International Journal of Cosmetic Science, 2015, 37, 31–40
doi: 10.1111/ics.12164
Optimization of cosmetic preservation: water activity reduction A. Kerdudo*,†, F. Fontaine-Vive†, A. Dingas*, C. Faure‡ and X. Fernandez† *SO.F.I.A. Cosmetiques, 1ere Avenue, 1ere Rue, 06514 Carros, France, †Universite de Nice Sophia Antipolis, ICN, UMR 7272, Parc Valrose, 06108 Nice CEDEX 2, France and ‡Laboratoire de Chimie et Biologie des Membranes et Nano-objets, Univ. Bordeaux, CBMN, UMR 5248, Allee Geoffroy St Hilaire, F-33600 Pessac, France
Received 26 May 2014, Accepted 20 September 2014
Keywords: formulation, glycol, microbiology, polymer, preservation, water activity
Synopsis OBJECTIVE: Preservation of cosmetics is a prerequisite for industrialization, and among the proposed solutions, self-preserved cosmetics are of great interest. One key influencing parameter in selfpreservation is water activity; its reduction can help to fight against microbial growth in cosmetic products. This work presents a study on the influence of humectants on water activity and its consequence on the preservation of cosmetic formulations. METHODS: First, water–humectants mixtures were considered. The influence of glycol and glycerin content, glycol chemical structure, glycerin purity and formulation process on the water activity of the binary mixture was studied. Molecular modelling was performed for a better understanding of the impact of glycol chemistry. Then, the results were applied to five different cosmetic formulations to get optimized products. Challenge test on five strains was carried out in that sense. RESULTS: We showed that the higher the humectants concentration, the lower the water activity. Glycol chemical structure also influenced water activity: propan-1,2-diol was more efficient than propan-1,3-diol, certainly because of a better stabilization in water of propan-1,2-diol as shown by DFT calculation. A drop by drop introduction of glycol in water favoured aw reduction. The best water activity loss was 6.6% and was reached on the cream formulation whose preservation was improved as evidenced by challenge test. CONCLUSION: Fabrication process as well as humectants concentration were shown to influence water activity. The hydroxyl group positions as well as the presence of an alkyl group on the glycol carbon chain impacted water binding as suggested by DFT calculation. Reducing aw improved the preservation of a cosmetic cream, inhibiting or slowing down the growth of bacteria and fungi. sume Re OBJECTIFS: La conservation des cosmetiques est un pre-requis a l’industrialisation et, parmi les solutions proposees, les cosmetiques auto-conserves sont d’un grand inter^et. Parmi les parametres clefs influencßant l’auto-conservation, la reduction de l’activite de l’eau lutter contre les developpements microbiens dans les peut aider a produits cosmetiques. Ce travail presente une etude sur l’influence d’humectants sur l’activite de l’eau et ses consequences sur la conservation des formulations cosmetiques. Correspondence: Pr. Xavier Fernandez, Parc Valrose, 06108 Nice CEDEX 2, France. Tel.:+334 92 07 69 61; fax: +334 92 07 61 89; e-mail: [email protected]
METHODES: Tout d’abord, des melanges eau-humectants ont ete consideres. L’influence de la teneur en glycols et en glycerine, de la structure chimique des glycols, de la purete de la glycerine, et du procede de formulation sur l’activite de l’eau de ces systemes binaires ont ete etudies. De la modelisation moleculaire a ete realisee afin de mieux comprendre l’impact de la chimie des glycols. cinq formulations cosmeEnsuite, les resultats ont ete appliques a tiques differentes afin d’obtenir des produits optimises. Un challenge test portant sur cinq souches a enfin ete realise. RESULTATS: Nous avons montre que plus la concentration en humectant est elevee, plus l’activite de l’eau est faible. La structure chimique des glycols influence l’activite de l’eau : le propan-1,2diol est plus efficace que le propan-1,3-diol, certainement de part une meilleure stabilisation dans l’eau du propan-1,2-diol, comme goutte du l’ont montre les calculs DFT. Une introduction goutte a glycol dans l’eau favorise une reduction de l’aw. La meilleure reduction de l’activite de l’eau obtenue fut de 6.6% et ce, pour une creme cosmetique dont la conservation a effectivement ete amelioree, comme l’a montre le challenge test. CONCLUSION: Il a ete demontre que le procede de fabrication ainsi que la concentration en humectants influencent l’activite de l’eau. La position des groupements hydroxyles ainsi que la presence d’un groupement alkyle sur la cha^ıne carbonee impactent les interactions avec les molecules d’eau, comme cela a ete montre par des calculs DFT. La reduction de l’aw a permis d’ameliorer la conservation de la creme cosmetique, par inhibition ou ralentissement de la croissance des bacteries et champignons. Objective/Introduction Preservation is one of the most challenging topics in the cosmetic area so that most of the efforts in formulation research are dedicated to that objective [1]. Preservatives have to protect formulation against several bacteria, yeast, moulds and against oxidative stress. They have to be stable and active during the cosmetics shelf life and compatible with the other ingredients of the formulation [2]. In Europe, the cosmetics regulation (EC 1223/2009) authorized the use of 56 synthetics preservatives to fight against micro-organism’s growth [3]. Among the most frequently used (Table I) are parabens, methylchloroisothiazolinone/methylisothiazolinone (MCIT/MIT) or phenoxyethanol [4]. However, in 2004, Dardre et al. published a study revealing that methylparaben was found in human breast tumours [5,6]. Parabens use then became polemical resulting in a large rejection from cosmetic consumers. More recently, restrictions on methylisothiazolinone use was recommended by the European commission.
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie
31
A. Kerdudo et al.
Optimization of cosmetic preservation
reduction in the total number of microbial cells [21] (NF EN ISO 29621). Numerous raw materials are known to impact and especially reduce water activity: humectants, inorganic salts, acids and bases, hydrocolloids or some surfactants [10,22]. Formulation process can also impact water activity, such as raw material temperature, incorporation order or the time of homogenization [5,20]. We present here a comprehensive study on the impact of humectants, namely glycols and glycerin, on the water activity in binary mixtures (water/humectant) as well as in five different cosmetics. Several glycols differing in their structure were compared and the influence of their concentration on aw was studied. The influence of formulation process on aw was also considered. Eventually, a challenge test was carried out on a cream optimized in terms of aw.
Cosmetic industries are then looking for preservation alternatives. Several ways are considered, such as the use of alcohol [5] or essential oils (EO) [7–9]. Nevertheless, each solution presents drawbacks as alcohol can be skin irritating and drying, and EO can cause allergic reaction [10]. Sterilization process using ultra-heat treatment (UHT) [11], gamma or beta radiation, ethylene oxide gaz [5] or supercritical carbon dioxide [12–14] are recent solutions. They, however, require specific and expensive equipment and packaging to limit the contact between the cosmetic product and the environment. Using specific packaging, such as single dose or airless pump, is another solution to reduce or suppress preservatives [15,16] but this is costly and not eco-aware. Good manufacturing practices, or the combination of judicious raw materials having anti-microbial properties, can be considered as alternatives to prevent microbial growth and especially to produce cosmetics requiring weaker preservative concentration. So, an interesting area to focus is the water activity reduction. To multiply, micro-organisms require the presence of bioavailable water [17], also called free water, and commonly measured using water activity (aw). Free water does not interact with raw materials contrarily to bound water that bind to raw materials when mixed together through different interactions such as hydrogen ones [5,18,19]. In general, a decrease in water activity slows down micro-organisms growth [20]. Indeed, to survive and develop, micro-organisms have to maintain turgescent state in cells thanks to osmosis with extracellular medium. A loss in turgescence leads to an increase in latency phase, a decrease in growth and a
Materials and methods Raw materials Five formulations were studied: a lotion (Table II), three oil-in water emulsions (Tables III, IV and VI), and a gel (Table V), compositions of which are given in the corresponding table. Each raw material was from cosmetic grade and was sampled by the suppliers given in the Tables (see Tables). Additional raw materials were also used for this study: natural glycerin was supplied by Interchimie (Compans, France),
Table I Structures of some common authorized preservatives
COOH
O H
H
O H
OH
Formol
Lactic acid
Formic acid
O
O
O
O HO
N OR
N
R=CnH2n+1, n=1, 2, 3, 4
3
S
OH
O
OH
Cl
R = CnH2n+1 avec n = 8, 10 12, 14, 16, 18
Benzalkonium chlorure
H O N O
Propionic acid Cl–+ N R
Triclosan
N N
OH
OH
Cl
H
N H
Imidazolidine urea
32
H
O
Bromonitropropanediol
HO
O
O
Glutaraldehyde
NO2 OH
O
Phenoxyethanol
H
O
Cl
Br
HO
CI
Citric acid
HO
CH3
Methylisothiazolinone and Methylchloroisothiazolinone COOH
O
CH
S
Parabens
HO
OH
H
H
H
N
N
N
CH2 Cl
NH
(CH2)6
NH 2
2
Chlorhexidine
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation Table II Lotion formulation
Phases
Ingredients Commercial Name
INCI name
A
– – D Panthenol 50P Camomile HG Arnica HG Flavoured rose water C red n°40 W 093
Aqua Butylene glycol Panthenol (and) propylene glycol Aqua (and) propylene glycol (and) chamomilla recutita extract Arnica Montana extract, propylene glycol Aqua (and) rosa damascena extract CI 16035 (0,2% solution) Parfum PPG-1-PEG-9 Lauryl Glycol Ether
B C
Eumulgin L
Supplier
Brenntag Alban Muller Alban Muller Gazignaire
AMI
Quantity (%)
91.79 4.00 0.05 1.00 0.50 2.00 0.03 0.03 0.60
Table III O/W emulsion (cold emulsification)
Phases
Ingredients Commercial Name
INCI name
A
– Glycerin 99.5% Petroleum jelly oil Salcare SC 91
Aqua Glycerin Paraffinum liquidum Sodium acrylates copolymer (and) parraffinum liquidum (and) PPG-1 trideceth-6
B C
Supplier
Interchimie
Quantity (%)
88.75 2.00 8.00 1.25
Table IV O/W emulsion (hot emulsification)
Phases
Ingredients Commercial Name
INCI name
Supplier
Quantity (%)
A
/ Glycerin 99,5% EDETA BD Cutina CP Cuitna GMS-V Eumulgin B1 Eumulgin B2 Lanette 16 Fluid petroleum jelly Fucogel 1.5P Sepigel 305
Aqua Glycerin Disodium EDTA Cetyl palmitate Glyceryl stearate Ceteareth-12 Ceteareth-20 Cetyl alcohol Paraffinum liquidum Biosaccharide gum-1 Polyacrylamide (and) C13-14 isoparaffin (and) laureth-7
Interchimie BASF ChemTrade Prod Hyg Ami Prod Hyg Prod Hyg Ami Interchimie Solabia Group Seppic
84.6 2.00 0.10 1.45 3.45 1.50 1.50 2.45 2.00 0.20 0.75
B
C D
Table V Gel formulation
Phases
Ingredients Commercial Name
INCI name
A
– Glycerin 99.5% Carbopol ultrez 10 TEA 99%
Aqua Glycerin Carbomer Triethanolamine
B C
3-methyl-1,3-butanediol was supplied by Kuraray Europe GmbH (Germany) under commercial name Isopentyldiol, and natural butan-1,3-diol was supplied from Kokyu Alcohol Kogyo under commercial name Haisugarcane.
