Chemical synthesis of environment-friendly nanosized yellow titanate pigments

Article abstract of DOI:10.1016/j.materresbull.2007.04.001

As yellow pigments, nanosized rutile structured Ni_0.1W_0.1Ti_0.8O₂ and priderite structured BaNiTi₇O₁₆ have been prepared through pyrolysis of precursor solution containing nickel nitrate, dimethyl tungstate, titanium oxalate, triethanolamine (TEA) for the former composition and barium nitrate, nickel nitrate, titanium oxalate, TEA for the later composition, respectively. In the reaction, TEA acts to minimize the agglomeration in the products through formation of a highly branched polymeric framework, which anchors the metal ions for producing nanocrystalline powders. The rutile and priderite structured titanates obtained on heat-treatment of the precursor powders at 800 °C and 850 °C, respectively, have been characterized by XRD, TGA-DTA, BET surface area measurement, UV-vis spectroscopy, CIE L^*a^*b^* colour parameter measurements, TEM and HRTEM. XRD reveals the purity of the resulting rutile and priderite phases. Their crystallite sizes, average particle sizes and specific surface areas determined from XRD, TEM, and BET surface area measurement are found to be in the range between 25 nm and 30 nm, 25-45 nm and 100-120 m²/g, respectively, for both compositions.

Full text of DOI:10.1016/j.materresbull.2007.04.001

Materials Research Bulletin 43 (2008) 665–675
www.elsevier.com/locate/matresbu
Chemical synthesis of environment-friendly nanosized
yellow titanate pigments
Soumya Kanti Biswas a, Debasis Dhak a,b, Amita Pathak a, Panchanan Pramanik a,
*
a Department of Chemistry, Indian Institute of Technology Kharagpur, Kharagpur 721302, India
b Department of Chemistry, Serampore College, Hoogly 712201, India
Received 26 September 2006; received in revised form 22 March 2007; accepted 2 April 2007
Available online 6 April 2007
Abstract
As yellow pigments, nanosized rutile structured Ni0.1W0.1Ti0.8O2 and priderite structured BaNiTi7O16 have been prepared
through pyrolysis of precursor solution containing nickel nitrate, dimethyl tungstate, titanium oxalate, triethanolamine (TEA) for
the former composition and barium nitrate, nickel nitrate, titanium oxalate, TEA for the later composition, respectively. In the
reaction, TEA acts to minimize the agglomeration in the products through formation of a highly branched polymeric framework,
which anchors the metal ions for producing nanocrystalline powders. The rutile and priderite structured titanates obtained on heat-
treatment of the precursor powders at 800 8C and 850 8C, respectively, have been characterized by XRD, TGA–DTA, BET surface
area measurement, UV–vis spectroscopy, CIE L*a*b* colour parameter measurements, TEM and HRTEM. XRD reveals the
purity of the resulting rutile and priderite phases. Their crystallite sizes, average particle sizes and specific surface areas determined
from XRD, TEM, and BET surface area measurement are found to be in the range between 25 nm and 30 nm, 25–45 nm and
100–120 m2/g, respectively, for both compositions.
# 2007 Elsevier Ltd. All rights reserved.
Keywords: A. Nanostructures; A. Oxides; B. Chemical synthesis
1. Introduction
The most known yellow inorganic pigments are praseodymium-zircon yellow, vanadium–zirconia yellow, tin–
vanadium yellow, chromates of alkaline earth metal ions (e.g. calcium, strontium, barium), lead and zinc, litharge (lead
oxide), naples yellow (lead antimonate), cadmium yellow, iron oxide yellow (a and g-FeOOH), and nickel–antimony
doped rutile phase TiO2. Among these inorganic yellow pigments, the application of iron oxide yellow gets
constrained due to thermal stability only up to 220 8C [1], while the toxicity [2] of chromium(VI), cadmium and lead
based yellow pigments restricts their commercial usage; and praseodymium-zircon yellow, vanadium–zirconia yellow
which are even if most popular in the market of yellow inorganic pigments, they have some limitations in bulk
colouration porcelainized stoneware at high temperature [3].
Nickel-doped TiO2 with rutile as well as priderite structures are well known intense and brilliant yellow ceramic
pigments [4]. They are non-toxic, chemically inert, thermally stable, photochemically inactive, and having high tinting
strength for the outdoor application and hence satisfy most of the characteristics requisite for pigment application. Still
* Corresponding author. Tel.: +91 3222 283322; fax: +91 3222 255303.
E-mail addresses: pramanik@chem.iitkgp.ernet.in, panchanan_123@yahoo.com (P. Pramanik).
