Corn Oil – Water Separation: Interactions of Proteins and Surfactants at Corn Oil/Water Interfaces
ABSTRACT
Purification and collection of industrial products from oil-water mixtures are commonly implemented processes. However, the efficiencies of such processes can be severely influenced by the presence of emulsifiers that induce formation of small oil droplets dispersed in the mixtures. Understanding of this emulsifying effect and its counteractions which occur at the oil/water interface is therefore necessary for the improvement of designs of these processes. In this article, we investigated the interfacial mechanisms of protein-induced emulsification and the opposing surfactant-induced demulsification related to corn oil refinement. At corn oil/water interfaces, the pH-dependent emulsifying function of zein protein, which is the major storage protein of corn, was elucidated by surface/interface-sensitive sum frequency generation (SFG) vibrational spectroscopy technique. The effective stabilization of corn oil droplets by zein protein was illustrated and correlated to its ordered amide I group at the oil/water interface. Substantial decrease of this ordering with the addition of three industrial surfactants to corn oil-zein solution mixtures was also observed using SFG, which explains the surfactant-induced destabilization and coalescence of small oil droplets. Surfactant-protein interaction was then demonstrated to be the driving force for the disordering of interfacial proteins, either by disrupting protein layers or partially excluding protein molecules from the interface. The ordered zein proteins at the interface were therefore revealed to be the critical factor for the formation of corn oil-water emulsion.
INTRODUCTION
Oil-water separation is a critical industrial process in a variety of fields such as foods, agricultural products, petrochemicals, textile and pharmaceuticals. In the corn refining industry, oil-water separation plays a significant role for the treatment and collection of corn oil products. Generally, the oil-water separation processes can be severely hindered by the presence of an emulsifier in the mixture, for example, protein molecules or surfactants. These molecules can migrate to the oil/water interface and stabilize smaller oil droplets distributed in aqueous solutions to form relatively homogeneous emulsions. However, when proteins and other artificial surfactants are both present in such systems, their individual emulsifying mechanisms might be incompatible, leading to competitive destabilization of small oil droplets. Therefore, the introduction of a surfactant into an oil-water emulsion stabilized by a protein can effectively disrupt the stability of the system, resulting in coalescence of oil droplets, leading to easier oil-water separation.
This collateral protein-surfactant effect can be observed in the dry grind ethanol process. Thin stillage, which is about ten percent solids and contains fiber, protein and corn oil, is the less dense material that is produced from centrifugation of the whole stillage after distillation of alcohol. The more dense, higher solid material produced in the centrifugation is called wet cake. The coproduct condensed distillers’ solubles, commonly referred to as corn syrup, is produced by evaporation of water from thin stillage to yield a material with a solids content of about twenty-five to fifty percent. Corn syrup is generally treated with a surfactant to enable efficient separation and increased yield of corn oil during subsequent centrifugation. Chemically, corn oil consists of triacylglycerols, lipids, free fatty acids and other minor components. The syrup after centrifugation and oil removal generally is recombined with the wet cake and dried to produce an animal feed called dried distillers grains with solubles. The enhanced corn oil separation efficiency by treating the corn syrup with surfactants is believed to be related to the presence of maize proteins in the system.
Zein protein is one of the most important natural products extracted from maize. Comprising approximately fifty percent of corn proteins, zein is regarded as a very versatile material that has shown extraordinary potentials as commercialized products in various industrial applications such as coatings, fibers, biodegradable films, drug delivery and tissue engineering. It is alcohol-soluble and considered to be largely hydrophobic. However, its hydrophobicity is tunable by adjusting the pH value in the environment, which renders it amphiphilic under controlled conditions. More specifically, acidic (pH less than one) or alkali (pH more than twelve) treatment can significantly enhance the solubility of zein in aqueous solutions by protonation and deprotonation of the protein functional groups respectively. Zein colloidal particles have therefore been utilized as an emulsifier in some reported studies. Interestingly, the deacidification of crude corn oil by alkali treatment is a commonly implemented process in corn oil refinement, where undesirable loss of considerable amount of corn oil (around twenty percent) was previously reported. Since the emulsification effect of zein protein can occur at oil/water interfaces under this condition, a thorough understanding of the interfacial behavior of zein is necessary. An analytical technique that can access the oil/water interface is therefore desirable for this research.
