Open AccessCCS ChemistryRESEARCH ARTICLE6 Jun 2022Water-in-Water Emulsions, Ultralow Interfacial Tension, and Biolubrication Yitong Wang, Jin Yuan, Yunpeng Zhao, Ling Wang, Luxuan Guo, Lei Feng, Jiwei Cui, Shuli Dong, Shanhong Wan, Weimin Liu, Heinz Hoffmann, Kiet Tieu and Jingcheng Hao Yitong Wang Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author , Jin Yuan Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author , Yunpeng Zhao Department of Orthopedics, Qilu Hospital, Cheeloo College of Medicine, Shandong University, Jinan, Shandong 250012 Google Scholar More articles by this author , Ling Wang Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author , Luxuan Guo Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author , Lei Feng Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author , Jiwei Cui Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author , Shuli Dong Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author , Shanhong Wan Faculty of Engineering and Information Sciences, University of Wollongong, Wollongong, NSW 2522 State Key Laboratory of Solid Lubrication, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000 Google Scholar More articles by this author , Weimin Liu State Key Laboratory of Solid Lubrication, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000 Google Scholar More articles by this author , Heinz Hoffmann Physikalische Chemie I, Bayreuth Universität, Bayreuth D-95447 Google Scholar More articles by this author , Kiet Tieu Faculty of Engineering and Information Sciences, University of Wollongong, Wollongong, NSW 2522 Google Scholar More articles by this author and Jingcheng Hao *Corresponding author: E-mail Address: [email protected] Key Laboratory of Colloid and Interface Chemistry (Ministry of Education) & State Key Laboratory of Crystal Materials, Shandong University, Jinan, Shandong 250100 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101028 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail A water-in-water (W/W) emulsion consists of droplets formed by the spontaneous liquid–liquid separation of two immiscible aqueous phases. The inherent properties of the W/W interfaces, low or ultralow interfacial tension (γW/W = 1–1000 μN/m) and a large thickness of several nanometers, beget the poor inherent stability of emulsions. Herein, we report a nanofibril emulsifier having Schiff base reactivity to generate a W/W emulsion. The W/W emulsion has superior stability (stable > 6 months) because collagen nanofibrils, acting as a stabilizer of W/W emulsions, can simultaneously satisfy the requirements of size and overall coverage ratio of the phase interfaces. W/W emulsions having γW/W ∼10 μN/m were used as synthetic synovial fluids, showing superior lubrication performance with a coefficient of friction in the range of 0.003–0.011, which has been demonstrated to be suitable for joint lubrication. An intraarticular injection assessment further confirmed this protective effect on articular cartilage in vivo. Our study reveals the mechanism of emulsion stabilization and opens up the possibility of osteoarthritis (OA) treatment using the biolubrication effects of W/W emulsions for lubricated artificial implant surfaces. Download figure Download PowerPoint Introduction The extremely low coefficient of friction (COF) between interfaces in relative motion in an organism facilitates their daily use.1,2 Among them, joint interfaces are typical biolubrication interfaces, and their excellent lubrication performance is mainly attributable to articular cartilage and joint synovial fluid (SF).3,4 Smooth and elastic articular cartilage can cushion the vibration between the connected bones during movements, and viscous SF also plays a vital role in lubricating cartilage surfaces. Lubrication failure induced from the degeneration of SF leads to cartilage wear and thus osteoarthritis (OA).5,6 The primary choice for OA treatment, artificial joint replacement surgery, not only has a high cost and causes great pain for patients, but may also limit implant rejection. In addition, abrasive particles generated by friction and degradation of artificial joint materials may cause the bones around the joint to dissolve or loosen.7,8 An ideal substitute material, bulk hydrogel, has been investigated because of its many similarities with articular cartilage.9,10 However, the challenges in meeting the requirements for both hydrogel lubrication and strength drastically limit