Wetting and electrokinetic properties of cholesterol—Revisited

Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

Contents lists available at ScienceDirect

Colloids and Surfaces A: Physicochemical and Engineering Aspects journal homepage: www.elsevier.com/locate/colsurfa

Wetting and electrokinetic properties of cholesterol—Revisited L. Holysz, A. Szczes, E. Chibowski ∗ Department of Physical Chemistry–Interfacial Phenomena, Faculty of Chemistry, Maria Curie-Sklodowska University, 20-031 Lublin, Poland

h i g h l i g h t s

g r a p h i c a l

 Surface free energy of cholesterol pellets and layers deposited on glass differ significantly.  The surface structures, pictures and images from optical and confocal microscope, SEM and profilometer, are presented.  The pellets structure are not uniform even at a large deposited amount, what reflects in their surface free energy.  The zeta potentials of silica with deposited cholesterol versus pH lie between those of bare silica and bare cholesterol.  The electrokinetic charge at pH 11 and zeta potential −50 mV of bare cholesterol is very low 0.596 ␮C/cm2 .

The electrokinetic charge at slip plane in 0.001 M NaCl, pH 11,  = −50 mV, ek = 0.596 ␮C/cm2 .

a r t i c l e

a b s t r a c t

i n f o

Article history: Received 22 June 2012 Received in revised form 23 October 2012 Accepted 24 October 2012 Available online 20 November 2012 Keywords: Cholesterol Contact angle Surface free energy Zeta potential

a b s t r a c t

The important role of cholesterol in human body is well known. Therefore its properties have been investigated in many aspects. However, still some of its surface physicochemical properties are not well understood. Among them are wetting and electrokinetic properties. Therefore the aim of this paper was to study these properties and to compare them with the published literature data. For this purpose the advancing and receding contact angles of water, formamide and diiodomethane on the surface of cholesterol pellets and on its layers deposited on glass from its solution in chloroform were measured and then the surface free energy ␥S of the layers was calculated from van Oss’ (LWAB) and Chibowski’s (CAH) models. Also the images of the surfaces from SEM, confocal and optical microscopes, and profilometer were obtained. It was found that the surface of cholesterol pellets is more hydrophobic than this of the deposited layers, even if up to 200 statistical monolayers have been deposited on glass surface. From the images it was seen that the deposited layers were not uniform and at a larger coverage the needle-like cholesterol crystals were formed on the surface. From the zeta potential of cholesterol as a function of pH in 10−3 M NaCl was found that cholesterol shows the isoelectric point at pH 2.65, which is within the range of the literature data 1.8–3. The hydroxyl ions OH– are the potential determining and probably adsorb on OH groups of cholesterol molecules. In the pH range 3–6 the literature and our results of the negative zeta potential of cholesterol are convergent, except for those measured on a cholesterol suspension precipitated in water from its solution in 1-propanol. The zeta potential of silica particles covered with statistical monolayer of cholesterol lie between those of pure silica and cholesterol suspensions. The calculated electrokinetic charge at the cholesterol surface is very low, e.g. at pH 11, where  = −50 mV, it amounts one excess negative charge (OH− ion) on the surface of 73 cholesterol molecules ( d = 0.422 ␮C/cm2 ). The results obtained in this paper partially confirm those obtained previously and bring new insight into the structures of cholesterol layers deposited on glass. © 2012 Elsevier B.V. All rights reserved.

∗ Corresponding author. Tel.: +48 81 537 5651. E-mail address: [email protected] (E. Chibowski). 0927-7757/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfa.2012.10.048

50

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

1. Introduction

2. Experimental

The important role of cholesterol in a human body is well known. Therefore its properties have been investigated for several decades in many aspects [1–13]. Chemically cholesterol is a lipid which is produced by the liver in living organisms and then transported in the bloodstream and it is recycled. A person having a weight of ca.70 kg synthesizes daily about 1 g of cholesterol, and total body content of it is about 35 g. In the liver, cholesterol is converted to bile, which is then stored in the gallbladder and excreted via the bile into the digestive tract. However, about 50% of the excreted cholesterol is reabsorbed by the small bowel back into the bloodstream. Moreover, in small amounts cholesterol is also found in plants and fungi. The molecular structure of cholesterol is following: The name ‘cholesterol’ originates from the Greek chole (bile) and stereos (solid) with the added suffix -ol for an alcohol. The name ‘cholesterine’ was used for the first time by chemist Eugène Chevreul in 1815, although it was identified much earlier in 1769 in a solid form by Franc¸ois Poulletier de la Salle in gallstones. In living cell membranes cholesterol is responsible for their fluidity and permeability. Through the interaction with the phospholipid fatty acid chains in the membranes, cholesterol increases their packing thus reducing the membrane fluidity. Simultaneously, the polar headgroups of phospholipids and sphingolipids interact with the hydroxyl group of cholesterol molecules. As a result, cholesterol plays an important structuring role regulating the membrane permeability too. Nevertheless extensive studies for many years upon cholesterol role and its properties [1–13], it is still an object of investigations [14–18] and some of its surface physicochemical properties are not well understood, among them are the wetting and electrokinetic ones. Looking at the chemical formula of cholesterol (Fig. 1), it can be seen that there is only one polar OH group, one double bond (␲ electron), and a large apolar sterol ring. Nevertheless it behaves as visibly polar substance and can orient with the polar OH group to a polar solid substrate [3]. Also its electrokinetic potential depends on the solution pH in which it is dispersed [6–9]. To our knowledge, so far there is no published paper in which both surface energy and electrokinetic potential were compared for the same sample of cholesterol. These two parameters may play crucial role in cholesterol particles adhesion to the blood vessels. Therefore it seemed to us interesting to conduct such investigations and revisit so far published results. Hopefully they would shed new light on surface properties of this biologically important compound which, unfortunately, among other things, blockes the blood vessels.