In recent years, sum frequency generation (SFG) vibrational spectroscopy has been developed into a powerful tool to study surfaces and interfaces. SFG has been extensively employed to study a variety of biological molecules such as proteins, peptides, lipid monolayers and bilayers at interfaces in situ. It is an intrinsically surface-sensitive second order nonlinear optical spectroscopy technique that can provide molecular level structural information for surfaces and interfaces. We have demonstrated the valuable molecular insights into interfacial protein structures probed by SFG in numerous reported studies, including those at solid/liquid interfaces. In this study, SFG spectroscopy was applied to investigate the interfacial molecular behavior of zein protein at the oil/water interface and reveal the mechanism of zein protein functioning as an oil-water emulsifier. Several surfactants were also tested as cleaning agents to interact with adsorbed zein protein at the oil/water interface or prevent zein protein from interfacial adsorption. To the best of our knowledge, this is the first report on the SFG studies on interfacial behavior of biological molecules and industrial surfactants in the oil-water separation process.
EXPERIMENTAL SECTION
Sample Preparation
Zein protein from maize, potassium hydroxide and hydrochloric acid were purchased from Sigma-Aldrich. Deuterated polystyrene was purchased from Polymer Source, Inc. Right angle calcium fluoride prisms were purchased from Altos Photonics (Bozeman, MT). Corn oil and corn syrup generated from the dry grind ethanol process were obtained from BASF. To investigate the interactions at corn oil surfaces, we prepared thin corn oil films on solid substrates. Corn oil solution in toluene at two percent volume to volume was spin-coated onto a clean calcium fluoride prism at 2000 revolutions per minute for thirty seconds with a P-6000 spin coater (Speedline Technologies) to make such thin corn oil films. The spin-coated corn oil samples were left under ambient condition for at least twenty-four hours before use to ensure complete toluene evaporation.
Alkaline zein protein solution was produced by dissolving zein protein in potassium hydroxide solution at 1.0 milligrams per milliliter and at pH equal to twelve. Acidic zein protein solution was produced by first partially dissolving zein protein in hydrochloric acid solution at pH equal to 4.5 until saturation. The supernatant with a concentration lower than 1 milligram per milliliter was then separated from the undissolved protein particles and used as zein protein solution at pH equals to 4.5. The pH values were measured using a Thermo Scientific Star A111 pH benchtop meter. Various surfactant solutions were made by diluting the original surfactant stock solutions in deionized water to different desired concentrations. The chemical nature of these surfactants could not be disclosed in this article because such information is proprietary. Deuterated polystyrene in toluene two percent weight to weight was spin-coated onto a clean calcium fluoride prism at 2000 revolutions per minute for thirty seconds to make thin deuterated polystyrene films. Alkaline treatment of corn syrup was carried out by adding 500 microliters of potassium hydroxide solution at pH equals to twelve to 100 microliters of corn syrup.
The corn oil-water dispersions were prepared by mixing the corn oil sample with water solutions at different pH values. The mixtures were then shaken using a vortex mixer for one minute and placed on a horizontal bench.
