The effects of synthesis pH and hydrothermal treatment on the formation of zinc aluminum hydrotalcites

Article abstract of DOI:10.1016/j.jssc.2004.07.010

Zn/Al hydrotalcites were synthesized by coprecipitation at increasing pH from 6.0 to 14.0, followed by hydrothermal treatment at 150°C for 7 days. The materials were characterized by X-ray diffraction (XRD), STEM, inductively coupled plasma-atomic emission spectroscopy (ICP-AES), thermal analysis, infrared spectroscopy and Raman spectroscopy. The XRD analysis for the samples prepared between pH 9.0 and 12.0 showed a pattern typical of hydrotalcite, with a c-axis distance of ～22.6 A. STEM showed that the pH of preparation affected the stability of the hydrotalcite and that instability, observed at pH 9.0, favored the formation of mixed phases when treated hydrothermally. It was also shown that treatment of a stable starting material increased the crystallinity and resulted in the formation of hexagonal plate-shaped particles. ICP-AES and thermal analysis showed that the Zn/Al ratio and thermal stability increased with pH. Thermal analysis showed three major weight losses corresponding to the loss of interparticle water, interlayer water and dehydroxylation of the hydroxide layers and decarbonization of the interlayer region.

Full text of DOI:10.1016/j.jssc.2004.07.010

ARTICLE IN PRESS
Journal of Solid State Chemistry 177 (2004) 4047–4057
www.elsevier.com/locate/jssc
The effects of synthesis pH and hydrothermal treatment on the
formation of zinc aluminum hydrotalcites
Ã
J. Theo Kloprogge , Leisel Hickey, Ray L. Frost
Inorganic Materials Research Group, School of Physical and Chemical Sciences, Queensland University of Technology, 2 George Street,
GPO Box 2434, Brisbane QLD 4001, Australia
Received 30 November 2003; received in revised form 20 May 2004; accepted 13 July 2004
Available online 7 October 2004
Abstract
Zn/Al hydrotalcites were synthesized by coprecipitation at increasing pH from 6.0 to 14.0, followed by hydrothermal treatment at
150 1C for 7 days. The materials were characterized by X-ray diffraction (XRD), STEM, inductively coupled plasma–atomic
emission spectroscopy (ICP–AES), thermal analysis, infrared spectroscopy and Raman spectroscopy. The XRD analysis for the
samples prepared between pH 9.0 and 12.0 showed a pattern typical of hydrotalcite, with a c-axis distance of ꢀ22.6 A. STEM
showed that the pH of preparation affected the stability of the hydrotalcite and that instability, observed at pH 9.0, favored the
formation of mixed phases when treated hydrothermally. It was also shown that treatment of a stable starting material increased the
crystallinity and resulted in the formation of hexagonal plate-shaped particles. ICP–AES and thermal analysis showed that the Zn/
Al ratio and thermal stability increased with pH. Thermal analysis showed three major weight losses corresponding to the loss of
interparticle water, interlayer water and dehydroxylation of the hydroxide layers and decarbonization of the interlayer region.
r 2004 Elsevier Inc. All rights reserved.
1. Introduction
A wide range of compositions are possible for
synthetic hydrotalcites, based on the general formula
b=n Á mH2O, where M2+ and
bþ
Anionic clays, such as hydrotalcites, are not as well
known and are rarer in nature compared to cationic
clays such as smectites. The first hydrotalcite was
discovered in Sweden around 1842 [1]. Anionic clays
have a structure electrically opposite to that exhibited by
cationic clays. Their structure can be derived from the
brucite structure, Mg(OH)2, where each Mg2+ ion is
octahedrally surrounded by six OHꢁ ions. The hydro-
talcite-type (HT) structure is obtained when some of the
Mg2+ ions, or other divalent cations, are replaced by
trivalent cations, with a similar radius [2]. The higher
charge of the trivalent cations imposes an overall
positive charge on the brucite-type layer, which is
compensated by the incorporation of generally hydrated
interlayer anions.
½M21þꢁxM3xþðOHÞ2Š ½Anꢁ
Š
M3+ are the divalent and trivalent cations in the
octahedral positions within the hydroxide layers, with
x normally ranging between 0.17 and 0.33, Anꢁ is an
interlayer anion with a negative charge n, b is the charge
of the layer and m is the number of water molecules.
Many anions or anionic complexes, both organic and
inorganic, can and have been incorporated into the
hydrotalcite structure. The wide range of materials of
this type that can be synthesized gives rise to a range of
applications including catalysts and catalyst precursors,
ion exchangers and adsorbers, anion scavengers and
stabilizers, etc. [3,4].
Several factors are important in the precipitation of
HT compounds, such as the nature of the cations, their
ratio, the nature of the anions, pH, temperature, aging,
and the precipitation method. Although generally
anionic clay particles are larger than their counterpart,
the cationic clays, there is still interest in the ability to
Ã
Corresponding author. Fax: +61-7-3864-1804.
E-mail address: t.kloprogge@qut.edu.au (J.T. Kloprogge).
0022-4596/$ - see front matter r 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2004.07.010

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4048
increase particle size, to improve their chemical and
physical properties.
pH from 6.0 to 6.5, 9.0, 9.5, 10.0, 10.5, 11.0, 11.5, and
414, and labelled 3–10. In all experiments, solutions
were added using a peristaltic pump at a rate of 40 mL/
min, under vigorous stirring. The pH was maintained
constant by adding HNO3 when necessary. When the
addition was complete, the mixture was stirred for a
further 1.5 h, during which time the pH was still
maintained constant. Two samples were selected based
on their purity, stability and crystallinity, and were
treated hydrothermally at autogenous water vapor
pressure at 150 1C, in Teflon-lined stainless steel Parr
Bombs for 7 days (samples 11 and 12). All samples were
filtered, thoroughly washed several times, in approxi-
mately 500 mL of deionized water each time to remove
excess soluble ions, and finally the samples were oven
dried at 60 1C for 1 h.