Supplier
Interchimie
Quantity (%)
96.87 2.00 0.53 0.60
Water activity measurement Aw was measured with a Novasina LabSwift-aw (Switzerland) in a 0.03 to 1.00 aw range with an accuracy of 0.010 aw in the
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
33
A. Kerdudo et al.
Optimization of cosmetic preservation
0.10–0.95 aw range. The 21.1 mL measuring chamber contained standardized sample dishes filled to the two-thirds with samples. Measurements were performed at room temperature and in triplicate. Apparatus calibration was performed with standard salts displaying relative humidity of 11%, 58%, 84% and 97%. Formulation processes All the formulations were prepared on a 100 g-basis final mass. Stirring was always performed with a propeller mixer (Turbotest laboratory mixer equipped with an axial flow propeller, Rayneri, VMI). Lotion formulation (Table II). The ingredients of phase A were first weighed separately. They were then mixed together and homogenized with the propeller mixer. Phase B was added under agitation at 600 rpm. Phase C was first homogenized with the propeller mixer before being incorporated in the mixture under stirring at 1000 rpm. The lotion contained 4% of butylene glycol which is the usual amount in commercialized product. Cold O/W fluid emulsion (Table III). Phase A was prepared under stirring at 1000 rpm. The emulsion was formed adding phase B under stirring at 1500 rpm. Phase C was then added to stabilize the emulsion under agitation at 2000 rpm. The glycerin content was fixed to 2% which is the usual amount in commercialized product. Hot O/W fluid emulsion (Table IV). Phase A and B were first prepared separately by mixing the ingredients under stirring with the propeller mixer (500–800 rpm). Then, they were separately heated at 75°C. This temperature was chosen to induce the melting of Phase B ingredients. Phase B (the lipophilic phase) was then incorporated in the hydrophilic phase (phase A) under stirring at 2000 rpm. Stirring was maintained and phase C was added at 30°C. Viscosity was then adjusted adding phase D. The glycerin
content was fixed to 2% which is the usual amount in commercialized product. Gel formulation (Table V). Ingredients of phase A were mixed together under agitation at 500 rpm and then heated at 55°C to favour polymer hydration. Phase B was dispersed in A under stirring at 1500 rpm. Phase C was then added for polymer hydration. The glycerin content was fixed to 2% which is the usual amount in commercialized product. O/W cream emulsion (Table VI). Phase A was heated at 55°C to favour polymer (Phase B) hydration. Phase B was dispersed in A under stirring at 1500 rpm. Stirring was maintained until complete polymer hydration. Phases A+B and C (all the phase C components were weighed together) were then separately heated at 75°C to get liquid phases. Emulsification was performed by addition of the phase C (lipophilic) in phases A+B (hydrophilic phase). Phases D and E were incorporated one after another at 60°C under stirring at 1500 rpm. Phases F, G, H and I were successively incorporated at 30°C under stirring at 1500 rpm. The glycerin content was fixed to 4% which is the usual amount in commercialized product. Challenge tests Antimicrobial preservative efficiency evaluation, realized by an external laboratory (Ideatest, Plouzane, France), was performed following European Pharmacopeia (Eur. Ph.). Five potential pathogenic germs were chosen for this study: Pseudomonas aeruginosa ATCC 9027, Staphylococcus aureus ATCC 6538, Escherichia coli ATCC 8739, Candida albicans ATCC 10231 and Aspergillus brasiliensis ATCC 16404. A contamination by these micro-organisms could cause illness such as skin infection or respiratory diseases. Tryptic soy agar and sabouraud dextrose was used for culture media for bacteria, moulds and yeasts, respectively. Cosmetic formulations were sampled in single-used sterile flasks, and every flask
Table VI Cream (O/W) formulation
Phases
Ingredients Commercial Name
INCI name
A
– Glycerin 99.5% Carbopol EDT 2020 Brij 72 Brij 721P Lanette 18 Cutina GMS-V Eutanol G Sophiderm Arlamol HD Virgin jojoba oil Refined wheat germ oil Refined shea butter Silicone oil baysilone M350 Eusolex 9020 Uvinul MC 80 Silicone oil SF1202 TEA 99% – D panthenol 50P Vitamin E acetate
Aqua Glycerin Acrylates/C10-30 alkyl acrylate cross-polymer Polyoxyethylene(2) (and) stearyl ether Steareth-21 (and) polyethoxylated alcohol Stearyl alcohol Glyceryl stearate Octyldodecanol Squalane Isohexadecane Simmondsia chinensis (Jojoba) seed oil Triticum vulgare (wheat) germ oil Butyrospermum parkii butter Dimethicone Butyl methoxydibenzoyl-methane Ethylhexyl methoxycinnamate Dimethicone Triethanolamine Isononyl isononanoate, titanium dioxide Panthenol (and) propylene glycol Tocopheryl Sodium hyaluronate Parfum
B C
D E F G H I
34
Supplier
Interchimie Gattefosse Croda France SAS Quimica masso Ami Ami Ami Sophim Quimica masso Olvea Olisud Interchimie Interchimie IES IES Brenntag specialties ^te d’Azur Brenntag Co – Laserson IES
Quantity (%)
59.34 4.00 0.15 2.00 2.50 1.50 3.00 3.00 2.50 3.00 1.50 1.50 2.00 1.50 1.00 1.50 2.00 0.11 1.25 0.20 0.50 5.85 0.10
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation Table VII European Pharmacopeia criteria (logarithmic reduction as a function of days D)
Aw evolution as a function of glycerin concentration 1.0
European Pharmacopeia
Logarithmic reduction
0.9 Criteria
D2
D7
D14
D28
Bacteria
A B A B
2 – – –
3 – – –
– 3 2 1
NI NI NI NI
Yeast and moulds
Table VIII Minimal water activity allowing micro-organisms growth, at 25°C [5,19,33]
Bacteria Pseudomonas aeruginosa Escherichia coli Staphylococcus aureus Yeast and fungi Aspergillus brasiliensis Candida albicans
Natural glycerin
0.7 0.6
y = –0.8094x2 + 0.0537x + 0.9654 R2 = 0.9978
0.5
NI, No Increase, –, no minimal reduction required.
Micro-organisms
Synthetic glycerin
0.8
aw
Strains
Minimal aw
0.97 0.95 0.86 0.77 0.87
was inoculated with strain suspension (at 22.5 2.5 °C in the dark). Inoculation concentration was fixed to 106 CFU mL1, 105 CFU mL1 and 104 CFU mL1, for bacteria, yeast and moulds, respectively. The validation of the neutralization of the preservative was realized on the 5 strains, in the 1/10th and 1/100th dilution, with LT100 broth. The microbial densities were counted by inclusion (results in CFU g1or mL) and compared with the logarithmic reduction criteria of the European Pharmacopeia (Table VII). DFT calculations The calculations were performed with the Gaussian 03 program [23]. The molecules structures were optimized, and energies were calculated by the DFT (Density Functional Theory) method with the exchange-correlation functional B3LYP [24–26] and the 631 + G(d,p) basis set for calculations. The effects of water as solvent were taken into account using the polarizable continuum model (PCM). Results and Discussion When measured, water activity (aw) values were systematically compared with the reference values, shown in Table VIII, as our aim was to reduce aw under these critical values to impede microorganisms growth. All raw materials having affinity with water molecules through hydrogen bonds, hydrophilic interactions or Van der Waals interactions are presumed to reduce aw: humectants, rheology modifiers, surfactants, fatty acids and their esters are recommended [27–29]. This study will focus on glycols and glycerin, two humectants, and to begin with on the influence of their concentration and their chemical structure on aw.
0.4 0%
20%
40%
60%
80%
100%
Glycerin concentration (wt %) Figure 1 Water activity as a function of synthetic glycerin (dark circles) concentration. Natural glycerin (gray squares) was compared at 16 wt%. Standard deviations are included in points.
Influence of humectants on aw Hydrogen bonds formation between glycerin or glycols and water molecules may reduce free water in finished product and thus water activity. In a first part of the study, glycerin was diluted in water and its impact on aw as a function of its concentration was examined. Glycerin impact on water activity: Concentration effect. Figure 1 shows water activity as a function of glycerin amount (from 4% to 80 wt %). The higher the glycerin concentration, the lower aw, as expected. 8% of glycerin was sufficient to reduce aw under 0.97, the minimal growth value for P. aeruginosa (Table VIII). With 16% of glycerin, aw was lower than 0.95, the minimal growth value for E. coli. Finally, more than 32% of glycerin was necessary to reduce other micro-organisms’ development. Let us note that such a huge amount is hardly conceivable in cosmetic formulation because of sensorial consequence (highly sticky products). Influence of glycerin source. The influence of a natural glycerin was also studied. The natural product was much more active than the synthetic one: 16 wt% of natural material was as efficient as 40 wt% of synthetic one, leading to 0.866 0.010 for aw value. This value was lower than the minimal aw allowing C. albicans growth (0.87). Such a difference in efficiency could be explained by the purity of the two sources of glycerin. The purity of the synthetic one was 99.5% (the remaining 0.5% being water), whereas that of the natural one, obtained from corn and soya fractionation, was 98%. Unknown residues from fractionation could explain our results. Glycols impact on aw: Influence of their chemical structure. Figure 2 shows the concentration dependency of water activity for two glycols differing in their chemical structure: butan-1,3-diol and 3methyl-1,3-butanediol. As expected, aw decreased with glycol concentration. However, 3-methyl-1,3-butanediol was more efficient than butan-1,3-diol, except for the highest concentration, 80 wt%. The additional methyl group could influence the stability of the hydrogen bond through an iceberg effect [30]. The additional methyl group decreases the polarity of the molecule so that ice
