Abstract
Particle adsorption at the oil-water interface is an important strategy for emulsion stabilization against coalescence, however, the particle occupation of the interface and the requirement of free interfaces in many applications is a contradiction. We report an emulsion type with low droplet surface coverage but enhanced stability by employing colloidal rings as emulsifiers. The formed emulsions provide a large accessible oil-water interface (>80%). The enhanced diffusion through the interface and highly efficient loading of catalytic nanoparticles at the interface result in much higher catalytic efficiency than the nanosphere-covered emulsions in both batch and continuous flow interface catalysis. The loading of plasmonic nanoparticles brings excellent performance in surface-enhanced Raman spectroscopy-based detection, which exhibits the lowest detectable concentration as low as 10−11 M using only 0.25 μL of analyte and 0.2 μg of Au nanoparticles. The ring-based Pickering emulsion provides freedom for designing interface structures and compositions for functional emulsions.
Similar content being viewed by others
Introduction
Pickering emulsions (PEs) are a dispersion formed by dispersing a liquid in another immiscible liquid where the droplets are stabilized by colloidal particles1,2 and provide a general application platform in many fields such as in interfacial catalysis3,4,5,6,7,8,9,10 and surface-enhanced Raman spectroscopy (SERS)11,12,13,14,15,16,17,18,19. PEs have good stability during storing because of the high binding energy of the particles to the liquid-liquid interface which drives the particles to form densely packed layers as mechanical barriers20. The particle layer usually occupies almost the whole emulsion interface. It is about 91% when the packing is ordered hexagonal monolayer; therefore, only ~9% of the emulsion interface is free and accessible. However, the newly emergent applications such as interfacial catalysis and SERS require sufficient free emulsion interface to load functional particles such as enzymes, and plasmonic nanoparticles (NPs).
A usual solution is to use the desired plasmonic NPs directly as the Pickering emulsifier11,12,13,14,16, which are mostly amphiphilic or omniphobic to stably attach to the oil-water interface. To make the NPs to be amphiphilic or omniphobic, the NPs have to be modified with additional molecules. These additional molecules inevitably hinder the access to active sites of these NPs during their use21,22,23. Meanwhile, forming a densely packed layer at the oil-water interface of PEs requires the use of a large number of NPs, which increases the cost of use. One could think that a better solution is to separate the two roles (emulsification and catalysis or SERS) into two different particles. By reducing the droplet coverage of the particle emulsifier, functional NPs are loaded onto the emulsion interface that was not occupied by the particle emulsifier15. This scheme leads to the problem of competition between emulsifiers and functional NPs in the interfacial area. Based on the droplet armoring mechanism, the stability of the emulsion will be significantly reduced due to the reduction of particle emulsifier coverage24. Therefore, it is highly desirable to develop other emulsion types to address these problems.
Here we present a colloidal ring-based PE (RPE) in which the emulsion droplets are covered by a densely packed colloidal ring monolayer. The droplets in RPE still have a very large accessible oil-water interface (~80%) but the emulsions show enhanced stability. We demonstrate that, compared to nanosphere-stabilized PEs (SPE), faster inward/outward diffusion in RPEs leads to high catalytic efficiency (CE) in emulsion interface catalysis. Moreover, large unoccupied emulsion interfaces allow for further loading of diversified NPs, and the resulting NP-loaded emulsions have excellent performance in continuous flow interfacial catalysis and SERS detection.
Results
Implementation and mechanical stability of RPE
RPEs were prepared with the help of a homogenizer (IKA S10N-5G) using ring-shaped SiO2 particles with different hydrophobicity as Pickering emulsifiers. Figure 1a schematically shows the structure of a single RPE droplet and the structure-related two properties: fast inward or outward diffusion and large free interfaces for NP loading. The SiO2 rings employed for preparing RPEs are shown in Fig. 1b. Before emulsion preparation, the rings were modified by phenyltrimethoxysilane (PTMS) to increase the amphiphilicity (Supplementary Fig. S1). The grafted density of PTMS was tuned from 0.35 to 11.9 nm−2 (Supplementary Table S1). In the system of toluene and water, the forming emulsion type (oil-in-water, O/W, or water-in-oil, W/O) depends on the grafting density of PTMS (Supplementary Fig. S1). Low grafting density (<2.10 nm−2) resulted in an O/W type (Fig. 1c). High grafting density (>5.64 nm−2) resulted in a W/O type (Supplementary Fig. S2). The grafting density of PTMS also affects the packing structures of SiO2 rings at the interface. In O/W emulsions, the rings formed hexagonal packed monolayer for low grafting densities (0.35–1.35 nm−2) but tubular structures for high grafting densities (1.70–2.1 nm−2) (Supplementary Fig. S3). The case in W/O emulsions is reversed (Supplementary Fig. S4). The hexagonal monolayer packing of rings in the O/W emulsions can be confirmed by optical microscopy (Fig. 1d). Confocal microscopy confirms that the rings (green) were partially embedded into the dispersed water droplet (Fig. 1e). The diameter of the emulsion droplets obtained is typically tens of microns, depending on the concentration of the rings and the homogenization speed (ω). For the toluene-in-water emulsion, as the ring concentration (Cring) increases from 5 to 50 mg mL−1, the Sauter mean diameter of droplets (DS) gradually decreased from about 160 to 40 μm (Fig. 1f). Similar phenomena have been reported in many previous studies25,26,27,28,29. A qualitative relationship between the droplet diameter and the added particle mass has been given in several studies25,27,28, which show that 1/Ds has a linear relationship (DS ~ Cp−1) to the particle concentration (Cp) when the volume of the phase that the particle was dispersed in and the interfacial coverage are fixed. That is, the product of Ds and Cp should be a constant under these premises. This relationship also holds for RPE except the constant will be changed. However, it was found that in our case the relationship can be only described by DS ~ Cring−0.63 (red line, Fig. 1f). The value of this power exponent deviates from −1. We think that there are two possible reasons: one is that the packing structure of rings is no longer monolayer, and the other is that in the preparation of the emulsions, the rings that are not attached to the interface increase. The former is in agreement with our observations that the rings formed the tubes when the ring concentration is greater than 40 mg mL−1 (Supplementary Fig. S5), and the latter can be evidenced by observing the bottom of the samples after homogenization (Supplementary Fig. S5). Previous studies have shown that a higher homogenization speed ω usually results in a smaller average droplet diameter30,31,32,33,34, which, as expected, is also the case for our RPEs. For ω from 8000 to 30000 revolutions per minute (rpm), the DS decreased from about 110 μm to 51 μm (Fig. 1g and Supplementary Fig. S6). The relationship between the homogenization speed ω and the DS can be well fitted by an exponential model (1):
where the constants a, b, and c were obtained by fitting and are in turn 321.94, 0.21, and 51.25, respectively. Using water-in-1-dodecene emulsions, with silica particles HDK®H20 as the Pickering emulsifier, Kempin et al. reported that the relationship between the homogenization speed ω, which showed the DS follows a power law model31,32. Differently, using fish oil-in-water emulsions, with regenerated silk fibroin as the Pickering emulsifier, Zhong et al. found the relationship follows an exponential model30. The difference in the fitting models may be attributed to different emulsion systems, particle types, and the selected range of ω. A comparison of our data between the power law model and the exponential model is shown in Supplementary Fig. S7, and the exponential model shows a better fit (Supplementary Table S2). Other RPEs with dichloromethane, cyclohexane, ethyl acetate, decane, hexane, octanol, octane, and anisole as the oil phase are shown in Supplementary Fig. S8.
a Schematic showing the structure of RPE and its two main properties. NPs represent nanoparticles. RPE represents ring-stabilized Pickering emulsion. b SEM image of the SiO2 rings used. c Fluorescence microscopy image of toluene-in-water RPE (Ring: 20 mg mL−1; toluene/water: 2/3, v/v). The toluene phase was labelled with the dye Nile red. d Optical microscopy image of a single toluene emulsion droplet covered by a monolayer of rings. e Confocal microscopy image of the emulsion interface. Note: The SiO2 rings are labeled green, and the water phases are labeled orange. f The Sauter mean droplet diameter Ds of toluene-in-water RPE versus the used ring concentration Cring. Cp is the particle concentration and β is the fitting parameter. Exp. represents experimental data. g Ds of toluene-in-water RPE versus the homogenization speed ω. The letters a, b, c are the fitting parameters. Exp. represents experimental data. In (f, g), the mean droplet diameters were obtained by measuring at least 300 droplets. Scale bars: 2 μm (b), 100 μm (c), 20 μm (d), 3 μm (e).
The RPEs exhibit good stability during long-term storage as the arithmetic mean droplet sizes of water-in-toluene PEs did not show significant changes after storing for one month (Supplementary Fig. S9). To further explore the stability of RPEs, centrifugation was utilized and the arithmetic mean droplet sizes were compared using water-in-toluene PEs. Meanwhile, we compared the stability of RPE and SPE by considering a simplified model (Fig. 2a). In this model, an actual ring Ra can be modeled as an equivalent-volume ring Rb, which consists of N interconnected equivalent-volume nanospheres, and the ring Ra and the ring Rb have the same diameter (Fig. 2a). The PEs stabilized by the ring Ra and Rb are RPE1 and RPE2, respectively (Fig. 2b). We speculate that RPE1 and RPE2 do not have significant difference on emulsion stability. Consequently, the comparison of the emulsion stability between RPE1 and SPE can be converted to a comparison between RPE2 and SPE. As long as the rings and the nanospheres have the same surface chemistry, the main difference between RPE2 and SPE is that, in RPE2, the SiO2 nanospheres first form Rb-like rings and then stabilize the PE. If the emulsion stability of RPE2 (or RPE1) is better than that of SPE, this indicates the necessity of using nanospheres to form rings. By this model, the diameter of nanospheres was determined to be about 351.6 nm. Therefore, we selected 353 ± 11 nm SiO2 nanospheres to prepare SPE for the comparison of emulsion stabilization. For both PEs, the added mass fraction of particles, oil-water fraction, and the used oil phase are the same. The initial droplet sizes are also similar as they are 112 ± 22 μm for RPE1 and 107 ± 17 μm for SPE. To avoid significant deviations caused by different degrees of modification, the modification with a wide range of PTMS grafting densities was performed for both particles. An optimized grafting density is chosen by selecting the grafting density that results in the smallest droplet size (Supplementary Fig. S10 and Supplementary Table S3). During centrifugation, for RPE1, no significant coalescence between the droplets was observed until 3302 g, while for SPE, the coalescence occurred at only 606 g (Fig. 2c, d). The results were confirmed by monitoring the arithmetic mean droplet diameters (Fig. 2e). The high stability of RPE may be attributed to the fact that the rings have higher interfacial desorption energy than nanospheres (Fig. 2f, g), which was calculated using a reported method35. In principle, if the nanospheres can form Rb-like rings at the emulsion interface, the obtained SPE has similar centrifugation stability to that of the RPE, although it seems very difficult to form this ring from the nanospheres. This, in turn, illustrates the necessity of preparing highly mechanically stable emulsions with rings. We further prepared SPEs using SiO2 spheres with a diameter of 92 ± 5 nm, 1105 ± 30 nm, and 2072 ± 38 nm (comparable to the diameter of SiO2 rings) and compared their centrifugation stability to the RPE. Without exception, these emulsions show worse centrifugation stability even with the optimum grafting density of PTMS (Supplementary Figs. S11–S14 and Supplementary Tables S4–S6).
