Microporous and Mesoporous Materials 132 (2010) 268–275

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials

journal homepage: www.elsevier .com/locate /micromeso

Preparation of hierarchically structured porous aluminas by a dual softtemplate method

Leandro Martins, Marinalva A. Alves Rosa, Sandra H. Pulcinelli, Celso V. Santilli *

Instituto de Química, UNESP – Univ Estadual Paulista, P.O. Box 355, 14800-900 Araraquara, SP, Brazil

a r t i c l e i n f o a b s t r a c t

Article history:Received 11 January 2010Received in revised form 4 March 2010Accepted 5 March 2010Available online 12 March 2010

Keywords:Macro–mesoporous aluminasMicelleEmulsionSol–gel and template

1387-1811/$ - see front matter � 2010 Elsevier Inc. Adoi:10.1016/j.micromeso.2010.03.006

* Corresponding author. Tel.: +55 16 33016645; faxE-mail address: santilli@iq.unesp.br (C.V. Santilli).

A general and potentially easy method of synthesizing aluminas possessing hierarchical pore structure ispresented in this paper. This method is based on the integration of the sol–gel process with micelles ofpoly(ethylene oxide)–poly(propylene oxide)–poly(ethylene oxide) block copolymer and oil droplets ofdecahydronaftalen (DHN) as dual pore templates. With the addition of n-pentanol as co-surfactant,DHN droplets diameter could be controlled within the range of ca. 2–12 lm. Mercury intrusion porosi-metry shows that the produced alumina ceramics possess hierarchical structure composed of two fam-ilies of pores, namely, the macro- (from 0.1 to 6 lm) and mesopores (from 8 to 10 nm). The relativepopulation of each family and the average size of macropores could be easily controlled by adjustingthe DHN quantity.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction

The conjugation of macropores with interconnected mesoporesframework and controlled micro texture are required in an increas-ing number of functional advanced materials [1]. For instance, incatalysis the larger surface density of active sites is often achievedby the high specific surface of the micropores (<2 nm), while themesopores (2-50 nm) and macropores (>50 nm) substantially facil-itate the mass flow and reduce the transport limitations [2]. In-deed, the transport of small molecules inside the macroporousmaterials has diffusion rates comparable to those achieved in openmedium. Additionally, in order to attain a higher performance,such material should possess macropores interconnected withmesoporous walls, i.e. hierarchically organized and not just segre-gated pores. This becomes particularly important for chemical pro-cesses involving large molecules or viscous liquids, in which thediffusion rates are low. Consequently, large molecules would notonly interact at their external surfaces, but throughout the bulkof the porous material.

Most synthetic methods used in the production of porous solidsare based on multi-step processes through repeated templating, inwhich a porous solid material, e.g. mesoporous silica MCM-48 orSBA-15 is used as a sacrifice hard template [3]. However, thesetemplates need to be previously synthesized by soft templatingprocedures that usually use long chain surfactant micelles, henceonly a limited range of pores size can be mimicked.

ll rights reserved.

: +55 16 33016692.

An emerging technique for the production of inorganic solidsshowing macroporous architectures involves the combination ofthe sol–gel process with emulsions as soft template [4]. This tem-plating method consists of the emulsification (dispersion of oildroplets in a continuous aqueous phase, see Scheme 1) of the inor-ganic precursors solution, followed by the gelation of the solaround the oil droplets and the subsequent conversion of the gelinto xerogel by drying [5]. This method takes advantage of the factthat the oil droplets are both highly deformable and easily remov-able. In addition, emulsification conditions can be adjusted to pro-duce different sized droplets. Irrespective of the self-assemblingsurfactants used to direct the structure of the mesopores [6], alarge extent can be reached, thereby allowing the independentcontrol of macro- and mesopore dimensions.

Emulsified systems are non-equilibrium assemblies that canundergo phase separation into two liquid layers. However, theycan be kinetically stabilized by surfactants that form films at thesurface (Scheme 1b) [7]. In this case, the surfactants play two mainfunctions: (1) to alter the energetic situation at oil/water interface(O/W) increasing the emulsion stability and (2) to form micelles,thus working as soft template for the mesopores formation(Scheme 1a). Furthermore, the addition of a co-surfactant, usuallyan alcohol of medium chain length, can also lead the emulsion to along-term kinetic stability or to a thermodynamically stable micro-emulsion [7].

