water interfaces

Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102

Contents lists available at SciVerse ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Surface activity of saponin from Quillaja bark at the air/water and oil/water interfaces Kamil Wojciechowski ∗ Department of Microbioanalytics, Faculty of Chemistry, Warsaw University of Technology, Warsaw, Poland

a r t i c l e

i n f o

Article history: Received 6 August 2012 Received in revised form 17 December 2012 Accepted 13 February 2013 Available online xxx Keywords: Saponins Adsorption Ionic surfactants Air–water Olive oilwater Tetradecane–water

a b s t r a c t Surface activity of Sigma’s Quillaja bark saponin (QBS) was studied by means of dynamic interfacial tension and surface dilational rheology at three fluid/fluid interfaces with the polarity of the non-aqueous phase increasing in the order: air/water, tetradecane/water and olive oil/water. The equilibrium interfacial tension isotherms were fitted to the generalized Frumkin model with surface compressibility for the air/water and tetradecane/water interfaces, whereas the isotherm for the third interface displays a more complex shape. Upon fast compression of a drop of concentrated “Sigma” QBS solution immersed in olive oil, a clearly visible and durable skin was formed. On the other hand, no skin formation was noticed at the air/water interface, and only a little at the tetradecane/water interface. Addition of a fatty acid, however, improved slightly the skin-formation ability of the QBS at the latter interface. The surface behavior of the QBS from Sigma was compared with that from Desert King, Int. (“Supersap”), employed in a recent study by Stanimirova et al. [22]. The two products exhibit different areas per molecule in the saturated adsorbed layer (0.37 nm2 vs. 1.19 nm2 for “Sigma” and “Supersap”, respectively). Also their surface rheology is different: although both QBSs form predominantly elastic layers, for “Sigma” the surface storage modulus, εr = 103 mN m−1 , while for “Supersap” εr = 73 mN m−1 at 10−3 mol l−1 (i.e., around their cmc). The two saponin products exhibit also different ionic character, as proven by the acid-base titration of their aqueous solutions: QBS from Sigma is an ionic surfactant, while the “Supersap” from Desert King is a non-ionic one. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Surfactants currently used in food and cosmetic industries are predominantly produced by chemical transformation of animal and plant fats or petroleum derivatives [1]. Recently, however, numerous studies are oriented toward biosurfactants, obtained directly from renewable natural material. Biosurfactants can be produced by numerous plants [2] and microorganisms using low-cost substrates, and even waste [3]. Especially saponins (Fig. 1) find several applications in food and cosmetic industry because of their foaming and emulsifying properties (the name “saponin” derives from the Latin sapo, that means “soap”). The most prominent sources of saponins are the Chilean tree Quillaja saponaria Molina [4] and Californian tree Yucca schidigera. Further examples include Chinese perennial Panax ginseng [5] and Southeast Asian shrub Camellia sinensis, known as “tea plant” [6]. The Quillaja bark saponin (QBS) extract is an approved ingredient for use in food and beverages as a flavoring agent. It also bears FEMA (Flavor and Extract Manufacturers’ Association) GRAS status. In the European Union, it

∗ Corresponding author. E-mail address: [email protected] 0927-7765/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfb.2013.02.008

is an approved ingredient (E999) for water-based non-alcoholic drinks and ciders. In Japan, the Quillaja extract is allowed for human consumption (as emulsifier and foaming agent) and for use in cosmetics. Because of their unique properties, which include sweetness or bitterness [7], saponins find applications also in other industries: in agriculture they are mainly used as an additive to animal feed [8], for soil remediation [9] and as natural pesticides [10]. Some important potential applications, partially related to their haemolytic activity [7], include cancer therapy (especially combined with other anticancer drug to enhance the growth inhibition of tumor) [11] or prevention of cardiovascular diseases. Saponins have proven useful for lowering the LDL fraction of cholesterol by its solubilization in micelles [12]. Other potential applications of saponins include: antidepressants [13], antioxidants [14], antibacterial and antiviral agents, immune system stimulators, and many others [7]. Despite the common name “Quillaja bark saponin”, the commercially available extracts vary greatly in composition, mostly due to an important variability of the raw material and differences in extraction protocols. This is the reason why several, often conflicting reports, can be found in the literature on both bulk and surface properties of Quillaja saponins. Any comparison with the literature data should therefore be made with caution and a

96

K. Wojciechowski / Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102

H3C

CH3

triterpene sapogenin O CH3

CH3

28

glucuronic acid 3

R

1

HO O

O

OH

O

HO HO

O

H3C

CHO

O

R 1

O

O

CH3

O O

H3C O

HO O

OH

OH

CH3

HOOC

O

O O R

3

R

H

2

OH

E=

OH

4

4

Fig. 1. General structure of QBS. R –R represent either hydrogen or carbohydrate group.

special attention should be paid to the origin of saponins under study. In this paper we provide a thorough characterization of surface activity of a QBS available commercially from Sigma (cat. No 85410). Dynamic and equilibrium interfacial tensions at the air/water, tetradecane/water and olive oil/water interfaces are discussed and compared with those reported recently for another commercially available QBS (“Supersap” from Desert King).