a Schematic SiO2 ring Ra, the equivalent-volume ring model Rb and SiO2 spheres. Each SiO2 sphere have the same volume to that in the ring Rb. b Schematic corresponding PEs stabilized by the three particles shown in (a). RPE1 is the Pickering emulsion stabilized by the ring Ra. RPE2 is the Pickering emulsion stabilized by the ring Rb. SPE is Pickering emulsion stabilized by SiO2 spheres. c, d Fluorescence microscopy images of (c) RPE1 and (d) SPE under an increasing centrifugal force. The PEs are water-in-toluene emulsions, and the centrifugation time is 8 min. The insets are the digital images of the RPE1 or SPE just after centrifugation. The water phase was labelled with the dye RhB. e The arithmetic mean diameter of the emulsion droplets for RPE1 and SPE as a function of centrifugal force. The mean droplets sizes were obtained by measuring at least 300 droplets. Data are expressed as mean ± standard deviation of n = 300 independent droplets. f Schematic showing the wetting of the ring Ra and a single nanosphere for the calculation of the desorption energy. The ring Ra has a rectangular section. The letter θ represents the contact angle of water in a W/O emulsion. γwo represents the oil-water interfacial tension. d1 is the diameter of the ring Ra and Rb. g Comparison of the desorption energy between a single ring Ra and a single nanosphere. The outer and the inner diameters of the rings used are 2.16 ± 0.04 μm and 1.88 ± 0.04 μm, and the height is 0.46 ± 0.08 μm. Consequently, SiO2 nanospheres with a diameter of 353 ± 11 nm are selected for SPE that was used in (d). The diameter of 353 nm was used for calculation in (g). At the beginning, both emulsions have similar mean diameters of about 110 μm. Scale bars: 100 μm.
Enhanced diffusion of RPE for enzyme catalysis
The immediate advantage of PEs with large unoccupied emulsion interfaces may be the rapid diffusion of the substance from the oil phase to the aqueous phase or from the aqueous phase to the oil phase. To examine this, water-in-cyclohexane RPE and SPE were prepared (diameter ~158 μm), and rhodamine B (RhB) was used as a probe molecule (Fig. 3a). The large mass density mismatch resulted in the deposition of ring-armored water droplets at the bottom. The cyclohexane phase in RPE during the preparation exhibits a redder color than SPE (stabilized by 353 ± 11 nm SiO2 spheres), suggesting that the RhB in RPE diffuses outward faster than SPE (Fig. 3a, b). To accelerate the diffusion, n-octanol (a good solvent of RhB) was added to the cyclohexane phase until 2.8% v/v. This addition did not disrupt the packing structure of rings at the interface (Supplementary Fig. S15) but significantly improved the solubility of RhB in the cyclohexane phase. Upon addition, the RhB concentration in the cyclohexane phase of RPE quickly increased to 3.74 times that of SPE (Fig. 3c), indicating that RPE has a faster diffusion rate. Over time, the increase in RhB concentration in the cyclohexane phase of RPE was pronouncedly faster than that of SPE, but the concentration ratio remained the same at about 3.8.
a Digital photo of the diffusion of RhB from the aqueous phase to the cyclohexane phase. b Schematic showing the difference in diffusion rate between RPE and SPE. The oil phase is a mixture of cyclohexane and octanol. RhB represents Rhodamine B. c Left: absorbance of RhB in cyclohexane phase as a function of time; right: the absorbance ratios (ARPE/ASPE) of RhB in cyclohexane phase between RPE and SPE as a function of time. d Schematic hydrolysis of p-NPP into p-nitrophenol catalyzed by lipase in RPE. e, f Comparison of (e) kinetic curves and (f) SA between different PEs for the hydrolysis of p-NPP into p-nitrophenol. p-NPP: 100 mg mL−1. Data are expressed as mean ± standard deviation of n = 3 independent experiments. g Schematic oxidation of pyrogallol into purpurogallin catalyzed by hemoglobin. h, i Comparison of (h) kinetic curves and (i) SA between different PEs for the oxidation of pyrogallol into purpurogallin. Pyrogallol: 130 mM. Data are expressed as mean ± standard deviation of n = 3 independent experiments. The SA of lipase was the activity per unit amount of the enzyme (U mg–1) in the reactions. The activity (U) was defined as 1 μmol product produced within 1 min. To obtain the SA, all reactions were repeated at least three times.
The fast diffusion makes RPE an excellent platform for catalytic reactions where the substrate is in one phase and the product is in another, for example, the hydrolysis of 4-nitrophenyl palmitate (p-NPP, oil-soluble) to produce p-nitrophenol (water-soluble). Toluene-in-water RPE was chosen and Aspergillus oryzae lipase was added to the water phase as the catalyst. The lipase was found to be interfacial active and tended to be located at the emulsion interface as shown by confocal microscopy images (Supplementary Figs. S16–S18). Meanwhile, these images suggested that some lipase possibly adsorbed on the SiO2 particle surface. The hydrolysis reaction of p-NPP requires the generated p-nitrophenol can quickly diffuse into the continuous water phase to promote the formation of more p-nitrophenol (Fig. 3d). An RPE with a diameter of 76 ± 22 μm was prepared. Several other emulsions including SPE, DPE (stabilized by SiO2 discs), and TPE (stabilized by tubes that are formed by stacked rings) with similar diameters have been used for comparison. During the comparison, apart from the shape and size of the particles, other experimental conditions such as the particle mass fraction, oil-water phase fraction, and the concentration of the added catalysts remain the same. The droplet sizes are similar (Supplementary Fig. S19). For SPE, SiO2 nanospheres with an arithmetic mean diameter of 270 ± 11 nm were chosen, as many previous studies using SPE for emulsion interfacial catalysis have used diameters of approximately 200–300 nm36,37,38,39. This diameter is also close to the size (351.6 nm) of SiO2 nanospheres in the considered model (Fig. 2a, b), which has been used for emulsion stability comparison. For DPE, the hollow SiO2 discs with a diameter of 2.19 ± 0.07 μm are used as this diameter is similar to the diameter of the used rings (2.16 ± 0.04 μm). With these size choices, we expect the comparisons to reflect the performance of the RPE as closely as possible. For each type of SiO2 particles, the modification with a wide range of PTMS or butyltrimethoxysilane (BTMS) grafting densities was performed and an optimized grafting density was selected by selecting the grafting density that results in the smallest droplet size (Supplementary Figs. S20 and S21 and Supplementary Tables S7 and S8).
It was found that the conversion reached about 80% for only 1 h with an A. oryzae lipase concentration of 0.041 mg mL−1 and was more than 90% for 2 h (Fig. 3e and Supplementary Figs. S22 and S23). SPE and DPE showed conversions of only 51% and 47% for 1 h. Obviously, the RPE shows the fastest hydrolysis rate. Interestingly, when the rings stacked themselves into tubes at the oil-water interface and the RPE became TPE, the conversion significantly decreased to 53% for 1 h. For RPE, the lipase showed a specific activity (SA) of 9.24 μmol min−1 mg−1 (Fig. 3f), which is 2.2, 2.4, and 2.1 times that of SPE, DPE, and TPE. The high SA for RPE is not limited to a specific p-NPP concentration (Supplementary Figs. S24–S26). To explore the reason that RPE has the highest SA, we first measured the loading amount of lipase at the toluene-water interface for the four types of PEs by measuring the lipase concentration in the continuous water phase after adding lipase to PEs and shaking the PEs (Supplementary Fig. S27). The obtained lipase amounts at the toluene-water interface are 32%, 36%, 41%, and 30% w/w for RPE, DPE, SPE, and TPE, respectively. These amounts are similar, but the free interface of RPE is much larger than that of the other three PEs, therefore, some lipase must be adsorbed on the surface of the SiO2 particles. Similar adsorption phenomena have previously been reported using other enzymes and PE systems40,41. According to previous studies42,43,44, the lipase could only be activated at the oil-water interface because, at the oil-water interface, the amphiphilic peptidic loops covering the active sites of the enzyme in solution undergo a conformational rearrangement, making the active site easily accessible. Our controlled experiment also shows that the conversion time remains almost the same even though 90% v/v of lipase was removed from the continuous water phase (Supplementary Fig. S28), also implying only those lipases at the emulsion interface or the three-phase lines could act as effective catalysts. Therefore, for the four types of PEs, RPE should provide the most accessible lipase catalytic sites. Because of the difficulty of having the same number of accessible active catalytic sites for these four types of PEs, we cannot distinguish between faster diffusion or more catalytic sites, which one contributes more to high SA.
