were centrifuged
2
Happy et al.
Table I. Experimental conditions for the preparation of the nanoparticles by homogeneous precipitation.
Concentration (M)
Sample
Variables
[Dy3+ ]
[Urea]
Aging time
(min)
1
2
3
4
5
6
Aging time
0.020
0.020
0.020
0.020
0.020
0.005
1.0
1.0
1.0
1.0
1.0
1.0
10
45
70
90
240
90
0.010
0.050
0.100
0.020
1.0
1.0
1.0
0.5
90
90
90
90
0.020
0.020
0.020
1.5
2.0
4.0
90
90
90
7
8
9
10
11
12
13
Metal Cation
Concentration
[Dy3+ ]
Urea Concentration
[Urea]
and washed thoroughly with deionized water (6×) and
ethanol (final stage). The precipitates were then dried in a
drying oven at 75 C for 18 hours. Dried precipitates were
calcined in air at different temperatures up to 700 C for
2 hours.
The morphology, particle size, and particle size distribution of the Dy2 O3 nano-particles were examined using
transmission electron microscopy (TEM JEOL 2010) and
field emission scanning electron microscopy (FESEM JSM
6340F). The TEM samples were prepared by dispersing
the precipitated dysprosium oxide powder in ethanol in an
ultrasonic bath and dropped onto the copper grid coated
with holey carbon. The SEM samples were prepared by
placing the powders on carbon tape and gold sputtered
for 50 s. Dynamic Light Scattering (DLS) was performed
using Malvern Zetasizer Nanoseries Instrument to measure
the particle size and its distribution. The particles were dispersed in DI Water and measured using 4 mW He–Ne laser
633 nm at 25 C. Simultaneous differential thermal and
thermogravimetric analyses (DTA/TGA) were carried out
in flowing air by a Netzsch thermal analyzer (STA449C).
Ramp rate used was 5 C·min−1 from ambient temperature
to 1000 C. X-ray analysis was performed using a Shimadzu XRD-6000 diffractometer, with Cu K-radiation
( = 15418 Å). Using Bragg-Brentano focusing geometry, the diffraction patterns were recorded with a scanning
speed of 2 = 2 · min−1 for a range of 2 = 10 − 100 .
Zeta potential measurements were done with a Zeta Plus
system. The particles were dispersed in 0.001 mole dm−3
KCl and the pH was adjusted with HNO3 or NaOH. The
sols were treated in an ultrasonic bath before the measurements were taken. The powder chemical composition was
monitored using Fourier Transform Infrared Spectroscopy
(Perkin-Elmer Spectrum GX). Samples were prepared by
mixing the powder with potassium bromide, which had
been baked in an oven at 75 C for 24 hours.
J. Nanosci. Nanotechnol. 7, 1–9, 2007
Happy et al.
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
3. RESULTS AND DISCUSSIONS
Figure 1 shows simultaneous temperature and pH rise of
the dysprosium nano-particles precipitated at different dysprosium ion concentrations. The onset of precipitation is
indicated on the graph, and this was observed by the slight
turbidity in the solution. All samples exhibited a similar
rise in pH, even though the dysprosium ion concentration
varied significantly. This was attributed to the decomposition of urea followed by hydrolysis of its products. The
urea decomposition has been discussed by various investigators and can be described as follows:11 13
k1
→ HNCO + NH3
H2 N − CO − NH2 −
(1)
HNCO + NH3 + 4H2 O → 2NH4 OH + H2 CO3
(2)
NH4 OH ↔
−
NH+
4 + OH
(3)
H2 CO3 ↔ H+ + HCO−
3
(4)
HCO−
3
(5)
↔H
+
+ CO2−
3
(a) 10 mins
(b) 45 mins
(c) 70 mins
(d) 90 mins
RESEARCH ARTICLE
It is proposed that the decomposition of urea in aqueous
medium yield ammonium (NH+
4 ) and cyanate ions
(CNO− ), where the rate-determining step (k1 ) is represented by Eq. (1) whilst Eqs. (2), (3), (4), and (5) describe
the conversion of cyanate ions to ammonium ions (NH+
4)
).