All interfacial tension () measurements were performed using a drop profile analysis tensiometer PAT-1 (Sinterface Technologies, Germany), as described in [15]. The glassware was cleaned with acetone and Hellmanex II solution (Hellma Worldwide) and rinsed with copious amounts of Millipore water. All the experiments were performed at constant temperature (21 ◦ C) controlled with a thermostatic bath. Two QBSs from different sources were used: 84510 “Sigma” (8–25% sapogenin, Premium quality) purchased from Sigma–Aldrich, and “Supersap” donated by Desert King, Int. Since the exact composition of these products is not known, the average molecular weight of 1650 g mol−1 was assumed for both QBSs. Tetradecane (Aldrich 172456, ≥99%) and olive oil (Sigma O1514, highly refined, low acidity) were purchased from Sigma–Aldrich, and were purified by shaking with Florisil® (Sigma–Aldrich) followed by centrifugation at 6000 rpm during 30 min and filtering through Whatman’s filter paper (No. 1). The criterion for the surface purity of the oil phase was the constancy of the corresponding interfacial tension during at least 3600 s (the timescale of the typical experiment) against pure water (whose surface purity had been confirmed prior to contacting with the oil phase). For the olive oil, a single shaking/centrifugation/filtering cycle using Florisil 30–60 mesh (Sigma–Aldrich, 288691) was sufficient, while for the tetradecane four cycles were necessary: two with Florisil 30–60 mesh, and two with 60–100 mesh (Fluka, 46385, for chromatography). Fresh Milli-Q (Millipore) water (18.2 × 106 cm) was used for all the measurements. The QBS aqueous solution drop (5–30 ␮l) was formed at the tip of a steel capillary immersed in a glass cuvette (10 ml), filled with air, tetradecane or olive oil. If not stated otherwise, the aqueous solutions of QBS were buffered at pH 7 with the phosphate buffer (I = 0.02 mol l−1 ). Most experiments were performed for at least 3600 s (except for the air/water interface, where they lasted 2700 s), in some cases followed by oscillations of the drop volume, and repeated at least three times (typically six-to-seven times). The equilibrium interfacial tension ( eq ) values were calculated using a long-time extrapolation of dynamic interfacial tension [16]: RT 2 c



d = εr (f ) + iεi (f ) d ln A/A0

(2)

where  is the interfacial tension, f is the frequency of oscillations, A and A0 are the actual and initial interfacial area, respectively. The surface rheology was probed by performing sinusoidal perturbations of the drop volume after 3600 s of adsorption, in the frequency range 0.005–0.1 Hz, with the amplitude of 5%. The (t) amplitude and the phase shift with respect to the generated oscillations were obtained from the Fourier transformation of the data, and were used to calculate the real (storage modulus, εr ) and imaginary (loss modulus, εi ) parts of the elasticity modulus [17]. 3. Results 3.1. Dynamic interfacial tension of “Sigma” QBS at fluid/fluid interfaces

2. Experimental

eq =  −

where  is the interfacial tension,  eq is the equilibrium interfacial tension, R is the gas constant, T is the temperature, D is the diffusion coefficient, t is the time and c is the QBS bulk concentration. Eq. (1) is valid for t → ∞, when the subsurface concentration approaches that in the bulk. Therefore, the dynamic interfacial tension data were plotted in  vs. t−1/2 coordinates, and the intercept with the ordinate gave the equilibrium interfacial tension,  eq . The real (εr ) and imaginary (εi ) parts of the dilatational elasticity modulus, E, were obtained from the response of (t) to periodic area perturbations:

 4Dt

(1)

The interfacial tension decays for the same set of “Sigma” QBS concentrations in the aqueous phase at three interfaces: air/water, tetradecane/water and olive oil/water are presented in Fig. 2 as dynamic surface pressures. In all cases, two groups of the dynamic curves can be clearly distinguished, with a sharp transition between them around the concentration of 1 × 10−5 mol l−1 . For the air/water and olive oil/water (both buffered and not buffered), the concentration of 1 × 10−5 mol l−1 is a boundary case with the initial behavior resembling that of the lower concentration solutions and a consecutive slow increase of surface pressure, characteristic for the higher concentration solutions, at later stages of adsorption. For tetradecane, this transition is less sharp and spans over two intermediate concentrations (4 × 10−6 mol l−1 and 10−5 mol l−1 ). The rate of interfacial tension decays clearly increases from nonpolar air to polar olive oil, where practically stable readings are obtained within minutes at higher “Sigma” QBS concentrations. 3.2. Interfacial tension isotherms of “Sigma” QBS at fluid/fluid interfaces The dynamic interfacial tension data were used to construct the interfacial tension isotherms for “Sigma” QBS by extrapolation of the dynamic data to t → ∞ (t−1/2 → 0), using the asymptotic solution of the Ward–Tordai equation. Fig. 3 shows a comparison of the three isotherms obtained for the air/water, tetradecane/water and olive oil/water interfaces. The trends observed in the dynamic curves (Fig. 2) are fully confirmed in the equilibrium data. For a given QBS concentration, the surface pressures are higher for both non-polar phases (air and tetradecane) than for the relatively polar olive oil. Interestingly, the presence of the phosphate buffer does not have any significant effect on the shape of the olive oil/water interfacial tension, in contrast to our previous results for the air/water interface [18]. At low QBS concentrations (<4 × 10−6 mol l−1 ) the isotherms for the air/water and olive oil/water show similar behavior, i.e., very little increase of the surface pressure. Above cQBS = 4 × 10−6 mol l−1 for both interfaces a sharp increase in the surface pressure is observed, which continues until 1 × 10−5 mol l−1 for the olive oil/water and 4 × 10−5 mol l−1 for

