J. Soc. Cosmetic Chemists, 18,367-384 (May 27, 1967) Tristimulus Color Measurements in Fading Studies and Other Color Changes of Pharmaceutical Dosage Forms A.M. RAFF, Ph.D.* Preser•ted November 30, 1966, New York City Synopsis--The concept of color as a psychological-physical phenomenon is reviewed, and examples are given illustrating this viewpoint. The need for and subsequent development of tristimulus color measurements are briefly described. A procedure for following the degrada- tion of colors is discussed, mad then methods of measuring the degree of color change are evaluated. These methods of measuring color change include R (reflectance), K/S (Kubelka- Munk function), and AE (change in color). INTRODUCTION Everyone, including those who may have some form of color blindness, will see some colors when viewing a "colored" object. Not everyone will see the same colors, but all will see colors. In Fig. 1 there is no color--just a curved line. It does not convey as much information as the object which generated the curve. Even with a number of transformations it can, at best, only start to convey the in- formation that is desired, namely, the color that the viewer will see when looking at the object which generated this curve. It should be noted at the onset that the subject of this paper is color, not dye or colorant concentration in a dosage form. * Pharmaceutical Development Dept., Sxnith Kline & French Laboratories, Philadelphia, Pa. 19101. 367

368 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS IOO 6o 4o i i i i i i i , i i i i 460 540 6ZO 700 WAVELENGTH (MILLIMICRONS) Figure 1. Reflectance measurement of a colored object Nickerson (1) said: "It is true that a spectral distribution curve may serve to define unambiguously the stimulus for the color of a light source, but the spectral-reflectance or spectral-transmittance curve of an object is not sufficient to define even the stimulus for its color. Its spectral properties must be combined with those of the light source under which it is to be viewed before even the stimulus for the object's color is de- fined. To define its color the concept of an observer must also be intro- duced." Essentially what Nickerson is saying is that to see an object as colored there are three requirements: the object, the observer, and an illuminant. If any one of these three is absent then there is no colored object in the true meaning of the word. This is more than a semantic nicety. USE OF SPECTROPHOTOMETRIC RESULTS A spectrophotometric curve alone will not serve to define a color. In order to describe color with any degree of adequacy it is therefore necessary to have some form of color space defined. Color space de-