2.1. Materials

Fig. 1. Chemical structure of cholesterol.

The cholesterol (>99%; Sigma) was used without further purification. The chloroform (POCH S.A. Poland) was used as a solvent to deposit of the cholesterol layers. As the solid support microscope glass slides 38 mm × 26 mm (Comex, Poland) were used. The probe liquids for contact angle measurements were water (18.2 M cm, Millipore, from a Milli-Q System), formamide (98%; Aldrich) and diiodomethane (99%; Aldrich). The silica gel (purity p. a.) powder was purchased from Fluka A.G. The specific surface area of the silica gel determined by Brunauer, Emmet, and Teller (BET) thermal desorption of nitrogen was 7.1 m2 /g. 2.2. Methods 2.2.1. Deposition of cholesterol (Chol) on glass slides or silica powder from the chloroformic solutions The glass slides were first cleaned in methanol in an ultrasonic bath for 30 min, then in Milli-Q water ultrasonic bath, and dried in a desiccator at 100 ◦ C. The cholesterol layers were prepared by pouring progressively 0.5 mL of the chloroformic solution on the glass slides and waiting until the solvent has evaporated. To obtain onestatistical-monolayer coverage 1.7 ␮L of Chol solutions, containing 1 mg/mL, was added to 0.5 mL of the chloroform. The statistical monolayer was calculated from the geometric surface area of the glass slides and the cross section of the lipid molecule which was determined from ␲–surface A isotherm 38 A˚ 2 for Chol [4,19]. The consecutive layer was formed on the first one, already dried, by pouring the same volume of the solution. The powdered samples of SiO2 ·xH2 O of known specific surface area were precovered with calculated statistical mono- (ML) or bi- (BL) layer of Chol. Then, the obtained solutions were mixed with 10 mg of silica placed in a beaker and then put under high vacuum in a vacuum dryer for 24 h, in order to remove the chloroform. 2.2.2. Contact angle measurements The contact angles of water, formamide and diiodomethane were measured by sessile drop method using contact angle meter (GBX Co., France) equipped with a video camera and computer software (WinDrop+++). They were calculated by fitting a mathematical expression to the shape of the drop. The advancing contact angles were measured for 6 ␮L droplets settled on the surface with a help of an automatic deposition system and the receding one after sucking into the syringe 2 ␮L of the liquid from the droplet. The receding contact angle (which is smaller than the advancing one) appears if the original volume of the droplet (here 6 ␮L) is reduced. The reason for this phenomenon is the surface roughness and/or the liquid film left behind the retreated three-phase contact line of the droplet. The readings of contact angle were taken for both sides of droplet for all three probe liquids at 20 ± 1 ◦ C. 2.2.3. Surface topography investigations Surface topography of the studied samples was imagined by using: optical microscope (Eclipse E600 POL, Nikon), confocal microscope Inverted Metallurgical Microscope (MA 200M, Nikon), Scanning Electron Microscope–Focused Ion Beam (SEM–FIB) (QuantaTM 3D FEG). The MA200 Metallurgical Microscope combines captured images with data on their observation settings for more comprehensive documentation. Nikon’s DS-U2 Camera Control Unit and NIS-Elements Imaging Software allow the users to perform everything from basic image capture to the measurement, analysis, and management of captured images. The QuantaTM 3D FEG is the most versatile high-resolution, low-vacuum SEM/FIB for 2D and 3D material characterization and analysis which gives clear

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

Fig. 2. Advancing and receding contact angles of water and diiodomethane on bare glass after chloroform evaporation, on glass plates pre-covered with 1–200 statistical monolayers of cholesterol, and on cholesterol pellets.

and sharp electron imaging. It accommodates the widest range of samples of any SEM system. The surface roughnesses were measured by optical profilometer (Contour GT-K1, Veeco), which is optical surface-profiling system measuring surface topography with high accuracy from sub-nanometer up to 10 mm size. 2.2.4. Preparation of suspensions The suspensions of silica and Chol were prepared by mixing with 100 cm3 of 10−3 M NaCl solution 10 mg of bare silica or Chol, or silica powder precovered with statistical mono- (ML) or bi- (BL) layer of Chol (further 1 ML Chol/SiO2 and 1 BL Chol/SiO2 , respectively). Before measurements of zeta potential, all the suspensions were sonicated using a sonicator 3000 (Misonix) for 3 min. To avoid sample heating the ultrasonic head was operated for 2 s with following 4 s breaks at the maximum power of 3.5 W. 2.2.5. Zeta potential measurements The zeta potentials for all the suspensions were determined as a function of the time, after 5, 15, 30 and 60 min since the moment of the suspension preparation using a Zetasizer Nano ZS, Malvern. The determined a values (where  is the Debye–Hückel parameter and a is the particle radius) indicated that Helmholtz–Smoluchowski equation can be applied with good approximation for the zeta potential calculation from the electrophoretic mobility data. All the experiments were replicated 3–5 times at 25 ± 1 ◦ C and arithmetic mean values were calculated. 3. Results and discussion 3.1. Advancing and receding contact angles To get more information about cholesterol surface properties of the studied samples the advancing and receding contact angles of water, formamide and diiodomethane were measured. Fig. 2 shows the advancing and receding contact angles of highly polar water and nonpolar diiodomethane measured on eight differently prepared surfaces of cholesterol and on the reference system, i.e. the glass surface rinsed with chloroform and dried. For the sake of clarity, the contact angles of polar formamide are not shown in Fig. 2. However, they were used to calculate the surface free energies. As can be seen in Fig. 2 the largest contact angle of water was measured on the two compressed pellets of cholesterol. The advancing contact angles are practically the same