K. Wojciechowski / Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102 45

40

40

35

35

30

30

/mN.m

/mN.m

-1

-1

45

25 20

25 20

15

15

10

10

5

5

0

0 0

500

1000

t /s

1500

2000

2500

45

45

40

40

35

35

30

30 -1

-1

25

/mN.m

/mN.m

97

20

500

1000

0

500

1000

t /s

1500

2000

2500

1500

2000

2500

25 20

15

15

10

10

5

5

0

0

0 0

500

1000

1500

2000

2500

t /s

t /s

Fig. 2. Dynamic surface pressures at (from left upper corner, clockwise): air/water (buffered), tetradecane/water (buffered), olive oil/water (buffered), olive oil/water (not buffered). “Sigma” QBS concentrations in the aqueous phase: 5 × 10−7 mol l−1 (), 1.5 × 10−6 mol l−1 (䊉), 4 × 10−6 mol l−1 (), 1 × 10−5 mol l−1 (), 4 × 10−5 mol l−1 (), 1 × 10−4 mol l−1 (), 4 × 10−4 mol l−1 (), 1 × 10−3 mol l−1 (

).

the air/water. For both interfaces a slow increase of the surface pressure continues until the cmc is reached around 4 × 10−4 mol l−1 , although for the olive oil system a significant reduction of the slope in this region of the isotherm can be noticed. The situation is clearly different in the case of tetradecane/water interface, where the isotherm starts to raise already at 1.5 × 10−6 mol l−1 and increases steeply until cQBS = 4 × 10−5 mol l−1 , followed by a slow increase up to the cmc, similarly to the other two interfaces. The maximum surface pressures (at cmc) reached at the three interfaces are 35, 45 and 23 mN m−1 , for the air, tetradecane and olive oil, respectively. In order to quantitatively compare the interfacial behavior of “Sigma” QBS at the air/water and tetradecane/water interfaces, the data from Fig. 3 were analyzed using the Isofit software by Aksenenko [19] and the best-fit parameters are presented in Table 1. The best results were obtained using the generalized Frumkin model [20], where the adsorption isotherm and the surface equation of state are given by the following equations: bc =

Fig. 3. Surface pressure isotherms of “Sigma” QBS at the air/water (䊉), tetradecane/water (), and olive oil/water (), buffered and (), not buffered) interfaces. Error bars represent the maximum value of the standard error for the procedure of extrapolation of the dynamic data from Fig. 2. The solid lines correspond to the best fits to the modified Frumkin model (Eqs. (3)–(5)) with the parameters from Table 1. Inset shows a comparison of the steep parts of the surface tension (air/water) isotherms for “Sigma” and “Supersap” QBSs with the corresponding d/dlog c slopes.

 exp(−2˛) 1−

(3)

Table 1 Best-fit parameters of “Sigma” QBS surface and interfacial tension isotherms (water–air, water–tetradecane) using the Frumkin model.

Air/water Tetradecane/water

ω/m2 mol−1

˛

ε/m mN−1

b/m3 mol−1

2.1 × 105 2.6 × 105

2.0 1.6

0.03 0.03

28.5 334.4

98

K. Wojciechowski / Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102

Fig. 4. Photographs of phosphate-buffered “Sigma” QBS aqueous solution drops in olive oil at different times after the compression.

˘ =−

 RT  ln(1 − ) + 2 ˛ ω

ω = ω0 (1 − ε˘)

(4) (5)

b is the adsorption constant, c is the bulk QBS concentration,  is the surface coverage ( =  ω),  is the adsorbed amount, ω is the area per mole of adsorbed QBS, ω0 is the area per mole of solvent (water), ˛ is the interaction parameter, ε is the relative twodimensional compressibility coefficient, R is the gas constant, T is the temperature, ˘ is the surface pressure. 3.3. Skin formation and interfacial rheology of QBS at olive oil/water interface When a drop of the phosphate-buffered concentrated aqueous “Sigma” QBS solution immersed in the olive oil was rapidly contracted by more than 10% of its initial area, a clearly visible skin could be observed on the drop surface. The skin showed a characteristic pattern of wrinkles in the neck region, all parallel to the symmetry axis, and remained visible for at least 109 min (when the observation was stopped), as shown in Fig. 4. In the absence of the phosphate buffer, the skin was also formed, but the drops were pulled back to the capillary and collapsed within a couple of minutes (see Fig. S1, Supporting Information). A similar, although much less pronounced and less stable skin was observed for “Sigma” QBS drops (both buffered and non-buffered) immersed in tetradecane. The wrinkles were hardly visible already one minute after the compression, and after two minutes the drop’s surface became smooth again (see Fig. S2). Interestingly, no skin formation was observed in air even at the highest “Sigma” QBS concentration, 1 × 10−3 mol l−1 (both buffered and non-buffered) although it has been observed earlier by Stanimirova et al. for “Supersap” QBS (from a different source, see below). The fact that the skin formation in solutions containing the “Sigma” QBS is enhanced in olive oil suggests that some chemical processes may be involved. The molecules constituting QBS extract contain several potentially reactive groups, e.g. OH, COOH