Nowadays various approaches regarding the synthesis of meso-porous alumina have been described including the self-assembly ofanionic, cationic or neutral surfactants. They all provide mesopor-ous aluminas that offer structural features attractive in several

mesopores

macropores

d

Cc

1. pH correction2. aging3. calcination

Protectionagainst

coalescence

Facilitated diffusion

(from macropores to mesopores)

b

oil droplet

micellesurfactant

sol a

1

2

1

2

Csol a

1

2

1

2

Scheme 1. Dual soft templating method used to obtain solids presenting hierarchical macro–mesopores: (a) Oil droplets and surfactant micelles dispersed in aqueous sol, (b)surfactant at the droplets interface, (c) porous solid obtained after gelation, aging and calcination and (d) macro–mesoporous solid presenting facilitated diffusion frommacropores to mesopores (from 1 to 2).

L. Martins et al. / Microporous and Mesoporous Materials 132 (2010) 268–275 269

applications [8]. Among various synthesis routes, those employingpoly(ethylene oxide)–poly(propylene oxide)–poly(ethylene oxide)block copolymers [(EO)x(PO)y(EO)x] as soft templates attract atten-tion because they are inexpensive, commercially available, and af-ford materials with relatively uniform pores. In the case of PluronicP123 (EO20PO70EO20), the hydrophilic PEO blocks are found to bemuch shorter as compared to the hydrophobic PPO portion, henceleading to a decrease in the repulsion among the PEO–PPO–PEOchains, and thereby favoring the formation of large micelles [9].

Herein, we present a straightforward method of synthesizingporous aluminas through the conjugation of the sol–gel route aswell as a dual soft template technique that consists of dispersedoil droplets and block copolymers micelles. By using this strategy,a series of aluminas with hierarchical macro–mesopores and pre-senting controlled amount of macroporosity are readily obtained.

2. Experimental

2.1. Macro–mesoporous aluminas synthesis

Pluronic P123 (Maverage = 5800 g/mol, EO20PO70EO20), aluminumiso-propoxide (AlðOPi

rÞ3) and decahydronaphthalene (C10H18, DHN)were purchased from Aldrich and Sigma–Aldrich, nitric acid(65% m/m) and n-pentanol from Vetec and ammonium hydroxidefrom Mallinckrodt. All of them were used as received.

In a typical synthesis, 1.36 g of the Pluronic P123 surfactant wasdissolved in 8.97 g of Milli-Q water at room temperature. Aftercomplete dissolution, 0.17 cm3 of n-pentanol was added as co-sur-factant. Then 3.18 g of aluminum iso-propoxide was further addedinto the above solution under magnetic stirring followed by theaddition of 2.25 g of HNO3 (caution: highly exothermic reaction!).The mixture was then covered with polyethylene film and stirredat room temperature for 5 h. HNO3 was used to catalyze the hydro-lysis of the alkoxide to produce aluminum hydroxide: Al(OR)3 +3H3O+ ? Al(OH)3 + 3ROH + 3H+ [10,11]. A homogeneous and clearsol was obtained with the following molar composition:

0.015Plunonic P123:1AlðOPirÞ3:0.1n-pentanol:1.5HNO3:35H2O

After the sol preparation, emulsification was performed by add-ing decahydronaphthalene (DHN) under stirring and dispersed for5 min (Scheme 1a). The quantity of DHN used was made to varybetween the range of 15–80%, according to the formula %DHN ¼

massDHNmassDHNþmasssol

� 100:

Gelation was induced by adding 1.4 cm3 of NH4OH solution(29 wt.%) drop by drop into the above medium under mechanicalstirring. This caused an increase in pH from �3 to �4, inducingthe sol–gel transition. The gelated emulsions were aged at 25 �Cfor 7 days in closed flasks, before drying at 50 �C for 2 days. Calci-nation was then carried out in a conventional muffle oven byincreasing the temperature 5 �C/min from room temperature to190 �C for 2 h (for the removal of DHN) and heating (1 �C/min) to600 �C for 2 h.

The molar quantity of the co-surfactant n-pentanol was made tovary as follows: 0.015, 0.1 and 0.3, resulting in n-pentanol/pluronicmolar ratio of 0, 1, 7 and 20, respectively. The quantity of DHN wasalso made to vary within the following range: 15, 30, 50, 60, 70 and80% of the sol mass (n-pentanol/pluronic molar ratio fixed to 7).Based on these quantities, the samples were named as Al-x, wherex denotes DHN weight percentage. Occasionally the samples weredenoted as Al-x (Ry), where y stands for n-pentanol/pluronic molarratio.