Since
the
formation
of
solid
preand carbonate (CO2−
3
cipitates in urea precipitation is governed by controlled
generation of carbonate ions, it is important to observe the
pH of the solution and relate it to precipitation process
during aging.
In this work, the onset of the precipitation process was
observed when the pH of the solution increased rapidly
to pH 4.7–5.1. This onset took longer as dysprosium ion
concentration increased because the hydroxide ions (OH− )
as a product of urea decomposition was used to neutralize
the excess hydrogen ions (H+ ) from dysprosium chloride
solution. As the aging process continued, the rate of pH
rise was decreased, indicating that the converted carbonate ions were used to precipitate the remaining dysprosium
ions from the solution. When the dysprosium ions were
depleted, the pH rose rapidly and leveled off at about 6.5.
No further change in the pH was observed when the solution at [Dy3+ = 0020 M was aged for 240 minutes.
Representative transmission electron micrographs of
dysprosium particles precipitated at different aging time
are shown in Figure 2. In general, the particles were spherical, uniform in size, and polycrystalline. Each particle
consisted of smaller crystals about 30–40 nm in size, as
shown in subset micrograph in Figure 2(d). Analyzing
the micrographs, the change in particle size with aging
time can be divided into three stages. In the initial state
when increasing aging time from 10 to 45 minutes, the
particles grew from 50 nm to 75 nm on average. Subsequently, no growth of the particles was observed around
75 nm when the aging time increased from 45 to 90 minutes. Holding the solution for a further 4 hours while
retaining its spherical morphology, the particles grew to
about 150 nm. Dynamic light scattering (DLS) measurement, shown in Figure 3, was done to confirm the particle
size and size distribution with TEM results. The measurement showed small increase in the average particle size
compared to TEM results, for example, particles aged at
(e) 240 mins
Fig. 1. pH and temperature dependence on time for different dysprosium ion concentration. Note X indicates the first visible sign of
precipitation.
J. Nanosci. Nanotechnol. 7, 1–9, 2007
Fig. 2. Effects of aging time on Dy2 O3 particles precipitated at constant
[Dy3+ = 002 M, [Urea] = 10 M and aging temperature of 90 C.
3
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
RESEARCH ARTICLE
Fig. 3. Dynamic Light Scattering measurement of dysprosium particles
precipitated at constant [Dy3+ = 0020 M, [Urea] = 10 M and aged at
90 C at different aging time.
10 minutes have average particle size of 59.1 nm and
width of 6.13 nm. Nevertheless, these results were not
unexpected since dynamic light scattering measures the
hydrodynamic radius of the particles in the solvent. As
observed, the growth of the particles with aging time can
also divided into three distinct regions, which corresponds
to the electron micrographs.
In addition, the amount of the precipitates (yield)
increased as aging continued from 10 to 90 minutes, which
means that new precipitates were being formed from the
solution as aging continued. This phenomenon is in agreement with pH curve is Figure 1, where the pH of the
solution increased slowly after the onset of turbidity, indicating new dysprosium precipitates were being formed by
consuming the converted carbonate ions from urea decomposition, which in turn increased the yield of the precipitation process.
If the short nucleation time model by LaMer was used
to explain the behavior, the particles would have grown
steadily when the solution was aged from 10 to 90 minutes, since the number of particles was constant and the
precipitate’s yield was increasing, which means more mass
were attached to the existing particles. Yet, the growth
of the particles seems discontinued and more precipitates
were being formed, which indicates nucleation of primary
particles was continuously occurring. Therefore, based on
this behavior, we can deduce that the aggregation model
is more appealing as a growth mechanism of particles in
homogeneous precipitation by urea.