0025-5408/$ – see front matter # 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.materresbull.2007.04.001

666
S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
they find little usage in the ceramic industry because of their partial solubility into the transparent glazes above
1000 8C [5]. However, due to their high thermal stability these pigments have gained considerable industrial interest in
recent years as potential alternatives to praseodymium zircon yellow pigment, essentially for bulk colouration of
´
porcelain stoneware tiles [6,7]. Recently Sorly et al. reported the structure and colour of nickel doped rutile phase
yellow pigment with compositions: NixA1 ꢀ 3xB2xO2 [A = Ti(IV), Sn(IV); B = Sb(V), Nb(V)] prepared by
conventional ceramic method [3] and also studied the role of nickel precursors on the colour of pigments prepared
from different nickel starting materials [8]. Matteucci et al. [5,9] studied the role of counterions (Mo, Nb, Sb, W) in
doped rutile ceramic pigments with compositions NixA1 ꢀ 3xB2xO2 and NixA1 ꢀ 2xBxO2 [A = Ti(IV), Sn(IV);
B = W(VI), Mo(VI)] and found the tungsten to be the most suitable counterion for best titanate pigments formulation
in glaze applications.
For the formation of TiO2 based yellow pigments, chromophorous ions, Ni2+ having ionic radii (ionic
4+
˚
˚
radius = 0.69 A) slightly higher than Ti (0.605 A) are incorporated into the host rutile (TiO2) lattice to generate
oxide solid solutions with stable colorimetric properties. In formulations such as TiIV(WVINiII)O2 the colouring effect
is produced when the Ti(IV), which is surrounded by six neighbouring oxygen atoms at the corners of a regular (but
slightly distorted) octahedron in the rutile lattice, is substituted by transition metal ions of lower valency Ni2+ (ionic
6+
˚
˚
radius = 0.69 A) while the higher valency ions in the lattice (e.g. A = W ; ionic radius = 0.60 A) compensate for the
charge off set such that their molar ratios ensure an overall electroneutrality in the oxide solid solution. In the present
paper, nickel doped rutile phase titanate pigment with composition TiIV(WVINiII)O2 has been reported as a prototype.
In this composition, molybdenum can play the same role of tungsten producing different yellow colour.
Solid solution of BaNiTi7O16 is also a recognized yellow pigment having a priderite structure. Priderite is a member
of hollandite family and may have tetragonal or monoclinic structure depending upon rB/rA ratio. The rB/rA = 0.48
corresponds to approximate tetragonal/monoclinic boundary (at room temperature), where rB is the average of the
radius of Ni2+ and that of Ti4+ and rA is the radius of Ba2+ ion [10]. For Bax(NixTi8 ꢀ x)O16 (with x = 1) composition,
the rB/rA value is less than 0.48 and the greenish-yellow compound can therefore be inferred to have a tetragonal
structure.
For concerning the toxicity, both TiIV(WVINiII)O2 and BaNiTi7O16 are safe for environment, because they are not
composed of toxic elements and they have low solubility in water and dilute mineral acids.
The ceramic pigments with grain sizes in nanoscale are commercially potent in pigment industries, because of their
high surface area which assures higher surface coverage; higher number reflectance points and hence maximum
scattering; and small particle size for uniform dispersion by homogenous mixing with the binders in paint formations,
which enhances the mechanical strength of the paint after drying. When properly dispersed, the nanosized pigments
exhibit superior effectiveness in critical abrasive and polishing applications also.
The common route to synthesize these coloured titanates is solid-state method, which requires high temperatures
and repeated mixing/grinding of raw materials and produces coarse particles. In the present paper, chemical synthesis
of nanosized nickel doped yellow solid solutions with compositions Ni0.1W0.1Ti0.8O2 and BaNiTi7O16 have been
reported through an aqueous-based chemical route using dimethylammonium tungstate and titanium oxalate as
innovative and stable water-soluble sources of tungsten and titanium, respectively. The method [11] involves
thermolysis of precursors obtained from evaporation of aqueous based precursor solution of soluble metal salts/
complexes and optimum amount of triethanolamine (TEA). Solution based on this method produces pure and single
phased nanosized pigments maintaining the required stoichiometry at temperature lower compared to the conventional
solid-state method [3,5]. This paper also describes the structural and colorimetric properties of these pigments.
2. Experiment
Ni0.1W0.1Ti0.8O2 and BaNiTi7O16 have been prepared from aqueous solutions of nickel nitrate, dimethylammo-
nium tungstate, titanium oxalate and TEA for the former composition; and nickel nitrate, barium nitrate, titanium
oxalate and TEA for the latter.