51

(ca. 77◦ ), however, the receding contact angles differ. Pellet-1 was compressed at ca. 100 kG/cm2 while Pellet-2 was compressed at ca. 1000 kG/cm2 . Application of a higher pressure causes smaller contact angle hysteresis, ca. 20◦ and 12◦ , respectively (Fig. 2). This may be a result of smaller roughness of the pellet surface, what can be expected for the sample compressed at a higher pressure. The water contact angles measured on the cholesterol layers deposited on glass do not differ much (the differences are within 5–7◦ ) showing practically the same hysteresis, ca. 13◦ , except for the very thick layer deposited from chloroform (200 statistical monolayers). The advancing contact angles are much smaller than those on the pellets and amount to 43–51◦ . It means, that in the range of 1–20 statistical Chol monolayers the wetting properties of these surfaces are similar irrespective of the deposited amount. However, the wetting properties of a very thick cholesterol layer, corresponding to 200 statistical monolayers, lie between those of the thinner layers (1–20) and the pellets. As for diiodomethane contact angles, there is no clear relationship between the contact angle and the layer thickness (Fig. 2). Again the biggest contact angle was measured on the pellet surfaces but practically the same value was found on the surface of 1- and 2-monolayer thick layers. Moreover, the contact angle of diiodomethane on the reference glass surface is actually the same as that on 5-, 10-, and 20-monolayer thick Chol layers, this indicates for similar strength of London dispersion interactions. 3.2. Surface free energy of cholesterol surface From the contact angle values alone it is difficult to conclude about kind and strengths of the surface interactions. More information can be obtained by calculating the surface free energy of the studied surfaces. The van Oss et al.’s model was used for calculations of apolar Lifshitz–van der Waals SLW and polar electron-donor S− and electron-acceptor S+ interactions [20–22]. This can be done if advancing contact angles of two polar liquids (e.g. water and formamide) and one apolar (e.g. diiodomethane) have been measured, for which the surface tension components are known. Simultaneous solution of three equations of type Eq. (1) gives the above mentioned components of solid surface free energy: WA = L (1 + cos ) = 2(SLW LLW )

1/2

+ 2(S− L+ )

1/2

+ 2(S+ L− )

1/2

(1)

These authors are aware of fact that this approach is controversial and it has been criticized in the literature [23–25]. However, as long as there is no better experimental method to determine the energy and its components, even these apparent and relative values are helpful for better understanding the energetic interactions occurring on cholesterol surface. The surface free energy and its components for cholesterol surface determined from Eq. (1) on compressed pellets have been reported in previously published papers [10–13,26]. These values together with those determined on the pellets in this paper are plotted in Fig. 3. For comparison the energy components for bare glass surfaces are also shown. As can be seen in Fig. 3 the apolar SLW (in fact London dispersion) component is well reproducible within less than 2 mJ/m2 range. Moreover, on a thick layer of cholesterol deposited on glass by evaporation of chloroform (unfortunately not well defined in the paper), SLW also fits in the range (Fig. 3). The electron-acceptor component S+ is small, below 1 mJ/m2 , what is typical for many solid surfaces. The electron-donor component S− is also small, but here the values are scattered in the range between 1.2 and 6.0 mJ/m2 . The value of S− determined by us on the pellets equals 6.0–6.2 mJ/m2 , and those obtained in the previously published papers amount 1.8

52

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

Fig. 3. Surface free energy components of glass/Chol (1), glass/Chol after 2 h contacted with water (2) [13] and cholesterol pellets ((3,4) our studies, (5) [12], (6) [11], (7) [10]).

Fig. 4. Surface free energy components of cholesterol layers deposited on glass plates.