The discontinued growth of the particles upon aging
from 10 minutes to 90 minutes could be related to the
stability of these particles in the solution. As mentioned
earlier, the stability could be contributed by steric and/or
electrostatic forces. Since no organic molecules were
present in our system, hence only electrostatic contribution
was effected in the stability of the particles. The thickness of this electrostatic barrier can be affected by the
4
Happy et al.
Fig. 4. Zeta potential of precipitated powder as a function of pH. Note
the particles were dissolved at lower pH.
type and concentration of ions present in the solution, such
as dysprosium, carbonate, and chlorine ions. In summary,
the dysprosium particles are formed from aggregation of
smaller (primary) particles and once a stable size of the
particles has been reached, no more primary particles were
attached onto it.
In contrast, on prolonged aging for 4 hours, the particles
grew from average size of 75 nm to 150 nm (TEM analysis) and 83.5 nm to 163 nm (DLS measurement), which
indicates a rather slow growth rate. This size increment
was also accompanied by broadening of size distribution
from 9.91 nm to 30.8 nm. This growth could be attributed
to the rearrangement of primary particles, which forms
secondary particles, into a more compact structure. Nevertheless, with regards to the growth mechanism of urea precipitation, these results are encouraging as the precipitation
process could be completed without any significant growth
during aging, achieving mono-dispersed nano-particles.
Since the stability of final (secondary) particles is
related to the electrostatic barrier surrounding the particles, a study on zeta potential of precipitated particles was
performed. Figure 4 shows the zeta potential curve of the
precipitate particles as a function of pH. The isoelectric
points (IEP) for the precipitated powders are shown at pH
value of 7.7. It is known that both zeta potential and electrical double layer thickness are affected by the concentration of ions present in the solution. Thus, changing the
synthesis parameters such as dysprosium ion concentration
will affect the electrical double layer, which will in turn
affect the particle growth and eventually particle size and
size distribution.
The effect of dysprosium ion concentration on the particle size and its distribution was examined using transmission electron microscopy, shown in Figure 5. Analyzing
the electron micrographs, we observed that as the dysprosium ion concentration increased from 0.005 M to 0.02 M,
the average particle size decreased from 205 nm to 75 nm
whilst retaining its spherical morphology. This result was
confirmed using DLS technique, shown in Figure 6, which
indicated a decrease in the average particle size from
J. Nanosci. Nanotechnol. 7, 1–9, 2007
Happy et al.
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
(a) [Dy 3+] = 0.005 M
(b) [Dy 3+] = 0.010 M
(c) [Dy 3+] = 0.020 M
(d) [Dy 3+] = 0.050 M
(e) [Dy 3+] = 0.100 M
227 nm to 83.5 nm and also accompanied by narrowing
size distribution from 32 nm to 9.91 nm. On the contrary, increasing the concentration further from 0.02 M
to 0.100 M, the particles were more agglomerated, as
Fig. 6. Dynamic Light Scattering measurement of dysprosium particles
precipitated at constant [Urea] = 10 M, aged at 90 C for 90 minutes at
different dysprosium ion concentration.
J. Nanosci. Nanotechnol. 7, 1–9, 2007
5
RESEARCH ARTICLE
Fig. 5. Effects of dysprosium ion concentration on Dy2 O3 particles precipitated at [Urea] = 10 M, aged at 90 C for 90 minutes.
seen in Figures 5(d) and (e). This agglomeration trend is
also accompanied by an increase average particle size and
widening of particle size distribution. Meanwhile, DLS
measurement showed similar trend, increase in the average particle size (413 nm for 0.050 M and 1180 nm for
0.100 M) and broadening size distribution (103 nm for
0.050 M and 224 nm for 0.100 M). This trend where the
particle size initially decreases and then increases as dysprosium ion concentration in increased can be related to
how the dysprosium ion concentration affects the electrostatic barrier surrounding the particles.