2.1. Raw materials used
Raw materials used were: (a) nickel nitrate hexahydrate (Merck India Ltd.), (b) sodium tungstate (Merck India
Ltd.), (c) titanium(IV) oxide (Merck India Ltd.), (d) oxalic acid dihydrate (Merck India Ltd.), (e) TEA (Merck India

S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
667
Ltd.), (f) di-ammonium oxalate monohydrate (Merck India Ltd.), (g) barium nitrate (Merck India Ltd.), (h) HCl (35%)
(Merck India Ltd.), (i) HNO3 (70%) (Merck India Ltd.), (j) NH4OH (25%) (Merck India Ltd.) and (k) dimethylamine
(40%) (Merck India Ltd.).
2.2. Preparation of aqueous solution dimethylammonium tungstate
Weighed amount of sodium tungstate was dissolved in deionized water and then acidified by drop wise addition of
dilute hydrochloric acid (>5N) to obtain a yellow precipitate of insoluble tungstic acid. The precipitate was filtered
and after washing with deionized water to remove the chloride ions it was dissolved in minimum volumes of
dimethylamine (40% solution) and finally the resultant solution was diluted to required volume with deionized water
to obtain the stock of dimethylammonium tungstate. The tungsten-content in the solution was confirmed by assay
process.
2.3. Preparation of aqueous solution of titanium oxalate
Aqueous solution of titanium oxalate was prepared starting from its hydrous oxide [TiO2ꢁnH2O]. To begin with,
weighed amounts of titanium oxide (TiO2) (1 M > 99%) was dissolved in HF (>7 M) by heating the mixture over a
water bath for 28/30 h to obtain a clear solution of the titanium–fluoride complex [TiF62ꢀ]. Hydrous titanium oxide
was then precipitated out from this clear solution by drop wise addition of dilute ammonia, and washed with 5%
ammonia to make it free from fluoride ions. The precipitate was then slowly dissolved in an aqueous mixture of oxalic
acid (0.1 M) and ammonium oxalate (0.1 M) with constant stirring to obtain a clear solution, which was diluted to
required volume with deionized water to get the stock solution of titanium oxalate. The aqueous mixture of oxalic acid
and ammonium oxalate (3:1 mole ratio) used for the dissolution was freshly prepared and always used in excess to
stoichiometric requirement of two moles per mole of Ti(IV) ion. The titanium content in the solution was estimated by
the assay method.
For the preparation of Ni0.1W0.1Ti0.8O2 solid solution, stoichiometric amount of aqueous solution of nickel nitrate
(0.1 M) was taken from a freshly prepared stock, added into the mixture of stoichiometric amount of aqueous solutions
of titanium oxalate and dimethylammonium tungstate taken from their respective stocks and mixed up with constant
stirring. The triethanolamine (TEA) was then added into the solution mixture to keep the metal ions in solution through
complexation and the pH of the solution was adjusted to ꢂ5 with the help of nitric acid to avoid precipitation. The
amount of TEA was always added in excess to the total cations present in the solution mixture (i.e., ꢂ4 moles with
respect to the total moles of the metal ions). The resultant precursor solution was homogenized by constant stirring and
finally set to dehydrate by heating at ꢂ200 8C. On complete evaporation of the precursor solution, the TEA and the
metal-complexes decomposed with the evolution of dense fumes and resulted in a voluminous, fluffy, black organic-
based mass. The fluffy carbonaceous mass was crushed to obtain the fine precursor powder, which was calcined for 2 h
in alumina crucible (at a heating rate of 5 8C per min.) in ambient atmosphere with temperatures varying from 450 8C
to 800 8C to obtain pure rutile phased solid solution.
For the preparation of BaNiTi7O16 solid solution, stoichiometric amount of aqueous solution of nickel nitrate
(0.1 M) was taken from a freshly prepared stock and added into the mixture of stoichiometric amount of aqueous
solutions of titanium oxalate taken from stock. Stoichiometric amount barium nitrate was complexed with equal mole
ethylene diamine tetra-acetic acid (EDTA) separately with the help of few drops of ammonia. The EDTA-complexed
solution was then added to the mixture of nickel nitrate and titanium oxalate solution. The TEA (ꢂ4 moles with respect
to the total moles of the metal ions) was added in the solution mixture. The resultant precursor solution was
homogenized by constant stirring and finally evaporated by heating at ꢂ200 8C. On complete evaporation of the
precursor solution, the TEA and the metal-complexes decomposed with the evolution of dense fumes and resulted in a
voluminous, black, fluffy, mass. The fluffy carbonaceous mass was crushed to obtain the fine precursor powder, which
was calcined for 2 h in alumina crucible (at a heating rate of 5 8C per min.) in air with temperatures varying from
450 8C to 850 8C to obtain the priderite phased solid solution.
