This second article in a four-part series concerning ceramic slips will focus on rheology and slip stability as measured by gelation (gel) curves, the effect of plasticity imparted by clays and associated chemistry, and casting properties.

Rheology and Slip Stability

Gel curves tell us about slip stability. When measuring a freshly stirred slip on a viscometer, rapid changes in viscosity occur for up to one minute as the slip recovers from mixing. The exceptions are a fully deflocculated slip or a measurement at high shear, where the breakdown of the slip from the mixing is preserved. Following the initial rapid changes, the viscosity decreases (thixotropy), increases (rheopexy) or becomes constant (see Figure 1).¹ The examples of rheopectic behavior are very typical of the shapes of gel curves.

The mark of slip stability is a flat gel curve. This is true for all slips, whether they are fine or coarse, narrow or broad in size distribution, flocculated or deflocculated, high or low viscosity, or high or low specific gravity. The best stability has been achieved when the gelation curve becomes flat very rapidly after a shear rate change. This means the particles/flocs have come to equilibrium within the liquid at the new shear rate.

In slip casting, slip stability is critical because it is indicative of the amount of variability that will occur in moisture content and shrinkage across the cast wall. If slip viscosity changes during casting, the amount of bonding in flocs and the relative positions of particles with respect to each other also changes, which means the inter-particle spacing (IPS) within the flocs is changing and the cast density is changing as the cast forms. A large percentage of drying and firing flaws are due to excess variations in moisture and shrinkage from inside to outside the cast caused by slip instability.

Gel curves for over-flocculated slip build to a maximum viscosity very rapidly and then decrease in viscosity with time, an indication of syneresis. Such a slip would cast quickly initially and to low density, and then theoretically would densify with increased thickness. Gel curves for properly flocculated slips build over a slightly longer time to a maximum and constant viscosity. The cast density is quite uniform for these slips. Gel curves for a slightly over-deflocculated slip build viscosity continuously after the first few seconds of measurement. The cast would become much softer as thickness increases. Gel curves for a totally deflocculated slip build up only slightly or not at all, and then are completely flat. The slip is stable, but casts will be thin and brittle due to lack of plasticity.

The rate of buildup (RBU) as measured on a Brookfield viscometer is a very good indicator of casting uniformity.¹ RBU is calculated as the percent change in viscosity in two minutes compared to the change in 10 minutes. It is measured at a low, but non-zero (e.g., 5 or 6 rpm), shear rate as an approximation to buildup at rest, which also allows an assessment of stability at the low shear rate.

The pseudoplasticity as measured on the viscometer is an indicator of casting rate and cast plasticity. It is calculated from gel curves by comparing the logs of low shear vs. high shear viscosity, each recorded after an equal number of minutes of measurement, which is typically 10 minutes at 100 (or 60) rpm, then 10 minutes at 10 (or 6) rpm. The calculated value is negative for pseudoplastic systems and positive for dilatant systems. Similar values can be calculated using data from graphs such as those shown in Figures 2 and 3.

An over-flocculated (syneretic) slip has a high pseudoplasticity (negative slope) and an RBU over 100%; that is, viscosity decreases after the first two minutes of measurement. A stable flocculated slip has a higher pseudoplasticity (negative slope) and an RBU range of 80-100% (and usually a very flat gel curve), but casts too fast and soft. Neither of these conditions should be used for casting. However, the second condition, preferably at 80% RBU with lower specific gravity (SG), can be used for filter pressing.

A partially deflocculated slip with good casting behavior has an RBU of about 75% with a flat gel curve after four to six minutes. An over-deflocculated slip (for casting) has a lower RBU with viscosity continually increasing, while a fully deflocculated slip has an RBU of 100% and is stable, but very slow casting and non-plastic.

If gel curves are run at each step during the determination of a deflocculation curve, the RBU is shown to decrease from near 100% when well flocculated to a minimum of 30-60% when partially deflocculated (actual minimum depending on a number of factors), and then to increase to 100% again at full deflocculation. A graph of RBU vs. % deflocculation shows that the slip will pass through 75% RBU twice, once while decreasing and once again while increasing. Particularly for casting whitewares, it is important that the slip be used at the first 75% RBU to give the best combination of good casting rate, plasticity and stability.

