doi.org/10.26434/chemrxiv.12106773.v2

Self-Assembly of Charged Oligomeric Bilayers at the Aqueous/OrganicInterface Regulated by Ion-Pair AssociationsGraham J. Taylor, Yingdong Luo, Kunlun Hong, Stephen A. Sarles, Robert L. Sacci, Benjamin Doughty, C.Patrick Collier

Submitted date: 14/04/2020 • Posted date: 16/04/2020Licence: CC BY-NC-ND 4.0Citation information: Taylor, Graham J.; Luo, Yingdong; Hong, Kunlun; Sarles, Stephen A.; Sacci, Robert L.;Doughty, Benjamin; et al. (2020): Self-Assembly of Charged Oligomeric Bilayers at the Aqueous/OrganicInterface Regulated by Ion-Pair Associations. ChemRxiv. Preprint.https://doi.org/10.26434/chemrxiv.12106773.v2

Phospholipid bilayer membranes show promise as biomolecular soft materials that mimic the ability of livingsystems to sense, respond and learn but are fragile. Amphiphilic charged oligomers(oligodimethylsiloxane-methylimidazolium cation, ODMS-MIM(+)), assembled into bilayers at the oil-aqueousinterfaces of droplet interface bilayers (DIBs), possessed similar size and functionality as phospholipidbilayers, but were stable. The ionic liquid headgroups (MIM(+)) of the oligomers were covalently bound toshort-chain hydrophobic tails (ODMS). Bilayer self-assembly was influenced both by the charged headgroups,constrained to two-dimensional diffusion at the liquid-liquid interface, which formed electric double layers inthe aqueous phase, and the tails in the organic phase. Bilayers formed spontaneously at low ionic strength butrequired an external voltage to form at higher ionicities. This switch in assembly behavior was due toion-pairing of the MIM(+) headgroups with chloride ions, resulting in an increase in the density of the chargedheadgroups at the interface and the ODMS hydrophobic tails in the oil phase as they were covalently graftedto the headgroups. Chain overlap led to repulsive disjoining pressures between droplets due to osmoticstress. The applied voltage caused an attractive electrocompressive stress that overcame the repulsion,enabling bilayer formation. Bilayer assembly at high ionic strength, while requiring a voltage to initiate, wasirreversible, and the resulting membrane was considerably more stable than those formed at lower values ofthe ionic strength. This switching of assembly behavior can be exploited as an additional mechanism forshort-term synaptic plasticity in neuromorphic device applications using soft materials.

File list (2)

download fileview on ChemRxivMIM-ODMS Collier Supporting Information.docx (58.65 KiB)

download fileview on ChemRxivMIM-ODMS oligomer bilayer membranes Collier_final.docx (2.72 MiB)

Supporting Information.

Values of the separation interval, D, needed for the estimation of the disjoining pressures in the oligomer bilayers, were determined with the generalized reduced gradient nonlinear optimization program “Solver” found in Microsoft Excel. The program minimized the differences between the membrane surface charge densities, σ < 10-9, from equation 6 and equation 7 of the manuscript by

varying the membrane surface potentials, ψ0, as functions of the ionic strength in the aqueous phase and oligomer concentration in the organic phase. These values are listed in Table S1 (c = 2.0 mg/mL) and Table S2 (c = 0.2 mg/mL) and displayed in Figure 4 of the manuscript. The other parameters are described in the main text.

Table S1.

c = 2.0 mg/mL

Kd = 0.275 M

σ0 = 4.00E-03 (C/m2)

D0 = 7.50 nm

I (M) 0.1 0.15 0.2 0.25 0.3 0.4 0.6 0.7 1.0ψ0

(mV)

12.53 9.19 7.23 5.93 4.99 3.75 2.42 2.03 1.32

σ (C/m2) 9.11E-03 8.14E-03 7.38E-03 6.76E-03 6.24E-03 5.40E-03 4.27E-03 3.86E-03 3.00E-03

1-α 0.37 0.44 0.49 0.53 0.57 0.63 0.71 0.73 0.79

D (nm) 6.15 5.66 5.35 5.13 4.97 4.73 4.46 4.38 4.21

1/κ (nm) 0.91 0.74 0.64 0.57 0.52 0.45 0.37 0.34 0.29

Table S2.

c = 0.2 mg/mL

Kd = 0.275 M

σ0 = 1.00E-03 (C/m2)

D0 = 16.15 nm

I (M) 0.1 0.15 0.2 0.25 0.3 0.4 0.6 0.7 1.0ψ0

(mV)

3.08 2.22 1.73 1.40 1.17 0.87 0.55 0.46 0.29

σ (C/m2) 2.22E-03 1.96E-03 1.76E-03 1.59E-03 1.46E-03 1.25E-03 9.67E-04 8.70E-04 6.68E-04

1-α 0.29 0.37 0.44 0.49 0.53 0.60 0.69 0.72 0.79

D (nm) 14.98 13.23 12.21 11.54 11.06 10.42 9.72 9.51 9.11

1/κ (nm) 0.91 0.74 0.64 0.57 0.52 0.45 0.37 0.34 0.29

The values of the Debye screening lengths, 1/κ (equation 2), which are functions of the ionic strength but not the membrane surface charges or potentials, nor the oligomer concentration, are included in thetwo tables for comparison to the values of D determined from the MIM-Cl ion-pair associations. These are plotted in Figure S3 below. Note that the electrostatic screening lengths are about an order of magnitude smaller than the minimum separation distances between oligomers due to excluded volume effects involving the tails in the oil phase. In addition, there is no oligomer concentration dependence to the Debye lengths, while a comparison of the data in Figure 5 of the manuscript clearly shows one (compare Fig. 5 (b) with 5 (c)).

