Figure 1 / The Effect of Decreasing Extender Particle Size on the Spacing of TiO2 Pigment
Titanium dioxide pigments provide both opacity and whiteness to paint films and influence a wide range of other film properties. The introduction and development of TiO2 throughout the 20th century enabled the production of high-quality and safe paints for both interior and exterior use. However, the manufacturing processes of TiO2 are relatively expensive, and so this pigment is often the most expensive raw material used in emulsion paints. Many techniques and additives have been developed over the years to help improve the utilization of TiO2 and thus reduce the amount used, while maintaining paint performance properties. For over 30 years extender manufacturers have been developing their products and promoting them as potential ways of saving TiO2, often proposing savings as high as 30% by volume.

It is true to say that, over a period of 30 or so years, paint formulators have made use of smaller-particle-size extenders in their paints and significantly reduced their TiO2 content. For example, a typical high-quality matte paint of the 1970s would often contain in excess of 18% by volume TiO2. Through use of small-size calcined clay extenders, and to some degree opaque polymers, this level has now been reduced to typically between 10 and 14% by volume.

It is difficult to see how modern paints, with the already reduced TiO2 volume concentrations (vc), could continue to benefit from extender manufacturers' continuing claims of potential savings of 20 to 30% TiO2 by using their fine-particle-size extenders.

This evaluation, therefore, looks at several current fine-particle-size extenders and some "new" flash calcined extenders to see how the manufacturers' claims hold up when the extenders are used in a relatively modern paint formulation.

Figure 2 / Theoretical Effect of Reducing Extender Particle Size on the TiO2 Nearest Neighbor Distance and Scattering

Improving TiO2 Utilization Through Improved Spacing

In decorative emulsion paints it has long been known that TiO2 pigment is often poorly dispersed and crowded by the extender and emulsion particles.1 Indeed the higher the particle volume concentration the worse the crowding can become. It is well understood that by reducing the size of the extender particles used, the spacing of the TiO2 can be significantly improved (Figure 1).

Certainly, replacing a 6æm extender/filler with a 2æm calcined clay would, and did, lead to improved TiO2 spacing, better opacity and, therefore, the possibility of reducing TiO2 levels. However, to further improve the spacing of the TiO2 to levels that could lead to significant improvements in scattering and opacity would ideally require the particle size of the extender to be an order of magnitude smaller at least. Figure 2 shows the theoretical effect on the TiO2 nearest neighbor distance when extender particle size varies from 0.01 to 1.0µm. This modeling program, developed by Temperley et al2, assumes that there is random packing between TiO2 and extender particles.

Figure 3 / Opacity Curves for Four Extenders Over a Range of PVCs
It can clearly be seen that further spacing of the TiO2 (i.e. increasing values in nearest neighbor distance) only begins to have a significant effect on opacity when the particle size of the extender reaches a size less than 0.5æm. The optimum TiO2 spacing and maximum increase in scatter occurs when the particle size reaches an average of between 0.1 and 0.2µm.

The majority of extenders that still claim to offer potential savings of 20 to 30% through TiO2 spacing are well in excess of this particle size and therefore are unlikely to be able to have such an impact on opacity using this mechanism.

Figure 4 / Opacity Curves for Four Extenders After the Application of a Glycol Film

Extender Evaluation

Today the paint formulator has a bewildering choice of additives from which to choose, and the choice of extenders is no different. For this evaluation a number of leading extender manufacturers were asked to provide fine-particle-size extenders that they would recommend for improving TiO2 utilization. Table 1 lists the extenders used in the evaluation with the chemical type and particle sizes quoted by the manufacturers. Extender A, a well-known and used calcined clay extender, was selected as the standard. It is known to have improved the utilization of TiO2 when substituted for a larger, filler-type product.