(see Fig. 1). On the other hand, the olive oil may contain components other than triglycerides, e.g. their hydrolysis products (free fatty acids, mono- and di-glycerides). Typical content of free fatty acids (mostly oleic, palmitic and linoleic acids) in olive oil does not exceed a few wt%, depending on preparation and degree of purification [21]. The fatty acids were chosen as components most likely responsible for the enhanced skin formation in olive oil. In order to verify if any reaction between the “Sigma” QBS constituents and fatty acids could affect the skin formation, the QBS drops formed in tetradecane solutions of palmitic acid (PA) were analyzed. Indeed, with increasing PA concentrations (0.2–0.5%) the skin was more visible and more durable in comparison to pure tetradecane (see Fig. S3). It is likely that the unsaturated fatty acids present in olive oil (oleic, linoleic and others at minor concentrations), would enhance the skin formation even further. In order to shed some light on the mechanical properties of skin-forming layers, the surface dilational rheology response of the adsorbed layers formed in concentrated “Sigma” QBS solutions (1 × 10−3 mol l−1 ) were compared. The 10–30 ␮l drops were formed from a stainless steel capillary and left in contact with the oil phase for 3600 s at constant volume (maintained by the feedback between the drop shape analysis in real time and a syringe pump connected to the capillary). Then, harmonic surface deformations were applied at five frequencies: 0.005 Hz, 0.01 Hz, 0.02 Hz, 0.05 Hz and 0.1 Hz. The corresponding storage and loss moduli (εr and εi ) were determined as described in the experimental part. Fig. 5 shows the oscillation frequency dependence of both the storage and loss moduli for “Sigma” QBS at the air/water, olive oil/water and tetradecane/water interfaces (with the aqueous phase buffered in all cases). The corresponding data for the “Supersap” QBS is also added for comparison (see below). In all cases, the rheological response of the adsorbed QBS layers is predominantly elastic, especially at higher oscillation frequencies. The storage modulus increases with an increase of the oscillation frequency, and decreases with an increase of the non-aqueous phase polarity. At high concentrations (where the skin formation is best evidenced), the high-frequency storage modulus for the olive oil

K. Wojciechowski / Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102

99

100 20 80

/mN m

10

i

60

r

/mN m

-1

-1

15

40 5 20 0 0,01

0,1

0,01

f /Hz

0,1

f /Hz

Fig. 5. Storage (εr ) and loss (εi ) moduli for 1 × 10−3 mol l−1 “Sigma” QBS solutions at the olive oil/water (), tetradecane/water (䊉), and air/water () as a function of oscillation frequency. For comparison, the results for the QBS from different supplier (“Supersap”) at the air/water interface are showed (). Error bars represent the standard deviation of the mean of at least 3 oscillatory surface rheology experiments.

3.4. Comparison between the “Sigma” and “Supersap” QBS The comparison of recently published data on two Quillaja bark saponins [18,22] clearly calls for some explanation of the significant differences observed for these supposedly similar products.

-1 r /mN m

60 40 20 0 0,01

0,1

f /Hz

-1 i /mN m

10 8 6 4 2 0 0,01

0,1

f/ Hz

-1 r, i /mN m

εr = 12 mN m−1 , while for the air, εr = 103 mN m−1 . This suggests that the skin formation is not directly related to the surface elasticity of the layer. The lack of a clear correlation between the skin formation and εr at high “Sigma” QBS concentration prompted us to study in more detail the surface rheological properties of this QBS at the olive oil/water interface. Even though, as pointed recently by Stanimirova et al., the drop shape analysis-based techniques (DSA) are not very suitable for surface rheological studies of highly elastic films, it is difficult to find a good alternative for DSA at liquid/liquid interfaces. Stanimirova et al. pointed to the anisotropy of drop deformation (evidenced by the appearance of the wrinkles), as the most probable source of error in DSA. In order to estimate the effect of this anisotropy, the 1 × 10−3 mol l−1 “Sigma” QBS solution drops of different size (15–30 mm2 ) were subject to the same surface deformation, ˛ = 20–50% (˛ is defined as a relative change of surface area of the drop). Using the appearance of the wrinkles as a criterion, we found that the smaller the drop, the more isotropic the deformation is. Hence, for the subsequent study of rheological properties of the “Sigma” QBS adsorbed layers, the smallest possible drops of the aqueous phase were used. The frequency dependence of the storage and loss moduli (Fig. 6) shows that for all “Sigma” QBS concentrations εr > εi , and that εr increases with increasing frequency of oscillations, while the opposite is observed for εi . Another characteristic feature of the system is a non-monotonous concentration dependence of both moduli. In the whole frequency range, εr and εi initially increase with increasing the QBS concentration until the maximum at cQBS = 1 × 10−5 mol l−1 . Upon further increase of the QBS concentration, both moduli decrease, and around their cmc reach the values that are practically independent of the modulation frequency: εr = 12 mN m−1 , εi < 1 mN m−1 .