2.2. Emulsion characterization

The oil droplets present in the emulsions were observed in anoptical microscope (Olympus BX41) before gelation. For observa-tion purposes, a portion was put on a sample holder and then ob-served directly without further treatment at room temperature.Droplet-size distributions were determined by evaluating at leastfifty droplets in microscopic images. The electrical conductivityof emulsified alumina sol with the desired composition was mea-sured at 22 �C using a Marconi conductivity meter (Model CA-150).

2.3. Macro–mesoporous alumina characterization

The crystalline phases present in calcined samples were ana-lyzed by X-ray diffraction powder (XRD) using a Siemens D5000diffractometer, Cu Ka radiation selected by a curved graphitemonochromator. The phase identification was done using the pro-gram X’Pert High Score and the crystallographic pattern files [70-238] and [04-0875] for c-Al(OH)3 and c-Al2O3, respectively.

The thermal behavior of the dried solids was investigated bythermo-gravimetric analysis (TGA), performed from room temper-ature up to 800 �C under oxygen flow (50 cm3/min), using aSDT600 TA Instruments at a heating rate of 10 �C/min.

Nitrogen adsorption–desorption isotherms were recorded at li-quid nitrogen temperature and relative pressure interval between0.001 and 0.998 on the equipment supplied by Micromeritics

270 L. Martins et al. / Microporous and Mesoporous Materials 132 (2010) 268–275

(ASAP 2010). Samples were evacuated prior to measurements at200 �C for 12 h under vacuum of 10 lPa. Surface areas were calcu-lated following the BET equation [12].

The pore size distribution was then determined from mercuryintrusion porosimetry using the AUTOPORE III equipment(Micromeritics). All the samples were degassed before analysis ata vacuum pressure below 50 lPa. The pore diameter was calcu-lated from the Washburn equation [13], using surface tensionand contact angle of 0.489 N/m and 135o, respectively. The aluminamacropores were observed by scanning electron micrographs,using a Philips XL 30 equipment. The samples were further depos-ited on aluminum sample holder and sputtered with gold.

3. Results and discussion

3.1. Effect of co-surfactant/surfactant ratio

A series of emulsions containing aluminum iso-propoxide and30% of DHN were synthesized and the kinetic stability of the O/W emulsions before gelation was monitored as a function of thePluronic P123 block copolymer addition. Prompt macroscopic

Fig. 1. (a) Optical micrographs of emulsions before gelation prepared with 30 wt.% of DHNdroplet diameter of emulsions as a function of n-pentanol/Pluronic P123 ratio.

phase separation was observed in the system emulsified in the ab-sence of this surfactant. In contrast, a huge increase in kinetic sta-bility of the DHN droplets was achieved due to the addition ofPluronic P123. Here, a macroscopic flotation of the polar phasewas found to occur only after 12 h of the emulsification step. A fur-ther increase in the emulsion stability can be achieved with theaddition of the co-surfactant n-pentanol.

The effect of the co-surfactant/surfactant molar ratio (n-penta-nol/Pluronic P123) on the structural feature of emulsions contain-ing aluminum iso-propoxide and 30% of DHN is depicted in theoptical micrographs shown in Fig. 1a. Clearly, an increase in then-pentanol content dramatically reduces the droplets averagediameter (Fig. 1b), thereby modifying their size distribution. Thisfeature evidences the amphiphilic nature of n-pentanol, i.e. a shorthydrophobic chain with terminal hydroxyl group. This makes it tointeract with surfactant monolayers at the interface of oil droplets,affecting the packing of the interface and in turn influencing thecurvature of the interface as well as the interfacial energy. Thenet result is a more densely packed interfacial layer with strongerlateral interactions between the surfactant and co-surfactant mol-ecules, which provides a barrier to the collision coalescence and,

and with different n-pentanol/Pluronic P123 M ratio of 0, 1, 7 and 20, and (b) mean

L. Martins et al. / Microporous and Mesoporous Materials 132 (2010) 268–275 271

thus, droplets with narrow distribution sizes are obtained as theproportion of n-pentanol is increased.

