Cross-Linked Monolithic Xerogels Based on Silica Nanoparticles

Subscriber access provided by University | of Minnesota Libraries

Article

Cross-Linked Monolithic Xerogels Based on Silica Nanoparticles Markus Börner, Theresa Noisser, and Gudrun Reichenauer Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/cm401717q • Publication Date (Web): 20 Aug 2013 Downloaded from http://pubs.acs.org on September 1, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Cross-Linked Monolithic Xerogels Based on Silica Nanoparticles Markus Börner, Theresa Noisser, Gudrun Reichenauer* Bavarian Center for Applied Energy Research (ZAE Bayern), Am Galgenberg 87, 97074 Wuerzburg, Germany E-mail: *Corresponding author: [email protected]

Abstract Low density silica xerogels were obtained by reducing the capillary forces and thus the shrinkage upon ambient pressure drying. This was achieved by increasing the particle and thus also the pore size at a given target density. To increase the particle size beyond the values characteristic for classical silica aerogel, particle growth and formation of the interconnected network were decoupled. Uniform silica nanoparticles of 20 to 30 nm in diameter synthesized by a Stöber process were subsequently cross linked. To strengthen the gel network additional treatments were applied including an organic coating of the gel structure. For that purpose 3Aminopropyltriethoxysilane (APTES) was used as a linker between the inner surface of the silica backbone and the organic component. Simultaneously, APTES serves as a catalyst for the gelation. Since the components for the organic coating of the silica backbone do not interfere with the network formation, a one-pot synthesis is feasible. Several bifunctional epoxides were investigated to obtain highly homogeneous low density xerogels. A further reduction in density was accomplished by the oxidative decomposition of the organic components in the xerogel state thus providing xerogels with a density of about 180 kg/m3 and a pore size of roughly 100 nm.

Keywords: silica xerogels; low density; surface modification; Stöber process; organic coating 1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1. Introduction Silica aerogels are typically characterized by very low densities with pores at the nanoscale thus providing low thermal conductivities and large, well accessible surface areas. These properties make them suitable for several applications. However, they are often strongly restricted by their poor mechanical properties (in particular the low tensile strength) and the noncontinuous process related to the supercritical extraction of the pore liquid. Several ways to improve the mechanical properties and thus to make silica gels withstand the mechanical stresses upon drying

[1]

and handling have been reported. Common strategies are the massive

strengthening of the gel backbone by reinforcement of the necks connecting adjacent backbone particles with additional monomers

[2]

and the reduction of irreversible drying shrinkage via

silylation of the silica backbone surface

[3]

, which is a time- and solvent-consuming process.

Another way is the modification of the network by means of an organic coating of the inner surface to strengthen the gels backbone and thereby reduce the shrinkage during drying to a great extent [4][5][6] . A variety of polymers such as polystyrenes [7], isocyanates [8] and epoxides [9] have been used in this strategy. Based on this approach, a basically continuous process for preparing nanoporous solids is presented here by drying under subcritical conditions. To significantly reduce drying shrinkage, the capillary forces have to be decreased and/or the gel structure has to be mechanically reinforced. The pressure in a cylindrical capillary is inversely proportional to the pore radius

[1]

. On the other hand, for a given gel backbone structure and

density the pore size is directly correlated with the size of the particles forming the backbone. Therefore larger particles help reducing the capillary forces and consequently decrease the

2


Page 2 of 30

Page 3 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


shrinkage upon drying. Since the particle sizes in a classical sol-gel process for silica based gels are typically below 10 nm [10][11][12][13][14][15][16] a novel, sol-gel-based approach is required. In this study we developed a new type of monolithic silica xerogel with both larger primary particles and an organic coating of the network. At first monodisperse silica nanoparticles were synthesized by the Stöber process [17], which are then crosslinked in a subsequent process step to form a gel. In addition to aging processes an organic coating of the inorganic backbone was applied to mechanically strengthen the gel. For this purpose we used several bifunctional epoxides that feature an appropriate solubility in the ethanol-based suspension. The fact that the reaction of the epoxides with the surface groups introduced as interface is very slow until the temperature of the wet gels is raised above 50°C

[9]

enables a one-pot synthesis where all

reactants are already present in the initial solution. A slightly different approach of an one-pot synthesis including a heat triggered cross-linking of the modified silica surface was reported by Leventis et al.

