Controlled Preparation and Mechanism Study of Zirconia-Coated

pubs.acs.org/Langmuir © 2009 American Chemical Society

Controlled Preparation and Mechanism Study of Zirconia-Coated Hematite Particles by Hydrolysis of Zirconium Sulfates Feng Zhao,†,‡ Yanfeng Gao,*,† and Hongjie Luo† †

Research Center for Industrial Ceramics, Shanghai Institute of Ceramics (SIC), Chinese Academy of Sciences (CAS), Dingxi 1295, Changning, Shanghai 200050, China, and ‡Graduate School of Chinese Academy of Sciences, Yuquanlu 19, Beijing 100049, China Received January 20, 2009. Revised Manuscript Received February 26, 2009

Zirconia-precursor-coated hematite particles were prepared by hydrolysis of zirconium sulfate in aqueous solution. The as-prepared zirconia-precursor shell was amorphous with a thickness of about several ∼30 nm that can be controllably achieved by varying the processing parameters and had a composition of Zr2(OH)6SO4, which crystallized to tetragonal ZrO2 after annealing at 700 °C. The focus of this work is to investigate in detail the process and to understand the key issues for surface coating in solution. The thermodynamic analysis on hydrolysis of zirconium sulfate was conducted, and a “surface-deposition region” for zirconia coating was suggested in this work. Furthermore, the kinetic study of the process was also described. The hydrolysis could be considered as a pseudo-second-order reaction at 50 °C, and the rate constant was calculated to be 0.61 L mol-1 s-1. The hydrolysis mechanism of zirconium salt was also interpreted from the viewpoint of structural chemistry. The influence of the surfactants on the coating process was also discussed.

Introduction In the past few decades, the preparation of core-shell-structured particles has attracted great attention because of the ever-increasing demands upon materials synthesis by performance design. The as-prepared composite particles exhibit a nice performance that greatly extends the application of the singlecomponent core materials at many aspects, such as sensors,1 cosmetics,2 pharmaceutics,3 pigments,4 etc. These processes termed particle engineering5,6 can fabricate core-shell colloids with tailored microstructures and surface properties. Among all of the strategies to coat an inorganic oxide shell, the :: well-known Stober method makes silica the most common shell material because it provides a general route for heterogeneously coating with surfactants.7 Meanwhile, more efforts have been devoted to zirconia-coated particles because of the unique properties of zirconia, which are not commutable. The products with a zirconia shell can be applied to affinity probes,8 ionic conductors,9 catalysts,10 and other aspects to inhibit the crystal growth11 or *To whom correspondence should be addressed. Telephone/Fax: +86-215241-5270. E-mail: [email protected]. (1) Chavez, J. L.; Wong, J. L.; Duran, R. S. Langmuir 2008, 24, 2064–2071. (2) Jaroenworaluck, A.; Sunsaneeyametha, W.; Kosachan, N.; Stevens, R. Surf. Interface Anal. 2006, 38, 473–477. (3) Riskin, M.; Basnar, B.; Huang, Y.; Willner, I. Adv. Mater. 2007, 19 2691–2695. (4) Lin, C. K.; Li, Y. Y.; Yu, M.; Yang, P. P.; Lin, J. Adv. Funct. Mater. 2007, 17, 1459–1465. (5) Caruso, F. Adv. Mater. 2001, 13, 11–22. (6) Katagiri, K.; Matsuda, A.; Caruso, F. Macromolecules 2006, 39, 8067–8074. (7) Graf, C.; Vossen, D. L. J.; Imhof, A.; van Blaaderen, A. Langmuir 2003, 19, 6693–6700. (8) Li, Y.; Leng, T. H.; Lin, H. Q.; Deng, C. H.; Xu, X. Q.; Yao, N.; Yang, P. Y.; Zhang, X. M. J. Proteome Res. 2007, 6, 4498–4510. (9) Omata, T.; Sasai, S.; Goto, Y.; Ueda, M.; Otsuka-Yao-Matsuo, S. J. Electrochem. Soc. 2006, 153, A2269–a2273. (10) Chen, H. R.; Gao, J. H.; Ruan, M. L.; Shi, J. L.; Yan, D. S. Microporous Mesoporous Mater. 2004, 76, 209–213. (11) Yu, J. C. J. Inorg. Mater. 2005, 20, 1054–1058. (12) Moon, J. W.; Lee, H. L.; Kim, J. D.; Kim, G. D.; Lee, D. A.; Lee, H. W. Mater. Lett. 1999, 38, 214–220. (13) Chen, D.; Liu, J. S.; Wang, P.; Zhang, L.; Ren, J.; Tang, F. Q.; Wu, W. Colloids Surf., A 2007, 302, 461–466.