A recent study has shown that the carbonate anion
concentration in the interlayer region of hydrotalcite
was affected by both temperature and aging [5]. It was
shown that the best crystalline material was obtained at
25 1C, with at least 5 h of aging. Generally, it is accepted
that a basic pH is required for the preparation of
hydrotalcites. However, the optimal pH depends on
which cations are used and where their hydroxide
precipitation curves cross. The crystallinity of the
precipitated hydrotalcite can be further increased by
hydrothermal treatment [6,7]. In 1980, Miyata showed
that using a higher temperature, 180 1C, for hydro-
thermal treatment increased the crystallite size [8].
Several studies have reported on the influence of the
synthesis parameters and show that pH is the major
factor in the formation of Zn/Al hydrotalcite [9–13]. Zn/
Al hydrotalcites have been used for anion exchange
reactions, pillaring and absorption [14–24] and a
number of catalytic applications, either with the
hydrotalcite structure or after calcination as the mixed
metal oxide [25–32].
This study forms part of an ongoing investigation on
the effects of reaction conditions during synthesis and
hydrothermal treatment of the hydrotalcite minerals
with various cations and anions to improve their
physical and chemical properties. The aim of this study
is to obtain insight into the nature of the Zn/Al
hydrotalcite-type materials when the experimental con-
ditions are varied. Samples were prepared over a wide
pH range, and the reagent addition was optimized.
Samples were selected, based on their purity, stability
and crystallinity, and treated hydrothermally to inves-
tigate (1) the effects of aging under the influence of
temperature, pressure and time, and (2) the effects of pH
of preparation on the crystallinity of the hydrothermally
treated hydrotalcite.
2.2. Characterization
2.2.1. Powder X-ray diffraction (XRD)
Powder XRD (XRD) diagrams were recorded on a
Phillips wide-angle PW 1050/25 vertical goniometer
equipped with a graphic diffracted beam monochroma-
tor. The d-values and intensity measurements were
improved by application of an in-house developed
computer-aided divergence slit system enabling constant
sampling area irradiation (20 mm long) at any angle of
incidence. The radiation applied was CoKa from a long
fine-focus Co tube operating at 40 kV and 40 mA. The
samples were measured in the step-scan mode from 5 to
7512y, with steps of 0.0212y and a counting time of 2 s.
Jade software package (Materials Data, Inc.) was used
to analyze the data; identification of the existing
crystalline phases was based on comparison with pre-
existing JCPDS diffraction data files. Pattern manipula-
tion and interpretation were performed using TracesV4
(Diffraction Technology Pty. Ltd.) and Microsoft
EXCEL.
2.2.2. Transmission electron microscopy (TEM)
2. Experimental
Electron micrographs of the samples were taken by
the TEM method at 200 kV with a Philips CM200
Analytical Scanning TEM instrument, equipped with
LaB6 filament, cryo vacuum options, and link thin-
window Si EDS detector. Microanalysis of selected
samples was carried out in the STEM mode using an
energy-dispersive X-ray spectrometer with Link ISIS
system. Samples were prepared by suspension in alcohol
(70%) and a small amount was placed on a carbon-
coated copper grid, prior to use in the TEM.
2.1. Preparation
Zn/Al hydrotalcites were synthesized by the copreci-
pitation method, based on a theoretical composition
of Zn6Al2(OH)16(CO3) Á 4H2O [33]. Two solutions
were prepared, solution 1 contained 2 M NaOH and
0.125 M Na2CO3, solution
2
contained 0.75 M
Zn2+ [Zn(NO3)2 Á 6H2O] along with 0.25 M Al3+
[Al(NO3)3 Á 9H2O]. For sample 1, solution 2 was added
to solution 1; for sample 2, solution 1 was added to
solution 2. The pH was maintained at ꢀ12, for both
methods. The best method was established on the basis
of crystallinity and purity. Using the optimum method,
the next series of samples were prepared by varying the
2.2.3. Inductively coupled plasma–atomic emission
spectroscopy (ICP–AES)
ICP–AES was used for elemental analysis of the Zn
and Al ions. A calibration series of standards for
determining the relative amounts of Zn and Al were

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4049
prepared at 10, 50, 100, and 200 ppm. Small amounts of
(003)
12, hydrothermally
treated
the samples were accurately weighed and digested in
3 mL of 6 M HCl along with 1 mL of 6 M HNO3. The
relative amounts of each atom were recorded on Varian
Liberty 2000 ICP–AES at wavelengths of 394.40 and
213.86 nm for Al and Zn, respectively.
11, hydrothermally
treated
(006)
(012)
(015)
(018)
10, pH > 14
1, pH 12
2, pH 12
9, pH 11.5
8, pH 11
7, pH 10.5
6, pH 10
2.2.4. Thermal analysis
Differential thermal analysis (DTA) and thermal
gravimetric analysis (TG) of the samples were recorded
on a Setaram 2000 (1992) unit. The samples were
measured under a stream of argon (20 mL/h) at a
heating rate of 21/min, from 30 1C to 1000 1C at a
heating rate of 21/min. Data manipulation and inter-
pretation was carried out using Microsoft EXCEL.
5, pH 9.5
4, pH 9
3, pH 6-6.5
82
12
22
32
42
52
62
72
92
˚2θ
Fig. 1. XRD patterns of all synthetically prepared Zn–Al hydrotalcite
samples (1–12) in comparison with each other.
2.2.5. Fourier transform infrared (FTIR) and Raman
spectroscopy
to the variation in synthesis conditions, where based on
the Zn solubility curves precipitation of ZnO starts at
pH ꢀ9 [35]. The addition of the hydroxide-containing
solution to the cations-containing solution yielded a
hydrotalcite-type material without the ZnO impurity.
Proceeding via this pathway is therefore preferred as it
progresses under acidic conditions that do not favor the
formation of ZnO [35]. Under strong alkaline condi-
tions, ZnO species are not soluble and compete with the
formation of Zn/Al hydrotalcite.
FTIR spectra were recorded using the KBr tablet
technique (2 wt% sample) on a Perkin-Elmer FT-IR
1000 spectrometer; 64 scans were taken to improve the
signal-to-noise ratio in the range of 400–4000 cmꢁ1, the
normal resolution was 4 cmꢁ1. The Raman spectra were
obtained using a Renishaw 1000 Raman microscope
system, which includes a monochromator, a filter system
and a charged-coupled device (CCD) as the detector.