[6][18]

. However the method pursued in our study significantly shortens the

processing time compared to a multistep synthesis by eliminating solvent exchanges and monomer diffusion. Within the framework of this study a screening with 6 different bifunctional epoxides (see Figure 1) was performed to identify the organic component that fulfills the requirements to minimize drying shrinkage best. They can be categorized in two groups: Bisphenol A propoxylate diglycidyl ether (BAPGE), bisphenol A digylcidyl ether (BADGE) and resorcinol diglycidyl ether (REDGE) show short chain lengths with one or two centered phenol rings. The remaining epoxides, poly(propylene glycol) diglycidyl ether (PPGGEn) and poly(ethylene glycol) diglycidyl ether (PEGGE), are characterized by longer chains with propylene or ethylene glycol units, respectively. 3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

FIGURE 1

To further reduce the density of the obtained xerogels the organic coating is considered being only a temporary mechanical support for the backbone upon drying. Initiating oxidative decomposition of the organic components finally yields a low density inorganic xerogel.

2. Experimental Section Chemicals.

Tetraethoxysilane

(TEOS)

was

purchased

from

Alfa

Aesar,

3-

aminopropyltriethoxysilane (APTES), bisphenol A propoxylate diglycidyl ether (BAPGE), bisphenol A digylcidyl ether (BADGE), resorcinol diglycidyl ether (REDGE), poly(propylene glycol) diglycidyl ether (PPGGE) and poly(ethylene glycol) diglycidyl ether (PEGGE) were purchased from Sigma Aldrich. Ethanol denat. was purchased from Nedalco Alcohol. All chemicals were used as received without further purification. Preparation of hybrid xerogels. All gels synthesized are based on monodisperse silica nanoparticles prepared in a Stöber process. The silica spheres were prepared by reacting tetraethoxysilane (TEOS) in ethanol in presence of water and ammonia as the catalyst. As suggested by Stöber et al. the particle size varies with the concentration of the catalyst [17]. Hence diameters smaller than 50 nm can be prepared by successively reducing the catalysts concentration to a certain value. The reactant concentrations in the base catalyzed sol-gel process were chosen in a way that neither gelation nor precipitation occurred. The final particle size was adjusted by an ammonia concentration of 0.147 mol/l. The concentrations of the other reactants were 0.26 mol/l for tetraethoxysilane (TEOS), 3.24 mol/l for water and about 15 mol/l for 4


Page 5 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ethanol. To prevent high local concentrations the reactants were separately dissolved in fractions of the total amount of the required ethanol prior to mixing. The solution was stirred for more than 24 hrs until the particle growth was complete. To enable the crosslinking (gelation) of the silica particles the suspension was reduced by 50 vol% via distillation of ethanol. Furthermore the ammonia was removed due to its lower boiling point. As a consequence the suspension was characterized by a neutral pH after distillation. The quantity of APTES to first catalyze the gelation and to cover the surface of the silica backbone thereafter was calculated by assuming 4.6 hydroxyl groups per square nanometer on the silica surface according to Iler et al.

[19]

and 0.49 square nanometer covered by an APTES

molecule [20]. The total surface area of all particles in the suspension was calculated by assuming that the TEOS monomers in the initial solution completely reacted to silica. The radius of the particles was determined by dynamic light scattering. In the experiments the amount of APTES to cover the silica surface was varied from 56 %, 113 %, 169 % up to 225 % of the theoretical value. This was sufficient to initiate the gelation of the suspension in all cases. To ensure a uniform distribution of the epoxides they were mixed with the distilled Stöber suspension prior to adding APTES. The molar ratio of the epoxide (E) and APTES (A) is denoted by E/A. Since the bifunctional epoxides are capable of crosslinking two amine sites, the respective stoichiometric ratio is E/A = 0.5. However, it turned out to be necessary to add an excess of epoxides to achieve uniformly covered surfaces. Upon synthesis the E/A value was therefore varied from 1.5 to 20, a range in which the solubility limit is not yet exceeded. The reaction between the amine and the epoxide took place during a thermal treatment of the gel for

