with High Specific Surface Area
Min Ku Lee
*, Young Rang Uhm, Chang Kyu Rhee and Yong Bok Lee
Nuclear Materials Research Division, Korea Atomic Energy Research Institute, P.O. Box 105, Yuseong, Daejeon 305-600, Korea
Sedimentation properties of the TiO2nanoparticles with very high specific surface area (about 180 m2/g) based on the BET method have been investigated in various organic solvents. The TiO2 nanocolloid in the isopropyl alcohol displayed a considerably high electrostatic repulsive force, compared to that in other solvents, with negligible coalescence between the particles. Both the backscattered light flux measurements and scanning electron microscopy confirmed that in the cases of water, methyl alcohol and ethyl alcohol, a progressive sedimentation of the TiO2particles was observed at the bottom due to a flocculation-induced particle growth, while a very stable TiO2dispersion was observed in isopropyl alcohol. [doi:10.2320/matertrans.M2010245]
(Received July 21, 2010; Accepted October 7, 2010; Published November 25, 2010)
Keywords: TiO2nanoparticles, specific surface area, dispersion stability, zeta potential, sedimentation, organic solvent
1. Introduction
Titanium dioxide (TiO2) has been widely used for white
pigment materials due to its good scattering ability from ultraviolet light. It has also been used for optical coatings, beam splitters, and anti-reflection coatings owing to its high dielectric constant, refractive index, good oil adsorption ability, tinting power, and chemical stability, even under strongly acidic or basic conditions.1–3) Especially, the potential applications of the TiO2 particles with well
dispersion property in organic solvents include solar energy conversion, degradation of organic pollutions, photo-chemical disinfection, manufacture of sensors, and paper-making.4–6)For its various applications the incorporation of
powders with a liquid medium, i.e., colloidal powder process, is required, since it provides a promising route not only to produce the fine materials optimally but also to produce a high-performance ceramic powder.
In recent years, study and preparation of stabilized colloidal nanoparticles have attracted considerable attention due to both their outstanding physical properties and potential application.7)Needless to say, a precise control of the shape or morphology of ceramic oxides at a nanometric scale are expected to offer an opportunity for both a tuning of the property of colloidal particle system and a fundamental scientific understanding. So far, many efforts have been paid to studies dealing with the dispersions and colloidal stability of the ceramic oxides,8) but relatively little work has been
carried out on dispersion and destabilization properties of the organic colloid nanoparticles, especially nanostructured TiO2 particles with a very high surface area.
In this study, colloidal dispersion and sedimentation properties of TiO2 nanoparticles with a unique structure
consisting of ultrafine aciculae (37nm thick) and substan-tially very high specific surface area (180m2/g) have been investigated in pure aqueous and various organic solvents including methyl alcohol, ethyl alcohol and isopropyl alcohol. The study was carried out by means of a zeta
potential measurement, which is often used to characterize the electric field around a colloid particle, and a back-scattered light flux analysis to characterize the sedimentation and destabilization behaviors.
2. Experimental Details
Transparent titanium tetrachloride (TiCl4, 3N, Aldrich,
Co.) was used as a starting material to fabricate the ultrafine TiO2 powder by using the homogeneous precipitation
process at low temperatures. The procedure involving a synthesis of crystalline rutile TiO2 powder from a stable
aqueous TiOCl2 solution has been described. Aqueous
TiOCl2 solution was made by adding ice pieces slowly to
TiCl4 while suppressing the vigorous hydrolysis reaction.
