Biocompatible Microemulsions Based on Limonene: Formulation

Feb 28, 2008 - The preparation of biocompatible (w/o) microemulsions based on R-(+)-limonene, water, and a mixture of lecithin and either 1-propanol o...
5 downloads 0 Views 75KB Size
Download PDF
3380

Langmuir 2008, 24, 3380-3386

Biocompatible Microemulsions Based on Limonene: Formulation, Structure, and Applications Vassiliki Papadimitriou,*,† Stergios Pispas,‡ Stauroula Syriou,† Anastasia Pournara,† Maria Zoumpanioti,† Theodore G. Sotiroudis,† and Aristotelis Xenakis† Institute of Biological Research & Biotechnology, Theoretical and Physical Chemistry Institute, National Hellenic Research Foundation, 48, Vassileos Constantinou AVenue, 11635, Athens, Greece ReceiVed NoVember 26, 2007. In Final Form: January 9, 2008 The preparation of biocompatible (w/o) microemulsions based on R-(+)-limonene, water, and a mixture of lecithin and either 1-propanol or 1,2-propanediol as emulsifiers was considered. The choice of the compositions of the microemulsions used was based on the pseudo-ternary phase diagrams of the four-component system determined at 30 °C for different weight ratios of the components. When 1-propanol was considered as co-surfactant, the area of the microemulsion zone was remarkably increased. Interfacial properties and the dynamic structure of the emulsifier’s monolayer were studied by electron paramagnetic resonance (EPR) spectroscopy using the spin-labeling technique. The rigidity and polarity of the interface were affected by the nature of the alcohol used as co-surfactant. When 1-propanol was used, the emulsifier’s interface was much more flexible, indicating a less tight packing of lecithin molecules than in the case of 1,2-propanediol. In addition, the membrane’s polarity was decreased when the diol was added as co-surfactant in the microemulsion system. To evaluate the size of the dispersed aqueous domains as a function of water content and other additives concentration, dynamic light scattering (DLS) measurements were carried out. Radii in the range from 60 to 180 nm were observed when 1-propanol was used as co-surfactant, and the water content varied from 0 to 12% w/w. Electrical conductivity measurements of R-(+)-limonene/lecithin/1-propanol/water microemulsions with increasing weight fractions of water indicated the appearance of a percolation threshold at water content above 4% w/w. Lipase from Rhizomucor miehei was solubilized in the aqueous domains of the biocompatible microemulsions, and the esterification of octanoic, dodecanoic, and hexadecanoic acids with the short-chained alcohols used as co-surfactants for the formulation of microemulsions was studied. The enzyme efficiency was affected by the chain length of the carboxylic acids and the nature of the alcohol. In the case of 1-propanol, a preference for the long-chain carboxylic acids was observed. On the contrary, when 1,2-propanediol was used formulation of the corresponding esters was not observed. This behavior could be possibly attributed to either the specificity of the lipase toward the alcohol employed for the esterification of the acids or the structural changes induced in the system when 1-propanol was replaced by 1,2-propanediol.

Introduction The potential application of highly biocompatible water-inoil (w/o) microemulsions to the food, cosmetic, and pharmaceutical industry as solubilization media of hydrophilic, hydrophobic, and amphiphilic functional materials has been of growing interest during the past few years.1-4 The potential technical and commercial applications of microemulsions are mainly linked to their unique properties such as thermodynamic stability, optical clarity, and high solubilization capacity. However, the most critical problem regarding the use of microemulsions in the food, cosmetic, and pharmaceutical fields is the toxicity of their partial components. Formulation and characterization of nontoxic microemulsion formulations based on biological amphiphiles and different oils has been studied for over a decade.5-7 In the present study we suggest the preparation of biocompatible microemulsions based on R-(+)-limonene as the nonpolar solvent, * To whom correspondence should be addressed. Phone: +302107273736. Fax: +302107273758. E-mail: [email protected]. † Institute of Biological Research & Biotechnology. ‡ Theoretical and Physical Chemistry Institute. (1) Lawrence, M. J.; Rees, G. D. AdV. Drug DeliVery ReV. 2000, 45, 89-121. (2) Garti, N. Curr. Opin. Colloid Interface Sci. 2003, 8, 197-211. (3) Flanagan, J.; Singh, H. Crit. ReV. Food Sci. Nutr. 2006, 46, 221-237. (4) Leser, M.; Sagalowicz, L.; Michel, M.; Watzke, H. AdV. Colloid Interface Sci. 2006, 123-126, 125-136. (5) Shinoda, K.; Araki, M.; Sadaghiani, A.; Khan, A.; Lindman, B. J. Phys. Chem. 1991, 95, 989- 993. (6) Kahlweit, M.; Busse, G.; Faulhaber, B. Langmuir 1995, 11, 1576-1583. (7) Kahlweit, M.; Busse, G.; Faulhaber, B.; Eibl, H. Langmuir 1995, 11, 41854187.