Raman spectra were excited on a Spectra-Physics model
127 He/Ne laser (633 nm), scanned for 15 s for 1000
scans and recorded in the range of 400–1800 and
2800–4000 cmꢁ1. Data manipulation, such as baseline
adjustment, smoothing and normalizing as well as
interpretation, was performed using the Spectracalc
software package GRAMS (Galactic Industries Cor-
poration, NH, USA) and Microsoft EXCEL. Band
component analysis was undertaken using the Jandel
‘‘Peakfit’’ software package, which enabled the type of
fitting function to be selected and allowed specific
parameters to be fixed or varied accordingly.
3.2. The effect of pH of synthesis
The pH of preparation is important in the formation
of all hydrotalcite-type materials and the optimum pH
depends on the cations used (Fig. 1). Sample 3 (pH
6.0–6.5) displays very weak, broad peaks at higher 2y
values (14.91) compared to sample 2 (pH 12, 13.61). The
XRD patterns of sample 3 (pH 6.0–6.5) show that no
hydrotalcite structure is formed under acidic synthesis
conditions. The broadness of the peaks indicates that
the sample is poorly crystalline, where the reflections
occur at similar 2y values as for brucite, indicating that a
low crystalline hydroxide similar to brucite is formed
instead of hydrotalcite. Samples prepared within the pH
range of 9.0–12.0 (samples 4–9 and 2) show patterns
similar to that of natural hydrotalcite [7], with the (003)
reflection corresponding to a c-axis of ꢀ22.5 A. The
difference between these samples is in the intensity of
the (00l) reflections. As pH increases, from 9.0 to 12.0,
the intensity of the reflections increases and the line
width decreases, corresponding to an increase in crystal-
linity. Samples 2 and 9 at pH 12 and 11.5, respectively,
have the most intense and sharpest peaks, and thus are
the most crystalline samples. Sample 10 (pH414) shows
a structure different from the previous samples; it is
composed of one major crystalline phase being ZnO and
a minor phase, sodium nitrate. Based on solubility
curves and predominance diagrams for Zn2+, at this
strong alkaline pH, zinc-hydroxy complexes, such as
3. Results and discussion
3.1. Effect of reagent addition
The XRD patterns of all samples are shown in Fig. 1.
The crystalline components in samples 1 and 2 (pH 12)
show some significant differences. Both show a sharp
symmetric (003) reflection at low 2y angles correspond-
ing to a c-axis of ꢀ22.8 A. Both spectra also show broad
asymmetric peaks at higher angles, which are character-
istic of clay minerals possessing a layered structure and
are similar to the pattern of natural hydrotalcite [7,34].
Sample 1 (pH 12), however, shows additional peaks that
are associated with the second major phase zincite
(ZnO) and sodium nitrate (NaNO3) as a minor phase.
The difference in the crystalline products formed is due

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4050
Table 1
Chemical composition of Zn/Al for samples 2, 4–9 with increasing
synthesis pH
hydrotalcites, become more soluble and ZnO, the
predominant phase that would be precipitated. It is
concluded that the hydrotalcite phase formed at lower
pHs during synthesis would redissolve at this high pH to
form ZnO [35].
Sample
ICP-AES analysis
(ppm)
Sample mass Concentration
in 200 mL (g) ratio Zn/Al
The TEM of sample 5 (pH 9.5) shows large particles,
120 nm across, with a distorted hexagonal shape
(Fig. 2a). The particle morphology was observed to be
regular and homogenous throughout. The electron
diffraction pattern appeared very weak and the beam
exposure was destructive to the sample. The STEM
analysis exhibits a bulk ratio of Zn/Al of ꢀ2.1 for a
cluster of crystals. This analysis also shows that a small
amount of nitrogen is present; however, the signal was
very weak and the crystalline phase could not be
identified in the TEM mode. Sample 9 (pH 11.5) shows
large particles of ꢀ180 nm across. The particle shape of
this sample was almost hexagonal and regular through-
out the sample. This sample displayed a ratio of Zn/
Al=2.6.
Al
Zn
4 (pH 9)
5 (pH 9.5)
6 (pH 10)
17.47
15.75
14.44
80.40
81.03
74.73
81.95
73.49
83.92
80.47
0.0509
0.0523
0.0490
0.0516
0.0487
0.0485
0.0498
1.89
2.11
2.13
2.11
2.14
2.16
2.25
7 (pH 10.5) 16.01
8 (pH 11) 14.04
9 (pH 11.5) 16.87
2 (pH 12) 14.76
that Al is only present in a trace amount (1/20). This is
in agreement with the XRD that indicated only ZnO
with a minor trace of sodium nitrate as the crystalline
component.
The TEM of sample 10 (pH414) is primarily
composed of rounded, elongated, flat crystals of
submicron size (27–58 nm). The particle shape is regular
and the diffraction pattern was very strong, indicating a
highly crystalline phase (Fig. 2b). The STEM analysis
showed that Zn and O are the major components, and
Table 1 summarizes the chemical composition derived
from the ICP–AES analyses for samples 2, 4–9 (pH 12,
9–11.5), with a precision of ꢀ5%. The Zn/Al ratio
ranges from 1.89 (sample 3, pH 6–6.5) to 2.25 (sample 9,
pH 11.5), and differs from the ratios obtained from
STEM as they are bulk ratios. It is apparent that, whilst
expecting a theoretical ratio of 3, all samples have a
ratio of ꢀ2, which is in accordance with the findings of
Velu et al. [36] and Roussel et al. [37,38] for Zn/Cr
hydrotalcite. It is postulated that the lower ratio
observed in both studies is due to preferential crystal-
lization of Zn-hydrotalcite as Zn4Al2(OH)10(CO3) Á
xH2O, rather than a hydrotalcite with a Zn/Al ratio of
3:1. Previous studies have indicated that the possible
formation of hydrotalcites involves the reaction of a
solution containing the divalent metal with the trivalent
metal hydroxide [13]. This would explain the results
obtained with samples 1 and 2 (pH 12). It also would
explain the increase in crystallinity with the pH. The pH
range between 9.0 and 12.0 is considered to be ideal for
the preparation of hydrotalcites, and is a commonly
used range for their synthesis.
