Specific Ion and pH Effects on Supramolecular Assembly of

Langmuir 2004, 20, 7517-7525

7517

Specific Ion and pH Effects on Supramolecular Assembly of Mesostructured V-Mg Oxides Zi-Sheng Chao† and Eli Ruckenstein* Department of Chemical Engineering, State University of New York at Buffalo, Amherst, New York 14260 Received February 23, 2004. In Final Form: May 11, 2004 Mesostructured V-Mg oxides were synthesized using the surfactant cetyltrimethylammonium bromide (CTAB) as template, V2O5, V(acac)3 (vanadium acetylacetonate), or NH4VO3 as vanadium source, and Mg(NO3)2, MgCl2, MgSO4, (MgCO3)4‚Mg(OH)2, Mg(CH3CO2)2, or Mg(C2H5O)2 as magnesium source. The factors that influence the formation of mesostructured V-Mg oxides, such as the pH, the natures of magnesium and vanadium sources, and the ionic strength, were identified. The formation of mesophases could be related to the presence of anionic vanadium species, to the electrostatic interactions between the oppositely charged vanadates and micellar headgroups, and to the nature of the counterion of Mg2+ in the magnesium source. The main role was played by the pH and only when the pH allowed the formation of vanadates was a mesostructure generated. The counterions of Mg2+ also played a role, which could be explained via specific ion effects and the formation of complexes between them and the vanadium-containing species, which are attracted by the headgroups of the micellar templates.

1. Introduction The synthesis of mesoporous materials templated by surfactant micelles has attracted so much interest that it became a separate research field. The mesoporous materials are characterized by their large and uniform pores, high specific surface areas and pore volumes, and excellent stabilities. They have high potential for applications in a variety of fields, such as catalysis, adsorption/separation processes, and host-guest systems.1-7 Particular attention was given to the synthesis of mesostructured silicates. By utilization of various surfactants and synthesis routes, numerous mesosilicates have been prepared, among which there are three main groups: (i) The M41S family,8,9 which includes the hexagonal MCM-41, the cubic MCM-48, and the lamellar MCM-50 phases, was prepared in basic media using ionic surfactants as templates. It was suggested that the formation of these mesophases follows a cooperative selfassembly mechanism, based on the electrostatic interactions between the oppositely charged micellar headgroups and polysilicate species. (ii) The hexagonal HMS10 and MSU-X11 families, which possess wormlike pore * Corresponding author: tel, (+1 716) 645-2911, ext 2214; fax: (+1 716) 645-3822; e-mail, [email protected] (E. Ruckenstein). † Present address: College of Chemistry and Chemical Engineering, Hunan University, Changsha 410012, P.R. China. (1) Trong On, D.; Desplantier-Giscard, D.; Danumah, C.; Kaliaguine, S. Appl. Catal. 2001, 111, 299. (2) Behrens, P. Adv. Mater. 1993, 5, 127. (3) Sayari, A. Chem. Mater. 1996, 8, 1840. (4) Corma, A. Chem. Rev. 1997, 97, 2373. (5) Brunel, D. Microporous Mesoporous Mater. 1999, 27, 329. (6) Ying, J. Y.; Mehnert, C. P.; Wong, M. D. Angew. Chem., Int. Ed. 1999, 38, 56. (7) Selvam, P.; Bhatia, S. K.; Sonwane, C. G. Ind. Eng. Chem. Res. 2001, 40, 3237. (8) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359, 710. (9) Beck, J. S.; Vartuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C. T. W.; Olson, D. H.; Sheppard, E. W.; McCullen, S. B.; Higgins, J. B.; Schlenker, J. L. J. Am. Chem. Soc. 1992, 114, 10834. (10) Tanev, P. T.; Pinnavaia, T. J. Science 1995, 267, 865. (11) Bagshaw, S. A.; Prouzet, E.; Pinnavaia, T. J. Science 1995, 269, 1242.

architectures, were prepared in neutral media by employing nonionic surfactants, long-chain primary amines, or poly(ethylene oxide) as templates. The hydrogen bonding interactions between the neutral surfactant and the oligomeric silica precursors were suggested to be responsible for the self-assembly process. (iii) The hexagonal SBA-X12 family was prepared in strong acidic media by employing amphiphilic di- or triblock copolymers as templates. The interactions between the protonated template molecules and the positively charged silicate oligomers mediated by the counterions were suggested to be responsible for the self-assembly. In applications, the mesosilicates were, however, ineffective, particularly in catalysis, because of the absence of active sites in the siliceous framework. To improve the performance of mesosilicates, isomorphous substitution, grafting, or deposition of non-silica components, Al, B, Ti, Zr, V, Pd, or Mn,6,7,13,14 etc., were employed. An alternative procedure was to synthesize non-silicate mesoporous materials. Until now, a few non-silicate mesoporous materials were prepared, involving either single or mixed oxides of Sb, W, Nb, Zr, Ti, Ta, Al, V, Al-P, and V-P.6,15-32 Most of these materials have a wormlike or lamellar structure (12) Huo, Q.; Margolese, D. I.; Ciesla, U.; Feng, P.; Gier, T. E.; Sieger, P.; Leon, R.; Petroff, P. M.; Schuth, F.; Stucky, G. D. Nature 1994, 368, 317. (13) Tuel, A. Microporous Mesoporous Mater. 1999, 27, 151. (14) Ciesla, U.; Schuth, F. Microporous Mesoporous Mater. 1999, 27, 131. (15) Huo, Q.; Margolese, D. I.; Ciesla, U.; Demuth, D. G.; Feng, P.; Gier, T. E.; Sieger, P.; Firouzi, A.; Chmelka, B. F.; Schuth, F.; Stucky, G. D. Chem. Mater. 1994, 6, 1176. (16) Ciesla, U.; Demuth, D. G.; Leon, R.; Petroff, P.; Stucky, G. D.; Unger, K.; Schuth, F. J. Chem. Soc., Chem. Commun. 1994, 1387. (17) Antonelli, D. M.; Ying, J. Y. Chem. Mater. 1996, 8, 874. (18) Antonelli, D. M.; Ying, J. K. Angew. Chem., Int. Ed. Engl. 1995, 34, 2014. (19) Antonelli, D. M. Microporous Mesoporous Mater. 1999, 30, 315. (20) Sun, T.; Ying, J. Y. Nature 1997, 389, 704. (21) Holland, B. T.; Isbester, P. K.; Blanford, C. F.; Munson, E. J.; Stein, A. J. Am. Chem. Soc. 1997, 119, 6796. (22) Tian, Z. R.; Tong, W.; Wang, J. Y.; Duan, S. G.; Krishnan, V. V.; Suib, S. L. Science 1997, 276, 926. (23) Terribile, D.; Trovarelli, A.; De Leitenburg, C.; Dolcetti, G.; Llorca, J.; Chem. Mater. 1997, 9, 2676. (24) Liu, P.; Liu, L.; Sayari, A. Chem. Commun 1997, 577. (25) Michael, S. W.; Ying, J. K. Chem. Mater. 1998, 10, 2067. (26) Luca, V.; Hook, J. M. Chem. Mater. 1997, 9, 2731.