water as the dispersed phase, and a mixture of lecithin and either 1-propanol or 1,2-propanediol as emulsifiers. R-(+)-Limonene, the main constituent of peel oil from citrus fruits, is an inexpensive monoterpenic hydrocarbon produced from a renewable source and widely available on the market. The distilled peel oil contains 95% R-(+)-limonene together with other terpenes, mainly myrcene. R-(+)-Limonene is one of the most commonly used fragrance materials in the production of fine fragrances and also many kinds of technical products. R-(+)-Limonene is also used as a solvent and industrial defatting agent.8 Soybean lecithin is a combination of naturally occurring phospholipids, which are extracted during the processing of soybean oil. During recent years, lecithin has been successfully used for the construction of various nontoxic microemulsion formulations.5,6,9,10 Because of lecithin’s high lipophilicity and strong tendency to form liquid crystalline structures, addition of co-surfactants such as short-chained alcohols was necessary in order to formulate (w/o) microemulsions.5,11 The present work mainly focuses on the formulation and structural characterization of (w/o) microemulsions based on R-(+)-limonene and nontoxic emulsifiers. Although preparation (8) Matura, M.; Goossens, A.; Bordalo, O.; Garcia-Bravo, B.; Magnusson, K.; Wrangsjo¨, K.; Karlberg, A. T. J. Am. Acad. Dermatol. 2002, 47, 709-714. (9) Avramiotis, S.; Lianos, P.; Xenakis, A. Biocatal. Biotransform. 1997, 14, 299-316. (10) Papadimitriou, V.; Sotiroudis, T.; Xenakis, A. Langmuir 2007, 23, 20712077. (11) Schurtenberger, P.; Peng, Q.; Leser, M. E.; Luisi, P. L. J. Colloid Interface Sci. 1993, 156, 43-51.

10.1021/la703682c CCC: $40.75 © 2008 American Chemical Society Published on Web 02/28/2008

Biocompatible Microemulsions Based on Limonene

of microemulsions is relatively easy, structural characterization at both the macroscopic and the microscopic levels remains difficult since application of a wide range of different techniques which act in a complementary way is generally required. Nevertheless, structural evaluation of microemulsions is essential for their successful commercial exploitation.1 In the present study, structural characteristics of R-(+)-limonene-based (w/o) microemulsions were evaluated by applying two direct techniques, a spectroscopic technique (electron paramagnetic resonance spectroscopy) and a scattering technique (dynamic light scattering), as well as an indirect technique based on electrical conductivity measurements. Interfacial properties and the dynamic structure of lecithin/ alcohol layers in biocompatible (w/o) microemulsions were studied by electron paramagnetic resonance (EPR) spectroscopy using the spin-labeling technique.10 The spin-labeled fatty acid 5-doxyl stearic acid (5-DSA) is a long amphiphilic molecule having the tendency to align with the surfactant molecules. EPR spectra of the interfacially located spin probe reflect the mobility of the probe and the rigidity of its environment. The mobility of the spin probe in the membrane is reflected by the rotational correlation time, τR, while the rigidity of the surfactants monolayer is expressed quantitatively by the order parameter S value.12,13 Dynamic light scattering (DLS) is a well-established scattering technique for studying self-organizing amphiphilic systems like microemulsions.14,15 This technique can provide valuable information on the size, shape, and diffusion dynamics of the dispersed aqueous domains in (w/o) microemulsions. In the present study DLS was applied to evaluate the size of water droplets in the microemulsions as a function of water content and other additives concentration. A well-known feature of (w/o) microemulsions is the sharp increase of several orders of magnitude in their electrical conductivity as the water volume fraction is increased in the oil-rich mixture.16 In the present study, the conductivity of R-(+)limonene-based (w/o) microemulsions formulated with lecithin and different co-surfactants was investigated at increasing water contents. The existence of a percolation threshold indicated the structural transition of the system in the bicontinuous form. During the last decades extensive research has been conducted on the potential biotechnological applications of microemulsions taking advantage of the microemulsions as microreactors for the effectuation of several enzymatic reactions.17,18 Lipases, a class of hydrolytic enzymes activated at the interface of hydrophobic and hydrophilic domains, are the most often studied enzymes in microemulsions since they have shown very good stability and activity in these systems.19 Lipases when hosted in reverse micelles can catalyze reactions that are not favored in aqueous media. Several products of high added value can be thus produced. In the present study, the catalytic behavior of lipase from Rhizomucor miehei in R-(+)-limonene microemulsions was (12) Kommaredi, N. S.; O’Connor, K. C.; John, V. T. Biotechnol. Bioeng. 1993, 43, 215. (13) Griffith, O. H.; Jost, P. C. Lipid Spin Labels in Biological Membrane. In Spin Labeling, Theory and Applications; Berliner, L. J., Ed.; Academic Press: New York, 1976; pp 454-484. (14) Rouch, J.; Safouane, A.; Tartaglia, P.; Chen, S. H. J. Chem. Phys. 1989, 90, 3756-3764. (15) Eastoe, J.; Young, W. K.; Robinson, B. H.; Steytler, D. C. J. Chem. Soc., Faraday Trans. 1990, 86, 2883-2889. (16) Lagues, M.; Sauterey, C. J. Phys.Chem. 1980, 84, 3503-3508. (17) Orlich, B.; Scoma¨cher, R. AdV. Biochem. Eng. Biotechnol. 2002, 75, 185-208. (18) Fadnavis, N. W.; Deshpande, A. Curr. Org. Chem. 2002, 6, 393-410. (19) Stamatis, H.; Xenakis, A.; Kolisis, F. N. Biotechnol. AdV. 1999, 17, 293318.