and 1.16 mJ/m2 . However, S− on the surface of the above mentioned cholesterol layer deposited on glass was 5.37 mJ/m2 , and curiously, after 2 h contact with water it decreased to 1.76 mJ/m2 (Fig. 3) [13]. This was explained by water adsorption on cholesterol surface via hydrogen bonds with cholesterol OH electron-donor groups thus reducing S− . The above shown differences in the S− of cholesterol may also be explained by different amounts of the adsorbed water depended on the method of cholesterol samples preparation. Moreover, if the Chol surface was contacted in 0.1 M solutions of several different cations and anions for 90–120 min, the electron-donor interaction S− increased linearly up to 16.4 mJ/m2 versus the entropy of hydration of these ions [13]. However, unexpectedly big values of S− have been calculated for statistical mono-, bi-, and multi- layers of cholesterol deposited on glass plates, which were obtained by pouring cholesterol solution in chloroform on the plates. On thus produced cholesterol layers S− lies between 26.4 mJ/m2 and 47.7 mJ/m2 , while SLW remains similar to that obtained on the pellets. The electronacceptor component S+ , similarly like on the pellet surface (Fig. 3) is small, but for 2 statistical monolayers it is bigger and amounts 2.1 mJ/m2 which is a result of stronger interaction of hydrogen atom of OH group of cholesterol molecule. These values are shown in Fig. 4 and at first glance S− values might look as incredible ones. To find any explanation, the surface structure of the samples was determined and the obtained results are helpful for the values justification. Also a bit bigger S+ for 2 statistical monolayers may result from visible difference in the structure in comparison to one monolayer. This is discussed in next section.

˚ while these for monohydrate amount: a = 12.39 A, ˚ and c = 10.481 A, ˚ and c = 12.41 A, ˚ where a, b, c are the lengths of the edges b = 34.36 A, of a unit cell describing the crystal structure. In the two forms there are 8 molecules in the unit cell [27,28]. However, the crystal structure of cholesterol is complex because its molecules tend to crystallize with more than one independent molecule in the unit. The crystals tend to form double layer structures of roughly parallel molecules arranged end-for-end of total thickness 3.39 nm, which is similar for both forms [29]. Because of the presence of hydroxyl group in the molecule, it can adsorb both on polar and nonpolar surfaces. From the cell parameters one can easily calculate that on 1 nm2 (100 A˚ 2 ) surface there are statistically 2.7 OH-end groups if oriented outside. In other words, the unit cell containing four OH end-group on each of its side (bilayer) occupies 1.485 nm2 area. This is an important information for the explanation of the large values of S− obtained for cholesterol layers deposited on glass (see above) and it will be discussed in detail later.

3.3. Crystal structure of cholesterol Seeking for a justification of thus determined values both suitable literature data were found and optical and spectroscopic pictures of the produced surface structures were also obtained. First, considering the crystal structure of both anhydrous and monohydrate cholesterol, it crystallizes as triclinic. In the case of ˚ b = 34.209 A, ˚ anhydrous, the lattice parameters are: a = 14.172 A,

3.4. Cholesterol layers deposited on a solid support Abendan and Swift [14] investigated monohydrate cholesterol crystals of ca. 4–5 mm in length and 0.5–1 mm thick grown by slow evaporation of its solution in 95% ethanol. Using Atomic Force Microscopy (AFM) and Chemical Force Microscopy (CFM) they found multiplies plate-like crystals of 34 ± 5 A˚ thick, which corresponded to the bilayer thickness, and in an aqueous environment the surface was terminated with hydroxyl groups, while immersed into anhydrous methanol the crystals were terminated with alkyl groups. This reflected in the contact angle of water, which was  = 87 ± 3◦ and  = 99 ± 3◦ , as well in the adhesion force measured by CFM (in water or ethylene glycol), which amounted 4.4 ± 1.9 nN and 2.0 ± 0.9 nN, respectively. Thus, depending on the interacting liquid phase the rearrangement of cholesterol surface molecules occurs. The water advancing contact angle measured in our paper on the cholesterol pellets is 78◦ , which indicates for significant amount of OH accessible on the cholesterol surface. Moreover, the contact

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

angles on the layers deposited on glass from chloroform solution are much smaller (Fig. 2). Then Grotenhuis et al. [3] using spin-coating technique have found that at 1750 rpm on silicon wafers up to seven height levels were deposited from 7 mg/ml of cholesterol in chloroform solution with steps about 37 A˚ between them, as determined from SFM images. Basing on this the authors concluded that cholesterol molecules were oriented vertically to the surface and each layer was roughly bimolecular layer thick. They also concluded that the structure of these layers was not rigid and could be destructed easily by the SFM tip. Lafont et al. [30] found that the Langmuir monolayer of cholesterol, transferred on a freshly cleaved mica surface at a pressure of 3 mN/m, produced homogenous monolayer ˚ However, if the film was compressed film having thickness 13 ± 2 A. to 20 mN/m and then transferred on mica, the film was about 10 layers thick showing elongated faceted crystallites. Moreover, some samples of thus deposited cholesterol were 32-layer thick. As was quoted above, the unit cell of cholesterol OH end-groups occupies 1.485 nm2 area. involving 2 × 4 Taking the value 21 kJ/mol for HO· · ·H hydrogen bond strength [31] and S− = 35.5 mJ/m2 (and negligible S+ determined for 20 statistical Chol monolayers) (Fig. 3) one can easily calculate that 21 kJ/mol = 3.49 × 10−20 J/OH, and S− = 35.5 mJ/m2 = 35.5 × 10−21 J/nm2 . Hence the density of OH groups originating from cholesterol surface has to be 35.5 × 10−21 J/nm2 )/(3.49 × 10−20 J/OH) = 1.02 nm2 /OH, while four OH groups of the unit cholesterol cell occupy only 1.485 nm2 . This means that even S− = 35.5 mJ/m2 for Chol surface can be theoretically proved. It is discussed below whether this value really originates from the OH groups oriented outward on the cholesterol surface deposited on glass. Therefore, to get better insight into the structure of the investigated surfaces an optical polarizing microscope, optical profilometer, scanning electron microscope (SEM–FIB), and confocal inverted metallurgical microscope were used. 3.5. Surface structure of the investigated cholesterol samples Fig. 5A–D present pictures and replicas of the pellet surfaces. The slate-like structure of the pellet is clearly seen, especially that on the inset in Fig. 5C of SEM image of the pellet edge. The data from profilometer (Fig. 5D and 5D ) show the micrometer-sized roughness of the compressed Pellet-2 surface, which for across entire measured array 156 ␮m × 117 ␮m are: average roughness Ra = 0.877 ␮m, root mean square (rms) Rq = 1.10 ␮m, and the peak-to-valley difference calculated over the entire measured array Rt = 24.6 ␮m. These parameters mean: Ra (roughness average) is the arithmetical mean of the departures of the profile from the mean line i.e. the main height as calculated over the entire measured length or area. It is useful for detecting general variations in overall profile height characteristics. Rq (root mean square) (rms) is average between the height deviations and the mean line/surface, taken over the evaluation length/area. It represents the standard deviation of the profile heights, and Rt (maximum height of the profile) is the vertical distance between the highest peak and the lowest valley along the assessment length of the profile. These parameters along the marked vertical and horizontal lines are shown in Fig. 5D . This pellet structure reflects in the contact angle hystereses of water and diodomethane (Fig. 2). Note smaller contact angle hystereses appearing on Pellet-2 surface. This is because Pellet-2 has been compressed at a higher pressure and therefore the surface roughness is smaller. Nevertheless, the advancing contact angles are very similar on both pellet surfaces and therefore the surface free energy and its components are practically the same (Fig. 3). The small values of electron-donor S− and electron-acceptor S+

