Nanostructures, Faceting, and Splitting in Nanoliter to Yoctoliter Liquid

Apr 15, 2019 - Contrary to everyday experience, where all liquid droplets assume rounded, near-spherical shapes, the temperature-tuning of liquid drop...
0 downloads 0 Views 10MB Size
Download PDF
Letter Cite This: Nano Lett. XXXX, XXX, XXX−XXX

pubs.acs.org/NanoLett

Nanostructures, Faceting, and Splitting in Nanoliter to Yoctoliter Liquid Droplets Shani Guttman,†,∥ Ellina Kesselman,‡,∥ Avi Jacob,§ Orlando Marin,† Dganit Danino,‡ Moshe Deutsch,† and Eli Sloutskin*,† †

Physics Department and Institute of Nanotechnology and Advanced Materials, Bar-Ilan University, Ramat Gan 529002, Israel Russell Berrie Nanotechnology Institute (RBNI) and Department of Biotechnology and Food Engineering, Technion, Israel Institute of Technology, 32000 Haifa, Israel § The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat Gan 52900, Israel Nano Lett. Downloaded from pubs.acs.org by UNIV OF LIVERPOOL on 04/22/19. For personal use only.

‡

S Supporting Information *

ABSTRACT: Contrary to everyday experience, where all liquid droplets assume rounded, near-spherical shapes, the temperature-tuning of liquid droplets to faceted polyhedral shapes and to spontaneous splitting has been recently demonstrated in oil-in-water emulsions. However, the elucidation of the mechanism driving these surprising effects, as well as their many potential applications, ranging from faceted nanoparticle synthesis through new industrial emulsification routes to controlled-release drug delivery within the human body, have been severely hampered by the micron-scale resolution of the light microscopy employed to date in all in situ studies. Thus, the thickness of the interfacially frozen crystalline monolayer, suggested to drive these effects, could not be directly measured, and the low limit on the droplet size still showing these effects remained unknown. In this study, we employ a combination of super-resolution stimulated emission depletion microscopy, cryogenic transmission and freeze-fracture electron microscopy, to study these effects well into the nanometer length scale. We demonstrate the occurrence of the faceting transition in droplets spanning an incredible 12 decades in volume from nanoliters to yoctoliters and directly visualize the interfacially frozen, few nanometer thick, crystalline monolayer suggested to drive these effects. Furthermore, our measurements allow placing an upper-limit estimate on the twodimensional Young modulus of the interfacial nanometer-thick surface crystal in the smallest droplets, providing insights into the virtually unexplored domain of nanoelasticity. KEYWORDS: Faceting, emulsion, cryo-electron microscopy, STED, interfacial freezing

L

trolled-release drug delivery in the human body,12,13 possibly through the temperature-tunable self-emulsification, and for new industrial-scale emulsification protocols.9 These effects also potentially provide a very simple experimental model of biosystems’ morphogenesis and protocell division.14 Further progress in elucidating the underlying mechanism of faceting and in developing its nanoscale applications has been severely limited by the fact that only conventional light microscopy has been used in all studies to date. For example, the droplet faceting has been attributed to either the formation of a 2 nm thick crystalline monolayer at the droplet’s surface and the domination of its elasticity over the vanishingly small interfacial tension of the droplet1−3 or to the formation of a 300 nm thick, 1 μm radius, tubular rotator crystal phase at the surface.4 Both suggested mechanisms also predict (different) low limits on the size of droplets that undergo faceting: the

iquid droplets, unperturbed by external forces, are spherical, since for a given volume this shape minimizes their surface area A and hence their surface energy. Yet, recent studies have demonstrated that water-dispersed, surfactantstabilized, emulsion droplets of many common fully saturated linear oils can be temperature-tuned to undergo at some T = Td a spontaneous transition to specific faceted polyhedral shapes while retaining their liquidity.1−4 At a slightly lower temperature, TSE < Td, the faceted droplets abruptly distort and undergo multiple spontaneous splitting transitions. These splitting events are reminiscent of the classical spontaneous emulsification, leading to the formation of microemulsions.5−8 However, both the sharply defined facets of the droplets and the precise temperature controllability of the observed transitions, not existing in conventional spontaneous emulsification, suggest that the two phenomena are driven by completely different physical mechanisms. The nature of the mechanism driving these effects in the present system is still under debate.1,9 Nevertheless, these effects clearly open new routes for synthesizing faceted nanoparticles10,11 for con© XXXX American Chemical Society

