FACTORS AFFECTING SPF AND PA VALUES OF SUNSCREEN CREAM FORMULATIONS 151 counter-brand sunscreen products in Figure 1, which were found to have a good correla- tion with a determination of coeffi cient (r2) of 0.901. Determination of UVA-PF according to the method of Ferrero et al. (8) In vitro UVA-PF was calculated with spectral range of 320–400 nm as in equation (2): λ λ % %λ œ œ − ⋅ 400 320 400 m 320 1 UVA PF Invitro T T (2), where Tm is the arithmetic mean of the transmittance data in the UVA range. UVA-PF was classifi ed according to the Japan Cosmetic Industry Association into four categories as follows: UVA-PF 2: no protection against UVA 2–4 (PA+): protection against UVA 4–8 (PA++): considerable protection against UVA and ≥8 (PA+++): the greatest protec- tion against UVA (2). Instrumental texture analysis of sunscreen formulations. Each sample was packed with a height of 7 cm in a 28-ml McCartney bottles (S Murray & Co., Surrey, UK) and kept at 25 ± 1°C for 48 h before performing texture profi le analysis. The texture profi les of each formulation were determined according to Jones et al. (17) with some modifi ca- tion using a texture analyzer (Model TA.XT Plus, Stable Micro Systems, Surrey, UK). A stainless steel probe of 1 cm in diameter (P/0.5R) was compressed twice into the sample at a defi ned rate of 6 mm⋅s-1 to a depth of 1.5 cm, with a delay period of 15 s between the two compressions. Data collection and calculation were performed using the XTRA Dimension software package of the instrument. The texture profi le analysis values of hardness, compressibility, adhesiveness, and cohesiveness of sunscreen sam- ples were calculated by determining the load and displacement at predetermined points on the texture profi le analysis curve (Figure 2). Hardness was the maximum force required to attain a given deformation of the sample during the fi rst compression. Compressibility was the work required to deform the sample during the fi rst compres- sion (A1). Adhesiveness (A3) was the work required to overcome the attractive force Figure 1. Correlation between labeled SPF and calculated SPF of tested counter-brand sunscreen products.

JOURNAL OF COSMETIC SCIENCE 152 between the surface of the sample and the surface of the probe. Cohesiveness (A2/A1) was the ratio of the area under curve for the second compression (A2) to that under curve for the fi rst compression (A1). Four replicate analyses were performed for each formulation at 25 ± 1°C. Physical stability study. Each sample underwent six cycles of freeze-thaw cycling. In each cycle, the sample was kept at 4 ± 1°C for 24 h following by keeping at 45 ± 1°C for 24 h. Before and after undergoing the freeze-thaw cycling, texture profi les, viscosity, and pH of each formulation using a fl at electrode (Metler Toledo in Lab surface, Greifensce, Switzerland) were examined. The viscosity of each sample was measured at 25 ± 1°C using a digital Brookfi eld Viscometer (Model DV-III+ Programmable Rheometer, Stoughton, MA) mounted on the helipath stand fi tted with a T-F spindle. The rheometer was set at 5 rpm. The apparent viscosity, expressed as centripoints (cPs), was an average of the data points collected for 30 s. Statistical analysis. Unless otherwise stated, all experiments were performed in triplicate. The effects of PVR, xanthan gum, and stearic acid on SPF and PA values as well as texture profi les among formulations were evaluated using regression analysis. The differences between before and after freeze-thaw cycling of each formulation were analyzed using a paired-samples t-test. p 0.05 denoted statistical signifi cance. All statistical analyses were performed using SPSS software. Mathematical models for 23 full factorial design were obtained, relating the responses with the experimental conditions as follows: 0 1 2 3 12 13 23 y b b A b B b C b AB b AC b BC (3), where the coeffi cients estimations were b1, b2, and b3 corresponding to the response ob- tained with the factors A, B, and C, respectively. The interaction coeffi cients were b12, b13, and b23 corresponding to the response obtained with interaction terms of AB, AC, and BC respectively. Figure 2. Texture profi le curve of a sample.