53

parameters indicate that the cholesterol molecules are oriented horizontally in the slates and hence the density of polar OH groups, which are able to form hydrogen bonding, is low. Conducting similar calculations, as shown above, and assuming the same strength of electron-donor S− and electron-acceptor S+ interactions (hydrogen bonding, Fig. 3), i.e. 21 kJ/mol, it is found that one hydrogen bond is formed on 1.97 nm2 of the pellet surface. Now, to explain the high electron-donor component S− values obtained on the cholesterol layers deposited on glass, the replicas and images of the surfaces which were obtained from SEM and profilometer should be very helpful. Fig. 6A–D show SEM images of bare glass, 1, 2 and 20 statistical monolayers of cholesterol deposited on glass surface. The bare glass surface (Fig. 6A) is really ‘clean’, while on the images showing the statistical monolayers the nonuniform coverage of the glass surface by cholesterol molecules is clearly seen. In the case of statistical mono- and bilayer there are islands (Fig. 6B and C), when 20 statistical monolayers were deposited the formed needle-like triclinic crystals still did not cover the surface uniformly. To get better insight into the surfaces structure in Fig. 7A–D are presented the replicas obtained with the help of the profilometer (the size of surfaces in Fig. 7 is 156 ␮m × 117 ␮m) together with the roughnesses parameters, Ra and Rq . For bare glass surface Ra = 0.4 nm and Rt = 8.88 nm (not shown on the picture), which indicates that there are some small protrusions (white dots in Figs. 6A and 7A) and less visible black holes on the bare glass surface. The differences in deposited cholesterol layers structure are also well seen in Fig. 7. The layers differ significantly in the roughness parameters. The Ra parameter increases from 1.96 nm (ML) to 2.88 nm (BL) and 199 nm for 20 statistical monolayers. Thus instead of an uniform cholesterol monolayer there are islands of mono and bilayer thicknesses (Fig. 7B), and in the case of statistical bilayer, the islands are of a bigger size, and though as the bilayers on the average, but because of Rt = 26.10 nm, there must be also several-bilayer-height islands (see also the chapter 3.4 above). It is worth mentioning that we have also produced the statistical mono- and bi- layer on glass surface from cholesterol solution in 1-propanol and the results were similar, the layers were not uniform in contrast to the results reported by Lafont et al. [30] that homogenous monolayer film on freshly cleaved ˚ was deposited if transferred at a mica, having thickness 13 ± 2 A, low pressure of 3 mN/m from water surface by Langmuir–Blodgett technique. Also Grotenhuis et al. [3] using spin-coating technique have found that at 2500 rpm from 1 mg/ml of chloroformic solution about 1 monolayer has been deposited on silicon wafers, as determined with the help of a scanning force microscopy (SFM), however there are no mention about the uniformity of this layer. Surprisingly, Gupta and Suresh [17] obtained uniform homogenous layer of cholesterol (16.3 A˚ thick) on hydrophobized glass surface (obtained by treating it with 1,1,1,3,3,3-hexamethyldisilazane, HMDS) by transferring the layer at 30 mN/m, while Lafont et al. [30] already at pressure of 20 mN/m the transferring the cholesterol film obtained 10 monolayers thick layer on mica surface, and even as thick as 32 layers, each of 13 ± 2 A˚ thick on some other samples. The results of our experiments showed that no uniform homogenous cholesterol film can be obtained by spreading cholesterol solution on hydrophilic glass surface (Figs. 6 and 7). As a consequence large values of the electron-donor S− components on these layers are obtained (Fig. 4), but they are about twice less than S− value for the bare glass surface. These values rather do not result from closely packed cholesterol molecules, with their OH polar groups directed outward, but from partial interaction of water and formamide molecules with the solid support (glass) surface. Even at 200 statistical monolayers deposited on the glass still some interactions with the glass surface seem to occur. However in this case they are four times weaker than on bare glass surface (Fig. 4).