However, using water-in-toluene PE for the oxidation of pyrogallol into purpurogallin by hemoglobin, the role of diffusion can be separated because the hemoglobin can catalyze the reaction both in the water phase and at the interface45,46. In this case, the accessible catalytic sites should be the same for all four PEs. The pyrogallol was dispersed in the dispersing water droplets. The product purpurogallin will also be produced in the dispersed water phase, but it has higher solubility in the continuous toluene phase45,47. Therefore, it requires that the generated purpurogallin can quickly diffuse from the aqueous phase to the toluene phase to facilitate the reaction. Again, the choice of the particle size is the same as that in Fig. 3d–f, and all the PEs have similar arithmetic mean droplet diameters (~70 μm) (Supplementary Fig. S29). The optimized modification by PTMS or BTMS grafting was selected for each PE (Supplementary Figs. S30 and S31 and Supplementary Tables S9 and S10). Except for the particle shape and modification, all the other experimental conditions are kept the same. Confocal microscopy images show hemoglobin stays both in the droplets and at the emulsion interface (Supplementary Fig. S32). With the same concentration of hemoglobin (0.079 mg mL−1), RPE still showed much faster conversion than the other three PEs (DPE, SPE, and TPE) (Fig. 3g and Supplementary Fig. S33). The SA of RPE (6.36 μmol h−1 mg−1) was 5.0, 4.6, and 5.7 times that of SPE, DPE, and TPE (Fig. 3h, i). Because of the same reactant concentration and accessible catalytic sites, the high SA should be attributed to faster diffusion of the products from the dispersing water phase to the continuous toluene phase. Note that high hemoglobin concentrations should increase the hemoglobin coverage at the interface, which also reduces the outward diffusion of the product purpurogallin. However, our control experiment showed that hemoglobin at the oil-water interface had no significant effect on the outward diffusion of purpurogallin at our selected low hemoglobin concentration (0.079 mg mL−1) (Supplementary Fig. S34). The quantitative relationship between the area of the available free interface and the productivity of RPE in catalysis still needs to be further studied in the future.
NP-loaded functional RPE
A large unoccupied emulsion interface can provide space for loading diversified NPs to construct functional emulsions. Because colloidal rings are responsible for stabilizing the emulsion, NPs can be added in the dispersing phase as long as NPs can spontaneously migrate to the interface. The self-assembly of NPs at the liquid-liquid interface has been achieved by reducing the surface charges or increasing the hydrophobicity48,49,50,51. We realized the spontaneous migration of citrate-stabilized Au NPs (20–130 nm) by simply reducing the number of ligands (citrate group), and the aggregation of NPs can be effectively reduced by adding a given amount of tetramethylammonium hydroxide (TMAH) (2.0 mM) (Supplementary Fig. S35). The same method has been applied to various NPs, such as nanospheres, nanorods, nanoplatelets, and nanostars, to obtain corresponding NP-loaded PE (Supplementary Figs. S36–S38). In the case of Pd NP-loaded PE, no TMAH was added and the Pd NPs did not show any aggregation. In the following, Pd NP-loaded PE will be used for batch and continuous flow interfacial catalysis, and Au NP-loaded PE will be used for SERS detection.
Pd NP-loaded RPE for batch and continuous flow interfacial catalysis
The continuous flow interfacial catalysis based on SPE has been demonstrated to have enhanced reaction efficiency in comparison to batch reactions52,53. In the case of SPE, the emulsion interface is occupied by nanospheres, and the catalyst stays in the interstice among the nanospheres and in the droplets. Since all the substrates are in the flow phase, only those catalysts at the interface can effectively catalyze the reaction. As a result, the limited accessible emulsion interface significantly reduced the catalytic active sites. For NP-loaded RPE, NPs dominantly situated at the interface (Fig. 4a), therefore, the flow substrate could contact the catalyst more efficiently (Fig. 4b). Consequently, the conversion of the substrate might be accelerated. To demonstrate this, Pd NP-loaded water-in-toluene RPE (186 ± 52 μm) was prepared (Supplementary Fig. S39). The emulsion shows a black appearance (Fig. 4c, left-down). On the droplet surface, the rings have a hexagonal close packing (Fig. 4c). Pd NPs formed a loose packing in the hollow cavity of the ring particles as well as the interstice among the ring particles (Fig. 4d, e). Cryo-SEM further confirmed that the ring formed a hexagonal packed monolayer on the droplets (Fig. 4f). The arrow in Fig. 4g indicates the layer of Pd NPs that covered the droplet. Inside the water droplet, Pd NPs are hardly observed, implying that Pd NPs dominantly stay at the interface. In the cavities of some rings, it is possible to clearly distinguish between areas covered or uncovered by Pd NPs (Fig. 4h). All these results give sufficient evidence that catalytic Pd NPs indeed dominantly situated at the unoccupied free oil-water interface of RPE. In comparison, Pd NPs can be observed only in very limited areas of interfaces of SPE (Fig. 4i).
a Schematic showing the Pd NPs located in the hollow cavities of rings. b Schematic showing the continuous flow catalytic oxidation of benzyl alcohol to benzaldehyde by Pd NP-loaded RPE. PE represents Pickering emulsion. TBHP represents tert-butyl hydroperoxide. c Optical microscopy image of a single Pd NP-loaded emulsion droplet. Left-down: the appearance of Pd-NP loaded PEs. Right-up: a zoomed-in image of the area shown in the white square. d, e TEM images of the emulsion interface of Pd NP-loaded RPE. f–h Cryo-SEM images show the emulsion interface of Pd NP-loaded RPE. i Schematic and cryo-SEM images indicate the location of Pd NPs at the interface of Pd NP-loaded SPE. j, k Comparison of kinetic curves and CE among RPE and SPE for the oxidation of BA in batch reaction. l, m Comparison of kinetic curves and CE for the oxidation of BA in continuous flow catalytic reaction. Scale bars: 20 µm (c), 1 µm (d), 25 nm (e), 50 µm (f), 2 µm (inset in f), 500 nm (g, h, i).
The oxidation of benzyl alcohol (BA) by Pd NPs was established as a model reaction in the RPEs to examine the performance for continuous flow interface catalysis. A mixture of toluene and decane (4/1, v/v) was used instead of toluene as the oil phase, and Pd NP aqueous dispersion was used as the dispersing phase. In a batch reaction, the BA solution containing tert-butyl hydroperoxide (TBHP) in toluene/decane was added to the RPE to initiate the reaction. Similar experiments with SPE of arithmetic mean droplet diameter (174 ± 38 μm) were performed for comparison. Likewise, SiO2 nanospheres of 270 ± 11 nm were chosen. The reaction was performed for 12 h. The conversion of the SPE was only 35.0%, however, the conversion of the RPE reached 59.5% (Fig. 4j). The CE of RPE increased by 70% compared to that of SPE (Fig. 4k). In a continuous flow reaction, Pd NP-loaded RPE (181 ± 42 μm) and SPE (177 ± 48 μm) (Supplementary Fig. S40) were packed into a glass column (inner diameter: 1.34 cm) with a height of 7.5 cm. A mixed toluene/decane solution of BA (0.065 M) and TBHP (0.5 M) was pumped into the inlet and exited from the outlet at a steady flow rate of 0.5 mL h−1. The obtained conversions were plotted as a function of time (Fig. 4l). In the first 4 h, the conversion of RPE quickly reached 84.6%, then dropped back slightly and remained around 81.5% for 40 h. The conversion of SPE was only around 45.8%. The CE for RPE was estimated to be 0.087 mol mol−1 h−1, which increased by 90% compared to that of SPE (Fig. 4m).
Au NP-loaded RPE for SERS detection
Another important application of RPE is for SERS detection. Plasmonic NP-stabilized PEs are an effective molecular sensing platform based on SERS11. The stabilization of the emulsions requires the NPs to be modified with amphiphilic or omniphobic ligands. Since the SERS signal is mainly coming from the molecules that are situated in the hot spots, the ligands inevitably disturb the detection. In principle, the large unoccupied free interface of RPE no doubt provides a great opportunity for loading plasmonic NPs with only weak bonded ligands, which probably greatly lowers the limit of detection (LOD) of the SERS. Au NP-loaded RPE was used for the following SERS detection. First, the SERS active area of the droplets was investigated with malachite green (MG, 10−6 M) as the model analyte by analyzing the SERS signal intensity along the diameter of a single droplet (Fig. 5a). The focus was on the focal plane of Au NPs at the emulsion interface. The intensity of the characteristic peak at 1616 cm−1 was recorded, which showed that the largest intensity occurs at the location near the edge (Fig. 5b). This may be because this location takes full advantage of the three-dimensional nature of the emulsion and provides the maximum hotspot volume11. Outside of the emulsion droplet, the signal intensity is almost zero. To further explore the intensity distribution of the signal over the whole droplet surface, we measured the signal intensity along three different lines through the geometrical center of the droplet (Supplementary Fig. S41a). Almost overlapped intensity curves suggest that the intensity distribution is radically symmetrical (Supplementary Fig. S41b). The largest intensity area is roughly between 0.36d and 0.44d (d: the droplet diameter), which is independent of the droplet sizes in the selected size range (Supplementary Fig. S42). Consequently, as long as the measurement is performed in this annular area, the strongest signal can be safely reproduced. Of course, the amount and accumulation of gold NPs at the interface, as well as the resulting number of hot spots, are closely related to the concentration of gold NPs in the aqueous phase when preparing the emulsion. To investigate this effect, crystal violet (CV, 10−7 M) was chosen as the model analyte. The intensity of the characteristic peak at 1615 cm−1 was plotted as a function of the concentration of Au NPs (Fig. 5c and Supplementary Fig. S43). Surprisingly, it was found that the SERS intensity increased first and then decreased when the concentration of Au NPs increased from 0.16 mg mL−1 to 2.4 mg mL−1. Increased NPs amounts at the interface increased the number of hotspots, thus increasing the signal intensity. Meanwhile, increased NPs amounts at the interface led to a decrease in the number ratio of analytes to NPs, thus decreasing the signal intensity. Therefore, the Au NP concentration dependence of the signal intensity is a combined result of the two causes.
a Schematic illustration of SERS detection. The image shows an emulsion droplet under a Raman microscope. b SERS intensity distribution of an emulsion droplet along the dashed line shown in (a). MG concentration in the droplets: 10−6 M. Data are expressed as mean ± standard deviation of n = 3 independent experiments. c SERS intensity of the characteristic peak of CV (10−7 M) at 1615 cm−1 for increasing Au NP concentrations. Data are expressed as mean ± standard deviation of n = 10 independent experiments. d SERS spectra of MG in RPEs for decreasing MG concentration. e The obtained linear relationship between the intensity and MG concentration. Data are expressed as mean ± standard deviation of n = 5 independent experiments. f SERS intensity for decreasing volume of aqueous phase containing MG (10−11 M). g–j SERS spectra of (g) CV (10−10 M), (h) MNBI (10−7 M), (i) ATP (10−6 M) and (j) adenine (10−8 M). The grey curves in (g–j) are SERS spectra when the analyte concentration is 0 M. The dashed lines indicate the positions of characteristic peaks. k Schematic dual phase detection of phoxim (in toluene, 10−4 M) and paraquat (in water, 10−4 M) using Au NP-loaded RPE. Scale bar in (a) is 100 µm.