Tables S3 through S5 contain the values for the pressure components that make up the data displayed in Figure 5 (a)-(d). These designations are further subdivided into sub-tables giving the components that comprise:

a) the disjoining pressures, Π, as functions of ionic strength, oligomer concentration, and membrane thickness,

b) the electrocompression pressures, Pec, as functions of these parameters and of voltage,

c) the sum of the two, P = Π + Pec.

These are displayed in Figure S6. Table S3 contains an additional sub-table (S3 (a2)) showing the dependence of Pec on voltage at a membrane thickness of hc = 3.95 nm relative to a threshold given by the disjoining pressure at that membrane thickness.

Table S3 (a)

I = 1.0 M, c = 2.0 mg/mL

D = 4.21 nm,hc = 3.95 nm

h (nm) Pmush (N/m2) PvdW (N/m2) Π (N/m2) = Pmush + PvdW

8.00 8.93E+03 -5.91E+03 3.03E+03

7.00 1.28E+04 -8.82E+03 3.94E+03

6.00 1.92E+04 -1.40E+04 5.24E+03

5.00 3.13E+04 -2.42E+04 7.09E+03

3.95 5.88E+04 -4.92E+04 9.60E+03

3.00 1.22E+05 -1.12E+05 1.02E+04

S3 (a2)

I = 1.0 M, c = 2.0 mg/mL

lPecl (N/m2) Π (N/m2)

V hc = 3.95 nm threshold

0.00 0.00E+00 9.60E+03

0.02 2.64E+02 9.60E+03

0.04 1.05E+03 9.60E+03

0.06 2.37E+03 9.60E+03

0.08 4.22E+03 9.60E+03

0.10 6.59E+03 9.60E+03

0.122 9.81E+03 9.60E+03

0.14 1.29E+04 9.60E+03

0.16 1.69E+04 9.60E+03

S3 (b)

Pec (N/m2)

h (nm) 0.02 V 0.04 V 0.06 V 0.08 V 0.10 V 0.122 V 0.14 V 0.16 V

8.00 -6.42E+01 -2.57E+02 -5.77E+02 -1.03E+03 -1.60E+03 -2.39E+03 -3.14E+03 -4.11E+03

7.00 -8.38E+01 -3.35E+02 -7.54E+02 -1.34E+03 -2.10E+03 -3.12E+03 -4.11E+03 -5.36E+03

6.00 -1.14E+02 -4.56E+02 -1.03E+03 -1.83E+03 -2.85E+03 -4.24E+03 -5.59E+03 -7.30E+03

5.00 -1.64E+02 -6.57E+02 -1.48E+03 -2.63E+03 -4.11E+03 -6.11E+03 -8.05E+03 -1.05E+04

3.95 -2.64E+02 -1.05E+03 -2.37E+03 -4.22E+03 -6.59E+03 -9.81E+03 -1.29E+04 -1.69E+04

3.00 -4.56E+02 -1.83E+03 -4.11E+03 -7.30E+03 -1.14E+04 -1.70E+04 -2.24E+04 -2.92E+04

S3 (c)

P (N/m2)

= Π + Pec

h (nm) 0.02 V 0.04 V 0.06 V 0.08 V 0.10 V 0.122 V 0.14 V 0.16 V

8.00 2.96E+03 2.77E+03 2.45E+03 2.00E+03 1.42E+03 6.40E+02 -1.16E+02 -1.08E+03

7.00 3.85E+03 3.60E+03 3.18E+03 2.60E+03 1.84E+03 8.20E+02 -1.68E+02 -1.42E+03

6.00 5.13E+03 4.78E+03 4.21E+03 3.41E+03 2.39E+03 9.96E+02 -3.49E+02 -2.06E+03

5.00 6.93E+03 6.44E+03 5.62E+03 4.47E+03 2.99E+03 9.82E+02 -9.54E+02 -3.42E+03

3.95 9.34E+03 8.55E+03 7.23E+03 5.38E+03 3.01E+03 -2.08E+02 -3.32E+03 -7.27E+03

3.00 9.71E+03 8.34E+03 6.06E+03 2.87E+03 -1.24E+03 -6.81E+03 -1.22E+04 -1.90E+04

Table S4 (a)