its practical application in articular cartilage replacement materials. In this case, developing highly efficient and biocompatible synthetic SFs provide an alternative solution to treat OA, but it remains a challenge. Herein, we put forward water-in-water (W/W) emulsions, which are formed by two incompatible polymer aqueous solutions, dextran and polyethylene glycol (PEG) with collagen nanofibrils as a stabilizer (Scheme 1), to serve as synthetic SFs (SSFs). Considering that no research on W/W emulsions as SSFs has been reported, the analysis of the possibility of W/W emulsions as SSFs is presented. (1) The artificial SSFs that have been reported, such as microgels11 or polymer brushes,12 are the products of covalent cross-linking of synthetic polymers, and their biosafety remains to be discussed, but W/W emulsions address this concern. Compared with oil-containing emulsions, the biocompatibility of W/W emulsions is favorable.13 (2) The ultralow interfacial tension and thermodynamic instability of all-aqueous emulsions significantly limit their application in various fields.14 However, in this work, W/W emulsions exhibited excellent stability for greater than 6 months. (3) Although there is no cross-linked elastic structure as in polymer-like microgels, the large amount of water inside the W/W emulsions and the smooth and elastic character of the stabilizing collagen nanofibrils can greatly buffer stress and protect articular cartilage. (4) It is well known that biopolyelectrolytes are the main components of SFs.15 Scheme 1 | Schematic of the W/W emulsions stabilized by mPEG-collagen nanofibrils for application of SFs. Download figure Download PowerPoint Analogous to biological systems, the combination of natural collagen protein, polysaccharide dextran, and hydrophilic PEG ligands might be a promising design for simultaneous biomimetic lubrication and OA treatment. Moreover, recent studies have shown that there are many liquid-like membraneless compartments in eukaryotic cells, which are formed through liquid–liquid phase separation (LLPS), driven by proteins, nucleic acids, and other biomacromolecules. After phase separation is formed, the biological macromolecules exist in two forms, one is a low concentration state in the solution, and the other is a higher concentration in the form of “droplets” whose morphology is similar to the W/W emulsion in this work. Thus, our study not only suggests a new strategy for the construction of arthritis treatment materials by incorporating natural proteins in an all-aqueous surrounding, expanding the types of biomimetic lubrication systems, but it might also provide a theoretical basis for intracellular liquid–liquid separation. Experimental Methods Materials and chemicals For the preparation of stock solution, dextran (500 kDa) and PEG (8 kDa) were obtained from J&K Scientific Ltd. (Beijing, China). Collagen was purchased from Chengdu Kele Biotechnology Co., Ltd. (Chengdu, China). Methoxy polyethylene glycol-acetaldehyde (mPEG-ALD) was synthesized by Shanghai Yayi Biotechnology Co., Ltd. (Shanghai, China). Fluorescein isothiocyanate (FITC)-dextran, tris, and polyacrylic acid (PAA) were purchased from Sigma-Aldrich (Shanghai, China). Sodium tripolyphosphate (TPP) was provided by Aladdin Reagent (Shanghai, China). Salts required in this system, such as CaCl2, NaCl, NaHCO3, KCl, K2HPO4·3H2O, MgCl2·6H2O, Na2SO4, NaN3, and NaH2PO4, were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All of the chemicals were of analytical grade and used without further purification. The pH was adjusted by adding aliquots of 1 M NaOH or HCl. The water used in the experiments was deionized water (18.25 MΩ, Milli-Q, Millipore, Shanghai, China). Synthesis and characterization of mPEG-collagen nanofibrils The mPEG-collagen nanofibrils were prepared using a Schiff base reaction. To obtain a uniform transparent viscous solution, 50 mg collagen was dissolved in 5 mL of 0.2 M acetic acid solution and stirred at room temperature for more than 48 h. After the pH of the mixture was adjusted to 5.0 (near isoelectric point (PI) of collagen, pH 4.9–5.0) by 1 M NaOH solution, 50 mg mPEG-ALD was added into the beaker and stirred at room temperature for 12 h. Then the reaction solution was sealed and maintained at 25 °C without disruption for the formation of nanofibrils. After reacting for 48 h, the resultant solution was dialyzed in a dialysis tube (molecular cut-off 8 kDa) for 3 days to remove