54

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

Fig. 5. Pictures and replicas of cholesterol pellet surfaces: (A, B) optical microscope (Eclipse E600 POL, Nikon); (C) SEM-FIB (QuantaTM 3D FEG); (D) optical profilometer (Contour GT, Veeco). D . Image of cholesterol pellet and profiles of roughness along line X and Y.

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

Fig. 6. Pictures from confocal microscope of bare glass surface (A) and glass surface covered by ML (B), BL (C) and 20 ML of cholesterol from chloroformic solutions.

Fig. 7. Pictures from optical profilometer of bare glass surface (A) and glass surface covered by ML (B), BL (C) and 20 ML of cholesterol from chloroformic solutions.

55

56

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

Fig. 9. Zeta potential of cholesterol, silica and silica covered by statistical mono(ML) and bilayer (BL) of cholesterol as a function of pH. Fig. 8. Zeta potential of cholesterol as a function of pH: (A) our studies; (B) [9], (C) [8]; (D) [15]; (E) [6].

3.6. Zeta potential of cholesterol 3.6.1. The literature and present data Zeta potential of cholesterol suspension in water or electrolyte solution as a function of pH has already been determined several times in the published papers [6,7,9,15]. In Fig. 8 are plotted the literature values of the zeta potential and those determined in this paper (calculated from Helmholtz–Smoluchowski equation). In the papers cited above information about a value and/or which equation was used to calculate the zeta potentials is not always given. The values determined in water by Salcedo et al. [7] in 1989, and Uskokovic´ and Matijevic´ [15] in 2007, as a function of pH differ significantly and pHiep determined there is about 3 and 1.8, respectively, while pHiep determined by us amounts 2.65 (Fig. 8). The microelectrophoretic values obtained by Salcedo et al. [7] were calculated from both Helmholtz–Smoluchowski and O’Brien and White equations and were practically the same. That sample was from Carlo Erba “USP-Cod Franc” quality. The results of Uskokovic´ and Matijevic´ [15] were obtained on cholesterol 99+% Alfa Aesar sample dissolved in 1-propanol and then precipitated in water. The particle size was about 1.85 ␮m but there is no mention about the measurement and calculation of the zeta potentials. The values determined in 10−2 M NaCl (the sample anal. grade from Serva, values from Helmholtz–Smoluchowski equation) [9] and at the ionic strength I = 0.05 (“purified sample”) [15], are similar in 4–8 pH range. In the original paper [15] are given the electrophoretic mobilities from which we have calculated the zeta potentials using Helmholtz–Smoluchowski equation. At pH 9 the literature values lie between −50 mV and −85 mV, and the value determined by us is −45 mV (Fig. 8). The negative zeta potentials determined by Uskokovic´ and Matijevic´ [15] are relatively large, ca.−45 mV at pH 4 and −85 mV at pH 10. However, these values seem to be out of the range and they may be due to the method of the cholesterol suspension preparation (see above). Despite the differences in the zeta potential values determined in different laboratories and dated from 1957 to 2007, it is clearly seen that for cholesterol surface OH− are stronger potential determining than H+ ions. The hydroxyl ions probably adsorb to the OH groups of cholesterol molecules. The observed differences in the zeta potentials may results from purity of the samples, the ionic strength, suitability of the equation used for the zeta potential calculation, and quality of the apparatus as well.

3.6.2. Zeta potentials of cholesterol covered silica Fig. 9 presents zeta potentials measured for cholesterol suspension, bare SiO2 , and the silica covered with statistical monolayer (ML) or bilayer (BL) of cholesterol. The values were calculated from Helmholtz–Smoluchowski equation because from a values it was found that this equation can be used with good approximation. The particle sizes were determined using Malvern ZetaSizer. The cholesterol sample used was of 99% purity. As can be seen in Fig. 9 the negative zeta potentials of cholesterol are smaller than those of the silica. As was mentioned above the isoelectric point of the cholesterol sample was found at pH 2.65, while at this pH the zeta potential of silica is still negative −15 mV. In the pH range 3.5–7 the changes of zeta potential of silica and cholesterol run roughly parallel and those of cholesterol are about 10 mV smaller. In the alkaline environment the zeta potential of silica is more negative and at pH 10.5 the difference amounts ca. 23 mV. The deposition of 1 or 2 statistical cholesterol monolayers on the silica surface from chloroform solution causes the zeta potentials to locate between those for bare silica and cholesterol (Fig. 9). Moreover, in the case of deposited 2 ML in pH 3–8 the zeta potential of the particles is the same as that of bare silica particles. This might be due to larger patches of the uncovered silica surface and bigger cholesterol islands present when two statistical cholesterol monlayers were deposited on the surface from cholesterol solution in comparison to deposited 1 ML (Fig. 7). Generally the results of zeta potential in Fig. 9 seem to be reliable and show relatively small effect of the cholesterol presence on electrokinetic behavior of the silica particle suspended. The zeta potential of cholesterol particles at pH 11 amounts −50 mV (Fig. 9). Using this value one can calculate the electrokinetic chare at the slip plane. At 25 ◦ C it can be calculated from following equation [32–34]: √  d = −11.73 C · sinh(0.0195z) (2) where: C – [mol/dm3 ];  – [mV]:  d – ␮C/cm2 . Here C = 10−3 M NaCl + 10−3 NaOH, because the pH 11.  d = 0.5964 ␮C/cm2 → 0.5964 × 10−20 C/nm2 → 0.0372 Hence (OH– exc )/nm2 → 1 (OH– exc )/27 nm2 . Because the surface area of 1 cholesterol molecule amounts 0.37 nm2 [28,29], therefore only about one excess electrokinetic negative charge is present in the slip plane at 73 cholesterol molecules (∼1 (OH– exc )/73 Chol molecule). This is depicted in Fig. 10. It should be kept in mind that the electrokinetic charge is only a small part of the countercharge that may be present at the surface and  ek amounts a few ␮C/cm maximally [33,34]. Moreover, Lyklema [34] concluded that both at uncharged (e.g. at zero point of