The DTA/TGA patterns of representative samples are
shown in Fig. 3. The hydrotalcite samples 5 and 9 (pH
9.5 and 11.5) show similar features in both patterns. The
TGA curve does not exhibit well-defined plateaus
between the decomposition steps, which extend over
relatively high temperature ranges. However, in light of
the DTA pattern, two major weight losses can be
identified, which is in agreement with the well-known
decomposition mechanism of such compounds, which
predicts two distinctive steps of weight loss [39–41]. The
first weight loss occurs around 200 1C and reflects the
loss of physically adsorbed (surface) and interlayer
water [36,42,43]. The loss of surface water begins as
soon as heating commences and is completed at ꢀ90 1C,
Fig. 2. TEM images of Zn/Al hydrotalcites: (a) sample 5, (b) sample
10, (c) hydrothermally treated sample 11 and (d) hydrothermally
treated sample 12.

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4051
12, hydrothermally
treated
11, hydrothermally
treated
5, pH 9.5
5
10, pH > 14
1, pH 12
9, pH 11.5
9
2, pH 12
9, pH 11.5
8, pH 11
11, hydrothermal
7, pH 10.5
6, pH 10
5, pH 9.5
DTA
4, pH 9
TGA
3, pH 6-6.5
12, hydrothermal
400
800
1200 1600 2000 2400 2800 3200 3600 4000
Wavenumber/cm-1
100
150
200
250
300
350
Temperature (˚C)
Fig. 4. FT-IR spectra of all synthetically prepared samples (1–12),
showing the effects of pH and hydrothermal treatment.
Fig. 3. Thermal analysis of Zn/Al hydrotalcite of samples 5, 9, 11, and
12: (a) TGA, (b) DTA.
the removal of the interlayer water is finished at
ꢀ190 1C. TGA-MS study of Mg/Al-hydrotalcite has
shown that up to 200 1C water is lost in several steps,
accumulating in a major loss around 195 1C [44]. The
decomposition step, starting at ꢀ200 1C, is due to
the decomposition (dehydroxylation) of the hydroxide
layers and the removal of interlayer anions (carbonate)
[44]. The end products of the thermal decomposition
have been shown to be metal oxides as well as mixed-
oxide spinel-type phases [43,45]. The patterns of various
samples differ with respect to the temperature at which
the weight losses occur, such that sample 5 (pH 9.5)
occurs at a lower temperature than for sample 9 (pH
11.5). Since stability is dependant on temperature,
sample 9 appears to be the most stable sample from
this series [46]. It is concluded, in comparison with the
XRD and TEM results, that the more crystalline
samples have the highest thermal stability.
hydrotalcites (Zn, Al) are distributed at random, it is
unlikely than an OH group is surrounded by three metal
cations of the same kind, especially Al3OH. In most
cases, one OH group will be coordinated by the two
different metal cations. In the following sections, the
vibrations will be assigned to the coordination which it
most likely resembles, e.g. ‘‘Al-OH’’. All spectra show a
very broadband between 2700 and 3750 cmꢁ1. Band
component analysis (Fig. 5) reveals three superimposed
peaks at 3025, 3348, and 3468 cmꢁ1. The 3025 cmꢁ1
band is generally interpreted as the CO3–H2O bridging
mode [7,47]. The 3348 cmꢁ1 band is ascribed to the H-
bonded interlayer H2O surrounding the anions. The
strong band at 3468 cmꢁ1 is assigned to the metal-OH-
stretching mode. However, the various metals (Zn, Al)
bonded to the OH group cannot be distinguished
between themselves [48].
Within the range of pH synthesis conditions, the
positions of the infrared bands change up to 50 cmꢁ1
and generally the intensity increases with increasing pH.
In addition, peaks become sharper with increasing pH.
Since increasing intensity corresponds to an increase in
the order, the samples prepared at pH 11.5 and 12 are
the most crystalline and the sample prepared at pH ꢀ6.5
The FT-IR spectra for samples synthesized at various
pH conditions are shown in Fig. 4, and their assignment
in Table 2. With the infrared assignment, it has to be
kept in mind that most of the OH groups are (when not
at the edges of the crystals) triply coordinated to the di-
and trivalent cations [33]. If the various metals in the

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4052
hydrotalcites. Band component analysis of sample 2 (pH
12) shows only two broad peaks at 1361 and 1397 cmꢁ1
corresponding to the splitting of antisymmetric stretch-
ing mode, n3, of carbonate to n3 and n3a (Fig. 6b).
Sample 3 (pH 6–6.5) also show signs of nitrate; however,
the carbonate antisymmetric stretching mode has
shifted. Sample 3 (pH 6–6.5) shows a very weak broad
peak at ꢀ1450 cmꢁ1, which is assigned to carbonate and
occurs at a higher wavenumber than for the other
samples. XRD has already shown that this sample did
not contain any crystalline hydrotalcite, which fits well
with the shift in the carbonate band in the infrared
spectrum. Sample 10 (pH414) is different as it has a
very sharp nitrate peak around 1388 cmꢁ1 and a very
weak, almost indistinguishable carbonate band. Con-
sidering that this sample was characterized as to be
primarily composed of ZnO, these results are consistent,
as no carbonate should be present.
Table 2 gives the infrared band positions, of selected
samples between 1000 and 400 cmꢁ1 and their assign-
ments. Band component analysis (Fig. 7) shows the
complexity of the spectrum in this region, where in most
cases a number of overlapping bands exist. The assign-
ment of these bands is consistent with the XRD results,
where most samples show a mixture of bands including
those of ‘‘Zn-OH’’, ‘‘Al-OH’’, CO23ꢁ and NO3ꢁ. Sample
2 (pH 12), for example, clearly shows in this region a
pattern typical of hydrotalcite, where it has a broad
shoulder at 868 cmꢁ1 and strong bands at 553 and
430 cmꢁ1 [51]. Sample 10 shows no ‘‘Al-OH’’ transla-
tions modes, as was expected.
-1
3468cm
-1
3348cm
-1
3025cm
Fig. 5. Band component analysis of the OH-stretching region for the
infrared spectrum, of synthetic Zn/Al hydrotalcite sample 2.
is the least crystalline. All samples have a weak broad
peak centered ꢀ1630 cmꢁ1. This peak is associated with
the deformation mode of water (n2).