The phase identification in the oxide solid solutions was performed using room temperature X-ray powder
˚
diffraction (XRD) (Philips X-ray Diffraction Unit, PW 1710) with Cu Ka radiation (l = 1.5406 A) and Ni filter. In the
XRD studies, the powders obtained at different heat treatment temperatures were mounted in the sample holder and
recorded in the 2u range from 208 to 708 for rutile phase and from 158 to 708 for priderite phase. The X-ray diffraction

668
S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
lines were analyzed using PCPDFWIN. The crystallite size (D) of the solid solution compositions was calculated using
Sherrer equation to the full widths at half maximum (FWHM) of the diffraction peaks introducing the correction for
instrumental broadening with respect to standard silicon. The actual FWHMs have been calculated by the following
equations considering the peaks to be of the Gaussian type:
0:9l
2
2
2
ðboÞ ¼ ðbmÞ ꢀ ðbsÞ ;
D ¼
bo cos u
where bo is the integral of FWHM, bm the FWHM considering the particle size effect, bs the FWHM due to strain, for
˚
the standard Si, u the half incident angle of the X-ray, l the wavelength of the target material (l = 1.5406 A), and D is
the crystallite size. The fraction of rutile phase was calculated from the relative intensities of the XRD peaks
corresponding to the anatase and the rutile phases in the Ni0.1W0.1Ti0.8O2 composition at various heat-treated
temperatures using the Spurr equation [12]:
1
FR ¼
1 þ 1:26ðIAð1 0 1Þ=IRð1 1 0ÞÞ
where FR is the mass fraction of rutile phase in the samples and IA(1 0 1) and IR(1 1 0) are the intensities of d1 0 1 and
d1 1 0 lines of the anatase and rutile phase, respectively. Simultaneously recorded thermogravimetric and differential
thermal analysis (TGA/DTA) of the precursor powders were carried out in air at a heating rate of 5 8C/min using Perkin-
Elmer instruments (Pyris Diamond TGA/DTA). UV–vis-NIR spectroscopy studies forthe calcined powders were carried
out using Ocean Optics SD 2000. The high-resolution transmission electron microscopy (HRTEM) measurements were
carried out with 200 keVelectron beam using JEOL 2010 (ultrahigh-resolution model). For the HRTEM experiment, the
samples were prepared by suspending the heat-treated powder in ethyl alcohol by sonication and taking a drop of the
suspension on 200-mesh carbon coated copper grid. The BET specific surface areas of the calcined powders (after out-
gassingthepowdersat150 8Cfor4 h)weredeterminedthroughN2 adsorptionisothermal at77 KusingBeckmanCoulter
SA3100 surface area analyzer. The colour of the pigments was assessed from CIE L*a*b* colour parameter measure-
ments, calculated from the diffuse reflectance spectra taken with a Perkin-Elmer colorimeter using a standard illuminant
D. According to the method the parameter L* represents the brightness of a sample; L* is a lightness axis [black (0) to
white (100)]; a* represents the green (ꢀ) ! red (+) axis and b* the blue (ꢀ) ! yellow (+) axis.
3. Results and discussion
Fig. 1a shows the TGA and DTA curves for the carbonaceous precursor powders of the Ni0.1W0.1Ti0.8O2
composition. The TGA curve clearly shows 8% weight loss of up to 200 8C, and a rapid weight loss of ꢂ90% between
200 8C and 570 8C. Beyond 570 8C the TGA curve remained constant and shows no further weight loss up to 950 8C.
The first weight loss step up to 200 8C in TGA possibly corresponds to the vaporization of absorbed water. The second
sharp weight loss between 200 8C and 570 8C in TGA is manifested by four exothermic peaks in the DTA curve in the
corresponding temperature range. This exothermic thermal effect could be assigned to the oxidation of the
carbonaceous precursors that probably retained some un-decomposed fraction of metal-complexes, nitrates and TEA
even after the dehydration process. The accompanying weight loss in the TGA curve could be attributed to the
evolution of gases (such as, water vapor, and oxides of carbon and nitrogen) due to decomposition/oxidation of the un-
decompose fractions along with that of the carbon retained in the precursor. Emergence of a broad exothermic peak
beyond 600 8C in the DTA curve, with no weight loss manifestation in the TGA curve in the corresponding region,
could be attributed to the gradual transformation of the anatase phase to the rutile form in the oxide solid solution.
