modifying the walls of mesoporous silicas prepared by supramolecular-templating

Ž .Current Opinion in Colloid & Interface Science 7 2002 92�106

Modifying the walls of mesoporous silicas prepared bysupramolecular-templating

Anne Davidson�

( )Reacti�ite de Surface UMR-7609 , Uni�ersite Pierre et Marie Curie, 75252 Paris Cedex 05, France´ ´ ´ ´

Abstract

The present review is not exhaustive and rather endeavors to illustrate ways to improve the wall stability of orderedŽmesostructured silicas in presence of water. In situ X-ray diffraction of the phase transformations of hybrid surfactant

. Žcontaining silicas show how synthesis temperature, pH and duration control wall chemistry thickness, condensation and. Ž . Žhydrolysis . Ex situ studies of calcined surfactant free, empty mesopores silicas show how wall stability in boiling water,

.under steam is affected by synthesis conditions and�or post-synthesis treatments. � 2002 Elsevier Science Ltd. All rightsreserved.

Keywords: Mesoporous; Silica; Supramolecular templating; Stability; Wall thickness

1. Introduction

ŽMesoporous silicas with regular mesopore diame-˚ .ters between 20 and 500 A, as defined by IUPAC

attract nowadays considerable interest in heteroge-� �neous catalysis 1�6 . Silica has no intrinsic catalytic

property but it can be used as support and thereforeŽfunctionalized either by framework substitution in-

. � �corporation of heteroelements during synthesis 5 ,Žor by post-synthesis surface modification impreg-

. � �nation for instance 6 . The main advantage of meso-Žporous silica supports, rather than zeolitic ones pore

˚.diameters lower than 20 A is their ability to treatbulkier reactants and obtain larger products. Anotheradvantage is the possibility to graft large catalytic

Ž .sites Keggin heteropolyanions for instance , whilekeeping access for the diffusion of reagents. Com-pared to amorphous silica, the advantage of meso-porous supports may come from the specificities ofthe silica surface in mesopores. For instance, a Ti-grafted MCM-41 catalyst has been shown to be much

� Tel.: �33-144-27-60-04; fax: �33-144-27-60-33.Ž .E-mail address: [email protected] A. Davidson .

more stable toward leaching than a Ti-grafted amor-� �� phous silica 7 .

Mesostructured silica materials are templated bysupramolecular aggregates of surfactant. Severalrecent reviews describe the mechanisms of their

� �formation from gel precursors 1�3 . Whatever themechanism involved, subsequent aggregate removalŽ .by calcination for instance generates well-orderedmesopores separated by amorphous silica walls. Thesensitivity of these walls to water and the fact thatsome of their siloxane Si�O�Si bonds may be irre-

� � �� versibly hydrolyzed, is well documented 8 ,9 . Be-Žcause of the role of water in catalysis as a solvent for

preparation by impregnation or for catalytic reactionsin the liquid phase, as a by-product of oxidation

.reactions in the gas phase , it is important to obtainmore information about how mesoporous silicas areaffected by water, depending upon the experimentalconditions used to obtain them. In this review, we willfocus mainly on recent progress in the synthesis andpost-synthesis modifications of mesoporous silicas thatminimize their structural degradation due to water.

Ž .We will discuss recent 1998�2001 experiments re-Ž .lated to i information about silica walls gathered by

1359-0294�02�$ - see front matter � 2002 Elsevier Science Ltd. All rights reserved.Ž .PII: S 1 3 5 9 - 0 2 9 4 0 2 0 0 0 1 1 - 0

( )A. Da�idson � Current Opinion in Colloid & Interface Science 7 2002 92�106 93

Table 1� �Abbreviations used in the literature for some mesoporous silicas obtained by supramolecular templating 1�3

Symbol Signification Symmetry

Ordered mesoporous MCM-50 Mobil Composition of Matter L �silicas templated bycationic surfactants MCM-41 41: 2-D hexagonal p6mm

MCM-48 48: cubic Ia3d

Ordered mesoporous SBA-15 University of California at Santa Barbarasilicas templated byneutral copolymers 15: 2-D hexagonal p6mmŽtriblock SBA-12 12: 3-D hexagonal P6 �mmc3

.copolymers SBA-16 16: cubic Ia3dSBA-11 11: cubic Pm3mSBA-1 1: cage-type, cubic Pm3n

Disordered KIT-1 Korea Advanced Institutemesoporous of Science and Technologysilicas 1: disordered material

MSU-X Michigan State University

in situ studies of the formation and phase transforma-Ž .tions of hybrid surfactant-containing MCM-41 mate-

Ž .rials, and ii information about silica walls in calcinedŽ .surfactant-free mesoporous silicas by ex situ charac-terization techniques.

The assembly of silica precursors and surfactantsgenerates hybrid mesostructured silicas of lamellar,

Ž� �hexagonal, cubic or disordered morphologies 1�3 ,.Table 1 . Whatever the morphology, the nature of the

interaction between silica precursors and surfactantsmainly determines wall thickness and further stability.

This will be illustrated here by comparing two hexago-nal mesoporous silicas: on the one hand MCM-41,templated by cationic surfactants; and on the otherhand, SBA-15, templated by neutral copolymers. Forinformation, the relative numbers of publications de-voted to these two kinds of mesoporous silicas duringthe period 1998�2001 are compared in Fig. 1. It isclear that the methods described here to modify thewalls of mesoporous silicas should not only apply tohexagonally ordered ones but most probably also toother morphologies.

Fig. 1. Up-dated plot of the number of papers published in the field of mesoporous silicas obtained by supramolecular templating between1998 and 2001. The above data were obtained from current content searches for the keywords: silica and MCM-41; silica and SBA-15; andsilica and ‘mesoporous or mesostructured’ in papers title, abstract and text.

( )A. Da�idson � Current Opinion in Colloid & Interface Science 7 2002 92�10694

2. Information about the silica walls of hybrid MCM-41

Hexagonal MCM-41 is templated by a cationic am-Žphiphile surfactant cetyltrimethylammonium,

� .CTMA , for instance . Due to the positive charge ofŽ .the surfactant head group, a basic pH approx. 10�11

significantly higher than the isoelectric point of silicaŽ . �2.5 is used to generate Si�O oligomeric silicates inthe early stages of the synthesis. The presence of

� �these oligomers, assumed in several publications 1�3� � �has been clearly demonstrated only recently 10 .

The pH also controls the silica condensation rate,which is maximum at pH 5�7 and abruptly decreasesat approximately pH 8�9.

A poorly condensed MCM-41 hybrid, formed in lessthan 2 h at room-temperature, undergoes phase

� �� transformations upon thermal treatments 11,12 ,13 .Depending upon experimental conditions, it can be

� �� transformed either into a lamellar phase 11,12 or� �into cubic MCM-48 13 . The MCM-41�lamellar and

MCM-41�MCM-48 phase transformations, althoughdriven by the surfactants, are only possible if some ofthe Si�O�Si bonds are broken. The tendency of agiven MCM-41 hybrid to undergo phase transforma-tion under varied conditions can be studied. One cantherefore understand how not only simple experimen-

Žtal factors thermal treatment temperature, pH and.duration , but also details of temperature profiles

Žcontrol the inorganic walls chemistry thickness, con-.densation and hydrolysis .