6940 DOI: 10.1021/la900237m

sintering12 of core materials. Most of the reported processes8-11,13 for zirconia-coating selected alkoxides as the starting material for zirconium, because the hydrolysis rate of metal alkoxides was controllable in the process. As a contrast, inorganic zirconium salts are seldom employed to fabricate core-shell materials in aqueous solution, which might be because the hydrolysis of inorganic zirconium salts often occurs in an acidic solution,14 although they are much cheaper and easier to handle than alkoxides. Typically, in our previous report,15 a 2-3 nm thick shell was fabricated on the hematite particles by compelling hydrolysis of zirconyl chloride in aqueous solution. However, the solution was strongly acidic, which was too rigorous to deposit a thick shell. Matijevic and co-workers synthesized the monodispersed zirconia colloid particles,16 and zirconia-coated polystyrene17 and hematite particles18 with zirconium sulfate successfully. Agarwal et al.19 also deposited zirconia films on self-assembled monolayers by the hydrolysis of zirconium sulfate. The chemical reaction formula of zirconium sulfate hydrolysis in aqueous solution was proposed.16,19 These results indicated that zirconium sulfate was a better choice for coating zirconia in aqueous solution. However, the coating mechanism of zirconium sulfate in aqueous solution was still obscure, and a further study was needed. In the present study, a process using Zr(SO4)2 as the starting material was developed to coat hematite particles homogeneously with ZrO2 precursor. Hematite is a kind of red ceramic pigment that is easy to react with some components in glazes and ceramic matrix, and the hematite particles cannot be directly used in the enamel. Coating the hematite particle with zirconia shell is to (14) Matsui, K.; Ohgai, M. J. Am. Ceram. Soc. 1997, 80, 1949–1956. (15) Zhao, F.; Gao, Y.; Li, W.; Luo, H. J. Ceram. Soc. Jpn. 2008, 116 1164–1166. (16) Aiken, B.; Hsu, W. P.; Matijevic, E. J. Mater. Sci. 1990, 25, 1886–1894. (17) Kawahashi, N.; Persson, C.; Matijevic, E. J. Mater. Chem. 1991, 1 577–582. (18) Grag, A.; Matijevic, E. J. Colloid Interface Sci. 1988, 126, 243–250. (19) Agarwal, M.; De Guire, M. R.; Heuer, A. H. J. Am. Ceram. Soc. 1997, 80, 2967–2981.

Published on Web 04/14/2009

Langmuir 2009, 25(12), 6940–6946

Zhao et al.

Article

Figure 1. (a) TEM image of zirconia-coated hematite particles and (b) EDS spectrum of the shell prepared at 30 °C for 1 h in 2.5 mM zirconium sulfate solution.

prepare an inclusion pigment that can protect the pigment from reaction. However, the focus of the paper is not on the production of pigments but on the investigation of the coating process. The focus of this work is to investigate in detail the process and to understand the key issues for surface coating in solution. The hydrolysis behavior of zirconium sulfate in an aqueous solution was investigated in detail by changing the process parameters, including the zirconium concentration, the pH value of solution, the temperature, and time for hydrolysis. In addition, the hydrolysis kinetics of the zirconium sulfate and the mechanism of the process were also discussed. We hope that the processing factors for the preparation of oxides from solution onto core particles can be clarified, and the understanding could be extended to other oxide systems or coat other matrix particles.