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
35
A. Kerdudo et al.
Optimization of cosmetic preservation
Aw evolution as a function of propan-1,2diol and propa-1,3-diol concentration
Aw evolution as a function of butan-1,3-diol and 3-methyl-1,3-butanediol 1
1.0
0.9
Propan1,2diol Propan1,3diol
0.8
0.7
aw
aw
0.9
y = –0.4773x2 – 0.0743x + 0.9736 R2 = 0.9974
0.8
y = –0.7255x2 + 0.024x + 0.9468 R2 = 0.9986
y = 0.0107x2 – 0.3244x + 0.9167 R2 = 0.9737
0.6
0.7
y = –0.3255x2 – 0.2852x + 0.9455 R2 = 0.9864
0.6
butan-1,3-diol
0.5
0.5
3-methyl-1,3-butanediol
0.4 0%
20%
40%
60%
80%
100%
0.4 0%
Glycols concentration (wt %)
organization of the water molecules in the first layer surrounding the molecule could be generated. This effect could also decrease water mobility and thus water activity [30]. The efficiency of 3methyl-1,3-butanediol was particularly pronounced at low concentration (8–16%): aw decreased to c. 0.85 for 16 wt%, value for which all micro-organisms, except A. niger, cannot grow (Table VIII). Two other glycols, differing also in their chemical structure, were compared. In this case, the difference came from the position of hydroxyl groups: propan-1,2-diol vs propan-1,3-diol. Figure 3 shows that they behaved similarly at extreme concentrations (≤5% and ≥65%). In between, propan-1,2-diol was more efficient than propan-1,3-diol. This suggests that the hydroxyl group position on a linear carbon chain can influence hydrogen bond formations. In that sense, propan-1,2-diol seems to be better stabilized in water than propan-1,3-diol. To confirm this assumption, molecular modelling was carried out. Molecular modelling of propanediols. The Density functional theory (DFT) is a very popular quantum chemistry method to investigate the electronic ground state of atoms and molecules. The underlying theory was developed by Hohenberg and Kohn [31] and later by Kohn, Sham and Pople who were awarded the Chemistry Nobel Prize in 1998.* Hohenberg and Kohn demonstrated that the electronic energy is a function of the electronic density, as well as being a function of the total many body wavefunction. Therefore, it reduces the number of variables in the Schr€odinger equation, energy being of function of the three-dimensional variables instead of 3N with N the number of electrons. The contribution of Kohn and Sham allowed treating interacting electrons as independent fermions moving in a mean field represented by other electrons and parameterized experimentally. This paradigm shift allows treating hundred of atoms and allows chemists to investigate molecules in solution as well as solid state chemistry with a reasonable computational cost. The DFT calculations, and widely quantum chemistry calculations, start by guessing the electronic density and then optimizing the density by following the so-called variational principle, ensuring to find the ground state correctly. To model solvation effects within the DFT method, the PCM method was developed by Tomasi
36
40%
60%
80%
100%
Glycols concentration (wt %) Figure 3 Water activity as a function of propan-1,2-diol (gray squares) and propan-1,3-diol (dark circles) concentration.
Propan-1,2-diol
Propan-1,3-diol
In water
Figure 2 Water activity as a function of butan-1,3-diol (dark circles) and 3-methyl-1,3-butanediol (gray circles) concentration.
20%
Figure 4 Molecular modelling of propan-1,2-diol and propan-1,3-diol in water.
and co-workers [32] and implemented in the GAUSSIAN software. This method replaces the solvent molecules by placing the solute in a cavity within the solvent reaction field, allowing the dielectric continuum to polarize the total electronic wavefunction. To confirm that propan-1,2-diol is better stabilized in water than propan-1,3-diol, the total energy (free enthalpy) of both molecules have been calculated in water and in vacuum as the dispersion or solubilization environment of a molecule influences its polarizability and therefore its geometry. The corresponding optimized structures are shown in Figs 4 and 5. Solvation energy was then calculated (see Eq. 1); the lower the energy, the more stable the molecule. Solvation energy corresponds to the solvent energy participating to the molecule stabilization. Esolv ¼ Em1ðwaterÞ Em1ðvacuumÞ
ð1Þ
where Esolv is the solvation energy, Em1(water) the total energy of molecule 1 in water and Em1(vacuum) the total energy of molecule 1 in vacuum. Modelling the propanediols in vacuum, we observed a modification of the propan-1,2-diol geometry in comparison with that obtained in water, whereas no change was observed for propan1,3-diol (Fig. 5). In vacuum, an intramolecular hydrogen bond was created between the hydrogen of hydroxyl group in C1 position and the oxygen of the hydroxyl group in C2 position. As
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation
Propan-1,2-diol
Propan-1,3-diol
Table X Description of the 5 compared protocols for water solution containing 16% (wt) of glycerin
In vacuum
Formulation protocols tested with a 16% (wt) water solution of glycerin
Protocol 1 Protocol 2
Figure 5 Molecular modelling of propan-1,2-diol and propan-1,3-diol in vacuum.
Protocol 3
Table IX Solvation energies of propan-1,2-diol and propan-1,3-diol Protocol 4 Esolv (kcal mol1) Protocol 5 14.82 13.52
this hydrogen bond formation increased the propan-1,2-diol stability, we could not consider the associated energy to calculate Esolv and compare it with propan-1,3-diol. Therefore, we calculated the energy of the molecule in vacuum considering that its geometry is that calculated in water. Solvation energies of molecules were then calculated as exposed in Table IX. DFT calculations demonstrated that propan-1,2-diol was better stabilized in water than propan1,3-diol confirming then our conclusion based on water activity measurement.
Protocol impact on water activity 0.96 0.94 0.92 0.90
aw
Propan-1,2-diol Propan-1,3-diol
Weigh water and glycerin Homogenize with propeller mixer at 500 rpm Weigh glycerin Add water drop by drop under stirring using a propeller mixer at 500 rpm Hold on stirring for 5 min Weigh water and glycerin Homogenize with propeller mixer Increase temperature at 75°C with water-bath Stir with propeller mixer for 10 min at 500 rpm Weigh water Add glycerin drop by drop under stirring with propeller mixer at 500 rpm Hold on stirring for 5 min Weigh water Add glycerin drop by drop under stirring with propeller mixer at 500 rpm Hold on stirring for 15 min
0.88
Influence of the preparation protocol on aw Literature reveals that processing method, processing rate, processing conditions, time interval between mixing and processing are part of critical factors influencing aw [20]. To have a further insight on the influence of processing on aw, we compared five protocols (Table X) on a simple, two-binary system: glycerin (16wt %)/water (84wt%). The modified parameters were the mixing process (P1/P2 and P1/P4), the order of both ingredients introduction (P2/P4), the mixing temperature (P1/P3) and the duration of stirring (P4/P5). These parameters were chosen as we believed they could play on the interactions between water molecules and glycerin. The mixture composition was chosen from previous section (1.a) because of its efficiency on aw and of its relative low glycerin amount compatible with cosmetic formulation constraints (too much glycerin would produce sticky products). As shown in Fig. 6, differences in water activity were obtained as a function of the applied protocol (P). In particular, the drop by drop addition of glycerin in water under mixing had a beneficial effect: a 7.5% and 6.5% decrease of aw was reached using P4 and P5, respectively, as compared to P1 (the reference protocol). This effect was less pronounced when water (P2), instead of glycerin (P4), was drop by drop added. In all cases, the successive addition of the raw materials under mixing (P2, P4 and P5) systematically decreases aw as compared with assays where both raw materials were mixed together (P1 and P3). Finally, increasing the time of mixing (P4/ P5), or heating during stirring when both components are initially mixed (P2/P1) slightly decreased aw.
0.86 0.84 0.82 P1
P2
P3
P4
P5
Figure 6 Water activity of glycerin solution (16% wt) as a function of applied protocol. All protocols were described in Table X.
To confirm that P4 had a positive impact on aw, other binary systems were studied: synthetic glycerin, natural glycerin 1, synthetic and natural butan-1,3-diol. Their concentration was fixed to 4 wt% as this is the usual polyol concentration in cosmetics. For all assays (Fig. 7), water activity was indeed reduced, from 1.2% to 2.3%, using P4 instead of P1. One can assume that drop by drop pouring of the polyol allows increasing the contact interface between water and polyol favouring interactions between them. Measurement of aw on cosmetic formulations Five cosmetic formulations were prepared following the protocol exposed in the material and method section (corresponding to protocol P1): a lotion, a gel, a cream and two oil-in-water emulsions. Their composition is given in Tables II–VI. The water activity of these formulations, measured 1 day after their preparation, is given in Table XI: all of them present a comparable aw value of c. 0.97.
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
37
A. Kerdudo et al.
Optimization of cosmetic preservation
Protocol impact on aw as a function of raw material
1.00
Aw of formulation as a function of protocol 0.98
Protocol 4
0.94
0.96
0.92
aw
0.94
aw
Reference
0.96
Protocol 1 Protocol 4
0.98
0.92
0.90 0.88
0.90
0.86
0.88
0.84
0.86
0.82 0.80
0.84
Glyc
erin Bu
,3tan-1
in 1 -diol lycer n-1,3 ral g buta l a r Natu Natu
diol
Figure 7 Water activity of glycol solutions as a function of applied protocol. All solutions contained 4% of corresponding glycol.
–3,8%
–2,8%
Lotion
Gel
–4,1%
O/W fluid emulsion (cold)
Figure 8 Water activity of cosmetic formulations as a function of applied protocol. The reference product was formulated following the protocol P1 described in the materials and methods section.
Table XI Water activity of the five references cosmetic products 1 day after formulation
Aw of formulation as a function of protocol 0.98 0.96
Data at day + 1
0.92
0.001 0.001 0.000 0.001 0.001
26.0 26.5 26.3 26.5 26.2
0.88 0.86
–3,8%
–4,8%
–5,4%
–4,3%
–6,6%
O/W cream emulsion
0.970 0.973 0.971 0.971 0.961
O/W fliud emulsion (hot)
Lotion O/W fluid emulsion (cold process) O/W fluid emulsion (hot process) Gel O/W cream
0.90
O/W fluid emulsion (cold)
T (°C)
Gel
std
Lotion
aw
aw
Formulation
0.94
0.84 0.82 0.80
Bold values corresponded to the water activity values of the cosmetic products
Based on the previous section, the protocol was slightly modified: the active ingredient of Phase A (glycerin or butylene glycol) was added drop by drop to water, following then P4 protocol. This led to a systematic decrease in aw (from 2.8% to 4.1%) for the three tested formulations (lotion, gel and cold emulsion, Fig. 8). In all case, aw was lower than 0.95 implying a better protection against P. aeruginosa and E. coli development. In a next step, all cosmetics were prepared following protocol 4 but with an increased glycol concentration as its positive action on aw was demonstrated above: 10% instead of 2 or 4%. The increase in glycol amount was then balanced by a decrease in water amount. Figure 9 predicts a better protection against P. aeruginosa and E. coli growth, especially for the cream for which aw was about 0.898 0.009 at 25°C, the only value lower than 0.90. However, this value is not low enough to expect a protection against S. aureus, yeast and mould (Table VIII). In a last formulation, the cream was prepared with 10% of natural glycerin instead of synthetic one, following protocol 4. Surprisingly, this had no effect on water activity (data not shown). Challenge test As the cream displayed the lowest aw value, it was chosen for performing a challenge test. The ‘reference’ cream (containing 4% of
38
Figure 9 Water activity of cosmetic formulations as a function of applied protocol. Deep gray: aw of references product. Light gray: aw of optimized product, formulated with protocol 4 and 10% of glycol.