I = 1.0 M, c = 0.2 mg/mL

D = 9.11 nm,hc = 3.95 nm

h (nm) Pmush (N/m2) PvdW (N/m2) Π (N/m2) = Pmush + PvdW

8.00 1.91E+03 -5.91E+03 -4.00E+03

7.00 2.72E+03 -8.82E+03 -6.09E+03

6.00 4.11E+03 -1.40E+04 -9.89E+03

5.00 6.68E+03 -2.42E+04 -1.75E+04

3.95 1.26E+04 -4.92E+04 -3.66E+04

3.00 2.61E+04 -1.12E+05 -8.59E+04

S4 (b)

Pec (N/m2)

h (nm) 0.02 V 0.04 V 0.06 V 0.08 V 0.10 V 0.122 V 0.14 V 0.16 V

8.00 -6.42E+01 -2.57E+02 -5.77E+02 -1.03E+03 -1.60E+03 -2.39E+03 -3.14E+03 -4.11E+03

7.00 -8.38E+01 -3.35E+02 -7.54E+02 -1.34E+03 -2.10E+03 -3.12E+03 -4.11E+03 -5.36E+03

6.00 -1.14E+02 -4.56E+02 -1.03E+03 -1.83E+03 -2.85E+03 -4.24E+03 -5.59E+03 -7.30E+03

5.00 -1.64E+02 -6.57E+02 -1.48E+03 -2.63E+03 -4.11E+03 -6.11E+03 -8.05E+03 -1.05E+04

3.95 -2.64E+02 -1.05E+03 -2.37E+03 -4.22E+03 -6.59E+03 -9.81E+03 -1.29E+04 -1.69E+04

3.00 -4.56E+02 -1.83E+03 -4.11E+03 -7.30E+03 -1.14E+04 -1.70E+04 -2.24E+04 -2.92E+04

S4 (c)

P (N/m2)

= Π + Pec

h (nm) 0.02 V 0.04 V 0.06 V 0.08 V 0.10 V 0.122 V 0.14 V 0.16 V

8.00 -4.06E+03 -4.25E+03 -4.58E+03 -5.02E+03 -5.60E+03 -6.39E+03 -7.14E+03 -8.10E+03

7.00 -6.18E+03 -6.43E+03 -6.85E+03 -7.43E+03 -8.19E+03 -9.21E+03 -1.02E+04 -1.15E+04

6.00 -1.00E+04 -1.03E+04 -1.09E+04 -1.17E+04 -1.27E+04 -1.41E+04 -1.55E+04 -1.72E+04

5.00 -1.77E+04 -1.82E+04 -1.90E+04 -2.01E+04 -2.16E+04 -2.36E+04 -2.56E+04 -2.80E+04

3.95 -3.69E+04 -3.77E+04 -3.90E+04 -4.08E+04 -4.32E+04 -4.64E+04 -4.95E+04 -5.35E+04

3.00 -8.64E+04 -8.77E+04 -9.00E+04 -9.32E+04 -9.73E+04 -1.03E+05 -1.08E+05 -1.15E+05

Table S5 (a)

I = 0.1 M, c = 2.0 mg/mL

D = 6.15 nm,hc = 4.41 nm

h (nm) Pmush (N/m2) PvdW (N/m2) Π (N/m2) = Pmush + PvdW

8.00 5.05E+03 -5.91E+03 -8.58E+02

7.00 7.21E+03 -8.82E+03 -1.61E+03

6.00 1.09E+04 -1.40E+04 -3.13E+03

5.00 1.77E+04 -2.42E+04 -6.51E+03

4.41 2.47E+04 -3.52E+04 -1.05E+04

3.00 6.90E+04 -1.12E+05 -4.30E+04

S5 (b)

Pec (N/m2)

h (nm) 0.02 V 0.04 V 0.06 V 0.08 V 0.10 V 0.122 V 0.14 V 0.16 V

8.00 -6.42E+01 -2.57E+02 -5.77E+02 -1.03E+03 -1.60E+03 -2.39E+03 -3.14E+03 -4.11E+03

7.00 -8.38E+01 -3.35E+02 -7.54E+02 -1.34E+03 -2.10E+03 -3.12E+03 -4.11E+03 -5.36E+03

6.00 -1.14E+02 -4.56E+02 -1.03E+03 -1.83E+03 -2.85E+03 -4.24E+03 -5.59E+03 -7.30E+03

5.00 -1.64E+02 -6.57E+02 -1.48E+03 -2.63E+03 -4.11E+03 -6.11E+03 -8.05E+03 -1.05E+04

4.41 -2.11E+02 -8.43E+02 -1.90E+03 -3.37E+03 -5.27E+03 -7.84E+03 -1.03E+04 -1.35E+04

3.00 -4.56E+02 -1.83E+03 -4.11E+03 -7.30E+03 -1.14E+04 -1.70E+04 -2.24E+04 -2.92E+04

S5 (c)

P (N/m2)