10.1021/la049528y CCC: $27.50 © 2004 American Chemical Society Published on Web 07/29/2004

7518

Langmuir, Vol. 20, No. 18, 2004

Chao and Ruckenstein

Table 1. The Phases Formed at Different pH Values Using Various V and Mg Sources and the d Spacings of the Identified Mesophasesa sources of V and Mg

phases identified at different pHs and their d spacings (Å)

V

Mg

pH ) 2

pH ) 4

pH ) 7

pH ) 10

pH ) 11.5

V(acac)3

Mg(NO3)2 3Mg(NO3)2 MgCl2 MgSO4 Mg(C2H5O)2 Mg(CH3CO2)2 Mg(NO3)2 Mg(NO3)2 MgCl2 (MgCO3)4‚Mg(OH)2

D.S. N.A. D.S. D. S. Amo Amo L (22.41) L (24.07) N.D. N.D.

Amo H (31.45) Mic Mic Mic Amo L (22.41) L (24.07) L (23.89) L (22.41)

L (34.46), H (22.72) H (32.83) Mic Mic Amo H (35.21) L (22.41) L (24.07) N.D. L (22.41)

H (36.95) H (35.60) H (36.95) H (37.96) H (36.95) H (37.66) H (43.48) 3D-H (42.07), L (28.26), L′ (24.13) N.D. L (22.41)

L (24.25) Amo L (24.61) Amo Amo Amo N.D. N.D. N.D. N.D.

NH4VO3 V2O5

a For the synthesis procedure and the batch compositions see the Experimental Section. Nomenclature: D.S., no solid product but dilute solution; N.D., not determined; Amo, amorphous; Mic, microporous phase; L and L′, lamellar mesophases; H, hexagonal mesophase; 3D-H, three-dimensional hexagonal mesophase. The numbers in parentheses provide the d spacings.

Table 2. Mg/V Molar Ratios of the Mesostructured V-Mg Oxides Synthesized from Batches with Different Compositions and pH Values Mg/V molar ratios of the mesostructured V-Mg oxides synthesized at different pH values pH ) 2

composition of initial batches V(acac)3/Mg(CH3CO2)2/0.25 CTAB/295 H2O/14 EtOH V(acac)3/Mg(C2H5O)2/0.25 CTAB/295 H2O/14 EtOH V(acac)3/MgCl2/0.25 CTAB/295 H2O/14 EtOH V(acac)3/Mg(NO3)2/0.25 CTAB/295 H2O/14 EtOH V(acac)3/3Mg(NO3)2/0.375 CTAB/ 442.5 H2O/14 EtOH V(acac)3/MgSO4/0.25 CTAB/295 H2O/14 EtOH NH4VO3/MgNO3/0.25 CTAB/295 H2O V2O5/Mg(NO3)3/0.25 CTAB/295 H2O V2O5/0.2(MgCO3)4‚Mg(OH)2)/0.25 CTAB/295 H2O V2O5/MgCl2/0.25 CTAB/295 H2O

