How do I get the most out of AEROSIL® fumed silica

Many factors influence the successful use of AEROSIL® fumed silica in epoxy resins. Let's take a look at how fumed silica works, how to choose the right grade for your system, how to use it properly and what to bear in mind about the interactions other components can have on its efficacy.

Hebei Silicon Research Electronic Materials Co., L contains other products and information you need, so please check it out.

AEROSIL® fumed silica has been a successful rheological additive in epoxy systems for decades. There is a good chance you will have worked with it at some point. Its success stems from the versatility it gives formulators in building viscosity, thixotropy and in controlling rheology over a wide range of shear and temperature conditions.

As a formulator of epoxy resins, you know that rheology control is an important consideration when designing a system. Take proper care over the rheology profile and you will have a product that resists sedimentation, has excellent shelf life stability, is easier and more accurate to blend (2k systems), can be applied with ease and has a high level of anti-sagging performance. Some of these factors are important before, during and after application &#; making your and your customer's lives easier.

How does fumed silica thicken liquid systems?

First of all, let&#;s take a look at how AEROSIL® fumed silica works as a thickening agent. Fumed silica is supplied in the form of agglomerates that need to be broken down under high shear into aggregates. The latter would ideally have a particle size in the region of 100-300 nm. To achieve this breakdown of particle size you would incorporate the fumed silica with a dissolver, planetary mixer, triple roll mill or kneader. Propeller stirrers are NOT suitable for this process! Dissolvers and planetary mixers are the most popular due to the ease of continuous incorporation and lower energy demands.

Once the particles have been broken down they will reform as a continuous 3D gel network, imparting a yield value and higher viscosity at rest. This is a reversible process - in that applying shear will break the network down and decrease the viscosity. Once again, the network can reform. This behaviour is what gives rise to the shear thinning and thixotropic behaviour AEROSIL® is so well known for in epoxy resins.

OK, so that&#;s how it works on a mechanistic level. To understand why some grades work in certain systems, but not in others, we need to look at the interactions between the fumed silica particles. AEROSIL® grades come both untreated (silanol surface groups) and with hydrophobic silane treatments (various options, some silanol groups remain). Interactions between the silanol groups are responsible for the formation of stable networks. In non-polar fluids, this interaction is maximised. When trying to thicken a polar fluid these molecules solvate the AEROSIL® and lead to a destabilisation of the gel network. This is why relatively large amounts of hydrophilic silica are required when thickening water or ethanol*.

How to thicken an epoxy resin with fumed silica

It is interesting to note that the presence of a small concentration of polar molecules in a non-polar fluid can have a synergistic effect in the thickening behaviour; through a bridging mechanism. Consider the use of a polyamine curing agent for example. This effect works the other way above a certain critical concentration due to the solvating effect mentioned above.

Hydrophilic fumed silica is often not suitable for use in epoxy resins as many grades are polar in nature, solvating the AEROSIL®. Epoxies also have curing agents added to them which can destabilise the gel network. To demonstrate this, we direct your attention to Figure 1.

Figure 1: Viscosity drop of a fumed silica thickened epoxy resin (ARALDIT M, Huntsman), both (left) before and (right) after the addition of curing agent and accelerator. Three grades of AEROSIL® are used at a loading of 5.6% by weight pre-addition and 3.8% post-addition: AEROSIL® R 202 (PDMS surface-treated, most hydrophobic), AEROSIL® R 805 (octylsilane, hydrophobic) and AEROSIL® 300 (highest surface area, hydrophilic).

Notice how the hydrophobic grades increase viscosity in this polar resin to the greatest extent. The addition of polar hardener molecules then massively boosts the viscosity of the AEROSIL® 300 due to the bridging effect mentioned earlier. This effect is short lived however, as the fumed silica particles are increasingly solvated by polymer molecules containing hydroxyl groups that originate during the curing reaction. In conclusion, if you want a stable viscosity profile through cure, then you need to use a hydrophobic AEROSIL® grade such as AEROSIL® R 202, AEROSIL® R 208 or AEROSIL® R 805. These grades have hydrophobic groups on the surface to shield the remaining silanol groups from undesired wetting. Note that both silanol groups and the hydrophobic chains contribute to thickening with these grades.