Samples 2, 4–9, and 12 (pH 12, 9–11.5 and hydro-
thermally treated) show a peak centered ꢀ1400 cmꢁ1 in
the infrared spectrum, associated with the presence of
CO23ꢁ species in the interlayer region. Assuming the
carbonate species is present as the free ion (D3h) and not
specifically involved in any bonds, only three bands
should be present in the infrared spectrum, n2, n3 and n4
[49,50]. Comparison of the spectrum carbonate modes
with that of the free ion discloses a shift to lower
wavenumbers, which is due to the interaction of the
carbonate ions with the water molecules in the confined
space between the hydroxide layers resulting in loss of
freedom, therefore leading to a lowering of the
carbonate symmetry to C2s or Cv [48]. The out-of-plane
bending mode (n2) and in-plane bending mode (n4) are
usually found at 874 and 680 cmꢁ1, respectively. Close
examination of the spectra reveals the activation of the
symmetric stretching vibration, n1, as a very weak
shoulder band at ꢀ1045 cmꢁ1, which is overlapped by
the metal-OH translational modes. The lowering of
symmetry is further confirmed by the splitting of the n3
mode.
In certain samples, the n3 antisymmetric stretching
mode at ꢀ1400 cmꢁ1 has a sharp peak at high intensity,
with a broad base at low intensity. Band component
analysis of sample 4 prepared at pH 9.0 shows the
presence of three overlapping bands: one sharp and two
broader bands (Fig. 6a). The broader bands at 1359 and
1399 cmꢁ1 are assigned to the antisymmetric stretching
mode (n3) of the CO23ꢁ species. The sharper band at
1384 cmꢁ1 is assigned to nitrate, as nitrate always gives a
very sharp and intense peak around this frequency. It
should be noted that with increasing pH the presence of
nitrate is no longer observed. The presence of nitrate in
the interlayer region of the hydrotalcite is caused by the
fact that metal nitrates were used to synthesize the
The examination of the samples using Raman
spectroscopy compliments the infrared study [52]. The
hydroxyl-stretching region (3000–4000 cmꢁ1) of the
spectra is quite broad and very weak. Fig. 8 shows the
Raman spectra between 350 and 1600 cmꢁ1 of all
samples.
The first group of samples 1, 2, 4–9, and 12 (pH 12,
9–11.5 and hydrothermally treated) display an intense
peak centered at ꢀ1064 cmꢁ1, assigned to the n1
symmetric stretching mode (A01) of carbonate and the
‘‘Al–OH’’ translation [49]. The antisymmetric stretching
mode (E0) n3 and in-plane bending mode (E0) n4 usually
lie at 1415 and 680 cmꢁ1, respectively; however, they are
eminently weak and are unable to be confidently
assigned [53]. Sample 3 has two peaks at 1052 and
1077 cmꢁ1, assigned to the nitrate symmetric stretching
mode (n1) and carbonate symmetric stretching mode,
respectively. The frequency of the carbonate symmetric
stretching mode has shifted to a higher frequency than
the free ion; this is due to the lowering of symmetry from
the free ion D3h. Sample 10 (pH414) shows a very weak
and broad 1068 cmꢁ1 peak that cannot be precisely
assigned.
At the lower end of the Raman spectrum, samples
that contained an identifiable hydrotalcite phase and

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4053
(a)
(b)
1361 cm-1
1384 cm-1
1359 cm-1
1399 cm-1
1397 cm-1
1250
1300
1350
1400
1450
1500
1250
1300
1350
1400
1450
1500
-1
Wavelength (cm-1
)
Wavelength (cm
)
(c)
1423 cm-1
1510 cm-1
1384 cm-1
1359 cm-1
1275
1325
1375
1425
1475
1525
1575
Wavelength (cm-1)
Fig. 6. Band component analysis of the CO23ꢁ antisymmetric stretching mode (n3) in the infrared spectrum of synthetic Zn/Al hydrotalcite samples:
(a) 4, (b) 2, and (c) 11.
exhibited similar carbonate infrared active stretching
modes (samples 1, 2, 4–9, 12) also have comparable
bands in the Raman spectra. The samples show a peak at
ꢀ553cmꢁ1 assigned to the ‘‘Al-OH’’ translation mode as
well as a peak ꢀ500 cmꢁ1 that can be assigned to ‘‘Zn-
OH’’ or ‘‘Al-OH’’ translation modes, but is probably a
combination of both. Both these peaks have shoulders
that are too weak to accurately assign. Sample 3 shows
no assignable peaks, consistent with XRD and IR, in this
region. Sample 10 (pH414) shows the same peak at
439 cmꢁ1 and is more intense and sharper that the former
sample. Furthermore, this sample contains no peaks at
ꢀ555cmꢁ1 assigned to ‘‘Al-OH’’ as expected.
3.3. The effect of hydrothermal treatment of Zn/Al–Ht
Two samples from the previous pH series were treated
hydrothermally; one synthesized at the high pH with
good crystallinity (sample 9, pH 11.5) and the other
at lower pH with poor crystallinity (sample 5, pH 9.5).