To examine the LOD of RPE-based SERS, MG solution with a concentration ranging from 10−6 M to 10−11 M was used as the target analyte. At 10−6 M, about 475 MG molecules bind to one NP. At 10−11 M, about 210 NPs share one MG molecule. The SERS spectra of these samples in the low concentration region of 10−11 ~ 10−9 M are shown in Fig. 5d and the SERS spectra in the high concentration region of 10−9 ~ 10−6 M are shown in Supplementary Fig. S44a. As expected, a decrease in analyte concentration leads to a decrease in SERS intensity. However, it was found that even for 10−11 M, the signal of MG can be stably observed. The corresponding analytical enhancement factor (AEF) is 8 × 108 (Supplementary Fig. S45). The relationship between the SERS intensity and the MG concentration shows two linear regions, one corresponds to the low concentration region of 10−11 ~ 10−9 M (Fig. 5e) and the other corresponds to the high concentration region of 10−9 ~ 10−6 M (Supplementary Fig. S44b). The former can be fitted with the equation y = 5.5 × 1011x + 216.8 (R2 = 0.992) and the latter can be fitted with the equation y = 3.6 × 1010x + 3370 (R2 = 0.985). Using the slope of the calibration curve at low concentration region, the LOD was determined to be 2.0 × 10−12 M using the formula LOD = 3δ/s, where δ is the standard deviation of the blank measurement (n = 10), and s is the slope of the linear fitting equation at the low concentration region54,55,56. These SERS spectra indicate a high signal reproducibility but the fittings provide standard curves for quantitative analysis. In addition, the analyte volume required for detection can be reduced to 0.25 μL without affecting the lowest detectable concentration (Fig. 5f). Such a small required analyte volume means that the needed amount of Au NPs is only 0.2 μg for a single detection. Sub-microliter volume detection is of great significance in the field of clinical diagnostics and toxin detection.
To demonstrate the generality of RPE for SERS detection, another four molecules, CV, mercapto-5-nitrobenzimidazole (MNBI), adenosine-5′-triphosphate (ATP) and adenine were used (Fig. 5g–j). The achieved lowest detection concentrations in turn reached 10−10 M, 10−7 M, 10−6 M, and 10−8 M, and the AEFs of CV, MNBI, and ATP were 9.0 × 107, 4 × 104 and 6.4 × 104 (Supplementary Figs. S46–S49). We did not calculate the AEF for adenine as the Raman signal of the adenine characteristic peak of the aqueous bulk solution cannot be obtained even at its saturated concentration (Supplementary S49d). We plotted the calibration curves for the four molecules at the low centration region (Supplementary Fig. S50). Using these calibration curves and a similar method to MG, we determined the LOD of the four molecules: 9.5 × 10−12 M (CV), 6.9 × 10−8 M (MNBI), 1.3 × 10−6 M (ATP), and 1.0 × 10−8 M (adenine). A significant advantage of using NP-loaded RPE as SERS sensors is their biphasic nature, which allows the construction of biphasic sensors. Specifically, plasmonic NPs situating at the water-oil interface allows analyte molecules to enter the hotspots from both water and oil phases, therefore, it is possible to simultaneously detect both water-soluble and oil-soluble analytes. This is important for the practical use of SERS testing, such as the detection of pesticide residues in vegetables or fruits. For example, we chose two widely used pesticides, paraquat (water-soluble) and phoxim (oil-soluble), as proof-of-principle (Fig. 5k). Paraquat is added into the water phase but phoxim is added into toluene phase for preparing Au NP-loaded RPE. From the obtained SERS spectra, when their concentrations are both only 10−4 M, their characteristic peaks can be observed and distinguished, demonstrating the feasibility of biphasic detection with RPE.
Discussion
We have presented an unconventional type of PE by stabilizing the emulsion droplets with a hexagonal monolayer of colloidal rings that have a special hollow cavity. Unlike conventional PEs, the hollow cavities of the rings afford the formed droplets a large, accessible unoccupied oil-water interface. In particular, this does not come at the expense of the mechanical stability of the PE, and our results even show that RPE has much better centrifugation stability than SPE under the given experimental conditions. Large unoccupied oil-water interface provides many possibilities for improving the performance of PEs in applications, as we demonstrated, emulsion interface catalysis and SERS. For the former, the CE of the emulsion using RPE is much higher than SPE. For the latter, the loading of plasmonic NPs onto the empty interface for SERS exhibits a lowest detectable concentration as low as 10−11 M using only 0.25 μL of analyte and 0.2 μg of Au NPs. Considering the practical use of RPE, we estimated the synthesis cost at about 117 CNY per gram of SiO2 ring (Supplementary Tables S11–S13). Of course, the actual cost will exceed this number, because this number only applies to raw materials. However, we may note that this number is not higher than the cost of those commercial SiO2 spheres. Considering the improved interfacial catalysis and SERS performance, we believe that the three-step synthesis of the SiO2 ring is quite worthwhile. Looking future, using rings to stabilize emulsions will make the droplets more like pure droplets with a clean oil-water interface, which, we strongly believe, will create freedom for designing the interface structures and compositions or the mass transfer between the inside and outside of the droplets or the communication between the droplets, and significantly expand the use of PEs in a wide variety of applications, ranging from advanced emulsions, smart capsules to biological detection and vaccination.
Methods
Materials
Malachite green (95%), phoxim solution (1 mg/mL in methanol), A. oryzae lipase (300000 U/g), albumin from Bovine Serum (BSA, 96%) and crystal violet (AR) were obtained from Aladdin Bio-chem (Shanghai). Hemoglobin (64.5 kDa) and tert-Butyl hydroperoxide solution (5.0–6.0 M in n-decane) were obtained from Macklin Biochemical (Shanghai). Pyrogallic (99%), adenine (99%) and adenosine triphosphate (95%) were obtained from Beijing Innochem. n-BTMS (97%) and PTMS (97%) were obtained from J&K Scientific. Dichlorodimethylsilane (DDS, 98.5%) was obtained from Sigma-Aldrich. Benzyl Alcohol (MBA, 99%) was obtained from Tokyo Chemical. 2-Mercapto-5-nitrobenzimidazole (MNBI) and 4-nitrophenyl palmitate (p-NPP, 98%) were obtained from Alfa-Aesar.
Synthesis of SiO2 colloidal rings
SiO2 rings were synthesized by a three-step procedure57. (1) Polystyrene (PS) particles (1.21 ± 0.04 μm) were first prepared following a reported procedure58. 3.78 g of polyvinylpyrrolidone (PVP, Mw: 40000), 0.20 g of azobis(isobutyronitrile) (AIBN), 180.00 g of anhydrous ethanol, and 22.0 0 mL of styrene were added into a flask (250 mL). The mixture was deoxygenized with nitrogen bubbles for 10 min. The polymerization was performed in an oil bath at 70 °C for 8 h with a magnetic stirring (400 rpm). The resulting PS particles were purified by centrifugation/re-dispersion cycles with ethanol for 3 times. The obtained PS particles were dispersed in 100 g of anhydrous ethanol for use. Afterward, the ethanol dispersion of the particles (100 g, 76.3 mg mL−1), 96.00 g of anhydrous ethanol, 0.30 g of AIBN, 3.78 g of PVP (Mw: 40000), 0.30 mL of [2-(methacryloyloxy)ethyl]trimethylammonium chloride solution (DMC, 80% in water), 6.00 mL of styrene (St) and 6.00 mL of water was added to a flask (250 mL). After the dispersion was deoxygenized with N2 flow for 15 min, the polymerization was performed at 70 °C for 1 h under magnetic stirring. The particles were sequentially purified with ethanol and water by centrifugation/re-dispersion cycles. Afterward, the particles were dispersed in water with a concentration (26.0%, w/w). (2) The above PS particles were transformed into PS-decane composite droplets. 0.10 g of AIBN, 0.40 g of PVP (Mw: 58000), 3.85 g of the above PS particle dispersion (contains 1.00 g of PS particles), 32.00 g of anhydrous methanol and 5.15 g of water were added to a flask (100 mL). Afterward, 1.21 g of 2-ethylhexyl methacrylate (EHMA) and 6.84 mL of decane were added. The dispersion was deoxygenized with nitrogen flow for 15 min. The polymerization was performed at 60 °C for 18 h under magnetic stirring. The dispersion was directly used in the next step without further treatment. (3) The dispersion obtained in (2) was used for the synthesis of SiO2 rings. 50.40 mL of the above PS droplet dispersion, 121.20 g of anhydrous methanol, 20.88 g of water, and 21.60 mL of ammonium hydroxide (25.0%, w/w) were mixed in a flask (250 mL). After stirring for 10 min, 2.40 mL of tetraethyl orthosilicate (TEOS) was added every 1 h up to 9.60 mL. The reaction was performed at 30 °C for 6 h under magnetic stirring. The resulting PS/SiO2 core/shell discs were washed with butanol by repeated centrifugation/re-dispersion cycles. The PS core was removed by tetrahydrofuran (THF), resulting in SiO2 rings. The obtained SiO2 rings have a diameter of 2.16 ± 0.05 μm.