pH ) 4

pH ) 7

3.13 × 10-2 5.24 × 10-3

and lose their mesostructure after the removal of the surfactant from the framework by either solvent extraction or calcination. Recently, mesostructured V-Mg oxides were synthesized in our laboratory for their potential application to catalysis. By use of various vanadium and magnesium sources, various templates, and batch compositions, a number of mesostructures and morphologies have been identified.33-37 Catalysts were prepared from mesostructured Mg-V oxides precursors and used in the oxidative dehydrogenation of ethane and propane.38,39 In the present paper, the effects of the pH and of the anions of the Mg compounds employed on the structure of mesoporous V-Mg oxides have been investigated by using a large variety of sources for V and Mg. An attempt was made to relate the structures obtained to specific ion effects via the Hofmeister series. 2. Experimental Section All reagents were purchased from Aldrich and were used without further purification. The vanadium and magnesium (27) Huo, Q.; Margolese, D. I.; Ciesla, U.; Demuth, D. G.; Feng, P.; Gier, T. E.; Sieger, P.; Firouzi, A.; Chmelka, B. F.; Schuth, F.; Stucky, G. D. Chem. Mater. 1996, 6, 1840. (28) Ayyappan, S.; Rao, C. N. R. Chem. Commun. 1997, 575. (29) Kimura, T.; Sugahara, Y.; Kuroda, K. Chem. Commun. 1998, 559. (30) Mizuno, N.; Hatayama, H.; Uchida, S.; Taguchi, A. Chem. Mater. 2001, 13, 179. (31) Ciesla, U.; Schacht, S.; Stucky, G. D.; Unger, K. K.; Schuth, F. Angew. Chem., Int. Ed. Engl. 1996, 35, 541. (32) Gougeon, R. D.; Bodart, P. R.; Harris, R. K.; Kolonia, D. M.; Petrakis, D. E.; Pomonis, P. J. Phys. Chem. Chem. Phys. 2000, 2, 5286. (33) Chao, Z. S.; Ruckenstein, E. Langmuir 2002, 18, 734. (34) Chao, Z. S.; Ruckenstein, E. Chem. Mater 2002, 14, 4611. (35) Ruckenstein, R., Chao, Z. S. Nano Lett. 2001, 1, 739. (36) Chao, Z. S.; Ruckenstein, E. Langmuir 2002, 18, 8535. (37) Chao, Z. S.; Ruckenstein, E. Langmuir 2002, 19, 4235. (38) Chao, Z. S.; Ruckenstein, E. J. Catal. 2004, 222, 17. (39) Chao, Z. S.; Ruckenstein, E. Catal. Lett. 2004, 94, 217.

1.05 × 10-2 2.31 × 10-2 2.25 × 10-2

8.23 × 10-2 2.94 × 10-2

pH ) 10 1.79 × 10-2 3.84 × 10-3 5.62 × 10-3 3.65 × 10-3 1.48 × 10-1 3.63 × 10-3 5.73 × 10-1 1.38 × 10-1

sources consisted of V2O5, V(acac)3 (vanadium acetylacetonate), and NH4VO3 and Mg(NO3)2, MgCl2, MgSO4, (MgCO3)4‚Mg(OH)2, Mg(CH3CO2)2, and Mg(C2H5O)2, respectively. A cationic surfactant, CTAB (cetyltrimethylammonium bromide), was used as template. First, a vanadium-containing solution was prepared using a procedure dependent on the vanadium source employed. V(acac)3 was dispersed into a C2H5OH/H2O (1/3, v/v) solution with intense overnight stirring, V2O5 was dissolved into a 2 M NaOH aqueous solution (molar ratio NaOH/V2O5 ) 2) with heating and stirring, and NH4VO3 was dissolved into water with stirring, the dissolution being promoted with several drops of 2 M NaOH aqueous solution. Second, the magnesium source, except (MgCO3)4‚Mg(OH)2, and the surfactant were dissolved together into water with stirring, the dissolution being promoted with several drops of 2 M HCl aqueous solution. (MgCO3)4‚Mg(OH)2 was dissolved into a concentrated HCl solution (molar ratio HCl/(MgCO3)4‚ Mg(OH)2 ) 10), and then the surfactant was added under stirring. Finally, the solution containing the vanadium source was dropwise introduced into that containing the magnesium source and the template with intense stirring. By use of dilute HCl or NaOH aqueous solutions, the pH of the mixture was adjusted to a selected value between 2 and 11.5, monitored by a Corning Chekmite pH-10 pH-meter (Corning Inc.) with a resolution of 0.02. The final batch had a molar composition 1.0 V source/x Mg source/0.25 CTAB/295 H2O/y C2H5OH, where x ) 0.2 when (MgCO3)4‚Mg(OH)2 was the magnesium source and 1.0 for the other magnesium sources and y ) 14 for V(acac)3 as vanadium source and 0 for the other vanadium sources. A specific batch with a molar composition 1.0 V(acac)3/3.0 Mg(NO3)2/0.375 CTAB/ 442.5 H2O/14 C2H5OH was also employed to examine the effect of the electrolyte concentration. The batches were allowed to age at room temperature for 4 days, and the solid products formed were recovered by filtration, washed with distilled water, and dried at 100 °C for 12 h. 2.1. Characterization. The phase structures of the specimens were determined by X-ray diffraction (XRD), using a Siemens D500 diffractometer and a Cu KR radiation of 1.5406 Å. The diffraction data were recorded for 2θ angles ranging between 1 and 15°, with a resolution of 0.02°. The structure and morphology

Assembly of Mesostructured V-Mg Oxides

Figure 1. XRD spectra of specimens prepared from a batch with the molar composition 1.0 V(acac)3/1.0 Mg(CH3CO2)2/0.25 CTAB/295 H2O/14 C2H5OH and pH values of (a) 2, (b) 4, (c) 7, (d) 10, and (e) 11.5.

Langmuir, Vol. 20, No. 18, 2004 7519

Figure 2. XRD spectra of specimens prepared from a batch with the molar composition 1.0 V(acac)3/1.0 Mg(C2H5O)2/0.25 CTAB/295 H2O/14 C2H5OH and pH values of (a) 2, (b) 4, (c) 7, (d) 10, and (e) 11.5.

of the specimens were determined by transmission electron microscopy (TEM) and selected area electron diffraction (SAED), using a JEOL-JEM-100CX transmission electron microscope equipped with a tungsten gun operating at an accelerating voltage of 80 keV. Before the TEM measurements, the specimens were ground in acetone and supported on holey carbon films located on Cu grids. The mesophases identified at various pH values, using various V and Mg sources, and their d spacings are summarized in Table 1. The compositions of the products were determined by the inductively coupled plasma (ICP) method on a Perkin-Elmer 3000 ICP OES system in the Gallbraith Laboratories, Inc., and the results are summarized in Table 2.