To achieve a minimum moisture content with maximum moisture stability in the cast, the lowest RBU that gives a flat gel curve after about five minutes at low shear rate should be chosen. This approximates an even faster equilibration at lower and at zero shear (buildup of yield stress). The amount and quality of plastic clays dictates how low the RBU can be set while still giving stability after five minutes. An RBU decrease to even 70% is normally too deflocculated, as indicated by a non-flat gel curve.

The root cause of slip stability is the interaction between the structured layers of water and chemistry at the surfaces of adjacent particles, which quickly slows and stops relative movement when shear is removed. Particles that are chemically bare undergo elastic collisions. This does not occur naturally. Each particle that has been exposed to the atmosphere has adsorbed a thin layer of water, even as a dry powder. Water bonded directly to the surface of a clay mineral possesses different physical properties than that of liquid water.²

This water layer is organized by hydrogen bonding with the solid surface, and the amount of organization depends on the amount of hydrogen bonding and the dimensional "fit" of water unit cells with unit cells of the crystalline structure at the surface of the particle. In water suspension, this layer thickens and becomes structured by further hydrogen bonding of water to itself, is enhanced by the effect of dissolved cations hydrated by the water (three-fourths of the hydrogen in the water molecules is involved in this structural bonding and one-fourth is available for bonds with other ions and to the clay surface), and then at its periphery is graded into the surrounding bulk water.

This interfacial layer has a high viscosity because of its extra bond density, similar to the surface tension of water. Its thickness is also related to the number and ionic size of the hydrated cations necessary to reduce the particle surface charge to zero according to colloid theory. Ions with a low charge/size ratio, such as Na⁺, reduce structure, while ions with a high ratio, such as Ca⁺², increase structure. A low conductivity slurry of clean particles deflocculated by NaOH to high pH gives the closest approximation to a bare surface. The thickness of the bound water fraction, as well as its integrity, is a major factor in stability because it affects the speed of re-formation of particle-particle bonds after shearing.

This re-formation of bonds is the same concept that leads to pseudoplasticity in slips and plasticity in a lower-water-content body. A breakdown of the bonding structure by the applied mechanical energy of mixing decreases viscosity. A buildup of the bonding on cessation of mixing (shear) increases viscosity.

"Plastic" Particles

The foregoing information applies to all particles in suspension. Chemically enhancing the structure of water around each particle improves pseudoplasticity and stability, but using specific types of particles also provides an improvement. This is one reason why we use clays. Most coarse and/or blocky particles are dilatant, rather than intrinsically plastic, and must be blended with other components to achieve desirable properties. Clays add stability to a suspension partly because of their high surface area, but also because of the similarity (affinity or fit) of the crystalline structure of the clay surface to the structure of water,³ which allows an "extension" of the particulate structure into the surrounding water.

The clays' silica layer has the closest structural affinity. Therefore, three-layer clays with two external silica layers give the most enhancement of stability. Also, the platy nature of clays allows for the extension of the enhanced structure (while at rest) over a long range (on a molecular scale) without curvature, while allowing easy slippage under shear. This is pseudoplasticity. In the case of two-layer clays, only the single silica layer has a strong structural affinity with water. The opposite side of the crystal is gibbsitic (alumina-like).⁴ Water adjacent to this layer needs the structural enhancement of additional chemistry to increase its viscosity and the stability of the system.

Methylene blue dye is a nitrogen-containing three-ring organic cation with a planar ring structure that easily adsorbs on structured clay surfaces, partly due to the cationic affinity for negatively charged particles and partly from the hydrophobic effect of its carbon structure. For these reasons, it also adsorbs better on clean or deflocculated surfaces. (Adsorption decreases slightly with flocculation, even using standard procedures. sup> 5) The Methylene Blue Index is used to characterize the amount of structured surface area for clays, and its ratio with surface area (MBI/SSA) has been called the intrinsic plasticity of clays,¹ a measure of clay quality.

Ball clays or kaolins of good plasticity will have an MBI/SSA ratio near 0.45, although the reasons for the high ratio vary. Plastic kaolins are usually well crystallized and enhance the surrounding water structure without additional chemistry, while plastic ball clays might be poorly crystallized but usually have associated organics that improve the structure of the surrounding water. Some clays with a fairly high SSA (20-25m²/g) and an MBI/SSA ratio higher than 0.5 contribute so much stability (and pseudoplasticity) to a suspension that they can become difficult to deflocculate at high concentrations in a body. Those with high SSA and a ratio of 0.3-0.5, such as some montmorillonites, might have similar or even stronger effects. Such clays will show some pseudoplasticity even when fully deflocculated. These clays cannot be used in high concentrations unless the concentration of other plastic components is low. Their expense might make this method uneconomical.