To evaluate the extender performance, it was necessary to produce a PVC ladder with constant TiO2 volume concentration and varying extender volume concentration. This evaluation allowed us to determine the critical pigment volume concentration (CPVC) for each type of extender in this paint system. It would not, however, easily allow multiple extender paints to be tested, and we therefore accepted that the evaluation would be limited and that we may lose any synergistic effects (good or bad) that using extenders in combination could give.

Figure 5 / Opacity Curve Showing Crowding Effect
Several paint properties were evaluated including opacity (contrast ratio dry and wet), measured using black and white charts and a number 6 K-Bar wire-wound applicator; color (CIELab); resistance to Gilsonite oil-based stain; and glycol opacity (reflectance after the application of a propylene glycol film). The performance of each extender was plotted against the extender volume concentration and compared to that of the standard Extender A. In this way the extenders were rated for performance as better than, similar to or worse than Extender A.

Figures 3 and 4 give examples of how three of the extenders performed in comparison to Extender A in terms of dry opacity and glycol opacity.

The use of a glycol film is to remove the effects of the increased porosity (dry hiding) and help confirm the position of the CPVC for each extender. If we consider the relationship of the curves and the CPVC levels, shown in Figures 5, 6 and 8, we can draw certain conclusions from their shape and direction. Take for example the curves for Extender A and Extender I in Figure 5. Both curves show a downward trend as the volume of extender increases, which suggests the degree of crowding is also increasing. The gradient of the decrease for Extender I is greater than that for Extender A and so it is crowding the pigment more and thus producing a more rapid fall-off in opacity. Given that the particle size of Extender I is 3.2æm this is as would be expected, since Extender A has a slightly smaller particle size and tighter distribution. By our rating method we would consider Extender I to have a worse performance and not offer any potential savings over Extender A.

Figure 6 / Opacity Curve Showing Crowding Effect of a Homoflocculated Extender on Opacity
Figure 6 compares the performance of Extender A with Extender H. This extender has a much smaller particle size of 0.18æm and, given the modeling data, should produce a much better opacity than Extender A. It can clearly be seen however that this is not the case. In fact, the performance below the CPVC is worse than Extender I. The reason for the unexpected result can be seen in the scanning electron micrograph of the paint, Figure 7. The micrograph clearly shows that the extender is badly flocculated, giving it a much larger effective particle size. It is therefore not surprising that the opacity is down and that this extender is performing worse than Extender A.

Figure 7 / Extender Flocculates
This result also highlights the important fact that the measured particle size of the extender may not truly reflect the particle size in application. It cannot be stressed enough that in order to obtain the best TiO2 utilization both the extender and the TiO2 must be well dispersed and stable.

The third extender is of the "new" flash calcined family. The manufacturing process produces an extender particle containing air voids (both interconnected and discrete), which are claimed to help scatter light, thus improving the opacity of the paint film. The particle size of this extender (Extender F) is very similar to Extender A and so we would not expect to gain any advantage through a reduction in pigment crowding.

Table 1/ Extenders and Their Particle Sizes Used in This Evaluation
Figure 8 shows the results for the comparison of the opacity generated by paints containing Extender A and Extender F. It can be seen that there is an improvement over and above the opacity generated using Extender A. As both extenders are of similar particle size, the additional opacity is most likely produced by the entrapped air, which is offsetting the crowding effect.

Table 2 / Ranking of Extender by Overall Performance

Ranking of Extenders

It is not possible to review all of the results obtained in this presentation. However, each property was carefully considered and compared to the performance of Extender A and the results used to rank the extenders in order of performance in terms of being better than, similar to or worse than Extender A. Table 2 shows the results.

From the results obtained, four extenders appeared to give an improved, and therefore potential TiO2-saving performance - extenders C and D and extenders E and F, the latter two both being flash calcined china clays. Examination of the four curves in Figure 9 shows that Extenders C and D are both gaining most of their advantage by reducing the CPVC of the paint, although small increases in opacity below the CPVC are also evident.

