Comparison of Surfactants Used to Prepare Aqueous Perfluoropentane Emulsions for Pharmaceutical Applications

Abstract

Perfluoropentane (PFP), a highly hydrophobic, non-toxic, non-carcinogenic fluoroalkane, has generated much interest in biomedical applications, including occlusion therapy and controlled drug delivery. For most of these applications, the dispersion within aqueous media of a large quantity of PFP droplets of the proper size is critically important. Surprisingly, the interfacial tension of PFP against water in the presence of surfactants used to stabilize the emulsion has rarely, if ever, been measured. In this study, we report the interfacial tension of PFP in the presence of surfactants used in previous studies to produce emulsions for biomedical applications: polyethylene oxide-co-polylactic acid (PEO-PLA, and polyethylene oxide-co-poly-ε-caprolactone (PEO-PCL). Since both of these surfactants are uncharged diblock copolymers that rely on the mechanism of steric stabilization, we also investigate for comparison’s sake use of the small molecule cationic surfactant cetyl trimethyl ammonium bromide (CTAB), and the much larger protein surfactant bovine serum albumin (BSA). The results presented here complement previous reports of the PFP droplet size distribution, and will be useful for determining to what extent the interfacial tension value can be used to control the mean PFP droplet size.

Introduction

The unique physical, chemical and biological properties of perfluoropentane (PFP – see Figure 1) have generated much interest in its use in several biomedical applications such as propellants for pressurized metered dose inhalers (pMDIs) for direct delivery of drugs to lungs^{1, 2}, ultrasound contrast enhancement³, and for transport of oxygen in vivo in artificial blood^4–7. PFP also shows promise in the field of drug delivery and gene transfection^8–11. Like almost all perfluorocarbons, PFP has a high capacity for absorbing oxygen. More importantly, the normal boiling point of PFP lies between room temperature and body temperature. This means that PFP can be injected in the form of liquid droplets dispersed in an aqueous medium, and then converted to bubbles using ultrasound. This conversion to bubbles can be used to release anti-cancer drugs such as doxorubicin at a site of interest¹¹, or to occlude blood vessels that supply oxygen to tumors¹².Obviously, the size of the blood vessels that can be occluded in such a way depends on the PFP droplet size in the injected emulsion. In addition, for passive targeting of hydrophobic drugs in PFP to tumors via the Enhanced Permeability and Retention (EPR) mechanism, the PFP droplets must have the correct size range (10 – 100 nm) so that they only pass through the walls of blood vessels that supply tumors¹¹. Thus for both of the applications, the ability to control the mean PFP droplet size is of paramount importance. The interfacial tension value, as determined by the surfactants employed, is most likely an important determinant.

Figure 1.

Chemical structures of (a) PFP, (b) PEO-PLA, (c) PEO-ε-PCL and (d) CTAB

Figure 1(a) shows the chemical structure of PFP (C₅F₁₂, Molecular weight 288g/gmol), a member of the perfluorocarbon family. PFP is a liquid at room temperature but is a vapor at body temperature with a normal boiling point of 29.2 °C (Fluoromed, L. P. Product Information). At 25 °C its density of 1.63 g/ml is much greater than that of water and its kinematic viscosity is much lower than that of water, at 0.4 cSt.^13–15 As a result of its non-polar, hydrophobic nature, the solubility of oxygen in PFP at 25 °C (mole fraction 5.8 × 10⁻³) is much higher than in water (mole fraction 2.3 × 10⁻⁵) (Fluoromed, L. P. Product Information). PFP has also proven to be a non-toxic, non-carcinogenic biocompatible material^{5, 7, 10}.

Due to the hydrophobic nature of PFP, it has a very high interfacial tension against water and does not readily disperse or dissolve in hydrophilic fluids. Hence, lowering the interfacial tension of PFP with surfactants is important for improving its dispersion in aqueous media. For delivery of doxorubicin in PFP droplets to tumors, Rapoport et al.¹¹ used the surfactants polyethylene oxide-co-polylactic acid (PEO-PLA, L-form) or polyethylene oxide-co-poly-ε-caprolactone (PEO-PCL), which are uncharged block copolymers. The hydrophobic block (PLA or PCL) is thought to adsorb on the PFP surface, and the hydrophilic PEO block is thought to form a “brush” that stabilizes the droplets against coalescence.