It can be seen from Fig. 1b that TGA/DTA curves for the BaNiTi7O16 precursors are found to be similar to that of
the Ni0.1W0.1Ti0.8O2 precursors and no significant difference is observed in the thermal behavior of the two precursors.
The thermal decomposition of the BaNiTi7O16 precursors also reflected weight loss in the TGA curve with
simultaneous exothermic thermal effect in the DTA curve up to the ꢂ600 8C due to oxidation of the un-burnt
carbonaceous mass and un-decomposed metal-complexes, nitrates and TEA retained in the precursors after complete
evaporation of precursor solution of Ba(NO3)2, Ni(NO3)2, titanium oxalate and TEA. There was no exothermic effect
or, weight loss observed beyond 600 8C signifying that the heat generated during the oxidation of the carbonaceous
precursor between 350 8C and 600 8C was sufficient for direct crystallization of the pure priderite phase from the
amorphous precursors.

S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
669
Fig. 1. Simultaneously recorded TGA/DTA plots of the carbonaceous precursors of solid solution composition: (a) Ni0.1W0.1Ti0.8O2 and (b)
BaNiTi7O16
.
Fig. 2a shows the characteristic X-ray powder diffraction (XRD) patterns for the Ni0.1W0.1Ti0.8O2 precursor
powder, when heat-treated at various temperatures between 450 8C and 800 8C for 2 h. The virgin precursor was
amorphous to X-ray and the crystallization of the oxide phase began at temperatures higher than 450 8C, which was
reflected by the emergence of broad, low-intensity diffraction lines. Up to 550 8C the precursor powder contained gray
tint indicating the presence of residual carbon. When heat-treated at 600 8C and above, the powder appeared yellow in
colour, which indicated the complete burnout of residual carbon from the precursor powders. Increase in the heat-
treatment temperatures up to 650 8C resulted only anatase phase in the solid solution indicated by the appearance of
d
1 0 1 diffraction line. Beyond 650 8C, the d1 1 0 plane of rutile began to appear and at 800 8C pure rutile phase was
formed. Thermodynamically rutile phase of titanium dioxide is stable over anatase phase. Therefore, the rutile phase is
desirable for pigment formation. In addition to this the photocatalytic efficiency of rutile phase is also diminished.
Gradual increase of calcination temperature resulted the crystal growth and increase in the rutile fraction in the
samples, which were evident by gradual sharpening of the d1 1 0 diffraction line of rutile phase with gradual increase its
intensity and simultaneous diminishing of the d1 0 1 diffraction line of anatase phase from X-ray diffractogram. On
heat-treatment of the precursors at 750 8C/2 h, the percentage of the rutile phase in Ni0.1W0.1Ti0.8O2 was found to be
around 95%, and the pure rutile phase was obtained at 800 8C/2 h (Table 1). The XRD peaks for the solid solution
composition were indexed using the standard data available for the rutile phase (JCPDS card no. 781510). The lattice
parameters (i.e., a and c) for the rutile phase Ni0.1W0.1Ti0.8O2 solid solution, heat-treated at 800 8C/2 h were calculated
˚
˚
following the reflection planes (2 0 0) and (0 0 2) and their values: a = b = 4.602 A and c = 2.962 A; the cell volume
3
62.74 A . The calculated values of a and c were found to be in close agreement with those reported for the standard
˚
rutile phased TiO2 (JCPDS data card no.781510). As there was no evidence of any additional phase in the samples at
Table 1
Percentage of rutile phase, relative XRD intensities of the anatase and rutile phases and tap densities of the Ni0.1W0.1Ti0.8O2 composition at different
calcination temperatures
Calcination temperature
Rutile phase (%) for
compositions
XRD intensities of the anatase and
rutile phases for compositionsa
Tap densities
(g/cm3)
600 8C/2 h
650 8C/2 h
750 8C/2 h
800 8C/2 h
4.0
12.0
A (s), R (w)
A (s), R (w)
R (s), A (vw)
R (s)
0.1367
0.1732
0.1855
0.2387
95.0
100.0
a
Relative XRD intensities of the anatase (A) and rutile (R) phases; s, strong; w, weak; mw, medium weak; vw, very weak.

670
S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
Fig. 2. X-ray diffractograms (using Cu Ka radiation) of (a) Ni0.1W0.1Ti0.8O2 and (b) BaNiTi7O16 precursor on calcination at various temperatures.
heat-treatment temperature 800 8C/2 h, and also because the unit-cell parameters fairly matched with the chemical
composition, therefore the structure of Ni0.1W0.1Ti0.8O2 solid solution (at heat-treatment temperature 800 8C) was
confirmed to be of purely rutile type.