60 40 20 0 1E-6

1E-5

1E-4

1E-3

-1

CQBS /mol L

Fig. 6. Frequency dependence of the storage (εr ) and loss moduli (εi ) for “Sigma” QBS concentrations in the aqueous phase, measured at the olive oil/water interface at 21 ◦ C: 5 × 10−7 mol l−1 (), 1.5 × 10−6 mol l−1 (䊉), 4 × 10−6 mol l−1 (), 1 × 10−5 mol l−1 (), 4 × 10−5 mol l−1 (), 1 × 10−4 mol l−1 (), 4 × 10−4 mol l−1 (), 1 × 10−3 mol l−1 ( ). The bottom graph shows concentration dependence of the storage () and loss moduli (䊉) at the frequency of 0.1 Hz, corresponding to the limiting (Gibbs) elasticity (see Section 4). Error bars represent the standard deviation of the mean of at least 3 oscillatory surface rheology experiments.

100

K. Wojciechowski / Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102

behaves as a non-ionic compound, the “Sigma” QBS behaves like a weak acid, with pKa = 6.1. The molar fraction of proton-dissociable groups on “Sigma” QBS estimated from the titration curve (Fig. 7) is 0.5.

11 10

pH

9

4. Discussion

8 7 6 5 0

500

1000

1500

2000

VNaOH (0,1 M) / L Fig. 7. Acid–base titration curves for “Sigma” in pure water () and in KCl solution, I = 2 × 10−2 mol l−1 (), as well as for “Supersap” QBS in pure water ( ).

The main and relatively easy to verify characteristic of the adsorbing layer is a slope of the interfacial tension isotherm, d/dlog c. According to the Gibbs equation, this slope determines the area occupied by one molecule at the interface (A): A=



−

NAv d 2, 303RT d log c

−1 (6)

where NAv is the Avogadro number, R is the gas constant and T is the temperature. In order to semi-quantitatively compare the “Sigma” QBS used in our previous [18,23] and current study (84510 Saponin from Sigma) with the one used in Ref. [22] (“Supersap” from Desert King, Int.), the surface tension of their buffered aqueous solutions was measured with the same instrument (DSA tensiometer, PAT-1 from Sinterface). The middle parts of the two surface tension isotherms, determining d/dlog c slopes in the Gibbs equation (Fig. 3, inset) clearly prove that the two samples behave differently in terms of their adsorption at the air/water interface. The slope for the “Sigma” QBS is much higher than that of the “Supersap”, suggesting that the area per molecule in the monolayer is lower in the former, as it was suggested in [22]. A rough estimation of the area per molecule obtained from Eq. (6) gives 0.37 nm2 and 1.19 nm2 , for “Sigma” QBS and “Supersap” QBS, respectively. Even though only three points of each isotherm were used for these estimations, agreement with the previously reported results from analyses of full isotherms is very good. The dilational rheology response of the two QBSs compared in Fig. 5 also point to significant differences in mechanical properties of their adsorbed layers at the air/water interface – at cQBS = 1 × 10−3 mol l−1 those formed by “Sigma” QBS are more elastic in the studied frequency range. The ionic character of saponins remains a subject of intense debate [2,24,25–27]. The most often employed criterion of ionic character is the effect of added electrolyte on the interfacial tension isotherm. Surprisingly, the contradictory conclusions have been drawn by different authors for presumably the same QBS products. In order to clarify this issue we performed the acid-base titrations for both “Sigma” and “Supersap” saponins (cQBS = 1 × 10−3 mol l−1 ) in pure water. The curve for “Sigma” QBS shows a typical sigmoidal shape characteristic for a weak acid, while that for “Supersap” resembles the curve for titration of pure water, i.e. with no or very little buffer capacity (Fig. 7). The “Sigma” QBS was additionally titrated in the presence of electrolyte (KCl) of the same ionic strength as that of the phosphate buffer used in the study (cKCl = 2 × 10−2 mol l−1 ). The two curves practically coincide, excluding the effect of an electrolyte as a possible source of the differences between “Supersap” and “Sigma”. While the former clearly

The slow decays of interfacial tension observed for “Sigma” QBS at all three interfaces (Fig. 2) suggest the existence of some adsorption barrier, in agreement with previous reports [18,22]. This barrier does not seem to be related critically to the QBS diffusion in the aqueous phase (where it is dissolved), since the rate of interfacial tension decays depends to a large extent on the nature of the contacting non-aqueous phase. The rate of surface pressure increase is the highest for the most polar olive oil/water interface, and the slowest for the air/water. The opposite trend can be observed in the values of surface storage and loss moduli (Fig. 5). However, the maximum surface pressures attained at each interface do not depend solely on the polarity of the non-aqueous phase: while the lowest values are found for olive oil, the highest – for tetradecane. Most likely, solvation of the hydrophobic parts of QBS by the solvent plays some role here. From the limited set of data available, one could speculate that solvation of QBS by the non-aqueous phase increases the lateral repulsion between the adsorbed “Sigma” QBS molecules, thus increasing the surface pressure. On the other hand, increasing the solvent polarity screens this repulsion, leading to a decrease of the surface pressure. The interfacial tension isotherms at the air/water and tetradecane/water interfaces shown in Fig. 3 have typical sigmoidal shape, while for the polar olive oil above cQBS = 1 × 10−5 mol l−1 the slope of the isotherm decreases. This would suggest that upon increase of the surface coverage the area per molecule in the adsorbed layer increases. Although such behavior has been observed for proteins, it is normally not expected for the low-molecular weight surfactants. On the other hand, the QBS molecules are bigger than the typical low-molecular weight surfactants, and it is possible that they behave in some aspects like proteins (for example they may undergo 2D condensation). Alternatively, the change of the slope could be an artifact related to either a partitioning of QBS to the non-aqueous phase, or a chemical reaction between the components of QBS and those of the olive oil. Whatever the origin of this behavior, it could only be observed at the polar oil/water interface. Although Stanimirova et al. showed that their experimental data can be better fitted with van der Waals-type isotherms (e.g., Volmer) than with the Langmuir-type ones, the QBS used in their study was different from the currently used “Sigma” QBS. The interfacial tension isotherms for “Sigma” QBS at two nonpolar interfaces are well described with a simple Frumkin model (Table 1), while the data for the olive oil cannot be fitted to neither Frumkin, nor to the more complex adsorption models, like reorientation or aggregation [28]. Given the fact that QBS is a mixture of several saponins, it should be borne in mind that the best-fit parameters from Table 1 in no case refer to any unique molecule of defined geometry. Nevertheless, they show that the air/water interface enables the most dense packing of “Sigma” QBS molecules (the lowest ω, close to the footprint of a single alkyl chain perpendicular to the interface). The fact that for this interface the interaction parameter, ˛, attained the physically reasonable maximum for the Frumkin model (for ˛ > 2 a phase transition is expected to take place in the monolayer) suggests that also the attractive interactions between the adsorbed molecules are the highest at the air/water interface. While the air/water interface favors the attractive interactions between the adsorbed “Sigma” QBS molecules, at the tetradecane/water interface, these molecules are probably solvated by tetradecane. This is evidenced by a significant increase of