5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

24 hrs at 85 °C. A temperature of 85°C was chosen to assure that all amine sites react with an epoxide within the 24 hrs of the thermal treatment. After the addition of epoxide and APTES under constant stirring the suspension was poured into HDPE molds with a volume of 5 ml each. All gels formed in less than two hours and were aged at 30 °C for 24 hrs followed by aging in monomer solution for another 24 hrs at 30 °C. For that purpose 1.5 g TEOS were added on top of each gel. The monomers were allowed to diffuse into the gel over 24 hrs to increase the stiffness of the network. Subsequently, the gels were heat treated to initiate the reaction of the epoxide with the amine for the organic coating and to accelerate the Ostwald ripening for a further increase in strength of the gel backbone. For this treatment the vials were kept closed and placed in a water bath to guarantee uniform heating. Afterwards the samples were slowly dried at 30 °C for 4 to 5 days while keeping about 75 % of the vial openings covered. Finally, residual pore liquid was removed within 90 min at 110 °C. Characterization. The diameter of the silica nanoparticles was determined by dynamic light scattering (DLS) using the particle size analyzer Horiba LB-550. The physical characterization of the xerogels was performed with respect to their density, linear shrinkage and Young’s modulus. The Young’s modulus was obtained by measuring the sound velocity via the runtime of an ultrasonic pulse across the xerogel. The dynamic modulus was calculated by sound velocity, density and the Poisson’s ratio[21] which was assumed to be 0.2 for silica gels [22]. The microstructure was investigated by scanning electron microscopy (SEM), nitrogen sorption measurements and small angle X-ray scattering (SAXS) experiments. A Zeiss Ultra Plus SEM was used to inspect the microstructure of the xerogels with respect to their backbone particle size and shape, pore size and uniformity. The investigation of the structural properties on 6


Page 6 of 30

Page 7 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the length scale < 100 nm was performed by a volumetric N2-sorption experiment (ASAP 2010 and ASAP 2020 by Micromeritics) at 77 K. The isotherms recorded were analyzed to determine the overall specific surface area SBET with the BET model

[23]

. This total surface area can be

separated into the contribution due to micropores (Smicro as obtained by the t-plot method [24]) and the external surface Sext. Moreover, the BET analysis yields the so-called C parameter that provides information about the interaction between adsorbate and surface of the gel backbone. This quantity was used to distinguish between organic and inorganic surfaces. Further analysis of the microstructure was performed by SAXS experiments at the beamline B1 (JUSIFA) of HASYLAB. Specimen with a thickness of about 2 mm were irradiated with an X-ray energy of 12 keV and the scattering intensity as a function of scattering angle (respectively scattering vector q) was recorded at two sample-detector distances of 0.9 m and 3.6 m. Due to the very narrow size distribution of the silica nanoparticles the resulting scattering intensity resembles the scattering of an ensemble of homogeneous spheres with identical radii scattering intensity of an ensemble of almost monodisperse spheres

[26]

[25]

. The theoretical

can be refined by

superimposing a Gaussian distribution of a certain width to the particles radius. This theoretical model was fitted to the scattering curves in the q-range between 0.15 nm-1 and 1 nm-1 to determine a mean particle radius. The oxidative decomposition of the organic components was carried out in air at a heating rate of 1 K/min and was monitored by simultaneous thermal analysis (STA) type 501 of Bähr thermal analysis GmbH.