SFG
SFG vibrational spectroscopy has been widely applied to study surfaces and interfaces, including polymer and biological surfaces and interfaces. SFG theories, experimental details, and data analysis methodology have been extensively reported and will not be repeated here. For the experiments carried out in this research, two input laser beams, a frequency fixed 532 nanometer visible laser beam and a frequency tunable infrared laser beam (wavenumber tunability from 1000 to 4300 cm⁻¹), were used for data collection. The visible beam was generated by frequency doubling the 1064 nanometer (20 picosecond pulsewidth) output from an EKSPLA Nd:YAG laser. The tunable IR beam was produced using an EKSPLA optical parametric generation, optical parametric amplification and difference frequency generation system with LBO and AgGaS2 crystals. These two input beams were overlapped on the sample surface or interface spatially and temporally (about 500 micrometers diameter). The sum frequency signal was then generated and collected using a photomultiplier normalized by the input visible and IR beam intensities. All SFG spectra presented in this article were collected using SSP polarization combination (S-polarized sum frequency output, S-polarized visible input, and P-polarized IR input). Many interfaces investigated in this article were not very ordered, resulting in weak SFG signals. All the results presented in this article were reproduced using multiple samples (at least three) for multiple times (at least three) to ensure the data reproducibility. The SFG fitting results are presented in Table S1 in the Supporting Information.
RESULTS AND DISCUSSIONS
Characterization of Corn Oil Surface in Air and in Water Using SFG
Before we present the results on interactions of proteins and surfactants at the corn oil interfaces, we first characterized the corn oil surface in air and in water to gain fundamental understanding of its surface properties. The corn oil thin films on solid substrates prepared as described above were used in this study. SFG spectra were successfully collected from both the corn oil/air interface and the corn oil/deionized water interface. For the sake of convenience, we will refer to these interfaces simply as oil/air and oil/water (and oil/zein for corn oil/zein protein solution interface below). It is noteworthy to mention that the corn oil sample from BASF was prepared by separating the corn oil from the corn syrup mixture at BASF, and we directly spin-coated such corn oil solution in toluene on a solid substrate to prepare the corn oil thin film surface. The corn syrup from BASF contains zein proteins, therefore likely some zein protein molecules that interacted strongly with corn oil in the corn syrup were still present in the corn oil sample. Some zein protein residues thus might be present at various interfaces of corn oil which we investigated in this research. We will address this issue further in the following discussions.
As shown in the results, the SFG signal from the corn oil sample in air exhibited three obvious vibrational peaks at 2850, 2880 and 2940 centimeters to the minus one in the C-H stretching frequency region. These three peaks are contributed by the methylene symmetric stretching, methyl symmetric stretching and methyl Fermi resonance modes, respectively. It can be inferred that the corn oil sample surface in air is dominated by ordered methyl groups and methylene groups. On the other hand, SFG signals collected from the oil/water interface in the C-H stretching region have very similar spectral features to those from the oil/air interface. However, the signal intensities of the peaks from the oil/water interface are much weaker compared to those from the oil/air interface. This is reasonable because of the different refractive indices of air and water, leading to varied Fresnel coefficients of the oil/air and oil/water interfaces (about 2.7 to 1). Additionally, the hydrophobic functional groups such as methyl and methylene groups likely form more favorable interactions with hydrophobic medium (air) than with hydrophilic medium (water), therefore they should be more ordered in air, resulting in stronger SFG signals.
SFG spectrum in carbonyl group frequency region collected from the corn oil/water interface shows a distinct signal at 1725 centimeters to the minus one. This peak is contributed by the carbonyl groups from corn oil, indicating that the carbonyl group is ordered at the corn oil/water interface. SFG spectrum collected from the corn oil in air shows negligible 1725 cm⁻¹ peak signal, indicating that the carbonyl groups either are not ordered or have much lower surface coverage in air. The above observations are reasonable: The carbonyl groups are hydrophilic, which interact favorably with water and exhibit an ordered orientation at the corn oil/water interface. Oppositely, hydrophilic carbonyl groups do not cover the corn oil surface or adopt random orientation in hydrophobic air due to unfavorable interaction between carbonyl groups and air.
No signal was observed from this interface between 1600 and 1700 cm⁻¹, indicating that either no protein is present at the interface or the interfacial proteins are disordered. Since we used a thin corn oil film in this experiment, likely the small number of proteins in such a film cannot form an ordered protein layer at the interface or even cannot achieve a substantial coverage at the interface.