K. Wojciechowski / Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102

the adsorption constant, b, and an area per molecule, most likely related to some tetradecane – QBS intermixing. The surface storage (εr ) and loss (εi ) moduli, in analogy to the bulk rheology, can be assigned to the corresponding elastic and viscous properties of the monolayer, respectively. With increasing frequency of modulations, the “Sigma” QBS monolayers at the olive oil/water interface clearly become more elastic, while the loss modulus decays practically to zero at 0.1 Hz. The maximum of the loss modulus for “Sigma” QBS monolayers lays at frequencies below 0.005 Hz, suggesting that the characteristic timescale of the process responsible for this behavior is rather long. Given the slow rate of interfacial tension decays observed for “Sigma” QBS, this long timescale can be assigned to the exchange of mass between the subsurface and surface (although it is much too long for the diffusion). At frequencies approaching 0.1 Hz, this mass exchange cannot follow the fast oscillatory changes of drop area, and the adsorbed “Sigma” QBS layers start to behave as insoluble ones. Such a behavior allows us to assign the value of the storage modulus at 0.1 Hz to the limiting elasticity, E0 [29]. In the framework of the Lucassen – van den Tempel (LT) model, the maximum of the concentration dependence of E0 seen in Fig. 6 (bottom) is assigned to the fact that with the increasing surface occupation, the concentration dependence of the adsorption becomes weaker. Unfortunately, a quantitative analysis of these data within the framework of the LT model was not possible because of the lack of a suitable adsorption model for QBS at the olive oil/water interface (see above). Skin formation has been previously reported for several proteins, both globular and random coil [30]. The “skin-like 2D films” appear upon adsorption at the surface of shrinking drops as a result of aging of the adsorbed layers. In the case of protein films, the skin formation can be enhanced by addition of low molecular weight amphiphiles, e.g., in protein/phospholipid mixtures [30], possibly through some interfacial complex formation. Recently, a 2D skin formation has been reported also for a QBS obtained from Desert King (“Supersap”) [22]. The comparison of the skin-forming behavior of the QBS used in our study (“Sigma”) with the one reported in Ref. [22] clearly shows a huge variability in QBS composition between different suppliers, and possibly also between different lots. The fact that in the present setup stable skin was not formed in the air but only in the olive oil (and to some extent in the fatty acid-containing tetradecane) might suggest that some chemical reaction is involved in the process. Possibly other QBS products (e.g., “Supersap”) contain enough of skin-forming material already present without reacting with the other phase components (e.g., fatty acids, mono- or diglycerides from the olive oil). It is interesting to compare the surface properties of the two QBS used in this study. The area per molecule obtained from the best-fit to Gibbs equation reported by Stanimirova et al. equals 1.11 nm2 for the “SuperSap” [22]. This practically coincides with the value estimated in this paper (1.19 nm2 ). On the other hand, for the “Sigma” QBS a much lower area per molecule (0.37 nm2 ) is obtained, in good agreement with our previous result from the full isotherm analysis [18]. Besides the differences in the slope of the isotherms, also the ionic character of both QBS differs significantly. The acid–base titration fully confirmed our earlier conclusions [18], as well as those of Stanimirova et al. [22]: “Sigma” QBS is an ionic surfactant, while “Supersap” is a nonionic one. In analogy to other groups of surfactants, e.g., fatty acids ([31,32]), also for “Sigma” QBS an increase of the ionic character improves the surface activity, even though only half of the molecules present in “Sigma” OBS have protondissociable groups. On the other hand, the presence of a single inflection point in the titration curves from Fig. 7 suggests that these groups are capable of dissociating only one proton each (e.g., COOH). Interestingly, the pKa of “Sigma” (6.1) is significantly higher than that expected for a typical carboxylic group (e.g. for