Received: February 10, 2019 Revised: April 8, 2019 Published: April 15, 2019 A

DOI: 10.1021/acs.nanolett.9b00594 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

being very effective in stabilizing droplets’ positions, glycerol does not interfere with the faceting transitions: the light microscopy resolvable droplets are observed to transform on cooling into the same geometrical shapes with and without the presence of glycerol in the aqueous phase. While at T > Td all the droplets are spherical, they adopt an icosahedral shape at T = Td. On further cooling, the icosahedra undergo an additional shape transition, forming platelet-like hexagonal, parallelogram, triangular, and rodlike shapes. In this low-temperature regime, many of the droplets develop tail-like protrusions, emanating from their vertices.1−4 Initially, the tails are sufficiently thick to be clearly resolved by conventional confocal microscopy. However, the tails’ lengths increase very quickly while their widths decrease commensurably fast and become unresolvable by conventional light microscopy. Therefore, it was hitherto unknown whether the radii of the tail-like protrusions can become smaller than4 ∼ 1 μm. In Figure 1a,b, we show STED images of two very thin and long tails. While thicker tails are easier to detect, we focus here

former because of the nanoscale applicability of continuum elastic theory1 and the latter because of the rotator phase’s thickness.4 As the sizes above are all well below classical Rayleigh-diffraction-limited light microscopy resolution, they could be inferred only indirectly1−4 rather than verified directly by measurements. The study presented here is a first experimental foray into the nanoscale structure of these faceted droplets, demonstrating the occurrence of droplet faceting down to subzeptoliter (>10−21 L) droplet volumes. We use several complementary experimental techniques to visualize the droplets’ size, shapes, and features in the nanometer regime and provide conclusive direct observations resolving some of the issues discussed above. Specifically, we show the existence of these polyhedral droplets in sizes extending down to ∼10 nm, implying the possibility of obtaining such solid particles by UV-polymerization. Moreover, the thickness of the interfacially frozen crystal, t, is measured to be a few nanometers, supporting the surface-frozen-monolayer-driven faceting mechanism discussed above1−3 but 2 orders of magnitude lower than the theoretical estimates appearing in some recent publications.4,15 Also, the radius of curvature at the faceted droplets’ edges, estimated in those studies4,15 to be of order ∼1 μm, is found here to be only a few nanometers. The very low t of the interfacially frozen crystals implies that their structure, and consequently their elastic properties, are inherently anisotropic. It is widely recognized that the elasticity of such quasi-two-dimensional (2D) crystals and their assemblies is fundamentally different from these of the 3D material,1 potentially allowing the bending and the stretching moduli to be controlled independently.16 Yet, because experimental elasticity determinations of quasi-2D crystals are very rare, even the most fundamental concepts of the nanoscale elasticity are not fully established, not allowing the great promise of such systems to be fully realized. Our measured t values and nanoscale droplet geometries provide a unique insight into the elastic properties of the interfacial crystals. The physical understanding of droplet faceting and splitting thus achieved should allow this process to be controlled and applied in new preparation strategies of faceted nanoparticles and emulsions of a high demand in nanotechnology. Results and Discussion. Elongated Tail Imaging by STED. To observe the hitherto unresolved nanoscale structural features of the water-dispersed droplets in situ and under fully controlled temperature conditions, we employ optical stimulated emission depletion (STED) microscopy,17 one of the least-invasive super-resolution light imaging methods. Following previous droplet faceting studies,3,15 we employ C16 [hexadecane, CH3(CH2)14CH3] oil droplets, stabilized in an aqueous medium by the C18TAB [trimethyloctadecylammonium bromide, CH3(CH2)17N(CH3)3Br] surfactant. For this combination of alkane and surfactant, the faceting transition occurs very close to room temperature3 at Td ≈ 20 °C, providing for a convenient control of the temperature and minimizing possible temperature gradients. For fluorescent imaging, the droplets are stained by the Bodipy 505/515 hydrophobic dye, which does not leak into the aqueous continuum and does not interfere with the droplet faceting phenomena.1 The high resolution of STED microscopy and its relatively slow scanning rate require minimizing fluctuations in the position of the solution-suspended emulsion droplets during imaging. To this end, we employ a 70% mixture of glycerol in water as our aqueous medium. Remarkably, while