Emulsifying Effect of Zein Protein
The function of zein protein as an emulsifier for corn oil phase and water phase is demonstrated further. When corn oil was mixed with deionized water, the free fatty acids mostly existed in their protonated, unionized forms. Zein protein was also not very soluble at this pH value. Clear separation between the oil layer and the water layer was then observed. However, when corn oil was mixed with alkaline water, more fatty acids became deprotonated and ionized to function as anionic surfactants. Zein protein also became very soluble in this basic environment. An oil-in-water emulsion was then produced. We hypothesize that in an alkaline environment, zein protein is soluble and amphiphilic. It migrates to the oil/water interface with the ionized fatty acids to function as an emulsifier. When extra zein protein was added into the above mixture, combined with the zein protein molecules already existed on the corn oil surfaces, they can function as additional emulsifiers, producing a more homogeneous emulsion. However, when the pH of the system was lowered to 4.5, the emulsifying effects of zein, the free fatty acids or other interface-active species such as monoglycerides were observed to be greatly reduced. When corn oil was mixed with water and zein solution at pH equals to 4.5, minimal oil dispersion occurred in the mixture, resulting in only mildly turbid solutions, instead of complete emulsions. This can be related to zein protein being not very soluble at pH equal to 4.5 in aqueous solutions, as well as the protonation of free fatty acids. Although the low solubility of zein at pH 4.5 might be slightly greater than that in deionized water to make zein function as an emulsifier, the emulsifying effect might be substantially weaker than that at pH 12. This research is focused on the investigations of the emulsifying effect of zein protein. SFG is therefore applied to examine the interfacial behavior of zein protein in the oil-water system.
Interfacial Behavior of Zein Protein at Corn Oil/Protein Solution Interface and Surfactant Effect
In order to elucidate the interaction mechanism between corn oil and zein protein, SFG experiments were undertaken to study the oil/zein protein solution interface. At first, the SFG spectrum collected from the corn oil/water (pH equals twelve) interface is very similar to that collected from the oil/deionized water interface presented earlier. Here the “water (pH = twelve)” is a potassium hydroxide solution with a pH of twelve. The pH value of twelve is substantially higher than the pKa values of all major fatty acids in corn oil. Therefore, the deprotonated carboxylate groups were expected to be present at the oil/water (pH=12) interface. Surprisingly, the peak centered at 1725 cm⁻¹ was observed, which is typically assigned to the unionized carboxyl groups. There are two possible explanations for this observation. The first hypothesis is that the studied acids exhibit higher pKa values at the interface than in the bulk to stay protonated, as previously reported for other long-chain fatty acids. The second hypothesis is that this 1725 cm⁻¹ peak originated from the ester group in the corn oil glycerides, whose typical vibrational frequency (about 1745 cm⁻¹) was shifted by hydrogen-bonding interactions at the interface. It is worth mentioning that the SFG spectrum in the C=O stretching region may also contain contributions from both unionized carboxyl group and the glyceride ester groups.
After the spectra collection, water (pH=12) in contact with oil was replaced by a zein protein solution (pH=12, zein concentration: 1 mg/mL) and SFG spectrum was collected from the oil/zein protein solution interface. Zein protein is fully soluble at this pH value. A strong SFG signal at about 1645 cm⁻¹ was detected from this interface which proves that the amide groups in zein proteins are ordered at the corn oil surface. The carbonyl group signal intensity at about 1725 cm⁻¹ at the oil/zein protein solution interface decreased substantially compared to that from the oil/water interface, which either indicates less ordering of carbonyl groups, or carbonyl groups lying down when zein protein was present at the oil/aqueous solution interface. The peak center also shifted slightly. Carbonyl groups on the corn oil surface are therefore likely to play some roles in oil-zein protein interactions. This result can be related to a previously reported adsorption study about zein protein, which demonstrates that zein protein showed higher affinity and higher mass adsorption on carboxyl than alkyl surfaces. The above findings provide the interfacial molecular information for zein-stabilized emulsions.