101

acetic acid, pKa = 4.8). This suggests that either the acidic character of the carboxylic group attached to the “Sigma” QBS molecule is weakened by some neighboring groups, or that some other than carboxylic groups (less acidic) are responsible for its ionic character. Finally, it should be stressed out that in our study both “Sigma” and “Supersap” QBSs display similar, high limiting dilational surface elasticity, E0 (0.1 Hz) for cQBS = 10−3 mol l−1 at the air/water interface: 103 and 73 mN m−1 , respectively. Interestingly, the latter value is smaller than that obtained using the DSA technique reported for “Supersap” in [22] (115 ± 15 mN m−1 ) and significantly smaller than the values obtained using other techniques in that work. The authors suggested that the difference between the DSA-derived results and those from other techniques might be attributed to the non-isotropic strain in elongated drops and possibly related to the formation of vertical wrinkles upon fast drop compression. Our current results suggest that the wrinkle formation is probably not the direct cause of the apparent reduction of E0 : despite the fact that we did not observe any wrinkles at the air/water interface even upon very large compressions (up to 50%), our experimental value is still ∼1/4 of the “correct” one, obtained from non-DSA techniques in [22]. Nevertheless, we fully support the hypothesis that the deformation in highly elongated (i.e., low interfacial tension and/or large volume) drops is not isotropic and may lead to significant errors in determination of surface dilational characteristics. 5. Conclusions The surface activity of Quillaja bark saponin, QBS (“Sigma” from Sigma–Aldrich, cat. No. 85410) was studied using dynamic interfacial tension at the air/water, tetradecane/water and olive oil/water interfaces. The rate of the interfacial tension decays is slow and depends on the nature and polarity of the contacting non-aqueous phase, suggesting the presence of an adsorption barrier. Interfacial tension isotherms also display marked dependency on the polarity of the non-aqueous phase. While the equilibrium data for the tetradecane/water and air/water interfaces could be well fitted using the Frumkin model, the analogous data obtained for the olive oil/water interface does not fit to any simple adsorption model. The surface rheological characteristics of “Sigma” QBS at high concentrations (1 × 10−3 mol l−1 ) is consistent with the formation of surface layers with dilational surface elasticities decreasing in order: air/water > tetradecane/water > olive oil/water. The systematic surface rheological study of the “Sigma” QBS layers at the olive oil/water interface revealed that the storage modulus passes through a maximum at concentrations corresponding to the saturation of the monolayer, while the loss modulus remains low and almost constant. At the olive oil/water (but not at the air–water) interface, the adsorbed layers of “Sigma” QBS are so thick and rigid, that the drops immersed in the oil phase wrinkle upon fast compression. This phenomenon has been observed previously for another QBS (Desert King’s “Supersap”), but at the air/water interface. Despite sharing the same name, the Quillaja bark extracts-derived saponins (QBS) may differ significantly in composition and consequently also in their surface properties. Thus, any comparison of the results from different studies should be done with care and only after verification that the products have similar composition. Further studies on bulk and interfacial behavior, as well as on the origin of differences between QBS from different sources are currently underway. Acknowledgements This work was funded by the Polish National Science Centre, grant no. DEC-2011/03/B/ST4/00780 and Warsaw University

102

K. Wojciechowski / Colloids and Surfaces B: Biointerfaces 108 (2013) 95–102

of Technology, Poland. J. Lewandowska and M. Piotrowski are acknowledged for performing the measurements of interfacial tension. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb. 2013.02.008. References [1] K. Holmberg, B. Jonsson, B. Kronberg, B. Lindman, Surfactants and Polymers in Aqueous Solution, 2nd ed., John Wiley & Sons, Inc., 2002. [2] J. Vincken, L. Heng, A. De Groot, H. Gruppen, Saponins, classification and occurrence in the plant kingdom, Phytochemistry 68 (3) (2007) 275–297. [3] N. Kosaric, Biosurfactants in industry, Pure Appl. Chem. 64 (11) (1992) 1731–1737. [4] R. San Martín, R. Briones, Industrial uses and sustainable supply of Quillaja saponaria (Rosaceae) saponins, Econ. Bot. 53 (3) (1999) 302–311. [5] J. Park, D. Rhee, Y. Lee, Biological activities and chemistry of saponins from Panax ginseng C.A. Meyer, Phytochem. Rev. 4 (2–3) (2005) 159–175. [6] K. Kobayashi, T. Teruya, K. Suenaga, Y. Matsui, H. Masuda, H. Kigoshi, Isotheasaponins B1–B3 from Camellia sinensis var. sinensis tea leaves, Phytochemistry 67 (13) (2006) 1385–1389. [7] S.G. Sparg, M.E. Light, J. Van Staden, Biological activities and distribution of plant saponins, J. Ethnopharmacol. 94 (2/3) (2004) 219–243. [8] G. Francis, H.P.S. Makkar, K. Becker, Quillaja saponins – a natural growth promoter for fish, Anim. Feed Sci. Technol. 121 (1/2) (2005) 147–157. [9] W. Chen, L. Hsiao, K.K. Chen, Metal desorption from copper(II)/nickel(II)-spiked kaolin as a soil component using plant-derived saponin biosurfactant, Process Biochem. 43 (5) (2008) 488–498. [10] B.P. Chapagain, Z. Wiesman, L. Tsror (Lahkim), In vitro study of the antifungal activity of saponin-rich extracts against prevalent phytopathogenic fungi, Ind. Crops Prod. 26 (2) (2007) 109–115. [11] S. Man, W. Gao, Y. Zhang, L. Huang, C. Liu, Chemical study and medical application of saponins as anti-cancer agents, Fitoterapia 81 (7) (2010) 703–714. [12] S. Kim, K. Park, Effects of Panax ginseng extract on lipid metabolism in humans, Pharmacol. Res. 48 (5) (2003) 511–513. [13] D. Zhou, H. Jin, H. Lin, X. Yang, Y. Cheng, F. Deng, J. Xu, Antidepressant effect of the extracts from Fructus Akebiae, Pharmacol. Biochem. Behav. 94 (3) (2010) 488–495. [14] J. Xiong, J. Guo, L. Huang, B. Meng, Q. Ping, Self-micelle formation and the incorporation of lipid in the formulation affect the intestinal absorption of Panax notoginseng, Int. J. Pharm. 360 (1/2) (2008) 191–196. [15] G. Loglio, P. Pandolfini, R. Miller, A.V. Makievski, F. Ravera, M. Ferrari, L. Liggieri, Drop and bubble shape analysis as a tool for dilational rheological studies of interfacial layers, Stud. Interface Sci. 11 (2001) 439–483.