The MCM-41�lamellar phase transformation is in-duced by heating treatments performed in NaOH

� �� aqueous solutions 11,12 . This phase transforma-tion is driven by the conformational disorder of thesurfactant alkyl tail. This tail motion increases withtemperature, so that the overall shape of the surfac-tant changes and its hydrophobic volume increases.Both volume and shape changes produce an expansiveforce that induces the phase transformation. In situ

Ž .X-ray diffraction at small angle SAXS has shownŽ .Ozawa method that an activation energy of 163�3kJ�mol was necessary to transform the MCM-41 hy-

Ž .brid at low pH NaOH, 0.115 M into a lamellarphase, whereas an activation energy of 106�3 kJ�mol

Ž .was found at higher pH NaOH, 0.225 M . This showsthat the wall thickness of MCM-41 silicas increaseswhen pH decreases since larger silicate oligomersinteract with organic aggregates when the silica solu-bility decreases. Furthermore, chemical changes areinduced within the silica walls by the heating process.Indeed, at the highest pH, the activation energy in-creases when heating processes are performed with

slower ramps, whereas at the lowest pH, the activa-tion energy decreases with slower ramps. Furtherconfirmations of this were obtained by ex situ 29 SiMAS NMR experiments. This technique revealed thatthe MCM-41 hybrid obtained at the lowest pH startsout with the most condensed silica framework. How-ever, at the onset of the MCM-41�lamellar phasetransformation, its degree of condensation decreasesmore rapidly than the one of the sample prepared atthe highest pH. It therefore contains more strainedsiloxane bridges.

The MCM-41�MCM-48 phase transformation canbe induced when the mesoscopic ordering of thehybrid MCM-41 is altered by alcohol molecules, forinstance by the 4 EtOH molecules liberated whentetraethoxyorthosilicate, TEOS, is used as a silica

� �source 13 . The time necessary to complete theMCM-41�MCM-48 phase transformation increaseswith temperature, which indicates an increasing acti-vation barrier as silica wall condensation proceeds.Furthermore, after calcination, the MCM-48 pro-duced at 190 �C has approximately the same pore

˚Ž .diameter as that produced at 100 �C 24�25 A ,whereas its lattice parameter is significantly larger

˚Ž .81.81 compared to 72.01 A . Silica walls are thensignificantly thicker in the material prepared at 190�C.

The phase stability of hybrid MCM-41 upon high� �� pressure has also been followed by SAXS 14 . Up

to 12 GPa, the mesoscopic hexagonal order is re-Žtained and pressure-induced distorsions unit-cell

.shrinkage are reversible. Beyond 12 GPa changesbecome irreversible. Stability tests have beenperformed before and after 4 days of hydrothermal

Ž .treatment HT at 100 �C. More condensed wallscorrespond to the higher bulk moduli. This increase isnot due to an improved ordering of Si atoms withinthe walls. In fact, the more condensed materials alsohave the broadest 29 Si MAS NMR peaks and there-fore contain a wider distribution of inter-tetrahedralSi�O�Si angles. In other words, the hydrothermallytreated sample contains more siloxane strained bondsthan the as-synthesized sample.

3. Information about the silica walls of calcinedmesoporous silicas

Determining the long-range order of mesoporousŽsilicas by X-ray diffraction is rather easy diffraction

angles and relative intensities, full-width at half maxi-.mum of diffraction peaks . Obtaining the condensa-

tion degree within the silica walls is also a relatively


easy task because 29 Si MAS NMR spectra provide the4 � Ž . � 3molar ratios of Q fully condensed Si OSi to Q4

� Ž . �partially uncondensed HOSi OSi species. The3larger the Q4�Q3 ratio, the higher the condensationdegree within the walls. Furthermore, a unit-cellshrinkage is systematically observed by XRD upon

Ž .calcination surfactant removal . This shrinkage issmaller for more condensed silica walls. Determiningthe thickness of silica walls is more involved becausethe mesopore diameters are required. These diame-ters are obtained by analyzing N sorption experi-2ments. Among the calculations methods used so far,

Žthe most popular one, called BJHd Barrett, Joynerand Halenda, applied to the desorption branch of the

.isotherm , which is based on the Kelvin equation, isknown to underestimate mesopore diameter by 20%� � Ž15 . Several corrections BJHa: BJH applied to the

.adsorption branch of the isotherm and more sophisti-Žcated methods BdB, Broekhoff, de Boer; KJS, Kruk,

Jaroniec, Sayari; HRS, hydrolytic radius concept, GM,geometrical model; DFT, based on non-linear func-

.tional density theory have been proposed recently� �� 15�17 ,18 . The mesopore diameters obtained byapplying these different methods differ by less than20% for MCM-41 silicas, the detected differencesreflecting mainly the physical assumptions introducedby each method. The mesopore diameters obtain forSBA-15 silicas more widely differ, due to the presenceof intra-wall porosity. In Tables 2 and 3 in whichstructural data on several mesoporous silicas are sum-marized, the method used to calculate mesopore di-ameters is mentioned. Wall thickness was calculatedfrom the difference between unit-cell parameter andmesopore diameter. The thickness thus obtained doesnot have any absolute meaning. Only comparisonsbetween thicknesses measured before and after agiven treatment and using the same method aremeaningful. This particularly applies to SBA-15 mate-rials.

Tables 2 and 3 also give information about stabilitytests in the presence of water, keeping in mind, how-ever, that there is no widely used procedure and thateach research group uses his own.

3.1. Hexagonal mesoporous silicas obtained by templatingwith ionic surfactants

Let us begin with the modifications of silica wallsintroduced by variations of the precursor gel chem-

Žistry in otherwise similar conditions identical temper-ature synthesis, pH, duration and same post-synthesis

. Ž .hydrothermal treatment, if any : i the chemicalŽ .nature of the cationic surfactant; ii the chemical

Ž .nature of the associated anion; iii the silica precur-Ž .sor; iv additives such as co-solvents that modify

surfactant chemistry or salts that modify the ionicŽ .strength; and v use of ultrasounds.

MCM-41 obtained with n-alkyltrimethylammoniumŽ .cations with different n-alkyl chains C �C have18 14

� �� been compared 9 ,19 . Longer alkyl chains yieldmore ordered and more stable mesoporous silicas butno difference in wall thickness is detected. This couldbe expected because longer alkyl chains, being morehydrophobic, produce more well-defined cylindricalmicelles. Also, since these surfactants have the samehydrophilic head group, they most probably interactsimilarly with silica precursors, which yields silicawalls of the same thickness. Preliminary resultsconcerning MCM-41 prepared with cationic surfac-

Žtants with larger head groups C -triethylammonium,22. � �for instance have also been recently reported 20 . To

obtain good quality MCM-41 with these surfactants aŽ .low pH 8 is required, indicating preferred associa-

tion with large silicate oligomers.The chemical nature and concentration of the an-

ions associated to the cationic surfactants is alsoimportant and is known to affect the long range order

� �of the mesostructured hybrids 21 . These anions areindeed partially incorporated in the hybrids, and boththeir nature and concentration controls the number

Ž . �of silanolate SiO Si�O groups required to balance3the positive charges of the surfactant head groups� � �22 ,23 . They participate to the charge matchingbetween silicate oligomers and organic aggregates. Aphase transition from lamellar to hexagonal phasetransition can be induced by mere addition of anions� � �22 . The solvation energy of the anions, in otherwords their hydrophobic character, affects the amountof water stabilized at the organic�inorganic interface.One may therefore expect that more hydrophobicanions generate a more hydrophobic interface, that isto say a lower silanolate density. This concept, first

� � �applied to inorganic anions 22 , has been recently� �extrapolated to organic anions 23 . The use of n-al-

kyltrimethylammonium cations, associated with vari-Žous aryl substituted anions toluene phosphonate, sul-

.fonate and carboxilate , yields mesoporous silicas of˚Ž .quite low wall thickness t�10�15 A that are re-

Žmarkably stable less of 20% loss in structural orderaccording to X-ray diffraction after 48 h in boiling

.water . The nature of the aryl group is probably animportant factor to control mesopore diameter sincethis hydrophobic group must be located within theorganic aggregates. Its effect on wall thickness andstability remains to be addressed.