3. Results and Discussion 3.1. Formation of V-Mg Oxides Mesophases. Figures 1-5 present the XRD patterns of the solid products prepared from batches with a molar composition of 1.0 V(acac)3/1.0 Mg source/0.25 CTAB/295 H2O/14 C2H5OH, where the magnesium source was MgCl2, Mg(NO3)2, MgSO4, Mg(CH3CO2)2, or Mg(C2H5O)2, respectively. In acidic media, no mesophases could be identified for any of the above magnesium sources. For pH < 4 and MgCl2, Mg(NO3)2, or MgSO2 as magnesium source, no solid products were obtained, while for Mg(CH3CO2)2 or Mg(C2H5O)2 the solid products were amorphous (Figures 1a and 2a). At pH ) 4, some strong diffraction peaks were identified for 2θ > 10° in the XRD patterns when Mg(C2H5O)2, MgCl2, or MgSO4 was used as magnesium source (Figures 2b, 3a, and 5a), revealing that microporous phases were generated; when Mg(NO3)2 or Mg(CH3CO2)2 was used as magnesium source, only amorphous solids were obtained (Figures 1b and 4a). In neutral media (pH ) 7), mesophases could be identified from the XRD patterns when Mg(CH3CO2)2 or Mg(NO3)2 was employed as magnesium source, because diffraction peaks in the lower 2θ range were present

Figure 3. XRD spectra of specimens prepared from a batch with the molar composition 1.0 V(acac)3/1.0 MgCl2/0.25 CTAB/ 295 H2O/14 C2H5OH and a pH of (a) 4, (b) 7, (c) 10, and (d) 11.5.

(Figures 1c and 4b). Mg(NO3)2 provided both a lamellar (d (Å) ) 22.72 (001), 11.68 (002), and 7.72 (003), the number in parentheses following the d spacing indicating the diffraction face) and a distorted hexagonal phase (d (Å) ) 34.46 (100), 19.18 (110), 16.88 (200)), whereas Mg(CH3CO2)2 provided only a hexagonal phase (d (Å) ) 35.21 (100), 19.41 (110), 17.13 (200)). MgCl2 or MgSO4 provided a microporous phase (Figures 3b and 5b) and Mg(C2H5O)2 an amorphous solid (Figure 1b).

7520



Figure 4. XRD spectra of specimens prepared from a batch with the molar composition 1.0 V(acac)3/1.0 Mg(NO3)2/0.25 CTAB/295 H2O/14 C2H5OH and pH values of (a) 4, (b) 7, (c) 10, and (d) 11.5.

Figure 6. TEM micrographs of specimens prepared from a batch with the molar composition 1.0 V(acac)3/1.0 Mg source/ 0.25 CTAB/295 H2O/14 C2H5OH and a pH of 7. The Mg sources were (a) Mg(NO3)2 and (b) Mg(CH3CO2)2.

Figure 5. XRD spectra of specimens prepared from a batch with the molar composition 1.0 V(acac)3/1.0 MgSO4/0.25 CTAB/ 295 H2O/14 C2H5OH and pH values of (a) 4, (b) 7, (c) 10, and (d) 11.5.

Weak and moderate basic media promoted the formation of mesophases. At pH ) 10, distorted hexagonal mesophases could be identified for all magnesium sources employed. They were characterized by a strong diffraction peak for 2θ between 2.32 and 2.39° (d100 ) 37.66, 36.95, 36.95, 36.95, and 37.96 Å), and two very weak diffraction peaks for 2θ between 4.06 and 4.25° (d110 ) 21.76-20.78 Å) and 2θ between 4.68 and 4.81° (d200 ) 18.89-18.35 Å), for Mg(CH3CO2)2, Mg(C2H5O)2, MgCl2, Mg(NO3)2, and MgSO4, respectively (Figures 1d, 2d, 3c, 4c, and 5c). For

pH ) 11.5, no mesophases could be identified when Mg(CH3CO2)2, Mg(C2H5O)2, and MgSO4 were used as magnesium sources. However, when MgCl2 and Mg(NO3)2 were used as magnesium sources, the distorted hexagonal mesophases formed at low pH values were transformed into lamellar phases at high pH values. Indeed, as shown in Figures 3d and 4d, equidistanced diffractions at 2θ ) 3.59, 6.89, and 10.20° and 2θ ) 3.64, 6.93, and 10.28° for MgCl2 and Mg(NO3)2, respectively, were present at pH ) 11.5. The above specimens were also examined by TEM and SAED, and some of the electron micrographs are presented in Figures 6 and 7. The specimens prepared in acidic media were amorphous. The specimens prepared in neutral media were also amorphous when MgCl2, MgSO4, or Mg(C2H5O)2 was employed as magnesium source. However, mesostructures were identified for Mg(NO3)2 or Mg(CH3CO2)2 as magnesium source. Both a lamellar and a distorted hexagonal (wormlike pore architecture) mesostructure were identified for Mg(NO3)2 (Figure 6a), and only a distorted hexagonal mesophase was identified for Mg(CH3CO2)2 (Figure 6b). For the specimens prepared in weak and moderate basic media, e.g., pH ) 10, distorted


Langmuir, Vol. 20, No. 18, 2004 7521

Figure 8. XRD spectra of the specimens prepared from a batch with the molar composition 1.0 V(acac)3/3.0 Mg(NO3)2/0.375 CTAB/442.5 H2O/14 C2H5OH and pH values of (a) 4, (b) 7, (c) 10, and (d) 11.5.