the unreacted mPEG-ALD. The collagen formed nanofibrils throughout the reaction. Preparation of W/W emulsions Stock dextran and PEG solutions were prepared by mixing certain amounts of PEG, dextran, and deionized water in a vessel and stirring until dissolved. A final concentration of 1–12 wt % dextran and 1.9–8.6 wt % PEG was dissolved into a 0.05–0.20 wt % nanofibril suspension. The volume fraction of the W/W emulsion phase was varied from 10% to 80% by changing the emulsion compositions. To form a uniform diameter of emulsion droplets, the stirring rate of homogenization was kept at 1500 rpm for 3 min. For fluorescence microscopy observation, FITC-dextran replaced the dextran. For the biomineralization of emulsion droplets, the deionized water prepared for the stock solution was replaced by synthetic body fluid solution (2.5 mM CaCl2, 136.8 mM NaCl, 4.2 mM NaHCO3, 3 mM KCl, 1 mM K2HPO4·3H2O, 1.5 mM MgCl2·6H2O, 0.5 mM Na2SO4, 50 mM Tris, 3.08 mM NaN3, 6 mM NaH2PO4, 1 mg/mL PAA, 2.5 wt % TPP, and pH 7.4), and the biomineralized emulsion droplets were formed after incubation for 72 h. Emulsion stability was monitored daily until demixing of the two bulk phases was observable under a microscope. Measurement of W/W interfacial tensions The interfacial tension between the coexisting dextran- and PEG-rich phases was measured using a spinning drop interface tensiometer (TX500, USA KINO Industry Co., Ltd., Boston, MA). The density of PEG phase was lower than the dextran phase, which served as the external phase for spinning. A 1–2 μL droplet of the PEG-rich solution was injected into a transparent glass tube with a capillary insert filled with the denser solution of the dextran-rich phase. Then the PEG-rich solution droplet continued to expand along the axis of rotation with the glass capillary rotating horizontally. The rotation speed was continuously adjusted until the length of the droplet was more than four times longer than the width. The interfacial tension, γ, between the dextran and phases was calculated from the Vonnegut equation by measuring the width of the cylindrical stick droplet in the capillary. Characterization 1H NMR (400 MHz) spectra were detected on a Bruker Avance 400 spectrometer (Bruker, Karlsruhe, Germany). The scanning electron images of freeze-dried emulsion droplets with gold coating were observed on a scanning electron microscopy (SEM; Zeiss G300, Shanghai, China). Atomic force microscopy (AFM) images were performed with a Bruker Dimension ICON operating in PeakForce Tapping Mode. All images were collected at a scan frequency of 1.5 Hz and a resolution of 512 × 512 pixels. The morphology of emulsion droplets was observed on a microscope (Zeiss Axioskop 40, Shanghai, China). The fluorescent images were recorded on a fluorescent microscope (Zeiss Axio observer 3, Shanghai, China). Wide-angle X-ray diffraction (WAXD) experiments were measured on a X’Pert3 powder diffractometer (PANalytical, Almelo, Netherlands) and XPK-900 in situ solid reaction cell (Anton Paar, Graz, Austria) in the range of 10° < θ < 70°. The friction test was performed on a conventional pin-on-disk reciprocating tribometer (Tribometer UMT-2, Center for Tribology, Beijing, China) recording the friction coefficient (μ). The formation of the frictional pair was realized by pressing the upper running ball against the lower settled disk. The distance of one sliding cycle was 6 mm at constant frequency of 1 Hz and the friction coefficient was obtained by applying different normal loads (from 1 to 10 N) assisted by the tribometer software. The upper ball is an elastomeric poly(dimethylsiloxane) (PDMS) hemisphere with a diameter of 10 mm. The lower settled disk was silica wafer-plated titanium. In this work, the friction coefficients are the average value of the obtained friction coefficients during each cycle. All friction tests were repeated three times to obtain an average coefficient. Biocompatibility of W/W emulsions on hMSCs The Enhanced Green Fluorescent Protein (EGFP)-human mesenchymal stem cells (hMSC)-TERT cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin (P/S). Cells were maintained in a 5% CO2 atmosphere at 37 °C. Besides, cell morphology and growth were visualized by an Invitrogen EVOS FL Auto Cell Imaging System. After 24, 48, and 72 h of cocultivation of W/W emulsion and cells in growth medium, a Cell Counting Kit-8 (CKK-8, Dojindo, Kumamoto, Japan) was used to evaluate cell proliferation on W/W