Different silica sources have been used as precur-sors. TEOS and fumed silica yield to MCM-41 materi-

� �als of similar surface areas 24 . The materialsŽobtained from TEOS are less ordered two broad

. Žpeaks in XRD instead of five and less stable fully

()

A.D

a�idson�

CurrentO

pinionin

Colloid

&Interface

Science7

200292

�10696

Table 2Ž .Synthesis conditions, structural parameters and stability tests reported between 1998 and 2001 on selected hexagonal MCM-41 silicas powders

� �Refs. S X I Synthesis Stability tests Complementary RemarksStructural parametersŽconditions % of porous information4 3D V S d t Q �Qp BET 100 .volume decrease3 2˚ ˚ ˚Ž . Ž . Ž . Ž . Ž .A cm �g m �g A A

�� Ž .8 C Cl TEOS Basic NaOH , RT BJHd16Ž . Ž .a Freshly calcined 30 1120 36.6 12.3 1.5 Sample a overnight Weight gain for three Upon water adsorptionŽ . Ž .b - a rehydrated 30 712 0.6 in moist air: 35% months of hydration irreversible formation of

3 2Ž �overnight 3 months in moist air; approximately 58% Q species Q alsoŽ . Ž . . Ž .�c - b calcined � 697 0.5 100% detected for sample cŽ .773 K, 1 h

�� Ž .9 C Br TEOS Basic NH , RT HRS After measuring a water n-Alkylammonium No significant change12 323.4 0.59 1006 29.9 11.1 sorption isotherm cationic S with in wall thickness

Ž .2�3 weeks at RT: longer alkyl chains100% yields to more

C 27.9 0.70 1040 33.9 11.2 60% ordered and more14C 39.2 0.88 1059 42.3 9.6 40% stable materials18

�� Ž . Ž .30 , C Br Fumed Basic TMAOH , GM Reflux in water 24 h : When MCM-41 is used16�31 silica RT, 20 h then 39.6 0.89 918 42.8 9.8 1.4 64% as a seed, thickening of

HT 150 �C, 48 h Steamed at 900 �C: 100% silica walls withoutŽ . Ž .MCM- Basic TMAOH , 39.6 0.79 700 44.6 12.0 3.4 Reflux in water 24 h : important changes of d100

41 RT, 20 h then 50%HT 150 �C, 48 hIdem but HT 96 h 44.1 0.83 806 48.1 11.4 No measured differenceIdem but HT 168 h 41.6 0.59 613 49.2 15.2 �

� �29 Sodium BJHaŽ . Ž . Ž . Ž .silicate a Aged 4.5 h RT 0.39 833 Am. � Condensation degree b , c and f samplesŽ .b Sonicated 3.5 h 24.8 0.79 853 38.8 20.1 of silica increased; retain hexagonal orderRT unit-cell after surfactant elimination,Ž . Ž .d Aged 28 h then 22.6 0.92 931 36.8 19.9 contraction upon not a .sonicated 3.5 h RT Reflux 6 h in boiling calcination limitedŽ .f Aged 56 h RT 23.5 1.10 1042 35.4 17.4 6.7 water: 35%

()

A.D

a�idson�

CurrentO

pinionin

Colloid

&Interface

Science7

200292

�10697

Ž .Table 2 Continued

� �Refs. S X I Synthesis Stability tests Complementary RemarksStructural parametersŽconditions % of porous information4 3D V S d t Q �Qp BET 100 .volume decrease3 2˚ ˚ ˚Ž . Ž . Ž . Ž . Ž .A cm �g m �g A A

�� Ž .32 C BR Fumed Basic TMAOH Identical Si�Al in the16silica 20 h at RT then HT three samples�

then calcination � proportion of Td Al alsografting of Al equivalent.Ž .target Si�Al�13 GM Al grafting associatedŽ .a HT: 150 �C, 48 h 38.1 0.78 907 42.2 10.6 Steamed 900 �C, 75% with thicker silica wallsŽ .b HT: 140 �C, 96 h 45.6 0.76 749 50.7 12.9 65% gives a better stabilityŽ .c HT: 145 �C, 96 h 50.2 0.78 734 55.6 14.0 38% under steam.

Ž .� symbol indicates that the porous volume after stability test has been found larger than before test, most probably due to the appearance of some textural porosity. BJHa, GM, HRC:˚ ˚Ž . Ž .symbols relative to the calculation methods used to obtain D A and t A . BJH: Barrett, Joyner and Halenda. GM: geometrical model. HRC; hydrolytic radius concept, see ref.

� �� 16,17 ,18 ,19,20 and references therein.S�: surfactant; X�: associated anion; I: inorganic precursor.

()

A.D

a�idson�

CurrentO

pinionin

Colloid

&Interface

Science7

200292

�10698

Table 3Ž .Synthesis conditions, structural parameters and stability tests reported between 1998 and 2001 on different hexagonally-ordered mesoporous silicas powders

templated by neutral copolymers

�Refs. S I Synthesis RemarksStructural parametersconditions 4 3D V S d t Q �Qp BET 100

3 2˚ ˚ ˚Ž . Ž . Ž . Ž . Ž .A cm �g m �g A A

�� Ž .34 EO PO EO TEOS HCl, 35 �C, 20 h BJHa:47 0.56 690 104 96 64 1.3 XRD unchanged after 24 h in boiling water20 70 20Ž .HCl, 35 �C, 20 h 77 1.03 820 103 97 38 � Larger hydrophilic volume of the surfactant

HT: 80 �C, 24 h yields to thicker wallsŽ .HCl, 35 �C, 20 h 89 1.17 850 105 104 31 � Wall thickness decreasing and mesopore

HT: 100 �C, 24 h diameter increasing with the HT temperatureŽ .EO PO EO HCl, 60 �C, 20 h 59 1.19 950 81 80 34 �13 70 13

�� Ž .36 EO PO EO TEOS HCl, 35 �C, 24 h BJHd:40 0.70 800 102 83 57 2.05 Shrinkage of unit-cell parameter upon20 70 20calcination reduced by HT, without

4 3Ž .HCl, 35 �C, 24 h 60 1.06 900 102 95 50 1.55 noticeable variation of the Q �QŽ .HT: 100 �C, 72 h molar ratio broader peaks

2�� Ž .38 Brij 56 TEOS pH 7, Co HK: 43 0.68 937 61.4 18.4 EO complexation by the metallic cationC EO H cations18 10

� �47 EO PO EO TEOS HCl, 45 �C, 8 h BJHa17 58 17Ž .I�S�60, 80 �C, 58 565 88 41 Thicker walls for higher S�I ratio

8 hŽ .I�S�120, 80 �C, 47 401 92 62 and lower synthesis temperature

8 hŽ .I�S�120, 100 �C, 64 436 88 40

8 h

� �41 C EO H Sodium HCl, 1 day RT, BdB:12 9Ž .EO PO EO metasili- 1 day 100 �C 38 � 1055 56 10 � Stability tests not described yet20 70 20Ž .cate 76 � 600 126 31

˚Ž . Ž .HT, hydrothermal treatment; d A were determined on uncalcined materials. The values indicated under brackets in the same column were obtained after calcination surfactant removal .100S, surfactant; and I, inorganic precursor. The wall thickness was calculated by the difference between the unit-cell parameter after calcination and the mesopore diameter.