Experimental Section Materials. All of the reagents were used as purchased without further purification. Because the zirconium sulfate solution became unstable and precipitated within 1 h even at room temperature, a fresh stock aqueous solution of 0.01 M zirconium sulfate (Zr(SO4)2 3 4H2O, 98%, Sinopharm) was made for each set of experiments. Hematite particles that are originally used as a red pigment are commercially available. In this work, different surfactants, such as hydroxypropyl cellulose (HPC, MW 1 000 000, Alfa Aesar), polyvinylpyrrolidone (PVP, K-30, Sinopharm), and sodium dodecyl benzene sulfonate (SDBS, AR, Sinopharm) were used as dispersants. Additionally, xylenol orange and hydroxylamine hydrochloride (AR, Sinopharm) were used to determine the zirconium concentration in the solution. Preparation. For a typical coating process, hematite particles were first dispersed in distilled water with 1 wt % HPC. To the above solution, a certain amount of 0.01 M zirconium sulfate was added. The amount of hematite in the solution was 0.4 g L-1, and the final concentration of obtained zirconium sulfate varied from 0.5 to 2.5 mM. After that, the dispersion was aged at a constant temperature (30-70 °C) for a certain time. The coated particles could be separated by centrifugation (4000 rpm) and washed with anhydrous ethanol. The amorphous zirconia shell deposited on the hematite particles was crystallized as tetragonal zirconia (t-ZrO2) by annealing at 700 °C. Characterization. The X-ray diffraction (XRD) patterns of the samples were determined by X-ray diffractometry (Model D/ Max 2550V, Rigaku). The morphology of the coated particles was characterized by a field emission scanning electron microscope (FE-SEM, JSM-6700F, JEOL). The structure and elements analysis of the coated powders were performed on a transmission electron microscopy (TEM) and an energy-dispersive spectrometer (EDS, JEM-2010F, JEOL), respectively. The concentration of the zirconium in the solution was determined by xylenol orange colorimetry, and the absorbance of the sample was measured Langmuir 2009, 25(12), 6940–6946

Figure 2. SEM image of the zirconia-coated particles prepared at 30 °C for 1 h in 2.5 mM zirconium sulfate solution. by a UV-vis-NIR spectrophotometer (U-4100, Hitachi, Japan). ζ-Potentials as a function of pH for hematite particles were measured with a zetaplus analyzer (Zetaplus, Brookhaven, Holtsville, NY).

Results and Discussion Characterization of Coated Particles. The scanning electron microscopy (SEM) and transmission electron microscopy (TEM) observation of hematite powders (see Figure S1 in the Supporting Information) showed that the original hematite particles were 100-200 nm in diameter and most of particles were irregular in shape and have multiple cores that could not be further separated via ultrasonication. Occasionally, monocore hematite particles were also observed. The TEM image of the coated particles synthesized by hydrolysis at 30 °C for 1 h is shown in Figure 1a. The starting concentration of zirconium sulfate was 2.5 mM. The image clearly demonstrates the outline of hematite particles and an about 20 nm thick zirconia shell outside. The EDS spectrum (Figure 1b) of the shell shows the chemical composition of the zirconia precursor. Approximate element analysis of the EDS indicated that the atomic ratio of Zr/S in the precursor was about 2. The zirconia-coated particles dispersed well, which was characterized by SEM in Figure 2. It is noteworthy that no acid or basic agents were introduced, and the hydrolysis temperature was near room temperature (30 °C). The hydrolysis conditions adopted were more moderate than that reported by Grag and Matijevic.18 DOI: 10.1021/la900237m

6941

Article

Figure 3. XRD patterns of (a) zirconia precursor and zirconiacoated hematite particles annealed at (b) 500 and (c) 700 °C for 2 h, respectively.

Zhao et al.

Figure 4. TEM image of zirconia-coated hematite particle after annealing at 700 °C for 2 h.

