This third article in a four-part series concerning ceramic slips will focus on simple and complex deflocculation systems.
Sodium hydroxide (NaOH) is a simple deflocculant. Clean clays or fine quartz slurries-or clay-based bodies that include both materials-can be deflocculated well with NaOH at pH 9-10, and feldspar slurries are naturally deflocculated just from solubilized alkali. The dissociated hydroxyl ions from the base react to form water with hydronium ions extracted from the particle surfaces, leaving a negative, almost purely electrostatic charge on the surface because the sodium cation is so soluble and active. This action sets up an electrical double layer that causes particles to repel each other, and the system deflocculates. However, this method results in higher sodium ion levels than deflocculants that are effective at low pH levels, which increases the possibility of over-deflocculation or reflocculation.
At low viscosity, the deflocculated slip will easily settle and form a hard cake. Even at a high slip density, casts made with only NaOH will be relatively hard and brittle, because little buffering through a plasticizing water layer is left on the particle surfaces. The sodium ion is so large that it destabilizes the bound water layer. This problem occurs with the simplest of deflocculants, since they do not have any steric repulsive or plasticizing effects. Using more plastic clays can improve the situation because such clays either have relatively high surface area or contain humate, which can act as a deflocculant with both electrostatic and steric effects. In both of these cases, the amount of the plasticizing layer increases. A good deflocculant minimizes the formation of hard cake, even at full deflocculation, by increasing the amount of the plasticizing water layer.
To better understand this correlation, one must understand the difference between electrostatic and steric repulsion. In electrostatic repulsion, the repulsive energy generated by a charged particle surface decreases with distance due to a higher concentration of counter-ions than co-ions near the particle surface, and reaches zero when the concentrations of both ions become constant. Steric repulsion energy is not due to the particle charge, but to the structure of the surface layer and its kinetic activity. If the surface is coated with a complete monolayer of surfactant, the new surface layer (surfactant) is essentially incompressible. When another particle approaches, the repulsion increases from zero to maximum over a very short distance. Deflocculation with sodium hydroxide increases the particle charge, producing electrostatic repulsion, but tends to remove surface structure that could give steric repulsion. Deflocculation with nonionic polymers adds structure to the surface, thus providing steric repulsion only, while ionic polymers provide both steric and electrostatic repulsion.
It has been shown that at least for one organic polymeric deflocculant (polyacrylic acid), the pH used determines how much the deflocculant will adsorb.1 This concept should be true for most ionic polymers of this type. A higher pH will extend the polymer more fully, allow more adsorption and produce a thicker adsorbed layer. The result is both increased electrostatic (more charge) and steric (more rigid structure) repulsion. A lower pH will give a lesser charge, which is probably more appropriate for a partially flocculated casting slip. It will also provide a steric structure more like a ball that has some resiliency, which is more appropriate for the formation of a plastic structure in slip. This ball-like structure has been seen in electron microphotographs of humic acids. The more familiar name for these steric polymers is "protective colloids."
Figure 1. The formation of an orthosilicate monomer.
Sodium silicate is normally produced by melting quartz with Na2
CO, then dissolving the resulting glass with steam. Carbon dioxide is volatilized during the melting process, leaving Na2O behind to react. Sodium silicate can also be made (very slowly) by dissolving quartz with NaOH solution (Na2O + H2O = NaOH).
In a solution, as the amount of dissolved quartz increases, the pH decreases because the hydroxyls in the solution are neutralized by surface silicate ions that have been dissolved. This is shown conceptually in one dimension in Figures 1 and 2. In a melt, the hydroxyls and water are non-existent or transitory; therefore, no ionization occurs. Silica (quartz) is treated in both figures as a polymer with a hydroxyl coating, and the resulting sodium silicate likewise has polymeric character.
Figure 2. The formation of a disilicate dimer.
Figure 2. The formation of a disilicate dimer.An orthosilicate monomer is formed in Figure 1, while a disilicate dimer is formed in Figure 2 because there is not enough NaOH to form two monomers. If the amount of silica is not limited, polymers will eventually form even if the initial reactions give only orthosilicate.