Adding the distilled water to the aqueous TiOCl2 solution
with 0.67 M of Ti4þ, crystalline TiO
2 ultrafine powder was
prepared by homogeneous precipitation with only heating the solution at temperatures of 17C for 4 h, followed by drying the precipitates at 50C for 12 h, after washing them thoroughly.9,10)
For a comparison, the commercially available P-25 (Degussa Co., Germany) TiO2 nanopowders were also
introduced. Pure organic solvents varying in both their viscosity and dielectric constants were chosen as a suspend-ing medium: deionized water, methyl alcohol (99.8%, Merck, Germany), ethyl alcohol (95%, Jin Chemical Co. Ltd., Korea), and isopropyl alcohol (99.5%, Showa Chemical Co. Ltd., Japan). Their viscosity and dielectric constants are summarized in Table 1. The electrolytes including CsCl (99%, Showa Chemical Co. Ltd., Japan), CaCl2(90%, Showa
Chemical Co. Ltd., Japan), FeCl3 (97%, Showa Chemical
Co. Ltd., Japan), and ZrCl4 (98%, Merck, Germany) were
also used to study the effect of a surface charge on TiO2
nanoparticles in a dilute condition below101M.
TiO2 dispersion was prepared by introducing 0.5 g of
TiO2 powder into a 100 ml solvent media, followed by an
ultrasonic treatment for 5 min. The zeta potentials on the particle surface were also measured to predict the suspension stability by using a zeta potential analyzer (BIC-90 Plus,
Brookhaven Instruments Co., USA). The dispersion strength of the particles is determined by the particle charge and the ionic strength of the suspension, as described by DLVO (Derjaquin-Landau-Verwey-Overbeek) theory.11) The
elec-trostatic potential for the surface of the particles consists of permittivity of the vacuum, relative permittivity of the suspension, and zeta potential, etc.11)A Hu¨ckel relationship
was introduced to estimate the zeta potential,, as follows: e¼ ð2"r"o=3Þ (e: electrophoretic mobility of particle,
"r: relative permittivity, "o: permittivity of free space, :
viscosity of liquid).12)A back scattered light flux measure-ment was analyzed to investigate the destabilization phe-nomena of the TiO2dispersion by using a Turbiscan MA Lab
optical analyzer (Formulaction, France). The principle of the measurement was based on a variation of the droplet volume fraction (migration) or diameter (coalescence), resulting in a variation of the backscattering and transmission signals.
3. Results and Discussion
The TiO2 nanoparticles used in this study were purely
rutile and mono-dispersed with a mean spherical diameter of 300 nm (Fig. 1(a)). Although their particle size was relatively large compared to the commercial ones with the mean diameter of2050nm (Fig. 1(b)), their specific surface area was as high as about 180 m2/g based on the BET method using N2adsorption, which was more than three times higher
than those of the commercial ones (5060m2/g). The observed high specific surface area was totally attributed to a
formation of a unique ‘‘chestnut bur’’ shaped nanostructure comprising of ultrafine acicular or needle-shaped with a thickness of 37nm. Detailed preparation and particulate properties for those unique TiO2 nanoparticles were given
elsewhere.9) Titania particles could be obtained at around
room temperature by heating and stirring an aqueous TiOCl2
solution, which was prepared by slowly adding distilled water ice pieces to TiCl4 below 0C. The low treatment
temperature makes it possible to produce powders with higher specific surface area and soft particle agglomeration, which will strongly affect the resulting physico-chemical behaviors.9,13,14)
As seen in Fig. 2(a), the zeta potential measurements for the TiO2 nanopowders with those unique morphological
features have shown that a broader stability of zeta potential was observed over the investigated range of the pH, which was quite distinct from the commercial P-25 TiO2powders.
The TiO2particles also generally exhibited a transition from
positive zeta potential at a low pH to negative zeta potential at a high pH. Positive sign of the zeta potential is due to the adsorption of the Hþions onto the surface, while at high pH regions Hþis released out of the surface to induce a negative sign of the zeta potential. For the TiO2nanoparticles used in
this study, the isoelectric point (IEP), i.e., the pH at which the zeta potential vanishes to zero, was found at about 2.31, which was remarkably low, when compared to that of the commercial P-25 powders (5.88) and those reported in a previous investigation (6:26:5 for rutile TiO2).15,16) The
IEP of oxide powders are determined by the potential determining ions (Hþ and OH ions in aqueous solvent) present on a surface at equilibrium. It is also known that the finer the oxide particle, the lower the IEP due to the easier desorption of the Hþ ions out of the surface.12)When
the particle size is decreased, the specific surface area is increased. As all, Hþ adsorption on the surface of the particles with high specific surface area is significantly increased. A remarkable shift of the IEP towards the acidic regions for the TiO2 nanoparticles used in this study
[image:2.595.128.469.74.246.2]originates from a marked increase in the surface area for the Hþadsorption due to the formation of very fine primary aciculae with the thickness of37nm.17)
Table 1 Viscosity and dielectric constants of the solvent media (20C).