3. Results and Discussion 3.1 Preparation of gels based on silica nanoparticles 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 30

To reduce shrinkage during drying the particle size had to be increased compared to the one typical for the classical sol-gel process. To preserve the gel properties in terms of e.g. a low thermal conductivity, the target diameter of the silica nanoparticles was set to 25 nm. For that purpose the ammonia concentration was set to 0.147 mol/l and the resulting particle size achieved was (27.2 ± 2.4) nm determined by DLS measurements. Since the Stöber process is very sensitive, the final particle size varied with the suspension volume in orders of about 10 %. Compared to the diameter of the spherical particles in classical sol-gel based gel, the particles prepared within the framework of this study are more than twice as large. In addition, the Stöber particles feature a very narrow size distribution therefore providing a high structural uniformity. After the volumetric reduction of the Stöber suspension by distillation, the gelation was catalyzed by APTES due to its basic amine group

[8]

which afterwards served also as the linker

between the inorganic network and the organic coating. Thus the amine groups are expected to be homogeneously distributed across the silica network whereas in a postsynthetic modification a monomer clustering may lead to a nonuniform distribution of functional groups

[27]

. The amount

of APTES normalized to the theoretical quantity required for a complete surface covering was varied from 56 % to 225 %. To determine the most suitable amount providing a uniform distribution without having to take into account the influence of other organic components, some reference gels were dried subcritically at this stage. The resulting monolithic xerogels show a homogeneous, highly porous microstructure (see Figure 2), a density of about 350 kg/m³ and a linear shrinkage of about 45 %. The slight trends in the xerogel properties such as density, linear shrinkage and Young’s modulus upon variation of the APTES concentration were found to be negligible. Due to the basicity of APTES the gelation times decreased significantly with increasing APTES concentration. When they become very short (less than a few minutes), 8


Page 9 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


inhomogeneities appear during the formation of the gel structure. Therefore a theoretical value for the surface coverage of 113 % was found to give the best results and was used for all further gels prepared. The amine modified gels with this amount of APTES were used as the reference samples. Moreover, the aging in monomer solution was adopted for all gels to strengthen the gel backbone which turned out to be inevitable for their mechanical stability.

FIGURE 2

The objective of crosslinking the inorganic network with epoxides was to reinforce the gel structure and reduce the irreversibility of shrinkage. The homogeneous distribution of the amine sites across the silica network was as crucial as the epoxides solubility in the suspension and their passive behavior during the formation of the network. Via heat treatment of the wet gel a uniform organic coating of the highly porous and homogeneous gel structure can be accomplished. The different bifunctional epoxides used to crosslink the modified network were all soluble in the suspension at a molar ratio of E/A = 3. REDGE e.g. exceeded the solubility limit at E/A = 5. Since REDGE, BADGE and BAPGE show similar results and there were no significant differences between PPGGE5 and PPGGE9, only three characteristic epoxides (BADGE, PPGGE9 and PEGGE) are chosen for a detailed analysis at E/A = 3. The properties of the monolithic epoxide crosslinked xerogels are shown in Table 1. All gels show a linear shrinkage of 40 % to 50 % independent of the epoxides used and E/A ratios. The only exception was the PEGGE coated sample with a linear shrinkage of only 31 %.

TABLE 1 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30

The molecular configuration of the different types of epoxides provides a first piece of information about the ability to crosslink the modified gel structure. BADGE is a short molecule with an estimated length of less than 1 nm. Considering the theoretical amount of amine sites per square nanometer, the BADGE molecules can only crosslink adjacent amine groups leading to a poor overall crosslinking of the silica network. The PEGGE and PPGGE9 molecules are long chains with a length of up to 3.5 nm. This allows them to crosslink amine groups over a longer range and to build a mesh-like coating. Once the reaction of the epoxide molecules and the amine groups on the surface is complete, the behavior depends on the properties of the poly(ethylene glycol) (PEG) and the poly(propylene glycol) (PPG) chains respectively. The PEG chains show special elastic properties. The presence of water in the pore liquid affects the formation of hydrogen bonds between the oxygen atoms of PEG and the water molecules

[28]

(see Figure 3).

These hydrogen bonds connect adjacent chain units of PEG which consequently stabilize the configuration

[29]

. At larger stretching forces rupture and relocation of the hydrogen bridges

weaken their influence until they disappear and only intramolecular interactions contribute to the elasticity of PEG

[28]

. For PPG the formation of hydrogen bonds might be hindered by the

hydrophobicity of the methyl group thus reducing the elastic properties of the PPG chain.

