A cosmetic product is a complex, nonequilibrium mixture of ingredients including polymers, small molecules, surface-active species, extracts and particulates, all distributed inhomogeneously throughout the bulk. In order to create the functionality expected in a given cosmetic product, several processes must be addressed and controlled; these include: the deposition of ingredients from the bulk to a surface, the timely and targeted delivery of ingredients while maintaining the appropriate physical form, the protection of an ingredient from the surrounding system until its functionality is activated, and the self-assembly or segregation of parts of a system in order to perform a specific operation. To monitor or anticipate what can happen when components are combined and as time passes is no trivial task especially when compounded with the psychophysical needs for the product to fit a specific concept and to spread, feel, pour and appear elegant.
Classical analytical methods will give the quantitative values of composition and concentration but at the expense of the spatial information. For example, in an emulsion stabilized by a lamellar structure, how the molecular species relate to each other is lost when the classical chemical analysis is performed.
Classical measurement and imaging, despite a plethora of surface sensitive techniques—e.g., electron microscopy and secondary ion emission, together with a host of different particle-surface bombardment techniques—have until recently seen a paucity of suitable noninvasive spatial methods that can examine materials in their natural, undisturbed state.
Modern image analysis and processing, coupled with fiberoptics and imaging detectors such as slow scan charge coupled devices (CCD cameras) or spectral cameras, and combined with modern spectrometers or the scanning tunnelling devices, are in the vanguard to rectify this. This is particularly true in the UV-visible (250 to 900 µ) and near-infrared (0.7 to 2.5 µ) regions.
Once captured in a computer-compatible form, the digitized image can be mathematically treated and significant features, such as boundaries, can be enhanced while other features of interest including fluctuations in concentration, color and shape can be extracted and quantified from the total image using various mathematical inversion techniques such as Fourier methods. Coupled to this, Monte Carlo methods can be used to feed basic physical equations for processes into a computer that then takes a stochastic pathway, i.e., a nondeterministic approach.