The size selection mechanism by aggregation model
does not only depend on the stability of the formed particles, but also on how the primary particles interact with
one another. Initially, the above trend shows a strong correlation between surface potential of the particles and dysprosium ion concentration. During the aging process, the
pH of the solution was at 4.2–5.3, as shown in Figure 1,
which suggests that the particles are positively charged
in this pH range. As the dysprosium ions are positively
charged, it seems that the increase in dysprosium ion concentration from 0.005 M to 0.020 M led to an increase in
the surface potential, which in turn led to more stabilized
particles and smaller particle size. In other words, dysprosium ion could be said as the potential determining ions
for these precipitates.
Although the stability of the particles increased as dysprosium ion concentration increased, the growth of the particles by aggregation model is also determined by primary
particle interactions. As dysprosium ion concentration
increased further from 0.020 M to 0.100 M, the effect of
primary particle collisions was more dominant than the
increase in the surface charge, such that it led to further
growth of the particles. These particle collisions could also
be observed from heavy agglomeration of the particles
such one has difficulties to distinguish one particle from
another and increase in the particle size. Furthermore, the
particle size distribution was also widened, which was
attributed to incomplete precipitation process. This incomplete precipitation can also be observed from pH curve
in Figure 1, where for dysprosium ion concentration of
0.050 M and 0.100 M, the pH curve level off at pH 4.9
and 4.7 respectively, instead of level off at pH 6.5. This
means that throughout the precipitation process, the carbonate ions were used to precipitate dysprosium ions, and
translated into incomplete formation of secondary particles
at the end of the process. Incomplete precipitation process at higher metal cation concentration was also reported
by Sordelet and Akinc,12 which observed lower yield at
higher metal cation concentration.
To study the effect of chlorine ion (Cl− ) on the particles
growth, a precipitation with condition of sample 4 in
Table I was done, with the initial pH of the solution
was adjusted to pH 2 with addition of concentrated
HCl. Figure 7 exhibits the change in the morphology of
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
(a) Without HCl addition
RESEARCH ARTICLE
(b) With HCl addition
Fig. 7. Effects of HCl addition on Dy2 O3 particles precipitated at
[Dy3+ = 0020 M, [Urea] = 10 M, aged at 90 C for 90 minutes.
precipitated particles with addition of HCl. It was noticed
that the particles with addition of HCl shows larger size
and wider particle size distribution and accompanied with
heavy agglomeration between the particles compared to
the particles without addition HCl. This agglomeration
morphology is similar to the one observed at high dysprosium ion concentrations (0.050 M and 0.100 M). Chlorine
ions can be said as indifferent electrolyte ions, which affect
the extension of the double layer out into the solution. During the precipitation, chlorine ions acted as counter ions
and reduced the thickness of the electrical double layer
around the particles since the precipitated particles was
positively charged, which in turn promoted aggregation
among the particles. A similar behavior was also observed
by Hu,21 who found that Cl− induced aggregation among
zirconia particles and responsible for particle growth. In
study on the effect of hydrochloric acid on electrical double thickness of Al2 O3 , Modi24 found that adding more
acid at first caused an increase in zeta potential, but at
higher acid concentrations the acid itself began to contribute to the ionic strength; such affected the extension of
6
Happy et al.
the double layer. In conclusion, the amount of Cl− in our
system was very high; such it reduced the electrical double layer and promoted aggregation of particles. This also
promoted growth of the particles and the effect will be
aggravated if coupled with high probability of collisions
among the particles; like in the case of high dysprosium
ion concentration where increased in dysprosium ion concentration was also accompanied by increased in chlorine
ion concentration.
Figure 8 shows the transmission electron micrographs
of dysprosium particles precipitated at different urea concentrations with dysprosium ion concentration kept at
0.020 M and aged at 90 C for 90 minutes, respectively.
As observed, there was no significant variation in terms
of morphology, particle size and size distribution, except
at urea concentration of 4.0 M where there was a slight
decrease in average particle size from 75 nm to 52 nm.
