JOURNAL OF COSMETIC SCIENCE 72 indicate that all the rheological parameters of synthetic polymers have much higher val- ues than cellulose derivatives, except the damping factor. The difference in rheological proper- ties defi nitely has an impact on the skin sensory attributes of the emulsions. For example, synthetic polymer-based emulsions have a richer skin feel than their cellulose-derived counter parts, whereas the latter are easier to spread into skin. Lower viscosity and yield stress indicate that the emulsion system has less resistance to fl ow, once the yield stress has been exceeded. SENSORY ANALYSIS Ten college students (age 19–24) of both sexes were recruited from Beijing Technology and Business University to participate in an evaluation of the sensory attributes of seven emul- sions. The selection of the panellists was based on their interest, willingness, background, experience in using cosmetic products, and their ability to describe and rate the selected sensorial attributes. The panellists were systematically trained on the sensory evaluation protocols. For each sensory attribute, four to fi ve commercial products were used as stan- dard reference points to defi ne the scales (0–10) used during the training. The panel’s sensory evaluation results for the seven emulsions are shown in Table V. For each sensory attribute, a one-way ANOVA of the Kruskal–Wallist tests was conducted to determine the overall signifi cant difference among the seven emulsions. The results are listed in the far-right column of Table V. The Tukey’s HSD tests were used to compare the mean be- tween each pair of the seven emulsions the results are shown as letters A–C. The same letter in one row indicates that the corresponding emulsions are not signifi cantly differ- ent for the attribute considered (p  0.05). The results showed that thickening agents play an important role in the sensory proper- ties of the fi nal formulas. All eight attributes for the factors of appearance, pick-up, and rub-in exhibited statistically signifi cant differences (p  0.05), whereas all fi ve attributes of the after-feel factors showed no statistically signifi cant differences ( p 0.05). As shown Table V Sensory Evaluation Results of Seven Emulsionsa Factors Sensory attributes Control HEC HMHEC HPMC PVP PAA Na PAA Signifi cance Appearance Gloss 6.8 ab 7.0 a 7.1 a 6.7 ab 4.7 c 5.9 b 6.4 ab 0.0001 Pick-up Ease of pick-up 4.3 b 4.5 b 4.3 b 3.8 b 6.2 a 5.8 a 4.3 b 0.0001 Peak after pick-up 2.6 c 3.6 b 3.4 bc 3.1 bc 5.5 a 5.1 a 4.6 a 0.0001 Firmness 2.0 c 3.3 b 3.1 bc 2.6 bc 6.6 a 6.1 a 5.6 a 0.0001 Rub-in Spreadability 6.7 a 5.5 ab 5.8 ab 5.8 ab 4.9 b 4.8 b 5.0 b 0.02 Hydration feel 6.9 a 5.4 ab 5.5 ab 6.3 ab 4.8 b 4.9 b 5.5 ab 0.01 Oil feel 4.2 ab 4.4 ab 4.1 b 4.3 ab 6.0 a 5.3 ab 4.4 ab 0.02 Absorbency 5.9 a 6.1 a 7.2 a 6.6 a 7.5 a 7.8 a 7.4 a 0.03 After-feel Tackiness 2.9 a 3.3 a 3.2 a 3.3 a 4.7 a 3.7 a 3.8 a 0.26 Gloss (after-feel) 3.4 a 3.5 a 3.5 a 3.5 a 4.8 a 4.3 a 3.8 a 0.32 Slipperiness 5.4 a 6.2 a 6.1 a 6.1 a 6.2 a 6.1 a 5.7 a 0.89 Greasiness 3.3 a 2.8 a 3.3 a 3.5 a 3.8 a 4.0 a 3.4 a 0.62 Moisture 4.5 a 4.6 a 4.8 a 4.4 a 5.1 a 4.6 a 4.7 a 0.26 a The same letter (a, b, c) in one row indicates that the corresponding emulsions are not signifi cantly different for the attribute considered (p  0.05).

RHEOLOGICAL AND SKIN SENSORY PROPERTIES OF COSMETIC EMULSIONS 73 in Table V, fi rmness was infl uenced most by thickening agents for all sensory attributes. The samples with cellulose derivatives had less fi rmness than the samples with synthetic polymers. This fi nding correlates well with the rheology results. Ease of pick-up and peak after pick-up were also highly infl uenced by the nature of the polymers, as the formulas with cellulose derivatives received lower scores due to lower viscosity. These samples fl ow easily, so a smaller quantity of these products could be picked up by fingers, and the peaks after pick-up were lower. On the other hand, the formulas with cellulose derivatives scored higher than synthetic polymers for spreadability because of the lower viscosity and yield stress. It is well known that PAA-type thickening agents lead to a breakdown in contact with existing electro- lytes on skin, which reduces the spreading force of a formulation. However, this break- down is not as signifi cant as the effect of their high viscosity and yield stress, which aligns with our spreadability prediction. The formulas with cellulose derivatives showed higher original gloss than the synthetic polymers (p  0.05), which is also reported by Savary (7). Notably, the scores for gloss (after-feel) on skin tended to be higher for the synthetic polymers than for the cellulose derivatives, although no statistically signifi cant difference was found. Compared with synthetic polymers, cellulose derivatives tended to have more hydration, less oil feel, and slower absorbency. This could be explained by the nature of cellulose, which has more hydrophilic groups (hydroxyl groups) in its structure. These hydroxyl groups help the formula to retain water and give a greater hydration feel during the rub- in phase. However, this also leads to a lower absorbency of the formula because the absor- bency is closely related to the moisture evaporation during the rub-in phase. In addition, of the synthetic polymers tested in this study, PAA has a relatively better gloss and hydration feel, which may be because of its chemical structure and process technology (Table V). There were no statistically signifi cant differences (p 0.05) among the seven emulsions for the after-feel factor (tackiness, gloss, slipperiness, greasiness, and moisture). Similar results were reported by Wang et al. (5), who found that the effi cacy of moisturizing was not necessarily linked to the hydrophilic polymers used in the formula. PCA ANALYSIS To illustrate the relationships among the eight rheological and 13 sensorial parameters, as well as the differences among the seven emulsions, PCA was conducted based on all the variables in Tables IV and V. As shown in Figure 2B, the fi rst two principal components account for 86.3% of the total variance (component 1 = 75% and component 2 = 11.3%). Except for the damping factor, the positions of all the rheological parameters are close, which means they are related to one another. Regarding the sensorial attributes, absor- bency, fi rmness, and peak after pick-up are closely related to the rheological parameters in the positive side of the Component 1 axis (PC1), whereas spreadability is closely re- lated to the rheological parameters in the negative side of PC1. Based on the PCA results, the seven emulsions studied can be divided into two groups—cellulose-derivative & con- trol and synthetic polymers(Figure 2A). The cellulose-derivative & control-based emul- sions are on the negative axis of Component 1, and tend to have lower fi rmness, G′, G″, yield stress and viscosity, more original gloss before use and less gloss on the skin after use, easier spread as lotions, more hydration feel and less oil feel during rub-in, and lower