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

57

Operational Program (contract no. POIG.02.01.00-06-024/09 Center of Functional Nanomaterials). We very much acknowledge financial support from Polish Ministry of Science and Higher Education, project no. N N204 272839.

References

Fig. 10. Schematic representation of the electrokinetic charge at the slip plane cholesterol/10−3 NaCl solution.

charge) hydrophilic and hydrophobic surface exists stagnant layer which is principally formed by adjacent water and the concentration of counterions in the layer is very low and actually they have negligible effect on the averaged stacking of water molecules. However, the presence of such layer cannot be detected electrokinetically. Based on these conclusions it can be expected that at the cholesterol/electrolyte solution interface a water layer is present which may possess a layer-like stacking with changing density (maxima and minima) where slip plane is located. Hence the calculated electrokinetic charge (Fig. 10) is really very low despite the relatively high negative zeta potential. 4. Summary and conclusions Cholesterol possesses a weakly polar surface exposing electrondonor and some electron-acceptor interactions (the hydrogen bonding) originating from OH group present in the molecule. Cholesterol deposited from chloroform solution on glass surface after the chloroform evaporation does not form any homogenous mono- or bilayer. It forms islands which are of bilayer or multilayer thicknesses. Even if a large amount of cholesterol is deposited on glass surface from chloroform solution the formed layer is not homogenous. It is built up of triclinic needle-like crystals. These structures reflect in the surface free energy of the layers which is much larger than that of the surface of compressed cholesterol pellet. The literature zeta potential values of cholesterol suspensions are convergent in the pH range 3–6 and the pHiep lies in between pH 1.8–3. The hydroxyl ions OH are potential determining ones, probably interacting with OH groups originating from the cholesterol molecules. The changes of negative zeta potential of silica and cholesterol as a function of pH run in a similar way. However, the values for cholesterol are smaller by 10–20 mV. The zeta potential of silica particles is affected by cholesterol adsorption and its value depends on the deposited amount of cholesterol, but surprisingly, it is affected more by the amount corresponding to one statistical monolayer than the bilayer. This is because of different island-like structures of these coverages.These results shed also a light on the mechanism of cholesterol-rich lipids (fat) deposition on the wall of a blood vessel, which showed that cholesterol prefers to deposit as islands than uniformly, thus developing plaque, causing a bulge that narrows the lumen of the artery and forming a connective tissue cap (scar tissue) over the developing plaque. Acknowledgments The research was carried out with the equipment purchased thanks to the financial support of the European Regional Development Fund in the framework of the Polish Innovation Economy