In contrast, cetyl trimethyl ammonium bromide (CTAB) is a well known cationic surfactant with a relatively low molecular mass. The chemical structures of the surfactants are shown in Figure 1 (b–d). Above a certain concentration known as the critical micelle concentration (CMC) these surfactants form spherical core-shell type micelles¹³. Typical CMC values for CTAB and BCP surfactants of similar molecular weights, from literature, are shown in Table 1.

Table 1.

Molecular weights of the surfactants used in this study and CMC values from literature for CTAB and for BCP surfactants of similar molecular weights.

Surfactant	Molecular weight	PDI	Literature CMC value in Distilled Water (mM)
PEO114-PLA65	9700	1.04	0.0071–0.0092a
PEO45-PCL23	4600	1.15	0.00014–0.0074b
CTAB	365	1.00	0.76–0.92c
BSA	66000	1.00	0.000076d

^aReference ¹⁴

^bReference ¹⁵, ¹⁶

^cReference ¹⁷–²¹

^dReference ²²

Bovine serum albumin (BSA) is a negatively charged globular protein and the most abundant protein in mammalian serum (4 weight % in human serum). It possesses a prolate ellipsoid tertiary structure with dimensions 4×4×14nm. BSA has been shown to adsorb and denature at air-water interface and reach an equilibrium surface tension of ~53 dyn/cm³⁸. Kripfgans et al studied PFP-BSA emulsion stability and droplet size distribution, for occlusion therapy and ultrasound contrast applications¹². They found that the PFP droplets in BSA solutions were stable against spontaneous vaporization at physiological temperature (37 °C) and ultrasonic scanning up to certain pressure thresholds and that their mean diameter and number density could be controlled by altering the concentration of BSA.

Due to their high molecular weights, BCP surfactants such as PEO-PLA and PEO-PCL possess a significantly lower CMC than the smaller, non-polymeric surfactant CTAB^14–21.

The biocompatibility and biodegradability of PEO-PLA and PEO-PCL have been exploited in several biomedical applications such as bio-resorbable sutures, orthopedic screws, coating materials, implants and drug delivery ^{14, 23–33}. Micelles formed by PEO-PLA and PEO-PCL are of particular interest because they can be used to sequester anti-cancer drugs, which are then delivered to tumors using PFP droplets¹¹. CTAB, a cationic surfactant, has been used in several cosmetic products, topical antiseptic creams and recently for the isolation of DNA from solutions containing other polysaccharides^{34, 35}.

No critical micelle concentration can be defined for BSA since it is not an amphiphilic surfactant. However, an apparent critical micelle concentration of ~0.076 µM is reported in literature²². BSA has been studied as a potential emulsifier for PFP for occlusion therapy and ultrasound contrast enhancement¹².

The interfacial behavior of hydrogenated or oxygenated fluoroalkanes in aqueous solutions containing low molecular weight surfactants, Pluronics^{1, 2}, and phospholipids^{9, 10, 36, 37} has been documented. However, the interfacial tension of perfluropentane in aqueous solutions containing surfactants of any type has not, to our knowledge, been reported in literature. Thus the goal of this work is to compare the interfacial tension of the PFP/water interface in the presence of the surfactants used in recent studies^11–12 to stabilize PFP emulsions for biomedical or pharmaceutical use.

Experimental section

Materials

Research grade PFP was purchased from Fluoromed L.P. and stored in the refrigerator at 0°C in a sealed container. PEO₁₁₄-b-PLA₆₅ diblock copolymer (Table 1) was purchased from Polymer Source Inc., Canada in a crystalline form. Because PEO-PLA crystals do not readily dissolve in water¹³, the aqueous solution of this BCP was prepared following a multi-step process.¹¹ First, 0.125 grams of the polymer were dissolved in 5 ml of THF with 20 ml of distilled water. Next, the organic solvent was removed by running the solution through a membrane tube with a molecular weight cut-off of 3500 Da (Spectra/Por, Spectrum Labs Inc., GA) followed by dialysis with water and 0.15M Dulbecco’s Phosphate Buffer Solution (PBS). Distilled water was used to replace the volume of the organic solvent removed during the previous step to obtain the final 0.5% by weight PEO-PLA solution. The solution was allowed to homogenize by gentle agitation and stored in the refrigerator when not in use. The solutions were diluted with 0.15M PBS to get desired concentrations. PEO₄₅-b-PCL₂₃ diblock copolymer (Table 1) was also obtained from Polymer Source Inc., Canada and micellar solutions were prepared in a similar fashion.