Suspending media
Molecular form
Viscosity,
(mPas)
Dielectric
constant," ="
Water H2O 1 80.37 0.0124
Methyl
alcohol CH3OH 0.55 33.64 0.0163 Ethyl alcohol C2H5OH 1.195 25.07 0.0477
Isopropyl
alcohol C3H7OH 2.86 18.3
0.156
: 25C
(b)
(a)
[image:2.595.48.290.314.410.2]Figure 2(b) shows the variation of the zeta potential of the TiO2nanoparticles prepared in this study as a function of the
FeCl3 electrolyte concentration in various solvent media
including water, methyl alcohol, ethyl alcohol, and isopropyl alcohol, where the pH was kept at around 67 initially belonging to the negative zeta potential regime (see Fig. 2(a)). As the concentration of the FeCl3 electrolyte
increases, typically the measured zeta potentials belong to a negative regime in a very dilute condition and then go through a sign reversal from a negative to a positive, followed by a saturation or slight decrease for various solvent media. In the absence of electrolytes the corre-sponding zeta potentials were 26:05mV, 26:11mV,
closely related to the physical properties of the solvent, implying that the higher the =" value of the medium, the larger the absolute zeta potential.
From Fig. 2(b) the critical coagulation concentrations (CCCs), i.e., the concentration of the electrolyte required to reverse the zeta potential, were also estimated for all the solvent media, which were 4:81105M, 3:78
106M,2:51105M, and8:731010M, for the water,
methyl alcohol, ethyl alcohol, and isopropyl alcohol, respec-tively. It is found that the higher the ="of the dispersion media, the smaller the CCC, which is due to a more significant charge neutralization at an oxide-liquid interface. The maximum positive zeta potentials as a function of the equivalent concentration for all the added electrolytes including CsCl, CaCl2, FeCl3, and ZrCl4are also determined
and the results are summarized in Fig. 2(c). It is worthwhile to note that the broader stability of the maximum zeta potential is seen in isopropyl alcohol dispersions compared to other dispersion media. When the electrolyte concentration is very high, the maximum positive zeta potential decreases due to the effect of charge neutralization as a result of the adsorption of Cl or OH. When the concentration of the electrolyte is increased, the pH is decreased. According to the results in Fig. 2(b), the influence of the anion (Cl) of electrolyte on charge reversal of TiO2 is hardly seen and is
likely to be negligible. This tendency becomes more evident for the water, methyl alcohol, and ethyl alcohol with the smaller="value.
To investigate the destabilization phenomena of the unique TiO2 nanoparticles in various solvents, the
time-dependent sedimentation behavior was investigated by the backscattering profile measurements using a Turbiscan. Figure 3(a) to 3(d) shows the backscattering profiles taken every 2 h for 24 h when the suspending media are water, methyl alcohol, ethyl alcohol, isopropyl alcohol, respectively. It was found that the backscattering intensity decreased at the sample top due to a clarification and increased at the sample bottom due to a sedimentation, as seen in Fig. 3(a) to 3(c). A progressive fall of the back-scattering signal, which was observed as a function of the time in the middle region of the sample, can be explained by a flocculation-induced particle growth. On the contrary, a very stable TiO2 dispersion was observed in isopropyl
alcohol without showing any clarification or sedimentation, as seen in Fig. 3(d). This implies that a flocculation due to a coalescing reaction between the TiO2 particles hardly
occurred in the isopropyl alcohol.