Fig. 2b shows the characteristic X-ray powder diffraction (XRD) patterns for the BaNiTi7O16 precursor powder at
various heat-treatment temperatures between 600 8C and 850 8C for 2 h. The virgin precursor was amorphous to X-ray
below 600 8C and initiation of phase formation of this titanate was found from XRD of the powder at 600 8C. The
greenish yellow coloured, pure priderite phase with the tetragonal structure of the compound was realized at 850 8C.
The XRD peaks for this solid solution were indexed using the standard data available for the priderite phase (JCPDS
card no. 741452). The calculated lattice parameters (i.e., a and c) for the BaNiTi7O16 solid solution, heat-treated at
˚
˚
850 8C/2 h following the reflection planes (2 0 0) and (0 0 2) were a = b = 10.074 A and c = 2.958 A; the cell volume
3
300.19 A . The values of a and c were found to be in close agreement with those reported for the standard priderite
˚
structure (JCPDS data card no. 741452). At heat-treatment temperature 850 8C/2 h, BaNiTi7O16 solid solution was
therefore confirmed to be of purely priderite type because of no evidence of any additional phase in the samples, and
also the unit-cell parameters matched with the reported data for priderite structure.

S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
671
The calculated crystallite size obtained from XRD studies for both the oxides were found to be in the range of 25–
30 nm. The EDAX analysis of both samples also showed an acceptable chemical homogeneity, which in these cases
were very close to the theoretical compositions formulated.
Finer microstructural information of the samples was obtained using transmission electron microscopy (TEM).
Fig. 3a shows a TEM image of Ni0.1W0.1Ti0.8O2 composition heat treated at 800 8C for 2 h. The smallest visible
particles in the bright field electron micrograph can be identified as single crystallites and/or their aggregates in the
Fig. 3. (a–c) Bright field TEM micrograph of the calcined (800 8C/2 h) Ni0.1W0.1Ti0.8O2 solid solution. (d) HRTEM image of the calcined
(800 8C/2 h) Ni0.1W0.1Ti0.8O2 solid solution. (e) Selected-area electron diffraction (SAED) pattern of the calcined (800 8C/2 h) Ni0.1W0.1Ti0.8O2
showing distinct rings corresponding to the rutile lattice.

672
S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
figure. The average diameters of the smallest visible isolated particle/crystallite agglomerate were found to range
between 30 nm and 60 nm. The bright field micrographs represented in Fig. 3b and c depict two sections of Fig. 3a taken
at higher resolutions (at 40k and 100k, respectively). Fig. 3d represents a typical HRTEM image (with a resolution 800k)
of the sample, depicting the lattice fringes in the crystal structure. From Fig. 3d, the lattice spacing was found to be
˚
ꢂ1.98 A, which corresponded to the (2 0 0) plane in the rutile lattice when the crystals were oriented along the (2 0 0)
plane. The selected-area electron diffraction (SAED) pattern of the samples is shown in Fig. 3e. The characteristics
SAED pattern of the heat-treated (800 8C/2 h) powder showed distinct rings corresponding to the (1 0 1), (2 0 0), (2 2 0)
planes in a rutile lattice, thus inferring the powders to be rutile structured nanosized polycrystalline.
Fig. 4a shows a bright field TEM image of BaNiTi7O16 solid solution composition heat-treated of the respective
precursors at 850 8C for 2 h. The particles were mostly spherical with average particle diameters ranging between
25 nm and 45 nm. Characteristics selected-area electron diffraction (SAED) pattern of BaNiTi7O16 powder, heat-
treated at 850 8C/2 h (Fig. 4b) showed the powders to be polycrystalline.
The BET surface areas of both the compositions were found to be high and they were in the range of 100–120 m2/g.
The tap densities of the Ni0.1W0.1Ti0.8O2 composition were found to increase with the increase in heat treatment
temperatures of the precursor powders and the corresponding values are summarized in Table 1.
Fig. 4. (a) Bright field TEM micrograph of the calcined (850 8C/2 h) BaNiTi7O16 solid solution. (b) Selected-area electron diffraction (SAED)
pattern of the calcined (850 8C/2 h) BaNiTi7O16 with a priderite structure.

S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
673
Fig. 5. UV–vis-NIR absorption spectra of the precursor powders of the solid solution composition: (1) Ni0.1W0.1Ti0.8O2 after the calcination at
800 8C/2 h and (2) BaNiTi7O16 after the calcination at 850 8C/2 h.