The zein protein solution was then replaced with water (pH=12) to remove proteins that are loosely adsorbed to corn oil. Another SFG oil/water spectrum was taken after this cleaning process. After washing with water, the protein amide I peak intensity decreased but was still much stronger than the intensity of carbonyl groups. This result indicates that zein proteins can be strongly adsorbed to the corn oil surface and cannot be easily removed by rinsing the surface using water (pH=12).
Experiments were then performed to see whether various surfactants can remove zein proteins from corn oil surface. Three surfactants, surfactants one, two, and three provided by BASF, were used in this study. Three corn oil surfaces after contacting zein solution and water washing discussed above were placed in contact with water solutions of three surfactants at 0.2 percent concentration respectively for approximately ten seconds. The corn oil surfaces were then washed using water to remove all the surfactants and were brought to contact with water again. After that, SFG spectra were collected again from the oil/water interface. SFG protein amide I signals decreased after contacting all the three surfactants, indicating either effective removal of protein molecules from the surface or disruption of interfacial protein ordering by the surfactants. Some weak SFG amide I signals can still be detected, showing that the surfactant cannot completely remove the zein protein from the interface or completely disorder the protein at the interface.
The above experiments were then repeated at lower surfactant concentrations (500 parts per million instead of 0.2 percent). SFG spectra were collected from each oil/water interface after replacing zein protein solutions with surfactant solutions then with water. After washing by surfactants one or three, SFG amide I signals from zein proteins can still be observed at the oil/water interface, showing that these two surfactants could not completely remove the zein proteins from the oil/water interface or cannot completely disorder the interfacial zein protein. In contrast, SFG spectrum collected from the oil/water interface after surfactant two cleaning shows no protein amide I signal. We believe that surfactant two was capable of completely cleaning zein protein from the interface or disordering the interfacial zein protein at this lower surfactant solution concentration. We therefore also believe that surfactant two will perform better compared to the other two surfactants for oil-water separation.
In the above studies, the protein solutions have a pH value of twelve. It is well known that proteins can behave very differently at interfaces when the contacting media have varied pH values. Here in addition to the protein solutions with pH of twelve presented above, we also studied various interfaces at pH of around 4.5. SFG spectra collected from the oil/water (pH=4.5) and oil/zein solution (pH=4.5) interfaces demonstrate that the carbonyl group signal at the oil/water interface was decreased when water was replaced by zein solution. This change is related to the presence of zein protein at the interface, which is consistent with the previously observed trend at pH=12. Very weak protein amide I SFG signal was detected from the oil/zein solution interface, indicating that the zein proteins at this interface might have a low coverage and/or were quite disordered. At this lower pH value, zein protein is much less soluble than that at pH=12, primarily due to less deprotonation of aspartic acid and glutamic acid in zein. We can therefore conclude that the ionization of carboxyl groups on zein protein molecules (i.e., the hydrophilicity of zein) is critical for their activity and ordering at corn oil/water interface. The surfactant two solution (500 parts per million) was also applied using the same method discussed above to remove surface adsorbed zein protein. After the cleaning procedure, SFG signal was collected again from oil/water interface to reveal an obvious increase of carbonyl group peak intensity. The very weak amide I signal disappeared. We believe that this is because of the removal of zein (at least some of the proteins if not all) from the interface which makes the interfacial oil molecules more ordered in the presence of surfactant molecules.
Surfactant-Induced Coalescence of Zein-Stabilized Corn Oil Droplets
The SFG results demonstrated a cleaning effect of the surfactants, where surfactants removed or disordered interfacial zein protein molecules. To visualize this surfactant effect in oil-water emulsion systems, surfactants were added to the mixtures of zein protein at pH=12, corn oil and water. Corn oil layer segregation to the top was observed for all the three mixtures with the respective surfactant, contrasting the relatively homogeneous emulsion where surfactant was absent in the mixtures. This experiment demonstrates the effective destabilization of protein-stabilized oil droplets by surfactants in oil-water emulsion system.
