Apparent contact angle of oleic acid and triolein on a reverse osmosis membrane in SC-CO2 environment
Graphical Abstract
Introduction
Reverse osmosis (RO) polymeric membranes can be used in combination with supercritical carbon dioxide (SC-CO2) as a solvent for the separation of complex liquid mixtures. By dissolving into the liquid mixture, SC-CO2 reduces its viscosity, allowing the membrane to fractionate viscous mixtures, which would otherwise not be possible at ambient pressure. In addition, the membrane adds a fractionation step to separate complex mixtures, which would not be possible with SC-CO2 alone when the solubilities of their components in SC-CO2 are similar. Despite the potential applications of this coupled SC-CO2-membrane technology, only a limited number of studies have been reported using this approach to separate lipids from other components such as triacylglycerols (TAG) from fatty acid (oleic acid or linoleic acid) [1], TAG from fatty acids and/or ethyl esters [2], squalene from oleic acid [3], oleic acid from α-tocopherol/β-sitosterol and TAG from oleic acid [4]; to separate essential oils from SC-CO2 such as nutmeg [5], [6], lemongrass [6], [7], orange [6], limonene [8], and patchouli [9]; and to separate/fractionate fatty acids from vegetable oils such as macauba oil [10], [11]. The separations observed were largely due to the affinity of the membrane to specific components in the mixture and the permeability of the membrane under the conditions used. Permeability of SC-CO2 across the membrane was dependent mostly on the type of membrane used and increased with increasing transmembrane pressure drop [3], [6], [7], [8], [10]. As a general rule, it has been observed that higher SC-CO2 permeabilities corresponded to lower retention factors [6].
When fractionation of lipids was attempted with these membranes, it was seen that free fatty acids (FFA) permeate preferentially over TAG [1], [2], [4] or α-tocopherol and β-sitosterol [4]. In the fractionation of a model mixture of squalene/oleic acid, a significant enrichment of squalene in the permeate side was achieved both with a polydimethyl siloxane (PDMS) and a polyamide (AD type) membrane [3]. Considering that RO membranes can retain salt (NaCl with MW = 58 g/mol) in conventional operations, it is interesting that much larger lipid molecules (for example, oleic acid, C18H34O2 with MW = 282.5 g/mol or squalene, C30H50 with MW = 410.7 g/mol) can permeate through these very tight membranes in the presence of SC-CO2. This is mainly due to the swelling of the polymer as a result of the solubilization of CO2 in the polymeric layer. RO membranes used in SC-CO2 fractionations have been characterized before and after exposure to SC-CO2 using various techniques [10], [12], [13], [14]. Depending on the specific type of membrane, differences have been observed on the membrane surfaces, demonstrating that the exposure to SC-CO2 modifies the membrane, impacting the separations observed.
Despite the reports on the characterization of membranes before and after contact with SC-CO2, no studies are available that describe the contact between the lipid mixture saturated with SC-CO2 (drop/lipid phase), SC-CO2 (surrounding medium) and the membrane surface at supercritical conditions. However, contact angle (CA) measurements using water have been performed under ambient conditions to evaluate the changes in membranes after treatment with SC-CO2, where depending on the membrane and the SC-CO2 conditions used, CA of water increased in most cases, indicating an increase in hydrophobicity after exposure of the membrane to SC-CO2 [10], [12], [14]. The exception is a polyamide membrane (SG type from General Osmonics Inc.), for which either no modification or a slight reduction in CA was observed, in part due to its cross-linked structure [12], [14]. In addition, whether the membrane was subjected to pretreatment to remove any residual impurities or not would also have an impact on such results. There is a lack of information about the phenomena occurring while the membranes are being exposed to pure or saturated SC-CO2 with target compound(s) during processing under high pressure, which could be influenced not only by structural modifications, but also by surface energy variations that may result in higher or lower affinity of the membrane towards the target compound(s). All of these modifications occurring in the polymeric surface of the membrane will alter its affinity towards the different components present in the mixture to be fractionated. The lower is the contact angle, the more affinity the membrane has to that component and the more easily can that component permeate through the membrane. It is then expected that the larger is the difference between the contact angles of two individual components on the surface of a membrane, the larger the separation factor will be, resulting in a better separation.