Figure 7. TEM micrographs and SAED patterns of specimens prepared from a batch with the molar composition 1.0 V(acac)3/ 1.0 Mg source/0.25 CTAB/295 H2O/14 C2H5OH and a pH of 11.5. The Mg sources were (a) Mg(NO3)2, (b) MgCl2 for particle morphology, and (c) MgCl2 for nanotube morphology.

hexagonal mesophases were identified by TEM for all magnesium sources employed. However, for the specimens prepared in media with high basicity, e.g., pH ) 11.5, only those prepared with Mg(NO3)2 or MgCl2 as magnesium source exhibited mesoporous features (Figure 7). Ordered lamellar mesostructures were identified in both cases (Figures 7a,b). For MgCl2, very regular nanotubes

together with particulates were also present in the specimen and one of the nanotubes is presented in Figure 7c. Figure 8 presents the XRD patterns of the specimens prepared by aging at room temperature a batch with a molar composition 1.0 V(acac)3/3.0 Mg(NO3)2/0.375 CTAB/ 442.5 H2O/14 C2H5OH. Within a broad range of pH values, viz., pH e 10, distorted hexagonal mesophases were formed. The d spacings increased with increasing pH, having the values d100 ) 31.45, 32.83, and 35.60 Å for pH ) 4, 7, and 10 (lines a, b, and c of Figure 8), respectively. The crystallinity of the mesophase increased in the sequence pH ) 7 < pH ) 10 < pH ) 4. A further increase of the pH did not generate a V-Mg oxide mesophase. Indeed, no diffraction peaks specific to a mesophase could be identified by XRD for a pH ) 11.5 (Figure 8d). In contrast, as already mentioned, at a lower Mg(NO3)2 concentration in the initial batch, the system behaved differently. The synthesis of mesostructured V-Mg oxides with NH4VO3 as vanadium source was examined using a batch with the molar composition 1.0 NH4VO3/1.0 Mg(NO3)2/ 0.25 CTAB/295 H2O. Figure 9 presents the XRD patterns of the specimens prepared at various pH values. One can see that in acidic and neutral media, the specimens had a lamellar structure (d (Å) ) 22.41 (001), 11.28 (002), and 7.51 (003)) (Figure 9a-c). Several weak diffraction peaks at 2θ ) 8-12° and 2θ > 13° were also present, which are probably due to the microporous structures of the layers themselves. Within the acidic and neutral range, the crystallinity of the V-Mg oxide mesophase increased with increasing pH. However, when the pH increased in the basic range, the lamellar mesophase disappeared, being replaced by a distorted hexagonal mesophase with a d spacing of 43.48 Å (Figure 9d). The synthesis of mesostructured V-Mg oxides with V2O5 as vanadium source was examined using a batch with the molar composition 1.0 V2O5/x Mg source/0.25 CTAB/295

7522


Figure 9. XRD spectra of the specimens prepared from a batch with the molar composition 1.0 NH4VO3/1.0 Mg(NO3)2/0.25 CTAB/295 H2O and pH values of (a) 2, (b) 4, (c) 7, and (d) 10.

H2O, where the Mg sources were Mg(NO3)2, MgCl2, or (MgCO3)4‚Mg(OH)2 with x ) 1.0 for the former two and 0.2 for the latter one. Figures 10 and 11 present the XRD pattern of the specimens prepared at various pH values. In neutral and acidic media, a lamellar mesophase was identified for all three magnesium sources (Figures 10a-c and 11a,b,d), with the d001 of 24.07, 23.89, and 22.41 Å for Mg(NO3)2, MgCl2, and (MgCO3)4‚Mg(OH)2 as magnesium source, respectively. With increasing pH, the crystallinity of the mesophase increased. In basic media, the same lamellar mesophase as that in neutral and acidic media was still identified when (MgCO3)4‚Mg(OH)2 was used as magnesium source, but the crystallinity of the lamellar mesophase was much reduced (Figure 11c). When Mg(NO3)2 was used as magnesium source, a 3D hexagonal mesophase with a d100 of 42.07 Å and c/a of 1.61, and two lamellar mesophases with the d001 and d′001 of 28.26 and 24.13 Å were identified (Figure 10d). 3.2. Roles of Ions and Electrolytes. Generally, to synthesize mesoporous materials templated by micelles of ionic surfactants, two main conditions must be satisfied: (i) the inorganic precursors should have the ability to form polyions that can interact with the template via electrostatic interactions and (ii) the polyions should be able to polymerize into a rigid structure. Additionally, for mesoporous materials of mixed components, the hydrolysis rates of the inorganic precursors should be comparable to generate a mesophase containing large proportions of all components. Due to these conditions, few mixed oxide mesoporous materials could be prepared. The condition of comparable hydrolysis suggests that the pH should be one of the main factors that affect the


Figure 10. XRD spectra of the specimens prepared from a batch with the molar composition 1.0 V2O5/1.0 Mg(NO3)2/0.25 CTAB/295 H2O and pH values of (a) 2, (b) 4, (c) 7, and (d) 10.

formation of mesophases. The pH affects not only the hydrolysis and polymerization of the inorganic species, particularly vanadium, but also the aggregation of the surfactant in solution and hence the interaction between the inorganic framework species and the surfactant aggregates. In solution, the vanadate species generate various polyvanadate anions in equilibrium, dependent on the pH of the medium. Under alkaline conditions, the polyvanadate anions have a low degree of polymerization, which can be represented by VxOyδ-, where x < 4, y < 12, and δ ) 1-4, while under acidic conditions, the polyvanadate anions involve the decavanadates V10O286-, HV10O285-, and H2V10O284-.40 As noted in section 3.1, the phase behavior of the mesostructured V-Mg oxides prepared using various V and Mg sources depends on the pH of the medium and to some extent on the nature and concentrations of ions, particularly the counterions of Mg2+. However, the effect of the nature of the anions was significant only in a relatively narrow range of pH values, namely, neutral and strong basic media, suggesting a more moderate effect on the formation of V-Mg oxide mesophases than that of the pH. A detailed discussion of these effects is provided below. 3.2.1. V(acac)3-(MgNO3, MgCl2, MgSO4, Mg(CH3CO2)2, Mg(C2H5O)2)-CTAB Systems. When V(acac)3 was used as vanadium source, polyvanadate ions could not be formed under acidic conditions, because the hydrolysis of V(acac)3 is a base-catalyzed reaction. Instead, (40) Kepert, D. L. In The Early Transition Metals; Academic Press: London and New York, 1972; p 181.