glycerin and prepared following process P1) and the optimized one (prepared following P4 with 10% of natural glycerin) were compared. Challenge tests were performed for a 14-day period. As shown in Table XII, the optimized product was systematically more efficient to fight against micro-organism growth. Its antibacterial activity was remarkable as the logarithmic reduction of P. aeruginosa and S. aureus are in accordance with the criteria B of the Eur. Ph. for 14 days. C. albicans and A. brasiliensis growth in the optimized cream was less marked than in the reference product, but still remained too elevated for the Eur. Ph. regulation. The growth of C. albicans was only inhibited in the reference cream, whereas its population was reduced by 0.2 log in the optimized one. A. brasiliensis was still growing in the reference cream after 14 days, whereas a slight inhibition was noted in the optimized product. For all strains, the optimized cream was thus more efficient than the reference one. These data could be explained thanks to water activity reduction for bacterial growth protection. However, in the best cases, only
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation
Table XII Logarithmic reduction by strains for the reference cream and the optimized one formulated following the protocol 4 with 10% of natural glycerin 1
Strains Days
Criteria A Criteria B Reference Interpretation aw optimized Interpretation
P. aeruginosa
S. aureus
E. coli
C. albicans
A. brasiliensis
D2
D7
D14
D2
D7
D14
D2
D7
D14
D2
D7
D14
D2
D7
D14
≥2 – 1.3 B 0.9 B
≥3 – >3 B >3 B
– ≥3 1.1 NC >3 B
≥2 – 0.2 B 1.1 B
≥3 – 0.0 B 2.0 B
– ≥3 0.2 NC >3 B
≥2 – 0.5 B 0.9 B
≥3 – 0.7 B 0.9 B
– ≥3 0.0 NC 1.1 NC
– – –
– – –
– – –
– – –
– –
– –
≥2 ≥1 0.0 NC 0.2 NC
– –
– –
≥2 ≥1 0.1 NC 0.1 NC
NC, not conform. DX, Day X.
the criteria B of the Eur. Ph.were fulfilled. This was not surprising as self preserved product with reduced water activity are not supposed to be bactericide or fungicide [19,20]. A low aw will only induce a slowdown in micro-organisms’ growth [20]. The lower the aw value, the longer the lag time before growth initiation. Low water activities may not be sufficient to kill micro-organisms. The survival of germs depends on other stresses imposed by the composition (lack of nutrients, membrane destabilizing surfactants or chelating agents) [19]. Eventually, it was not surprising that antibacterial activity was better than anti-fungal one as it is more difficult to protect a product against mould and yeast than against bacteria [5,19,33] (Table VIII).
esting way could be to enlarge DFT calculation to correlate water activity with physical data. Manufacturing process was also shown to impact water activity: a slow introduction of glycol in water under mixing improves water binding. Thinking about scaling up this process, we could imagine that glycols could be added as a dribble rather than drop by drop, reducing then time processing, and presumably, decreasing aw by comparison with a direct mixing. These results were applied in cosmetic formulations. A final 6.6% water activity reduction was obtained for the cream formulation. The challenge test showed an improved preservation of the optimized cream. Acknowledgements
Conclusion As a conclusion, significant impact of glycerin and glycols concentration on water activity was demonstrated. Differences were observed in function of raw material origin. Natural products were more efficient to reduce aw than synthetic ones, maybe because of lower purity. The hydroxyl group position had an impact on water binding. Water activity measurement and molecular modelling showed that propan-1,2-diol was better stabilized in water than propan-1,3-diol. Then, the supplementary methyl in C3 position of 3-methyl-1,3-butanediol increases water binding ability in comparison with butane-1,3-diol, except at high concentration. An inter-
The authors are grateful to SO.F.I.A. Cosmetiques and Association Nationale de la Recherche et de la Technologie (ANRT) for financial support of this project. The authors wish to thank FEDER and PACA area for financial support. This project was supported by the University of Nice-Sophia Antipolis, the Bordeaux Polytechnic National Institute, and the CNRS.
ENDNOTE * Nobel Prize in chemistry 1998.
References 1. Dweck, A. Natural preservative – an update. In Formulating Natural Cosmetics (Kozlowski, A.C., ed.), pp. 107–130. Alluredbooks, USA (2011). 2. Amaral, L.F.B., Camilo, N.S., Pereda, M.D.C.V., Levy, C.E., Moriel, P. and Mazzola, P.G. Evaluation of antimicrobial effectiveness of C-8 xylitol monoester as an alternative preservative for cosmetic products. Int. J. Cosm. Sci. 33, 391–397 (2011). 3. EU regulation, Cosmetic products regulation (1223/2009). 4. Lundov, M.D., Johansen, J.D., Zachariae, C. and Moesby, L. Low-level efficacy of cosmetic preservatives. Int. J. Cosm. Sci. 33, 190–196 (2011).
5. Fernandez, X., Merck, F. and Kerdudo, A. Conservateurs pour cosmetiques – Generalites et conservateurs antimicrobiens. Techniques de l’Ingenieur, J2284, 1–21 (2012). 6. Dardre, P.D., Aljarrah, A., Miller, W.R., Coldham, N.G., Sauer, M.J. and Pope, G.S. Concentrations of parabens in human breast tumours. J. Appl. Toxicol. 24, 5–13 (2004). 7. Hammer, K. A, C.F. Carson and Riley, T.V. Antimicrobial activity of essential oils and other plant extracts. J. Appl. Microbiol. 86, 985–990 (1999). 8. Kunicka-Styczynska, A., Sikora, M. and KaLembz, D. Antimicrobial activity of lavender, tea tree and lemon oils in cosmetic preser-
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
9.
10.
11. 12.
vative systems. J. Appl. Microbiol. 107, 1903–1911 (2009). Baratta, M.T., Dorman, H.J.D., Deans, S.G., Figueiredo, A.C., Barroso, J.G. and Ruberto, G. Antimicrobial and antioxidant properties of some commercial essential oils. Flavour Fragr. J. 13, 235–244 (1998). Kabara, J.J. Aroma chemicals as preservatives. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice, (Kabara, J.J. and Orth, D.S. ed.), pp. 181– 208. Marcel Dekker, Phoenix (1996). Lopez, D. and Lopez, C., Cosmetic product sterilization, WO 2007/148022 A2. Oule, K.M., Dickman, M. and Arul, J. Microbicidal effect of pressurized CO2 and the
39
A. Kerdudo et al.
Optimization of cosmetic preservation
13.
14.
15.
16.
17.
18.
40
influence of sensitizing additives. Eur. J. Sci. Res. 41, 569–581 (2010). Spilimbergo, S., Mantoan, D., Quaranta, A. and Della-Mea, G. Real-time monitoring of cell membrane modification during supercritical CO2 pasteurization. J. Supercrit. Fluids 48, 93–97 (2009). White, A., Burns, D. and Christensen, T.W. Effective terminal sterilization using supercritical carbon dioxide. J. Biotechnol. 123, 504–515 (2006). Brannan, D.K. The role of packaging in product preservation. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice, (Kabara, J.J. and Orth, D.S., ed.), pp. 227–242, Marcel Dekker, Phoenix (1996). Fernandez, X., Merck, F. and Kerdudo, A. Conservateurs pour cosmetiques – Antioxydants et Anti-UV. Techniques de l’ingenieur, J2285, 1–19 (2012). Berthele, H., Sella, O., Lavarde, M., Mielcarek, C., Pense-Lheritier, A.-M. and Pirnay, S. Determination of the influence of factors (ethanol, pH and aw) on the preservation of cosmetics using experimental design. Int. J. Cosm. Sci. 36, 54–61 (2014). Faiveley, M. L’eau et la conservation des aliments. Notions de biochimie alimentaire et alimentation humaine. Techniques de l’Ingenieur, F1011, 1–20 (2003).
19. Kabara, J.J. Principles for Product Preservation. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice (Kabara, J.J. and Orth, D.S., ed.), pp. 1–14. Marcel Dekker, Phoenix (1996). 20. Enigl, D.C. and Sorrells, K.M. Water Activity and Self-Preserving Formulas. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice, (Kabara, J.J. and Orth, D.S., ed.), pp. 45–73. Marcel Dekker, Phoenix (1996). 21. NF EN ISO 29621 Cosmetiques - Microbiologie - Lignes directrices pour l’appreciation du risque et l’identification de produits a faible risque microbiologique (2011) 22. Varvaresou, A., Papageorgiou, S., Tsirivas, E., Protopapa, E., Kintziou, H., Kefala, V. and Demetzos, C. Self-preserving cosmetics. Int. J. Cosm. Sci. 31, 163–175 (2009). 23. Frisch, M. J., Trucks, G. W., Schlegel, H. B. et al. Gaussian03, Revision C.02, Gaussian, Inc., Wallingford CT (2004). 24. Becke, A.D. Density-functional exchangeenergy approximation with correct asymptotic behavior. Phys. Rev. A 38, 3098–3100 (1988). 25. Becke, A.D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 98, 5648–5652 (1993). 26. Lee, C., Yang, W. and Parr, R.G. Development of the Colle-Salvetti correlation-energy
27.
28.
29.
30.
31.
32.