Figure 5. TEM images of zirconia-coated hematite particles with zirconium starting concentration of (a) 0.5 mM and (b) 1.5 mM, respectively. The white arrows point out the spontaneous precipitates of the zirconia precursor. The insets show coating detail information of the same samples in large magnification.

Figure 3 shows the XRD patterns of the zirconia precursor particles and zirconia-coated hematite particles after annealing at 500 and 700 °C, respectively. The patterns indicate that the zirconia precursor crystallized as tetragonal ZrO2 after annealing at 700 °C. This result is consistent with the report by Akein et al.16 The TEM image of the sample annealed at 700 °C is shown in Figure 4. In comparison to the as-prepared particles (Figure 1a), the shell after annealing contained a large amount of tiny crystallites. According to the characterizations above and related literatures,16,17,19 the zirconia precursor seems to have a chemical composition of Zr2(OH)6SO4. These results suggest that the zirconia precursor could be deposited on hematite particles easily only if the hydrolysis process is under strict control. Control of the Hydrolysis Process. In principle, the driving force for zirconia coating is that the heterogeneous nucleation on the surface of hematite particles has lower surface free energy from a thermodynamic viewpoint. According to the LaMer diagram,20 heterogeneous nucleation occurs at a relatively low supersaturation degree than that of homogeneous nucleation. At a very high supersaturation degree, the nucleation rate is high and the deposition process completes quickly. In this case, a large amount of homogeneously nucleated particles may form in solution. As a result, only a little deposit can be grown on core particles. Therefore, when the supersaturation degree is moderately high, homogeneous and heterogeneous nucleations occur

simultaneously to obtain a mixture colloid. Therefore, to control the thickness of the shell and to increase the purity of the products by suppressing the homogeneous formation of particles in solution, it is crucial to keep the supersaturation degree within a narrow “window” to restrain spontaneous precipitation of the zirconia precursor.21 From the experimental results, isolate precipitates of zirconia precursor could be easily found in the TEM images (see Figures 1a and 4). As the first step to suppress the homogeneous precipitation of zirconia, the effect of the concentration of starting materials was investigated. For this purpose, solutions with different starting concentrations, typically 0.5 and 1.5 mM zirconium sulfate solutions, were prepared by adding a certain amount of the 0.01 M ZrIV stock solution to the hematite dispersion without pH regulation. The hydrolysis time was 1 h at 30 °C for all experiments. As illustrated in the TEM images (Figure 5), the amount of isolated zirconia precipitates increased as the concentration of ZrIV decreased. However, the shell thickness was also decreased. This phenomenon was due to the pH difference between the two solutions. The pH of 0.5 mM solution (pH 3.21) was higher than that of 1.5 mM solution (pH 2.71); i.e., the former one had a higher supersaturation degree correspondingly, which leads to the easiness in the spontaneous precipitation of zirconia precursor in 0.5 mM solution.

(20) LaMer, V. K.; Dinegar, R. H. J. Am. Ceram. Soc. 1950, 72, 4847–4853.

(21) Kapolnek, D.; De Jonghe, L. C. J. Eur. Ceram. Soc. 1991, 7, 345–351.

6942 DOI: 10.1021/la900237m

Langmuir 2009, 25(12), 6940–6946

Zhao et al.

Article

Figure 6. TEM images of zirconia-coated hematite particles prepared by hydrolysis of 1.5 mM zirconium sulfate solutions with starting pH of 1.75 (a) and 2.00 (b), respectively.

Accordingly, the equilibrium constant Ke for this hydrolysis process could be expressed as follows: Ke ¼ ½Zr4þ 2 ½OH- 6 ½SO4 2- 

ð1Þ

4+ ] before In the zirconium sulfate solution, [SO24 ] is twice [Zr hydrolysis begins. During the hydrolysis process, the consuming 4+ rate. To simplify eq 1, [SO2rate of SO24 is half the Zr 4 ] could be considered as a constant and one obtains

Ke 0 ¼ ½Zr4þ 2 ½OH- 6

ð2Þ

Ke 00 ¼ log ½Zr4þ  þ 3pH

ð3Þ

and subsequently Figure 7. Solubility curve of zirconium sulfate in aqueous solution calculated from eq 3.