Fig. 1 shows that the two hydrothermally treated
samples (11 and 12) have different structures. Sample
11 appears to have lost all of the hydrotalcite structure
and is composed of one major crystalline phase, ZnO,
and two or even three other minor crystalline phases,
containing various amounts of Zn2+, Al3+, CO32ꢁ, and
OHꢁ. The minor phases are associated with the

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4054
Table 2
Infrared frequencies and assignments of the various modes of vibrations between 1000 and 400 cmꢁ1 [48,49,51]
1
2
10
11
12
NOꢁ3 CO23ꢁ
Natural Gibbsite
HT
Hydrozincite
(Al(OH)3) (Zn5(CO3)2(OH)6) (ZnO)
Zincite Suggested
assignment
(pH 12) (pH 12) (pH 414) (HT) (HT)
950
908
942
868
948
958
‘‘Al-OH’’ deformation
‘‘Zn-OH’’ deformation
n2-CO23ꢁ
883
868
900–880
879–866
870
839
791
747
866
834
793
865
n2-CO23ꢁ in a HT
n2-NOꢁ3
833
839
692
785
739
800
749
738
‘‘Al-OH’’ translation
‘‘Al-OH’’ translation
‘‘Zn-OH’’ translation
n4-NOꢁ3
732
690
695
690
676
686
680
670
555
n4-CO23ꢁ
644
710
‘‘Zn-OH’’ translation
‘‘Zn-OH’’ translation
‘‘Al-OH’’ translation
‘‘Al-OH’’ translation
‘‘Zn-OH’’ translation
‘‘Zn-OH’’ or ‘‘Al-OH’’
translation
619
579
552
620
591
552
617
614
587
553
615
585
562–540
545
515
557
535
518
523
521
515
420
520
460
488
440
483
430
‘‘Zn-OH’’ translation
‘‘Zn-OH’’ translation
‘‘Al-OH’’ translation
Unknown
450
426
440
430
435
412
422
-1
430 cm
-1
785 cm
-1
552 cm
-1
-1
620 cm
591 cm
-1
869 cm
-1
942 cm
11, hydrothermal
12,hydrothermal
-1
738 cm
-1
686 cm
-1
523 cm
400
450
500
550
600
650
700
750
-1
800
850
900
950
1000
Wavenumber (cm
)
10, pH > 14
1, pH 12
Fig. 7. Band component analysis of the lower frequency region,
between 1000 and 400 cmꢁ1, of the infrared spectrum of synthetic Zn/
Al hydrotalcite sample 2.
2, pH 12
reflections at 15.0 and 22.612y, indicating a small
amount of hydrozincite, Zn5(OH)6(CO3)2, and/or zinc
carbonate hydroxide hydrate, Zn4CO3(OH)6 Á xH2O.
The reaction mechanism that plays a major role in the
decomposition of the low-crystallinity hydrotalcite is
unclear at this stage. Sample 12 yielded a highly
crystalline hydrotalcite-type compound, where the peaks
are more intense and sharper than the sample from
which it was prepared. However, there are several small
peaks, which are just resolved that are associated with
sodium nitrate.
9, pH 11.5
8, pH 11
7, pH 10.5
6, pH 10
5, pH 9.5
4, pH 9
3, pH 6-6.5
350
550
750
950
1150
1350
1550
The TEM of sample 11 (Fig. 2c) shows numerous
crystalline phases. The most obvious phase is thick,
needle-like and the largest crystal type in this sample.
The type is extremely crystalline, exhibiting a strong
Wavenumber/cm-1
Fig. 8. Raman spectra of samples, showing the changes brought about
by pH and hydrothermal treatment.

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4055
electron diffraction pattern, and according to the STEM
is primarily composed of Zn and O. This phase is
assigned to the hydrozincite, even though carbon could
not be observed, but the structure is different from the
ZnO phase of sample 10 (pH 14). The next phase is
structurally similar to the other ZnO phase observed
with elongated crystals that are rounded at the top. This
phase differs from the first phase in terms of the
morphology; it is thinner and edges are rounded, and is
composed of Zn and O in a different ratio. This phase is
assigned to ZnO. The third phase has a hexagonal shape
with a particle size of ꢀ77 nm across. The diffraction
pattern is strong and the STEM analysis revealed a
ratio of Zn/Al of ꢀ2. The fourth phase is also highly
crystalline according to the diffraction pattern,
and particles have cubic-type morphology. The particle
size is approximately the same as the hexagonal-shaped
phase but smaller than the first two phases. This phase
is quite common and appears to make up the bulk
of the sample, and is primarily composed of Al/Zn with
a ratio of 5/4. This was not expected since the XRD
showed that no crystalline Al phase was likely to be
present. Although this cubic phase makes up the bulk of
the sample, the initial two ZnO phases are highly
crystalline and exhibit very strong diffraction patterns,
which probably suppress the diffraction of the smaller
less crystalline phases. The final identifiable phase
is a flat flaky structure, which exhibits a diffraction
pattern and is also a crystalline phase. The STEM
reveals that this phase is composed of Zn, Al, C and
O in varying compositions, indicating a hydrotalcite-like
phase.
The TEM of sample 12 (Fig. 2d) showed large, flat
particles (260 nm) similar to those of the untreated
samples. Although the morphology of most particles is
fairly regular, there does appear to be another phase.
The predominant phase has a distorted hexagonal shape
where some of the edges are rounded, similar to sample
9, indicating signs of dissolution during the hydrother-
mal treatment. The electron diffraction pattern shows a
pattern stronger than sample 9. The STEM analysis
found a ratio of Zn/Al ratio of 2.5–3.2, compared to an
initial ratio of 2.6 in sample 9. The TEM also showed
smaller crystals with a cubic morphology, on the
rounded surfaces of larger hexagonal-like crystals,
which were too small for microanalyses. Thus, hydro-
thermal treatment increases the crystallinity and allows
time for the classical hexagonal shape of the hydrotalcite
to form [2,54]. The presence of the smaller cubic-shaped
particles indicates that with aging predominant crystal
growth occurs on the edges, resulting in relatively thin
hexagonal plate-shaped crystals with rounded edges.
These smaller particles are believed to be the initial
stages of growth of a second phase that was found to be
present. This second phase consisted of rounded
elongated crystals that are composed of Zn and O, in
a ratio of 1:1. It is concluded that these particles grew as
a result of the dissolution of the larger hexagonal-
shaped, crystalline phases. However, the contribution
from this phase to the total sample is small as only a few
crystals were observed.
The TGA/DTA of sample 12 is shown in Fig. 3. This
sample exhibits similar thermal properties as the two
starting samples previously discussed. The primary
difference between all these samples is the thermal
stability. Sample 12 is the most thermally stable, where
all decomposition steps occur at higher temperatures
than for the other samples. Furthermore, this sample
confirms that the more crystalline samples are the most
thermally stable.
The infrared spectrum of sample 11 is eminently
different from the hydrotalcite samples (Fig. 4, Table 2).