Casting Properties

The casting rate normally decreases as wall thickness increases, and this is a function of the amount of drag on the water moving from the slip through the cast layer into the mold. If the IPS and slip chemistry are constant, the casting rate depends only on the thickness of the cast layer. Porosity in the mold is usually considerably larger than the porosity in the cast, so the mold porosity is not the bottleneck, or rate-limiting, factor. The most drag comes from the smallest pores in the cast, usually associated with a dense initial layer formed from slip that has not had time to build to a stable viscosity. This dense layer has been observed on SEM photos for cross sections of slipcast fine SiC, even after firing.

To minimize this layer, even with a stable slip, the mold must be filled slowly enough to allow the buildup to start even before the slip stops moving. A stable slip produces a cast with a minimum variation in IPS and therefore a minimum variation in shrinkage. An unstable slip produces a cast with a changing IPS and the attendant shrinkage variation, leading to more losses.

If a slip is too flocculated, the casting rate is initially high. However, syneresis densifies the flocs and bring them closer together in the cast, slowing the casting rate and producing a density gradient. If the slip is somewhat deflocculated, the initial casting rate is very slow, and the initial cast layer dense. The slip builds in viscosity over time, which allows a faster casting rate if the pores in the initial cast are large enough. If they are not, the subsequent cast layer is soft. If the slip is deflocculated to minimum viscosity, there is no buildup over time. The casting rate is very slow, and the cast is dense, hard and brittle, with little or no plastic strength.

Under some conditions, a gel curve looks stable or, in some cases, over-flocculated and syneretic at a higher shear rate, but partially deflocculated (continuous buildup) at a lower shear rate. I once spent a number of months working on a particular set of foreign materials intended for sanitary casting slip and kept seeing this type of result. I was never able to achieve a good gel curve under any conditions. The casts were initially fast and firm, but then became slow and very soft. The problem turned out to be excess neutral salts in solution. In another similar situation, one type of slurry could not be deflocculated at all. This condition should be countered by washing the materials to remove the excess neutral salts or by adding salts that retire the excess and decrease the slip conductivity, which will also reduce deflocculant use.

When casting properties are tested on the bench with varying states of deflocculation, the quality criteria usually include cast thickness, drain, plasticity, firmness and cracking. A slip with low clay content, or low plasticity clays and low neutral salt content (conductivity), often deflocculates easily. If the rheology is set to a proper RBU for rapid casting and good drain, the cast might still be more brittle than desired, poor in plasticity and might crack. Increasing the clay content or the quality (MBI/SSA) of the clays can improve plasticity, brittleness and cracking behavior, but it will likely slow the casting rate unless the surface area of the body can be maintained.

The other option is to artificially add water of plasticity by flocculating and then deflocculating. Typically, this is done with gypsum (CaSO₄) salts. The added flocculation increases the requirement for a deflocculant. The additional deflocculant coats the particle surface, and the combination increases the thickness of the bound water layer associated with the surface. The casting viscosity might be the same, but the cast thickness is slightly higher due to the additional thickness of water at the particle surface. Additionally, the cast's moisture content increases, and plasticity is enhanced while brittleness decreases. This is exactly the opposite of the situation described previously, in that neutral salts have been added (in this case Na⁺ from the deflocculant and SO₄^-2 from the flocculant) and conductivity increased.

This whole scenario indicates the importance of measuring specific ion concentrations as one tool for understanding why problems arise with materials. When measurements are available, compensating changes can be made for variations in materials. This is also true for particle size distribution and surface area measurements.

In the foregoing case, additional flocculation and deflocculation improves casting behavior. Normally, we consider that plasticity is decreased by deflocculation, and this is often true when a deflocculant such as NaOH is used. However, it is important to realize that the mechanism of deflocculation can actually improve plasticity if the anion of the deflocculant has an appropriately complex structure that enhances the surface layer, or if more than a minimum of deflocculant is necessary. The flocculation/deflocculation method above increases the necessary amount of appropriate deflocculant as described. The plasticity is improved by the addition of flocculant, deflocculant and water.

The next article in this series will focus on simple and more complex deflocculation systems, with a focus on sodium silicate.

Links

• J.E. Funk Jr. e-mail