Synthesis of SiO2 hollow discs
SiO2 hollow discs were synthesized using a three-step procedure. The first two steps are the same as the synthesis of SiO2 rings. The difference is in the late stage of the second step. The composite PS-decane droplets were broken by adding butanol to directly obtain PS discs. In the third step, the synthesized PS discs were used as a template to prepare hollow SiO2 discs. 125.00 mg of hexadecyltrimethylammonium bromide (CTAB), 256.00 g of ethanol, 64.00 mL of H2O, and 15.00 mL of disc butanol dispersion (contains 1.00 g PS discs) were mixed in a flask (500 mL). The flask was put in an ultrasonic bath for 20 min. 2.40 mL of ammonium hydroxide (25.0%, w/w) was added and the mixture was magnetically stirred for 10 min. 1.50 mL of TEOS was added and the reaction was performed at 30 °C for 3 h. The resulting PS@SiO2 core/shell discs were purified by centrifugation/re-dispersion cycles with ethanol. This process can be repeated one time to grow a thicker SiO2 shell. THF was used to remove the core PS discs, resulting in SiO2 hollow discs. Finally, a final SiO2 shell with a thickness of 40 nm was obtained.
Synthesis of SiO2 nanospheres
The synthesis follows a procedure reported by a previous study59. 9.00 mL of TEOS and 180.00 mL of anhydrous ethanol were mixed in a flask (250 mL) under magnetic stirring. Then a mixed solution of 15.40 mL of ammonium hydroxide (25.0%, w/w) and 4.46 mL of water was slowly added. The reaction was performed at 30 °C for 12 h. Afterward, the SiO2 nanospheres were purified by centrifugation/re-dispersion cycles in turn with water and ethanol. The resulting SiO2 nanospheres have a mean diameter of 270 ± 11 nm. To prepare SiO2 nanospheres with a mean diameter of about 350 nm, when the above reaction was performed for 10 h, an extra 7.00 mL of TEOS and 5.00 mL of ammonium hydroxide (25.0%, w/w) were added. The reaction was performed at 30 °C for another 10 h.
Synthesis of SiO2 rings labeled by FITC
The synthesis of SiO2 rings labeled by fluorescein isothiocyanate (FITC) was similar to that of non-fluorescent SiO2 rings, and the only difference was in step (2) using a mixture of 2.4 mL TEOS and 0.5 mL of FITC solution instead of pure TEOS. The FITC solution was prepared as follows: 20.00 mg of FITC and 80.00 μL of (3-aminopropyl) trimethoxysilane (APTES) was dissolved in 4.00 mL of anhydrous methanol, and the solution was shaken at room temperature for 6 h in the dark. To avoid the interference of the dye molecules on the surface, an extra layer of pure SiO2 was grown. To this end, 100.00 mg of fluorescent SiO2 rings were dispersed in 160.00 g of ethanol containing 40.00 g of water and 2.50 mL of ammonia (25.0%, w/w). The dispersion was stirred for 10 min and then 0.30 mL of TEOS was added. The reaction was performed at 30 °C for 4 h under magnetic stirring. The resulting SiO2 rings were purified by centrifugation/re-dispersion cycles with ethanol. Similar methods were used for labeling SiO2 discs and SiO2 nanospheres.
Modification of SiO2 particles for Pickering emulsifier
The separated SiO2 particles by centrifugation from water were redispersed in 18 mL of 30% H2O2 in a round flask, and then 42 mL of H2SO4 (98%) was slowly added under magnetic stirring. The flask was put in an oil-bath at 90 °C for 3 h (Caution: Piranha solution reacts violently with organic materials, and it must be handled with extreme care). The SiO2 particles were washed in turn with water and ethanol by centrifugation/re-dispersion, and finally dispersed into toluene. For silane modification, a given amount of PTMS was added to 12.00 mL of ring toluene dispersion (containing about 120 mg of rings). The dispersion was refluxed at 120 °C for 12 h under magnetic stirring. The resulting particles were washed by repeated centrifugation/re-dispersion cycles in turn with toluene and ethanol. The obtained modified SiO2 rings were dried in an oven at 120 °C for 3 h. The modification degree was controlled by controlling the added amount of PTMS. For the used amount of 0.03 μL, 0.15 μL, 0.30 μL, 1.50 μL, 3.00 μL, 6.00 μL, 10.0 μL, 15.00 μL, 30.00 μL, 60.00 μL, and 120.00 μL, the obtained SiO2 rings were named sequentially as R1-R11. When other silanes such as BTMS and dichlorodimethylsilane (DDS) were used, a similar procedure was employed only using BTMS or DDS instead of PTMS.
Labeling lipase or hemoglobin with RhB
10.00 mg of A. oryzae lipase or hemoglobin was added to 5.00 mL of freshly prepared PBS (phosphate buffer: 0.1 M Na2HPO4-0.1 M NaH2PO4, pH 7.0). 100.00 μL of RhB solution (1 mg·mL−1 in dimethyl sulfoxide) was added to the suspension under magnetic stirring. After stirring 12 h at room temperature in the dark, the suspension was purified by a dialysis bag (14000 Da) to remove the excess RhB. After 4 days of dialysis, RhB-labeled lipase or RhB-labeled hemoglobin was obtained.
Determination of the concentration of lipase or hemoglobin (Bradford method)
20.00 mg of Coomassie brilliant blue G-250 (CBB) and 20.00 mL of phosphoric acid (85%, w/w) were dispersed in 10.00 mL of ethanol (95%, w/w). The solution was diluted to 200.00 mL with deionized water to obtain the CBB solution. 1.00 mL of bovine serum albumin (BSA) dispersion was added into 4.00 mL of CBB solution. The resultant mixture was incubated for 10 min at room temperature for UV–vis determination (595 nm). A blank experiment was also conducted in the absence of protein. A UV–vis standard curve for BSA assay was obtained using different BSA concentrations with 0.02, 0.04, 0.06, 0.08, and 0.10 mg mL−1. For the test of lipase and hemoglobin, the same procedure was used except for using lipase and hemoglobin dispersion instead of BSA dispersion. The peak at 595 nm was used to analyze the lipase or hemoglobin concentration based on the standard curve.
Determination of the concentrations of p-Nitrophenol and purpurogallin
p-Nitrophenol was dissolved in 0.50 M Na2CO3 solution for UV–vis determination (400 nm). The blank experiment was also conducted in the absence of p-nitrophenol. A UV–vis standard curve for p-nitrophenol assay was obtained using p-nitrophenol concentrations of 0.02, 0.04, 0.06, 0.08, 0.10, 1.20, and 1.40 mg L−1. Purpurogallin was dissolved in toluene for UV–vis determination (425 nm). The blank experiment was conducted in the absence of purpurogallin. A UV–vis standard curve for purpurogallin assay was obtained using purpurogallin concentrations of 0.011, 0.017, 0.023, 0.028, 0.034, and 0.045 mM.
Synthesis of metal NPs
Au NPs were synthesized based on the Turkevich-Frens method with slight modifications60. Briefly, 0.05 g of HAuCl4·3H2O was dissolved in 50.00 mL of water, and the solution was heated until boiling under vigorous stirring. 5.6 mL of sodium citrate solution (1.0%, w/w) was added all at once. The solution was boiled for another 15 min.
Au nanostars were synthesized following a protocol reported in Ref. 61. The seed solution was prepared by adding 15.00 mL of sodium citrate solution (1.00%, w/w) to 100.00 mL of boiling HAuCl4 solution (1.00 mM) under vigorous stirring. After boiling for 15 min, the solution was cooled and filtered using a nitrocellulose membrane (0.22 µm). For the synthesis of nanostars, 100.00 µL of the seed solution was added to 10.00 mL of 0.25 mM HAuCl4 solution (with 10.00 µL of 1 M HCl) in a 20 mL glass vial at room temperature under moderate stirring (700 rpm). Quickly, 100.00 µL AgNO3 solution (3.00 mM) and 50.00 µL of ascorbic acid (AA, 100 mM) were added simultaneously. The solution was stirred for 30 s so that its colour quickly changed from light red to blue or greenish-black. Immediately afterward, centrifugation at 955 g in a 15 mL tube for 15 min was performed to terminate nucleation. Immediately after synthesis, 1.00 mL of PVP solution (0.10%, w/w, Mw: 40000) was introduced into the colloid and stirred for 5 min to form PVP-stabilized nanostars.
Au nanospheres were synthesized following a protocol reported in Ref. 62. 150.00 mL of 2.20 mM sodium citrate solution was put in a round-bottom flask (250 mL) equipped with a condenser and heated with a heating mantle for 15 min under vigorous stirring. After boiling had commenced, 1.00 mL of HAuCl4 solution (25.00 mM) was injected. The solution color changed from yellow to bluish gray and then to soft pink in 10 min. At this time point, the solution temperature was allowed to lower to 90 °C. Then, 1.00 mL of sodium citrate solution (60.00 mM) and 1.00 mL of HAuCl4 solution (25.00 mM) were sequentially injected (time delay 2 min). By repeating this addition for five times, Au nanospheres of about 30 nm were obtained.
Au nanorods were synthesized based on Ref. 63 with slight modifications. The seed solution was prepared by adding HAuCl4 aqueous solution (0.01 M, 0.25 mL) into cetyltrimethylammonium bromide aqueous solution (CTAB, 0.1 M, 10 mL) in a 15 mL plastic tube under gentle mixing. A freshly prepared, ice-cold NaBH4 solution (0.01 M, 0.60 mL) was then injected quickly into the mixture solution, resulting in a seed solution. The seed solution was kept at room temperature for at least 2 h before use. To grow Au nanorods, HAuCl4 aqueous solution (0.01 M, 2.00 mL), AgNO3 aqueous solution (0.01 M, 0.40 mL), and CTAB aqueous solution (0.10 M, 40.00 mL) were mixed in a 50 mL plastic tube. HCl solution (1.00 M, 0.80 mL) was then added to adjust pH to be 1–2, followed by the addition of AA solution (0.10 M, 0.32 mL). Finally, 20.00 μL of seed solution was injected. The solution was gently mixed for 10 s and left undisturbed at room temperature for at least 6 h.