Figure 8 / Opacity Curve Showing Improved Opacity From a Flash Calcined Extender
The brightness of these two particular extenders is very high, and it is possible that their shape and size make them suitable for reflecting rather than scattering light, as is the case with some platelet-type extenders. Extenders E and F, although reducing the CPVC slightly, still show an overall increase in opacity across the PVC ladder. Considering Extender E, it can be seen in Figure 10 that the increase in opacity below the CPVC is quite substantial and should allow for a reasonable reduction in TiO2 loading.

Figure 9 / Extenders Giving Improved Performance

TiO2 Reductions

In order to determine how much TiO2 could potentially be saved and to determine what, if any, detrimental effects this would have on the other paint properties, a series of paints were made with incremental reductions in TiO2. As shown in Figure 10, a PVC of 30% was chosen for the evaluation as this was below the CPVC, thus avoiding complications of dry hiding, and also showed a substantial improvement in opacity over Extender A. Figure 11 gives the opacity results obtained for a given reduction in TiO2vc.

As can clearly be seen, a reduction in TiO2vc of 20% is possible if a match in opacity is required. If, however, we consider the reflectance over black, as in Figure 12, we see that the equivalent reflectance is at a slightly lower saving of between 10 and 15% TiO2vc. This also suggests that some of the opacity advantage exhibited by Extender E (and F) is coming from absorption rather than scattering.

Wet opacity is of less importance than it once was and, as one would expect, a reduction in TiO2 loading results in a decrease in the wet opacity. Indeed, only the smallest level of TiO2 reduction could be made without a significant decrease in wet opacity (Figure 13).

Figure 10 / Extenders Giving Improved Performance
Since the wet opacity is falling quickly, it confirms that the TiO2 is not being spaced sufficiently well and also suggests that the voids within the flash calcined extender are not discrete and only begin to scatter light when the paint dries.

The presence of greater porosity in the paint with Extender E was confirmed by the use of Mercury porosimetry. While not completely ideal for assessing the total porosity of paint films, the method does appear to work well in systems where any voids are likely to be interconnected rather than discrete. Figure 14 shows the plot from Mercury porosimetry for paints with Extender A and Extender E.

Figure 11 / Reducing TiO2 Content Using Extender E
The porosity of the paint made with Extender E is approximately 5% greater than that for Extender A. Also of interest is the fact that most of this additional porosity appears to come in the size range between 0.1 and 1.0æm, the exact size range given for the voids within the flash calcined clays. This range of sizes also covers the ideal range of void sizes for the scattering of light, although much of the additional porosity indicated is above 0.5æm in size.

Given that we had identified the existence of greater porosity in the Extender E paint film we examined the level of scrub resistance to see if there were any adverse effects in reducing the TiO2 level and replacing it with extender. Figure 15 shows the results obtained. In Figure 15 it can be seen that simply exchanging Extender A for Extender E has led to a slight decrease in scrub resistance. On reducing the level of TiO2 and replacing it with Extender E, the scrub resistance continues to fall and fails the requirements of <5mg/cm2 loss when more than 5% TiO2vc is replaced. This again falls well short of the claimed 30% saving in TiO2.

Figure 12 / Reducing TiO2 Content Using Extender E

The Effect of Entrapped Air on TiO2 Utilization

As we have discovered so far, the use of fine-particle-size extenders can help to improve the TiO2 utilization in emulsion paints, although the levels of 20 to 30% claimed by the manufacturers are currently only possible at the expense of other paint properties. We have seen how flash calcined extenders with their air void structures can improve the opacity of emulsion paints through some additional scattering, although we have shown that the majority of the voids are not of a suitable size to preferentially scatter light. This suggests that there is another component in the equation that is adding to the opacity of the paint and is imparted by the extender.