Figure 1. (a,b) STED images of two pairs of very thin tail-like liquid droplet protrusions demonstrate their thickness to be far smaller than the classical light microscopy resolution limit. A thicker protrusion (slightly out-of-focus and hence less bright) is visible as well roughly parallel to the image diagonal in (b). (c) The average intensity profile (symbols), transverse to the rightmost tail in (b), has a Lorentzian shape (see solid red line fit) with a HWHM of 30 ± 5 nm. Since all previous measurements of such shape-transforming droplets have been limited by the conventional light microscopy resolution, the existence of such thin tail-like protrusions remained unnoticed, leading to incorrect assumptions in some theoretical estimates (see text).

on the thinnest tails which are still clearly visible by STED. Intensity profiling across the tails yields a profile well-matched by a Lorentzian function (Figure 1c) with a HWHM (half width at half-maximum) of 30 ± 5 nm. Note that the experimental intensity profile is a convolution of the actual fluorescence signal with the instrumental resolution function. In STED systems in general, this function is Lorentzian18 and has in our system a HWHM of a few tens of nanometers. Thus, the experimental HWHM of the tail profile is the upper limit of the actual radii of these tails (see additional STED tail profiles in the Supporting Information, SI). Note that previous estimates4 suggested the tails’ radii to be ∼1000 nm, well above the currently established upper limit of 30 nm. Correspondingly, the calculated interfacial crystals’ thickness B

DOI: 10.1021/acs.nanolett.9b00594 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters obtained in that study,4 t ≈ 300 nm, is far too large for the crystals to fit into the tails. Finally, STED intensity profiles across thicker tails are found to be structureless, suggesting that the interfacial crystals, in either the thin or the thick protrusions, are far thinner than estimated in the aforementioned studies.4 While STED microscopy allows the shape transformations of these droplets to be followed in situ during the temperature scans, the measurements are limited by the STED’s resolution. Moreover, STED relies on the fluorescence from the object imaged. The fluorescence intensity scales with the droplet’s volume. Thus, while very long slowly diffusing tails are optimal for STED measurements, the imaging of small polyhedral droplets with this technique is quite challenging because of the weak signal from the small droplets. This, and the increased rotational and translational diffusion of the small droplets do not allow submicron liquid polyhedra to be imaged with our present STED setup. For imaging such droplets, we employed, therefore, cryo-electron microscopy. These measurements are detailed in the following subsection. Polyhedra Imaging by CryoTEM. For the cryo-transmission electron microscopy (cryoTEM) measurements, the sample is first kept at T0, a few degrees below Td, to ensure that all the droplets have undergone the faceting transition. The sample is then vitrified by plunging into liquid-nitrogen-cooled metastable ethane,19 ∼78 K, and kept throughout the measurements at that temperature. This procedure fixes and preserves the faceted droplet, allowing the measurement of its shape and features by transmission electron microscopy20 (see Experimental Methods). Control cryoTEM measurements of pure surfactant solutions without oil droplets show no faceted objects (see Figure S2 in the SI). All micrometer size droplets observed in our CryoTEM experiments are faceted (Figure 2a,b). As they are too large to fit into the 100−200 nm-thick CryoTEM sample (see Experimental Methods), they appear flattened. Yet, their faceting is very clear, and their shape closely matches the faceted liquid droplet shapes observed by light microscopy.1,4 Note the clear contrast between the oil inside the droplet and the surfactant solution outside of the droplet in the typical CryoTEM image of a faceted droplet shown in Figure 2a,b. Remarkably, the internal part of all droplets looks completely homogeneous. In some of the recent studies of self-shaping emulsion droplets, it has been proposed that the spontaneous faceting of the droplets is driven by tubular rotator crystals, forming along the faceted droplet’s edges.4,9,15 The tubules’ wall thickness has been estimated4 as ∼300 nm and their radii as ∼1000 nm. No sign of such tubules or of other rotator crystal formations is observed in Figure 2a,b at the droplet’s interface. The only feature observed there is a clearly defined interfacial layer, which is only a few nanometers thick. To obtain the thickness of the interfacial layer, we measured the intensity profiles across the interfaces of the faceted droplets (Figure 2c). The width of the dip in this plot, averaging 6 ± 3 nm, yields the thickness of the interfacial layer. This, however, is a value inflated by contributions from several artifacts, for example, the TEM’s basic resolution and the subtleties of the electron optics, such as the depth of focus, a possible tilt of the droplets’ interfaces with respect to the optical axis, and a possible slight defocusing. Thus, a better way to estimate the real interfacial layer thickness is to measure the apparent thickness of the interface as derived from images collected at different magnification factors, M. We note that