At high frequency (3000–4000 cmꢁ1), a broad feature
exists that is assigned to OH-stretching modes. At mid-
frequency, several bands are prominent that are assigned
to the presence of carbonate. Band component analysis
of this region shows the CO23ꢁ antisymmetric stretching
modes are less superimposed, occurring at frequencies of
1423 and 1510 cmꢁ1. These bands occur at different
frequencies as they are not residing in the interlayer, as
with the hydrotalcites. Instead, it is assumed that
carbonate is present in a covalent metal-to-anion bond,
Zn–CO3, where the individual sites occupied by Zn, C
and O have symmetries C3I=S6, D3, and C2, respec-
tively [20,24]. It is also noticed that they occur at higher
frequencies; thus the carbonate species present in this
sample is symmetric than those in the hydrotalcites. The
band at 1384 cmꢁ1 is associated with the presence of
nitrate. Furthermore, this sample shows another weak,
broad, shoulder peak at 1359 cmꢁ1, which can be
assigned to a vibrational mode of hydrozincite, usually
found at 1340 cmꢁ1, which is consistent with previous
studies [51]. At lower frequency, sample 11 shows a
different pattern compared to all other samples. Band
component analysis reveals numerous overlapping
bands, where the predominant bands are associated
with ‘‘ZnOH’’ translational modes (Table 2). Unlike
sample 10 (pH414), also primarily composed of ZnO,
this sample shows CO32ꢁ and NOꢁ3 in-plane and out-of-
plane bends (n4 and n2), that are associated with the
hydrozincite phase, which lie at higher wavenumbers
than for the hydrotalcites.
The infrared spectrum of samples 12 (Fig. 4) is very
similar to those of the other hydrotalcites, although
there are several weak bands at low frequency that the
other hydrotalcites do not show; these bands lie at a
frequency associated with nitrate, 834 and 690 cmꢁ1
.
The primary difference between the spectra of these
hydrothermal samples and the untreated samples is in
the intensity; such that sample 12 has more intense,
sharper and better-resolved peaks than the starting
material, sample 9 (pH 11.5), and all other untreated

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4056
hydrotalcites. This sample has experienced an increase
in crystallinity, with treatment [2,54].
Acknowledgments
The Raman spectrum of sample 11 (Fig. 3) shows an
intense, broad peak at 1280 cmꢁ1 with a shoulder at
1466 cmꢁ1. In this sample, the carbonate is not present
as the free ion. The position of the band is at a higher
wavenumber than the other hydrotalcite samples; thus,
this peak is assigned to a carbonate species with a lower
symmetry than the other samples and the free ion. The
band component analysis of this region shows four
bands at 1222, 1265, 1319, and 1458 cmꢁ1. These bands
are associated with carbonate symmetric and antisym-
metric stretching modes, n1 and n3. The peaks at 1265
and 1319 cmꢁ1 are the n1 symmetric stretching modes
and 1222 and 1468 cmꢁ1 are the n3 antisymmetric
stretching modes. The antisymmetric stretching modes
are assigned to the weaker, broader peaks, as typically
these modes are difficult to assign in the Raman
spectrum [53]. This sample has another small peak at
1058 cmꢁ1 that is assigned to the nitrate symmetric
stretching mode, n1. At the lower end of the spectrum,
this sample shows the same ‘‘Zn-OH’’ translational peak
at 438 cmꢁ1, but, however, no identifiable ‘‘Al-OH’’
peak, as it is probably too weak. Sample 12 shows
features similar to the other hydrotalcites, but, however,
is more ordered, as the peaks are more intense and
sharper.
The authors thank Tony Raftery for assistance with
X-ray diffraction. Sharyn Price, of the Natural Re-
sources Facility at QUT, is acknowledged for her
assistance with the elemental analysis using ICP-OES.
Thor Bostrom, of the Analytical Electron Microscopy
Facility at QUT, is thanked for his assistance with the
STEM analytical phase analysis. The contribution of
Shane Russell with the DTA/TGA is greatly appre-
ciated. The financial and infrastructural support of the
QUT, Inorganic Materials Research Program, is grate-
fully acknowledged.
References
[1] F. Trifiro, A. Vaccari, in: T. Bein (Ed.), Comprehensive
Supramolecular Chemistry, Vol. 7, Pergamon, Oxford, 1996,
p. 251.
[2] W.T. Reichle, Solid State Ion. 22 (1986) 135.
[3] A. Vaccari, Catal. Today 41 (1998) 53.
[4] M.K. Titulaer, J.B.H. Jansen, J.W. Geus, Clays Clay Miner. 42
(1994) 249.
[5] M. Badreddine, M. Khaldi, A. Legrouri, A. Barroug, M.
Chaouch, A.D. Roy, J.P. Besse, Mater. Chem. Phys. 52 (1998)
235.
[6] F.M. Labajos, V. Rives, M.A. Ulibarri, J. Mater. Sci. 27 (1992)
1546.
[7] S. Miyata, Clays Clay Miner. 23 (1975) 369.
[8] S. Miyata, Clays Clay Miner. 28 (1980) 50.
[9] F. An, Huagong Keji 11 (2003) 35.
4. Conclusion
[10] M. Badreddine, A. Legrouri, A. Barroug, A. De Roy, J.-P. Besse,
Collect. Czech. Chem. Commun. 63 (1998) 741.
[11] U. Costantino, F. Marmottini, M. Nocchetti, R. Vivani, Eur.
J. Inorg. Chem. (1998) 1439.
XRD, STEM, ICPAES, thermal analysis, infrared
and Raman spectroscopy are powerful methods em-
ployed in the characterization of hydrotalcite-type
materials [2]. The effect of reagent addition on the
purity and crystallinity of the hydrotalcite materials
showed that adding the base (NaOH) to the acid,
proceeding via an acidic pathway, favors the formation
of the hydrotalcite. Under strong alkaline conditions,
ZnO species are not soluble and are found to compete
with the formation of Zn/Al hydrotalcites.
The variation of pH showed that at low pH no
hydrotalcite-type structure formed and that with in-
creasing pH, up to pH ꢀ12, the crystallinity also
increased. However, at high alkaline pH (14) ZnO
forms, probably from the redissolution of the hydro-
talcite materials formed at lower pH. Furthermore, the
stability of these compounds was observed to increase
with increasing pH.
[12] A. Ennadi, A. Legrouri, A. de Roy, J. Pierre Besse, J. Mater.
Chem. 10 (2000) 2337.
[13] J.W. Boclair, P.S. Braterman, Chem. Mater. 11 (1999) 298.
[14] M. Badreddine, M. Khaldi, A. Legrouri, A. Barroug, M.