However, sodium silicate solutions never have a neutral pH as received. Unlike carboxylate systems, which can be ionized and quite soluble at pH<7, there must always be excess NaOH present in sodium silicate (water glass) to maintain the high pH in sodium silicate solutions to prevent the solution from turning to a gel due to the strength of the O-H covalent bond. Therefore, in a solution, the concept shown in Figure 2 must be modified somewhat to include either some degree of hydrolysis, while still keeping the sodium silicate soluble in water, or excess NaOH to give a higher pH and complete ionization. Conceptual examples are shown in Figures 3 and 4.
Figure 3. In a solution, the concept shown in Figure 2 must be modified somewhat to include some degree of hydrolysis, while still keeping the sodium silicate soluble in water.
The solution viscosity will increase over time with the dissolution of silica. The additional silica can dissolve until the mole ratio of silica to soda reaches about 4+:1, and viscosity is very high. At a 4:1 ratio, on average four silicate anions with 16 Si-O bonding sites are held in solution by only two solubilizing sodium cations. The other Si-O sites bond with each other or hydronium to form Si-O-Si or Si-OH, forming a higher or lower polymer, respectively. More time is necessary to produce increased ratios, and a higher temperature is necessary to maintain the solution at a reasonable working viscosity with a higher ratio. The example in Figure 5 has a 4:1 ratio with the maximum hydrolysis, and hence the minimum molecular weight, for that ratio. It will exist in a diluted state but in a three-dimensional form.
Figure 4. Alternatively, the concept shown in Figure 2 could be modified to include excess NaOH to give a higher pH and complete ionization.
Commercial water glass, which has a high pH and is therefore highly ionized, can be gelled by acidifying with or without diluting with de-ionized water. The acid decreases pH by neutralizing hydroxide (and slightly diluting silica content), while in Figures 3 and 4, increasing the dissolved silica decreases pH by converting hydroxide to water. The acid ion displaces the sodium as the neutralizing cation, with the sodium becoming part of a neutral salt in solution with Cl-, NO3-, or SO4-2, depending on the acid. Therefore, the SiO2.Na2O ratio in the polymer effectively increases. Below pH 10.5, the result is increasing gel formation and viscosity with additional acid. Adding NaOH destroys the gel again.
Figure 5. A solution with a 4:1 ratio with the maximum hydrolysis, and hence the minimum molecular weight, for that ratio.
Gelation is the self-polymerization or condensation of soluble silicate structures to form a hydrous, amorphous gel structure of silicate. Precipitation of silicate is the crosslinking of silicate molecules by multi-valent cations (i.e., Ca+2, Mg+2, Al+3, Fe+2, etc.). (In slip, in the presence of multi-valent cations, the precipitation of silicate salts is in competition with the formation of the gel due to the near neutral slip pH.) Sodium orthosilicate (Na4SiO4) forms an insoluble precipitate with multi-valent cations molecule by molecule. However, with increasing polymerization, a higher cationic concentration can be tolerated without precipitating the whole molecule due to the solubilizing effect of all the remaining sodium associated with the polymer. At the same time, the cation acts as the crosslinking agent to further increase the molecular weight of the polymer. This is one reason why gypsum additions are effective, but neutral salts should still be minimized.
Sodium silicate is available in a number of silica/soda ratios and solution concentrations, with the viscosity decreasing as the ratio or concentration decreases. Higher ratios are more powerful deflocculants. Silica (SiO2) and soda (Na2O) are equal in molecular weight; therefore, their molar ratios and weight ratios are equal. The N-brand of sodium silicate (supplied by PQ Corp., headquartered in Valley Forge, Pa.) used in many ceramic applications has a silica/soda ratio of about 3.3:1 and a high viscosity. The viscosity is due to its polymeric nature, which actually begins to appear at ratios well below 1:1. Silicon29 nuclear magnetic resonance and infrared spectroscopy have been used to determine relative concentrations of the various structures as a function of silica/soda ratio,2 as shown in Figure 6.3 The occurrence of dimers among the predominant monomer begins at a ratio as low as 1:20 and is significant at a ratio 1:10 to 1:5. Trimers and tetramers maximize between ratios of 0.52:1 and 1.65:1, the range where monomer would be expected to be high. Larger ring structures dominate at ratios of 1.5:1 to 2.5:1, and more complex structures dominate at higher ratios. The larger ions are two- or three-dimensional condensation products of the silicate monomer, SiO4-. In more siliceous silicate solutions, i.e., SiO2:Na2O > 2.0, some of the silicate condenses to polymeric (colloidal) silica.4
Figure 6. The relative concentrations of the various structures as a function of silica/soda ratio, as determined by silicon29 nuclear magnetic resonance and infrared spectroscopy.