Au nanoplatelets (NPLs) were synthesized following a protocol reported in Ref. 64. The seed solution was prepared by adding HAuCl4 aqueous solution (0.01 M, 1.00 mL) and sodium citrate aqueous solution (0.01 M, 1.00 mL) into 36.00 mL of water, followed by adding a freshly prepared, ice-cold NaBH4 solution (0.10 M, 1.00 mL) under vigorous stirring. After stirring for 2 min, the solution was aged for 2–6 h in order to the complete hydrolysis of unreacted NaBH4. Meanwhile, three growth solutions were prepared for the growth of Au NPLs. The growth solution (1) was made by mixing CTAB (0.05 M, 9.00 mL), HAuCl4 (0.01 M, 0.25 mL), NaOH (0.10 M, 0.05 mL), KI (0.01 M, 0.05 mL), and ascorbic acid (0.10 M, 0.05 mL) sequentially. The growth solution (2) was made the same to the growth solution (1). The growth solution (3) was made by mixing CTAB (0.05 M, 90.00 mL), HAuCl4 (0.01 M, 2.50 mL), NaOH (0.10 M, 0.50 mL), KI (0.01 M, 0.50 mL), and ascorbic acid (0.10 M, 0.50 mL). 1.00 mL of the seed solution was added to the growth solution (1), and the whole solution was gentle shaken for 5 s. 1.00 mL of the resultant solution was then added to the growth solution (2), and the whole solution was gentle shaken for 5 s. All of the resultant solution was then added to the growth solution (3). The obtained solution was gentle shaken and then left undisturbed at room temperature for 24 h. Triangular Au NPLs precipitated to the bottom of the reaction vessel. The supernatant was gently poured out and 40. 00 mL of water was added to redisperse the triangular Au NPLs.
Ag nanospheres were synthesized following a protocol reported in Ref. 65. (1) The Ag NP seeds were synthesized. 1.00 mL of sodium citrate aqueous solution (1.00%, w/w), 0.25 mL of AgNO3 aqueous solution (1.00%, w/w), and 0.20 mL of NaCl aqueous solution (20.00 mM) were successively added to 1.05 mL of water under stirring at room temperature. After 5 min, the citrate-silver-NaCl premixture was quickly added into 47.50 mL of the boiling water. Note that 80.00 μL of AA aqueous solution (0.10 M) has to be added into the boiling water 1 min before the addition of the citrate-silver-NaCl premixture. After heating and stirring for 1 h, the solution was cooled to room temperature. An Ag NP dispersion was finally obtained. (2) A solution of silver-ammonia complex was prepared by mixing 2.00 mL of AgNO3 aqueous solution (1.00%, w/w) and 0.80 mL of ammonia solution (25–28%, w/w). 0.20 mL of the original solution of Ag NP seeds was added into 4.73 mL of water under stirring in a 10 mL of glass vial at room temperature. Subsequently, 70.00 μL of silver-ammonia solution (43.00 mM) and 2.00 mL of AA aqueous solution (2.50 mM) were added into the Ag seed solution. After stirring for 1 h, Ag NPs were obtained.
Pd NPs were synthesized based on the method in Ref. 66. 0.165 g of anhydrous PdCl2 was dissolved in 20.00 mL of 1.00 M HCl solution and then diluted to 1 L with deionized water. The concentration is 9.30 × 10−4 M. 100.00 mL of the as-prepared PdCl2 solution was mixed with 200.00 mL of sodium citrate solution (1.00%, w/w). The mixed solution was diluted to 500.00 mL with water and then refluxed for 6 h.
General procedure for the preparation of Pickering emulsion and NP-loaded RPE
Pickering emulsions were prepared by using a homogenizer (ULTRA-IKA TURRAX T10) with a speed of 14500 rpm for 2 min. For O/W emulsions, the particles were dispersed in water before homogenization; for W/O emulsions, the particles were dispersed in oil before homogenization. For polydisperse NP-loaded RPE, 5.00 μL of Au NP dispersion (6.40 mg mL−1), 5.00 μL of TMAH solution (7.50 mM), and 0.40 mL of SiO2 ring dispersion in cyclohexane (2 mg mL−1) were mixed in a centrifuge tube. The mixture was shaken for 5 min to form emulsions. For other monodisperse NP-loaded RPE, 10.00 μL of NP dispersion (6.40 mg mL−1) and 0.40 mL of SiO2 ring dispersion in cyclohexane (2 mg mL−1) were mixed in a centrifuge tube. The mixture was shaken for 5 min to form emulsions. The sizes of the emulsion droplets were measured using an open-source software (ImageJ) based on optical microscopy images of the emulsions.
Stability evaluation of Pickering emulsion
(1) To evaluate the long-term storage stability, a water-in-toluene Pickering emulsion was employed. When investigating the effect of the dispersing water phase contents, the concentration of SiO2 rings were fixed to be 30 mg mL−1, and the volume ratios of toluene/water were changed from 0.2 to 0.7. When investigating the effect of the SiO2 ring concentrations, the volume ratios of toluene/water were fixed to be 2:3, and the concentrations of SiO2 rings were changed from 10 mg mL−1 to 80 mg mL−1. For each condition, 1 mL of emulsion were sealed in a 1.5 mL glass bottle. After storing for 6 h and one month, the emulsions were observed using optical microscopy. Before each sampling, the bottle was slightly shaken and then flipped up and down 5 times to ensure that the emulsion droplets of different particle sizes were evenly distributed in the bottle. (2) To evaluate the centrifugation stability, the water-in-toluene emulsions were used. The volume ratios of toluene to water were fixed to be 3:2 and the concentrations of SiO2 rings were 20 mg mL−1. 1.5 mL of Pickering emulsion was directly prepared in a 2 mL centrifugation tube and the centrifugation was performed with a given centrifugation force. The centrifugation time was fixed to 8 min. After centrifugation, the centrifugation tubes were slightly shaken. By this operation, the emulsion droplets were found to redisperse into the continuous phase again. The centrifugation tube was then flipped up and down 5 times to ensure that the emulsion droplets of different particle sizes were evenly distributed in the tube. Afterward, a random sample is taken for optical microcopy.
Calculation of the attachment energy of Pickering emulsifiers
For the nanospheres at oil-water interface, we calculated the energy (E) that was required to remove the particles into the oil phase using a reported Eq. (2)35:
where R is the radius of nanospheres and γow is the oil-water interfacial tensions. For the rings at oil-water interface, we assume the rings have a rectangular cross-section as shown in Fig. 2f (left) and the three-phase line situates at the intersection of the side and bottom surfaces. Therefore, we could use the following Eq. (3):
where do and di are the outer and inner diameter of the rings, respectively. The γow we use is 36 × 10−3 N m−1, and the temperature is 298.15 K.
Lipase-loaded RPE for hydrolysis of p-NPP
Based on the procedure above, 1.20 mL of RPE with a toluene/water (O/W) ratio of 1:2 was prepared. In it, the concentration of p-NPP in toluene is 100 mg mL−1 or 250 mg mL−1, and the ring concentration in water is 20 mg mL–1. 1.56 mL of PBS solution (10 mM, pH 7.4) was added. Afterward, 8.18 mL of aqueous dispersion of A. oryzae lipase was added (0.053 mg mL−1, dispersed in PBS buffer). The emulsion was slightly shaken for 1 min. The whole volume of the reactive system is 10. 94 mL. The hydrolysis reaction was performed at 30 °C under an undisturbed condition. At every given time point, 0.5 mL of the emulsion was sampled for analyzing the conversion. The sampled emulsion is quickly centrifuged at 1685 g for 2 min so that the emulsion droplets float on the top. After centrifugation, a clear emulsion layer and water layer can be observed. 100 μL of aqueous phase was sampled out from the water layer and further diluted to 10 mL (for p-NPP concentration: 100 mg mL−1) or 20 mL (for p-NPP concentration: 250 mg mL−1) with 0.5 M Na2CO3 solution. The absorbance of the diluted samples was measured by UV–vis spectroscopy and the peak at 400 nm was used to analyze the product concentration based on the standard curve. The specific activity (SA) of the lipase was the activity per unit amount of the lipase (U mg–1) in the reactions within the first 10 min. The activity (U) was defined as 1 μmol product produced within 1 min. To obtain the SA, all reactions were repeated at least three times.
Hemoglobin-loaded RPE for oxidation of pyrogallol
Based on the procedure above, 2.00 mL of RPE with a water/toluene (W/O) ratio of 2:3 was prepared. The water phase consists of 0.40 mL of citrate buffer solution (0.5 M, pH 3.0) containing pyrogallol (260 mM) and H2O2 (260 mM) and 0.40 mL of citrate buffer solution (0.5 M, pH 3.0) containing hemoglobin (0.158 mg mL−1). The toluene phase is 1.20 mL with a ring concentration of 20 mg mL–1. Another 7.00 mL of toluene was added to the emulsion so that the whole emulsion system was 9.00 mL. The oxidation reaction was performed at 30 °C under an undisturbed condition. At every given time point, 0.6 mL of the emulsion was sampled for analyzing the conversion. The sampled emulsion is quickly centrifuged at 1685 g for 2 min for the sedimentation of the emulsion droplets. After centrifugation, a clear toluene layer and emulsion layer can be observed. 0.45 mL of toluene phase was sampled out and further diluted to 1.35 mL with pure toluene. The absorbance of the diluted samples was measured by UV–vis spectroscopy and the peak at 425 nm was used to analyze the product concentration based on the standard curve. The SA was calculated based on the activity per unit amount of the hemoglobin (U mg–1) in a period of 4 h.