.destroyed after calcination at 1000 �C than the onesprepared from fumed silica. A possible explanation isthat, with TEOS, only small silicate oligomers mayform in the early synthesis stages and then condensa-tion within silica walls remains low. By contrast withfumed silica, that is itself composed of large oligomeric

Ž .silica particles of nanometer range larger silicateoligomers may form, yielding an increased condensa-tion degree within the silica walls.

The effect of organic additives such as butylamine� Žon a synthesis based on CTMA and TEOS basic

. � �pH, 35 �C, 1 h has been investigated in detail 25 .For a BuNH �CTMA� molar ratio ranging between2

2 and 15, a progressive decrease of unit-cell parame-Ž .ter was observed no phase transformation . This be-

havior results from a competition between protonatedamines and CTMA� head groups for adsorption ontonegatively charged silicates oligomers. A parallel slightdecrease in wall thickness is due to the pH increasedue to the amine protonation. Other short-chainamines and alcohols have been shown to increase ordecrease pore wall thickness. The corresponding arti-cles, devoted to formation mechanism elucidation, donot give further information about the influence ofadditives on materials stability. However, quick syn-theses performed at approximately room temperature,most probably yield MCM-41 of poor stability in thepresence of water. The same remark applies to thedisordered silicas with mesopore diameters as large as

˚135 A, which have been obtained by mesopore expan-Žsion using long-chain amines dimethyldecylamine and

.dimethylhexadecylamine . With these peculiar silicas,expensive structural collapse occurs upon calcination� �26 .

Adding salts during synthesis has also been shown� �� to increase wall thickness 27 ,28 . Whatever the salt

Ž .nature EDTANa for instance the ionic strength of4

the aqueous solution increases, inducing a strongscreening of the electrostatic interaction between sili-cate oligomers and surfactants. Consequently, there isa branching of micelles and a transition betweenordered MCM-41 and disordered KIT-1 is observed.

ŽUpon further salt addition salt�Si molar ratios larger.than 50 , only amorphous silica is obtained. The inter-

est of the MCM-41�KIT-1 transformation appearswhen comparing the stability of the calcined materi-als. A non-salted MCM-41 reference is decomposedafter heating in water for 12 h. In similar conditions, aKIT-1 sample, treated with salt during 1 day, resists12 h, whereas a KIT-1 sample treated during 10 days,has an even improved hydrothermal stability. Con-sidering the different behaviors of these two lastsamples, the stability increase is not ascribed to dis-order but rather to salting duration. The stability

increase due to salt addition is therefore stronglytime-dependent. It was even found that if the saltaddition procedure is properly controlled, mesophase

Ž .morphologies MCM-41, MCM-48 can be preserved� �� while silica walls are slowly restructured 27 . De-

spite stability differences between salted and non-salted samples, no significant change was observedeither in wall-thickness or in Q4�Q3 ratio, which isrelatively puzzling.

Well-organized MCM-41 can also be synthesizedfrom sodium silicate and CTMA� in a relatively short

Ž . � �time Table 2 by using ultrasounds 29 . The stabilityŽis improved due to the formation of thick approx. 20

˚ 4 3. Ž .A and well condensed Q �Q ratios as high as 6.7walls. The thickening of silica walls has been corre-lated with the rapid heating to crystallization temper-ature, due to volumetric heating, which results inhomogeneous nucleation and by the formation of hotspots within the organic�inorganic interface, whichaccelerate silica condensation.

Let us now consider now the influence on wall thick-ness of post-synthesis treatments, applied before surfac-tant remo�al by calcination. The effect of a 150 �C HTon a secondary MCM-41 prepared from a primary

Ž .MCM-41 Table 2 as a silica source has been investi-� �� gated by Mokaya 30 ,31 . Longer HT increases the

Žunit-cell parameter and wall thickness by more than.50% after a 168-h HT and decrease mesopore di-

ameter. Q4�Q3 molar ratios as high as 4.5 areobserved by 29 Si MAS NMR after HT. Interestingly,the NMR peaks observed after a 96-h HT are notbroader than those detected in absence of HT. Thick-ness and condensation improvements have then beenassociated with the addition of silicate oligomers tothe MCM-41 scaffolding via a diffusion-controlledprocess. It has even been proposed that the addedsilicate oligomers heal defects, resulting in a morestable and less strained silica framework.

The combined effects of hydrothermal treatment� �and pH have been studied by Maeda et al. 24 . A

hybrid was thermally treated, either in its motherŽ . Ž .solution pH 11�12 or in water pH 8�9 . The thick-

est walls were observed for the longest water treat-˚Žments applied at the highest temperature 13.8 A

.after 3 days HT at 165 �C . The unit-cell shrinkagedue surfactant removal by calcination is reduced for

Žthe thicker and more condensed walls this point has29 .been confirmed by Si MAS NMR . The higher

condensation degree is associated with an enhancedlong-range order, which suggests a further growth orliking of existing small MCM-41 crystals.

Finally, let us now discuss the post-synthesis modifi-cations applied to mesoporous silicas already calcined.Silylation techniques have been used to reduce the


number of silanols groups that decorate the internal� �surface of the mesopores 9 . The degree of silylation

increases the wall thickness and hydrophobocity andalso leads to an enhanced stability in presence ofwater. Another way to reduce the number of silanolsgroups of calcined MCM-41 silicas is to graft het-eroelements on their surface. An Al-grafted MCM-41

Ž .material obtained by using aluminum chlorohydrolhas been shown to have an improved stability in thepresence of water. For a given amount of grafted Alspecies and a given proportion of tetrahedral Al sites,thick silica walls were, however, necessary to observe

� � �a significant stability 32 . The influence of othergrafted heteroelements onto stability is still an openquestion.

3.2. Hexagonal mesoporous silicas obtained by templatingwith neutral copolymers

In the 1990s, Pinnavaia et al. used non-ionic surfac-tants in neutral conditions to prepare disorderedMSU-X mesoporous materials, with worm-like chan-

� �nels 1�3 . In 1998, the use of neutral surfactants hasbeen extended to ordered mesostructured silicas by

� �� Stucky et al. 33,34 . Several morphologies, eithersimilar to the ones observed with cationic surfactants

� � Ž .or different, have been described 33 Table 1 . Thehexagonal SBA-15 silicas are prepared in acidic con-

Ž .ditions HCl , TEOS being introduced as silica source�and a tri-block copolymer polyethyleneoxide�poly-

Ž .propyleneoxide�polyethyleneoxide EO �PO �EOn m n

possessing two medium length EO hydrophilic blocksn�surrounding a long and less hydrophilic PO block asm

� �� surfactant 34 . At low concentration in water, thesecopolymers form cylindrical aggregates, with the de-hydrated PO blocks in their cores, surrounded by am

Ž .corona formed by the hydrated EO blocks Fig. 2 .nŽ .At the low pH used for synthesis �2 , the EO

groups are protonated. Their interactions with posi-tively charged silicate oligomers are mediated by Cl�

Ž .anions. Depending mainly upon the n�m ratio ofthe tri-block copolymer and the synthesis tempera-ture, the SBA-15 mesopore diameter and their wallthickness were found to vary respectively within the

˚ �range 47�100 and 31�64 A three examples are de-scribed in Table 3�note that addition of a swelling

Ž .agent like trimethylbenzene TMB induces a phasetransition from the SBA-15 hexagonal structure to aporous cage one with uniform cage dimensions of up

˚ � ��to 360 A 35 . After calcination, SBA-15 silicas werefound to be extremely stable. The unit-cell parameterof a calcined SBA-15 remains unchanged after 24 h in

Fig. 2. Schematic representations of MCM-41 and SBA-15 in theirŽ . Žhybrid surfactant-containing and calcined surfactant free, meso-

.pores empty forms. For SBA-15, two distinct models are presented.� � �In model 1 43 , microporous connections between mesopores are

explained by connections of EO chains belonging to vicinal aggre-n� � �gates. In model 2 36 , the existence of a microporous corona

surrounding mesopores is proposed, based on the quantitativeŽexploitation of X-ray diffraction data the schemes illustrating

.model 2 have been drawn on scale .

� �� boiling water 34 , and changes by less than 2%� � �after 1 year of contact with moist air 36 . Replacing

� � Žcationic surfactants, a strong S I S, surfactant; I,.inorganic electrostatic interactions by neutral tri-

block copolymers, and weak S�Cl��I� interactions,is therefore an easy way to induce an increase in thewall thickness of mesoporous silicas and to stabilizethem.

The experimental conditions described in Zhao et� �� al. 34 are simple to reproduce, but they limit the

applicability of SBA-15 silicas, at least for catalysis. Itis indeed difficult to introduce heteroelements in

� � Ž .acidic conditions 5 . Using a larger pH range 0�9 ispossible, but requires a good control of the relativerates of hydrolysis and condensation of silica precur-

� � �sors 37 . Large enough oligomeric silicates must beŽformed from fully hydrolyzed monomeric silica pre-.cursors before interacting with tri-block copolymers.

From TEOS, it is possible to obtain these large


oligomers when pH is lower than 4 and when silicaoligomerization is a rapid process. At higher pH,there is a competition between the condensation ofpartially hydrolyzed silica species and the hydrolysisof remaining alkoxysilane moieties. Residual organicgroups weaken the interactions between the hy-drophilic part of the block copolymers and silicaoligomers. This results in a poorly organized hybridmaterial. Up to pH 4, hydrolysis can be accelerated bychanging the alkoxide nature. For instance, TEOS canbe changed for tetramethylorthosilicate TMOS. At agiven pH, TMOS hydrolysis is more rapid than that ofTEOS, because of the lower steric hindrance betweenmethoxide moieties and the increased solvation ofresulting alcohol. From pH 4 to 9, fluoride additioncan be used to promote TMOS hydrolysis and silicacondensation. Another method that avoids strongacidic conditions consists in using a non-ionic diblock

� Ž . �surfactant Brij 56: C H EO H complexed by a16 33 10Ž 2� . � �� divalent metallic cation Co for instance 38 .

Note that if the structural parameters of the hexago-nally ordered materials thus obtained are similar tothose of SBA-15 silicas prepared in acidic conditions,their walls are thinner. This is due to the short lengthof the hydrophilic part of the Brij 56 surfactant and toits complexation by Co2� cations, which decreases itshydrophilic character.

The use of expensive Si alkoxides may also hamperthe commercial use of SBA-15 silicas. Several groupshave worked with other silica sources. Guth et al.have reported that sodium silicate solutions could be

Ž .templated by poly ethylene oxide-based surfactantsŽ .Triton100 for instance in the pH range 3�10.5� �39,40 . However, the materials described by this groupexhibit irregular or disordered channels and rather

� �broad pore size distribution. Stucky et al. 41 havereported that well-ordered hexagonal SBA-15 can besynthesized from sodium metasilicate and different

�block copolym ers tri-block and di-blockŽ . �C H EO H, for instance . Note that the wall12 25 9

Ž .thickness of the C H EO H-templated silica is12 25 9˚Ž .quite low 10 A , but no stability tests have been

reported on it yet.SBA-15 materials cannot be considered as extra-

thick wall MCM-41 analogs. Several peculiarities af-fect their N sorption curves: a non-zero origin is2

detected on comparative t-plots and the DS�V ratioŽD, S and V are the mesopore diameter, the overall

.specific surface area and porous volume, respectivelyis generally larger than 4.4, which is the value ex-pected for a regular array of hexagonally packedmesopores. These peculiarities indicate the presenceof additional micropores and small mesopores within

� � �� silica walls 36 ,42 ,43 ,44 ,45 ,46 . On the basis of

� � � � �� XRD 36 , SAXS 42 , N sorption 42 ,2� �� 43 ,44 ,45 and NMR 46 data, this micro-

porosity has been associated with the partial occlusionof EO groups into the walls.

Several groups have tried to control the microp-orosity of SBA-15. The effect of synthesis tempera-ture and of the TEOS�EO PO EO molar ratio17 58 17

˚� � Ž .have been described 47 . Thicker walls 65 A andhigher amounts of intrawall pores were observed withthe lowest synthesis temperature and the highest

Ž .TEOS�EO PO EO molar ratio Table 3 . HT17 58 17

have also been introduced and were found to increasemesopore diameter, decrease wall thickness and par-tially eliminate intrawalls pores without changing the

� � � Ž .overall aspect of the SBA-15 grains 36 Fig. 3 . Allthese observations are consistent with a temperature-induced decrease in the hydrophilic character of theEO chains, which yields to their progressive expulsion

Ž .from silica walls Fig. 2 . Note that after HT, someintrawalls porosity remains. Indeed, modeling of XRDintensities reveals that a microporous corona still

Žsurrounds the mesopores density of 1.21 compared to3.that of amorphous silica, 2.2 g�cm . However, re-

maining micropores are no longer detected on com-parative t-plots. This could indicate that the surfaceof the mesoporous material is not of the same nature

� � �as that of the non-porous reference material 43 .Other methods to decrease the microporous volumehave also been reported such as thermally treatinghybrid SBA-15 in diluted alcohol solutions of block

� �copolymer 48 , microwave-assisted HT in the pres-� �ence of salts 49 and high-temperature calcinations

� � �� 50 ,51 .Recent works of Ryoo et al. on the preparation of

platinum and carbon replicas in SBA-15 silicas havesuggested that the intrawall micropores may form acontinuous network that connects adjacent mesopores� � �51 ,52,53 . Fig. 4 illustrates the importance of micro-pores, and of the resulting connectivity that they mayintroduce between mesopores, for the functionaliza-tion of SBA-15 silicas. This figure presents METimages of the crystallization of a manganese oxidewithin two different SBA-15 supports. When an as-

Žsynthesized SBA-15 35 �C�24 h followed by calcina-˚.tion at 500 �C�6 h, BJHd mesopore diameter of 40 A

is used as a support, the crystallization is fairly homo-Ž .geneous and oxide nanowires in black are formed in

Ž .nearly all mesopores silica in grey . This is not theŽcase when a SBA-15 hydrothermally treated 100

�C�72 h before calcination, BJHd mesopore diameter˚.of 70 A is used as a support. The observed differ-

ences are most probably due to connections betweenmesopores introduced by micropores. In favor of thishypothesis, an inset clearly shows that the fringes


Ž . Ž .Fig. 3. TEM images of typical SBA-15 grains with the electron beam parallel bottom and perpendicular top to the main axis of the grain:� Ž .�as-synthesized SBA-15 TEOS, EO PO EO , acidic conditions, 35 �C�24 h, followed by calcination 500 �C, 6 h ; with 3 days HT at 100 �C20 70 20

before calcination; and with 15 days HT at 100 �C before calcination.

observed in two adjacent manganese oxide nanowiresare perfectly parallel.