This shows that particle size and size distribution are less
sensitive to the variation in the urea concentration. Ammonium and carbonate ions, as a by-product of urea decomposition seems do not play a critical role in determining
zeta potential of the particles or ionic strength of the solution, which in turn will have little effect on the electrical
double thickness surrounding the particles. Nevertheless,
in the case of excess urea (4.0 M), it was also observed
through pH measurements that the precipitation process
was completed at a much earlier stage as indicated by the
leveling off of the pH curve at about pH 6.5. Consequently,
this could affect the kinetics of the particle growth, stabilizing particles at a shorter time and leading to a slight
reduction in particle size.
Simultaneous thermogravimetric and differential thermal analysis (TGA/DTA) traces of the dried dysprosium
precipitates are shown in Figure 9. There was continuous
weight loss during heating from room temperature to about
750 C, with no stepwise reaction being indicated. No
further weight loss was observed upon further heating
to 1000 C. The DTA trace indicates a broad endothermic peak at around 177 C with a shoulder at 100 C.
It has been reported previously18 that these endothermic
reactions correspond to the dehydration of crystallization
water in the dysprosium hydroxide carbonate precipitates,
Dy(OH)CO3 · xH2 O. During further heating, the weight
loss continues and was accompanied by the appearance
for an exothermic peak at about 600 C. This weight
loss corresponds to the conversion of dysprosium hydroxide carbonate to dysprosium oxide carbonate (Dy2 O2 CO3 ),
where the exothermic peak represents the crystallization
of the oxide carbonate form. The observed experimental
weight loss of 14.2% is close to the calculated value of
11.9%. The weight loss continued with an endothermic
peak around 632 C, which represent the conversion of
oxide carbonate to the oxide form, Dy2 O3 .
Figure 10 shows the XRD patterns of the samples
calcined at different temperatures. The dried precipitates
J. Nanosci. Nanotechnol. 7, 1–9, 2007
Happy et al.
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
(a) [Urea] = 1.0 M
Fig. 9. TGA/DTA curves of the dried precipitates.
(b) [Urea] = 2.0 M
Fig. 8. Effects of urea concentration on Dy2 O3 particles precipitated at
constant [Dy3+ = 0020 M, aged at 90 for 90 minutes.
showed an amorphous structure, as indicated by the broad
humps in the patterns in Figure 10(a), which was identified as a form of dysprosium hydroxide carbonate. The
position of the amorphous hump around 30 and 45 is the
only indication of dysprosium hydroxide compound which
has strong reflections around corresponding angles, based
J. Nanosci. Nanotechnol. 7, 1–9, 2007
Fig. 10. XRD patterns of (a) dried precipitates (b) calcined at 250 C
for 2 hr (c) calcined at 500 C for 2 hr and (d) calcined at 700 C for
2 hr (e) Dy2O3 cubic ICSD #962080.
7
RESEARCH ARTICLE
(c) [Urea] = 4.0 M
on ICSD #89605. Heating these precipitates further to
250 C did not change the structure of the samples, which
confirms that the weight loss around 100 C corresponds
to the loss of hydrated water. Annealing the samples at
500 C resulted in the formation of two crystalline phases,
namely dysprosium oxide and dysprosium oxide carbonate (Fig. 10(c)), which is in agreement with Hussein.25
The dysprosium oxide in the pattern corresponds to cubic
structure (ICSD #962080). Pure dysprosium oxide with a
crystallite size of 30 nm can be obtained by annealing the
precipitates at 700 C, as no other reflection was observed
in the spectrum.
A High Resolution TEM image of one of the crystallites in calcined Dy2 O3 particles is shown in Figure 11(a).
The image was reconstructed with the aid of crystallographic image processing program, CRISP. Fast Fourier
Transform (FFT) was performed to obtain simulated electron diffraction (ED) pattern, which was used to do lattice
refinement. It was found that the symmetry p2m has the
highest calculated figure of merit for amplitude (RA%)
and phases (
Res). This symmetry was the symmetry
projected along [110] as reported in International Tables
for Crystallography.26 Thus, the experimental HRTEM
image was Fourier processed and imposed with the p2m
symmetry to yield a highly symmetry image as shown in
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
Happy et al.