Pd NP-loaded RPE for interface oxidation of BA
Continuous flow reactions
7.5 mL of aqueous Pd NP dispersion (4.5 mg mL−1), and 10.0 mL of toluene containing 120 mg of SiO2 rings were used to produce a Pickering emulsion. The emulsion was gently filled into a glass column reactor (diameter: 1.34 cm) with a sand filter (pore diameter: 10–20 µm) at the bottom. The column reactor was put into an oven at 80 °C for reaction. A mixed solution of toluene/decane (4/1, v/v) containing BA (0.065 M), TBHP (0.5 M), and dodecane (0.065 M, internal standard for GC) was continuously added into the column reactor while maintaining a constant liquid height. The outflow from the column bottom was sampled for GC analysis at given intervals.
Batch reactions
3 mL of aqueous Pd NP dispersion (4.5 mg mL−1) and 3 mL of the mixed solvent of toluene and decane (4/1, v/v) containing SiO2 rings (60 mg) were used to produce Pickering emulsion. Afterward, 1.5 mL of the upper oil phase was removed and replaced by the same volume of the mixed toluene/decane (4/1, v/v) solution containing BA (0.13 M), TBHP (1 M) and dodecane (0.13 M, internal standard for GC). The batch reactions were performed at 80 °C. The same volume of emulsion was sampled at regular intervals and the conversion was monitored with GC.
For all these reactions, catalytic efficiency (CE) was defined as the moles of the converted reactants per mole of Pd per hour. For batch reactions, CE is calculated based on the conversion within 12 h. For the continuous flow reaction, CE is calculated based on the averaged conversion after 5 h.
Au NP-loaded RPE for SERS detection
RPE-based SERS detection was performed using a HORIBA confocal laser Raman Microscope (LabRAM HR Evolution) with an excitation wavelength of 633 nm (10× lens, ND Filter 100%, accumulation time: 20 s or 5 s). An appropriate amount of Pickering emulsion was poured into a hydrophobic polymer container, and the laser was focused on the interface of a Pickering emulsion droplet to obtain SERS signals. The reproducibility of the measurements was confirmed by at least five independent measurements and an average value of the five signal intensities was used.
(1) Water-soluble analytes. Typically, 2.50 µL of aqueous dispersion of Au NPs (3.2 mg mL−1), 2.50 µL of TMAH solution (5 mM), 5.00 µL of analyte aqueous solution, and 0.40 mL of SiO2 ring dispersion in cyclohexane (2 mg mL−1) were in a centrifuge tube (2 mL), and then the emulsion was prepared by hand shaking for 5 min. For the detection of small-volume analyte solution, the ratio of the components of the aqueous phase is the same and the volume of the continuous cyclohexane phase remains 0.4 mL. Note that the accumulation time in general is 20 s. For the measurements in Figs. 5B and S40–S41, the accumulation time of the detection was adjusted to be 5 s.
(2) Oil-soluble analytes. When oil-soluble analytes were analyzed, toluene was used instead of cyclohexane and the analytes were directly dissolved in toluene.
Characterization
Scanning electron microscopy (SEM) images were obtained using a field emission scanning electron microscope (FE-SEM, JEOL 6700 F) at an operating voltage of 5 kV. The samples were dispersed in ethanol and then dropped on a glass substrate. Before measurements, the samples were sputtered with a Pt layer (~ 10 nm) using a JFC-1600 auto fine coater with a current of 20 mA for 2 min.
Cryogenic-scanning electron microscopy (cryo-SEM) analysis was performed on a FE-SEM (Hitachi Regulus 8230) at an operating voltage of 5 kV. Liquid nitrogen was used in all heat exchange units of the cryogenic system. A small droplet of Pickering emulsion was placed on the sample holder, between two rivets, quickly frozen in liquid nitrogen for a few seconds, and transferred to the preparation chamber where it was fractured (at −140 °C). The revealed fractured surface was sublimed at −110 °C for 10 min to eliminate condensed ice. Before measurements, the samples were sputtered with tungsten with a current of 10 mA for 2 min. The temperature of the sample was kept at −140 °C during measurements.
Transmission electron microscopy (TEM) images were obtained by a transmission electron microscope (JEM-2200FS) at an accelerating voltage of 200 KV. To observe the packing structure of NPs at the interface, RPE with millimeter-scale droplets was prepared. Taking out some oil phase containing only several droplets to an open vessel, and evaporating the oil phase gradually until the brown film on the top of the droplets just exposes the air. The film was transferred to the carbon-coated copper grid for TEM.
Thermogravimetric analysis was performed on the equipment PerkinElmer Pyris 1. The measurements were performed under the N2 atmosphere. Before measurements, all the samples were dried at 100 °C for 36 h. The heating rate was 10 °C min−1. The samples were firstly heated from room temperature to 100 °C, and the temperature was held at 100 °C for 60 min, and then the samples were continuously heated to 800 °C.
Optical microscopy images were obtained on an Olympus IX73 microscope that was equipped with a CMOS camera.
Confocal microscopy was carried out on a laser confocal scanning microscope (Leica TCS SP8) with a 10× air lens and 63× oil lens. Imaging of FITC was done using a solid-state laser with 488 nm light, and the emission was detected in a range of 493–546 nm. Imaging of the RhB signal was done using a solid-state laser with 552 nm light, and the emission was detected in the range of 630–800 nm.
Ultraviolet-visible spectra (UV–vis) were measured on ultraviolet-visible spectrophotometer TU-1901. The wavelength range was from 250 nm to 800 nm at 1 nm increments.
Gas chromatography (GC) analysis was carried out on an Agilent GC system (7890B) equipped with a flame ionization detector (FID) on an HP-INNOWAX column (30 m × 320 μm × 0.5 μm). The starting temperature is 50 °C, and then the temperature increases to 240 °C at a rate of 20 °C min–1. The temperature was held at 240 °C for 10 min. The set temperatures of the injector and FID are 250 °C and 300 °C. The split ratio is set to 20:1.
Data availability
All data supporting this study and its findings are available within the article and Supplementary Information. Source data are provided with this paper. Data of this study are available from the corresponding author on request. Source data are provided with this paper.
References
Pickering, S. U. CXCVI.-Emulsions. J. Chem. Soc. T 91, 2001–2021 (1907).
Ramsden, W. Separation of solids in the surface-layers of solutions and ‘suspensions’ (observations on surface-membranes, bubbles, emulsions, and mechanical coagulation).—Preliminary account. Proc. R. Soc. Lond. 72, 156–164 (1997).
Li, K., Zou, H., Ettelaie, R., Zhang, J. & Yang, H. Spatial localization of two enzymes at pickering emulsion droplet interfaces for cascade reactions. Angew. Chem. Int. Ed. 62, e202300794 (2023).
Chen, G. et al. Regioselective surface assembly of mesoporous carbon on zeolites creating anisotropic wettability for biphasic interface catalysis. J. Am. Chem. Soc. 145, 9021–9028 (2023).
Sun, Z., Hubner, R., Li, J. & Wu, C. Artificially sporulated Escherichia coli cells as a robust cell factory for interfacial biocatalysis. Nat. Commun. 13, 3142 (2022).
Han, C. et al. Electrocatalytic hydrogenation of alkenes with Pd/carbon nanotubes at an oil–water interface. Nat Catal 5, 1110–1119 (2022).
Ni, L., Yu, C., Wei, Q., Liu, D. & Qiu, J. Pickering emulsion catalysis: interfacial chemistry, catalyst design, challenges, and perspectives. Angew. Chem. Int. Ed. 61, e202115885 (2022).
Chang, F., Vis, C. M., Ciptonugroho, W. & Bruijnincx, P. C. A. Recent developments in catalysis with Pickering Emulsions. Green Chem 23, 2575–2594 (2021).
Jiang, H., Sheng, Y. & Ngai, T. Pickering emulsions: versatility of colloidal particles and recent applications. Curr. Opin. Colloid Interface Sci. 49, 1–15 (2020).
Zhang, M. et al. Pickering emulsion droplets and solid microspheres acting synergistically for continuous-flow cascade reactions. Nat Catal 7, 295–306 (2024).
Phan-Quang, G. C., Lee, H. K., Phang, I. Y. & Ling, X. Y. Plasmonic colloidosomes as three-dimensional SERS platforms with enhanced surface area for multiphase sub-microliter toxin sensing. Angew. Chem. Int. Ed. 54, 9691–9695 (2015).
Lee, H. K. et al. Plasmonic liquid marbles: a miniature substrate-less SERS platform for quantitative and multiplex ultratrace molecular detection. Angew. Chem. Int. Ed. 53, 5054–5058 (2014).
Phan-Quang, G. C. et al. Online flowing colloidosomes for sequential multi-analyte high-throughput SERS analysis. Angew. Chem. Int. Ed. 56, 5565–5569 (2017).
Xie, Y. C. Z., Chen, M. Y., Liu, X. Y., Su, X. M. & Li, M. Artificial intelligence-assisted label-free spectroscopic quantification of global DNA cytosine methylation in a miniature plasmonic pickering emulsion. Adv. Funct. Mater. 33, 2307091 (2023).
Zhang, Y. et al. General approach to surface-accessible plasmonic Pickering emulsions for SERS sensing and interfacial catalysis. Nat. Commun. 14, 1392 (2023).
Ouyang, L., Li, D., Zhu, L., Yang, W. & Tang, H. A new plasmonic Pickering emulsion based SERS sensor for in situ reaction monitoring and kinetic study. J. Mater. Chem. C 4, 736–744 (2016).
Cecchini, M. P., Turek, V. A., Paget, J., Kornyshev, A. A. & Edel, J. B. Self-assembled nanoparticle arrays for multiphase trace analyte detection. Nat. Mater. 12, 165–171 (2013).
Cecchini, M. P. et al. Heavy metal sensing using self‐assembled nanoparticles at a liquid–liquid interface. Adv. Opt. Mater. 2, 966–977 (2014).
Klein, H. et al. Soft surface-enhanced Raman scattering sensing platform based on an oil-in-water emulsion stabilized by silver nanoparticles. Colloids Surf. B: Biointerfaces 245, 114278 (2025).
Aveyard, R., Binks, B. P. & Clint, J. H. Emulsions stabilised solely by colloidal particles. Adv. Colloid Interface Sci. 100–102, 503–546 (2003).