Evonik&#; have found the same trend of thickening performance across all tested epoxy resins. An example of this behaviour is shown in Figure 2.

Figure 2: Comparison of the thickening effect of different AEROSIL® fumed silica grades at a loading of 4% by weight in a plasticized DGEBA resin at room temperature.

Fumed silica surface area usually has a significant effect on the thickening performance of liquid systems, however - when it comes to epoxy resins - surface chemistry is far more important.

How do silanes influence fumed silica rheology in epoxy resins?

Bi-functional silanes are frequently employed in epoxy adhesives. You will likely be familiar with names such as GLYMO and other Dynasylan® products. These molecules can have a huge impact on the rheology of fumed silica thickened epoxy resins - if the wrong grade is used. Silanes are known to decrease the viscosity in pure epoxy resin through a dilution effect. Please see Figure 3, comparing the left-hand side with the right.

Figure 3: Comparison of viscosity changes of an epoxy resin (EPIKOTE 216) thickened with various AEROSIL® fumed silica grades, as a function of storage time at 50°C. (left) Without silane addition. (right) with Gamma-glycidocypropyl-trimethoxy-silane, Dynasylan® GLYMO.

It is clear that the addition of silane increases the thickening effect of hydrophilic fumed silica dramatically. Over time this viscosity plummets back down again because of the reaction between silanol groups and the methoxy groups of the silane, resulting in methanol release and the subsequent solvation of the silica; disrupting the gel network.

Hydrophobic grades such as AEROSIL® R202 and R805 are only marginally impacted by the addition of silane. Note the excellent storage stability offered by hydrophobic fumed silica grades, AEROSIL® R 805 in particular.

AEROSIL® fumed silica as an anti-settling agent

Fillers are frequently employed in the formulation of epoxy systems, either to improve performance like shrinkage or to reduce cost. These particles are prone to settlement if not properly stabilised. AEROSIL® fumed silica can slow down, or even prevent, the settling of fillers and pigments. After extended storage times, some sediment may be seen but it is often soft and easy to resuspend. This effect is attributable to the increased yield value that fumed silica imparts. Either AEROSIL® R805, R202 or R208 would perform best in this function.

Fumed silica for the thickening and thixotropy of hardeners for epoxy resins

By adding fumed silica to the less viscous hardener component, blending mistakes can be avoided and a more homogenous mixture obtained. Note that fillers can be suspended in the hardener component with the proper choice and use of AEROSIL® fumed silica.

Now here is where the recommendation can change from those made so far, depending on the polarity of your hardener (see Figure 4). Non-polar polyaminoamides like VERSAMID 140 are thickened best by hydrophilic fumed silica grades (AEROSIL® 300). In high polarity mercaptan hardeners, like CAPCURE 3-800, hydrophobic grades are the best (R805, R202 and R208). It is worth noting that the hydrophobic grades often have better storage stability though.

Figure 4: Thickening of various epoxy hardeners with AEROSIL® 300, AEROSIL® R202 and AEROSIL® R805 at room temperature.

Want more information on hydrophobic silica? Feel free to contact us.

Summary

If you think that one of the samples mentioned above would be interesting to look at then you can submit a sample request on our AEROSIL® fumed silica page. Product brochures and datasheets can also be found on this site. Our sales team is comprised entirely of chemists with decades of industry experience between us so please get in touch if you would like to discuss your project further &#; we are here to help.

* Note that it is possible with a special grade of AEROSIL® called COK 84. Please get in touch if this is of interest.

3.1. Dispersibility of HFS in Slurry and Aqueous Dispersion

3.1.1. Particle Morphology of HFS

In order to obtain clearer particle size results, the microscopic morphology and particle size distribution of HFS were observed via TEM. The results are shown in Figure 2.

Figure 2.