The average molecular weight (MW) for a 1 molar solution of a 3.3:1 ratio is in the range of 190-250 (which is lower than the MW of the example above) and 350-400 for a 6M solution.2 However, a substantial fraction of the molecules have a higher MW, as indicated by Figure 6. Both of these solutions would be at a high pH. These MWs are low compared to the MW ranges most effective for organic polyelectrolyte deflocculants, and in slip the silicate concentration is much less than 1M. Therefore, the polymerization due to low pH and crosslinking by multi-valent cations must be necessary to increase steric repulsion in the deflocculant. If there is not enough (or too much) polymerization, the effectiveness of the deflocculant will be limited.
Theoretically, a silica/soda ratio of 1:2 is necessary to produce orthosilicate monomer (2Na2O+SiO2 = Na4SiO4) molecules, where every oxygen atom in the silica is associated with one sodium atom. A higher ratio should give a polymer, because removing any sodium requires new Si-O-Si to replace the Si-O-Na+ bonds. Spectroscopy indicates that polymerization by condensation begins at a much higher soda content-a ratio of only 1:20. This effect occurs because some of the ionic bonding changes to covalent bonding, which is much stronger and therefore thermodynamically favored. This change indicates the strong tendency for polymerization by condensation, which in solution produces NaOH as a byproduct, giving increased pH.
On that basis, the polymer should exist in an anhydrous melt of SiO2 and Na2O in proper ratio. But the polymerization must actually be enhanced by the presence of water molecules, which tend to hydrolyze Na2O to NaOH, extracting some soda from the monomer or the existing polymer. This is a reverse explanation for the high pH of all sodium silicate solutions.
Extracting Na2O from the sodium orthosilicate monomer increases the silica/soda ratio, but for ratios between 1:2 and 1:1, the resulting average polymer must be linear. Single ring polymers have a 1:1 ratio, in which the monomeric unit is Na2SiO3, while higher ratios result in complex-ring or three-dimensional polymers. Each ratio has an associated range of polymer types due to kinetic factors. These concepts are also in accord with the information from spectroscopy.
In the presence of a constant amount of water, as the silica/soda ratio increases, the polymer becomes more complex and of a higher molecular weight. If the system is treated as polymerizing over time, the covalent bonding becomes internal and ionic bonding with the remaining sodium ion remains external. At a high pH the polymer is soluble in water.
As indicated by the molecular weights noted previously, adding water to dilute sodium silicate can alter the polymerization state. Water breaks down the polymer by hydrolyzation of more Si-O-Si bonds. Stabilization of the new polymer structure after a change in water content, silica/soda ratio or acidity occurs in a matter of only minutes to hours, depending on the concentration, temperature and agitation of the mixture.3 Stability of the molecular structures is improved with increased complexity (a high silica/soda ratio) and/or increased solution concentration. Therefore, if the highest molecular complexity is desirable, sodium silicate should not be diluted. If dilution is necessary, as it normally is for ceramic slips, the starting solution should have the highest possible silica/soda ratio to maintain the polymeric integrity and gel structure.
Sodium silicate has a high hydroxide content and pH from hydrolysis of some of the soda, which keeps the silicate in solution. If it is acidified to a pH below 10.5, a strong poly-silicic acid gel forms, leaving sodium ion in solution with the anion of the acid. The same change occurs if the silicate is added to a clay system with pH of 5-8. Some multi-valent cations are precipitated as silicate salts, and some are included in the polymer. The hydroxide is neutralized to form water, with hydronium ions displaced from the clay by the sodium ion (at a lower pH if organics are associated), and the replacing sodium associates loosely with the surface of the clay or organic colloid (its effective anion), increasing the net surface charge. The remaining sodium silicate forms mostly poly-silicic acid gel, which coats the particle surfaces, and some of the remaining sodium acts as the charge producing cation. The charge generates electrostatic repulsion, but the polymer itself also produces steric repulsion.