For further investigation of the effect of pH, hematite dispersions with 1.5 mM zirconium sulfate solution were prepared with different starting pH values of 1.75 and 2.00. The acid used to tune the pH value was hydrochloric acid. The hydrolysis conditions were the same as above: 30 °C for 1 h. TEM observations (Figure 6) indicated that homogeneous precipitation of isolate zirconia precursor was greatly restrained in both solutions. The results also showed that nearly no zirconia coatings were formed around the hematite particles at pH 1.75; in contrast, a shell was clearly observed for the sample obtained at pH 2.00. Moreover, this shell was much thinner than that of the sample without pH regulation (pH 2.71). This result suggests that pH has obvious effects on the hydrolysis of zirconium salts. According to the above discussion, the supersaturation degree of zirconium solution was determined by both ZrIV concentration and pH and there exists a narrow window of the supersaturation degree for heterogeneous deposition of zirconia precursor, which is critical and needed to understand surface coating.21 In this case, it is necessary to clarify the solubility and the hydrolysis process of zirconium sulfate solution. Thermodynamic Study on Hydrolysis. Before hydrolysis, dissolved zirconium sulfate first exists as ions or ion clusters in aqueous solution. Obviously, these ions or ion clusters are unstable and tend to complex and condense during the hydrolysis process, which results in the formation of zirconia precursor with the composition of Zr2(OH)6SO4, as discussed previously. Langmuir 2009, 25(12), 6940–6946

where Ke0 and Ke00 are both constants and can be expressed as Ke 0 ¼ Ke =½SO4 2- 

Ke 00 ¼

1 log Ke 0 þ 42 2

ð4Þ

ð5Þ

Ke can be estimated by a quasi-equilibrium constant Ka, calculated from [Zr4+] and the pH value, which were obtained from a complete hydrolysis, until the pH value was invariant. The preferred unit for solution concentration is mol L-1 (M). For a typical process, 2.5 mM zirconium sulfate solution hydrolyzed at 30 °C for more than 24 h. From [Zr4+] and the pH value measured, Ka was calculated to be 1.19  10-79; thus, Ke00 can be estimated to be 3.95. A solubility curve was thus obtained from eq 3, which is shown in Figure 7. The curve divides the chart into two areas with respect to precipitation and solution regions, at above and below the solubility curve, respectively. In the solubility curve (Figure 7), a coordinate in the chart is determined by both the ZrIV concentration and pH of the solution. Heterogeneously coated particles can be obtained when the coordinate locates near the solubility curve. If the coordinate is set far below the curve, the supersaturation degree of solution is low and almost no coatings could be formed. Contrarily, if the coordinate is far above the curve, homogeneous precipitation of zirconia precursor occurs inevitably. From our experimental results, the narrow window of the supersaturation degree in terms DOI: 10.1021/la900237m

6943

Article

Zhao et al.

Figure 8. Ke0 and Ke00 curves of zirconium sulfate hydrolysis in aqueous solution.

Figure 9. Second-order kinetic plot for the hydrolysis of zirconium sulfate solution at 50 °C, with [Zr4+]0 = 2.5 mM.

of “surface-deposition region” is below the solubility curve: the edge line below has an intercept of 0.8Ke00 (see Figure 7). In addition, it has to be pointed out that the solubility curve shown in Figure 7 is only suitable for the hydrolysis of zirconium sulfate in acidic aqueous solution. This is because the state of zirconium ion in aqueous solution was quite complex for different salts and pH of solution.22 The structure of zirconium complexes affects the hydrolysis process greatly, which will also be discussed below. Kinetic Study. It is well-known that the solubility varies as the temperature changes. Similarly, the solubility constants at other temperatures were also calculated. Figure 8 shows the Ke0 and Ke00 values at different temperatures; actually, Ke00 represents the intercept of the solubility curve on the log [Zr] axis. Besides solubility, the hydrolysis rate of zirconium sulfate was also determined by temperature. In general, a rate equation can be expressed as follows:23