When the solid surface where the drop is deposited is not ideal (e.g., has roughness), the angle formed is referred to as the apparent contact angle (CAapp) [15]. The interactions between the mixture components and the membrane under SC-CO2 environment can be studied by determining the CAapp of a drop on the membrane surface and by the spreading coefficient (SL/S) of the liquid (L) on the solid (S) surface. When a liquid is in contact with a surface, it forms an angle θ with that surface, which is the result of the mechanical equilibrium among the three interfacial tensions, e.g., the liquid-gas interfacial tension (γLG), the solid-gas interfacial tension (γSG), and the liquid–solid interfacial tension (γSL), as related through the Young’s equation [15]:
Contact angles have been measured under high pressure for different liquids on various surfaces. The CA of water in CO2 atmosphere on a glass surface [16] and on polystyrene thin films [17] increased significantly with pressure up until pressures close to the vapor pressure of CO2 at 23 °C (61.2 bar in the case of the glass surface [16] and 59 bar in the case of the polystyrene thin films [17]), and only moderately with a further pressure increase. Saraji et al. [18] reported that both advancing and receding contact angles of a CO2 bubble on a quartz surface in water atmosphere increased with transition from the subcritical to supercritical CO2, with a sharp increase near the critical point. Sutjiadi-Sia et al. [19] determined the CA of water, ethanol and their mixtures on rough PTFE, glass and stainless steel surfaces in the presence of CO2 at 1–270 bar and 40 °C. The wettability of water was poor, intermediate and good for PTFE, stainless steel and glass, respectively. Recently, Santos et al. [20] reported the complete wetting of corn oil on polished stainless steel surfaces in air, SC-N2 and SC-CO2, based on CA values lower than 10º. Moreover, bouncing of corn oil drops on a completely wettable surface was observed under the SC-CO2 environment and this phenomena was related to the formation of a thin vapor film of SC-CO2 saturated with oil (also called Leindenfrost phenomena). Santos et al. [20] highlighted that the formation of such a film could modify the interaction of the lipid drop on the surface as well as the surface energy of the solid, resulting in the bouncing effect.
The CAapp can be easily determined by applying the B-spline snake method to captured video frames over the first moments of contact of the drop on the surface, while the same method is applied to the pendant drop at the tip of the needle before detaching to determine γLG [21]. This method applies parametric spline curves (piecewise-polynomial functions consisting of concatenated polynomial segments) to moving images to find the contour of the drop and calculate both CAapp and γLG. When the drop of the liquid under investigation first contacts the surface, it may or may not adhere to it. The work necessary to separate the drop from the surface is referred to as the work of adhesion (WA), which represents the affinity between the liquid and the solid surface [22]. It can be determined from Eq. (2):
However, the behavior of the liquid on the solid surface is determinated not only by the work of adhesion, but also by the work of cohesion (WC). WC represents the work required to produce two unit areas of interface from an original unbroken column of the liquid, e.g., how likely a droplet is to remain intact [22]. It is expressed by:
Finally, the spreading coefficient is obtained as the difference between WC and WA (Eq. (4)) and represents the wettability of the surface by the liquid. If SL/S is larger than zero, the liquid spreads spontaneously on the solid surface to form a thin film [22].