FIGURE 3

The microscopic effects of the organic coating step were analyzed by comparing nitrogen sorption and SAXS measurements. As a reference the microscopic structure of the bare silica nanoparticles was investigated. To ensure that the same particle structure is analyzed as it appears 10


Page 11 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


in the gel, a PEGGE-coated xerogel was chosen and the organic components were completely removed by oxidative decomposition. This oxidation took place at 310 °C, a temperature where all organic components are removed without sintering the sample as verified by STA. The analysis of the adsorption data of the silica particles reveals a high overall surface area of the inorganic gel structure of SBET = 453 m²/g with a micropore contribution of more than 70 %. This highly microporous structure of the silica spheres contradicts the theory of LaMer et al. concerning the growth of monodisperse nonporous dense silica particles [30]. The formation of the small silica spheres presented here rather involves the aggregation of diminutive clusters thus resulting in a porous structure with a density below that of nonporous silica and a significant specific micropore surface area

[1]

. The information about the microporous structure of the gels

backbone is essential for a detailed understanding of the organic coating mechanisms. The samples coated with different epoxides were characterized with respect to their surface area through nitrogen sorption measurements and their final particle size by SAXS experiments. Part of the normalized scattering intensity of the reference and the three epoxide coated samples is shown in Figure 4. The change in the mean particle size upon coating can be estimated by applying the relation ∝ 1/ at the scattering minima, with the scattering vector q

[31]

. The

curves of the BADGE and the PPGGE9 samples are very similar to the scattering curve of the reference sample, thus the particles can be assumed to have almost identical radii. The radii are derived by a least squares fit of a spherical function [25] on the scattering data shown in Figure 4.

FIGURE 4

11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

The corresponding overall surface area and its fraction due to microporosity derived from the nitrogen sorption measurements are shown in Table 1. A decrease of the overall surface area was expected for all samples as the crosslinking of the amine groups is enhanced in the necks formed by adjacent particles of the inorganic network. The crosslinking of neighboring amine groups in the neck region leads to a reduction of Sext upon coating, which is observed for all samples. The coating with short epoxide molecules like BADGE causes a negligible increase in particle size whereas SBET simultaneously decreases by more than 50 % compared to the reference. These results suggest that the BADGE molecules with a length of less than 1 nm react either with the amine groups in the micropores or cover the micropore entrance. As a consequence the microporosity inside the silica particles becomes inaccessible for the nitrogen molecules thus causing a significant decrease of Smicro. A similar observation has been reported recently while studying the location of the polymer on a cross-linked silica aerogel [32]. Obviously, the additional organic component does not affect the strength and stability indicating a poor crosslinking with the gel structure. The short BADGE molecules are not able to form interparticular bridges to support the weak necks of the inorganic network. The data of the sample crosslinked with PPGGE9 show a coating thickness of 1 nm and the measured surface areas are similar to the values determined for the BADGE sample. The longer PPGGE9 molecules seem to build a dense mesh-like coating that covers the micropores almost completely thus strongly reducing Smicro. Since the molecules configuration is not stable enough to form interparticular bridges over a longer distance, the PPG chains are arranged closely along the surface of the modified gel structure. This lack of interparticular bridges to strengthen the gel structure leads to a large amount of shrinkage during drying. The fit of a particle size distribution to the PEGGE sample reveals a larger mean particle size and slightly broader size distribution compared to the 12


Page 13 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


reference, BADGE and PPGGE9. This can be concluded from the characteristics of the scattering data shown in Figure 4: the minima of the PEGGE data are slightly less pronounced and are shifted to smaller q-values thus indicating larger particles. Compared to the reference the overall surface area SBET decreases only slightly and Smicro remains nearly constant. This suggests that the PEGGE molecules build a mesh-like coating which does not cover the micropores. Instead, the molecules form a branched network of PEG chains each stabilized by hydrogen bridges. The PEG configuration (Figure 3) on the surface of the silica network keeps the microporous structure of the modified silica particles accessible for the nitrogen adsorption (see Smicro, Table 1). The PEG chain configuration in combination with a high amount of interparticular bridges result in a significantly increased mean coating thickness. Unlike the other epoxide coatings the deviations from the spherical shape of the particles and the smoothed necks become visible in the SEM image (Figure 5). The reduced shrinkage during drying observed for the PEGGE coated sample (Table 1) is caused by a high amount of interparticular bridges due to PEGGEs length and the stabilization of the configuration through hydrogen bridges. Although the modulus E is similar (Table 1), a small spring-back effect [3][33] was observed for the PEGGE-coated sample compared to the irreversibly shrinking reference gel.