Analytical grade CTAB was purchased from Sigma. A 1mM solution of CTAB in 0.15M PBS was prepared and mixed by gentle agitation for 1–2 minutes and stored at room temperature.

Purified BSA (RIA and ELISA grade) was purchased from Calbiochem (San Diego, USA). Solutions were prepared in 0.15M PBS, stored at ~0 °C and used within 24 hours of preparation.

Surface and Interfacial Tension Measurements

Gas tight glass syringes (250µl), flat-tip stainless steel needles (14 gauge) and a two-way stainless steel valve were purchased from the Hamilton Company.

The syringe, needle and valve were thoroughly cleaned before use and dried. The valve was used to control the flow of PFP from the syringe to the needle and control the size of the liquid drops. A glass environmental cell was cleaned thoroughly with distilled water and allowed to dry completely. For interfacial tension measurements, this cell was filled with surfactant solution.

The needle was then placed into the airtight glass cell as shown in Figure 2. For interfacial tension measurements, the needle tip was submerged in surfactant solution and the temperature in the cell was constantly monitored using a wire thermocouple in direct contact with the solution.

Figure 2.

Experimental apparatus for surface/interfacial tension measurements. A: Light source, B: Syringe, C: Two-way valve, D: Air tight glass cell, E: Microscope, F: CCD Camera, G: Temperature Monitor and H: Computer

All measurements were performed at room temperature (24 °C ± 1 °C).

PFP drops were injected into the surfactant solution and images were captured using the CCD camera (Fig. 3.). The first image was captured 5–10 seconds after the formation of the drop. The images were then converted into pixel coordinates, then into coordinates in centimeters using Image 1.37 software.

Figure 3.

Image of a PFP drop in PEG-PLA surfactant solution captured by the CCD camera

The images were next analyzed using an Axisymmetric Drop Shape Analysis-Profile (ADSAP) program³⁸ that fits the drop coordinates to the Young-Laplace equation (Y-L equation) of capillarity.

The Y-L equation as described here is specifically for pendant drops and is derived by a force-balance over the suspended drop. This equation relates the pressure difference Δp across an interface between two fluids to the curvature of the interface, caused by the phenomenon of surface tension.

$$\mathrm{\Delta }\mathrm{p}=\mathrm{\gamma }(1/{\mathrm{R}}_1+1/{\mathrm{R}}_2)$$ (1)

where γ is the surface/interfacial tension and R1, R2 are the principle radii of curvature. The ADSAP program uses numerical integration to develop theoretical drop profiles to match the experimental drop profiles (Fig. 4.). Surface/Interfacial tension, drop volume, drop surface area and the radius of curvature of the drop were obtained by running the ADSAP program.

Figure 4.

Comparison of theoretical and experimental interface profile of a drop obtained through ADSAP

Results and Discussion

Bond Numbers

The Bond number (Bo) is a dimensionless group which represents the ratio between gravitational force and surface/interfacial tension forces acting on a pendant drop^{39, 40} and is equal to:

$$\text{Bo}=\mathrm{\Delta }\mathrm{\rho }{*\mathrm{g}*\mathrm{R}}_{\mathrm{o}}{}^2/\mathrm{\gamma }$$ (2)

where Δρ is the density difference at the surface or interface, g is the acceleration due to gravity, R_o is the radius of curvature of the drop at the apex and γ is the surface or interfacial tension.

Spherical drops are obtained when this ratio is small.

In the presence of surfactants the drop assumes a pendant like ellipsoidal shape rather than a spherical shape due to adsorption of molecules at the interface resulting in a decrease in interfacial tension and an increase in the Bond number.

For an accurate pendant drop experiment it is necessary that the drops have a pendant like or ellipsoidal shape rather than a spherical shape. For spherical drops, the fit between the experimental and theoretical drop profile becomes independent of the surface/interfacial tension values^{38, 39}. This occurs when Bo << 0.1.

Bond numbers were calculated for the pendant drops for all the systems studied. The average values are listed in Tables 2 and 3. For accurate results the Bond number should lie between 0.1 and 0.6⁴⁰. For all our systems, the Bond numbers for the pendant drops were within this range.

Table 2.

Literature and experimental values of equilibrium surface tensions of surfactants at 25°C measured in this study.