Figure 11. XRD spectra of the specimens prepared from a batch with the molar composition 1.0 V2O5/0.2 (MgCO3)4‚Mg(OH)2/0.25 CTAB/295 H2O and pH values of (a) 4, (b) 7, and (c) 10. (d) Prepared from a batch with the molar composition 1.0 V2O5/1.0 MgCl2/0.25 CTAB/295 H2O and a pH of 4.

V(acac)3 could be protonated, thus acquiring a positive charge. In this case, hardly any V-Mg oxide mesophase can be formed, because the inorganic species and the surfactant have the same charge. Under neutral conditions, V(acac)3 is neutral and no polyvanadate ions, which can be attracted by the surfactant micelles, can be formed. When MgCl2, MgSO4, or Mg(C2H5O)2 was used as magnesium source, no mesophase could be identified by XRD; however, when Mg(NO3)2 or Mg(CH3CO2)2 was employed, mesophases were generated. This observation indicates that besides the pH, the anion of the magnesium salt can affect the formation of V-Mg oxide mesophases. Ionic effects are frequently encountered in colloid science, and their influence on the shape and size of micelles was studied.41 The specific ion effects are a result of the modifications they cause in the structure of water. The ions can be structure making (or cosmotropic), when they stimulate the organization of water, and structure breaking (chaotropic), when they disorganize the structure of water. The specific ion effects are important because they affect the electrostatic interactions. The anions can be ordered in a series, the Hofmeister series, and the location of these anions in the series can be correlated with their hydration in pure water.42 The Hofmeister series orders the ions in increasing structure-breaking potency: SO42-, OH-, HCOO-, CH3COO-, Cl-, Br-, NO3-, I-, ClO4-. The ions on the left of Cl- are structure makers, whereas those at the right are structure breakers. Numerous researchers43-48 observed experimentally that particularly (41) Leontidis, E. Curr. Opin. Colloid Interface Sci. 2002, 7, 81. (42) Manciu, M.; Ruckenstein, E. Adv. Colloid Interface Sci. 2003, 105, 63.

Langmuir, Vol. 20, No. 18, 2004 7523

the counterions of the cationic surfactants affect the formation of mesophases. Can the different behaviors of various magnesium sources be related to the Hofmeister series? The empirical Hofmeister series is not expected to be completely applicable to the present case, not only because it was initially formulated for the protein solubility but also because of the complex interactions between the numerous ions involved in the present case. However, sometimes it provides some insight. Being a chaotropic ion, NO3- can break the structure of water near the positively charged CTA+ and has, therefore, a higher affinity for the cationic headgroups of the surfactant than the other anions of the Mg compounds employed, Cl-, CH3CO2-, C2H5O-, and SO42-, which (with the exception of Cl- which is located at the intersection between cosmotropic and chaotropic) are cosmotropic (structure makers). NO3- can, therefore, better screen the electrostatic repulsion between the surfactant headgroups, thus stimulating the formation of rodlike micelles. Under acidic conditions, V(acac)3 becomes positively charged and, as a result, the vanadium and magnesium species cannot interact with the positively charged rodlike micelles and no mesophase can be generated. Under neutral conditions, V(acac)3 can interact with the positively charged micelles, perhaps because of the coordination between vanadium and NO3-. The further condensation of V(acac)3 results in a mesostructured V-Mg oxide. Because only trace amounts of magnesium can deposit on the surface of the micelles, a mesophase rich in vanadium will be formed. Under basic conditions, V(acac)3 can be transformed in polyvanadates which strongly interact with the micelles generating a mesophase rich in vanadium. It is noteworthy that while CH3COO- is positioned in the Hofmeister series as a structure-making anion, its behavior regarding the formation of a mesophase is similar to that of NO3-. In neutral conditions, this probably happened because of the coordination of CH3COO- to V(acac)3 and the interactions between the negatively charged species thus formed and the positively charged micelles. As a result of a cooperative assembly, a V-Mg oxide mesophase was formed. The C2H5O-, also a structure-making anion, which has a structure similar to that of CH3COO-, could not provide under neutral conditions a mesophase, probably because the coordination of C2H5O- to the vanadium species was much weaker than that of CH3COO-. In moderate basic media (pH ) 10), the V-Mg oxide mesophases could be obtained with any of the magnesium sources employed, regardless of whether the anions were chaotropic or cosmotropic. This occurred because in basic media, V(acac)3 underwent hydrolysis generating polyvanadate anions. The electrostatic interactions between the polyvanadate anions and the surfactant micelles (CTA+) template are responsible for the formation of V-Mg oxides mesophases. However, at high basicities (pH g 11.5), the vanadium species are present in solution not as polyvanadate ions but as small VO3- ions.40 In the absence of an effective coupling between the vanadium species and the surfactant micelles, no mesophases of V-Mg oxides could be formed when MgSO4, Mg(CH3CO2)2, or (43) Lin, H. P.; Kao, C. P.; Mou, C. Y. Microporous Mesoporous Mater. 2001, 48, 135. (44) Lin, H. P.; Kao, C. P.; Mou, C. Y.; Liu, S. B. J. Phys. Chem. B 2000, 104, 7885. (45) Echchahed, B.; Morin M.; Blais, S.; Badiei, A. R.; Berhault, G.; Bonneviot, L. Microporous Mesoporous Mater. 2001, 44, 53. (46) Badiei, A. R.; Cantournet, S.; Morin M.; Bonneviot, L. Langmuir 1998, 14, 7087. (47) Newalkar, B. L.; Komarneni, S. Chem. Mater. 2001, 13, 4573. (48) Zana, R.; Frasch, J.; Soulard, M.; Lebeau, B.; Patarin, J. Langmuir 1999, 15, 2603.