Yang, N. et al. Influence of stabilizers on the performance of Au/TiO2 catalysts for CO oxidation. ACS Catal 11, 11607–11615 (2021).
Xi, W., Shrestha, B. K. & Haes, A. J. Promoting intra- and intermolecular interactions in surface-enhanced raman scattering. Anal. Chem. 90, 128–143 (2018).
Ye, Z. et al. Surfactant-free synthesis of spiky hollow Ag-Au nanostars with chemically exposed surfaces for enhanced catalysis and single-particle SERS. JACS Au 2, 178–187 (2022).
Adkins, S. S., Gohil, D., Dickson, J. L., Webber, S. E. & Johnston, K. P. Water-in-carbon dioxide emulsions stabilized with hydrophobic silica particles. Phys. Chem. Chem. Phys. 9, 6333–6343 (2007).
Arditty, S., Whitby, C. P., Binks, B. P., Schmitt, V. & Leal-Calderon, F. Some general features of limited coalescence in solid-stabilized emulsions. Eur. Phys. J. E Soft Matter 11, 273–281 (2003).
Wang, H. & Hobbie, E. K. Amphiphobic carbon nanotubes as macroemulsion surfactants. Langmuir 19, 3091–3093 (2003).
Gautier, F. et al. Pickering emulsions with stimulable particles: from highly- to weakly-covered interfaces. Phys. Chem. Chem. Phys. 9, 6455–6462 (2007).
Kalashnikova, I., Bizot, H., Cathala, B. & Capron, I. New Pickering emulsions stabilized by bacterial cellulose nanocrystals. Langmuir 27, 7471–7479 (2011).
Binks, B. P., Philip, J. & Rodrigues, J. A. Inversion of silica-stabilized emulsions induced by particle concentration. Langmuir 21, 3296–3302 (2005).
Huang, S. D. et al. Regenerated silk fibroin for the stabilization of fish oil-loaded Pickering emulsions. Colloid Surface A 678, 132523 (2023).
Kempin, M. V., Kraume, M. & Drews, A. W/O Pickering emulsion preparation using a batch rotor-stator mixer - Influence on rheology, drop size distribution and filtration behavior. J. Colloid Interface Sci. 573, 135–149 (2020).
Kempin, M. V. & Drews, A. What governs Pickering emulsion properties during preparation via batch rotor‐stator homogenizers? Chem. Ing. Tech. 93, 311–317 (2020).
Ding, M. Z. et al. Effect of preparation factors and storage temperature on fish oil-loaded crosslinked gelatin nanoparticle pickering emulsions in liquid forms. Food Hydrocolloids 95, 326–335 (2019).
Xie, X., Wang, Y., Li, X., Wei, X. & Yang, S. Pickering emulsions stabilized by amphiphilic carbonaceous materials derived from wheat straw. Colloids Surf. A Physicochem. Eng. Asp. 558, 65–72 (2018).
Binks, B. P. & Lumsdon, S. O. Influence of particle wettability on the type and stability of surfactant-free emulsions. Langmuir 16, 8622–8631 (2000).
Li, Z. et al. Light-responsive, reversible emulsification and demulsification of oil-in-water Pickering emulsions for catalysis. Angew. Chem. Int. Ed. 60, 3928–3933 (2021).
Fapojuwo, D. P., Oseghale, C. O., Akinnawo, C. A. & Meijboom, R. Bimetallic PdM (M = Co, Ni) catalyzed hydrogenation of nitrobenzene at the water/oil interface in a Pickering emulsion. Colloids Surf. A Physicochem. Eng. Asp. 619, 126513 (2021).
Zhang, F. et al. In situ surface engineering of mesoporous silica generates interfacial activity and catalytic acceleration effect. ACS Omega 1, 930–938 (2016).
Yang, H., Zhou, T. & Zhang, W. A strategy for separating and recycling solid catalysts based on the pH-triggered Pickering-emulsion inversion. Angew. Chem. Int. Ed. 52, 7455–7459 (2013).
Wei, W. et al. Co-compartmentalization of enzymes and cofactors within Pickering emulsion droplets for continuous-flow catalysis. Angew. Chem. Int. Ed. 61, e202211912 (2022).
Liu, L., Jiang, L., Xie, X. & Yang, S. Amphiphilic carbonaceous microsphere-stabilized oil-in-water Pickering emulsions and their applications in enzyme catalysis. Chempluschem 81, 629–636 (2016).
Schmid, R. D. & Verger, R. Lipases: interfacial enzymes with attractive applications. Angew. Chem. Int. Ed. 37, 1608–1633 (1998).
Hanefeld, U., Gardossi, L. & Magner, E. Understanding enzyme immobilisation. Chem. Soc. Rev. 38, 453–468 (2009).
Verger, R. Interfacial activation’ of lipases: facts and artifacts. Trends Biotechnol 15, 32–38 (1997).
Wang, Q. et al. A supramolecular-hydrogel-encapsulated hemin as an artificial enzyme to mimic peroxidase. Angew. Chem. Int. Ed. 46, 4285–4289 (2007).
Li, F. et al. Hemoglobin: a new biocatalyst for the synthesis of 2‐substituted benzoxazoles via oxidative cyclization. ChemCatChem 11, 1192–1195 (2019).
Pan, Y. et al. Assembling two‐phase enzymatic cascade pathways in Pickering emulsion. ChemCatChem 11, 1878–1883 (2019).
Wang, M., Zhang, Z. & He, J. A SERS study on the assembly behavior of gold nanoparticles at the oil/water interface. Langmuir 31, 12911–12919 (2015).
Reincke, F., Hickey, S. G., Kegel, W. K. & Vanmaekelbergh, D. Spontaneous assembly of a monolayer of charged gold nanocrystals at the water/oil interface. Angew. Chem. Int. Ed. 43, 458–462 (2004).
Duan, H., Wang, D., Kurth, D. G. & Möhwald, H. Directing self-assembly of nanoparticles at water/oil interfaces. Angew. Chem. Int. Ed. 43, 5639–5642 (2004).
Andala, D. M., Shin, S. H., Lee, H. Y. & Bishop, K. J. Templated synthesis of amphiphilic nanoparticles at the liquid-liquid interface. ACS Nano 6, 1044–1050 (2012).
Zhang, M. et al. Compartmentalized droplets for continuous flow liquid liquid interface catalysis. J. Am. Chem. Soc. 138, 10173–10183 (2016).
Zhang, X. et al. Pickering Emulsion-derived liquid-solid hybrid catalyst for bridging homogeneous and heterogeneous catalysis. J. Am. Chem. Soc. 141, 5220–5230 (2019).
Shrivastava, A. & Gupta, V. Methods for the determination of limit of detection and limit of quantitation of the analytical methods. Chron. Young Sci. 2, 21–25 (2011).
Huang, Y. H. et al. Sensing antibiotics in wastewater using surface-enhanced raman scattering. Environ. Sci. Technol. 57, 4880–4891 (2023).
Su, X. et al. Quantitative label-free SERS detection of trace fentanyl in biofluids with a freestanding hydrophobic plasmonic paper biosensor. Anal. Chem. 95, 3821–3829 (2023).
Shi, S. et al. High-yield synthesis of colloidal carbon rings and their applications in self-standing electrodes of Li-O(2) batteries. J. Am. Chem. Soc. 145, 27664–27671 (2023).
Qiao, S., Li, S., Song, Q. & Liu, B. Shape-tunable biconcave disc-like polymer particles by swelling-induced phase separation of seeded particles with hydrophilic shells. Langmuir 39, 1190–1197 (2023).
Cao, W., Huang, R., Qi, W., Su, R. & He, Z. Self-assembly of amphiphilic janus particles into monolayer capsules for enhanced enzyme catalysis in organic media. ACS Appl. Mater. Interfaces 7, 465–473 (2015).
Frens, G. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nat. Phys. Sci. 241, 20–22 (1973).
Yuan, H. et al. Gold nanostars: surfactant-free synthesis, 3D modelling, and two-photon photoluminescence imaging. Nanotechnology 23, 075102 (2012).
Bastus, N. G., Comenge, J. & Puntes, V. Kinetically controlled seeded growth synthesis of citrate-stabilized gold nanoparticles of up to 200 nm: size focusing versus Ostwald ripening. Langmuir 27, 11098–11105 (2011).
Ni, W., Kou, X., Yang, Z. & Wang, J. Tailoring longitudinal surface plasmon wavelengths, scattering and absorption cross sections of gold nanorods. ACS Nano 2, 677–686 (2008).
Qin, F. et al. Thickness control produces gold nanoplates with their plasmon in the visible and near‐infrared regions. Adv. Opt. Mater. 4, 76–85 (2015).
Xing, L., Xiahou, Y., Zhang, P., Du, W. & Xia, H. Size control synthesis of monodisperse, quasi-spherical silver nanoparticles to realize surface-enhanced raman scattering uniformity and reproducibility. ACS Appl. Mater. Interfaces 11, 17637–17646 (2019).
Turkevich, J. & Kim, G. Palladium: preparation and catalytic properties of particles of uniform size. Science 169, 873–879 (1970).
Acknowledgements
This work was financially supported by the National Key R&D Program of China (Grant No. 2021YFA0716700 to Y.W. and B.L.), the National Natural Science Foundation of China (Grant No. 22172175 to B.L.), and the National Science Foundation of China (Grant No. 21988102 to Y.W.). The authors would like to thank Prof. Jiajia Zhou at South China University of Technology for the discussion of the calculation of the desorption energy.
Author information
Authors and Affiliations
Contributions
B.L. designed the research. M.L. performed all the experiments except the synthesis of nanoparticles, which was performed by Q.S. All authors analyzed and discussed the data. M.L., Y.W., and B.L. wrote the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Ming Li and the other, anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Li, M., Song, Q., Wang, Y. et al. Pickering emulsions with low interface coverage but enhanced stability for emulsion interface catalysis and SERS-based detection. Nat Commun 16, 2490 (2025). https://doi.org/10.1038/s41467-025-57914-3
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-025-57914-3
This article is cited by
-
Rings make more room
Nature Reviews Chemistry (2025)