to 1.22; thus, it gives k0 = 0.61 L mol-1 s-1. Unfortunately, the kinetic plots for the hydrolysis at 30 and 70 °C (not shown here) are nonlinear, which is probably due to two reasons: the measurement error of [Zr4+] and the simplifying treatment. For one thing, the [Zr4+] was measured by xylenol orange colorimetry, which is not an in situ method. Hence, the concentrations collected were smaller than real ones. For the other, eq 7 was considered as a pseudo-second-order kinetic by omitting the influence of [OH-]. Actually, the pH variation is drastic, and [OH-] cannot be considered as a constant, especially at the hydrolysis temperature as high as 70 °C. The hydrolysis process proceeded very quickly and approached equilibrium with pH decreasing observably. Thus, if the hydrolysis process occurred at a high temperature (supersaturation degree correspondingly), a mixture colloid composed of coated particles and homogeneous precipitates would be obtained definitely, which has been pointed out previously. In addition, the kinetics study also indicates that the hydrolysis at high temperature is so fast that it takes only a few minutes for the fresh prepared solution to approach the equilibrium state. This hydrolysis rate is obviously higher than that of zirconium alkoxides.24 Hydrolysis Mechanism of Zirconium Salts in Aqueous Solution. Zirconium is an amphoteric element, salts of which can be dissolved in both acidic and basic solutions.22,25 The aqua Zr4+ in water can form various ion clusters depending upon the chemical environments. Besides zirconium sulfate, the hydrolysis of other zirconium inorganic salts, such as ZrOCl226 and ZrO(NO3)2,27 has been reported elsewhere. Because of the difference in the state of zirconium aqua ions, some interesting experimental phenomena have been observed, which are helpful to illustrate the hydrolysis mechanism. Table 1 shows the results of two pairs of contrast experiments by mixing two solutions with the same volumes of 10 mL but with different raw materials, and results were obtained by observing immediately after mixing. Although the amounts of Zr4+ and SO24 are the same, the zirconium sulfate samples (numbers 3 and 4 in Table 1) had no precipitates, while the zirconyl chloride contained samples (numbers 1 and 2 in Table 1) precipitated immediately. Further investigations also found that sulfate radical stabilized the zirconium sulfate solution from hydrolysis, while a chloride ion accelerates the hydrolysis process. Obviously, these results mentioned above are related to the different chemical structures of zirconium aqua ions.

rate ¼ d½C=dt ¼ k½Zr4þ 2 ½OH- 6 ½SO4 2- 

ð6Þ

where [C] and k represent the concentration of the product and rate constant, respectively. The hydrolysis is considered as an irreversible reaction. The concentration units were all mol L-1, and the unit for k depends upon the order of the rate equation. However, eq 6 is very complicated and needs to be simplified. The rate-limited factor can be assumed to be [Zr4+], if the amounts of SO24 and OH are considered to be adequate and available for reactions within a short time. The rate equation can be simplified as a pseudo-second-order one, as shown in eq 7 rate ¼ d½Zr4þ =dt ¼ 2k0 ½Zr4þ 2

ð7Þ

which integrates to 2k0 t ¼ 1=½Zr4þ  -1=½Zr4þ 0

ð8Þ

where k0 = k[OH-]6[SO24 ] is the apparent rate constant and [Zr4+]0 represents the initial concentration of Zr4+. The solution concentration has a unit of mol L-1, and the time unit is in seconds. A plot of 1/[Zr4+] versus t yields a straight line following the second-order kinetics. k0 can be evaluated by a graphical solution. Figure 9 shows the second-order kinetic plot for the hydrolysis at 50 °C, with [Zr4+]0 = 2.5 mM. The fitting line has a slope equal (22) Jolivet, J.-P.; Henry, M.; Livage, J.; Bescher, E. Metal Oxide Chemistry and Synthesis, 3rd ed.; John Wiley and Sons: Chichester, U.K., 2000; Chapter 5. (23) Brezonik, P. L. Chemical Kinetics and Process Dynamics in Aquatic Systems, 1st ed.; CRC Press: Boca Raton, FL, 1994; Chapter 2.