(a)
2 nm
(b)
RESEARCH ARTICLE
(c)
Fig. 11.
High Resolution TEM image taken along [110] axis.
Figure 11(a) and (b). To illustrate the observed structural
projection, simulated atomic structure of the cubic dysprosium oxide was performed using crystallographic drawing
program (ATOMS). The simulated atomic structure was
superimposed on the calculated HRTEM image so that the
contrast in the HRTEM image could be represented by the
overlaid atoms and polyhedrals as shown in Figure 11(c).
The IR spectra of the dried precipitates and calcined
samples at different temperatures are shown in Figure 12.
The dried precipitates showed a broad band at 3400 cm−1 ,
which can be attributed to the stretching mode of hydroxyl
groups (–OH) of bound water. The IR spectrum also provided evidence for the carbonate ions in the precipitates.
The absorption band at 1520 cm−1 is attributed to the 3
mode of CO2−
ion. The splitting of the band could be
3
due to the carbonate ions located at a crystallographically
8
Fig. 12. Infrared spectra of the samples (a) dried precipitates (b) calcined at 250 C for 2 hr (c) calcined at 500 C for 2 hr and (d) calcined
at 700 C for 2 hr.
non-equivalent site. The other bands at 1078, 842, and
725 cm−1 are assigned to the 1 , 2 , and 4 modes of
the carbonate ion, respectively.27 Calcination of the precipitates to 250 C did not result in any change the IR
spectra, with the exception of a reduction in intensity of
the hydroxyl group, which again supports the idea of the
evaporation of adsorbed water at 100 C. Upon further
heating to 500 C, the absorption band of carbonate ions
were reduced and the absorption band of C-type Dy2 O3
started to appear at 560 cm−1 and 461 cm−1 .28 This is in
agreement with the corresponding XRD spectrum where
reflections for dysprosium oxide and weak reflections of
dysprosium oxide carbonate were observed together. Further calcination of the precipitates to 700 C did not eliminate the presence of carbonate ions completely. There was
a slight absorption of 3 mode of CO2−
3 ion, though there
were no additional peaks observed in the corresponding
XRD spectrum except cubic dysprosium oxide. Therefore,
further calcination at higher temperature or longer duration
is required to eliminate traces of the carbonate ions.
4. CONCLUSIONS
In this paper, homogeneous precipitation technique using
urea is reported to prepare Dy2 O3 nano-particles at low
processing temperature and high yield. The as-precipitated
particles had an amorphous microstructure, which was
converted into dysprosium oxide by heating in air at
700 C for 2 hours. The aggregation mechanism has been
offered to explain the growth of the nano-particles by urea
precipitation, where the growth of the particle is influenced
by stability of secondary particles and collisions among the
primary particles. The stability of the secondary particles
can be related to the electrical double thickness surrounding the particles, where the thickness is affected by the
concentration of ions present in the solution. Dysprosium
ion concentration was found to have a significant effect on
the size and size distribution of the nano-particles, while
urea concentration and aging time had less significance. In
addition, chlorine ion was found to promote aggregation
J. Nanosci. Nanotechnol. 7, 1–9, 2007
Happy et al.
Homogeneous Precipitation of Dy2 O3 Nanoparticles—Effects of Synthesis Parameters
among the particles since it reduces the electrical double
thickness.
Acknowledgments: We would like to thank the following people for their contributions to this work
• Dr. René Hübner from University at Albany, New York
for the discussions on the XRD work.
• Mr. Y. Setiawan for help in taking High Resolution TEM
image.
• The Agency for Science, Technology & Research
(A∗ Star), Singapore for providing SERC grant 022 105
0025 to fund this research project.
References and Notes
Received: 27 July 2006. Revised/Accepted: 1 September 2006.