4. Conclusions

We have shown that thicker walls and fewer strainedsiloxane bridges improve the stability of mesoporoussilicas prepared by supramolecular templating, andthat the wall thickness is considerably affected by thenature of the surfactant and the mechanism of themesostructure formation.

With cationic surfactants, the walls of mesoporous˚Ž .silicas are relatively thin �20 A thick . Their thick-

ness can be modified either by synthesis or by post-synthesis treatment. Modifications by synthesis re-quire the optimization of several experimentalparameters and are difficult to reproduce. Modifica-

Žtions by post-synthesis heating treatments before cal-

.cination are easier but time-consuming. Modifica-tions by silylation after calcination or by carefulselection of the anion associated with the cationicsurfactant are probably simpler, but are only helpfulif applications concern hydrophobic processes. Fi-nally, modifications by grafting are possible butstabilizing the mesoporous silica, which is used as asupport is still necessary.

With neutral copolymers, thicker silica walls inwhich hydrophilic organic parts are, at least partially,incorporated are obtained. Most of the SBA-15 sili-cas, templated by tri-block copolymers, have thicker

˚Ž .walls thickness up to 63 A than MCM-41 and areconsequently, more stable. However, SBA-15 silicascannot be only considered as ultra-thick wall analogsof MCM-41. Their porosity is indeed bimodal, andtheir mesopores are surrounded by microporous silicawalls. Future research strategies will aim either at

Žobtaining mesopores only for fundamental work de-.pending on the mesopore diameter or at designing


Fig. 4. Microtomic slides illustrating the crystallization of manganese oxide nanowires in two different SBA-15 supports: SBA-15 supportŽ .as-synthesized TEOS, EO PO EO , acidic conditions, 35 �C�24 h in which an intrawall porosity is detected on a comparative t-plot20 70 20

˚� Ž .inset: the correlation between the fringes observed in two adjacent manganese oxide nanowires inter-fringe distance �3.64 A , suggests that� Ž .the mesopores are connected by micropores . SBA-15 support hydrothermally treated at 100 �C, 3 days before calcination in which a

comparative t-plot fails to detect an intrawall porosity.

the pore sizes and connectivities of multimodalporosities. New characterization techniques must alsobe developed to compare the properties of the silicasurface in the different porosities. This appears cru-cial to optimize synthesis and post-synthesis parame-ters for the functionalization of SBA-15 supports.

Noted added in proof

After submitting this paper for publication, we have�become aware of a recent paper by Lee et al. Synthe-

sis of mesoporous silicas of controlled pore wall thick-ness and their replication of ordered nanoporous car-


bons with various pore diameters, J. Am. Chem. Soc.Ž .2002, 124 7 , 1156�1157 by Lee, JS, Joo, SH and R.

�Ryoo . The authors developed a synthesis strategywhich allows a systematic control of pore-wall thick-ness in mesoporous silicas. Their strategy is based onthe use of mixtures of ionic, hexadecyltrimethylam-

Ž .monium bromide HTAB, and neutral, C16 EO 2OHŽ .and C16 EO 10OH, surfactants as templates: the

thickness of silica walls is controlled by theHTAB�neutral surfactant molar ratio, i.e., by thenumber of functional gEO groups able to interactwith silicate precursors. The obtained silicas havebeen used to template 2D hexagonally ordered meso-porous carbons.

Acknowledgements

The author is deeply indebted to Prof. L. Bon-neviot, Prof. GD. Stucky, Prof. M. Che and Dr M.Breysse for their constant help and support. The

Ž .author also thanks P. Beaunier and MD. Appay LRSfor their help in MET measurements and analysis.

References and recommended reading

� of special interest�� of outstanding interest

� �1 Ying JY, Mehnert CP, Wong MS. Synthesis and applicationsof supramolecular-templated mesoporous materials. AngewChem Int Ed 1999;38:56�77.

� �2 Raimondi EM, Seddon JM. Liquid crystal templating ofporous materials. Liq Cryst 1999;26:1�35.

� �3 Ciesla U, Schuth F. Ordered mesoporous materials. Microp¨Mesop Mater 1999;27:131�149.

� �4 Bonneviot L, Giasson S, Kaliaguine S, Stocker M. Proceed-¨ings of the Second International Symposium on Mesoporous

Ž .Molecular Sieves. Microp Mesop Mater Editorial 2001:44�45.

� �5 Tuel A. Modifications of mesoporous silicas by incorporationof heteroelements in the framework. Microp Mesop Mater

Ž .1999;27 2�3 :151�169.� �6 Trong On D, Desplantiers-Giscard D, Dahunac C, Kaliaguine

S. Perspectives in catalytic applications of mesostructuredŽ .materials. Appl Catal A 2000;222 1�2 :299�357.

� �7 Fraile JM, Garcia JI, Mayoral JA, Vispe E, Brown DR,Naderi M. Is MCM-41 really advantageous over amorphoussilica, the case of grafted titanium epoxidation catalysts. ChemCommun 2001;16:1510�1511.

� �8 Zhao XS, Audsley F, Lu GQ. Irreversible change of pore� structure of MCM-41 upon hydration at room temperature. J

Phys Chem B 1998;102:4143�4146.The walls of mesoporous silicas templated by CTMA� are shownto be damaged by hydrolysis, full structure collapse occurring after3 months of air exposure.� �9 Carrott MMLR, Estevao-Candeias AJ, Carrott PJM, Unger

�� KK. Evaluation of the stability of pure silica MCM-41 towardwater vapor. Langmuir 1999;15:8895�8901.

MCM41 silicas obtained with different n-alkyltrimethylammoniumcations are compared using XRD, water vapor sorption isotherms

at RT and N sorption at 77 K. It is shown that water adsorption2induces alterations of silica walls. It is also confirmed that silylationis effective in stabilizing silica walls.� �10 Frasch J, Lebeau B, Soulard M, Patarin J. In situ investiga-

� tion on cetyltrimethylammonium surfactant�silicate systems,precursors of organized mesoporous MCM-41-type siliceous

Ž .materials. Langmuir 2000;16 23 :9049�9057.ŽFluorescence probing techniques with pyrene or dipyrenylpropane

.probes are used to show that the interactions between silicateoligomers and surfactant micelles are weak in the early stages ofthe formation of an MCM-41 silica in basic conditions. The firstand more important parameter in the synthesis is the formation of

Žsiliceous prepolymers. As these prepolymers grow, they bind multi-.dentate binding with an increasing number of surfactants of oppo-

site charge.� �11 Gross AF, Ruiz EJ, Tolbert SH. Effect of framework po-

lymerization on the phase stability of periodic silica�surfac-tant nanostructured composites. J Phys Chem B 2000;104:5448�5461.

� �12 Gross AF, Van Le H, Kirsch BL, Riley AE, Tolbert SH.�� Correlations between silica chemistry and structural changes

in hydrothermally treated hexagonal silica�surfactant com-posites examined by in situ X-ray diffraction. Chem Mater2001;13:3571�3579.