FIGURE 5

To investigate the influence of the amount of PEGGE on the resulting gel structure and the shrinkage during drying, the E/A value was varied from 1.5, 3, 5, 10 up to 20. The properties of the resulting xerogels are shown in Table 2.

13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 30

TABLE 2

The density, linear shrinkage and Young’s modulus decrease as the amount of PEGGE increases. The density is shown as a function of the E/A value in Figure 6. It decreases almost linearly despite the significantly rising mass of the sample. This originates from a reduced shrinkage as a result of an increased degree of crosslinking. Since a complete crosslinking of all amine groups on the surface of the network is not expected for the theoretical E/A value of 0.5, the degree of crosslinking was expected to rise with an increasing E/A value. However, from a theoretical point of view one can expect that for E/A > 5 no amine groups are remaining for the reaction with the epoxide. The further decrease of the shrinkage can be caused by different effects including secondary amines that increase the crosslinking, adsorption of PEG chains at the necks of the network and the influence of the viscosity of PEGGE during drying. The reaction of PEGGE and an aminopropyl group on the surface of the silica particles generates a secondary amine of high basicity. These groups are able to react with an epoxide group of another PEGGE molecule, thus twice the number of PEGGE molecules can participate in the crosslinking. The excess of epoxide molecules can be adsorbed on the surface without a covalent bond. Adsorption of PEGGE molecules at the necks of the network contributes to the mechanical stability of the gel structure during ambient pressure drying as they are not likely to decompose or evaporate during this process step. The epoxide groups are likely to react with the water in the pore liquid, resulting in inert PEG chains that support the gel structure similar to organic nanofibers. Moreover the viscosity of the unlinked PEGGE molecules can influence the shrinkage during drying

[1]

. It

represents one component of the pore liquid whereas the main part is ethanol. The favored 14


Page 15 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


evaporation of ethanol results in a graded concentration on the surface of the gel, which itself causes diffusion of ethanol from the interior. The tension in the pore liquid that causes a net flow toward the interior can compensate this diffusive flow. In case of a total compensation, no pressure occurs in the pores causing shrinkage during drying [1].

FIGURE 6

Regarding the overall network surface area, SBET decreases between E/A = 3 and E/A = 20 (see Table 2). This suggests that the structure becomes less porous indicating the presence of larger clusters as seen in Figure 7. Hence, a very high amount of PEGGE affects the gelation leading to a clustering of the silica nanoparticles and consequently to a less homogeneous and less porous structure. If the organic coating is considered to be a temporary support of the gel structure during drying, it can be removed afterwards. A temperature of 310 °C is suitable for a complete removal of the organic components by oxidative decomposition without having an impact on the inorganic backbone (e.g. sintering). This procedure yields an inorganic xerogel with a decreased density accompanied by additional, however negligible shrinkage. Starting with a density of 218 kg/m³ for a PEGGE-coated sample with E/A = 20, the pure inorganic sample features a density of 179 kg/m³ after the oxidative decomposition (Figure 6). Therefore the specific application of an organic coating and the subsequent removal can be used to obtain inorganic low density silica gels.

4. Conclusion 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 30

Highly porous silica based xerogels with large primary particles were obtained by using a new synthesis procedure that reduces shrinkage during drying. The gel structure is based on silica nanoparticles with a uniform size of about 30 nm and a very narrow size distribution. The inorganic network was modified by APTES to enable a covalent bonding of an organic coating. An epoxide screening revealed the lowest shrinkage for samples coated with poly(ethylene glycol)diglycidyl ether due to the stabilized configuration of the PEG chains by hydrogen bonds. An increase in the epoxide amount led to a further reduction of the density and the shrinkage. The lowest obtained values were 218 kg/m³ at a linear shrinkage of only 9 %. A subsequent oxidative decomposition of the organic components at 310°C yields an inorganic xerogel structure with a density as low as 179 kg/m³.