J. Nanosci. Nanotechnol. 7, 1–9, 2007
9
RESEARCH ARTICLE
1. K. E. Gonsalves, S. P. Rangarajan, and J. Wang, Hand Book of
Nanostructured Materials and Nanotechnology, edited by H. S.
Nalwa, Academia Press, San Diego (1999), Vol. 1, p. 1.
2. R. P. Rao, J. Electrochem. Soc. 150, H165 (2003).
3. S. Collusi, C. D. Leitenburg, G. Dolcetti, and A. Trovarelli, J. Alloy
Compd. 374, 387 (2004).
4. B. Antic, M. Mitric, and D. Rodic, J. Phys. Condens. Matter. 9, 365
(1997).
5. H. Kishi, Y. Mizuno, and H. Chazono, Jap. J. Appl. Phys. 42, 1
(2003).
6. G. Wakefield, H. A. Keron, P. J. Dobson, and J. L. Hutchison,
J. Colloid Interf. Sci. 215, 179 (1999).
7. M. H. Lee, S. G. Oh, and S. C. Yi, J. Colloid Interf. Sci. 226, 65
(2000).
8. M. Andersson, E. Bakchinova, K. Jansson, and M. Nygren, J. Mater.
Chem. 9, 265 (1999).
9. A. I. Y. Tok, L. H. Luo, and F. Y. C. Boey, Mater. Sci. Eng. A 383,
229 (2004).
10. J. Z. Gao, Y. C. Zhao, W. Yang, J. N. Tian, F. Guan, Y. J. Ma,
J. G. Hou, J. W. Kang, and Y. C. Wang, Mater. Chem. Phys. 77, 65
(2002).
11. E. Matijević and W. P. Hsu, J. Colloid Interf. Sci. 118, 506
(1987).
12. D. Sordelet and M. Akinc, J. Colloid Interf. Sci. 122, 47 (1988).
13. W. H. R. Shaw and J. J. Bordeaux, J. Am. Chem Soc. 77, 4729
(1955).
14. A. C. Taş, P. J. Majewski, and F. Aldinger, J. Am. Ceram. Soc. 85,
1421 (2002).
15. M. Akinc and D. Sordelet, Adv. Ceram. Mater. 2, 232 (1987).
16. B. Aiken, W. P. Hsu, and E. Matijeviæ, J. Am. Ceram. Soc. 71, 845
(1988).
17. V. K. LaMer and R. H. Dinegar, J. Am. Chem. Soc. 72, 4847 (1950).
18. P. Jeevanandam, Yu Koltypin, O. Palchik, and A. Gedanken,
J. Mater. Chem. 11, 869 (2001).
19. M. L. Panchula and M. Akinc, J. Europ. Ceram. Soc. 16, 833
(1996).
20. Y. D. Jiang, Hand Book of Nanophase and Nanostructured Materials,
edited by Z. L. Wang, Y. Liu, and Z. Zhang, Kluwer Academic,
NewYork (2003), Vol. 1, p. 55.
21. M. Z. C. Hu, M. T. Harris, and C. H. Byers, J. Colloid Interf. Sci.
198, 87 (1998).
22. G. H. Bogush and C. F. Zukoski, J. Colloid Interf. Sci. 142, 19
(1991).
23. V. Privman, D. V. Goia, J. S. Park, and E. Matijević, J. Colloid
Interf. Sci. 213, 36 (1999).
24. H. J. Modi and D. W. Fuerstenau, J. Phys. Chem. 61, 640 (1957).
25. G. A. M. Hussein, W. J. Kroenke, B. Goda, and K. Miyaji, J. Anal.
Appl. Pyrol. 39, 35 (1997).
26. H. Theo (Ed.), International Tables for Crystallography, Springer,
Netherlands (2005), Vol. A, p. 633.
27. R. P. Turcotte, J. O. Sawyer, and L. Eyring, Inorg. Chem. 8, 238
(1969).
28. N. T. McDevitt and W. L. Baun, Spectrochim. Acta 20, 799 (1964).