emulsions. According to the manufacturer’s protocol, CCK-8 was diluted at a ratio of 1:10 with growth medium and then used to incubate the cell with W/W emulsions for 1 h in a cell incubator. Afterwards, the medium was collected from each well and their absorbance at 450 nm was measured by a plate reader (Victor X5, PerkinElmer, Akron, OH). For all samples, 12 replicates were tested. After 48 h of culture, the activity levels of lactate dehydrogenase (LDH), a cytosolic enzyme normally used as an indicator of cellular toxicity, in the collected culture media were measured to evaluate cell damage. According to the manufacturer’s protocol (Roche Diagnostics, Mannheim, Germany), the LDH activity was obtained by assessing the absorbance of formazan product at 490 nm via the plate reader (Victor X5). The absorbance of the culture medium without cells was used as the background. The LDH activity in the medium of cells cultured on 96-well plate was used as a low toxicity control, while that of cells cultured on 96-well plate and treated with 1% Triton X-100 was marked as a high-toxicity control. The relative toxicity levels in percentages were calculated using eq. 1: Cytotoxicity = exp . value − low control high control − low control × 100 % (1) Experimental OA animal models Male Sprague-Dawley rats (7 weeks of age, 200–250 g) were purchased from Charles River Co. (Beijing, China). The animals were used after 1 week of acclimatization. The animals were housed in pairs in a temperature-controlled room (21–22 °C) with a 12-h light/12-h dark cycle. OA was induced in male Sprague-Dawley rats by surgical destabilization of the medial meniscus (DMM) of the right knee. Specifically, the DMM surgery was performed by surgical sectioning of the medial meniscotibial ligament and the sham operation was performed by incision of the cutaneous and muscular planes at baseline. All mice that had undergone DMM surgery or sham surgery were randomly divided into three groups: (1) sham group, (2) phosphate-buffered saline (PBS) group, and (3) W/W emulsion group. The mice were given intraarticular injections of 25 μL PBS or W/W emulsions. After the surgery, all animals were killed at the 4-week mark. Dissected joints were processed for either histopathological analysis. All experiments were performed in accordance with Institutional Animal Care and Use Committee (IACUC) guidelines and approved by the IACUC committees of Shandong University. Results and Discussion Stabilization of the W/W emulsion by mPEG-collagen nanofibrils Despite having been used in numerous applications including artificial bioreactors, food formulations, delivery of bioactive agents, and biomimetic templates for the synthesis of microgel particles, W/W emulsions are still limited to ultralow interfacial tension and low coverage of amphiphilic stabilizers adsorbed at W/W interfaces.16 Instead of traditional amphiphilic molecules, larger colloidal particles, such as protein-conjugate particles,17 cellulose nanocrystals,18 or pH-sensitive microgels,19 have been reported to adsorb at W/W interfaces even under conditions of low interfacial tension.20 However, the geometrical constraints of large colloidal particles packed at W/W interfaces often generate low overall coverage ratios,21 while the stability of W/W emulsions is still a challenge. In this paper, to produce W/W emulsions, we employed a protein nanofibril produced by a Schiff base reaction of collagen and mPEG-ALD at a pH near the PI of collagen (4.9∼5.0).22 Supporting Information Figure S1 and Figure 1a demonstrate the success of the Schiff base reaction and the resulting protein nanofibrils (mPEG-collagen) whose periodic structure can be clearly seen in the AFM image. The line profiles of five mPEG-collagen nanofibrils marked in different colors in Figure 1a are shown in corresponding colors of Figure 1b. The periodicity of each nanofibril, ranging from 64.9 to 68.5 nm, is calculated from the quotient of the fiber length and bump number. The width distribution of the mPEG-collagen nanofibers and Gaussian fitting (Figure 1c) show that the width of these nanofibers is around 65.9 nm, which is consistent with the well-known D-banded structure (the periodicity of around 68 nm) assembled from tropocollagen.23 This Schiff base reaction process not only induces the collagen molecules to form uniform D-banded nanofibrils, but can also improve the water solubility and preserve the bioactivity of the collagen during PEGylation. The