6944 DOI: 10.1021/la900237m

(24) Ogihara, T.; Mizutani, N.; Kato, M. J. Am. Ceram. Soc. 1989, 72, 421–426. (25) Tsukada, T.; Venigalla, S.; Morrone, A. A.; Adair, J. H. J. Am. Ceram. Soc. 1999, 82, 1169–1174. (26) Matsui, K.; Ohgai, M. J. Am. Ceram. Soc. 2001, 84, 2303–2312. (27) Gao, Y. F.; Masuda, Y.; Ohta, H.; Koumoto, K. Chem. Mater. 2004, 16, 2615–2622.

Langmuir 2009, 25(12), 6940–6946

Zhao et al.

Article Table 1. Contrast Experiments by Adding Acids or Sodium Salts Solution to the Solution of Zirconium Saltsa

number

ZrOCl2 (M)

1 2

0.02 0.02 Zr(SO4)2 (M)

3 4

H2SO4 (M)

Na2SO4 (M)

0.04 0.04 HCl (M)

pH evolution

precipitate

1.80-1.70 1.80-2.35

yes yes

1.80-1.72 1.80-1.92

no no

NaCl (M)

0.02 0.04 0.02 0.04 a Correspondingly, the concentrations are listed in the table, and the volumes are all 10 mL.

Scheme 1. Chemical Structure of Zirconium Ions in (a) Zirconium Sulfate and (b) Zirconyl Chloride Solution, Respectivelya

a

The molecular structure was captured from ref 28. The gray arrows represent the attacks of ions, such as Cl- or OH-.

Figure 10. TEM images of coated hematite particles using (a) PVP and (b) HPC as surfactants, respectively.

It is well-known that the zirconium ions in zirconyl chloride aqueous solution has a stable structure as cyclical tetramer.22 However, the zirconium ion in zirconium sulfate solution is quite different, which has an open or closed trimer structure (Scheme 1).28 This structure containing sulfate radicals is unstable and attacked by anions, such as OH- and others. In comparison to sulfate radical, small anions, such as OH- or Cl-, tend to replace sulfate radicals. This replacement destroys the stable structure and accelerates hydrolysis. Thus, the sulfate radicals will facilitate the stability of the structure, whereas chloride ions accelerate the hydrolysis. Actually, experiments have shown that increasing the [Cl-] accelerates the hydrolysis of zirconium sulfate (see Figure S2 in the Supporting Information).

Influence of Surfactants. In a general surface-deposition process, various surfactants are widely used for the modification of the physicochemical properties of particle surfaces,5,7,29 because these surfactants cannot only improve the chemical affinity of the core and shell materials7 but also stabilize the core particles in the solution during the process.29 Because the affinity of hematite to zirconia precursor is high, the dispersion property of the hematite particles in the solution is very important for coating.18 In most cases, dispersion is a prerequisite for coating formation on core particles. Specifically, aggregation of hematite particles would decrease the effective area for zirconia deposition and left a lot of uncoated particles after ultrasonic processing. Moreover, the freshly coated particles also tend to aggregate

(28) Kanazhevskii, V. V.; Shmachkova, V. P.; Kotsarenko, N. S.; Kolomiichuk, V. N.; Kochubei, D. I. J. Struct. Chem. 2006, 47, 860–868.

(29) Goon, I. Y.; Lai, L. M. H.; Lim, M.; Munroe, P.; Gooding, J. J.; Amal, R. Chem. Mater. 2009, 21, 673–681.

