2006 ANNUAL SCIENTIFIC MEETING 189 3.5 Deodorant Efficacy Testing Controls Since they are the primary measurement tool, the process for selecting, training, monitoring and maintaining the odor judges is essential to a successful program. Equally important is to provide a quiet, odor free evaluation area and controlling evaluation intervals to limit judge fatigue. Another important control is the subject selection process and monitoring subject restrictions during testing. These aspects will be discussed along with other factors that can impact judge performance. 4. CONCLUSION Understanding applicable historical infonnation and having a clear definition of the project objectives are important when designing a deodorant or antiperspirant efficacy study. Knowledge of which basic procedures must be followed, how they can be modified, and applying the appropriate statistical techniques is critical to the design of a successful study. Bibliography 1 Food and Drug Administration, "Antiperspirant drug products for over-the-counter human use (Final Monograph)," Federal Register, 68 FR 34273 (June 9, 2003). 2 Bowman, J.P., Oddo, L.P., Bates, T., Wild, J.E., "New Screening Technique for Antiperspirant Testing Utilizing Sweat Collections and Instrumental Evaluations", poster session presented at the Cosmetic Toiletries and Fragrance Association, 1998. 3Majors, P., Carabello, F, "Hill Top Research Method of Antiperspirant Evaluations and General Discussion of Results", presented to the OTC Panel for Antiperspirants, August 14, 1975. 4 ASTM Standard E 1207-02, Standard Practice for the Sensory Evaluation of Axillary Deodorancy, Nov. 2002.

190 JOURNAL OF COSMETIC SCIENCE KEYNOTE AWARD LECTURE SPONSORED By RUGER CHEMICAL CORPORATION INTERPARTICLE INTERACTIONS IN COLOR COSMETICS Lorna M. Kessell, Ph.D. Uniqema R&D, United Kingdom Pigments are the primary ingredient of any color cosmetic and the way in which these particulate materials are distributed within the product will determine many aspects of product quality including functional activity (color, opacity, UV protection), stability, rheology and skin feel. The particulate distribution depends on many factors such as particle size/shape/surface characteristics, processing and other formulation ingredients but ultimately is determined by the interparticle interactions. The fundamental principles of preparation of pigment dispersions are briefly described. These include wetting, dispersion (or wet milling) and stabilisation. This is followed by a section on stability for both aqueous and non­ aqueous media. Finally the use of rheometry in assessing particulate dispersions will be described. Introduction Simple mixing of inorganic powder (eg titanium dioxide or zinc oxide) can produce a fluid dispersion even at 50% solids. However this is not necessarily an indication of a 'well dispersed' material and indeed a particle size analysis (and for UV attenuators spectral analysis) demonstrates that particle dispersion is not optimised. The addition of mechanical energy (through milling) can break the aggregates (present in all powders) into their discrete nano-particles but the level of added dispersant/stabiliser can be critical in achieving a usable and stable dispersion. A dispersion could be highly stable but be too thick to stir or so thin it sediments. Rheometry can be a very useful tool in determining stability and on the other hand the rheological characteristics can be modified to give the required aesthetics by a suitable choice of stabilising agents. Stabilisation All particles experience attractive forces on close approach. The strength of this van der Waals attraction VA (h) depends upon the distance h between particles of radius R and is characterised by the Hamaker constant, A. The strength of this attraction(A) depends upon the dielectric properties of the material and for some materials such as TiO2 this is exceptionally high so that (in non-aqueous media at least) despite their small size a dispersant is always needed to achieve colloidal stabilisation. a. Steric stabilisation This is usually obtained using adsorbed layers of polymers or surfactants. The most effective molecules are the A-8 or ABA block or BA0 graft polymeric surfactants4 where 8 refers to the anchor chain. For a hydrophilic particle this may be a carboxylic acid, an amine or phosphate group or other larger hydrogen bonding type block such as polyethylene oxide. The A chains are referred to as the stabilising chains which should be highly soluble in the medium and strongly solvated by its molecules. For non-aqueous dispersions the A chains could be polypropylene oxide, a long chain alkane or oil soluble polyester. When two particles with an adsorbed layer of hydrodynamic thickness � approach to a separation distance h that is smaller than 26, repulsion occurs as a result of two main effects: (i) unfavourable mixing of the A chains when these are in good solvent condition (ii) reduction in configurational entropy on significant overlap. b. Electrostatic stabilisation For some powders we have a second option. Inorganic oxides have ionisable groups on their surface which means that in aqueous media they can therefore develop a surface charge depending upon pH, which affords an electrostatic stabilisation to the dispersion. This stabilisation is due to the interaction between the electric double layers surrounding the particles (2). This double layer repulsion depends upon the pH and electrolyte concentration. When the pH of the disfersion is well above or below the isoelectric point or the electrolyte concentration is less than 10· 2 mol dm· 1 :1 electrolyte, repulsion is often sufficient to produce a dispersion without the need for added dispersant However, in practice this condition cannot be reached since at high solids content the ionic concentration from the counter and co-ions of the double layer is high and the surface charge is not uniform. Therefore a dispersant such as sodium polyacrylate is required to achieve this high solids content. This produces a more uniform charge on the surface and some steric repulsion due to the high molecular weight of the dispersant. Under these conditions the dispersion becomes stable over a wide range of pH at moderate electrolyte concentration. This is electrosteric stabilisation. A schematic representation of the energy -distance curves can help us visualise the stabilisation and how to improve it. A positive energy barrier on close approach is required to prevent aggregation. As the distance between the particles, h, decreases this is achieved in all three cases. A small G 0 (attraction) may remain in some cases and can be beneficial as will be described later.