33.
formula into a functional of the electron density. Phys. Rev. B 37, 785–789 (1988). Bergasson, G., Arnfinnsson, J., Steingrimmson, O. and Thormar, H. In vitro killing of Candida albicans by fatty acids and monoglycerides. Antimicrob. Agents Chemother. 45, 3209–3212 (2001). Kabara, J.J. Fatty acids and esters as multifunctional components. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice. (Kabara, J.J. and Orth, D.S., ed.), pp. 119–137. Marcel Dekker, Phoenix (1996). Ingram, L.O. Mechanism of Lysis of Escherichia coli by Ethanol and Other Chaotropic Agents. J. Bacteriol. 146, 331–336 (1981). Dill, K.A., Truskett, T.M., Vlachy, V. and Hribar-Lee, B. Modeling water, the hydrophobic effect and ion solvation. Annu. Rev. Biophys. Biomol. Struct. 34, 173–199 (2005). Hohenberg, P. and Kohn, W. Inhomogeneous Electron Gas. Phys. Rev. 136, B864– B871 (1964). Tomasi, J., Mennucci, B. and Cances, E. The IEF version of the PCM solvation method: an overview of a new method addressed to study molecular solutes at the QM ab initio level. J. Mol. Struct. (Theochem) 464, 211– 226 (1999). Orth, D.S. Handbook of cosmetic Microbiology. pp. 1–108, Marcel Dekker, New York (1993)
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
doi: 10.1111/ics.12164
Optimization of cosmetic preservation: water activity reduction A. Kerdudo*,†, F. Fontaine-Vive†, A. Dingas*, C. Faure‡ and X. Fernandez† *SO.F.I.A. Cosmetiques, 1ere Avenue, 1ere Rue, 06514 Carros, France, †Universite de Nice Sophia Antipolis, ICN, UMR 7272, Parc Valrose, 06108 Nice CEDEX 2, France and ‡Laboratoire de Chimie et Biologie des Membranes et Nano-objets, Univ. Bordeaux, CBMN, UMR 5248, Allee Geoffroy St Hilaire, F-33600 Pessac, France
Received 26 May 2014, Accepted 20 September 2014
Keywords: formulation, glycol, microbiology, polymer, preservation, water activity
Synopsis OBJECTIVE: Preservation of cosmetics is a prerequisite for industrialization, and among the proposed solutions, self-preserved cosmetics are of great interest. One key influencing parameter in selfpreservation is water activity; its reduction can help to fight against microbial growth in cosmetic products. This work presents a study on the influence of humectants on water activity and its consequence on the preservation of cosmetic formulations. METHODS: First, water–humectants mixtures were considered. The influence of glycol and glycerin content, glycol chemical structure, glycerin purity and formulation process on the water activity of the binary mixture was studied. Molecular modelling was performed for a better understanding of the impact of glycol chemistry. Then, the results were applied to five different cosmetic formulations to get optimized products. Challenge test on five strains was carried out in that sense. RESULTS: We showed that the higher the humectants concentration, the lower the water activity. Glycol chemical structure also influenced water activity: propan-1,2-diol was more efficient than propan-1,3-diol, certainly because of a better stabilization in water of propan-1,2-diol as shown by DFT calculation. A drop by drop introduction of glycol in water favoured aw reduction. The best water activity loss was 6.6% and was reached on the cream formulation whose preservation was improved as evidenced by challenge test. CONCLUSION: Fabrication process as well as humectants concentration were shown to influence water activity. The hydroxyl group positions as well as the presence of an alkyl group on the glycol carbon chain impacted water binding as suggested by DFT calculation. Reducing aw improved the preservation of a cosmetic cream, inhibiting or slowing down the growth of bacteria and fungi. sume Re OBJECTIFS: La conservation des cosmetiques est un pre-requis a l’industrialisation et, parmi les solutions proposees, les cosmetiques auto-conserves sont d’un grand inter^et. Parmi les parametres clefs influencßant l’auto-conservation, la reduction de l’activite de l’eau lutter contre les developpements microbiens dans les peut aider a produits cosmetiques. Ce travail presente une etude sur l’influence d’humectants sur l’activite de l’eau et ses consequences sur la conservation des formulations cosmetiques. Correspondence: Pr. Xavier Fernandez, Parc Valrose, 06108 Nice CEDEX 2, France. Tel.:+334 92 07 69 61; fax: +334 92 07 61 89; e-mail: [email protected]
METHODES: Tout d’abord, des melanges eau-humectants ont ete consideres. L’influence de la teneur en glycols et en glycerine, de la structure chimique des glycols, de la purete de la glycerine, et du procede de formulation sur l’activite de l’eau de ces systemes binaires ont ete etudies. De la modelisation moleculaire a ete realisee afin de mieux comprendre l’impact de la chimie des glycols. cinq formulations cosmeEnsuite, les resultats ont ete appliques a tiques differentes afin d’obtenir des produits optimises. Un challenge test portant sur cinq souches a enfin ete realise. RESULTATS: Nous avons montre que plus la concentration en humectant est elevee, plus l’activite de l’eau est faible. La structure chimique des glycols influence l’activite de l’eau : le propan-1,2diol est plus efficace que le propan-1,3-diol, certainement de part une meilleure stabilisation dans l’eau du propan-1,2-diol, comme goutte du l’ont montre les calculs DFT. Une introduction goutte a glycol dans l’eau favorise une reduction de l’aw. La meilleure reduction de l’activite de l’eau obtenue fut de 6.6% et ce, pour une creme cosmetique dont la conservation a effectivement ete amelioree, comme l’a montre le challenge test. CONCLUSION: Il a ete demontre que le procede de fabrication ainsi que la concentration en humectants influencent l’activite de l’eau. La position des groupements hydroxyles ainsi que la presence d’un groupement alkyle sur la cha^ıne carbonee impactent les interactions avec les molecules d’eau, comme cela a ete montre par des calculs DFT. La reduction de l’aw a permis d’ameliorer la conservation de la creme cosmetique, par inhibition ou ralentissement de la croissance des bacteries et champignons. Objective/Introduction Preservation is one of the most challenging topics in the cosmetic area so that most of the efforts in formulation research are dedicated to that objective [1]. Preservatives have to protect formulation against several bacteria, yeast, moulds and against oxidative stress. They have to be stable and active during the cosmetics shelf life and compatible with the other ingredients of the formulation [2]. In Europe, the cosmetics regulation (EC 1223/2009) authorized the use of 56 synthetics preservatives to fight against micro-organism’s growth [3]. Among the most frequently used (Table I) are parabens, methylchloroisothiazolinone/methylisothiazolinone (MCIT/MIT) or phenoxyethanol [4]. However, in 2004, Dardre et al. published a study revealing that methylparaben was found in human breast tumours [5,6]. Parabens use then became polemical resulting in a large rejection from cosmetic consumers. More recently, restrictions on methylisothiazolinone use was recommended by the European commission.
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie
31
A. Kerdudo et al.
Optimization of cosmetic preservation
reduction in the total number of microbial cells [21] (NF EN ISO 29621). Numerous raw materials are known to impact and especially reduce water activity: humectants, inorganic salts, acids and bases, hydrocolloids or some surfactants [10,22]. Formulation process can also impact water activity, such as raw material temperature, incorporation order or the time of homogenization [5,20]. We present here a comprehensive study on the impact of humectants, namely glycols and glycerin, on the water activity in binary mixtures (water/humectant) as well as in five different cosmetics. Several glycols differing in their structure were compared and the influence of their concentration on aw was studied. The influence of formulation process on aw was also considered. Eventually, a challenge test was carried out on a cream optimized in terms of aw.
Cosmetic industries are then looking for preservation alternatives. Several ways are considered, such as the use of alcohol [5] or essential oils (EO) [7–9]. Nevertheless, each solution presents drawbacks as alcohol can be skin irritating and drying, and EO can cause allergic reaction [10]. Sterilization process using ultra-heat treatment (UHT) [11], gamma or beta radiation, ethylene oxide gaz [5] or supercritical carbon dioxide [12–14] are recent solutions. They, however, require specific and expensive equipment and packaging to limit the contact between the cosmetic product and the environment. Using specific packaging, such as single dose or airless pump, is another solution to reduce or suppress preservatives [15,16] but this is costly and not eco-aware. Good manufacturing practices, or the combination of judicious raw materials having anti-microbial properties, can be considered as alternatives to prevent microbial growth and especially to produce cosmetics requiring weaker preservative concentration. So, an interesting area to focus is the water activity reduction. To multiply, micro-organisms require the presence of bioavailable water [17], also called free water, and commonly measured using water activity (aw). Free water does not interact with raw materials contrarily to bound water that bind to raw materials when mixed together through different interactions such as hydrogen ones [5,18,19]. In general, a decrease in water activity slows down micro-organisms growth [20]. Indeed, to survive and develop, micro-organisms have to maintain turgescent state in cells thanks to osmosis with extracellular medium. A loss in turgescence leads to an increase in latency phase, a decrease in growth and a
Materials and methods Raw materials Five formulations were studied: a lotion (Table II), three oil-in water emulsions (Tables III, IV and VI), and a gel (Table V), compositions of which are given in the corresponding table. Each raw material was from cosmetic grade and was sampled by the suppliers given in the Tables (see Tables). Additional raw materials were also used for this study: natural glycerin was supplied by Interchimie (Compans, France),
Table I Structures of some common authorized preservatives
COOH
O H
H
O H
OH
Formol
Lactic acid
Formic acid
O
O
O
O HO
N OR
N
R=CnH2n+1, n=1, 2, 3, 4
3
S
OH
O
OH
Cl
R = CnH2n+1 avec n = 8, 10 12, 14, 16, 18
Benzalkonium chlorure
H O N O
Propionic acid Cl–+ N R
Triclosan
N N
OH
OH
Cl
H
N H
Imidazolidine urea
32
H
O
Bromonitropropanediol
HO
O
O
Glutaraldehyde
NO2 OH
O
Phenoxyethanol
H
O
Cl
Br
HO
CI
Citric acid
HO
CH3
Methylisothiazolinone and Methylchloroisothiazolinone COOH
O
CH
S
Parabens
HO
OH
H
H
H
N
N
N
CH2 Cl
NH
(CH2)6
NH 2
2
Chlorhexidine
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation Table II Lotion formulation
Phases
Ingredients Commercial Name
INCI name
A
– – D Panthenol 50P Camomile HG Arnica HG Flavoured rose water C red n°40 W 093
Aqua Butylene glycol Panthenol (and) propylene glycol Aqua (and) propylene glycol (and) chamomilla recutita extract Arnica Montana extract, propylene glycol Aqua (and) rosa damascena extract CI 16035 (0,2% solution) Parfum PPG-1-PEG-9 Lauryl Glycol Ether
B C
Eumulgin L
Supplier
Brenntag Alban Muller Alban Muller Gazignaire
AMI
Quantity (%)
91.79 4.00 0.05 1.00 0.50 2.00 0.03 0.03 0.60
Table III O/W emulsion (cold emulsification)
Phases
Ingredients Commercial Name
INCI name
A
– Glycerin 99.5% Petroleum jelly oil Salcare SC 91
Aqua Glycerin Paraffinum liquidum Sodium acrylates copolymer (and) parraffinum liquidum (and) PPG-1 trideceth-6
B C
Supplier
Interchimie
Quantity (%)
88.75 2.00 8.00 1.25
Table IV O/W emulsion (hot emulsification)
Phases
Ingredients Commercial Name
INCI name
Supplier
Quantity (%)
A
/ Glycerin 99,5% EDETA BD Cutina CP Cuitna GMS-V Eumulgin B1 Eumulgin B2 Lanette 16 Fluid petroleum jelly Fucogel 1.5P Sepigel 305
Aqua Glycerin Disodium EDTA Cetyl palmitate Glyceryl stearate Ceteareth-12 Ceteareth-20 Cetyl alcohol Paraffinum liquidum Biosaccharide gum-1 Polyacrylamide (and) C13-14 isoparaffin (and) laureth-7
Interchimie BASF ChemTrade Prod Hyg Ami Prod Hyg Prod Hyg Ami Interchimie Solabia Group Seppic
84.6 2.00 0.10 1.45 3.45 1.50 1.50 2.45 2.00 0.20 0.75
B
C D
Table V Gel formulation
Phases
Ingredients Commercial Name
INCI name
A
– Glycerin 99.5% Carbopol ultrez 10 TEA 99%
Aqua Glycerin Carbomer Triethanolamine
B C
3-methyl-1,3-butanediol was supplied by Kuraray Europe GmbH (Germany) under commercial name Isopentyldiol, and natural butan-1,3-diol was supplied from Kokyu Alcohol Kogyo under commercial name Haisugarcane.