Figure 2. (a,b) CryoTEM images of the edges of a vitrified faceted droplet. No internal structure is visible inside the faceted liquid droplets, and the interfacially frozen layer (marked by an arrow) that is only a few nm thick is clearly visible. The large dark vertical band on the left side of panel a is the carbon grid supporting the droplet. The small objects throughout the images are surfactant micelles and/or nanodroplets, residing in the aqueous continuum outside of the large alkane droplet. A zoomed-out image of the same droplet is shown in Figure S3b of the SI, together with other similar droplets. (c) The intensity profile across the interface of the droplet in panel b (symbols). The solid red curve is a fit to a negative Gaussian function, residing on a tilted linear background. Measurements at different magnification factors M yield thickness values inversely proportional to M, as shown in the inset. Linear extrapolation to M−1 = 0 (solid red line) yields the layer thickness as t = 2.2 ± 0.9 nm, consistent with the thickness of a single interfacially frozen monolayer of fully extended, mixed, C16 and C18TAB molecules, oriented normal to the interface. The t(M) values were obtained by averaging over ∼20 droplets at each M. The error bars are the corresponding standard deviations.

the thickness variation at a given M is far smaller than the Mdependence shown in the inset of Figure 2c. In particular, we do not observe any significant dependence on the drop size. Within the admittedly significant experimental scatter, the interfacial thickness seems to scale linearly with M−1 (inset of Figure 2c) and extrapolate to 2.2 ± 0.9 nm at the M−1 → 0 limit. While the linear extrapolation to M−1 → 0 is purely phenomenological, the obtained value agrees within the combined statistical error with t = 3.5 ± 1 nm, which we obtain for our highest-magnification images (×30 000). Note also that while the statistical error of these measurements is 0.9−1 nm, the experimental accuracy of this method is limited, likely reaching 1−2 nm. Therefore, within the combined errors, the obtained value of 2.2−3.5 nm is in a very good agreement with the values obtained by X-ray reflectivity for planar liquid−liquid and liquid−vapor interfaces of similar molecules,21,22 where a monolayer of fully extended cocrystallized alkane and surfactant molecules has been detected. In particular, the X-ray measured thickness of the monolayer with the surfactants’ headgroups and terminal CH3 groups included was reported as 2.7 ± 0.1 nm for the cocrystallized C16TAB and C14 molecules. These molecules are identical to the ones studied here except that here both alkanes and surfactants have alkyl tails longer by two methylenes. This adds 0.127 × 2 = 0.254 nm to the total surface monolayer thickness.22 Thus, a thickness of ∼3.0 ± 0.1 nm obtained from this calculation for our present C16/C18TAB system agrees well within the C