mPEG-collagen nanofibrils, as a stabilizer of W/W emulsions, can simultaneously satisfy the requirements of size and the overall interface coverage ratio. Figure 1 | W/W emulsions stabilized by mPEG-collagen nanofibrils. (a) A typical AFM image of the mPEG-collagen nanofibrils. (b) The profiles of five colored lines in (a). (c) The fiber width distribution of the collagen fiber in (a). (d) OP images of W/W emulsions at cPEG = 8.0 wt % with varying cdex from 0.05 (1), 0.10 (2), 0.15 (3), and 0.20 wt % (4). (e) γW/W between two aqueous phases with and without adsorption of mPEG-collagen nanofibrils. (f) The average diameter of stabilized W/W emulsion droplets with different concentrations of mPEG-collagen, increases linearly with the volume fraction of the dextran phase (top). The total interfacial area stabilized by mPEG-collagen nanofibrils with changing fibril concentration and volume fraction of the dextran phase (bottom). Download figure Download PowerPoint The mPEG-collagen level needed to stabilize the W/W emulsions was determined by varying the concentrations of mPEG-collagen at 0.05, 0.10, 0.15, and 0.20 wt %. Using optical microscopy (OP), emulsion droplets with well-defined spherical structures were observed and their average droplet sizes were monitored as a function of mPEG-collagen concentration. In Figure 1d, one can observe that the average size of the emulsion droplets becomes smaller with increasing concentration of mPEG-collagen; this principle is applicable for different volume fractions. There is a question whether mPEG-collagen will stick at W/W interfaces like surfactants absorbing at oil/water interfaces, even though the interfacial tension is extremely low. We measured the γ values between the coexisting dextran- and PEG-rich phases by a spinning drop interface tensiometer, following the Vonnegut equation (eq 2): γ = 1 / 4 Δ ρ ω 2 R 3 (2) The details of the measurement and calculation process are shown in Supporting Information Note 1 and Figure S2. A series of γW/W values, at a constant cPEG = 8.0 wt % and varying dextran concentrations from 5.5 to 18.0 wt %, were measured. As shown in Figure 1e, with increasing dextran weight fraction, the γW/W values increase. The γW/W values remain almost constant at 10−5 N/m from cdextran = 5.5 to 14 wt %. Figure 1e also shows that mPEG-collagen, behaving like a surfactant, can effectively reduce interfacial tension. It can be concluded that the mPEG-collagen adsorbs at the W/W interface to stabilize the W/W emulsions. The reduction of interfacial tension can be attributed to the directional adsorption of mPEG-collagen. To further establish the stabilizing effect of mPEG-collagen on the formation of W/W emulsions, we also examined the use of no emulsifier and use of native collagen protein as an emulsifier. The macroscopic appearance of the mixtures is shown in Supporting Information Figure S3a. The blank mixtures of PEG and dextran showed phase separation within 10 min. The dextran-rich phase is in the bottom and the PEG-rich phase is at the top. A similar case was observed in the W/W emulsions with native collagen, implying that native collagen cannot influence the emulsification tendency of PEG/dextran mixtures. A stabilizer undergoes dynamic adsorption and desorption at the interface. The key for stable adsorption at W/W interfaces is to overcome the Brownian movement of the stabilizer. Interfacial adsorption energy (ΔG) is related to the adsorption capacity of the stabilizer at the interfaces. The ΔG depends on the size of the stabilizer (R), the γW/W, and the contact angle (θ), as shown in eq. 3: Δ G = π R 2 γ W / W ( 1 − | cos θ | ) 2 (3) Only if the ΔG of the stabilizer is much larger than the thermal kinetic energy (ΔE) of the stabilizer can the stabilizer be firmly bonded to the W/W interfaces and stabilize the emulsions. The adsorption of stabilizer at the W/W interfaces is a prerequisite for the preparation of W/W emulsions. For the findings in Supporting Information Figure S3, the most extreme case for the adsorption energy of native collagen molecules is given by ΔG = πr2 γW/W = 3.14 × (1.5 nm/2)2 × 10−5 N/m = 1.767 × 10−23 J. It is two orders of magnitude lower than the Brownian thermal kinetic energy (ΔE = kT = 1.38 × 10−23 × 300 = 4 × 10−21 J) because the size of the stabilizer is as small as 1.5 nm.24 The amorphous morphology of native collagen shown in Supporting Information Figure S3b further confirmed this point. Native