In this studied system, the surfactant and the co-surfactantagents do not offer long-term stability against coalescence, how-ever, they initially stabilize the O/W emulsions for the period oftime necessary to proceed with gelation by increasing the pH from3 to 4. Once the abrupt addition of ammonium hydroxide solutionresults in fast reaction, consequently leading to uncontrolled phaseseparation of the organic and inorganic phases as well as the con-siderable precipitation of aluminum hydroxide. This step was car-ried out slowly drop by drop until gelation. No macroscopic phaseseparation was observed in the emulsion within a time period ofmore than one month after emulsification and gelation. Thislong-term stability can be attributed to the accumulation of con-densed aluminum hydroxide shell at the O/W interface [14,15].As condensation of the aluminum hydroxide proceeds during gela-tion with ammonium hydroxide, the viscosity of the emulsion in-creases, thereby improving the kinetic stability of the emulsifiedsystem.

Fig. 2 shows the thermogravimetric curve and the weight deriv-ative for the partially dried (50 �C) Al-30 (R7) sample. Weight lossbelow 70 �C corresponds to propanol and water evaporation. Thissample exhibited about 50% weight loss at around 190 �C, whichcorresponds to the DHN boiling point. From 190 to 600 �C

Fig. 2. (a) Thermogravimetric curve and (b) weigh

Fig. 3. Pore size distribution and cumulative pore volume determined by Hg-porosimetrand (d) 20.

(Fig. 2b), the weight loss was close to 15%. This was attributed pri-marily to the decomposition of the surfactant (250 �C) as well asthe release of water or iso-propanol formed from the secondarycondensation reaction of aluminum hydroxide and/or aluminumiso-propoxide present in the alumina framework (condensationto form Al–O–Al bridges at 360 �C). These two decomposition stepswere confirmed by testing a reference sample (Al-ref), which wassynthesized without surfactant and DHN. X-ray diffraction pat-terns of samples before and after thermal treatment at 600 �C(not shown) also confirmed the transition of c-Al(OH)3 to c-Al2O3.

Considering the TGA result and the relatively high differencebetween the vapor pressure of DHN (1.5 mm Hg at 22 �C) andwater (23.8 mm Hg at 25 �C), we conclude that the free waterand alcohol present in the emulsified gel can be evaporated by dry-ing at 110 �C, while keeping DHN in the pores. As the pores arefilled with surfactant and DHN, careful calcination is required inorder to avoid a catastrophic damage of the porous structure.Moreover, further condensation of Al–O–Al bridges that occursduring firing is probably decisive to achieve stable porous alumina.This is the reason why we have decided to perform slow samplecalcinations, i.e. at 1 �C/min.

The pore volume and size distribution of the calcined aluminasevaluated using mercury porosimetry revealed three intrusionsteps (Fig. 3), being two in the domain of macropores (i) >7 lm

t derivative for the dried Al-30 (R7) sample.

y of Al-30 series with different n-pentanol/Pluronic P123 M ratio: (a) 0, (b) 1, (c) 7,

272 L. Martins et al. / Microporous and Mesoporous Materials 132 (2010) 268–275

and (ii) from 0.1 to 1.1 lm and a third one in the domain of mes-opores (iii) from 3 to 15 nm. Two reference samples prepared with-out surfactant and DHN (Fig. 4a) and only with surfactant (Fig. 4b)were also included in this analysis. The former one presents only alow mercury intrusion volume in the region of 7–50 lm, which isalso present in all of the samples. It indicates that this large porefamily does not arise from the micelle/emulsion template process,being probably associated with the presence of cracks or air bub-bles formed during the xerogel processing. Furthermore, the poresize distribution of the sample prepared with the addition of sur-factant (Fig. 4b) occurs within the mesopore range, nonetheless,the modal pore size obtained is equivalent to that of the value ob-served for all the samples. Regarding these features, we attributedthe origin of the second macropores family as well as the mesop-ores to the oil droplets and to the micelle templating process,respectively. This provides evidence for the additive contributionof O/W emulsion and surfactant micelle on the total porosity, thusdemonstrating the viability of this method in the production ofceramic materials with hierarchical porous structure.

Table 1 shows some characteristics of the porous aluminas suchas macro/mesopore volume and size, including the BET surfacearea. The aluminas macropore sizes that were generated as a resultof the presence of DHN have been adjusted in the range of 0.1–1.1 lm, but these observed pore diameters are very below thanthose exhibited by the emulsions before gelation. Two hypothesescan be outlined to explain this result: (1) after calcinations, theshrinkage of large pores occurs thereby reducing the overall poros-ity and (2) the emulsion droplets size produced at the end of gela-tion probably diminishes as a consequence of the drying shrinkageas well as the condensation of aluminum hydroxide shell aroundthe DHN droplets. Unfortunately, the last hypothesis could not beconfirmed yet because the samples turned to opaque after gelationand the droplets could not be observed by optical microscopy.