Ž .pH-driven silica chemistry condensation�hydrolysis is shown tocontrol structural changes in hybrid MCM�41 silicas heated underhydrothermal conditions. The degree of silica condensation con-trols surfactant loss and framework flexibility and determines mate-rials expansions and contractions.� �13 Landry CC, Tolbert SH, Gallis KW et al. Phase transforma-

� tions in mesostructured silica�surfactants composites. Mech-anism for change and applications to materials synthesis.Chem Mater 2001;13:1600�1608.

Ž .A model is proposed to explain the hexagonal MCM-41 to cubicŽ .MCM-48 phase transformation and the existence of a lamellarintermediate during the transformation.� �14 Wu J, Liu X, Tolbert SH. High-pressure stability of ordered

�� mesoporous silicas: rigidity and elasticity through nanometerscale arches. J Phys Chem B 2000;104:11837�11841.

The compressibility of hybrid MCM-41 silicas is examined at highpressure by small angle X-ray diffraction. It is shown that hy-drothermal treatments greatly enhance mechanical properties.� �15 Galarneau A, Desplantier D, Dutartre R, Di Renzo F. Mi-

celle templated silicates as a test bed for methods of meso-pore size evaluation. Microp Mesop Mater 1999;27:297�308.

� �16 Rouquerol F, Rouquerol J, Sing K. Adsorption by powdersand porous solids: principles, methodology and applications.San Diego, CA: Academic Press, 1999.

� �17 Kruk M, Jaroniec M. Gas adsorption characterization of�� ordered organic�inorganic nanocomposite materials. Chem

Ž .Mater 2001;13 10 :3169�3183.Gas adsorption methods and their use to characterize mesoporousmaterials are reviewed with special emphasis on the differentmethods which can be use to determine mesopore diameters. Theopportunities in the adsorption on specific adsorption sites and inthe determination of bonding density of given groups on a surfaceare presented. Some recommendations are made for reporting ofadsorption data to facilitate a comparison of experimental resultsfrom different laboratories.� �18 Ravikovitch PI, Neimark AV. Characterization of micro- and

� mesoporosity in SBA-15 materials from adsorption data bythe NLDFT method. J Phys Chem B 2001;105:6817�6823.

A method based on the non-local density functional theory ofcapillary condensation and hysteresis in open cylindrical pores isused to evaluate the pore wall thickness and the amount of in-trawall porosity in SBA-15 silicas. It is show that, unless speciallytreated, an appreciable volume of intrawall pores, which may con-


stitute up to 30% of the overall porous volume, can be detected. Itis shown also that an SBA-15 silica calcined at 1273 K has a regular

Žstructure no intrawall porosity�roughness�see also Matos et al.� �.50 .� �19 Cheng YR, Lin HP, Mou CY:. Control of mesostructure and

morphology of surfactant-templated silica in mixed surfactantsystems. Phys Chem Chem Phys 1999;1:5051�5058.

� �20 Ryoo R, Ko CH, Park IS. Synthesis of highly ordered MCM-41by micelle-packing control with mixed surfactants. ChemCommun 1999;15:1413�1414.

� �21 Schulz-Ekloff G, Ratousky J, Zukal A. Controlling of mor-phology and characterization of pore structure in orderedmesoporous silicas. Microp Mesop Mater 1999;27:273�285.

� �22 Echchahed B, Morin M, Blais S, Badiei AR, Berhault G,� Bonneviot L. Ion mediation and surface charge density of

micelle templated silica. Microp Mesop Mater 2001;44�45:53�63.

It is shown that silica mesophases contains guest anions whichaffect the charge distribution within the organic�inorganic inter-face.� �23 Bonneviot L, Morin M, Badiei A. Mesostructured metal and

non-metal oxides and method for making them. Patent WO01�55031 A1.

� �24 Chen L, Horiuchi T, Mori T, Maeda K. Postsynthesis hy-drothermal restructuring of M41S mesoporous molecularsieves in water. J Phys Chem B 1999;103:1216�1222.

˚� �25 Kleitz F, Blanchard J, Zibrowius B, Schuth F, Agren P,¨Linden M. Influence of co-surfactants on the properties ofmesostructured materials, in press.

� �26 Kruk M, Jaroniec M, Sayari A. New insights into pore-sizeexpansion of mesoporous silicates using long chain amines.Microp Mesop Mater 2000;35�36:545�553.

� �27 Kim JM, Jun S, Ryoo R. Improvement of hydrothermal�� stability of mesoporous silica using salts: reinvestigation for

time-dependant effect. J Phys Chem B 1999;103:6200�6205.It is shown that the salt-effect which improve the hydrothermalstability of MCM-41 silicas is strongly time-dependent and requiresat least 10 days at 373 K.� �28 Das D, Tsai CM, Cheng S. Improvement of hydrothermal

stability of MCM-41 mesoporous molecular sieves. ChemCommun 1999;5:473�474.

� �29 Tang X, Liu S, Wang Y, Sominski E, Palchick, Koltypin Y,Gedanken A. Rapid synthesis of high quality MCM-41 silicaw ith u ltra so u n d ra d ia t io n . C h e m C o m m u n2000;21:2119�2120.

� �30 Mokaya R. Improving the stability of mesoporous MCM-41�� silica via thicker more highly condensed pore walls. J Phys

Chem B 1999;103:10204�10208.Thick walls MCM-41 silicas are prepared either by extended HT

Žtreatments or by using a calcined MCM-41 as a ‘seed’ secondary.crystallization . They are shown to be remarkably stable in boiling

Žwater information about stability under steam are given in the next.reference . A diffusion-controlled mechanism in which silicate

oligomers are added to the growing surfactant�silicate aggregatesŽ . Ž .prolonged HT , or to the ‘seed’ crystals secondary crystallizationis proposed.� �31 Mokaya R. Hydrothermally stable restructured mesoporous

� silica. Chem Commun 2001;11:933�934.Mesoporous silicas, prepared by using a calcined MCM-41 as a‘seed’, are shown to have very thick walls and to be remarkablystable both in boiling water and under steam.� �32 Mokaya M. Influence of pore wall thickness on the steam

� stability of Al-grafted MCM-41. Chem Commun 2001;7:633�634.

� �33 Zhao D, Huo Q, Feng J, Chmelka BF, Stucky GD. Non-ionic� triblock and star diblock copolymer and oligomeric surfactant

syntheses of highly ordered hydrothermally stable meso-porous silica structures. J Am Chem Soc 1998;120:6024�6036.

ŽA family of ordered mesoporous silicas mesopore diameters˚.between 20 and 300 A are obtained by using non-ionic

Ž . Ž .alkylpoly ethylene oxide and poly alkylene block copolymers. Peri-odic arrangements with cubic Im3m, cubic Pm3m, 3-dim. hexagonalŽ . Ž . Ž .P6 �mmc , 2-dim. p6mm and lamellar L � symmetries are3described.� �34 Zhao D, Feng J, Huo Q et al. Triblock copolymer syntheses

˚�� of mesoporous silica with periodic 50 to 300 A pores. Science1998;279:548�552.

First report showing that amphiphilic tri-block copolymers can beused to template highly ordered hexagonal SBA-15 silicas in acidicconditions. SBA-15 with a wide range of mesopore diameters

˚ ˚Ž . Ž .46�100 A and wall thicknesses 31�64 A can be prepared, using avariety of tri-block copolymers and changing the synthesis tempera-ture. The tri-block copolymers can be recovered by solvent extrac-tion with ethanol and reused. The obtained SBA-15 mesoporoussilicas are stable in boiling water.� �35 Lettow JS, Han YJ, Schmidt-Winkel P et al. Hexagonal to

mesocellular foam transitions in polymer-templated meso-porous silicas. Langmuir 2000;16:8291�8295.