Langmuir, Vol. 24, No. 7, 2008 3381

Figure 1. Pseudo-ternary phase diagram for the four-component system R-(+)-limonene/lecithin/alcohol/water at constant weight ratio of lecithin/alcohol (1:1): (0) 1-propanol and (9) 1,2-propanediol. Compositions are in weight ratios. The temperature was 30 °C.

investigated. Synthesis of different fatty acid esters with potential application in the food industry was studied. Experimental Section Materials. R-(+)-Limonene 97% was purchased from SigmaAldrich, Germany, and Alfa Aesar, USA. Soybean lecithin (Epikuron 200) containing 96% phosphatidylcholine was purchased from Lucas Meyer Hamburg, Germany. 1,2-Propanediol was purchased from Sigma-Aldrich, Germany. 1-Propanol was purchased from Merck, Darmstadt, Germany. 5-Doxyl stearic acid [5-(1-oxyl-2,2-dimethyloxazolidin) stearic acid] was obtained from Sigma-Aldrich, Germany. Lipase from R. miehei was supplied by Fluka. The enzyme preparation had a specific activity of 2.1 U/mg of protein (1 U corresponds to the amount of enzyme which liberates 1 µmol of butyric acid per minute at pH 7.5 and 40 °C using tributyrin as substrate). The enzyme was solubilized in 0.2 M Tris/HCl buffer, pH 7.5, and stored frozen. Octanoic acid (caprylic acid) was obtained from Merck, dodecanoic acid (lauric acid) from Sigma, and hexadecanoic acid (palmitic acid) from MP Biomedicals. High-purity water was obtained by a Millipore Milli Q plus water purification system. Phase Diagrams. Four-component systems consisting of oil/ surfactant/co-surfactant and water can be described with pseudoternary phase diagrams. These diagrams were prepared as follows. A mixture of emulsifiers containing equal quantities of lecithin as surfactant and either 1-propanol or 1,2-propanediol as co-surfactants was blended with limonene (oil phase) at determined weight ratios. The resulting mixture was titrated with water until it turned turbid. The volume of water added was recorded. The procedure was repeated several times at different weight ratios of limonene to emulsifiers. The temperature was kept constant at 30 °C. The pseudo-ternary phase diagram was constructed by plotting the amounts of oil, surfactant/co-surfactant, and water (including the traces of water present in limonene and lecithin, as determined by Karl Fischer titrations) used in the experiment. The microemulsion region (monophasic area) was identified as shown in Figure 1. Preparation of Microemulsions for Structural Studies. The composition of the (w/o) microemulsions used for structural studies were chosen to correspond to the monophasic area of the pseudoternary phase diagrams determined at 30 °C (Figure 1). A typical microemulsion was prepared by mixing R-(+)-limonene with a mixture of emulsifiers, either (1:1) lecithin/1-propanol or (1:1) lecithin/1,2-propanediol, at predetermined weight ratios. Then appropriate amounts of water or aqueous buffer solution were added to obtain a clear isotropic reverse micellar solution. Preparation of Microemulsions for Enzymatic Studies. A typical microemulsion for enzymatic studies was prepared by mixing R-(+)-limonene with a mixture of emulsifiers, either (1:1) lecithin/ 1-propanol or (1:1) lecithin/1,2-propanediol, at predetermined weight

3382 Langmuir, Vol. 24, No. 7, 2008

Papadimitriou et al.