From Fig. 3(a) to 3(d), variations of the backscattering intensity as a function of time were determined and the results are presented in Fig. 4(a). The estimated backscattering rates or slopes were about 0:98%/h,
1:20%/h,1:01%/h, and0:20%/h for the water, methyl alcohol, ethyl alcohol, and isopropyl alcohol, respectively. Figure 4(b) shows the characteristics of the migration velocity of the TiO2 particles as determined from Fig. 3(a)
to 3(d). The migration velocity of the TiO2 particle
were measured at about 6.15mm/min, 12.53mm/min,
2 -45 -30 -15 0 15
*Commercial TiO2
**TiO2 in this study
Zeta potential (mV)
pH in water
10-11
10-10
10-9
10-8
10-7
10-6
10-5
10-4
10-3
10-2
10-1
-45 -30 -15 0 15 30 45
Water Methyl
alcohol
Ethyl alcohol Isopropyl
alcohol
Zeta potential (mV)
FeCl3 concentration (M) (b)
0 10-10
10-8
10-6
10-4
10-2
0 10 20 30 40 50 60
water methyl alcohol ethyl alcohol isopropyl alcohol
Maximum positive zeta potential (mV)
Equivalent concentration (M)
(c)
10 8
6 4
Fig. 2 Variation of the zeta potential of the TiO2nanoparticles for the TiO2 dispersions in various solvents as a function of (a) the pH in water (: P-25, : TiO
[image:3.595.76.265.67.545.2]6.51mm/min, and 0.18mm/min for the water, methyl alcohol, ethyl alcohol, and isopropyl alcohol, respectively, showing a remarkably slow migration of the TiO2 particles
in the isopropyl alcohol. The above zeta potential properties demonstrated those variations of the dispersion stabilities of the TiO2 particles with the various solvents. For the
organic solvents, both the measured migration velocity and zeta potentials were found to be correlated well, that is, the migration velocity decreases as the absolute value of the zeta potential increases. It is concluded that the higher the
=" value of the organic solvent, the larger the absolute
zeta potential and the smaller the migration velocity. In order to investigate a degree of flocculation of the TiO2
nanoparticles in the corresponding solvents, the powders taken from the test solution after 24 h were analyzed by SEM, as shown in Fig. 5. It was found that a coalescence or agglomeration between the particles was hardly observed in the isopropyl alcohol as shown in Fig. 5(d), while a number of the aggregated TiO2particles were observed in the water,
methyl alcohol, and ethyl alcohol solvents (Fig. 5(a), 5(b), and 5(c)). These results are direct evidence for demonstrating a stable dispersion of the TiO2 nanoparticles in isopropyl
alcohol, showing a good correlation to the measured zeta potentials.
4. Conclusion
In summary, the colloidal stability and sedimentation behavior of the TiO2 nanoparticles consisting of 37nm
thick ultrafine aciculae (specific surface area of 180 m2/g)
0 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40
Delta Backscattering (%)
Height (mm) (a) Clarification t=2hr t=24hr t=0hr 0 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 t=2hr t=24hr
Delta Backscattering (%)
Height (mm) (b)
Particle size increase t=0hr 0 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 t=0hr t=2hr t=24hr
Delta Backscattering (%)
Height (mm) (c) 0 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 t=0hr t=24hr t=2hr
Delta Backscattering (%)
Height (mm) (d) 50 40 30 20
10 10 20 30 40 50
50 40 30 20 10 50 40 30 20 10
Fig. 3 Backscattering profiles as a function of the time for the dispersions of the TiO2powders at 25C: (a) water, (b) methyl alcohol, (c) ethyl alcohol, and (d) isopropyl alcohol.