TEM images in different magnifications: (a) 30,000×, (b) 60,000×, (c) particle distribution of HFS dried powder, and (d) the schematic illustration of HFS agglomerate formation.

The TEM image in Figure 2 can more clearly characterize the shape and particle size of HFS. It can be seen from Figure 2a,b that HFS particles have many pore structures. HFS agglomerates have particle sizes around 200 to 300 nm [31]. It is formed by several 50~150 nm aggregates connected by &#;bridges&#; [32,33]. These so-called &#;bridges&#; are actually formed by physical bonds between the edges of the aggregate and other aggregates (Figure 2d). The particle size distribution curves of HFS obtained via data fitting are shown in Figure 2c. Before dispersion, the HFS was formed by many aggregates with a size of about 200~300 nm. The aggregates were linked by physical bonds, such as hydrogen bonds, London forces, and other interaction forces, forming a larger agglomerate. These physical bonds could be reduced by surface property change and broken by high shear force. It can be seen from Figure 2c that the distribution curve of the particle size number of HFS primary particles is normally distributed, with D50 = 13.6 nm. However, the strong chemical bonds could hardly be broken by dispersing; it would take more energy to break the aggregates into primary particles. Ideally, HFS particles may only be dispersed to a particle size of 50~150 nm without any chemical bond breakage [32].

3.1.2. The HFS Dispersibility with Different Dispersants

In order to study the impact of different dispersants on HFS dispersibility, several representative dispersants were selected. According to the formula in Table 2, HFS was dispersed by using dispersant A (sodium hexametaphosphate), dispersant B (sodium dodecyl benzene sulfonate), dispersant C (cationic polymer dispersant), dispersant D (sodium polycarboxylate, anionic polymer dispersant), and dispersant E (non-ionic polymer dispersant). When the addition of the dispersant was lower than 6 wt%, no matter which dispersant was used, the HFS slurry seemed like gel. The results are shown in Figure 3.

Figure 3.

HFS slurry prepared with A (anionic dispersant); B (anionic dispersant); C (cationic polymer dispersant); D (anionic polymer dispersant); and E (non-ionic polymer dispersant).

From Figure 3, it can be confirmed that the HFS slurries prepared with dispersants A, B, and C are white or yellowish bulk solids, and HFS cannot be wetted and dispersed in the aqueous phase system. The HFS slurry prepared by using dispersants D and E is a light-yellow liquid. The slurry prepared by using these two dispersants has good fluidity, indicating that HFS particles are sufficiently wetted. For the former, HFS has a strong adsorption effect on the long-chain alkyl of the dispersant [34]. The negatively charged carboxyl group of the dispersant is exposed to the solution and forms a solvation chain. The non-ionic polymer dispersant was adsorbed on the surface of the particles through hydrophobic chain segments [35]. However, hydrophilic chain segments extend in the solution, forming a steric effect to prevent agglomeration between particles [29,36,37].

In order to accurately distinguish the wetting state of different dispersant solutions on the surface of HFS particles, the water contact angle was measured. The results are shown in Figure 4.

Figure 4.

The water contact angle of different dispersant solutions (6 wt%) on the surface coated with HFS powder. A (anionic dispersant); B (anionic dispersant); C (cationic polymer dispersant); D (anionic polymer dispersant); and E (non-ionic polymer dispersant).

Due to the alkylation of the HFS surface, water is totally unable to wet the HFS surface, and the water contact angle is even larger than 170°. After the addition of the dispersant (A, B, or C), the wetting condition was significantly improved, and the contact angle was reduced to 110~130°. With the addition of dispersant D, the contact angle was further reduced to 56°, indicating a transition from non-wettable to wettable. The water contact angle of non-ionic dispersant E is only 46°, and the wetting performance is significantly better than that of dispersant D. Furthermore, a large number of anionic surfactants are easy to aggregate in the waterborne system and form ion channels inside the coatings, which will affect the initial water resistance and long-term durability of the coatings, especially for waterborne industrial coatings with high initial water resistance requirements [38]. Therefore, it is reasonable to use a non-ionic dispersant (E) to disperse HFS.