ratios. Then the appropriate fatty acid was added in the above limonene solution to achieve a final concentration of 85 mM. The esterification reaction was started by adding 10 µL of an aqueous buffer solution (0.2 M Tris/HCl buffer, pH 7.5) containing the enzyme. The final enzyme concentration in the reaction medium was 83 µg/mL. Esterification of either 1-propanol (2.8 M) or 1,2propanediol (2.3 M) with three different fatty acids, namely, octanoic, dodecanoic, and hexadecanoic acids catalyzed by lipase from R. miehei entrapped in the above-mentioned microemulsions, was investigated. At fixed time intervals samples of 10 µL each were taken before less than 8% of the substrate was consumed and analyzed by GC.19,20 The low conversions were used to minimize inhibition by the water that is produced by the esterification reaction. EPR Measurements. To obtain the desired concentration of the lipid spin label, 5-doxyl stearic acid (5-DSA), in the microemulsions, 1 g of each microemulsion was added to a tube into which the appropriate amount of 5-DSA had been deposited previously. This was done by placing 10 µL of a stock 5-DSA solution in ethanol (7.8 × 10-3 M) in the tube and by further evaporating the ethanol. EPR measurements were carried out at room temperature using a Bruker ER 200D spectrometer operating at the X-band. The spectrometer was equipped with a Double Rectangular Cavity ER 4105 DR, and samples were taken up in 734-PQ-8, thin wall suprasil, EPR sample tubes (Wilmad Glass Co., Buena, NJ). Typical settings were as follows: center field, 3460 G; scan width, 100 G; time constant, 500 ms; microwave power, 2.4 mW; microwave frequency, 9.81 GHz; modulation amplitude, 1 G; gain, 20 000; phase 90°. Data collection was performed using the computerized program DAT-200 (Data Acquisition Program, University Lubeck, Germany) and analyzed with the GEP (Graphic Evaluation Program version 1.2) program for a personal computer. Interpretation of the EPR Data. In the present study, EPR spectra of 5-DSA were analyzed to provide information regarding molecular motion, lipid order, and polarity across the surfactants monolayer. For this purpose, experimental results were analyzed in terms of rotational correlation time, τR, order parameter, S, and isotropic hyperfine splitting constant, R′0, of the spin probe. The order parameter S provides a measure of the spin probe’s arrangement in a supramolecular assembly and varies from 0 to 1, with S ) 1 for the completely ordered state and S ) 0 for the completely random state. The rotational correlation time, τR, is relevant to the spin probe’s rotational motion. We must say that for the fast motion region (τR < 3 × 10-9s) the τR values are more accurate since they depict precisely the spin probe’s molecular motion. For the slow motion region (τR > 3 × 10-9 s) determination of τR values is difficult and inaccurate. In the last case the order parameter S is a much better approach to the immobilization degree. The hyperfine splitting constant values, R′0, are sensitive to the polarity of the environment of the spin probe and increase when the polarity of the medium is increased. Calculation of the Rotational Correlation Time, τR. The rotational correlation time, τR, was calculated by the following relationship12 τR ) 6 × 10-10[(h0/h+1)1/2 + (h0/h-1)1/2 -2]∆H0

(1)

where ∆H0 is the width (in Gauss) of the central peak and h+1, h0, and h-1 are the intensities of the low, center, and high field peaks, respectively (Figure 2). The relationship of eq 1 is applicable in the fast motion region, i.e., τR < 3 × 10-9 s. Calculation of the Order Parameter, S. From the spectral characteristics we calculated two more parameters: the order parameter, S, and the isotropic hyperfine splitting constant, R′0. The order parameter, S, is defined as13 S ) (A| - A⊥)/[AZZ - 1/2(AXX + AYY)](R0/R′0)

(2)

where AXX, AYY, and AZZ are the single-crystal values of the spin probe equal to 6.3, 5.8, and 33.6 G, respectively, and are indicative for doxyl derivatives. A|| corresponds to the half distance of the (20) Stamatis, H.; Xenakis, A. J. Mol. Catal. B: Enzym. 1999, 6 399-406.

Figure 2. EPR spectra of 5-DSA in (w/o) microemulsions of (A) R-(+)-limonene/lecithin-1-propanol/water and (B) R-(+)-limonene/ lecithin-1,2-propanediol/water. The weight ratio of limonene to emulsifiers was 1.5 and of lecithin to alcohol was 1. The water content was 1.7% w/w in both systems. [5-DSA] ) 7.8 × 10-5 M. outer maximum hyperfine splitting, Amax (Figure 2). A⊥ is calculated from the following: A⊥ ) Amin + 1.4(1 - Sapp) and Sapp ) (Amax Amin)/[AZZ - 1/2(AXX + AYY)], where Amin is equal to the half distance of the inner minimum hyperfine splitting (Figure 2). The ratio R0/R′0 is the polarity correction factor (hyperfine splitting constants), where R0 ) 1/3(Azz + Axx + Ayy) and R′0 ) 1/3(A|+ 2A⊥). R0 is the isotropic hyperfine splitting constant for the nitroxide molecule in the crystal state, and R′0 is the isotropic hyperfine splitting constant for the spin probe in the membrane. Dynamic Light Scattering Measurements. Dynamic light scattering (DLS) measurements were performed in the angular range from 20° to 150° by a ALV/CGS-3 Compact Goniometer System (ALV GmbH, Germany) using a JDS Uniphase 22mW He-Ne laser, operating at 632.8 nm, and an avalanche photodiode detector, interfaced with a ALV-5000/EPP multi-tau digital correlator with 288 channels and a ALV/LSE-5003 light scattering electronics unit for stepper motor drive and limit switch control. Autocorrelation functions, from DLS measurements, were collected five times at each observation angle for each sample, and they were analyzed by the cumulants method and the CONTIN routine. Fits to the correlation functions were made using the software provided by the manufacturers. Apparent hydrodynamic radii, Rh, at different water concentrations were calculated by aid of the Stokes-Einstein equation Rh ) kT/6πηoDapp

(3)

where k is the Boltzmann constant, T the absolute temperature, ηo the solvent viscosity, and Dapp the diffusion coefficient calculated from the analysis of the correlation functions. Microemulsions were filtered through 0.45 µm hydrophilic PTFE filters (Millex-LCR from Millipore) before light scattering measurements. Conductivity Measurements. The conductivity measurements were performed with a Metrohm 644 conductometer using a thermostated microcell (25 °C). The cell constant, c, was equal to 0.1 cm-1. R-(+)-Limonene-based (w/o) microemulsions were prepared as described above using lecithin as surfactant and either 1-propanol or 1,2-propanediol as co-surfactant. The weight ratio of limonene to emulsifiers was equal to 1.5 and of lecithin to alcohol equal to 1 throughout the experiment. Increasing amounts of a Tris/ HCl (0.2 M) buffer solution were added to the system to obtain the desired final aqueous contents.