Supplier
Interchimie
Quantity (%)
96.87 2.00 0.53 0.60
Water activity measurement Aw was measured with a Novasina LabSwift-aw (Switzerland) in a 0.03 to 1.00 aw range with an accuracy of 0.010 aw in the
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
33
A. Kerdudo et al.
Optimization of cosmetic preservation
0.10–0.95 aw range. The 21.1 mL measuring chamber contained standardized sample dishes filled to the two-thirds with samples. Measurements were performed at room temperature and in triplicate. Apparatus calibration was performed with standard salts displaying relative humidity of 11%, 58%, 84% and 97%. Formulation processes All the formulations were prepared on a 100 g-basis final mass. Stirring was always performed with a propeller mixer (Turbotest laboratory mixer equipped with an axial flow propeller, Rayneri, VMI). Lotion formulation (Table II). The ingredients of phase A were first weighed separately. They were then mixed together and homogenized with the propeller mixer. Phase B was added under agitation at 600 rpm. Phase C was first homogenized with the propeller mixer before being incorporated in the mixture under stirring at 1000 rpm. The lotion contained 4% of butylene glycol which is the usual amount in commercialized product. Cold O/W fluid emulsion (Table III). Phase A was prepared under stirring at 1000 rpm. The emulsion was formed adding phase B under stirring at 1500 rpm. Phase C was then added to stabilize the emulsion under agitation at 2000 rpm. The glycerin content was fixed to 2% which is the usual amount in commercialized product. Hot O/W fluid emulsion (Table IV). Phase A and B were first prepared separately by mixing the ingredients under stirring with the propeller mixer (500–800 rpm). Then, they were separately heated at 75°C. This temperature was chosen to induce the melting of Phase B ingredients. Phase B (the lipophilic phase) was then incorporated in the hydrophilic phase (phase A) under stirring at 2000 rpm. Stirring was maintained and phase C was added at 30°C. Viscosity was then adjusted adding phase D. The glycerin
content was fixed to 2% which is the usual amount in commercialized product. Gel formulation (Table V). Ingredients of phase A were mixed together under agitation at 500 rpm and then heated at 55°C to favour polymer hydration. Phase B was dispersed in A under stirring at 1500 rpm. Phase C was then added for polymer hydration. The glycerin content was fixed to 2% which is the usual amount in commercialized product. O/W cream emulsion (Table VI). Phase A was heated at 55°C to favour polymer (Phase B) hydration. Phase B was dispersed in A under stirring at 1500 rpm. Stirring was maintained until complete polymer hydration. Phases A+B and C (all the phase C components were weighed together) were then separately heated at 75°C to get liquid phases. Emulsification was performed by addition of the phase C (lipophilic) in phases A+B (hydrophilic phase). Phases D and E were incorporated one after another at 60°C under stirring at 1500 rpm. Phases F, G, H and I were successively incorporated at 30°C under stirring at 1500 rpm. The glycerin content was fixed to 4% which is the usual amount in commercialized product. Challenge tests Antimicrobial preservative efficiency evaluation, realized by an external laboratory (Ideatest, Plouzane, France), was performed following European Pharmacopeia (Eur. Ph.). Five potential pathogenic germs were chosen for this study: Pseudomonas aeruginosa ATCC 9027, Staphylococcus aureus ATCC 6538, Escherichia coli ATCC 8739, Candida albicans ATCC 10231 and Aspergillus brasiliensis ATCC 16404. A contamination by these micro-organisms could cause illness such as skin infection or respiratory diseases. Tryptic soy agar and sabouraud dextrose was used for culture media for bacteria, moulds and yeasts, respectively. Cosmetic formulations were sampled in single-used sterile flasks, and every flask
Table VI Cream (O/W) formulation
Phases
Ingredients Commercial Name
INCI name
A
– Glycerin 99.5% Carbopol EDT 2020 Brij 72 Brij 721P Lanette 18 Cutina GMS-V Eutanol G Sophiderm Arlamol HD Virgin jojoba oil Refined wheat germ oil Refined shea butter Silicone oil baysilone M350 Eusolex 9020 Uvinul MC 80 Silicone oil SF1202 TEA 99% – D panthenol 50P Vitamin E acetate
Aqua Glycerin Acrylates/C10-30 alkyl acrylate cross-polymer Polyoxyethylene(2) (and) stearyl ether Steareth-21 (and) polyethoxylated alcohol Stearyl alcohol Glyceryl stearate Octyldodecanol Squalane Isohexadecane Simmondsia chinensis (Jojoba) seed oil Triticum vulgare (wheat) germ oil Butyrospermum parkii butter Dimethicone Butyl methoxydibenzoyl-methane Ethylhexyl methoxycinnamate Dimethicone Triethanolamine Isononyl isononanoate, titanium dioxide Panthenol (and) propylene glycol Tocopheryl Sodium hyaluronate Parfum
B C
D E F G H I
34
Supplier
Interchimie Gattefosse Croda France SAS Quimica masso Ami Ami Ami Sophim Quimica masso Olvea Olisud Interchimie Interchimie IES IES Brenntag specialties ^te d’Azur Brenntag Co – Laserson IES
Quantity (%)
59.34 4.00 0.15 2.00 2.50 1.50 3.00 3.00 2.50 3.00 1.50 1.50 2.00 1.50 1.00 1.50 2.00 0.11 1.25 0.20 0.50 5.85 0.10
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation Table VII European Pharmacopeia criteria (logarithmic reduction as a function of days D)
Aw evolution as a function of glycerin concentration 1.0
European Pharmacopeia
Logarithmic reduction
0.9 Criteria
D2
D7
D14
D28
Bacteria
A B A B
2 – – –
3 – – –
– 3 2 1
NI NI NI NI
Yeast and moulds
Table VIII Minimal water activity allowing micro-organisms growth, at 25°C [5,19,33]
Bacteria Pseudomonas aeruginosa Escherichia coli Staphylococcus aureus Yeast and fungi Aspergillus brasiliensis Candida albicans
Natural glycerin
0.7 0.6
y = –0.8094x2 + 0.0537x + 0.9654 R2 = 0.9978
0.5
NI, No Increase, –, no minimal reduction required.
Micro-organisms
Synthetic glycerin
0.8
aw
Strains
Minimal aw
0.97 0.95 0.86 0.77 0.87
was inoculated with strain suspension (at 22.5 2.5 °C in the dark). Inoculation concentration was fixed to 106 CFU mL1, 105 CFU mL1 and 104 CFU mL1, for bacteria, yeast and moulds, respectively. The validation of the neutralization of the preservative was realized on the 5 strains, in the 1/10th and 1/100th dilution, with LT100 broth. The microbial densities were counted by inclusion (results in CFU g1or mL) and compared with the logarithmic reduction criteria of the European Pharmacopeia (Table VII). DFT calculations The calculations were performed with the Gaussian 03 program [23]. The molecules structures were optimized, and energies were calculated by the DFT (Density Functional Theory) method with the exchange-correlation functional B3LYP [24–26] and the 631 + G(d,p) basis set for calculations. The effects of water as solvent were taken into account using the polarizable continuum model (PCM). Results and Discussion When measured, water activity (aw) values were systematically compared with the reference values, shown in Table VIII, as our aim was to reduce aw under these critical values to impede microorganisms growth. All raw materials having affinity with water molecules through hydrogen bonds, hydrophilic interactions or Van der Waals interactions are presumed to reduce aw: humectants, rheology modifiers, surfactants, fatty acids and their esters are recommended [27–29]. This study will focus on glycols and glycerin, two humectants, and to begin with on the influence of their concentration and their chemical structure on aw.
0.4 0%
20%
40%
60%
80%
100%
Glycerin concentration (wt %) Figure 1 Water activity as a function of synthetic glycerin (dark circles) concentration. Natural glycerin (gray squares) was compared at 16 wt%. Standard deviations are included in points.
Influence of humectants on aw Hydrogen bonds formation between glycerin or glycols and water molecules may reduce free water in finished product and thus water activity. In a first part of the study, glycerin was diluted in water and its impact on aw as a function of its concentration was examined. Glycerin impact on water activity: Concentration effect. Figure 1 shows water activity as a function of glycerin amount (from 4% to 80 wt %). The higher the glycerin concentration, the lower aw, as expected. 8% of glycerin was sufficient to reduce aw under 0.97, the minimal growth value for P. aeruginosa (Table VIII). With 16% of glycerin, aw was lower than 0.95, the minimal growth value for E. coli. Finally, more than 32% of glycerin was necessary to reduce other micro-organisms’ development. Let us note that such a huge amount is hardly conceivable in cosmetic formulation because of sensorial consequence (highly sticky products). Influence of glycerin source. The influence of a natural glycerin was also studied. The natural product was much more active than the synthetic one: 16 wt% of natural material was as efficient as 40 wt% of synthetic one, leading to 0.866 0.010 for aw value. This value was lower than the minimal aw allowing C. albicans growth (0.87). Such a difference in efficiency could be explained by the purity of the two sources of glycerin. The purity of the synthetic one was 99.5% (the remaining 0.5% being water), whereas that of the natural one, obtained from corn and soya fractionation, was 98%. Unknown residues from fractionation could explain our results. Glycols impact on aw: Influence of their chemical structure. Figure 2 shows the concentration dependency of water activity for two glycols differing in their chemical structure: butan-1,3-diol and 3methyl-1,3-butanediol. As expected, aw decreased with glycol concentration. However, 3-methyl-1,3-butanediol was more efficient than butan-1,3-diol, except for the highest concentration, 80 wt%. The additional methyl group could influence the stability of the hydrogen bond through an iceberg effect [30]. The additional methyl group decreases the polarity of the molecule so that ice