The UV–vis-NIR absorption spectra for the calcined Ni0.1W0.1Ti0.8O2 (at 800 8C/2 h) and BaNiTi7O16
(at 850 8C/2 h) solid solutions are shown in Fig. 5(1) and (2), respectively. Both the samples showed three distinct
absorption bands between 250 nm and 1200 nm corresponding to the three spin allowed d–d transitions from the Ni2+
ions in octahedral environment. For the Ni0.1W0.1Ti0.8O2 composition, the three distinct humps are positioned in
3
3
the ranges: 350–570 nm, 600–950 nm, and 970–1100 nm. These bands were assigned as: A2g(F) ! T1g(P)
3
3
3
3
(350–570 nm), A2g(F) ! T1g(F) (600–950 nm) and A2g(F) ! T2g(F) (970–1100 nm) [13]. Similarly, the three
3
3
spin allowed d–d transitions for the BaNiTi7O16 composition are assigned as: A2g(F) ! T1g(P) (275–550 nm),
3
3
3A2g(F) ! T1g(F) (570–950 nm) and 3A2g(F) ! T2g(F) (950–1100 nm) [13]. For both the compositions the energies
of the last two electronic spin allowed transitions exist in the visible light energy range and hence they contribute
towards the colour of the samples while the first transition lies is in the ultraviolet range and therefore does not affect
the colour of the samples. Besides these three spin allowed d–d transitions from the Ni(II) ion in octahedral
environment there may be additional absorption bands in both the solid solutions (i.e., Ni0.1W0.1Ti0.8O2 and
BaNiTi7O16) owing to charge transfers between the constituent metal ions [Ni(II), Ti(IV), W(VI)], and between ligand
and metal ions. The absorption bands due to Ti4+ $ O2ꢀ charge transfer in undoped rutile are very close to the
3
absorption bands corresponding to 3A2g(F) ! T1g(P) transitions in the solid solution compositions. The absorption
bands corresponding to metal–metal and metal–ligand charge transfers are therefore difficult to distinguish in either
of the solid solution compositions as they are expected to overlap with the spin allowed d–d transitions of the Ni2+ in
the UV–vis spectral range. The colour of the rutile phased titanate pigments are generally attributed to crystal
field transitions [14] along with some contribution from the Ti–O metal–ligand charge transfer (MLCT) [4], the band
gap [4] or, defects in crystals created by the varying sizes of the constituent metal ions (e.g. chromophores and
counterions) [5].
The CIE L*a*b* measurements of Ni0.1W0.1Ti0.8O2 and BaNiTi7O16 solid solutions heat treated at their calcination
temperatures summarized in Table 2 represent a yellow colour of both the samples (i.e., high b* and low a* parameters)
and a light colour (L* by ꢂ83–91) in agreement to the UV–vis-NIR spectra of the samples. The BaNiTi7O16
composition shows higher b* but lower a* parameter (more yellow intensity but with greenish shades) and lightness
while Ni0.1W0.1Ti0.8O2 has some reddish shade. CIE L*a*b* colorimetric parameters for both the samples after heat
treatment at 1250 8C/1 h are also given in Table 2. Both the samples exhibit slight decrease in L* values indicating the
decrease in lightness at 1250 8C/1 h. On the other hand increase in b* values of two solid solutions represent the
enhancement of yellow amount and decrease in a* value indicates the development of slight green amount in the solid
oxides.
The novelty of the reported chemical route for the preparation of Ni0.1W0.1Ti0.8O2 and BaNiTi7O16 lays in the
introduction of dimethyl ammonium tungstate and titanium oxalate as respective stable water-soluble sources of

674
S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
Table 2
The colour parameters (CIE L*a*b*) for Ni0.1W0.1Ti0.8O2 and BaNiTi7O16 solid solution compositions at different temperatures
Pigment composition
Ni0.1W0.1Ti0.8O2
Heat treatment temperature/duration
L*
a*
b*
800 8C/2 h (phase formation temperature)
1250 8C/1 h
87
83
ꢀ5
58
62
ꢀ6
BaNiTi7O16
850 8C/2 h (phase formation temperature)
1250 8C/1 h
91
89
ꢀ11
67
71
ꢀ13
tungsten and titanium, and in the use of TEA as a competent complexing agent for the generation of nanosized
particles at reduced external heat treatment temperatures. TEA is a tetradentate chelating ligand that can readily form
complexes with a number of metal ions in the aqueous medium thereby ensures their homogeneously dispersion in the
precursor solution through formation of stable, water-soluble coordinated complexes with TEA. The chelation of
metal ions with TEA in the precursor solution is dependent on the solution parameters, e.g. pH, concentrations of TEA
to the metal ions, and temperature. Maintaining the amount of TEA in excess to the stoichiometric requirement in the
precursor solution probably leads to polymerization of the free amount [11], which provides a highly branched
polymeric framework for anchoring the metal ions/metal coordinated complexes. This not only prevents segregation/
precipitation of the metal ions from the homogeneous precursor solution during evaporation but also provides a
voluminous, fluffy carbonaceous matrix in which metal ions embed on complete evaporation of the precursor solution.