Results and Discussion Phase Diagrams. Phase diagrams of the four-component systems, (A) R-(+)-limonene/lecithin/1-propanol/water and (B) R-(+)-limonene/lecithin/1,2-propanediol/water, were constructed

Biocompatible Microemulsions Based on Limonene

to determine the extent of the monophasic area that corresponds to (w/o) microemulsions. As it can be observed from Figure 1, when 1,2-propanediol was used as co-surfactant, (w/o) microemulsions can be obtained within a rather narrow range of concentrations. At lecithin/1,2-propanediol concentration lower than 20% w/w, formulation of (w/o) microemulsions was not observed since water incorporation in the system was not possible. At the same time, at lecithin/1,2-propanediol concentration higher than 70% w/w formulation of gels was observed upon addition of small amounts of water. Most of these gels retain their structure after heating for 24 h at 30 °C in a water bath. The same phase diagram also shows that maximum water incorporation in the system was achieved at 32% w/w limonene and 68% w/w of the emulsifiers’ mixture. When 1-propanol was used as co-surfactant, the area of the microemulsion zone was remarkably increased. As it can be seen from Figure 1, water incorporation in the system starts from lecithin/1-propanol concentration 5% w/w and continues over the whole range of concentrations. In that case, formulation of gels was not observed. Although 1,2-propanediol is expected to be more soluble in the organic continuous phase than 1-propanol as verified experimentally by the increased viscosity measured in the former case (η ) 152.9 cp in comparison to η ) 2.76 cp in the case of 1-propanol), its participation in the interface cannot be negligible. It is well known that formulation of reverse micelles is not possible when lecithin is considered as surfactant unless a short chain alcohol is added in the system.5,11 EPR Studies. EPR spectroscopy using the spin-labeling technique was undertaken to study the interfacial properties of the surfactant monolayer in (w/o) microemulsions based on limonene. The spin-labeled fatty acid 5-DSA is a long amphiphilic molecule having a tendency to align with the surfactant molecules of the interface. In the present study, the surfactant monolayer consists of a mixture of emulsifiers, either lecithin and 1-propanol or lecithin and 1,2-propanediol. By calculating the rotational correlation times, τR, and order parameters, S, of 5-DSA when located in the interface, the mobility of the probe and rigidity of the interface can be estimated in both systems. Figure 2 shows the EPR spectra of 5-DSA in two different (w/o) microemulsions formulated with either 1-propanol or 1,2propanediol at constant water content (1.7% w/w) and constant limonene to emulsifiers’ weight ratio (1.5). In both cases a threeline EPR spectrum, characteristic of nitroxides, was obtained although small changes in spectral characteristics could be observed depending on the nature of the alcohol. When 1-propanol was considered as co-surfactant (A), the peaks were narrower than in the case of 1,2-propanediol (B). At the same time, the signal intensity ratios of the center to high-field peak (h0/h-1) and center to low-field peak (h0/h+1) were higher in the case of 1,2-propanediol than in the case of 1-propanol. Both observations are indicative of a more restrictive movement of the spin probe on the surfactants monolayer when 1,2-propanediol was used as co-surfactant instead of 1-propanol. To investigate the effect of the water content on the dynamics of the micellar interface of (w/o) microemulsions based on limonene, both rotational correlation time and order parameter of 5-DSA were calculated. Table 1 shows the rotational correlation times, τR, of 5-DSA in biocompatible microemulsions based on (A) R-(+)-limonene/lecithin/1-propanol/water and (B) R-(+)limonene/lecithin/1,2-propanediol/water for various water contents. In the case of R-(+)-limonene/lecithin/1-propanol/water microemulsions, as the water content of the system increases from 0.5% to 7.9% w/w, the τR values of 5-DSA slightly decrease from 0.62 to 0.50 ns (Table 1). The observed decrease of the rotational correlation times with water content is indicative of

Langmuir, Vol. 24, No. 7, 2008 3383 Table 1. Rotational Correlation Times, τR, and Order Parameters, S, of 5-DSA in (w/o) Microemulsions of (A) R-(+)-Limonene/Lecithin/1-Propanol/Water and (B) R-(+)-Limonene/Lecithin/1,2-Propanediol/Water at Different Water Contentsa system

water (% w/w)

τR (ns)