Material	Concentration (mM)	Solvent	Equilibrium Surface Tension (dyn/cm)	Literature Value (in distilled water)	Bond Number
PFP (99% purity)	--	--	9.41	9.42a	0.42
CTAB	1 1 0.4	Distilled Water 0.15M PBS 0.15M PBS	29 27 27	35–38b	0.46 0.41
PEO-PLA	0.15 0.26	0.15M PBS	43 47	45–46c	0.41 0.41
PEO-PCL	0.54	0.15M PBS	47	49–52d	0.39
BSA	0.015	0.15 M PBS	43	53e	0.40

^aReference ⁴⁸–⁵¹,

^bReference ¹⁷, ²¹,

^cReference ⁴¹, ⁵²,

^dReference ⁴⁵, ⁴⁶,

^eReference ³⁸, ⁴⁷ Average standard deviation in measured surface tensions is 1 dyne/cm.

Table 3.

Interfacial tensions of PFP against water or 0.15M PBS containing various surfactants at the concentrations given in the table.

Interface	Surfactant	Surfactant Concentration (mM)	Interfacial Tension (dyn/cm) (±1–4 dyn/cm)	Bond Number
PFP-PBS	--	--	49	0.14
PFP-Water	--	--	54 56a	0.15
PFP-PBS	PEO-PLA	0.0103 0.052 0.103 0.26	38 31 29 27	0.17 0.15 0.17 0.18
PFP-PBS	PEO-PCL	0.54	30	0.17
PFP-PBS	CTAB	0.05 0.1 0.2 0.4 0.6 0.8 1	17.7 15.2 12.6 8.5 10 10.7 10	0.42 0.44 0.46 0.52 0.48 0.46 0.58
PFP-PBS	BSA	0.015 0.045 0.075 0.152 0.227	31.1 33.4 28.0 31.9 30.8	0.58 0.58 0.52 0.56 0.56

^aReference ⁵³

The highest Bond numbers were obtained for systems involving CTAB surfactants, which were also the systems with the lowest surface/interfacial tension values.

Surface Tension of PFP at Liquid/Air Interface

The surface tensions of PFP, distilled water, 0.15M PBS, and various surfactant solutions were all measured against saturated air at room temperature. These measurements, which are not novel, are compared to values reported by previous authors as a check on our materials and experimental techniques. The measurements were repeatable with a low standard deviation (Table 2). Table 2 lists the measured equilibrium surface tension of the surfactant solutions (in PBS buffer) as well as literature values for solutions containing the same or similar (for the BCP) surfactants at their critical micelle concentrations in distilled water. Our measured values are in reasonable agreement with the literature values, with the possible exception of the BSA solution. As reported previously,³⁹ the surface activity of proteins often varies with the supplier and also changes after repeated freeze-thaw cycles. The surface tension of distilled water (68 dyn/cm) was also found to be a little lower than the literature value of 72 dyn/cm, which we attribute to accumulation of contaminants in water over the course of the experiment or fluctuations in room temperature. The lowest surface tension value is obtained using the surfactant CTAB, which is unsurprising because it has the highest CMC value. As mentioned in Table 1, the literature values for the CMCs of PEO-PLA and PEO-PCL BCPs in aqueous solutions are very low^{14, 15, 24, 26, 29, 41–46}. CTAB, on the other hand, has a CMC value which is at least two orders of magnitude larger. Above the CMC, the surface tension becomes approximately independent of concentration, presumably because the concentration of unimers stays close to the CMC value, and unimers are much more surface-active than micelles. Hence, even though our surfactant solution bulk concentrations were greater than the CMC value, the equilibrium surface tension values are nearly the same as the values reported at the CMC (Table 2). Similarly, for BSA, an apparent CMC of ~0.076µM has been reported in the literature²² with a constant equilibrium surface tension of 53–55mN/m at higher concentrations^{38, 47}.

Interfacial Tension of PFP against aqueous solutions

In the absence of any surfactants, the interfacial tension of PFP against pure distilled water was measured to be 54 dyn/cm, in excellent agreement with the literature value of 56 dyn/cm ⁵³. Against 0.15M PBS, the measured value was 49 dyn/cm. These very high values reflect the hydrophobicity of PFP, and hence surfactants are needed to disperse PFP in water.