Despite the insight obtained to date into the behavior of RO membranes in the fractionation of lipid mixtures using SC-CO2, little is known about the wetting behavior of lipids occurring on the membrane surface under high pressure CO2 environment. Thus, the aim of this study was to determine the CAapp of oleic acid (OA) and triolein (TO), representing different lipid classes, under atmospheric (22 °C/1.01 bar) and SC-CO2 (40 °C/120 bar) environments on a commercial polyamide membrane (SG type) over time (0–1 s). This parameter was correlated with the spreading coefficient through the work of adhesion and cohesion of the system to establish the wetting regime (partial or complete wettability) of OA and TO on SG membrane in SC-CO2 environment. The type of membrane and SC-CO2 conditions were chosen based on a previous study by Araus and Temelli [4] where the benefits of coupling SC-CO2 and cross-flow polyamide membrane technologies to fractionate lipid mixtures were demonstrated. Specifically, it was shown that the polyamide membrane (SG type) at 120 bar, 40°C and ΔP of 10 bar exhibited the best stability, selectivity and separation factors, where the OA separation factor was higher than 1 and those for TAG and α-tocopherol/β-sitosterol were less than 1 [4].
Section snippets
Materials
A commercial polyamide reverse osmosis membrane was provided by GE Osmonics Inc. (membrane type SG, Minnetonka, MN, USA), referred to as the SG membrane throughout the text. OA (purity of 100%) and TO (purity ≥75%) were purchased from VWR International (Mississauga, ON, Canada). Both OA and TO were stored in amber bottles under nitrogen at 4 °C. Liquid CO2 (purity of 99.9%, moisture content < 3 ppm) and nitrogen (purity of 99.998%) were obtained from Praxair Canada Inc. (Mississauga, ON,
CA at atmospheric conditions
The methodology was validated by measuring the CAapp of milli-Q water on washed SG membrane (the membrane was washed according to Akin and Temelli [13] at atmospheric conditions). The CAapp of water on washed SG membrane was 65.99° ± 0.98°, which was in agreement with 69.3° ± 2° reported by Akin and Temelli [13]. The CAapp was also measured on un-washed SG membranes at atmospheric conditions; however, it was lower (56.70° ± 1.45°) compared to that on washed membrane. The increased
Conclusions
The CAapp for OA and TO at atmospheric (air at 22 °C/1.01 bar) and supercritical (SC-CO2 at 40 °C/120 bar) conditions was measured by using the B-spline snake method for a sessile drop on un-washed SG membrane. The system and methodology were validated by measuring the CAapp for milli-Q water on washed SG membrane at atmospheric conditions and comparing to literature values. The CAapp of milli-Q water on un-washed SG membrane was lower that on washed SG membrane. This was attributed to the
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
The authors acknowledge the Natural Sciences and Engineering Research Council of Canada (NSERC)-Discovery Grants Program (RGPIN-2017-04384) for financial support and GE Osmonics Inc. for providing the membranes.
References (32)
- et al.
Purification of structured lipids using SCCO2 and membrane process
J. Memb. Sci.
(2007) - et al.
Supercritical carbon dioxide fractionation of the model mixture squalene/oleic acid in a membrane contactor
Sep. Purif. Technol.
(2008) - et al.
Separation of major and minor lipid components using supercritical CO2 coupled with cross-flow reverse osmosis membrane filtration
J. Memb. Sci.
(2018) - et al.
Separation of nutmeg essential oil and dense CO2 with a cellulose acetate reverse osmosis membrane
J. Memb. Sci.
(2001) - et al.
Performance of reverse osmosis membranes in the separation of supercritical CO2 and essential oils
J. Memb. Sci.
(2004) - et al.
Separation of d-limonene from supercritical CO2 by means of membranes
J. Supercrit. Fluids
(2005) - et al.
Performance of reverse osmosis and nanofiltration membranes in the fractionation and retention of patchouli essential oil
J. Supercrit. Fluids
(2016) - et al.
Effect of dense CO2 on polymeric reverse osmosis and nanofiltration membranes and permeation of mixtures of macauba oil (Acrocomia aculeata) and CO2
J. Memb. Sci.
(2015) - et al.
Characterization and performance of reverse osmosis and nanofiltration membranes submitted to subcritical and supercritical CO2
J. Supercrit. Fluids
(2017) - et al.
Effect of supercritical CO2 flux, temperature and processing time on physicochemical and morphological properties of commercial reverse osmosis membranes
J. Supercrit. Fluids
(2011)