16


Page 17 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References

(1) Brinker, C.J. und Scherer, G.W. Sol-Gel Science. s.l. : Academic Press, NY, 1990. (2) Einarsrud, M. J. Non-Cryst. Solids. 1998, 225, 1. (3) Prakash S. S., Brinker C. J., Hurd A. J., Rao S. M., Nature 1995, 374, 439. (4) Leventis N., Palczer A., McCorkle L., Zhang G., Sortiriou-Leventis C., J. Sol-Gel Sci. Technol. 2005, 35, 99. (5) Leventis, N. Acc. Chem. Res. 2007, 40, 874. (6) Leventis N., Sadekar A., Chandrasekaran N., Sotiriou-Leventis C., Chem. Mater. 2010, 22, 2790. (7) Ilhan, U.F., Fabrizio, E.F. und McCorkle, L. J. Mater. Chem. 2006, 16, 3046. (8) Zhang, G., et al. J. Non-Cryst. Solids. 2004, 350, 152. (9) Meador, M.A.B., Fabrizio, E.F. und Ilhan, F. Chem. Mater. 2005, 17, 1085. (10) Fidalgo, A., Rosa, M.E. und Ilharco, L.M. Chem. Mater. 2003, 15, 2186. (11) Fidalgo, A. und Ilharco, L.M. Microporous Mesoporous Mater. 2005, 84, 229. (12) Yoldas, B.E., Annen, M.J. und Bostaph J. Sol-Gel Sci. Technol. 2000, 12, 2475. (13) Phalippou, J., et al. Opt. Mater. 2004, 26, 167. (14) Igarashi, K., Tajiri, K. und Tai, Y. Z. Phys. D Atom. Mol. Cl. 1993, 26, 207. (15) Wang, P., et al. Microporous Mesoporous Mater. 1991, 24, 777. (16) Stangl, R., et al. Sol. Energy Mater. Sol. Cells 1998, 54, 181. (17) Stöber, W. J. Colloid Interface Sci. 1968, 26, 62. (18) Mulik S., Sotiriou-Leventis C., Churu G., Lu H., Leventis N., Chem. Mater. 2008, 20, 5035. (19) Iler, R. K. The Chemistry of Silica. s.l. : J. Wiley & Sons, New York, 1979. (20) Vrancken K. C., Van der Voort P., Possemiers K., Vansant E. F., J. Colloid Interface Sci. 1995, 174, 86. (21) Forest, L., Gibiat, V. und Woigner J. Non-Cryst. Solids. 1998, 225, 287. 17



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(22) Gross, J., Reichenauer, G. und Fricke J. Phys. D: Appl. Phys. 1988, 21, 1447. (23) Brunauer, S., Emmett, P.H. und Teller J. Am. Chem. Soc. 1938, 60, 309. (24) Lippens, B.C. und Boer J. Catal. 1965, 4, 319. (25) Glatter, O. und Kratky, O. Small Angle X-Ray Scattering. s.l. : Academic Press, London, 1982. (26) Pontoni, D., Narayanan, T. und Rennie Langmuir. 2002, 18, 56. (27) Gartmann, N., et al. J. Phys. Chem. Lett. 2010, 1, 379. (28) Heymann, B. und Grubmüller, Chem. Phys. Lett. 1999, 307, 425. (29) Oesterhelt, F., Rief, M. und Gaub, H.E. New J. Phys. 1999, 1, 6.1. (30) LaMer, V.K. und Dinegar, R.H. J. Am. Chem. Soc. 1950, 72, 4847. (31) Boukari, H., Long, G.G. und Harris, M.T. J. Colloid Interface Sci. 2000, 229, 129. (32) Mohite D. P., Larimore Z. J., Lu H., Mang J. T., Sotiriou-Leventis C., Leventis N., Chem. Mater. 2012, 24, 3434. (33) Smith, D.M., et al. J. Non-Cryst. Solids. 1995, 186, 104.