It was also found out that the co-surfactant addition in theemulsion did not considerably affect the final size of the macroporefamily templated by oil droplet. However, the molar ratio of 7 wasthe one that optimized the formation of macropores (Vmacro/Vme-

so = 0.7) and gave the highest BET surface area (339 m2/g). The low-

Fig. 4. Pore size distribution and cumulative pore volume determined by Hg-porosimsurfactant.

Table 1Pore volume of Al-30 samples synthesized with different n-pentanol quantity.

Sample Vmeso (cm3/g) Vmacro (cm3/g) VmacroVmeso

R0 0.36 ± 0.02 0.20 ± 0.01 0.6R1 0.59 ± 0.03 0.28 ± 0.01 0.5R7 0.51 ± 0.03 0.36 ± 0.02 0.7R20 0.42 ± 0.02 0.16 ± 0.01 0.4

a Pore diameter determined at maximum DV/Dlog(D).

er macroporosity was obtained for the sample prepared with co-surfactant/surfactant molar ratio of 20.

3.2. Effect of DHN quantity

The influence of the volume of DHN on the droplet diameter andsize distribution was also investigated. The co-surfactant/surfac-tant molar ratio was kept constant at 7. The average droplet diam-eter was found to increase from ca. 4 to 82 lm as the quantity ofDHN increases from 15% to 80% (Fig. 5). Also, for the higher DHNquantity, the droplet diameter shows a broader distribution(Fig. 6) while the relative amount of droplets was found to dimin-ish. Since the macroscopic separation into two liquid layers wasnot observed even for the highest DHN addition, it leads us to be-lieve that the O/W structure changes to a W/O structure, where theaqueous droplets are dispersed in the DHN continuous phase.

In order to confirm the compositional domains of W/O and O/Wemulsions formation, the electrical conductivity was measuredafter each DHN addition. These measurements show the occur-rence of three different behaviors of the specific conductivity as afunction of the amount of DHN, as clearly shown in Fig. 7. ForDHN < 70%, the specific conductivity decreases continuously withthe oil addition. This behavior is typical of the decreasing amountof continuous water phase in O/W structure. The conductivity pre-sents a small linear decrease with the DHN addition in the70% < DHN < 77% region (see the amplified region shown in the in-set of Fig. 7). This kind of behavior is typical of emulsions present-ing a bi-continuous structure for which aqueous and oil domainsare randomly interconnected to form a sponge like microstruc-tures, where both oil and water act as continuous phases whileremaining distinct from each other. Bi-continuous phases may ex-ist in the intermediate phase region between O/W and W/O emul-sions. The further increase in the amount of added DHN produces asharp decrease in the specific conductivity, attesting to the factthat the aqueous phase becomes discontinuous rightly as expectedof a W/O emulsion structure.

The effect of DHN quantities on the cumulative and differentialpore size distribution of fired aluminas is shown in Fig. 8, while the

etry of the reference samples (a) without surfactant and DHN and (b) only with

Dmacroporea (lm) Dmesopore (nm) SBET (m2/g)

0.40 ± 0.02 8.0 ± 0.4 225 ± 70.30 ± 0.02 8.0 ± 0.4 254 ± 80.70 ± 0.04 8.0 ± 0.4 339 ± 100.30 ± 0.02 10.0 ± 0.5 247 ± 7

Fig. 5. Optical micrographs of emulsions before gelation prepared with different weight percentage of DHN and fixed n-pentanol/Pluronic P123 M ratio of 7.

Fig. 6. Mean droplet diameter of emulsions as a function of the DHN weightpercentage.

Fig. 7. Normalized electrical conductivity of emulsion as a function of the DHNquantity. The insert shows the amplified region of DHN > 70%.

L. Martins et al. / Microporous and Mesoporous Materials 132 (2010) 268–275 273

values of the porous structure characteristic parameters are dis-played in Table 2. The contribution of the DHN in the formationof macropores becomes detectable by mercury intrusion porosime-

try for the sample with initial DHN contents above 15%. The aver-age macropores diameter increases monotonically from 0.7 to1.7 lm as the initial amount of DHN in the emulsion increasesfrom 30% to 60 %, respectively. As already mentioned, the macro-pore diameters are very below than those exhibited by the micro-structures of emulsions before gelation. Otherwise, in this DHNrange, the average mesopores size formed by the micelles templateprocess stays essentially independent on the amount of oil in theinitial emulsion. This demonstrates how viable the fine-tuning ofthe structural feature of the macropores family is, once it doesnot produce significant changes on the mesoporous sub-structure.