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
35
A. Kerdudo et al.
Optimization of cosmetic preservation
Aw evolution as a function of propan-1,2diol and propa-1,3-diol concentration
Aw evolution as a function of butan-1,3-diol and 3-methyl-1,3-butanediol 1
1.0
0.9
Propan1,2diol Propan1,3diol
0.8
0.7
aw
aw
0.9
y = –0.4773x2 – 0.0743x + 0.9736 R2 = 0.9974
0.8
y = –0.7255x2 + 0.024x + 0.9468 R2 = 0.9986
y = 0.0107x2 – 0.3244x + 0.9167 R2 = 0.9737
0.6
0.7
y = –0.3255x2 – 0.2852x + 0.9455 R2 = 0.9864
0.6
butan-1,3-diol
0.5
0.5
3-methyl-1,3-butanediol
0.4 0%
20%
40%
60%
80%
100%
0.4 0%
Glycols concentration (wt %)
organization of the water molecules in the first layer surrounding the molecule could be generated. This effect could also decrease water mobility and thus water activity [30]. The efficiency of 3methyl-1,3-butanediol was particularly pronounced at low concentration (8–16%): aw decreased to c. 0.85 for 16 wt%, value for which all micro-organisms, except A. niger, cannot grow (Table VIII). Two other glycols, differing also in their chemical structure, were compared. In this case, the difference came from the position of hydroxyl groups: propan-1,2-diol vs propan-1,3-diol. Figure 3 shows that they behaved similarly at extreme concentrations (≤5% and ≥65%). In between, propan-1,2-diol was more efficient than propan-1,3-diol. This suggests that the hydroxyl group position on a linear carbon chain can influence hydrogen bond formations. In that sense, propan-1,2-diol seems to be better stabilized in water than propan-1,3-diol. To confirm this assumption, molecular modelling was carried out. Molecular modelling of propanediols. The Density functional theory (DFT) is a very popular quantum chemistry method to investigate the electronic ground state of atoms and molecules. The underlying theory was developed by Hohenberg and Kohn [31] and later by Kohn, Sham and Pople who were awarded the Chemistry Nobel Prize in 1998.* Hohenberg and Kohn demonstrated that the electronic energy is a function of the electronic density, as well as being a function of the total many body wavefunction. Therefore, it reduces the number of variables in the Schr€odinger equation, energy being of function of the three-dimensional variables instead of 3N with N the number of electrons. The contribution of Kohn and Sham allowed treating interacting electrons as independent fermions moving in a mean field represented by other electrons and parameterized experimentally. This paradigm shift allows treating hundred of atoms and allows chemists to investigate molecules in solution as well as solid state chemistry with a reasonable computational cost. The DFT calculations, and widely quantum chemistry calculations, start by guessing the electronic density and then optimizing the density by following the so-called variational principle, ensuring to find the ground state correctly. To model solvation effects within the DFT method, the PCM method was developed by Tomasi
36
40%
60%
80%
100%
Glycols concentration (wt %) Figure 3 Water activity as a function of propan-1,2-diol (gray squares) and propan-1,3-diol (dark circles) concentration.
Propan-1,2-diol
Propan-1,3-diol
In water
Figure 2 Water activity as a function of butan-1,3-diol (dark circles) and 3-methyl-1,3-butanediol (gray circles) concentration.
20%
Figure 4 Molecular modelling of propan-1,2-diol and propan-1,3-diol in water.
and co-workers [32] and implemented in the GAUSSIAN software. This method replaces the solvent molecules by placing the solute in a cavity within the solvent reaction field, allowing the dielectric continuum to polarize the total electronic wavefunction. To confirm that propan-1,2-diol is better stabilized in water than propan-1,3-diol, the total energy (free enthalpy) of both molecules have been calculated in water and in vacuum as the dispersion or solubilization environment of a molecule influences its polarizability and therefore its geometry. The corresponding optimized structures are shown in Figs 4 and 5. Solvation energy was then calculated (see Eq. 1); the lower the energy, the more stable the molecule. Solvation energy corresponds to the solvent energy participating to the molecule stabilization. Esolv ¼ Em1ðwaterÞ Em1ðvacuumÞ
ð1Þ
where Esolv is the solvation energy, Em1(water) the total energy of molecule 1 in water and Em1(vacuum) the total energy of molecule 1 in vacuum. Modelling the propanediols in vacuum, we observed a modification of the propan-1,2-diol geometry in comparison with that obtained in water, whereas no change was observed for propan1,3-diol (Fig. 5). In vacuum, an intramolecular hydrogen bond was created between the hydrogen of hydroxyl group in C1 position and the oxygen of the hydroxyl group in C2 position. As
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation
Propan-1,2-diol
Propan-1,3-diol
Table X Description of the 5 compared protocols for water solution containing 16% (wt) of glycerin
In vacuum
Formulation protocols tested with a 16% (wt) water solution of glycerin
Protocol 1 Protocol 2
Figure 5 Molecular modelling of propan-1,2-diol and propan-1,3-diol in vacuum.
Protocol 3
Table IX Solvation energies of propan-1,2-diol and propan-1,3-diol Protocol 4 Esolv (kcal mol1) Protocol 5 14.82 13.52
this hydrogen bond formation increased the propan-1,2-diol stability, we could not consider the associated energy to calculate Esolv and compare it with propan-1,3-diol. Therefore, we calculated the energy of the molecule in vacuum considering that its geometry is that calculated in water. Solvation energies of molecules were then calculated as exposed in Table IX. DFT calculations demonstrated that propan-1,2-diol was better stabilized in water than propan1,3-diol confirming then our conclusion based on water activity measurement.
Protocol impact on water activity 0.96 0.94 0.92 0.90
aw
Propan-1,2-diol Propan-1,3-diol
Weigh water and glycerin Homogenize with propeller mixer at 500 rpm Weigh glycerin Add water drop by drop under stirring using a propeller mixer at 500 rpm Hold on stirring for 5 min Weigh water and glycerin Homogenize with propeller mixer Increase temperature at 75°C with water-bath Stir with propeller mixer for 10 min at 500 rpm Weigh water Add glycerin drop by drop under stirring with propeller mixer at 500 rpm Hold on stirring for 5 min Weigh water Add glycerin drop by drop under stirring with propeller mixer at 500 rpm Hold on stirring for 15 min
0.88
Influence of the preparation protocol on aw Literature reveals that processing method, processing rate, processing conditions, time interval between mixing and processing are part of critical factors influencing aw [20]. To have a further insight on the influence of processing on aw, we compared five protocols (Table X) on a simple, two-binary system: glycerin (16wt %)/water (84wt%). The modified parameters were the mixing process (P1/P2 and P1/P4), the order of both ingredients introduction (P2/P4), the mixing temperature (P1/P3) and the duration of stirring (P4/P5). These parameters were chosen as we believed they could play on the interactions between water molecules and glycerin. The mixture composition was chosen from previous section (1.a) because of its efficiency on aw and of its relative low glycerin amount compatible with cosmetic formulation constraints (too much glycerin would produce sticky products). As shown in Fig. 6, differences in water activity were obtained as a function of the applied protocol (P). In particular, the drop by drop addition of glycerin in water under mixing had a beneficial effect: a 7.5% and 6.5% decrease of aw was reached using P4 and P5, respectively, as compared to P1 (the reference protocol). This effect was less pronounced when water (P2), instead of glycerin (P4), was drop by drop added. In all cases, the successive addition of the raw materials under mixing (P2, P4 and P5) systematically decreases aw as compared with assays where both raw materials were mixed together (P1 and P3). Finally, increasing the time of mixing (P4/ P5), or heating during stirring when both components are initially mixed (P2/P1) slightly decreased aw.
0.86 0.84 0.82 P1
P2
P3
P4
P5
Figure 6 Water activity of glycerin solution (16% wt) as a function of applied protocol. All protocols were described in Table X.
To confirm that P4 had a positive impact on aw, other binary systems were studied: synthetic glycerin, natural glycerin 1, synthetic and natural butan-1,3-diol. Their concentration was fixed to 4 wt% as this is the usual polyol concentration in cosmetics. For all assays (Fig. 7), water activity was indeed reduced, from 1.2% to 2.3%, using P4 instead of P1. One can assume that drop by drop pouring of the polyol allows increasing the contact interface between water and polyol favouring interactions between them. Measurement of aw on cosmetic formulations Five cosmetic formulations were prepared following the protocol exposed in the material and method section (corresponding to protocol P1): a lotion, a gel, a cream and two oil-in-water emulsions. Their composition is given in Tables II–VI. The water activity of these formulations, measured 1 day after their preparation, is given in Table XI: all of them present a comparable aw value of c. 0.97.
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
37
A. Kerdudo et al.
Optimization of cosmetic preservation
Protocol impact on aw as a function of raw material
1.00
Aw of formulation as a function of protocol 0.98
Protocol 4
0.94
0.96
0.92
aw
0.94
aw
Reference
0.96
Protocol 1 Protocol 4
0.98
0.92
0.90 0.88
0.90
0.86
0.88
0.84
0.86
0.82 0.80
0.84
Glyc
erin Bu
,3tan-1
in 1 -diol lycer n-1,3 ral g buta l a r Natu Natu
diol
Figure 7 Water activity of glycol solutions as a function of applied protocol. All solutions contained 4% of corresponding glycol.
–3,8%
–2,8%
Lotion
Gel
–4,1%
O/W fluid emulsion (cold)
Figure 8 Water activity of cosmetic formulations as a function of applied protocol. The reference product was formulated following the protocol P1 described in the materials and methods section.
Table XI Water activity of the five references cosmetic products 1 day after formulation
Aw of formulation as a function of protocol 0.98 0.96
Data at day + 1
0.92
0.001 0.001 0.000 0.001 0.001
26.0 26.5 26.3 26.5 26.2
0.88 0.86
–3,8%
–4,8%
–5,4%
–4,3%
–6,6%
O/W cream emulsion
0.970 0.973 0.971 0.971 0.961
O/W fliud emulsion (hot)
Lotion O/W fluid emulsion (cold process) O/W fluid emulsion (hot process) Gel O/W cream
0.90
O/W fluid emulsion (cold)
T (°C)
Gel
std
Lotion
aw
aw
Formulation
0.94
0.84 0.82 0.80
Bold values corresponded to the water activity values of the cosmetic products
Based on the previous section, the protocol was slightly modified: the active ingredient of Phase A (glycerin or butylene glycol) was added drop by drop to water, following then P4 protocol. This led to a systematic decrease in aw (from 2.8% to 4.1%) for the three tested formulations (lotion, gel and cold emulsion, Fig. 8). In all case, aw was lower than 0.95 implying a better protection against P. aeruginosa and E. coli development. In a next step, all cosmetics were prepared following protocol 4 but with an increased glycol concentration as its positive action on aw was demonstrated above: 10% instead of 2 or 4%. The increase in glycol amount was then balanced by a decrease in water amount. Figure 9 predicts a better protection against P. aeruginosa and E. coli growth, especially for the cream for which aw was about 0.898 0.009 at 25°C, the only value lower than 0.90. However, this value is not low enough to expect a protection against S. aureus, yeast and mould (Table VIII). In a last formulation, the cream was prepared with 10% of natural glycerin instead of synthetic one, following protocol 4. Surprisingly, this had no effect on water activity (data not shown). Challenge test As the cream displayed the lowest aw value, it was chosen for performing a challenge test. The ‘reference’ cream (containing 4% of
38
Figure 9 Water activity of cosmetic formulations as a function of applied protocol. Deep gray: aw of references product. Light gray: aw of optimized product, formulated with protocol 4 and 10% of glycol.