DOI: 10.1021/acs.nanolett.9b00594 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters combined error with the value of 2.2 ± 0.9 nm derived from the present M-dependent CryoTEM measurements. Remarkably, while the freezing of liquid−vapor and liquid− liquid interfaces has been widely studied for a few decades by X-rays, nonlinear optics, and thermodynamical methods, the interfacially frozen layers have never been directly visualized on a nanometric scale23 as in Figure 2a,b. The frozen liquid− liquid interface CryoTEM imaging capabilities demonstrated here may prove useful in future studies of interfacial ordering phenomena. The CryoTEM is particularly advantageous for curved buried interfaces, such as those of the droplets studied here, because these interfaces are practically inaccessible by conventional surface-specific X-ray techniques. In the absence of any direct experimental methods for measuring the interfacial thickness, the attribution of the faceting transitions to the elastic properties of the interfacial monolayer remained inconclusive and controversial.3,15 Our present CryoTEM measurements demonstrate that the interfacial thickness is consistent with the formation of a single monolayer, composed of fully extended, cocrystallized, C16 and C18TAB molecules.1−3 Clearly, this result also agrees with our STED measurements, which cap the crystalline surface layer thickness at 20 min. To create thin films on the grid for CryoTEM, excess solution was removed by manual blotting with a filter paper from the uncoated side of the grid. In independent tomography experiments, the thickness of the liquid films was measured to be ∼100−200 nm. The carbon grid was then cooled from room temperature to T0 ≈ 16.5 ± 1 °C and equilibrated for ∼5 min at this temperature. The blotted grid was then plunged into liquid ethane, cooled to its melting point of −183 °C with liquid nitrogen (LN2), and the formed vitrified specimen was transferred to LN2 and stored there until the measurement. The vitrified specimen was loaded onto a Gatan 626 cryo holder, cooled with liquid nitrogen to below −170 °C, and examined in a Tecnai T12 G2 TEM (FEI) at 120 kV. Images were recorded at low-dose conditions on a US1000 2k × 2k cooled CCD camera (Gatan) with the Digital Micrograph software using imaging procedures developed in-house, as previously described.20 Freeze-Fracture CryoTEM. Freeze-fracture replicas were prepared in a BAF-060 system (Leica, Germany), as previously described.32 An ∼8 μL drop of each emulsion was placed on a bare copper TEM grid and sandwiched between two gold planchettes. The “sandwich” was equilibrated in the CEVS at T ≈ 16.5 ± 1 °C, then plunged into liquid ethane, transferred to liquid nitrogen, and inserted into a LN2 precooled sample fracture block. The block was then inserted into the BAF chamber, maintained at −160 °C to −170 °C under vacuum, and upon opening two fractured surfaces were exposed.32 Those were first etched at −110 °C for 30 s and then shadowed at a 45° angle with a 2 nm layer of platinum− carbon, followed by a 20 nm thick carbon layer. The samples were thawed to remove the organic material, and the replicas were mounted on TEM grids, dried, and examined in a Tecnai T12 G2 TEM (FEI) at 120 kV at room temperature.32 SEM of UV-Polymerized Micron-Scale Droplets. For imaging of the UV-polymerized droplets (Figure 5e), we used as the oil phase hexadecyl acrylate [HA, H2C CHCO2(CH2)15CH3] instead of the C16 alkane. The droplets also included a small amount (∼1% w/v) of diphenyl(2,4,6trimethylbenzoyl)phosphine oxide, to photoinitiate the polymerization of the acrylate by UV, after cooling to T < TSE. Both C16 and HA have identical-length fully saturated alkyl tails. This tail enables their interfacial cocrystallization with the C18TAB surfactant,10 which induces the faceting of the droplets. The continuous phase is an aqueous ∼1 mM C18TAB solution. Once polymerized, the particles are removed from their native suspending medium, redispersed in ethanol, deposited on an aluminum stub for SEM measurements, and coated by Au evaporation. The coating layer thickness is ∼2.5

■

ASSOCIATED CONTENT

S Supporting Information *

This material is available free of charge via the Internet at http://pubs.acs.org/. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.9b00594.

■

Additional figures and information; video captions (PDF) Bright-field microscopy video demonstrating the splitting of a triangular platelet-like droplet in slow motion at T < TSE (AVI) Simulated spontaneous splitting of a triangular plateletlike droplet (AVI)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Moshe Deutsch: 0000-0002-1061-2276 Eli Sloutskin: 0000-0002-7109-6893 Author Contributions ∥

S.G. and E. K. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research is supported by Israel Science Foundation Grant 1779/17. We thank Alexander V. Butenko (Bar-Ilan University) and Mingming Zhang (Technion - Israel Institute of Technology) for technical assistance. We also thank Luca Giomi and Ireth Garcia-Aguilar (Leiden University) for discussions.

■

REFERENCES

(1) Guttman, S.; Sapir, Z.; Schultz, M.; Butenko, A. V.; Ocko, B. M.; Deutsch, M.; Sloutskin, E. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 493−496. (2) Guttman, S.; Ocko, B. M.; Deutsch, M.; Sloutskin, E. Curr. Opin. Colloid Interface Sci. 2016, 22, 35−40. (3) Guttman, S.; Sapir, Z.; Ocko, B. M.; Deutsch, M.; Sloutskin, E. Langmuir 2017, 33, 1305−1314. (4) Denkov, N.; Tcholakova, S.; Lesov, I.; Cholakova, D.; Smoukov, S. K. Nature 2015, 528, 392−395. (5) Solans, C.; Morales, D.; Homs, M. Curr. Opin. Colloid Interface Sci. 2016, 22, 88−93. (6) Zhang, H.; Zhu, Y.; Qu, L.; Wu, H.; Kong, H.; Yang, Z.; Chen, D.; Mäkilä, E.; Salonen, J.; Santos, H. A.; Hai, M.; Weitz, D. A. Nano Lett. 2018, 18, 1448−1453. (7) Eastoe, J.; Hetherington, K. J.; Dalton, J. S.; Sharpe, D.; Lu, J. R.; Heenan, R. K. J. Colloid Interface Sci. 1997, 190, 449−455. (8) de Molina, P. M.; Zhang, M.; Bayles, A. V.; Hegleson, M. E. Nano Lett. 2016, 16, 7325−7332. (9) Tcholakova, S.; Valkova, Z.; Cholakova, D.; Vinarov, Z.; Lesov, I.; Denkov, N.; Smoukov, S. K. Nat. Commun. 2017, 8, 15012. (10) Marin, O.; Alesker, M.; Guttman, S.; Gershinsky, G.; Edri, E.; Shpaisman, H.; Guerra, R. E.; Zitoun, D.; Deutsch, M.; Sloutskin, E. J. Colloid Interface Sci. 2019, 538, 541−545. (11) Ruan, J.; Kjellman, T.; Sakamoto, Y.; Alfredsson, V. Langmuir 2012, 28, 11567−11574. G