S

(A) R-(+)-limonene/lecithin/ 1-propanol

0.5 0.9 1.3 1.7 2.5 4.3 6.1 7.9

0.62 0.62 0.61 0.60 0.61 0.55 0.48 0.50

0.10 0.09 0.08 0.08 0.10 0.09 0.10 0.09

(B) R-(+)-limonene/lecithin/ 1,2-propanediol

0.5 0.8 1.1 1.4 1.7 1.9 2.2

2.45 2.33 2.26 2.23 2.19 2.20 2.12

0.20 0.18 0.18 0.17 0.16 0.16 0.17

a The weight ratio of limonene to emulsifiers was 1.5 and of lecithin to alcohol was 1. [5-DSA])7.8 × 10-5 M.

a faster movement of the probe in the membrane. Nevertheless, the above observation clearly shows that the mobility of the spin probe in the interface is not so much affected by the water content of the system when 1-propanol was used as co-surfactant. When (w/o) microemulsions based on R-(+)-limonene/lecithin/1,2propanediol/water were considered the calculated rotational correlation times of 5-DSA were remarkably higher, indicating a much slower movement of the probe in the interface. As the water content of the system increases from 0.5% to 2.2% w/w, the τR values of 5-DSA decrease from 2.45 to 2.12 ns. Such a behavior clearly shows an increase of the spin probe’s mobility by increasing the water content of the microemulsions. In every case the rotational correlation time values are in the fast motion region (τR< 3 × 10-9 s) which indicates that the spin probe tumbles free at all directions. Similar behavior was observed when olive oil microemulsions based on lecithin were considered and also in the case of lecithin organogels containing tiny amounts of water.10,21 In both cases, increasing the water quantity induced an increase of the polar head area per lecithin molecule and consequently an expansion of the total interface. Such an expansion lowers the rigidity of the surfactants monolayer, and subsequently, the spin probe can rotate more easily. Table 1 also shows the values of the order parameter, S, of 5-DSA in (w/o) microemulsions based on (A) R-(+)-limonene/ lecithin/1-propanol/water and (B) R-(+)-limonene/lecithin/1,2propanediol/water, respectively, for various water contents. These order parameter values reflect the rigidity of the membrane from the depth of the membrane where the doxyl ring of the 5-DSA is located, namely, the fifth carbon atom. By increasing the water content of the system from 0.5 to 1.7% w/w, in the case of R-(+)-limonene/lecithin/1-propanol/water microemulsions the order parameter was slightly decreased from 0.10 to 0.08 (Table 1). Further increase of the water content, up to 7.9% w/w, induced an increase and finally stabilization of the order parameter values. When 1-propanol was used as co-surfactant, the motional degree of freedom of 5-DSA was not so much affected by the water content of the microemulsions. When R-(+)-limonene/lecithin/ 1,2-propanediol/water microemulsions were considered, the order parameter values were remarkably higher than in the case of R-(+)-limonene/lecithin/1-propanol/water microemulsions. Such (21) Avramiotis, S.; Papadimitriou, V.; Hatzara, E.; Bekiari, V.; Lianos, P.; Xenakis, A. Langmuir 2007, 23, 4438-4447.

3384 Langmuir, Vol. 24, No. 7, 2008

Figure 3. Hydrodynamic radius of water droplets versus water content in R-(+)-limonene/lecithin/1-propanol/water (w/o) microemulsions.

an observation indicates formulation of a more flexible membrane when lecithin was mixed with 1-propanol instead of 1,2propanediol. By increasing the water content of the system from 0.5% to 2.2% w/w, the order parameter was decreased from 0.20 to 0.16. Such a variation indicates a decrease in the rigidity of the interface, which is in agreement with the increased spin probe’s mobility mentioned above. From the hyperfine splitting constants, R′0, of 5-DSA when located in the interface, the polarity of the membrane at the depth of the doxyl ring can be evaluated in both 1-propanol- and 1,2propanediol-formulated microemulsions. When 1-propanol was considered as co-surfactant the hyperfine splitting constant was 14.76 ( 0.07 G, whereas in the case of 1,2-propanediol the same constant was lower, 13.99 ( 0.16 G. As mentioned above, the hyperfine splitting constant of the spin probe reflects the polarity of its microenvironment and increases as the polarity of the medium is increased. In the present study we can conclude that the membrane is more polar when 1-propanol is used as co-surfactant instead of 1,2-propanediol. The dynamic structure of the surfactants monolayer was strongly affected by the nature of the co-surfactant used for formulation of biocompatible (w/o) microemulsions. When 1-propanol was considered, the emulsifiers interface was much more flexible, indicating a less tight packing of lecithin molecules. In that case, the system was able to incorporate higher quantities of water. On the contrary, when 1-propanol was replaced by a diol, namely, 1,2-propanediol, the interface became more rigid and water solubilization in the system was considerably lower. Dynamic Light Scattering Studies. Dynamic light scattering measurements were carried out in order to evaluate the size of water droplets in the microemulsions as a function of water contents and other additives concentration. For the R-(+)limonene/lecithin/1-propanol/water microemulsions (series A) an increase of the hydrodynamic radius of the droplets was observed by increasing water content in the microemulsions (Figure 3) and up to the higher possible water content (12% w/w). Radii in the range of 60-180 nm were observed leading to average diameters between 120 and 360 nm for the water droplets. A parallel increase in the scattering intensity was also observed as a result of the scattering contrast increase due to an increase in the water concentration. Droplet sizes are in accordance with previous investigations on similar systems of comparable water contents.22 The larger sizes observed at the upper limits (22) Schurtenberger, P.; Cavaco, C. Langmuir 1994, 10, 100-108.