= Π + Pec

h (nm) 0.02 V 0.04 V 0.06 V 0.08 V 0.10 V 0.122 V 0.14 V 0.16 V

8.00 -9.23E+02 -1.12E+03 -1.44E+03 -1.89E+03 -2.46E+03 -3.25E+03 -4.00E+03 -4.96E+03

7.00 -1.69E+03 -1.94E+03 -2.36E+03 -2.95E+03 -3.70E+03 -4.73E+03 -5.72E+03 -6.97E+03

6.00 -3.24E+03 -3.59E+03 -4.16E+03 -4.95E+03 -5.98E+03 -7.37E+03 -8.72E+03 -1.04E+04

5.00 -6.68E+03 -7.17E+03 -7.99E+03 -9.14E+03 -1.06E+04 -1.26E+04 -1.46E+04 -1.70E+04

4.41 -1.07E+04 -1.14E+04 -1.24E+04 -1.39E+04 -1.58E+04 -1.84E+04 -2.09E+04 -2.40E+04

3.00 -4.34E+04 -4.48E+04 -4.71E+04 -5.03E+04 -5.44E+04 -6.00E+04 -6.53E+04 -7.22E+04

Figure S6. Pressure components acting at the oligomer bilayer between two aqueous droplets in hexadecane: the disjoining pressure, which is the sum of repulsive interactions between oligomers due to osmotic pressure and attractive van der Waals forces, and the electrocompressive stress, which is an attractive interaction which scales with voltage (shown at 122 mV), and is associated with the thinning of the membrane and removal of trace residual oil from the bilayer. The total pressure is the sum of these two contributions: P = + Pec.

download fileview on ChemRxivMIM-ODMS Collier Supporting Information.docx (58.65 KiB)

Self-Assembly of Charged Oligomeric Bilayers at theAqueous/Organic Interface Regulated by Ion-Pair AssociationsGraham J. Taylor†, Yingdong Luo‡, §, Kunlun Hong§, Stephen A. Sarles⁑, Robert L. Sacci¶, Benjamin Doughty ¶, C. Patrick Collier*, §

†Bredesen Center for Interdisciplinary Research and ⁑Department of Aerospace, Mechanical and Biomedical Engineering, The University of Tennessee, Knoxville, Tennessee 37996, United States¶ Chemical Sciences Division and §Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United StatesKEYWORDS: ionic polymer, droplet interface bilayer, electric double layer, ion pair, electrocompression, disjoining pressure

ABSTRACT: Phospholipid bilayer membranes show promise as biomolecular soft materials that mimicthe ability of living systems to sense, respond and learn but are fragile. Amphiphilic charged oligomers(oligodimethylsiloxane-methylimidazolium cation, ODMS-MIM(+)), assembled into bilayers at the oil-aqueous interfaces of droplet interface bilayers (DIBs), possessed similar size and functionality asphospholipid bilayers, but were stable. The ionic liquid headgroups (MIM(+)) of the oligomers werecovalently bound to short-chain hydrophobic tails (ODMS). Bilayer self-assembly was influenced both bythe charged headgroups, constrained to two-dimensional diffusion at the liquid-liquid interface, whichformed electric double layers in the aqueous phase, and the tails in the organic phase. Bilayers formedspontaneously at low ionic strength but required an external voltage to form at higher ionicities. Thisswitch in assembly behavior was due to ion-pairing of the MIM(+) headgroups with chloride ions, resultingin an increase in the density of the charged headgroups at the interface and the ODMS hydrophobic tailsin the oil phase as they were covalently grafted to the headgroups. Chain overlap led to repulsivedisjoining pressures between droplets due to osmotic stress. The applied voltage caused an attractiveelectrocompressive stress that overcame the repulsion, enabling bilayer formation. Bilayer assembly athigh ionic strength, while requiring a voltage to initiate, was irreversible, and the resulting membranewas considerably more stable than those formed at lower values of the ionic strength. This switching ofassembly behavior can be exploited as an additional mechanism for short-term synaptic plasticity inneuromorphic device applications using soft materials.

Neuromorphic devices assembled from nanoscalestimuli-responsive biomolecules and lipid bilayermembranes offer unique advantages fordeveloping low-power, reconfigurable circuitrycapable of sensing, signal processing, learning,and memorization involving many types ofphysical and chemical signatures not possiblewith solid-state systems.1 Droplet interfacebilayers (DIBs) that form between two or moreaqueous droplets in oil take advantage of thetendency of lipid bilayers to form spontaneouslyat the liquid/liquid interface between immisciblephases,2 as shown in Figure 1a. Recently, ourgroup demonstrated the potential for using lipidbilayers in DIBs with peptide ion channels assynapse-inspired memristors,1 and withoutpeptides as memcapacitors.3 Both can be

integrated into neuromorphic computingapplications. While promising, the fragility of lipid-based membranes ultimately limits the extent towhich biomolecular memristors with lipids canassemble to form extended neuromorphicnetworks large enough to elicit brain-likecomputation.