DOI: 10.1021/acs.nanolett.9b00594 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters (12) Tibbitt, M. W.; Dahlman, J. E.; Langer, R. J. Am. Chem. Soc. 2016, 138, 704−717. (13) Park, K. J. Controlled Release 2014, 190, 3−8. (14) Gordon, R.; Hanczyc, M. M.; Denkov, N. D.; Tiffany, M. A.; Smoukov, S. K. Emergence of polygonal shapes in oil droplets and living cells: the potential role of tensegrity in the origin of life. In Habitability of the Universe Before Earth; Gordon, R., Sharov, A., Eds.; Academic Press, 2018; pp 427490. (15) Denkov, N.; Cholakova, D.; Tcholakova, S.; Smoukov, S. K. Langmuir 2016, 32, 7985−7991. (16) Androulidakis, C.; Zhang, K.; Robertson, M.; Tawfick, S. 2D Mater. 2018, 5, 032005. (17) Vicidomini, G.; Bianchini, P.; Diaspro, A. Nat. Methods 2018, 15, 173−182. (18) Merino, D.; Mallabiabarrena, A.; Andilla, J.; Artigas, D.; Zimmermann, T.; Loza-Alvarez, P. Biomed. Opt. Express 2017, 8, 2472−2482. (19) Dubochet, J.; McDowall, A. W. J. Microsc. 1981, 124, 3−4. (20) Danino, D. Curr. Opin. Colloid Interface Sci. 2012, 17, 316−329. (21) Tamam, L.; Pontoni, D.; Sapir, Z.; Yefet, S.; Sloutskin, E.; Ocko, B. M.; Reichert, H.; Deutsch, M. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 5522−5525. (22) Sloutskin, E.; Sapir, Z.; Bain, C. D.; Lei, Q.; Wilkinson, K. M.; Tamam, L.; Deutsch, M.; Ocko, B. M. Phys. Rev. Lett. 2007, 99, 136102. (23) Jai, C.; Aimé, J. P.; Mariolle, D.; Boisgard, R.; Bertin, F. Nano Lett. 2006, 6, 2554−2560. (24) Garcia-Aguilar, I.; Fonda, P.; Giomi, L.; Sloutskin, E. Bull. Am. Phys. Soc. 2019. (25) Dubois, M.; Deme, B.; Gulik-Krzywicki, T.; Dedieu, J.; Vautrin, C.; Desert, S.; Perez, E.; Zemb, T. Nature 2001, 411, 672−675. (26) Hunter, R. J. Foundations of Colloid Science; Oxford University Press, 2009. (27) Epstein, B. J. Franklin Inst. 1947, 244, 471−477. (28) Gibaud, T.; Barry, E.; Zakhary, M. J.; Henglin, M.; Ward, A.; Yang, Y.; Berciu, C.; Olden-Bourg, R.; Hagan, M. F.; Nicastro, D.; Meyer, R. B.; Dogic, Z. Nature 2012, 481, 348−351. (29) Tu, F.; Park, B. J.; Lee, D. Langmuir 2013, 29, 12679−12687. (30) Würger, A. Phys. Rev. Lett. 2000, 85, 337−340. (31) Datta, S. S.; Kim, S.-H.; Paulose, J.; Abbaspourrad, A.; Nelson, D. R.; Weitz, D. A. Phys. Rev. Lett. 2012, 109, 134302. (32) Danino, D.; Kesselman, E.; Saper, G.; Petrache, H. I.; Harries, D. Biophys. J. 2009, 96, L43−L45.

H

DOI: 10.1021/acs.nanolett.9b00594 Nano Lett. XXXX, XXX, XXX−XXX