0 -30 -25 -20 -15 -10 -5 0 (a)
Back scattering (%)
Time (hr) water methyl alcohol ethyl alcohol isopropyl alcohol 25 20 15 10 5 0 0 5 10 15 20 25 30 water methyl alcohol ethyl alcohol isopropyl alcohol (b)
Migration velociry (
µ m/min) Time(hr) 25 20 15 10 5
[image:4.595.108.488.72.380.2] [image:4.595.75.264.428.745.2]have been investigated in pure aqueous and various organic solvents including methyl alcohol, ethyl alcohol and iso-propyl alcohol. The zeta potential behavior of those unique TiO2 nanoparticles was quite different from the commercial
spherical TiO2 nanoparticles, showing a broader stability
over the investigated range of the pH with a remarkably low IEP of 2.31. For all the solvents, addition of FeCl3electrolyte
leads to a charge reversal from a negative to a positive on the surface of the TiO2 particle. In the absence of electrolytes
the measured zeta potentials were26:05mV, 26:11mV,
28:66mV, and 45:37mV for the cases of the water,
methyl alcohol, ethyl alcohol, and isopropyl alcohol dis-persions, respectively. The backscattered light flux measure-ments revealed that in the case of the water, methyl alcohol and ethyl alcohol, a progressive sedimentation of the TiO2
particle was observed at the bottom due to a flocculation-induced particle growth, while a very stable TiO2dispersion
was observed in isopropyl alcohol without showing any clarification or sedimentation. Both the measured back-scattering rate and zeta potentials were correlated well each other.
Acknowledgements
This work was supported by the R&D program under the Ministry of Education Science & Technology (MEST), Republic of Korea.
REFERENCES
1) K. L. Siefering and G. L. Griffin: J. Elecrochem. Soc.137(1990) 814– 818.
2) A. Bally, K. Prasad, R. Sanjines, P. E. Schmid, F. Levy, J. Benoit, C. Barthou and P. Benalloul: Mat. Res. Soc. Symp.424(1997) 471–475. 3) T. Fuyuki and H. Matsunami: Jpn. J. Appl. Phys.25(1986) 1288–1291. 4) X.-Q. Chen and W.-H. Shen: Chem. Eng. Technol.31(2008) 1277–
1281.
5) R. L. Ziolli and W. F. Jardim: J. Photochem. Photobiol. A: Chemistry
147(2002) 205–212.
6) R. Pelton, X. Geng and M. Brook: Adv. Colloid Interf. Sci.127(2006) 43–53.
7) F. Bonet, V. Delmas, S. Grugeon, R. Herrera Urbina, P.-Y. Silvert and K. Tekaia-Elhsissen: Nanostruct. Mater.11(1999) 1277–1284. 8) A. C. van Dyk and A. M. Heyns: J. Colloids Interf. Sci.206(1998)
381–391.
9) S. J. Kim, S. D. Park, Y. H. Jeong and S. Park: J. Am. Ceram. Soc.82
(1999) 927–932.
10) S. J. Kim, S. D. Park, K. H. Kim, Y. H. Jeong and I. H. Kuk: U.S. Patent No. 6001326.
11) B. Schafer, M. Hecht, J. Harting and H. Nirschl: J. Colloids Interf. Sci.
349(2010) 186–195.
12) D. J. Shaw:Introduction to Colloid and Surface Chemistry, (Butter-worth, London, 1997) p. 157.
13) S. Yin, R. Li, Q. He and T. Sato: Mater. Chem. Phys.75(2002) 76–80. 14) S. S. Park, Y. H. Cho, W. W. Kim and S.-J. Kim: J. Solid State Chem.
146(1999) 230–238.
15) M. Pattanaik and S. K. Bhaumik: Mater. Lett.44(2000) 352–360. 16) M. J. Jaycock, J. L. Pearson, R. Counter and F. W. Husband: J. Appl.
Chem. Biotechnol.26(1976) 370–374.
17) S. J. Kim, E. G. Lee, S. D. Park, C. J. Jeon, Y. H. Cho, C. K. Rhee and W. W. Kim: J. Sol-Gel Sci. Technol.22(2001) 63–74.
500nm 500nm
(c)
500nm
(d)
500nm
[image:5.595.117.483.73.355.2]