Amphiphilic polymers self-assemble into syntheticnanoscale membranes with tunable behaviorsreminiscent of lipid-based membranes with morerobust interfaces.4,5 While directional attractiveinteractions resulting from shape anisotropy andenthalpic forces (hydrogen bonds and π-stacking)drive the spontaneous self-assembly of lipidbilayers at hydrophobic-hydrophilic interfaces justa few nanometers thick, for amphiphilic polymers,membrane formation often involves a more subtle

interplay between enthalpic and entropicforces. Most polymer-based membranes areformed from block copolymers, which self-assemble into well-characterized morphologicalphases.5-8 However, these self-assembledphases can be large and complex whileexhibiting long-range repulsive forces – basedon osmotic stress and other excluded volumeinteractions – that hinder membrane formation.Moreover, block copolymer membranes arecompressible and up to an order of magnitudethicker than lipid bilayers, features whichcomplicate the search for design rules affectingassembly and the emergence of neuromorphicbehaviors in memristors and memcapacitorsthat are meaningful and have predictive power.

For example, the voltage-dependent formationof nonionic triblock copolymer membranesformed between aqueous droplets under oil inDIBs has been recently described where themembranes consisted of two hydrophilicpolyethylene oxide (PEO) blocks flanking acentral hydrophobic polydimethylsiloxane(PDMS) linkage.9 The aqueous dropletsfunctionalized with these polymers did notspontaneously adhere to each other becausethe hydrophobic oils commonly used for theformation of DIBs were thermodynamicallygood solvents for the hydrophobic PDMS middleblock. Entrapped residual oil between the twodroplets in the hydrophobic block of thecopolymer due to osmotic stress with the bulkoil phase resulted in an energy barrier that hadto be overcome before membrane assemblycould proceed.

Figure 1. (a) Bright-field image of DIB suspendedon two agarose-coated Ag/AgCl wire electrodes.Scale bar 100 µm. (b) Structure of oligodimethyl-siloxane imidazolium mesylate: ODMS-MIM(+) OMs(−).

Our strategy to address these challenges is todevelop simpler, short chain homopolymersthat resemble lipid molecules more closely than

do bulky block copolymers. We synthesizedoligomers that were highly amphiphilic due tocharged ionic headgroups (methylimidazoliumcation, MIM(+)), covalently bound to short,hydrophobic oligodimethylsiloxane (ODMS)chains (number of monomer segments N = 13)that were monodisperse (all moleculesidentical),10 as sketched in Figure 1b. Thechoice of a small oligomeric molecule allowedus to avoid distributions of molecular weightsand other molecular properties typically seenwith larger polymers. Membrane thicknesseswere determined from specific capacitancemeasurements to be about 4 nm, like lipidbilayers.

We have recently reported on pendant droptensiometry (PDT) and vibrational sumfrequency generation (vSFG) measurements tocharacterize the dynamic self-assembly ofODMS-MIM(+) monolayers at hexadecane-aqueous interfaces. We found that the aqueousphase ionic strength controlled the kinetics andstructures of the amphiphiles in the organicphase as they assembled at the interface.11

Our results showed that hydrogen bonding andelectrostatic interactions between the ionicliquid head groups in the aqueous phaseinfluenced the ODMS oligomer tailconfigurations in the neighboring oil phasebecause of the covalent bond between theheads and the tails.

In this paper, we applied the DIB technique toelectrically probe the formation and stability ofpolymer bilayer membranes formed betweentwo aqueous droplets immersed in a mixture ofhexadecane and ODMS-MIM(+) oligomer. Wediscovered that the specific capacitance of thebilayer was not just a conventional geometricalcapacitance, which is dependent only on thedielectric function and membrane thickness,but also included a term dependent on the ionicstrength.12 Surprisingly, we also discoveredthat while oligomer bilayer membranes formedspontaneously on droplet contact at moderateionic strength values (≤ 400 mM NaCl), voltagethresholds had to be surpassed in order toinduce formation of polymer bilayermembranes at higher ionic strength (≥ 400mM), even though oligomer monolayers werestabilized under the same conditions.“Spontaneity” in this context refers to theformation of a DIB pair between droplets uponcontact, each decorated with a monolayer ofthe ionic oligomer in the absence of an appliedelectric field.

We determined in this work that elevated ionicstrength resulted in the formation of specificion pairs between the methylimidazoliumcation headgroups at the interface and chlorideanions in the aqueous solution of the droplets,beyond the purely coulombic interactionstypically associated with electrostatic Debyescreening decay lengths that assumes all ions

dissociate in the double layer.13-15 Ion pairassociation of the MIM(+) head groups with Cl (-)

subsequently increased the density of theODMS hydrophobic tails in the oil phase sincethey were covalently grafted to theheadgroups. At high ionic strength, the taildensity in each monolayer had increased to thepoint where long-range repulsive forcesbetween the hydrophobic chains in the oilphase dominated the interactions between thetwo droplets, due to the creation of excludedvolume as the chains began to overlap. Thisresulted in an osmotic stress that droveresidual solvent into the bilayer, leading to anet repulsive disjoining pressure16 whichprevented the spontaneous formation of a DIB.