Chaouch, A. de Roy, J.P. Besse, Mater. Chem. Phys. 52 (1998)
235.
[15] M. Badreddine, A. Legrouri, A. Barroug, A. de Roy, J.P. Besse,
Mater. Lett. 38 (1999) 391.
[16] L.F. da Silva, J. Tronto, H.P. Oliveira, J.B. Valim, J. Inclusion
Phenom. Macrocyclic Chem. 46 (2003) 187.
[17] D.P. Das, J. Das, K. Parida, J. Colloid Interf. Sci. 261 (2003) 213.
[18] M. Guenane, C. Forano, J.P. Besse, Mater. Sci. Forum 152–153
(1994) 343.
[19] J.L. Guimaraes, R. Marangoni, L.P. Ramos, F. Wypych,
J. Colloid Interf. Sci. 227 (2000) 445.
[20] E. Kanezaki, J. Mater. Sci. 30 (1995) 4926.
[21] E. Kanezaki, J. Inclusion Phenom. Macrocyclic Chem. 46 (2003)
89.
[22] F. Kooli, W. Jones, Inorg. Chem. 34 (1995) 6237.
[23] F. Kooli, C. Depege, A. Ennaqadi, A. de Roy, J.P. Besse, Clays
Clay Miner. 45 (1997) 92.
Comparing hydrothermal samples showed that treat-
ment of unstable, confirmed by STEM and thermal
analysis, hydrotalcites favors the formation of ZnO and
other non-hydrotalcite phases. Where stable, starting
materials produce highly crystalline hydrotalcite-type
products, which maintain the classical hexagonal-
shaped particle.
[24] S. Velu, V. Ramkumar, A. Narayanan, C.S. Swamy, J. Mater. Sci.
32 (1997) 957.
[25] A. Andreev, V. Kafedzhiiski, T. Khalachev, B. Kunev,
M. Kalchev, Appl. Catal. 78 (1991) 199.
[26] K. Bahranowski, R. Dula, M. Gasior, M. Labanowska, A.
Michalik, L.A. Vartikian, E.M. Serwicka, Appl. Clay Sci. 18
(2001) 93.

ARTICLE IN PRESS
J.T. Kloprogge et al. / Journal of Solid State Chemistry 177 (2004) 4047–4057
4057
[27] S. Bhattacharjee, J.A. Anderson, Chemical Communications,
Cambridge, UK, 2004, p. 554.
[43] V. Rives, in: V. Rives (Ed.), Layered Double Hydroxides: Present
and Future, Nova Science Publishers Inc., New York, 2001,
p. 115.
[28] A. Cervilla, A. Corma, V. Fornes, E. Llopis, P. Palanca, F. Rey,
A. Ribera, J. Am. Chem. Soc. 116 (1994) 1595.
[29] J. Das, K. Parida, React. Kinet. Catal. Lett. 69 (2000) 223.
[30] Y.-H. Ge, W.-M. Lu, J.-X. Mao, Wuji Huaxue Xuebao 18 (2002)
1143.
[44] J.T. Kloprogge, J. Kristof, R.L. Frost, in: E.A. Dominguez, G.R.
Mas, F. Cravero (Eds.), A Clay Odyssey, Proceedings of the 12th
International Clay Conference, Elsevier, Bahia Blanca, Argentina,
2001, p. 451.
[31] A.H. Padmasri, A. Venugopal, V. Durga Kumari, K.S. Rama
Rao, P. Kanta Rao, J. Mol. Catal. A 188 (2002) 255.
[32] R. Zhao, C. Yin, H. Zhao, C. Liu, Fuel Process. Technol. 81
(2003) 201.
[45] F.M. Labajos, V. Rives, M.A. Ulibarri, J. Mater. Sci. 27 (1992)
1546.
[46] S. Morpurgo, M.L. Jacono, P. Porta, J. Solid State Chem. 122
(1995) 324.
[33] J.T. Kloprogge, R.L. Frost, Phys. Chem. Chem. Phys. 1 (1999)
1641.
[47] W. Kagunya, R. Baddour-Hadjean, F. Kooli, W. Jones, Chem.
Phys. 236 (1998) 225.
[34] B. Houri, A. Legrouri, A. Barroug, C. Forano, J.P. Besse, Collect.
Czech. Chem. Commun. 63 (1998) 732.
[48] J.T. Kloprogge, R.L. Frost, in: V. Rives (Ed.), Layered Double
Hydroxides: Present and Future, Nova Science Publishers, Inc.,
New York, 2001, p. 139.
[35] C.F. Baes, R.E. Mesmer, The Hydrolysis of Cations, Wiley, New
York, 1976.
[49] V. Farmer, The Infrared Spectra of Minerals, Mineralogical
Society, England, 1974.
[36] S. Velu, V. Ramkumar, A. Narayanan, C.S. Swamy, J. Mater. Sci.
32 (1997) 957.
[50] K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, fifth ed., Wiley, USA, 1997.
[51] J.A. Gadsden, Infrared Spectra of Minerals and Related
Inorganic Compounds, Butterworths, England, 1975.
[52] B. Schrader, Infrared and Raman Spectroscopy: Methods and
Applications, Weinheim, New York, 1995.
[37] H. Roussel, V. Briois, E. Elkaim, A. de Roy, J.-P. Besse, J.-P.
Jolivet, Chem. Mater. 13 (2001) 329.
[38] H. Roussel, V. Briois, E. Elkaim, A. de Roy, J.P. Besse, J. Phys.
Chem. B 104 (2000) 5915.
[39] E.C. Kruissink, L.L.V. Reijen, J.R.H. Ross, J. Chem. Soc.
Faraday Trans. I 77 (1981) 649.
[40] M. Lal, A.T. Howe, J. Solid State Chem. 39 (1981) 368.
[53] S.D. Ross, Inorganic Infrared and Raman Spectra, McGraw-Hill
Book Company, London, 1972.
[41] N. Gutmann, B. Muller, J. Solid State Chem. 122 (1996) 214.
¨
[42] S. Yun, T.J. Pinnavaia, Inorg. Chem. 35 (1996) 6853.
[54] W.T. Reichle, S.Y. Kang, D.S. Everhardt, J. Catal. 101 (1986)
352.