� �36 Imperor-Clerc M, Davidson P, Davidson A. Existence of a� microporous corona around the mesopores of silica-based

SBA-15 materials templated by triblock copolymers. J AmChem Soc 2000;122:11925�11933.

The XRD intensities, as measured on hybrids and calcined SBA-15silicas templated by EO PO EO and EO PO EO , are20 70 20 17 60 17compared and quantitatively exploited. It is shown that the SBA-15silicas cannot be regarded as an ideal hexagonal lattice of poresimbedded in a uniform silica matrix. Cylindrical organic micellesare surrounded by a ‘corona’ region of low electronic density andthis ‘corona’ becomes microporous upon calcination. Note thatafter a 100 �C HT of 3 days, the micropores are no longer detected

Ž .by analyzing N sorption experiments zero origin on a �-plot , but2still remain as indicated by the electronic density of the ‘corona’,

Žwhich is still significantly lower than that of amorphous silica 1.23.compared to 2.2 g�cm .

� �37 Kim JM, Han YJ, Chmelka BF, Stucky GD. One-step synthe-� sis of ordered mesocomposites with non-ionic amphiphilic

block copolymers: implications of isoelectric point, hydrolysisrate and fluoride. Chem Commun 2000;23:2437�2438.

Mesoporous SBA-15 silicas can be prepared in the presence ofŽ .fluoride over a wide pH range 0�9 , by controlling the rates of

hydrolysis and condensation of tetramethoxysilane.� �38 Zhang W, Glomski B, Pauly TR, Pinnavaia TJ. A new non-

�� ionic surfactant pathway to mesoporous molecular sieve sili-cas with long range framework order. Chem Commun1999;18:1803�1804.

Ž .Hexagonally ordered mesoporous silicas are templated by R EO Hn

surfactants. Electrostatic forces are introduced by the complexationŽ � 2� 2� 2�of the EO groups by small metallic cations Li , Co , Ni , Mn

2�.and Zn .� �39 Sierra L, Lopez B, Gil H, Guth JL. Synthesis of mesoporous

Ž .silica from sodium silicate solutions and poly ethylene oxidebased surfactants. Adv Mater 1999;11:307�310.

� �40 Sierra L, Guth JL. Synthesis of mesoporous silica with tun-able pore size from sodium silicate solutions and a polyethy-lene oxide surfactant. Microp Mesop Mater 1999;27:243.

� �41 Kim JM, Stucky GD. Synthesis of highly ordered mesoporoussilica materials using sodium silicate and amphiphilic blockcopolymers. Chem Commun 2000;13:1159�1160.

� �42 Golltner CG, Smarsly B, Berton B, Antonietti M. On the¨�� microporous nature of mesoporous molecular sieves. Chem

Mater 2001;13:1617�1624.


Small angle X-ray scattering and qualitative porosimetry are ap-plied to characterize mesoporous silicas templated with different

Ž .surfactants containing a poly ethylene oxide chain. It is shown thatat low concentration, the water and silica-compatible EO groupscreate pores within silica walls of a size comparable to the diameterof hydrated EO chains. From specific surface area measurements, itis shown that a substantial amount of intrawall porosity may beexpected when block copolymers are used as surfactants.� �43 Galarneau A, Cambon H, Di Renzo F, Fajula F. True micro-

� porosity and surface areas of mesoporous SBA-15 silicas as afu n ction of syn th esis tem peratu re . L an gm u ir

Ž .2001;17 26 :8328�8335.Interaggregate interactions between the EO groups of adjacenttriblock co-polymer aggregates are proposed to be at the origin ofintrawall porosity in SBA-15 silicas. A geometrical model, using adensity of 2.2 g�cm3 for the solid part of silica walls, is used tocalculate mesopore diameters.� �44 Feng P, Bu X, Pine DJ:. Control of pore sizes in mesoporous

�� silica templated by liquid crystals in block copolymers�cosur-factants water systems. Langmuir 2000;16:5304�5310.

The synthesis of optically transparent triblock copolymersŽ . ŽEO PO EO and EO PO EO �cosurfactant butanol,20 70 20 106 70 106

.pentanol or hexanol �silica monoliths, using a direct liquid crystaltemplating approach, is described. It is shown that, because of itslarge hydrophilic volume, EO PO EO gives rise to much106 70 106

˚ ˚Ž .thicker walls than EO PO EO 53 A compared to 29 A .20 70 20� �45 Kruk M, Jaroniec M, Ko CH, Ryoo R. Characterization of

� the porous structure of SBA-15. Chem Mater 2000;12:1961�1968.

XRD and N sorption measurements are used to confirm that the2diameter of mesopores in SBA-15 silicas can be tailored by thechoice of synthesis temperature, and that their walls contain dis-ordered micropores and small mesopores.� �46 Melosh NA, Lipic P, Bates FS et al. Molecular and mesos-

copic structure of transparent block copolymers silicamonoliths. Macromolecules 1999;32:4332�4343.

� �47 Miyazawa K, Inagaki S. Control of microporosity within the

walls of ordered mesoporous silica SBA-15. Chem Commun2000;21:2121�2122.

� �48 Sun JH, Moulijn JA, Jansen KC, Maschmeyer T, CoppensMO:. Alcothermal synthesis under basic conditions of anSBA-15 with long-range order and stability. Adv Mater2001;13:327�331.

� �49 Newalkar BL, Komarneni S. Control over microporosity of� ordered microporous�mesoporous silica SBA-15 under mi-

crowave-hydrothermal conditions: effect of salt addition.Chem Mater 2001;13:4573�4579.

Under microwave-hydrothermal conditions, the microporousvolume of SBA-15 silicas is found to decrease with increasing saltŽ .NaCl concentrations.� �50 Matos JR, Mercuri L, Kruk M, Jaroniek M. Toward the

� synthesis of extra large pore MCM-41 analogues. Chem Mater2001;13:1726�1731.

� �Like in Shin et al. 51 , calcination above 1200 K is shown todecrease the intrawall porosity of SBA-15 silicas. An initial SBA-15with a low content of connecting micropores and sufficiently largeprimary mesopores has to be selected.� �51 Shin HJ, Ryoo R, Kruk M, Jaroniec M. Modifications of

�� SBA-15 pore connectivity by high temperature calcinationinvestigated by carbon inverse replication. Chem Commun2001;4:349�350.

SBA15 silicas calcined at 1153 and 1243 K are compared. Microp-ores connecting mesopores are still present at 1153 K and elimi-nated at 1243 K. Indeed, carbon replication yields to 2-Dim. or-dered CMK-3 carbon with the first sample, whereas only disorderedcarbon replicas are obtained with the second one.� �52 Ryoo R, KO CH, Kruk M, Antochshuk V, Jaroniec M. Block

copolymers templated ordered mesoporous silica: array ofuniform mesopores or mesopore�micropore network? J PhysChem B 2000;104:11465�11471.

� �53 Jun S, Joo SH, Ryoo R et al. Synthesis of new nanoporouscarbon with hexagonally ordered mesostructure. J Am ChemSoc 2000;122:10712�12713.

modifying the walls of mesoporous silicas prepared by supramolecular-templating

Documents