7524


Mg(C2H5O)2 was used as magnesium source. However, when MgCl2 or Mg(NO3)2, containing the anions Cl- and NO3-, was employed as magnesium source, lamellar mesophases of V-Mg oxides were generated. The high affinity of the chaotropic anion NO3- for the headgroups of the surfactant reduced the repulsion between the headgroups of surfactants, stimulating the formation of rodlike micelles. The VO3- anions were attracted by the positively charged micelles, and the condensation of the vanadium species generated a vanadium-rich mesophase. While Cl- is located at the intersection between the chaotropic and cosmotropic ions, its behavior regarding the formation of mesophases appears to be similar to that of the chaotropic NO3-. The transformation of the distorted hexagonal mesophases (with wormlike pore architectures) to lamellar ones when the pH was increased from 10 to 11.5 might have been caused by its effect on the dissociation of the headgroups and the change in the packing parameter of the surfactant. The packing parameter is defined as Vh/ lcao, where Vh is the volume occupied by the hydrophobic group of the micelle, lc is the length of the hydrophobic group, and ao is the cross-sectional area occupied by the hydrophilic group at the micelle-solution interface. The increase of the pH decreased the dissociation of the headgroups, hence the repulsion between them, thus decreasing ao and increasing the value of the packing parameter. While for nonspherical micelles the packing parameter is between 1/3 and 1/2, for bilayers it is between 1/2 and 1.49 Consequently, when the pH was varied from 10 to 11.5, the shape of the surfactant aggregates changed and the mesophases templated by them changed from hexagonal to lamellar. The formation of mesostructured V-Mg oxides is expected to be dependent on the concentration of Mg(NO3)2, because of the higher concentration of the NO3structure breaking ion and increased ionic strength when the concentration of Mg(NO3)2 increases. The increase in the ionic strength decreases the repulsion between similarly charged systems and the attraction between oppositely charged systems. As a result, the increase of the ionic strength decreases the rate of adsorption of an ion on an oppositely charged adsorbent and increases the rate of adsorption of an ion on a similarly charged adsorbent. Thus, for kinetic reasons, in acidic media, mesophases of V-Mg oxides could be generated because of the increased rate of adsorption of the protonated V(acac)3 species onto the positively charged CTA+ surfactant micelles with increasing ionic strength. In contrast, at low Mg(NO3)2 concentrations the amount adsorbed was very low and no mesophase could be generated (compare Figures 2 and 8). 3.2.2. (V2O5, NH4VO3, V(acac)3)-Mg(NO3)2-CTAB Systems. In basic media, similar distorted hexagonal mesophases were formed with either NH4VO3 or V(acac)3 as vanadium source, suggesting that vanadium underwent first hydrolysis, generating polyvanadate anions. In the case of V(acac)3, a vanadium-rich mesophase was generated, because of the electrostatic interactions between the oppositely charged polyvanadates and surfactant micelles. In the case of NH4VO3, the mesophase contained a relatively large proportion of Mg. Probably that in this case the polyvanadates interacted first with MgNO3 generating negatively charged magnesium polyvanadates, which interacted with the micelles forming V-Mg mesophases. Lamellar mesophases were formed in neutral and (49) Israelachvili, J. Intermolecular and Surface Forces, 2nd ed.; Academic Press: London, 1992.