Figure 13 / Reducing TiO2 Content Using Extender E
As stated at the beginning of this paper, it is well known that the TiO2 dispersion in emulsion paint can be very poor. In comparison to a solventborne alkyd paint, the TiO2 pigment in an emulsion paint is flocculated and under-utilized. Yet in many cases, for an equal loading of TiO2 the opacity of the emulsion paint can be equal to, or surpass that, of the alkyd. To explain this apparent contradiction we must look at the differences in refractive index of the two systems.

Alkyd resins have a typical refractive index of around 1.51, while an acrylic emulsion will be slightly lower at approximately 1.48. By modeling the scattering power of TiO2 as a function of resin refractive index (RI) we can show that this small difference of 0.03 units in RI is sufficient to produce an uplift in opacity of 0.5 units contrast ratio. It can also be shown by simple calculation that a level of 5% air, entrapped within the paint film, will approximately account for a similar reduction in refractive index and thus increase opacity.

Figure 14 / Porosity Measurement
From earlier Mercury porosimetry work, it is believed that a typical emulsion paint contains between 5 and 15% of entrapped air, mainly due to incomplete film coalescence. This quantity of air would account for an additional 1.5 units of contrast ratio and would be sufficient to offset the opacity loss due to flocculation when compared to the solventborne alkyd paint. Table 3 gives an indication of the relative increases in opacity that can be achieved by reducing the refractive index of the resin system used.

If, on the addition of certain extenders, more air is entrapped within the paint, then this would have the effect of reducing the perceived refractive index of the paint medium and thus increase the opacity of the paint. It is this process that we believe is contributing to the opacity advantage seen when using Extender E.

Further work is now being undertaken at Huntsman Tioxide to better determine the levels of air entrapped in emulsion paints, the "particle size" of the air voids, how these interact with the paint film and how, if at all, it is possible to control the inclusion of air in the paint.

Figure 15 / Reducing TiO2 Content Using Extender E


Having evaluated a range of fine-particle-size extenders we have concluded that some small improvements in TiO2 utilization are still possible using these products, although the claims of 20 to 30% saving in TiO2 could not be fully substantiated when all of the major paint properties were considered. Current extenders with a particle size in excess of 0.5µm have been shown, both theoretically and through experimentation, not to improve the spacing of TiO2 and thus will not improve the TiO2 utilization by this method. Those extenders with particle sizes of less the 0.5æm were not easily dispersed down to their primary size and tended to homo-flocculate, creating large, multi-particle extenders causing increased crowding and lower opacities.

The use of flash calcined china clays did appear to offer better TiO2 utilization, but again when other paint properties were included the benefits achieved were markedly reduced. The incorporation of air voids within the paint film by these extenders was shown to have a positive effect on opacity, although the size of the voids suggest that light scattering may not be the principle method by which the improvement in opacity is achieved. The inclusion of air voids is thought to reduce the apparent refractive index of the medium and this reduction is contributing to the increase in scattering of the TiO2 pigment.

Table 3 / Theoretical Effect of Entrapped Air on Refractive Index and Opacity


The author would like to express his thanks to Dr. Les Simpson, John Temperley, Kathryn Fullerton and Frank Holden for their help in preparing this paper and carrying out the necessary laboratory work. I would also like to thank Mahomed Maiter, Group Marketing Director and Brian Thomas, Group HR & Communications Director, for their support in publishing this paper.

For more information, contact Paul Dietz, Huntsman Tioxide, Haverton Hill Road, TS23 1 PS, Great Britain; phone 0044 1642 376467; fax 0044 1642 676777; e-mail


1 Cutrone, L; and Becherel, D. Interaction Between Fine Particle Extenders and Titanium Dioxide in Paints; Technical Report D9202GC; Huntsman Tioxide.
2 Temperley, J; Westwood, MJ; Hornby, MR; Simpson, LA. Use of a Mathematical Model to Predict the Effects of Extenders on Pigment Dispersion in Paint Film. J. Ctgs. Tech., 1992, Vol. 64 No.809, 33-40.

This paper was presented at the 7th Nurnberg Congress, European Coatings Show, April 2003, Nurnberg, Germany.