3.1.3. The Impact of Dispersant Dosage on HFS Dispersibility

In order to study how dispersant E addition affects the particle size of HFS aqueous dispersion, the dilution ratio was set at 10 times. Five minutes of ultrasonic dispersion was performed to obtain sufficient dispersion energy. The particle size is shown in Figure 5.

Figure 5.

The relationship between the addition of dispersant E and the particle size.

As can be seen from Figure 5, after adding 2% of dispersant E, the Z-average particle size decreased to 230 nm. In contrast, the HFS was completely unable to disperse without the use of a dispersant. As can be seen from Figure 4, the HFS surface is totally unable to be wetted by water. When the dispersant E addition is around 4~10%, the particle size of the HFS slightly decreases with the increase in usage, showing a nearly linear trend. When the addition of dispersant E reaches 10 wt%, the particle size is the smallest, and the Z-average particle size is 126.4 nm. Although the dispersant addition increased to 12 wt% and 14 wt%, the particle size did not change any further. The reasons are as follows. (1) When the addition of the dispersant increased, the adsorption of the dispersant molecules on the HFS particles surface also increased. At the same time, the physical bond strength between the aggregates was weakened. As a result, the HFS became more and more easily dispersed, and the particle size became smaller after ultrasonic dispersion. (2) The dispersant reached saturation adsorption on the HFS surface [39,40]. The excess dispersant formed micelles in the solution and had no effect on the particle dispersion. Therefore, the particle size no longer decreased. (3) The strong chemical bonds between the primary particles could not be broken by adding a dispersant. However, the dispersant successfully broke the physical bonds. This study also revealed that the chemical bonds could not be broken via high-speed dispersion and ultrasonic dispersion. As shown in Figure 2, the HFS primary particles&#; average size is 19.7 nm. After dispersion, the size of the agglomerates only decreased to 126.4 nm. In conclusion, when the amount of dispersant is controlled at 10 wt%, the slurry has the best dispersion effect in a waterborne system. No obvious difference in particle size between each HFS aqueous dispersion (10~14 wt% dispersant addition) means that it is hard to reduce particle size by adding any more dispersant.

It can also be reconfirmed from Figure 6 that the dried HFS powder is a large agglomerate of 5~100 μm. When the amount of dispersant addition is not enough (2 wt%), the solution is not enough to moisten large HFS particles. Only when the dispersant addition exceeds 10 wt% can HFS be fully wetted and dispersed, and finally form a uniform slurry.

Figure 6.

Microscope images of HFS in (a) dried powders, (b) 2 wt% dispersant solution, and (c) 10 wt% dispersant solution.

3.1.4. The Storage Stability of HFS Aqueous Dispersion with Different Dilution Ratios

The HFS slurry containing 10 wt% dispersant E was diluted 10 times, 100 times, times, and 10,000 times with the DMEA solution (pH = 8.5). In order to further investigate the storage stability of HFS aqueous dispersion after dilution, the dispersive solution was ultrasonic for 5 min and stayed at a constant temperature (25 °C) and humidity (30%) for 0 d, 3 d, and 47 d. They were used to observe whether there were stratifications, settlements, or other phenomena.

As can be seen from Figure 7, the color of the slurry changed from light yellow to a white turbid zed liquid after 10 dilutions. With the increase in dilution ratio, the dispersion gradually became clear and transparent. After standing for 3 d, the slurry and its diluted dispersion did not change significantly. Even standing for 47 d, stratification, settlement, and other phenomena did not appear in the slurry. This indicates that after the HFS slurry is diluted and dispersed in the waterborne system, the dispersant can be evenly adsorbed on the surface of the particles. Due to the steric hindrance effect between particles increasing, it is difficult to form agglomerates again [41]. Therefore, the storage of the slurries becomes stable. Moreover, it seems that the dilution only has a slight influence on the dispersity and settlement of HFS. Therefore, in order to not add too much water into the coatings, the dilution ratio was selected as 10 times in the preparation of HFS aqueous dispersion.