acidic media, when NH4VO3 was used as the vanadium source; a distorted hexagonal mesophase was formed in neutral media and no mesophase in acidic media, when V(acac)3 was used as the vanadium source. In neutral and acidic media, lamellar mesophases were generated from both V2O5 and NH4VO3, confirming again the importance of the formation of the polyvanadate ions. In basic media, a distorted hexagonal mesophase for NH4VO3 + Mg(NO3)2 and a 3D hexagonal and two lamellar mesophases for V2O5 + Mg(NO3)2 were identified. Because V2O5 was first dissolved into a NaOH aqueous solution, before being employed in the synthesis of mesostructured V-Mg oxides, and then the basicity was neutralized with a solution of HCl containing the surfactant and magnesium, large amounts of Cl- and Na+ were present in the V2O5 + Mg(NO3)2 system. It appears that NH4+ favored the formation of a hexagonal phase, whereas Na+ favored the formation of a lamellar mesophase. By comparing the phases formed under basic conditions when V2O5 and V(acac)3 were used as vanadium sources, one can conclude that the organic compounds which were either introduced in the initial batch (ethanol) or formed during the process (acetyl acetone), promoted the formation of hexaagonal mesophses. The effect of Mg(NO3)2 concentration on the formation of V-Mg oxide mesophases from V2O5 + Mg(NO3)2 + CTAB was somewhat different than that from V(acac)3 + Mg(NO3)2 + CTAB. For the former,36 we found that, with increasing Mg/V mole ratio in the initial batch, lamellar mesophases were formed regardless of the pH of the medium until a Mg/V mole ratio of 10 was attained. Above this ratio, such as Mg/V ) 20, cubic and octamer mesostructures were formed in acidic and basic media, respectively, and a lamellar mesophase was generated in neutral media. In contrast, for V(acac)3 + Mg(NO3)2 + CTAB, the increase of the mole ratio Mg/V in the initial batch from 1 to 3 resulted in the transformation from a lamellar mesophase to an amorphous phase in strong basic media and in the transformation from an amorphous phase to a hexagonal mesophase in acidic media. 3.2.3. V2O5-(Mg(NO3)2, MgCl2, (MgCO3)4‚Mg(OH)2)CTAB Systems. Under neutral and acidic conditions, lamellar mesophases were identified when V2O5 was used as the vanadium source and Mg(NO3)2, MgCl2, or (MgCO3)4‚Mg(OH)2 as magnesium source. Because (MgCO3)4‚Mg(OH)2 was first dissolved in a HCl aqueous solution before being employed in the synthesis of the mesostructured V-Mg oxides, and then the pH of the batch was adjusted with a NaOH aqueous solution, V2O5 + (MgCO3)4‚Mg(OH)2 was transformed into a V2O5 + MgCl2 + Na2CO3 system. This suggests that the formation of mesophases in acidic and neutral media was not much affected by the nature of the anion (NO3-, Cl-, or CO32-) employed. In basic media, 3D hexagonal and lamellar mesophases were generated from V2O5 + Mg(NO3)2 and only a lamellar mesophase from V2O5 + (MgCO3)4‚Mg(OH)2. Because a hexagonal mesophase was formed from NH4VO3 + Mg(NO3)2, where the concentration of Na+ was much lower than that in V2O5 + MgCl2 or (MgCO3)4‚Mg(OH)2 systems, one may conclude that a higher concentration of Na+ in the system favored the formation of lamellar mesophases and hindered the formation of the hexagonal ones. 3.3. Presence of Mg in the Mesostructured V-Mg Oxides. The above results have shown that V-Mg oxide mesophases can be formed as a result of electrostatic interactions between the polyvanadate anions and the micellar cations (CTA+) in acidic, neutral, and basic media. They were dependent on the vanadium source and affected


to some extent by specific ion effects and ionic strength. Whatever routes were involved in the formation of the mesostructures, the hydrolysis of the vanadium species at the beginning and the subsequent condensation into an inorganic framework represent the key steps. During these processes, magnesium can enter into the inorganic framework through the interaction between the oppositely charged Mg2+ and polyvanadate anions. Table 2 lists the Mg/V mole ratios in mesostructured V-Mg oxides synthesized from batches of various compositions. As wellknown, the size of the polyvanadates is mainly controlled by the pH of the medium: the higher the pH, the smaller the size of the polyvanadates. The amount of magnesium contained in the mesostructured V-Mg oxides is also mostly dependent on the pH of the medium. With increasing pH from acidic to neutral and basic, the Mg/V mole ratio in the synthesized mesostructured V-Mg oxides increased by nearly 1 order of magnitude in each step. The Mg/V ratio in the synthesized specimen also increased with increasing Mg concentration in the initial batch. The nature of the magnesium source had a relatively low effect on the Mg/V mole ratio in the synthesized specimens. Conclusion Factors such as the nature of V and Mg sources, the pH of the synthesis medium, and the ionic strength affected the formation of V-Mg oxide mesophases templated by the surfactant CTAB. When V(acac)3 was employed as V source, hexagonal mesophases were formed from media with moderate basicities, regardless the Mg source employed (Mg(NO3)2, MgCl2, MgSO4, (MgCO3)4‚Mg(OH)2, Mg(CH3CO2)2, or Mg(C2H5O)2). Lamellar mesophases were formed in strong basic media when the Mg source was associated with the chaotropic anion NO3- or in neutral

Langmuir, Vol. 20, No. 18, 2004 7525

media when the anion of the Mg source was either strongly chaotropic (like NO3-) or able to coordinate to the vanadium species (like CH3COO-). In all the other cases, either microphases, or amorphous solids, or even dilute solutions were formed. When V2O5 or NH4VO3 was employed as V source, lamellar mesophases were formed in acidic and neutral media regardless of the Mg source employed. A hexagonal mesophase for the NH4VO3 + Mg(NO3)2 system and 3D hexagonal and/or lamellar mesophases for the V2O5 + Mg(NO3)2 and V2O5 + (Mg(CO3)2)4‚ Mg(OH)2 systems were generated in basic media. The increase of the ionic strength of the medium had a much stronger influence on the formation of mesophases from V(acac)3 + Mg(NO3)2 + CTAB than from V2O5 + Mg(NO3)2 + CTAB. The amount of Mg contained in the mesostructured V-Mg oxide was dependent mainly on the pH of the synthesis medium and the Mg/V mole ratio in the initial batch and less on the nature of the Mg source. The V sources NH4VO3 and V2O5 generated mesostructured V-Mg oxides with much higher Mg contents than V(acac)3. The results obtained indicated that among the various factors, the pH of the synthesis medium played the main role, by controlling the hydrolysis/polymerization of vanadium species. The cooperative self-assembly based on electrostatic interactions between V species and surfactant micelles affected the mesostructures formed. In the cooperative self-assembly process, several factors, such as the Hofmeister anion series, the ionic strength, and the complexes formed between the vanadiumcontaining species and the anions of the Mg compounds influenced, to some extent, the formation of mesostructured V-Mg oxides. LA049528Y

Specific Ion and pH Effects on Supramolecular Assembly of

Recommend Documents