The application of an electric potential above athreshold value of 122 mV reversed this trendand resulted in irreversible DIB formation (onceformed in this manner, the bilayer remainedstable, even at 0 mV). This behavior was due toelectrocompression of the bilayer by a voltage-dependent dielectric stress, which overcamethe repulsive osmotic pressure by thinning themembrane and expelling residual solventsufficiently for shorter-range, attractive van derWaals interactions to become important.12 Onceformed, the oligomer bilayer at high ionicstrength was more stable, with a higheradhesive energy density,17 and its formationwas irreversible, meaning the bilayer remainedintact, even when the transmembrane voltagereturned to 0 mV. The voltage thresholdrequirement disappeared by lowering theaqueous droplets’ ionic strength (I < 400 mM)or by reducing the concentration of the ODMS-MIM(+) oligomers in the oil phase. Both changesreduced the ODMS graft density in eachmonolayer enough to decrease the osmoticstress sufficiently that a net attractive forcedeveloped between the two apposedmonolayers, resulting in spontaneous bilayerformation. However, the resulting bilayer wasless stable.

RESULTS AND DISCUSSION

The backbone tails of the ODMS-MIM(+) oligomerwere too hydrophobic to be miscible in theaqueous droplets, and so the molecule had tobe dissolved in the oil phase, where it formedinverse micelles.18-20 Due to the charged ionicliquid head groups, the oligomers were highlyamphiphilic, and monolayers spontaneouslyformed at the oil-aqueous liquid/liquidinterface, with the ionic liquid head grouplocated on the aqueous side of the interfaceand the hydrophobic ODMS tails on the oil side.Previous PDT and vSFG measurements ofmonolayer formation showed that the ionicstrength-dependent kinetics of self-assemblywere comprised of two distinct regimes whereoligomers first adsorbed and reoriented onrelatively fast time scales followed byconformational sampling and frustrated

packing at longer times. The rates of theseevents

Figure 2. DIB formation at two different values of the ionic strength in the aqueous droplets. (a) At I =0.1 M, bilayer formation was spontaneous after allowing monolayers of the ODMS-MIM(+) ionic oligomerenough time to form on both droplets to an equilibrium packing density before contact. The top panelsschematically show the DIB formation process, while the bottom three panels show the increase incapacitive current corresponding to the formation of a bilayer. (b) At I = 1.0 M, bilayer formation was notspontaneous until a threshold voltage (122 mV) was reached. The bottom three panels on the right-handside show that bilayer formation was irreversible, even at zero bias voltage. The capacitive currents of thebilayer were higher than at low ionic strength, corresponding to a thinner membrane for (b) than for (a).

scaled with the ionic strength (I → 1.0 M NaCl)and the nature of the counter ions.11 Thisinvolved a structural transition of the ODMStails, from directed mainly parallel to theinterface with low packing density at low ionicstrength, to extending out more into the oilphase in order to increase oligomer packingdensity at the interface with elevated ionicstrength, consistent with lower monolayertensions at elevated ionic strength measuredwith PDT. This is significant in that it indicatedthat the hydrophobic tail backboneconfigurations in hexadecane were controlledby electrostatic interactions occurring in theaqueous phase, even though the electrolyteswere not in direct contact with the tails. Thesechanges also coincided with disruption of thehydrogen-bonding network of water at theinterface with increasing ionic strength, asindicated by vSFG spectral changes in the -OHstretching region of water.11

Initial droplet interface bilayer experiments ofODMS-MIM(+) with 100 mM NaCl in the dropletsresulted in spontaneous membrane formationas illustrated in Figure 2a. 300 nL aqueousdroplets were pipetted onto electrodes in the 2mg/mL oligomer-hexadecane mixture andincubated for 10 minutes to allow monolayerassembly of the ionic oligomer around eachdroplet to reach equilibrium packing density.The droplets were then brought into contactwhile monitoring changes in capacitive currentindicative of bilayer formation. Within a few

minutes of bringing the droplets into contact,the current amplitude increased from thebaseline RMS noise level (± 10 pA indicatedwith an asterisk, ‘*’), corresponding to nobilayer, to ± 50 pA (indicated with a doubleasterisk, ‘**’), as shown in the third and fourthpanels from the top in Fig. 2a. Simultaneousbright-field images captured with the CCDcamera on the inverted optical microscopeshowed two droplets that were adjoined by aplanar interface. The current trace after bilayerformation actually had two components(bottom panels), a capacitive currentcomponent indicated by a periodic, rectangulartrace in phase with the driving voltage from thefunction generator, and what appeared to be aweaker ohmic current component, indicated bythe “shark fin” tops and bottoms at eachmaximum in the current trace, instead of theflat tops and bottoms for purely capacitivecurrents. These background “leakage” currentsare in the pA range, corresponding to GΩ

r r r r dmemb ane esistances, compa able to those f om lipir r d r bilaye memb anes an high enough fo single ion channel

r rd rr d eco ing to be ca ie out with high sensitivity.