Additional study for the surfactant effect at pH=4.5 was also conducted using surfactant two solution. Slight oil segregation was observed, likely due to the reduced emulsification by zein at this pH value. The amphiphilic surfactants usually reside at oil/water interfaces to function as cleaning agents. However, the presence of zein protein makes the exact mechanism of the observed oil droplets destabilization more difficult to be defined.
The mechanism of the oil droplet destabilization in water by protein plus surfactant mentioned above may be complicated because the system involves intertwined oil-protein, oil-surfactant and protein-surfactant interactions simultaneously. Therefore, in addition to the interfacial oil-protein interaction with and without the presence of surfactants in contact with the interface discussed above, we also present here interfacial analysis for the other two interactions. First, the interactions between the three surfactants and corn oil (without external zein protein adsorbed) were investigated using SFG. SFG signals were first collected from the corn oil/deionized water interface. Water was then replaced by three surfactant solutions in water (500 parts per million) respectively and SFG spectra were collected from the oil/surfactant solution interfaces. The SFG signals contributed by carbonyl group and C-H groups from both surfactant and oil were observed. The carbonyl group signal remained almost unchanged for surfactants one and two, and slightly decreased for surfactant three. This result contrasts with that collected from the oil/zein interface, where substantial decrease of carbonyl group signal was observed compared to that at the oil/water interface. Therefore, the oil-protein interaction and oil-surfactant interaction are different at the oil interface, which can be differentiated by the different interfacial behavior of the carbonyl group. We believe that the surfactants can competitively interact with corn oil against zein protein at the corn oil/water interface, likely because surfactant and zein interact with the carbonyl groups on corn oil differently. Additionally, SFG C-H stretching signals at the oil/surfactant two and oil/surfactant three interfaces were much stronger than that at the oil/water interface, while C-H signal at the oil/surfactant one interface was much weaker compared to the other two surfactants. This shows that likely different surfactants also have different interactions with oil at the interface.
Further study to investigate the mechanism of protein-surfactant competitive adsorption was conducted by pre-mixing one of the three surfactants (500 parts per million) with zein protein solution (1 mg/mL) at one to one volume ratio, and then contacting the spin-coated corn oil surface with this mixture solution, mimicking the chemical environment discussed above. No obvious SFG amide I signal was observed from the oil/mixture solution (pH equals to twelve) interfaces for all the three protein-surfactant mixtures, different from the SFG signal collected from the oil/pure zein solution interface where both signals from the oil and zein were seen. The absence of protein SFG amide I signal could be caused by either the exclusion of zein from the interface, or the disruption and disordering of zein at the interface caused by the surfactants. On the other hand, the SFG signals from carbonyl group signals were much weaker compared to those from the oil/surfactant solution interface. These spectral differences between oil/surfactant solution interface and oil/(zein plus surfactant) solution interface indicate that the likely presence of zein protein in the solution system exerts noticeable influence on the interfacial surfactants.
We propose that zein protein molecules were not completely excluded from the oil/water interface. Instead, surfactants and zein protein molecules formed strong interactions (or perhaps complexes) at the interface (certainly in bulk solution as well), which could severely disrupt the protein-stabilized oil droplets. We believe that the surfactant additions can possibly cause both the disordering of zein molecules and the decrease of zein molecule number at the interface, but the disordering effect is likely to play the dominant role. The disordering of interfacial zein protein at this pH environment is likely related to the destabilization of oil droplets. This result obtained from the study on oil – zein protein plus surfactant mixture interactions can also be correlated to the cleaning mechanism of surfactants removing or disordering already adsorbed proteins on corn oil surface mentioned above. Nevertheless, clearly the working mechanism for surfactants for oil-water separation is because the surfactants can influence the interfacial protein behavior by removing or more likely disordering the interfacial zein proteins.