Langmuir 2009, 25(12), 6940–6946

DOI: 10.1021/la900237m

6945

Article

Zhao et al.

hematite particles shown in Figure 11, the particles have positive charges on the surface in the region of pH ∼ 2-3. Hence, SDBS should be a nice surfactant for hematite dispersion in water.33 However, the zirconium sulfate hydrolyzed in aqueous SDBS solution very quickly, which might be because SDBS has an active radical of sulfonate, triggering the nucleation of zirconia precursor easily. Therefore, HPC is used for hematite dispersion in all of the current experiments.

Conclusions

Figure 11. ζ-Potentials as a function of pH for hematite particles in water.

because of high surface activity, which made it difficult to increase the thickness of shell. In our process, besides HPC, PVP and SDBS were also used to disperse hematite particles. PVP was reported by Grag and Matijevic18 and Xuan et al.30 to be able to effectively disperse spindle-like hematite particles, but the current experiments showed that PVP could not prevent the freshly coated particles from aggregating, especially when the soaking time was longer than 1 h (see Figure 10). This result might be because the reported spindle-like hematite particles were freshly synthesized and the ligands on the surface, especially hydroxyls, were more active, which are helpful for PVP molecule adsorption. Figure 10 shows the TEM images of coated particles prepared by the same process using PVP and HPC as dispersion surfactants, respectively. Obviously, HPC is a better choice because it adsorbed onto the particles physically and provided thermodynamic stability,31 and most importantly, it did not act as heterogeneous nucleation sites, which has been also proven by other researchers.32 This difference of the two surfactants on dispersion might be due to their macromolecular structures. The monomer of HPC is larger than that of PVP and therefore has a large steric resistance for aggregation.31 As an anionic surfactant, SDBS should be easily adsorbed to the positively charged hematite surfaces, which has been confirmed (Figure 11). From the ζ-potential curve of the (30) Xuan, S. H.; Fang, Q. L.; Hao, L. Y.; Jiang, W. Q.; Gong, X. L.; Hu, Y.; Chen, Z. Y. J. Colloid Interface Sci. 2007, 314, 502–509. (31) Jean, J. H.; Ring, T. A. Colloids Surf. 1988, 29, 273–291. (32) Moon, Y. T.; Park, H. K.; Kim, D. K.; Kim, C. H.; Seog, I. S. J. Am. Ceram. Soc. 1995, 78, 2690–2694.

6946 DOI: 10.1021/la900237m

Zirconia-coated hematite particles were prepared by hydrolysis of zirconium sulfate in aqueous solution. Freshly deposited zirconia precursor had a chemical composition of Zr2(OH)6SO4, which crystallized to t-ZrO2 after annealing at 700 °C for 2 h. The process parameters, such as the starting concentration and pH of the solution, the reaction time, and temperature, were investigated, and the following conclusions were obtained: (1) The hydrolysis of zirconium sulfate in aqueous solution has an equilibrium point at a certain temperature, and the equilibrium constant at 30 °C was 1.19  10-79. A kinetic study shows that the hydrolysis of zirconium sulfate has a pseudosecond-order kinetic at 50 °C within a short time and the rate constant k0 = 0.61 L mol-1 s-1. (2) There exists a narrow window of supersaturation degree called “surface-deposition region”, which was determined in the solubility chart. In this region, only zirconia-coated particles were obtained, while the homogeneous precipitation was restrained. (3) The hydrolysis behavior of zirconium salts in acidic aqueous solution greatly depends upon the chemical structure of zirconium aqua ions. In comparison to PVP, HPC is a better choice for hematite dispersion. Acknowledgment. This work was financially supported by the National Natural Science Foundation of China (50772129). Y. Gao thanks the Century Program (One-Hundred-Talent Program) of the Chinese Academy of Sciences for special funding support. Supporting Information Available: SEM and TEM images of hematite particles and the absorbance curves of freshly prepared zirconium sulfate solutions (0.01 mM) after hydrolyzation. This material is available free of charge via the Internet at http://pubs.acs.org. (33) Bhagat, R. P. Colloid Polym. Sci. 2001, 279, 33–38.

Langmuir 2009, 25(12), 6940–6946