In the presence of the block copolymer surfactants, the interfacial tension of PFP against water was measured in our lab to be about 27 dyn/cm for PEO-PLA, and about 30 dyn/cm for PEO-PCL (Table 3). All concentrations studied were above the CMC of the block copolymers, which probably explains why no concentration dependence is observed in Figure 5.

Figure 5.

Interfacial tension between 0.15M PBS and PFP as a function of bulk surfactant concentrations at 24°C. (●)PEO-PLA, (▲) BSA and (■) CTAB. The vertical arrow gives the location of the literature CMC value for CTAB; the CMC values for BSA and PEO-PLA are both below 0.01 mM. Note: Some error bars are not visible as they are smaller than or equal to the size of the symbols used due to very small standard deviations (< 1 dyn/cm)

In the presence of BSA, the interfacial tension of PFP against water was measured as about 31 dyn/cm (Table 3) at all concentrations studied. In the presence of CTAB, a cationic surfactant, the interfacial tension of PFP against water reaches its lowest value, about 11 dyn/cm, at a concentrations about one-half of the reported CMC value (Figure 5). As with the surface tension, the differing IFT behavior between CTAB on one hand and BSA, PEO-PLA, and PEO-PCL on the other hand can be explained by the much larger CMC value of CTAB. In all the surfactant solutions, the unimer concentration should be close to the CMC value, and this is at least two orders of magnitude larger for CTAB than for the macromolecular surfactants. Thus more surface-active species are available for adsorption on the PFP surface when in contact with the CTAB solution. Given the very low CMC values of PEO-PLA and PEO-PCL, they are surprisingly effective in lowering the tension of the PFP/water interface. This suggests that the molecular size of the adsorbed species is also an important consideration. Globular BSA has dimensions of 4 nm by 4 nm by 14 nm⁵⁴, spherical random coils of PEO-PLA have a hydrodynamic radius of about 35nm¹³, whereas the fully-extended CTAB tail has a length of 2.2nm and width of 0.3–0.5nm⁵⁵.

Significance to PFP Pharmaceutical Emulsion Preparation

Aqueous emulsions containing perfluorocarbons such as PFP^{11, 12}, perfluorobutane ⁵⁶ or decafluoropentane ⁵⁷ suitable for biomedical applications such as occlusion and drug delivery have been prepared using the surfactants PEO-PCL, PEO-PLA, BSA, and sodium cholate^{11, 12, 57} without knowledge of the interfacial tension value. The results presented here show that all three of the macromolecular surfactants give similar values for the IFT value at high surfactant concentrations, approximately 30 dyne/cm. Perhaps this is the IFT value which produces the optimum drop size distribution (10–100 nm for EPR¹¹), a proposition that we are checking by preparing PFP emulsions using CTAB as a surfactant. The results presented here also can be used to infer the minimum amount of surfactant that can be used to stabilize PFP emulsions. For example, Kripfagans et al has reported using BSA concentrations of 0.015–0.24 mM (~1–16 mg/ml)¹². The results in Figure 5 suggest that this is excessive, because the IFT reaches its asymptotic value at a BSA concentration below 0.015 mM.

Conclusions

The very high interfacial tension value of PFP against water or PBS can be substantially lowered by all the surfactants studied: the BCP surfactants PEO-PLA and PEO-PCL, the low molecular weight cationic surfactant CTAB, and the anionic protein surfactant BSA. All of the macromolecular surfactants have been used previously to prepare stable PFP emulsions, whereas molecular-size CTAB has not. All of the macromolecular surfactants give a similar value for the PFP/water tension (≈30 dyne/cm) which is independent of concentration, at least above very low reported CMC values. CTAB, which has a CMC value that is orders of magnitude larger, gives a limiting value for the tension of the PFP/water interface that is lower (10 dyne/cm). Given its low CMC value, PEO-PLA is surprisingly effective in lowering the tension of the PFP/water interface, probably because of its colloidal size. Furthermore, adsorption of the BCP likely gives rise to a PEO “brush” that stabilizes PFP droplets against coalescence or protein adsorption. The reduction in the surface tension of PBS buffer (against air) by PEO-PLA is also large. This may be relevant to biomedical applications in which the PFP liquid droplet is converted to a vapor within the body using ultrasound¹¹, and the PFP vapor contains a significant mole fraction of oxygen due to dissolution of oxygen from the surrounding aqueous solution into the PFP bubble⁵⁸.