Fig. 9 compares the effect of the initial amount of DHN in theemulsified sol on the contribution of each pores family to the totalpores volume. For DHN <60% the mesopore volume of the sampleswas effectively independent on the amount of DHN. However, themacropore volume increases slightly for the DHN range between0% and 50% and strongly between 50% and 70%. The total pore vol-ume and the specific surface area (SBET in Table 2) reached maxi-mum values for the DHN amount of 60%. The transformation ofO/W into W/O emulsion structure verified for the DHN additionsuperior to 70% may explain the apparent reduction in the macro-pore volume, which is in fact the consequence of the dominantcontribution of the pore size superior to 10 lm (Fig. 8) of the totalporosity.

Representative scanning electron microscopy (SEM) images ofthese aluminas (Fig. 10) provide evidence for the presence of mac-ropores for the samples prepared with DHN contents superior to15%. Alumina morphology is composed of circular cross sectionpores with diameter varying from 0.1 to 2 lm. The surface densityof macropores increases continuously as the DHN amount in-creases from 30% to 70%. In SEM images, it is interesting to notethat after calcination in air at 600 �C, the macropores of thesematerials were retained and no significant amounts of crackingwere observed. Furthermore, the structure of the walls separatingthe macropores reveals the presence of a porous texture, consistentwith the existence of the mesoporous family of approximately9 nm (Table 2). This hierarchical pore structure with intercon-nected macropores and texture walls can enhance the permeabilityof fluids through the porous ceramic, making this material verysuitable for several applications.

It is noteworthy that this hierarchical pore structure presents agood thermal-stability as revealed by the analysis of sample Al-60

Fig. 8. Differential pore size distribution and cumulative pore volume determined by mercury porosimetry of samples prepared with different DHN quantity: (a) Al-15, (b) Al-30, (c) Al-50, (d) Al-60, (e) Al-70 and (f) Al-80.

Table 2Pore volume of samples synthesized with different DHN quantity (co-surfactant/surfactant molar ratio of 7).

Sample VDHN/VSola Vmacro

VmesoDmacropore

b (lm) Dmesopore (nm) SBET (m2/g)

Al-ref 0 0.1 nd nd 362 ± 11Al-0 0 0.3 nd 8.0 ± 0.4 352 ± 11Al-15 0.2 0.3 nd 9.0 ± 0.5 276 ± 8Al-30 0.5 0.7 0.7 ± 0.1 8.0 ± 0.4 339 ± 10Al-50 1.1 1.2 1.2 ± 0.1 8.0 ± 0.4 267 ± 8Al-60 1.7 3.1 1.7 ± 0.1 9.0 ± 0.5 473 ± 14Al-70 2.6 30.1 1.3 ± 0.1 nd 417 ± 13Al-80 4.4 6.8 >10 nd 391 ± 12

a Volume DHN per volume of alumina sol.b Macropore diameter determined at maximum DV/Dlog(D).

Fig. 9. Pores volume as a function of DHN weight percentage, for samples preparedwith n-pentanol/Pluronic P123 M ratio of 7.

274 L. Martins et al. / Microporous and Mesoporous Materials 132 (2010) 268–275

treated at different temperatures until 1100 �C. Despite the c-Al2O3

to a-Al2O3 transformation occurring near to 1000 �C, the macro-and mesopores families were stable up to this temperature. At1100 �C the mesopores collapse, however the shrinkage of themacroporous framework does not take place according to mercuryintrusion porosimetry measurements.

As expected from the W/O structure of emulsion prepared with80% of DHN, the Al-80 sample shows a different morphology,apparently composed of small agglomerates. Besides, the opticalmicroscopy images of emulsions Al-70 have shown the presenceof sparse multiple emulsions domains (Fig. 11). Multiple emulsifi-

cation (designed as W/O/W) is composed of water droplets dis-persed in larger DHN droplets, which are then dispersed in thecontinuous phase (water) [8]. Due to the variety of interfaces, mul-tiple emulsions are inherently less stable compared to the simpleemulsions [8], nonetheless, this emulsion microstructure takesplace when the emulsion nears the borderline that defines the do-

Fig. 10. Scanning electron micrograph of alumina samples prepared with different DHN mass percentage.