Papadimitriou et al.

of water concentration range should be considered logical if one takes into account the relatively large water capacity of the present system. From a more conservative point of view the calculated radii can be considered as generalized hydrodynamic correlation lengths,22 which give information on the structural changes taking place in the systems as water content increases. This point of view may be more useful if one considers the possibility for formation of nonspherical structures in the particular microemulsions. In other words, the large hydrodynamic sizes observed may also be correlated with formation of, e.g., wormlike structures, with large effective diameters.22 If a buffer solution is used instead of pure water for microemulsion formation at a concentration of 1.3% w/w, water droplet sizes are found to have Rh ) 80 nm, somehow increased from the values expected from the line determined from the water microemulsions. This is something to be expected since modification of interfaces should be different in the case of water and buffer solution due to the presence of salts in the latter case. When lipase (lipase from R. miehei) is present in the system (at a buffer concentration of 1.3% w/w) droplet hydrodynamic radius decreases to 62 nm. This change may indicate a different modification of the buffer/matrix interface by the enzyme, since the protein may also play the role of a macromolecular surfactant in the system. The presence of enzyme is also followed by an increase in the scattering intensity measured from the microemulsions, since the mass of the droplets now increases due to incorporation of the protein. In the case of R-(+)-limonene/lecithin/1,2-propanediol/water, microemulsions (series B) droplet radius seems to remain constant in the water range studied (0.5-2.1% w/w) and equal to about 8 nm (diameter equal to 16 nm). Addition of buffer solution at a 1.3% w/w concentration, followed by addition of lipase, does not change the observed size of the droplets (within the resolution and experimental error of DLS). Currently there is no clear explanation for the observed behavior, but it may be associated with the greater viscosity of the B series compared to the A series. To conclude, when (w/o) microemulsions based on R-(+)limonene, lecithin, 1-propanol, and water were studied by DLS, an almost linear increase of the hydrodynamic radius of the dispersed aqueous domains by increasing the water content was observed. Such a dependence was not obvious when 1-propanol was replaced by 1,2-propanediol in the same microemulsion systems. In the former case, the size of the aqueous domains was affected when either salts or enzymic molecules were solubilized in them. Such a phenomenon was not observed in the case of 1,2-propanediol-containing microemulsions. Conductivity Measurements. The conductivity of biocompatible (w/o) microemulsions based on R-(+)-limonene as the continuous oil phase, lecithin as surfactant and either 1-propanol or 1,2-propanediol as co-surfactants, was measured at constant temperature, 25 °C, with increasing weight fractions of water. Using an appropriate buffer solution, 0.2 M Tris/HCl, ions were incorporated in the water phase, rendering it electrically conducting. As reported elsewhere, (w/o) microemulsions exhibit small macroscopic conductivity as the aqueous droplets containing the ions are separated by the continuous oil phase.16 However, a sharp increase in conductivity is observed beyond a certain value of water weight fraction indicating the structural transition of the system in the bicontinuous form. The water content of the microemulsion which points to the above-mentioned transition is called the percolation threshold. Figure 4 shows the variation of conductivity as a function of the water content in two different microemulsion systems: (A)

Biocompatible Microemulsions Based on Limonene

Figure 4. Variation of the conductivity of (w/o) microemulsions based on R-(+)-limonene/lecithin/alcohol/water as a function of the water content (0.2 M Tris/HCl buffer solution). The weight ratio of limonene to emulsifiers was 1.5 and of lecithin to alcohol was 1: (0) 1-propanol and (9) 1,2-propanediol. The temperature was 25 °C.