[16] V.B. Fainerman, A.V. Makievski, R. Miller, The analysis of dynamic surface tension of sodium alkyl sulphate solutions, based on asymptotic equations of adsorption kinetic theory, Colloid Surf. A 87 (1) (1994) 61–75. [17] E.H. Lucassen-Reynders, Surface elasticity and viscosity in compression/dilation, Surf. Sci. Ser. 11 (1981) 173–216. [18] K. Wojciechowski, M. Piotrowski, W. Popielarz, T.R. Sosnowski, Short- and mid-term adsorption behaviour of Quillaja Bark Saponin and its mixtures with lysozyme, Food Hydrocolloid 25 (4) (2011) 687–693. [19] ISOFIT, Available from: http://www.thomascat.info/thomascat/Scientific/AdSo/ AdSo.htm [20] V.B. Fainerman, S.V. Lylyk, E.V. Aksenenko, J.T. Petkov, J. Yorke, R. Miller, Surface tension isotherms, adsorption dynamics and dilational visco-elasticity of sodium dodecyl sulphate solutions, Colloid Surf. A 354 (1–3) (2010) 8–15. [21] Y.Z.H.Y. Hashim, M. Eng, C.I.R. Gill, H. Mcglynn, I.R. Rowland, Components of olive oil and chemoprevention of colorectal cancer, Nutr. Rev. 63 (11) (2005) 374–386. [22] R. Stanimirova, K. Marinova, S. Tcholakova, N.D. Denkov, S. Stoyanov, E. Pelan, Surface rheology of saponin adsorption layers, Langmuir 27 (20) (2011) 12486–12498. [23] M. Piotrowski, J. Lewandowska, K. Wojciechowski, Biosurfactant–protein mixtures: Quillaja bark saponin at water/air and water/oil interface in presence of b-lactoglobulin, J. Phys. Chem. B 116 (2012) 4843–4850. [24] Ö. Güc¸lü-Üstünda˘g, G. Mazza, Saponins: properties, applications and processing, Crit. Rev. Food Sci. Nutr. 47 (3) (2007) 231. [25] S. Mitra, S.R. Dungan, Cholesterol solubilization in aqueous micellar solutions of quillaja saponin, bile salts, or nonionic surfactants, J. Agric. Food Chem. 49 (1) (2001) 384–394. [26] S. Mitra, S.R. Dungan, Micellar properties of Quillaja saponin. 2. Effect of solubilized cholesterol on solution properties, Colloids Surf. B 17 (2) (2000) 117–133. [27] S. Mitra, S.R. Dungan, Micellar properties of Quillaja saponin. 1. Effects of temperature, salt, and ph on solution properties, J. Agric. Food Chem. 45 (5) (1997) 1587–1595. [28] V.B. Fainerman, E.H. Lucassen-Reynders, R. Miller, Adsorption of surfactants and proteins at fluid interfaces, Colloids Surf. A 143 (2/3) (1998) 141–165. [29] E.H. Lucassen-Reynders, A. Cagna, J. Lucassen, Gibbs elasticity, surface dilational modulus and diffusional relaxation in nonionic surfactant monolayers, Colloids Surf. A 186 (1/2) (2001) 63–72. [30] Q. He, Y. Zhang, G. Lu, R. Miller, H. Möhwald, J. Li, Dynamic adsorption and characterization of phospholipid and mixed phospholipid/protein layers at liquid/liquid interfaces, Adv. Colloid Interface Sci. 140 (2) (2008) 67–76. [31] K. Lunkenheimer, W. Barzyk, R. Hirte, R. Rudert, Adsorption properties of soluble, surface-chemically pure n-alkanoic acids at the air/water interface and the relationship to insoluble monolayer and crystal structure properties, Langmuir 19 (15) (2003) 6140–6150. [32] J.R. Kanicky, A.F. Poniatowski, N.R. Mehta, D.O. Shah, Cooperativity among molecules at interfaces in relation to various technological processes: effect of chain length on the pKa of fatty acid salt solutions, Langmuir 16 (1) (1999) 172–177.