collagen cannot adsorb at the W/W interfaces to form an effective barrier to prevent droplet fusion. Thus, only when the collagen was formed into a nanofibril resulting in interfacial adsorption and simultaneously satisfying the requirements of size and overall coverage ratio did this lead to stable W/W emulsions. This is confirmed by the microphotograph displaying the uniform emulsion droplet morphology and complete absence of droplet coalescence ( Supporting Information Figure S3c). The results with different volume fractions (φdex) of the dextran-rich phase from 0.1 to 0.5 and different concentrations of mPEG are shown in Supporting Information Figure S1f. At low mPEG-collagen concentrations, the nanofibrils cannot cover the entire droplet. When two small emulsion droplets coalesce, the nanofibrils gather at the interface of the merged droplets, resulting in larger emulsion droplets. When mPEG-collagen concentration is sufficiently large, the mPEG-collagen spreads out evenly over the droplet interfaces before the droplets coalesce, which delays the occurrence of Ostwald ripening and results in smaller emulsion droplets. These findings follow the basic rule of emulsion systems. Using the emulsion droplet size at different φdex values, the total interfacial area (S) can be calculated, as in eq 4: S = 6 ϕ dex / R ′ (4)where R′ is the average size of emulsion droplets. As shown in Figure 1f, the total interfacial area ( A, m2) stabilized by mPEG-collagen is independent of φdex with varying amounts of mPEG-collagen. The reason why the average size of the droplets increases with increasing φdex can also be revealed. At a specific mPEG-collagen concentration, the A stabilized by mPEG-collagen is constant, and as the φdex increases, the amount of mPEG-collagen at the droplet interface is insufficient to cover the entire emulsion droplet. Droplet fusion occurs and R′ becomes larger. A linear increase of A with mPEG-collagen concentration was observed in Figure 1f, providing the following two estimated values: per gram of added mPEG-collagen, the stabilized interface area is 53.49 m2, and the coverage of 1 m2 interfacial area requires 0.01869 g of mPEG-collagen. Stabilization mechanism of a W/W emulsion To determine the effect of composition on the stability of W/W emulsions, emulsions at different φdex values in the presence of different amounts of mPEG-collagen and FITC-dextran were prepared. Emulsion phase inversion occurred when the φdex was over 50% because of droplet stacking. The fluorescence images of W/W emulsions show that green emulsion droplets are formed in continuous phase, dextran-in-PEG emulsions, for a φdex of <50% (Figure 2a). In contrast, for a φdex of more than 50%, emulsion droplets in a green continuous phase, PEG-in-dextran emulsions, are formed (Figure 2b). Figure 2 | Fluorescence images of the W/W emulsions of FITC-modified dextran phase in the presence of 0.2 wt % nanofibrils: cPEG = 6.3 wt % and cdex = 4 wt % (a) and cPEG = 1.9 wt % and cdex = 12 wt % (b). OP images of W/W emulsions in the presence of 0.20 wt % nanofibrils dextran-in-PEG emulsion (cPEG = 6.3 wt % and cdex = 4 wt %) for 1 day (c) and 40 days (d) and PEG-in-dextran emulsion (cPEG = 1.9 wt % and cdex = 12 wt %) for 1 day (f) and 40 days (g). Size distribution of emulsion droplets, for 1 day and 40 days, of dextran-in-PEG emulsion (e) and PEG-in-dextran emulsion (h), respectively. Download figure Download PowerPoint We explored the stability of dextran-in-PEG emulsions and PEG-in-dextran emulsions. For the PEG-in-dextran emulsions, after being incubated for 7 days, the volume of the emulsion phase was reduced, even when the concentration of mPEG-collagen was relatively high (∼0.20 wt %). For the dextran-in-PEG emulsions, there was little change in the volume of the emulsion phase ( Supporting Information Figure S4). This is also observed in the OP images shown in Figures 2c, 2d, 2f, and 2g). The dextran-in-PEG emulsions maintained regular droplet structures, while the PEG-in-dextran emulsion droplets began to fuse. The enhanced stability of dextran-in-PEG emulsions can be explained by the free energy of mPEG-collagen during coalescence. The interfaces of two adjacent emulsion droplets must be in contact with each other before they fuse into one emulsion droplet. If the coalescence takes place, the nanofibrils, as the physical barriers, must enter into the PEG-rich phase or the dextran-rich phase, as shown in Figure 3. For