Experiments with Sodium Silicate and NaOHAs described previously for sodium silicate, sodium hydroxide is a gel breaker. In almost every water-based slurry or slip I have worked with, sodium hydroxide was included in the recipe as a modifier for the deflocculant. The materials ranged from coal to silicon carbide to whiteware bodies, and from coarse to fine particle size ranges, but the minimum viscosity achievable with the deflocculant could always be lowered further with additions or substitutions of NaOH. There are three reasons for this effect. The first is that hydroxide is effective as a deflocculant on clean surfaces, producing a high charge density without adsorbing large anionic species that crowd the inter-particle spacing. The second is that, for some cations, hydroxides are less soluble than other anions, such as sulfates; therefore, hydroxide will remove some soluble cations by precipitation, leaving behind a more highly charged surface. The third reason is that hydroxide will react with some deflocculants and make them stronger by neutralizing acidic sites in the molecular structure of the deflocculant. Sometimes more than one of these effects occurs simultaneously with a deflocculant addition.
Experience has shown that while there is often a benefit to using NaOH, employing it to give minimum viscosity is usually not good, and excess use can lead to disastrous results even if not at minimum viscosity. The gel formed by the polymeric deflocculant contributes the steric repulsion, and NaOH destroys that gel-and along with it, some potential plasticity in the slip and strength in the body.
In one instance, a low silica/soda ratio water glass was tested in a well deflocculated silicon carbide (SiC) used to cast large, thin plates. When a slip with an equivalent slip specific gravity and viscosity was cast, the strength of the dried ware decreased so much that it could no longer be handled at all. Needless to say, the original deflocculant was reinstated. In this case, the strength imparted by the more highly polymerized deflocculant, which was also acting as a binder, was very necessary.
In another type of technical slip, three different deflocculants were used: sodium silicate for strength, a high viscosity organic deflocculant to give a type of plasticity, and sodium hydroxide as the modifier for both. That situation was difficult to control because the modifier had different effects on the other two deflocculants. This challenge was made more complex by the aging of the slip, as well as any variation in raw materials. The pH was raised with NaOH to improve the organic deflocculant, but the resulting pH was too high for sodium silicate. A simple sodium silicate deflocculant with a proper silica/soda ratio would probably have worked better.
Figure 7. An over-deflocculated slurry at a higher shear will have a dilatant flow at the same viscosity at which a less deflocculated slurry has a pseudoplastic flow.
In sanitaryware casting, a less than fully deflocculated slip is desirable to achieve a high casting rate, as well as to improve the uniformity and plasticity of the cast. NaOH will deflocculate the slip well but imparts little strength, and in combination with water glass decreases dried strength. Because less deflocculation increases the moisture content, the porosity of the cast and its dry strength, the silica content of the water glass should be kept high so that it is more polymerized and can act better as binder. Usually NaOH or Na2CO3 is added to ball clay to digest included organics. However, it should not be added where no organics are present because it decreases the binding effectiveness of the water glass, and it should not be necessary if the feldspar is fine.
Figure 8. Flow curves relating apparent viscosity to spindle speed for suspensions of Old Mine #4 at 19.4 vol. % solids (1.30 g/cc) at various amounts of electrolyte additions and measured directly following a period of high shearing.
Coal-water slurries are a different system requiring different deflocculant types but using similar principles. They are formulated for maximum density to minimize water content for pumping and/or direct burning. Organic deflocculants give a minimum viscosity, but NaOH will usually decrease it slightly, or will react with soluble humates in the coal to form an inexpensive replacement deflocculant. In both cases, better pumping rheology can be achieved by cutting back on deflocculant. This is in line with Figure 7,4 which shows that an over-deflocculated slurry at a higher shear will have a dilatant flow at the same viscosity at which a less deflocculated slurry has a pseudoplastic flow. The same phenomenon would be seen if Figures 8 and 9 were measured at higher shear rates. Coal slurries are not as pseudoplastic as clays. For this reason, the onset of dilatancy should occur at a lower shear rate with coal slurries, but dilatancy is also pushed to a higher shear rate for these clays because of the lower solids content. Pseudoplastic flow is preferred in each case so that the viscosity will increase and the slurry will stabilize (with no separation of coarse and fine particles) when the flow stops.
Figure 9. Flow curves relating apparent viscosity to spindle speed for suspensions of Edgar Plastic Kaolin at 19.4 vol. % solids (1.30 g/cc) at various amounts of electrolyte additions and measured directly following a period of high shearing.
The fourth and final article in this series will focus on other deflocculants and aging effects.For more information about slips and their behaviors, contact Funk at firstname.lastname@example.org .