R-(+)-limonene/lecithin/1-propanol/water and (B) R-(+)limonene/lecithin/1,2-propanediol/water. The weight ratios of limonene to emulsifiers and of lecithin to alcohol were kept constant throughout the experiment, 1.5 and 1, respectively. In both systems, by increasing the water content, the conductivity is increased, indicating an interconnection of the aqueous droplets. When the system R-(+)-limonene/lecithin/1-propanol/water was considered, a sharp increase in conductivity was observed at water content above 4% w/w, indicating a structural transition in the bicontinuous form. In the case of R-(+)-limonene/lecithin/ 1,2-propanediol/water microemulsions, water incorporation in the system was limited (up to 1.7% w/w), and as a consequence, the percolation threshold was not observed since the system separates before the percolation region is reached. Lipase-Catalyzed Reactions. Esterification of either 1-propanol or 1,2-propanediol with three different carboxylic acids, namely, octanoic, dodecanoic, and hexadecanoic, catalyzed by lipase from R. miehei solubilized in the aqueous domains of microemulsions formulated with R-(+)-limonene and lecithin has been evaluated. In that case, the short-chained alcohols (1-propanol or 1,2-propanediol) used as co-surfactants for formulation of microemulsions also played the role of substrate for esterification of the carboxylic acids. In both cases, alcohol consumption for generation of the carboxylic acid esters was too small and did not affect the stability of the microemulsion system. Figure 5 shows the percent conversion yield of the carboxylic acids (octanoic, dodecanoic, hexadecanoic) during esterification with 1-propanol, catalyzed by lipase from R. miehei in R-(+)limonene (w/o) microemulsions at 30 °C. As it can be observed from Figure 5, the higher conversion yield was obtained for esterification of 1-propanol with hexadecanoic acid, 99% after 142 min, followed by that of dodecanoic acid, 95% after 142 min. In the case of octanoic acid the highest conversion yield, 87%, was observed after 94 min followed by a decrease to 81% after 166 min of reaction. From the experimental results mentioned above there seems to be a preference for the long-chain carboxylic acids when esterification with 1-propanol catalyzed by lipase from R. miehei in R-(+)-limonene microemulsions was considered. Although there is no clear evidence yet as for why some lipases show similar behavior concerning the rate of the esterification reaction when solubilized in microheterogeneous systems, the same behavior has been reported in the literature by other researchers as well. In 1993 Stamatis

Langmuir, Vol. 24, No. 7, 2008 3385

Figure 5. Conversion yield (%) for esterification of (9) octanoic acid, (0) dodecanoic acid, and (2) hexadecanoic acid (85 mM) with 1-propanol (2.8 M), catalyzed by lipase from R. miehei (83 µg/mL) solubilized in (w/o) microemulsions of R-(+)-limonene/lecithin/1propanol/water. The temperature was 30 °C.

and co-workers studied the activity of different lipases in esterification reactions of various substrates and related the enzyme selectivity to its localization in the reverse micellar system.23 More recently, Zhou and collaborators reported a similar behavior for a lipase immobilized in microemulsion-based organogels (MBGs).24 According to them, the chain length of the acids is related to their dissociation and the consequent variation of the ionic strength within the water domains. The dissociation degree is higher when shorter fatty acids are considered. Such a variation of the ionic strength possibly affects the activity conformation of the protein. When lipase from R. miehei was solubilized in the aqueous domains of R-(+)-limonene/lecithin/1,2-propanediol/water microemulsions in the presence of the same carboxylic acids, namely, octanoic, dodecanoic, and hexadecanoic, under identical experimental conditions, formulation of the corresponding esters was not observed. This behavior could be possibly attributed to the specificity of the lipase toward the alcohol employed for esterification of the acids. On the other hand, this phenomenon could be also related to the structural changes observed in the R-(+)-limonene-based microemulsions when 1-propanol was replaced by 1,2-propanediol. In the latter case, the increased rigidity of the interface followed by an increase of the reaction medium’s viscosity could affect the ability of the lipase to perform the esterification reaction. From the enzymatic studies mentioned above we can conclude that (w/o) biocompatible microemulsions formulated with R-(+)limonene, water, and a mixture of lecithin and 1-propanol as emulsifiers can be used as reaction media for the effectuation of esterification reactions catalyzed by lipase. The specificity of the enzyme as well as the structural characteristics of the reaction medium strongly affects the performance of the enzymic reaction.

Conclusion Biocompatible (w/o) microemulsions can be obtained using R-(+)-limonene, water, and a mixture of lecithin as surfactant and either 1-propanol or 1,2-propanediol as co-surfactant. Such systems can serve as microreactors for the effectuation of lipasecatalyzed esterifications. In the present study, synthesis of different propyl esters was reported. The efficiency of the lipase to convert (23) Stamatis, H.; Xenakis, A.; Provelegiou, M.; Kolisis, F. N. Biotechnol. Bioeng. 1993, 42, 103-110. (24) Zhou, G. W.; Li, G. Z.; Xu, J.; Sheng, Q. Colloids Surf. A. 2001, 194, 41-47.

3386 Langmuir, Vol. 24, No. 7, 2008

octanoic, dodecanoic, and hexadecanoic acids to the corresponding propyl esters was related to the structural characteristics of the reaction medium. To study the interfacial properties of the emulsifier’s monolayer electron paramagnetic resonance (EPR) spectroscopy was applied. The dynamic structure of the surfactant’s monolayer was strongly affected by the nature of the co-surfactant used. When 1-propanol was considered, the emulsifier’s interface was much more flexible and higher water quantities were incorporated in the system than in the case of 1,2-propanediol. The size of the dispersed aqueous domains in (w/o) microemulsions was evaluated by dynamic light scattering

Papadimitriou et al.

(DLS). A linear increase of the hydrodynamic radius of the dispersed aqueous domains by increasing the water content was observed when 1-propanol was used as co-surfactant. On the contrary, droplet radius remained constant in the water range studied when 1,2-propanediol was considered. In the latter case the size of the aqueous domains was considerably smaller. During the percolation transition and in percolated microemulsions the dispersed aqueous phase possibly retains its closed structure and ions are transferred through transient merging of these structures. LA703682C