glycerin and prepared following process P1) and the optimized one (prepared following P4 with 10% of natural glycerin) were compared. Challenge tests were performed for a 14-day period. As shown in Table XII, the optimized product was systematically more efficient to fight against micro-organism growth. Its antibacterial activity was remarkable as the logarithmic reduction of P. aeruginosa and S. aureus are in accordance with the criteria B of the Eur. Ph. for 14 days. C. albicans and A. brasiliensis growth in the optimized cream was less marked than in the reference product, but still remained too elevated for the Eur. Ph. regulation. The growth of C. albicans was only inhibited in the reference cream, whereas its population was reduced by 0.2 log in the optimized one. A. brasiliensis was still growing in the reference cream after 14 days, whereas a slight inhibition was noted in the optimized product. For all strains, the optimized cream was thus more efficient than the reference one. These data could be explained thanks to water activity reduction for bacterial growth protection. However, in the best cases, only
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
A. Kerdudo et al.
Optimization of cosmetic preservation
Table XII Logarithmic reduction by strains for the reference cream and the optimized one formulated following the protocol 4 with 10% of natural glycerin 1
Strains Days
Criteria A Criteria B Reference Interpretation aw optimized Interpretation
P. aeruginosa
S. aureus
E. coli
C. albicans
A. brasiliensis
D2
D7
D14
D2
D7
D14
D2
D7
D14
D2
D7
D14
D2
D7
D14
≥2 – 1.3 B 0.9 B
≥3 – >3 B >3 B
– ≥3 1.1 NC >3 B
≥2 – 0.2 B 1.1 B
≥3 – 0.0 B 2.0 B
– ≥3 0.2 NC >3 B
≥2 – 0.5 B 0.9 B
≥3 – 0.7 B 0.9 B
– ≥3 0.0 NC 1.1 NC
– – –
– – –
– – –
– – –
– –
– –
≥2 ≥1 0.0 NC 0.2 NC
– –
– –
≥2 ≥1 0.1 NC 0.1 NC
NC, not conform. DX, Day X.
the criteria B of the Eur. Ph.were fulfilled. This was not surprising as self preserved product with reduced water activity are not supposed to be bactericide or fungicide [19,20]. A low aw will only induce a slowdown in micro-organisms’ growth [20]. The lower the aw value, the longer the lag time before growth initiation. Low water activities may not be sufficient to kill micro-organisms. The survival of germs depends on other stresses imposed by the composition (lack of nutrients, membrane destabilizing surfactants or chelating agents) [19]. Eventually, it was not surprising that antibacterial activity was better than anti-fungal one as it is more difficult to protect a product against mould and yeast than against bacteria [5,19,33] (Table VIII).
esting way could be to enlarge DFT calculation to correlate water activity with physical data. Manufacturing process was also shown to impact water activity: a slow introduction of glycol in water under mixing improves water binding. Thinking about scaling up this process, we could imagine that glycols could be added as a dribble rather than drop by drop, reducing then time processing, and presumably, decreasing aw by comparison with a direct mixing. These results were applied in cosmetic formulations. A final 6.6% water activity reduction was obtained for the cream formulation. The challenge test showed an improved preservation of the optimized cream. Acknowledgements
Conclusion As a conclusion, significant impact of glycerin and glycols concentration on water activity was demonstrated. Differences were observed in function of raw material origin. Natural products were more efficient to reduce aw than synthetic ones, maybe because of lower purity. The hydroxyl group position had an impact on water binding. Water activity measurement and molecular modelling showed that propan-1,2-diol was better stabilized in water than propan-1,3-diol. Then, the supplementary methyl in C3 position of 3-methyl-1,3-butanediol increases water binding ability in comparison with butane-1,3-diol, except at high concentration. An inter-
The authors are grateful to SO.F.I.A. Cosmetiques and Association Nationale de la Recherche et de la Technologie (ANRT) for financial support of this project. The authors wish to thank FEDER and PACA area for financial support. This project was supported by the University of Nice-Sophia Antipolis, the Bordeaux Polytechnic National Institute, and the CNRS.
ENDNOTE * Nobel Prize in chemistry 1998.
References 1. Dweck, A. Natural preservative – an update. In Formulating Natural Cosmetics (Kozlowski, A.C., ed.), pp. 107–130. Alluredbooks, USA (2011). 2. Amaral, L.F.B., Camilo, N.S., Pereda, M.D.C.V., Levy, C.E., Moriel, P. and Mazzola, P.G. Evaluation of antimicrobial effectiveness of C-8 xylitol monoester as an alternative preservative for cosmetic products. Int. J. Cosm. Sci. 33, 391–397 (2011). 3. EU regulation, Cosmetic products regulation (1223/2009). 4. Lundov, M.D., Johansen, J.D., Zachariae, C. and Moesby, L. Low-level efficacy of cosmetic preservatives. Int. J. Cosm. Sci. 33, 190–196 (2011).
5. Fernandez, X., Merck, F. and Kerdudo, A. Conservateurs pour cosmetiques – Generalites et conservateurs antimicrobiens. Techniques de l’Ingenieur, J2284, 1–21 (2012). 6. Dardre, P.D., Aljarrah, A., Miller, W.R., Coldham, N.G., Sauer, M.J. and Pope, G.S. Concentrations of parabens in human breast tumours. J. Appl. Toxicol. 24, 5–13 (2004). 7. Hammer, K. A, C.F. Carson and Riley, T.V. Antimicrobial activity of essential oils and other plant extracts. J. Appl. Microbiol. 86, 985–990 (1999). 8. Kunicka-Styczynska, A., Sikora, M. and KaLembz, D. Antimicrobial activity of lavender, tea tree and lemon oils in cosmetic preser-
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
9.
10.
11. 12.
vative systems. J. Appl. Microbiol. 107, 1903–1911 (2009). Baratta, M.T., Dorman, H.J.D., Deans, S.G., Figueiredo, A.C., Barroso, J.G. and Ruberto, G. Antimicrobial and antioxidant properties of some commercial essential oils. Flavour Fragr. J. 13, 235–244 (1998). Kabara, J.J. Aroma chemicals as preservatives. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice, (Kabara, J.J. and Orth, D.S. ed.), pp. 181– 208. Marcel Dekker, Phoenix (1996). Lopez, D. and Lopez, C., Cosmetic product sterilization, WO 2007/148022 A2. Oule, K.M., Dickman, M. and Arul, J. Microbicidal effect of pressurized CO2 and the
39
A. Kerdudo et al.
Optimization of cosmetic preservation
13.
14.
15.
16.
17.
18.
40
influence of sensitizing additives. Eur. J. Sci. Res. 41, 569–581 (2010). Spilimbergo, S., Mantoan, D., Quaranta, A. and Della-Mea, G. Real-time monitoring of cell membrane modification during supercritical CO2 pasteurization. J. Supercrit. Fluids 48, 93–97 (2009). White, A., Burns, D. and Christensen, T.W. Effective terminal sterilization using supercritical carbon dioxide. J. Biotechnol. 123, 504–515 (2006). Brannan, D.K. The role of packaging in product preservation. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice, (Kabara, J.J. and Orth, D.S., ed.), pp. 227–242, Marcel Dekker, Phoenix (1996). Fernandez, X., Merck, F. and Kerdudo, A. Conservateurs pour cosmetiques – Antioxydants et Anti-UV. Techniques de l’ingenieur, J2285, 1–19 (2012). Berthele, H., Sella, O., Lavarde, M., Mielcarek, C., Pense-Lheritier, A.-M. and Pirnay, S. Determination of the influence of factors (ethanol, pH and aw) on the preservation of cosmetics using experimental design. Int. J. Cosm. Sci. 36, 54–61 (2014). Faiveley, M. L’eau et la conservation des aliments. Notions de biochimie alimentaire et alimentation humaine. Techniques de l’Ingenieur, F1011, 1–20 (2003).
19. Kabara, J.J. Principles for Product Preservation. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice (Kabara, J.J. and Orth, D.S., ed.), pp. 1–14. Marcel Dekker, Phoenix (1996). 20. Enigl, D.C. and Sorrells, K.M. Water Activity and Self-Preserving Formulas. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice, (Kabara, J.J. and Orth, D.S., ed.), pp. 45–73. Marcel Dekker, Phoenix (1996). 21. NF EN ISO 29621 Cosmetiques - Microbiologie - Lignes directrices pour l’appreciation du risque et l’identification de produits a faible risque microbiologique (2011) 22. Varvaresou, A., Papageorgiou, S., Tsirivas, E., Protopapa, E., Kintziou, H., Kefala, V. and Demetzos, C. Self-preserving cosmetics. Int. J. Cosm. Sci. 31, 163–175 (2009). 23. Frisch, M. J., Trucks, G. W., Schlegel, H. B. et al. Gaussian03, Revision C.02, Gaussian, Inc., Wallingford CT (2004). 24. Becke, A.D. Density-functional exchangeenergy approximation with correct asymptotic behavior. Phys. Rev. A 38, 3098–3100 (1988). 25. Becke, A.D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 98, 5648–5652 (1993). 26. Lee, C., Yang, W. and Parr, R.G. Development of the Colle-Salvetti correlation-energy
27.
28.
29.
30.
31.
32.
33.
formula into a functional of the electron density. Phys. Rev. B 37, 785–789 (1988). Bergasson, G., Arnfinnsson, J., Steingrimmson, O. and Thormar, H. In vitro killing of Candida albicans by fatty acids and monoglycerides. Antimicrob. Agents Chemother. 45, 3209–3212 (2001). Kabara, J.J. Fatty acids and esters as multifunctional components. In Preservative-Free and Self-Preserving Cosmetics and Drugs: Principles and Practice. (Kabara, J.J. and Orth, D.S., ed.), pp. 119–137. Marcel Dekker, Phoenix (1996). Ingram, L.O. Mechanism of Lysis of Escherichia coli by Ethanol and Other Chaotropic Agents. J. Bacteriol. 146, 331–336 (1981). Dill, K.A., Truskett, T.M., Vlachy, V. and Hribar-Lee, B. Modeling water, the hydrophobic effect and ion solvation. Annu. Rev. Biophys. Biomol. Struct. 34, 173–199 (2005). Hohenberg, P. and Kohn, W. Inhomogeneous Electron Gas. Phys. Rev. 136, B864– B871 (1964). Tomasi, J., Mennucci, B. and Cances, E. The IEF version of the PCM solvation method: an overview of a new method addressed to study molecular solutes at the QM ab initio level. J. Mol. Struct. (Theochem) 464, 211– 226 (1999). Orth, D.S. Handbook of cosmetic Microbiology. pp. 1–108, Marcel Dekker, New York (1993)
© 2014 Society of Cosmetic Scientists and the Societe Francßaise de Cosmetologie International Journal of Cosmetic Science, 37, 31–40