Fig. 11. Multiple emulsification observed in the preparation of sample Al-70.

L. Martins et al. / Microporous and Mesoporous Materials 132 (2010) 268–275 275

mains of W/O and O/W emulsions formation. Therefore, the imagedisplayed in Fig. 5 for the sample Al-80 is related to a W/O system,i.e. droplets of sol dispersed in DHN, conforming to the VDHN/VSol

higher than 4 as presented in Table 2. This inverted emulsion leadsto the formation of small alumina agglomerates with large interag-glomerate pores, which was confirmed with the mercury intrusionresults shown in Fig. 8f.

4. Conclusions

Macro–mesoporous aluminas presenting hierarchical structureof pores were synthesized using one-pot pathway based on thedual templating process of pores by micelles and oil droplets. Plu-ronic P123 micelles and decahydronaphthalene oil droplets wereused as soft template agents for the meso- and macropores struc-turing, respectively. A series of emulsions containing aluminumiso-propoxide and decahydronaphthalene were synthesized andthe emulsions micrographies have presented different oil dropletssize that was dependent on n-pentanol/ Pluronic P123 molar ratioand on decahydronaphthalene quantity. The templated bimodalpores size distributions were retained after xerogel–ceramic con-

version by firing at 600 �C. The well-defined mesopores familyhas shown an average size of 9 nm irrespective of the surfactant/co-surfactant ratio as well as the oil content used in the initial tem-plate step. The average pore size and the pore volume correspond-ing to the macropore family could be adjusted within the range of0.7–1.7 lm and 0.15–1.63 cm3/g, respectively, varying the initialoil content from 15% to 60%. The phase inversion from W/O to O/W structure occurs at 78% of oil leading to drastic structural mod-ifications of the porous alumina ceramic. The high specific surfacearea (�470 m2/g) associated to the hierarchical pore structure ofsuch alumina material may be useful in applications such as catal-ysis and separation processes, especially in macromolecules andviscous systems where large pores are needed to improve the masstransport throughout the pore structure. Results reported here areonly based on alumina, however, it should be possible to extendthis dual template method to other ceramic materials.

Acknowledgments

We are deeply grateful to CAPES/COFECUB, CNPq, and FAPESPfor the financial support offered.

References

[1] R. Backov, Soft Matter 2 (2006) 452.[2] A. Vantomme, A. Leonard, Z.Y. Yuan, B.L. Su, Colloids Surf. A 300 (2007) 70.[3] M. Tiemann, Chem. Mater. 20 (2008) 961.[4] M.A.A. Rosa, E.P. Santos, C.V. Santilli, S.H. Pulcinelli, J. Non-Cryst. Solids 354

(2008) 4786.[5] I. Akartuna, A.R. Studart, E. Tervoort, L.J. Gauckler, Adv. Mater. 20 (2008) 4714.[6] Z.Y. Yuan, B.L. Su, J. Mater. Chem. 16 (2006) 663.[7] D. Myers, Surfactant Science and Technology, third ed., John Wiley & Sons,

2006.[8] C. Marquez-Alvarez, N. Zilkova, J. Perez-Pariente, J. Cejka, Catal. Rev. – Sci. Eng.

50 (2008) 222.[9] S.A. Bagshaw, E. Prouzet, T.J. Pinnavaia, Science 269 (1995) 1242.

[10] N.Y. Turova, E.P. Turevskaya, V.G. Kessler, M.I. Yanovskaya, The Chemistry ofMetal Alkoxides, Kluwer, Academic Publishers, 2002.

[11] Q. Yuan, A.-X. Yin, C. Luo, L.-D. Sun, Y.-W. Zhang, W.-T. Duan, H.-C. Liu, C.-H.Yan, J. Am. Chem. Soc. 130 (2008) 3465.

[12] S. Brunauer, P.H. Emmet, E. Teller, J. Am. Chem. Soc. 60 (1938) 309.[13] E. Washburn, Phys. Rev. 17 (1921) 273.[14] R. Aveyard, B.P. Binks, J.H. Clint, Adv. Colloid Interface Sci. 100 (2003) 503.[15] B.P. Binks, Modern Aspects of Emulsion Science, The Royal Society of

Chemistry, 1998.