INFLUENCE OF PACKAGING ON COSMETIC EMULSION DURING STORAGE 253 highest changes in color were observed for samples stored for 4 mo in G and SAN jars, followed by Ac/PP. It is noteworthy that the changes in FT-NIR spectra were especially visible only for samples stored in G jars. Taking into consideration all experimental data from FT-NIR spectroscopy and color measurements, it can be concluded that the cream samples changed depending on the packaging type and storage conditions. It is known that the permeation of light and gases through polymer packaging is particularly diverse and sometimes very high (21), therefore not all kinds of polymer packaging can be used for storage of food products and cosmetics. It is also known that polymers, as organic compounds, are characterized by a high coeffi cient of linear thermal expansion in contrast to many other materials such as glass. Under the infl uence of temperature, the density of polymers and, as a consequence, other properties associated with it can be changed. Therefore, the effect of temperature must be taken into account when polymer properties are examined. SAN copolymer is a material with high transparency for light, but with good chemical and thermal resis- tance. Aluminum is the metal used for vacuum metalizing of packaging materials. Orig- inally, this process was used for decorative purposes. Currently, metallization is a widely Table II The L*, a*, b*, and ΔE* Values of Cream Stored at Refrigerated Temperature Month L* a* b* ΔE* F 0 80.50 ± 0.11 –2.73 ± 0.01 2.16 ± 0.04 0 Al/PP 2 80.48 ± 0.12 –2.74 ± 0.01 2.10 ± 0.16 0.16 ± 0.12 4 80.80 ± 0.02a –2.83 ± 0.02a 2.18 ± 0.04 0.31 ± 0.02 Ac/PP 2 80.43 ± 0.12 –2.72 ± 0.01 2.18 ± 0.05 0.13 ± 0.04 4 80.75 ± 0.11a –2.83 ± 0.04a 2.30 ± 0.02a 0.31 ± 0.09 G 2 80.50 ± 0.03 –2.69 ± 0.01a 2.16 ± 0.01 0.05 ± 0.01 4 80.82 ± 0.03a –2.81 ± 0.02a 1.87 ± 0.01a 0.44 ± 0.02 PP 2 80.58 ± 0.08 –2.70 ± 0.03 2.14 ± 0.10 0.15 ± 0.04 4 81.15 ± 0.05a –2.87 ± 0.02a 2.27 ± 0.02a 0.68 ± 0.04 SAN 2 80.51 ± 0.07 –2.72 ± 0.03 2.14 ± 0.06 0.09 ± 0.01 4 80.98 ± 0.03a –2.88 ± 0.02a 2.15 ± 0.03 0.50 ± 0.03 a Means in the column are signifi cantly different from that of fresh sample (LSD test, α = 0.05). Table III The L*, a*, b*, and ΔE*ab Values of Cream Stored at Room Temperature Month L* a* b* ΔE*ab F 0 80.50 ± 0.11 –2.73 ± 0.01 2.16 ± 0.04 0 Al/PP 2 79.66 ± 0.06a –2.44 ± 0.02a 2.35 ± 0.13a 0.92 ± 0.03 4 79.43 ± 0.05a –2.35 ± 0.03a 2.44 ± 0.08a 1.17 ± 0.03 Ac/PP 2 79.41 ± 0.23a –2.29 ± 0.01a 2.28 ± 0.10 1.19 ± 0.21 4 79.34 ± 0.23a –2.09 ± 0.03a 2.53 ± 0.03a 1.38 ± 0.20 G 2 79.66 ± 0.09a –2.30 ± 0.01a 2.32 ± 0.05a 0.96 ± 0.07 4 79.27 ± 0.21a –1.98 ± 0.02a 2.43 ± 0.02a 1.46 ± 0.04 PP 2 79.72 ± 0.07a –2.44 ± 0.02a 2.37 ± 0.04a 0.86 ± 0.07 4 79.70 ± 0.05a –2.27 ± 0.02a 2.60 ± 0.02a 1.02 ± 0.03 SAN 2 79.56 ± 0.15a –2.30 ± 0.02a 2.11 ± 0.07 1.03 ± 0.14 4 79.20 ± 0.08a –2.12 ± 0.02a 2.19 ± 0.04 1.44 ± 0.07 a Means in the column are signifi cantly different from that of fresh sample (LSD test, α = 0.05).

JOURNAL OF COSMETIC SCIENCE 254 used method to modify packaging to improve its barrier properties against moisture and gases. As shown in Figure 3B, the samples stored at room temperature in Al/PP were the closest to fresh samples (F). It is probably because of the better barrier properties of this packaging against moisture and gases (at room temperature for at least 4 mo) than other packagings used in the present study. The barrier properties of all packagings during 2-mo storage of cream samples in refrigerated conditions seem to be similar. Regression analysis using PLS1 was performed to check the possible relationship between FT-NIR spectra (12,500–4,000 cm-1) and color parameters. The PLS1 regression relation- ships with no FT-NIR data processing, FT-NIR data after MSC, and the fi rst derivative of FT-NIR data are presented in Table IV. All presented relationships, with an excep- tion of relationship for b*, had acceptable R2 for calibration (at least 0.85) and validation (higher than 0.84). The best relationships, expressed as R2 for validation, were found between L* or a* and FT-NIR data used with no spectral data processing (for L* value) or after MSC modifi cation (for a* value). The fi rst derivative of FT-NIR data did not improve the R2 value. CONCLUSIONS Altogether, the results of the present study showed that diffuse refl ectance FT-NIR spectroscopy and multivariate calculating methods, such as PCA, could be successfully used to determine the effect of packaging materials on overall changes in cosmetic creams during storage. The changes in FT-NIR spectra were probably associated with the changes in the water level in tested samples and they were correlated with their lightness L*, a* parameter, and total color difference ΔE*ab. This study also demon- strated that FT-NIR spectroscopy is a rapid and powerful technique that could be used for quality control of cream emulsions without any sample pretreatment, thus meeting the requirements of “green chemistry.” It could be useful for producers to make the best choice of packaging, protecting the original quality of cosmetic emulsion during storage. Table IV Coeffi cients of Determination (R2) for the PLS1 Regression between FT-NIR Spectra and Colour Parameters of Cosmetic Cream Samples Stored in the Cold and at Room Temperature Parameter Mean ± SD Method Factors R2cala R2vala L* n = 105b 80.12 ± 0.64 No data processing 4 0.919 0.911 MSC 5 0.923 0.911 First derivative 5 0.949 0.910 a* n = 105 2.53 ± 0.28 No data processing 4 0.926 0.918 MSC 7 0.975 0.931 First derivative 5 0.957 0.924 b* n = 105 2.5 ± 0.18 No data processing 7 0.724 0.498 MSC 6 0.684 0.506 First derivative 4 0.653 0.461 ΔE*ab n = 105 0.69 ± 0.49 No data processing 3 0.851 0.838 MSC 6 0.890 0.845 First derivative 3 0.871 0.835 a Coeffi cient of determination for calibration (cal) and for validation (val). b Number of measurements.