Unlike the case with the 100 mM NaCl dropletsdescribed above, there was no electrical orvisual indication of spontaneous bilayerformation involving ODMS-MIM(+) betweendroplets containing 1M NaCl during the first 10minutes of being brought into contact, asillustrated in Figure 2b (panel labeled with asingle asterisk, ‘*’). A linear bias voltage ramp

(~ 5 mV/s) was applied to explore thepossibility of voltage-dependent interfaceformation (panel labeled with a double asterisk,‘**’). The current remained at its baseline leveluntil the bias voltage exceeded a threshold ofabout 122 mV, at which point the capacitivecurrent amplitude increased abruptly, signalingthe formation of an oligomer bilayer membraneconnecting the two droplets in a DIBconfiguration (labeled with a triple asterisk,‘***’). The amplitude of the capacitive currentcomponent was higher than for the DIB shownin Figure 2a at low to moderate ionic strength,indicating that a thinner bilayer membrane hadformed. This increase in current wasirreversible (without mechanically pulling thedroplets apart), and the bilayer remained intacteven when the applied bias voltage wasreturned to zero mV (as shown in the bottomthree panels on the right-hand side of Figure2b). This behavior was similar to what wasobserved in the case of the PEO-PDMS-PEOtriblock copolymer in Reference 9 describedearlier in the introduction, except for the factthat bilayer formation in that case wasreversible: once the voltage was reduced belowthe threshold, the bilayer disappeared, and theDIB split into two separate droplets.

The ionic strength-dependent behaviordescribed thus far suggests that NaClconcentration affected bilayer formation viaelectrostatic interactions. In order to rule outthe possibility that the loss of spontaneity ofbilayer formation was caused simply bycolligative effects that lowered the chemicalpotential of the water with increasing NaClconcentration, experiments were performed inwhich droplets contained 0.1 M NaCl + 0.6 Mglucose to simulate the total osmolarity of a 0.4M NaCl droplet (0.8 M osmolarity), but at lowionic strength, I = 0.1 M, well below thethreshold value where voltage-activation wasrequired for bilayer formation. Unlike the casewith 0.4 M NaCl, the droplets containing 0.1 MNaCl + 0.6 M glucose (also 0.8 M osmolarity)promoted spontaneous membrane thinningwithin only a few minutes of the droplets beingbrought into contact. This control experimentindicated that increased concentration ofosmolytes alone cannot explain the diminishedspontaneity of bilayer formation observed withNaCl ≥ 0.4 M.

Further testing of the hypothesis that net ionicstrength dominated the mechanism governingmembrane formation, and not the number ofions (i.e., a colligative effect) involved the useof the divalent cation Ca2+, which can be usedto prepare solutions with the same ionicstrength, but at lower concentrations than NaCldue to its higher valency. The same ionicstrength could be reached with CaCl2 as withNaCl, but at 1/3 the concentration: 33 mM CaCl2substituted for 100 mM NaCl, below the

threshold ionic strength where voltageactivation is required for DIB formation, and133 mM CaCl2 substituted for 400 mM NaCl,above the threshold. The same ionic strengthdependence was seen for both singly anddoubly charged ions, corroborating thehypothesis that electrostatic interactions werepredominantly responsible for the voltageactivation requirement for bilayer formation.We also examined the response with a reducedconcentration of the ionic oligomer in the oilphase. When the concentration of ODMS-MIM(+)

in the oil phase was reduced by an order ofmagnitude, from 2.0 to 0.2 mg/mL, DIBformation regained spontaneity at all values ofthe ionic strength.

Taken together, these observations suggestthat the packing density of the adsorbedoligomers at the liquid/liquid interface playedan important role in whether the application ofa voltage was necessary for bilayer formation,promoted either by:

1. Changing the ionic strength in the aqueous phase changed the electrostatic screening of the charged MIM(+) head groups at the interface and changed their equilibrium separation distance, D.

2. Changing the concentration of the oligomer in the oil phase changed its diffusive flux and steady-state density at the interface.

In general, membrane thickness values can beobtained by specific capacitance

measurements, given by cm=ϵ rϵ 0

d, where

ϵ r is the dielectric constant of the

hydrophobic tail region (~2.3) and ϵ 0 is theelectric permittivity (8.85 x 10-12 F/m). For thisoligomer, however, in addition to thecapacitance of the bilayer, the applied voltagealso charged the electric double layer thatformed at the oil-aqueous interface where theionic liquid head groups were located. Whenthis happens, the true membrane thickness isdetermined by these capacitances in series,resulting in12

d=( ϵ rϵ 0

4 π )( 1cm

−8 π κ−1

ϵ wϵ 0)

(1)

where ϵw is the dielectric constant of theaqueous solution (~70 at elevated saltconcentration) and κ−1 is the Debyescreening length,

κ−1=( ϵw ϵ 0 kTI

2e2 )12

(2)

where I is the ionic strength ([NaCl] in molarity)kT is the product of the Boltzmann constantand the absolute temperature, and e is the unitof elementary charge (1.60 x 10-19 C). Thevalues of the Debye screening length weredependent on the ionic strength, I, of theaqueous solution, given by κ−1 = 0.9 nm at

