ALTERNATIVES TO COSMETICS PRESERVATION 121 to the solubility problems of glyceryl caprylate. Furthermore, the enrichment of system V with (0.1% w/w) anisic acid (system VI) or (0.3% w/w) levulinic acid (system VII) led to the separation of the phases in the cases of the cleansing milk and the anticellulite cream after 20 days. In contrast, addition of (0.1% w/w) levulinic acid (system VIII) did not infl uence the stability of the cleansing milk. WATER ACTIVITY Water activity (aw) or equilibrium relative humidity quantifi es the active part of the moisture content or “free water” as opposed to the total moisture content, which also includes “bound water.” It indicates the amount of water in the total water content that is available to microorganisms. Each species of microorganism has its own minimum aw value below which growth is no longer possible (6). The results of water activity measure- ments of the tested formulations are presented in Table IV. DISCUSSION All the tested antimicrobial systems (I–VIII) have exerted excellent activity against Gram-positive and Gram-negative bacteria in the acidic (pH=5.5) environment used. They protected effi ciently the emulsifi ed and aqueous formulations against Gram-positive and Gram-negative bacteria in challenge tests (criterion A of E. Ph.) and in in-use study (intact products and following three weeks of use). Antimicrobials I–VIII proved to be effective against Gram-positive bacteria, although 0.86 is the lowest aw value permitting S. aureus growth. Acidic pH conditions may contribute to the increase in the minimum aw value for this microorganism (6) and therefore improve the performance of alternative systems used. Of course, manipulation of aw is only part of the preservative system. The activity against Gram-negative bacteria could be partially attributed to the relatively low water activity values of the products (0.865–0.932) (Table VI), since water activity values lower than 0.95 prevent the growth of Gram-negative microorganisms (6). We note that these microorganisms are known to be very persistent and often are recovered in the in-use state, probably from the hands of the consumers (27), even in products contain- ing effective traditional preservatives such as parabens and phenoxyethanol (28). Figure 2. Results of the challenge test regarding the activity of preservative systems I–IV against A. niger in the case of tonic lotion.

JOURNAL OF COSMETIC SCIENCE 122 On the other hand, fungi were less susceptible to Lonicera extracts (system I) and Lonicera extracts/glyceryl caprylate (system V). Although, system I in the case of the tonic lotion showed moderate activity against A. niger and C. albicans in challenge tests (criterion B of E. Ph.), signifi cant levels of mold were recovered following use. The moderate effi cacy of system I in the preservation of the tonic lotion could be ascribed to the inability of Lonic- era extracts (0.2% w/w) to inhibit mold in this formulation. System V protected the aqueous shampoo and shower gel against A. niger, where criterion A of E. Ph. was ful- fi lled. Probably the antimicrobial potencies of Rosmarinus offi cinalis leaf water in the shampoo and Lavender angustifolia leaf water in the shower gel enhanced the antifungal activity of system V. System V satisfi ed marginally criterion B of E. Ph against A. niger in all the emulsifi ed formulations except the conditioning cream. The greater ability of system V in this cosmetic form could be ascribed to cationic agents, which might rein- force the antimicrobial activity (28,29). Furthermore, system V was unable to preserve the anticellulite cream, cleansing milk, and peeling cream, since A. niger was detected after three weeks of consumer use. Another factor that might enhance contamination risk in the case of the peeling cream is the jar-container, which allows the entry of microorganisms into the product. The lack of effi cacy in system V against A. niger in some emulsifi ed formulations is in accordance with the fi ndings reported previously that the preservative performance of glyceryl capry- late against molds depends on the formulation (18,19). Regarding the fragrance ingredients, although the addition of (0.1% w/w) anisic acid (systems II and VI) or (0.3% w/w) levulinic acid (systems III and VII) signifi cantly improved antifungal activity, in some cases it caused stability problems. The reduction of the con- centration of levulinic acid to 0.1% w/w resulted in microbiologically and physicochem- ically stable products. CONCLUSIONS The results demonstrate that natural origin ingredients such as Lonicera extracts seem to be promising as antimicrobial substances for producing self-preserving cosmetic prod- ucts. The addition of multifunctional ingredients such as glyceryl caprylate, levulinic acid or p-anisic acid and/or ethanol was benefi cial in the majority of the products. Ethanol at low concentration (i.e., 5% w/w) may contribute to the performance of the antimicro- bial. An interesting observation is that the products that fulfi lled criterion B in the chal- lenge tests proved to be inadequately preserved after of three weeks of consumer use. We note that this is not surprising since criterion B is more lenient than criterion A. We suggest that E. Ph. should be changed to recognize only criterion A for adequately preserved products in multiple-use containers. Furthermore, challenge tests should be performed not only during the preparation of cosmetic products, but should also be used to evaluate the protection effi cacy of the preservative systems following periods of use. REFERENCES (1) H. Epstein, Cosmetics preservation: Sense and nonsense, Clin. Dermatol., 24, 551–552 (2006). (2) J. Y. Legendre, I. Schnitzler, Q. Y. Li, C. Hausen, M. Huart, G. S. Luengo, M. L. Abella, and M. Roreger, Formulation, characterization, and effi cacy of an adenosine-containing dissolvable fi lm for a localized anti-wrinkle effect, J. Cosmet. Sci., 58, 147–155 (2007).