Slow heat-treatment of the carbonaceous residue through calcination in dynamic air, results in the nano-particles of the
respective mixed-oxide and the heat generated through combustion of the carbonaceous mass facilitates the reduction
in the external temperatures required for the formation of desired oxide phase. The entire oxidative decomposition
process develops a large amount of gases (such as water vapor, oxides of carbon and nitrogen), which helps the
precursor material to dissipate the heat of combustion and thus inhibits sintering of the fine particles during the
process.
4. Conclusions
Nanosized rutile structured Ni0.1W0.1Ti0.8O2 and priderite structured BaNiTi7O16 yellow pigments were prepared
through pyrolysis of precursor solution. The preparation method establishes the use of dimethyl ammonium tungstate
and titanium oxalate as respective stable water-soluble sources of tungsten and titanium, and TEA as a competent
complexing agent. Pure rutile phase and priderite phase titanates were realized through heat-treatment of the precursor
powders at 800 8C and 850 8C, respectively. The crystallite sizes, average particle sizes and specific surface areas for
the solid solution compositions were found to be in the range between 25 nm and 30 nm, 25–45 nm and 100–120 m2/g,
respectively. The UV–vis-NIR spectra for both solid solutions exhibit three absorption bands due to d–d transitions of
d-electrons of Ni2+ ions. The measurement of the colour parameters (CIE L*a*b*) reflected an intense yellow colour
with greenish shades for the BaNiTi7O16 composition, while the Ni0.1W0.1Ti0.8O2 sample exhibited yellow colouration
with reddish shades.
Acknowledgement
The authors express thanks to Dr. P.V. Satyam, Institute of Physics, Bhubaneswar for HRTEM experiment and the
Department of Science and Technology, Government of India, for providing the financial support for the research
work.
References
[1] R.M. Cornell, U. Schwertmann, The Iron Oxides; Structure, Properties, Reactions, Occurrence and Uses, VCH, 1996.
[2] J.A. Badenes, M. Llusar, M.A. Tena, J. Calbo, G. Monro’s, J. Eur. Ceram. Soc. 22 (2002) 1981.
´
[3] S. Sorly, M.A. Tena, J.A. Badenes, J. Calbo, M. Llusar, G. Monro’s, J. Eur. Ceram. Soc. 24 (2004) 2425.
[4] J. Maloney, in: H.M. Smith (Ed.), High Performance Pigments, Wiley-VCH, 2002, pp. 53–73.
[5] M. Dondi, G. Cruciani, G. Guarini, F. Matteucci, M. Raimondo, Ceram. Int. 32 (4) (2006) 393.
[6] F. Bondioli, T. Manfredini, G.C. Pellacani, Interceram 48 (6) (1999) 414.

S.K. Biswas et al. / Materials Research Bulletin 43 (2008) 665–675
675
[7] Italian Ceramic Society, Colour Pigments and Colouring in Ceramics, SALA, Modena, 2003, pp. 1–129.
[8] S. Sorlg’, M.A. Tena, J.A. Badenes, M. Llusar, G. Monro’s, Br. Ceram. Trans. 103 (1) (2004) 10.
[9] F. Matteucci, G. Cruciani, M. Dondi, M. Raimondo, Ceram. Int. 32 (4) (2006) 385.
[10] M.L. Carter, Mater. Res. Bull. 39 (2004) 1075.
[11] A. Pathak, S. Mohapatra, S. Mohapatra, S.K. Biswas, D. Dhak, N.K. Pramanik, A. Tarafdar, P. Pramanik, Am. Ceram. Soc. Bull. 83 (8) (2004)
9301.
[12] R.A. Spurr, H. Myers, Anal. Chem. 29 (1957) 760.
[13] A.B.P. Lever, Inorganic Electron Spectroscopy, Elsevier, Amsterdam, 1968.
[14] R.A. Eppler, J. Am. Ceram. Soc. 70 (4) (1987) C64–C66.