Zein Protein Behavior at Corn Syrup Interfaces
Instead of using zein protein solution and corn oil, the corn syrup sample was also analyzed to understand the interfacial behavior of zein protein in its original, unpurified environment. Clean calcium fluoride prism and deuterated polystyrene spin-coated onto calcium fluoride prism were used as two substrates that represent hydrophilic surface and hydrophobic surface respectively. SFG spectra were collected from the calcium fluoride/corn syrup and deuterated polystyrene/corn syrup interfaces. Strong C-H vibrational signals assigned to the alkyl chains in corn oil were detected from the deuterated polystyrene/syrup interface, but not from the calcium fluoride/syrup interface. This result can correspond well with previous spectra collected from the oil/air interface where C-H functional groups tend to be more ordered when contacting a hydrophobic environment than a hydrophilic environment. Additionally, in amide I spectral region, no noticeable SFG signal could be observed from either interface, indicating that proteins either did not move to the interface, or assumed a random orientation distribution at the interface. This is likely to be caused by the low solubility of zein protein in water at the pH value of corn syrup (about 4.5). We hypothesize that the increase of zein protein solubility will enhance its mobility to move to interfaces and then function as an emulsifier. To test this hypothesis, alkaline water containing potassium hydroxide was added into the syrup mixture to increase the pH value. SFG spectra were again collected from the calcium fluoride/syrup and deuterated polystyrene/syrup interfaces (with higher pH syrup). SFG amide I signal was then detected from the deuterated polystyrene/syrup interface but not from the calcium fluoride/syrup interface. The deuterated polystyrene/syrup interface spectrum is crowded with multiple peaks, likely due to the complicated composition of the syrup mixture. The presence of protein in corn syrup can be validated based on these results. This study again demonstrates that the protein interfacial behavior is heavily influenced by the pH value of the environment, due to its varied pH dependent solubility (hydrophilicity) in the mixture.
The very different SFG results obtained from the deuterated polystyrene/syrup and calcium fluoride/syrup interfaces at alkaline condition indicate that zein protein (in syrup) interacts with the relatively hydrophobic polystyrene more favorably to form more ordered structures. The abundance of hydrophobic domains in zein protein is likely to contribute to this phenomenon. Further tests were conducted to elucidate the behavior of zein protein at different surfaces using zein protein solution (pH=12) instead of corn syrup. Clean calcium fluoride prism and deuterated polystyrene spin-coated on calcium fluoride prism were again used as the two substrates to contact zein protein solution. Strong SFG amide I signal was successfully detected from the deuterated polystyrene/zein protein solution interface, while no discernible signal was observed at the calcium fluoride/zein solution interface. These results are well correlated to those detected from the corn syrup interfaces. The above presented interfacial zein protein behavior clearly demonstrates that zein protein molecules form ordered structures at both polystyrene and corn oil surfaces, likely due to hydrophobic nature of these two surfaces.
CONCLUSIONS
In this study, we systematically investigated the oil/aqueous interfaces using SFG spectroscopy to provide fundamental understanding on oil-water separation. The effects of solution pH, zein protein, and surfactant on such interfaces were elucidated. In air, only alkyl chains are ordered on the corn oil surface. In water, both alkyl chains and carbonyl groups of corn oil on the surface have some ordering. With the presence of zein protein in the solution, they can aggregate to the oil/water interface, making the separation of water and oil difficult. The solution pH plays an important role in determining the ordering of zein protein at oil/water interfaces. At pH of twelve, zein protein molecules are ordered at the oil/water interface, and some zein protein molecules are strongly adsorbed to the oil/water interface, which cannot be removed by water washing. Oppositely, at pH of 4.5, interfacial zein protein molecules adopt random orientation distribution.
Three surfactants were applied in the oil-water-protein system to reduce the emulsification effect of proteins for the oil-water system. It was found that all of them can remove or disorder some of the zein proteins at the interface by surfactant-protein interactions, leading to easier oil-water separation. Among the three studied surfactants, it seems that surfactant two has the best interfacial protein removal or disordering effect.
We also examined zein protein behavior at the interface with corn syrup. It was found that zein protein has different interfacial behavior at hydrophobic and hydrophilic interfaces.
Supporting Information
The Supporting Information is available on the ACS Publications website. SFG spectra in C-H stretching frequency region, photo of oil-water emulsion at pH=4.5, spectral fitting results.