Figure 7.

The storage stability of HFS slurry with different dilution ratios.

3.1.5. The Impact of Ultrasonic Dispersion Time on the HFS Dispersibility

To confirm the impact of the ultrasonic dispersing time on the particle size of HFS aqueous dispersion, a study was conducted on this. The dispersant usage and dilution ratio were set at 10 wt% and 10×, respectively. The results are shown in Figure 8.

Figure 8.

The relationship between the ultrasonic dispersion time and the particle size.

As can be seen in Figure 8, after high-speed dispersion, the particle size of the HFS slurry was around nm. After 2 min of dispersion, the particle size of the HFS aqueous dispersion sharply decreased to ~135 nm, which is consistent with the results in the literature [23]. After 10 min of dispersion, the particle size of HFS aqueous dispersion was only 126.4 nm. The dispersion of HFS is a dynamic equilibrium process. From the perspective of dynamics and thermodynamics, this process is mainly divided into three stages: To reduce the surface energy, the initial particle form aggregates with other particles to stay at a low energy state. With the increase in ultrasonic dispersion time (0~2 min), the agglomerates absorb mechanical energy and heat and split into smaller secondary particles. At this stage, the energy provided via ultrasonic dispersion is much larger than that consumed by the dispersion of HFS. When the ultrasonic time is 2~5 min, the curve of the particle size showed a gentle downward trend. This is because the surface area of particles increases rapidly after they split into primary particles, and the increased surface energy &#;G is much larger. When the ultrasonic time is more than 5 min, the particle size of silica has almost no obvious change. This is due to the limited energy provided via ultrasonic dispersion of a certain power. In this state, the agglomeration of aggregate particles forms a dynamic equilibrium with the dispersion of agglomerate particles. Therefore, due to the limited power of ultrasonic dispersion, the particle size gradually decreases with the increase in ultrasonic time but eventually establishes equilibrium and stays steady. From the perspective of reducing energy consumption, it seems that ultrasonic dispersion for 5 min is the most reasonable, as both energy consumption and particle size are already small enough at this time. It can be inferred that, after the ultrasonic dispersion, the particle size will form agglomeration again due to the energy loss of the system. In addition, if the temperature of the system decreases, the settlement speed will become faster. Larger particles will wrap the smaller particles in the settlement process, which will also lead to an increase in particle size. Storage at the slurry status and applying dilution and ultrasonic dispersion before using water dispersion may effectively avoid this problem, which is strongly recommended in this study.

The energy consumed per kilogram of HFS dispersion can be roughly calculated based on the HFS aqueous dispersion mass, ultrasonic dispersion time, and instrument power. Excitingly, the energy consumption of HFS aqueous dispersion prepared in this paper is 120 kJ/kg, which is much lower than the results (165.6 kJ/kg) in reference [26].

3.1.6. Particle Size Distribution of HFS under Optimized Conditions

The optimized process parameters (dilution 10×, ultrasonic dispersion time 10 min) obtained from the above analysis were used to prepare HFS aqueous dispersion. The particle size of the HFS aqueous dispersion was measured by using a laser particle size analyzer. The results are shown in Figure 9.

Figure 9.

Particle size test results under optimal dispersion conditions: (a) storage after 0 d, (b) storage after 47 d.

According to Figure 9a, under the optimal dispersion conditions, the particle size (D50) of hydrophobic HFS particles is 27.2 nm and PDI = 0.326. The Z-average particle size is 126.4 nm. However, there is still a gap between the particle size distribution calculated via TEM in Figure 2. Under optimized conditions, the particle size of the agglomerates decreases significantly, even very close to that of the aggregates. However, it still cannot be dispersed to the size of primary particles due to the high energy needed to break chemical bonds. After a 47 d storage, the D50 value of HFS aqueous dispersion is 31.0 nm, which is still close to that before storage. The Z-average particle size is 129.8 nm, which is also close to the value before storage. However, the PDI slightly increased, which means that flocculation was happening after storage for 47 d.