[1] M. Craven, Crystal structure of cholesterol monohydrate, Nature 260 (1976) 727–729. [2] H.-S. Sheih, L.G. Hoard, C.E. Nordman, A structure of cholesterol, Acta Crystallogr. B37 (1981) 1538–1543. [3] E. ten Grotenhuis, W.J.M. van der Kemp, J.G. Blok, J.C. van Miltenburg, J.P. van der Eerden, Scanning force microscopy of cholesterol multilayers prepared with the spin-coating technique, Colloids Surf. B 6 (1996) 209–218. [4] S. Lafont, H. Rapaport, G.J. Sömjen, A. Renault, P.B. Howes, K. Kjaer, J. AlsNielsen, L. Leiserowitz, M. Lahav, Monitoring the nucleation of crystalline films of cholesterol on water and in the presence of phospholipid, Phys. Chem. B 102 (1998) 761–765. [5] H. Ohvo-Rekilä, B. Ramstedt, P. Leppimäki, J.P. Slotte, Cholesterol interactions with phospholipids in membranes, Progr. Lipid Res. 41 (2002) 66–97. [6] H.W. Douglas, D.J. Shaw, Electrophoretic studies on model particles. Part 1. Mobility, pH and ionic strength relations for droplets having protein, lipid or polysaccharide surfaces, and for certain complexes, Trans. Faraday Soc. 53 (1957) 512–522. [7] J. Salcedo, A. Delgado, F. Gonzàlez-Caballero, Electrokinetic and stability properties of cholesterol in aqueous NaCl and NaCl + bile salt solutions, Progr. Colloid Polym. Sci. 79 (1989) 64–69. [8] J. Salcedo, A. Delgado, F. Gonzàlez-Caballero, The effect of pH on the electrical properties of the cholesterol/water interface. I. ␨ potential, J. Colloid Interface Sci. 133 (1989) 278–281. [9] F. Carrique, J. Salcedo, V. Gallardo, A.V. Delgado, An experimental study of the stability of cholesterol in inorganic electrolyte and bile salt solutions, J. Colloid Interface Sci. 146 (1991) 573–575. [10] M.L. Kerkeb, F. Gonzàlez-Caballero, B. Janczuk, T. Białopiotrowicz, Components of surface free energy of cholesterol in the presence of bile salts, Colloids Surf. 62 (1992) 263–272. ´ M.L. Kerkeb, T. Białopiotrowicz, F. Gonzàlez-Caballero, Surface free [11] B. Janczuk, energy of cholesterol and bile salts from contact angles, J. Colloid Interface Sci. 151 (1992) 333–342. [12] M.L. Kerkeb, F. Gonzàlez-Caballero, E. Chibowski, Changes in cholesterol surface free energy components due to its phase transition, J. Colloid Interface Sci. 159 (1993) 439–443. [13] E. Chibowski, M.L. Kerkeb, F. Gonzàlez-Caballero, Effect of inorganic ions on changes in the energetic heterogeneity of the cholesterol surface, Langmuir 9 (1993) 2493–2495. [14] R.S. Abendan, J.A. Swift, Surface characterization of cholesterol monohydrate single crystals by chemical force microscopy, Langmuir 18 (2002) 4847– 4853. ´ E. Matijevic, ´ Uniform particles of pure and silica-coated choles[15] V. Uskokovic, terol, J. Colloid Interface Sci. 315 (2007) 500–511. ´ Insights into morphological nature of precipitation of cholesterol, [16] V. Uskokovic, Steroids 73 (2008) 356–369. [17] R.K. Gupta, K.A. Suresha, AFM studies on Langmuir-Blodgett films of cholesterol, Eur. Phys. J. E 14 (2004) 35–42. [18] M. Golabek, M. Jurak, L. Holysz, E. Chibowski, The energetic and topography changes of mixed lipid bilayer deposited on glass, Colloids Surf. A 391 (2011) 150–157. [19] A. Szcze´s, M. Jurak, E. Chibowski, Stability of binary model membranes–prediction of the liposome stability by the Langmuir monolayer study, J. Colloid Interface Sci. 372 (2012) 212–216. [20] C.J. van Oss, R.J. Good, M.K. Chaudhury, The role of van der Waals forces and hydrogen bonds in hydrofobic interactions between biopolymers and low energy surfaces, J. Colloid Interface Sci. 111 (1986) 378–390. [21] C.J. van Oss, R.J. Good, M.K. Chaudhury, Additive and nonadditive surface tension components and the interpretation of contact angles, Langmuir 4 (1988) 884–891. [22] C.J. van Oss, M.K. Chaudhury, R.J. Good, Interfacial Lifshitz–van der Waals and polar interactions in macroscopic systems, Chem. Rev. 88 (1988) 927–941. [23] J. Lyklema, The surface tension of pure liquids. Thermodynamic components and corresponding states, Colloids Surf. A 156 (1999), 413-412. [24] D.Y. Kwok, D. Li, A.W. Neumann, Evaluation of Lifshitz–van der Waals/acid-base approach to determine interfacial tensions, Langmuir 10 (1994) 1323–1328. [25] D.Y. Kwok, The usefulness of the Lifshitz–van der Waals/acid-base approach for surface tension components and interfacial tensions, Colloids Surf. A 156 (1999) 191–200. [26] B. Janczuk, M.L. Kerkeb, F. Gonzàlez-Caballero, E. Chibowski, Wettability of cholesterol by bile salt solutions, Langmuir 10 (1994) 1012–1017. [27] H.-S. Shieh, L.G. Hoard, C.E. Nordman, The structure of cholesterol, Acta Crystallogr. B 37 (1981) 1538–1543. [28] B.M. Craven, Crystal structure of cholesterol monohydrate, Nature 260 (1976) 727–729. [29] H.-S. Shieh, L.G. Hoard, C.E. Nordman, Crystal structure of anhydrous cholesterol, Nature 267 (1977) 287–289.

58

L. Holysz et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 440 (2014) 49–58

[30] S. Lafont, H. Rapaport, G.J. Sömjen, A. Renault, P.B. Howes, K. Kjaer, J. AlsNielsen, L. Leiserowitz, M. Lahav, Monitoring the nucleation of crystalline films of cholesterol on water and in the presence of phospholipid, J. Phys. Chem. B 102 (1998) 761–765. [31] J. Emsley, Very strong hydrogen bonds, Chem. Soc. Rev. 9 (1980) 91–124.

[32] J. Lyklema, Fundamentals of Interface and Colloid Science, V.II, Academic Press, San Diego, 2001, p. 3.21. [33] J. Lyklema, Molecular interpretation of electrokinetic potentials, Curr. Opin. Colloid Interface Sci. 15 (2010) 125–130. [34] J. Lyklema, Surface charges and electrokinetic charges: distinctions and juxtapositionings, Colloid Surf. A 376 (2011) 2–8.