18


Page 18 of 30

Page 19 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table Captions Table 1. Summary of the data derived from the epoxide screening with BADGE, PPGGE9, PEGGE at a ratio of E/A =3 and a reference sample. The samples were analyzed with respect to their density ρ, linear shrinkage ∆l/l and Young’s modulus E. Furthermore the overall specific surface area SBET and the specific surface area due to the micropores Smicro were determined by nitrogen sorption. Table 2. Summary of the data derived for samples prepared with different amounts of PEGGE (E/A values of 1.5, 3, 5, 10 and 20). The samples were examined with respect to their density ρ, linear shrinkage ∆l/l and Young’s modulus E. Furthermore the overall surface area SBET and the contribution of the micropores to the surface area were determined by nitrogen sorption.

19



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 30

Tables Table 1 Epoxide Reference BADGE PPGGE9 PEGGE

ρ [kg/m³] 353 ± 18 372 ± 19 368 ± 18 329 ± 16

∆l/l [%] 44 ± 2 48 ± 2 43 ± 2 31 ± 2

E [MPa] 30 ± 9 48 ± 14 67 ± 20 38 ± 11

SBET [m²/g] 189 ± 9 84 ± 4 82 ± 4 168 ± 8

20


Smicro [m²/g] 63 16 9 62

Page 21 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2 E/A 1.5 3 5 10 20

ρ [kg/m³] 354 ± 18 329 ± 16 316 ± 16 269 ± 13 218 ± 10

∆l/l [%] 34 ± 2 31 ± 2 29 ± 1 20 ± 1 9±1

E [MPa] 71 ± 21 38 ± 11 26 ± 8 12 ± 4 10 ± 3

SBET [m²/g] 94 ± 5 168 ± 8 104 ± 5 34 ± 2

21


Smicro [m²/g] 17 62 26 7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 30

Figure Captions Figure 1. silica gels.

Bifunctional epoxides used to crosslink the amine modified surface of the

Figure 2. Scanning electron microscopy of a xerogel based on silica nanoparticles with a diameter of 27.2 nm. The gelation was catalyzed by an APTES amount that theoretically corresponds to a 113 % coverage of the surface. Figure 3. Schematic representation of a section of a PEG chain. The hydrogen bonds (gray) between the oxygen atoms of PEG and water molecules of the pore liquid stabilize the chain configuration. Figure 4. Normalized scattering intensity as a function of the scattering vector q measured by SAXS experiments on BADGE, PPGGE9 and PEGGE coated samples as well as a reference sample. The numbers correspond to the particle radii derived by fitting the experimental data with the scattering form factor of spheres. Figure 5. Scanning electron microscopy of a xerogel based on silica nanoparticles with a size of 27.2 nm. The APTES modified gel structure was coated with PEGGE at an E/A ratio of 3. Figure 6. The density of the PEGGE coated samples as a function of the E/A value. An example for the further reduction in density by an oxidative decomposition of the organic components is shown for an E/A value of 20. Figure 7. Comparison of scanning electron micrographs of PEGGE coated xerogels at an E/A value of 3 (a) and 20 (b).

22


Page 23 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figures Figure 1

23



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2

24


Page 24 of 30

Page 25 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3

25



Figure 4

scattering intensity [a.u.]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Reference 16.5 nm BADGE 16.6 nm PPGGE9 17.5 nm

4

10

PEGGE

22.1 nm

3

10

2

10

1

10

0.1

-1

q [nm ]

1

26


Page 26 of 30

Page 27 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5

27



Figure 6 400 350 ρ [kg/m³]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

300 250 200 0 0

Oxidative Decomposition

5

10 E/A

15

20

28


Page 28 of 30

Page 29 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7

29



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

TOC graphic

30


Page 30 of 30

Cross-Linked Monolithic Xerogels Based on Silica Nanoparticles

Recommend Documents