I = 0.10 M, and κ−1 = 0.3 nm at I = 1.0 M.The measured value of the specificcapacitance, following the experimentalprotocol described in the methods section, was0.39 μF/cm2 at an ionic strength value of I = 1.0M, resulting in a thickness d = 3.95 nm fromEquation (1), and 0.32 μF/cm2 at I = 0.1 M,corresponding to a thickness d = 4.41 nm. Ifthe charging of the electric double layer in theaqueous phase had been ignored, themeasured specific capacitance value of 0.39μF/cm2 at I = 1.0 M would have resulted in avalue of 5.26 nm for the membrane thickness.A membrane thickness at that value wouldhave been inconsistent with the results shownin Figure 2.

Figure 3. Representation of an ODMS-MIM(+)

bilayer membrane between two aqueous dropletsin hexadecane in the dilute mushroom limit. Theoligomers can diffuse freely in 2D along theaqueous-oil interfaces. Both the critical bilayerthickness between the two aqueous-oil interfaces,hc, and the distance of closest approach betweenoligomers in the plane of an interface, D0, are ~2xthe Flory radius (RF = 1.87 nm). MIM-Cl ion pairsare highlighted with purple dashed circles. Scalebar 1.00 nm.

Because of their strong amphiphilicity, theODMS-MIM(+) oligomers were effectively“terminally grafted” at the interface, eventhough the interface was between two liquids.This means that the cationic MIM(+) headgroupsdynamic grafting points were restricted todiffuse in two dimensions parallel to theinterface. Further, since the ODMS tails werecovalently bonded to the headgroups, theywere effectively grafted at the liquid-liquidinterface as well. Our previous PDT and vSFGresults on monolayers at the oil-aqueousinterface indicated that the amphiphilic oligomers had adsorbed atthe monolayer as dilute “mushrooms”,

characterized by the hydrophobic tails adoptingcoiled conformations with segments undergoinga

random walk, tracing out non-overlappinghemispheres in the oil phase. Figure 3illustrates the interactions between apposingmushrooms in the bilayer that occur at a criticalgap distance between aqueous droplets of h =hc ≈ 2 x RF, where RF is the Flory radius of theoligomer accounting for excluded volumeinteractions between the segments:

RF=a4 /5b1 /5 N3/5 = 1.87 nm.21,22 Here a is

the monomer size (0.3 nm), b is the Kuhnlength (1.30 nm), and N is the number ofmonomers (13). For the ODMS-MIM(+) bilayer, 2x RF = 3.75 nm, which was close to the valuefor the equilibrium membrane thicknessdetermined earlier from the specificcapacitance at 1.0 M ionic strength, d = 3.95nm. This number also corresponds to the in-plane distance of closest approach betweenoligomer grafting sites at an interface, D0 = 2xRF = 3.75 nm (Figure 3). This limiting intervalcorresponds to a maximum possible numberdensity, ρ0, of 0.088 oligomers/nm2.

There were also electrostatic repulsiveinteractions between the positively chargedheadgroups at the oil-aqueous interface, forboth the monolayer and the bilayer. If all theelectrolytes had been dissociated in theaqueous phase, then the electrostaticinteractions between the headgroups wouldhave been effectively screened at the Debyelengths given above (< 1 nm), however, it isknown for the similar MIM(+):hexane moleculethat the ionizable MIM(+) surface sites were notfully dissociated, but were only partiallyneutralized due to specific binding of Cl(-)

anions, with a dissociation constant Kd = 0.275M.23 This resulted in a Langmuir-type adsorptionprofile as a function of the ionic strength. Atequilibrium, the effective charge density at theinterface was regulated by the law of massaction:14 MIM(+) + Cl(-) ⇋ MIM-Cl at the oil-aqueous interface. If [MIM(+)]0 and [Cl(-)]0

represent the fully dissociated ionic liquidheadgroup and chloride concentrations at theinterface, and [MIM-Cl]0 the undissociated(neutralized) sites, the interface dissociationconstant for the reaction is represented by14

+¿¿

−¿¿¿0

¿MIM−Cl¿0

¿Cl¿

¿MIM ¿

¿K d=¿

(3)

−¿¿¿C l¿

¿σ0α

σ0(1−α)¿

(4)

−¿¿¿C l¿

¿ α(1−α )

¿

(5)

Here, σ0 =0.014 C/m2 represents the surface charge density corresponding to the maximum number density given earlier: ρ0 = 0.088 oligomers/nm2, α represents the fraction of headgroup sites that were dissociated, ψ0 is

the electrostatic surface potential, and

−¿¿¿C l¿

¿

is the concentration of the chloride anion in thebulk aqueous solution. Combining this equationwith the Grahame equation relating the surfacecharge with the surface potential, derived from Gouy-Chapman theory with the assumption of electroneutrality gives10

σ1=σ0α=σ 0Kd

(Kd+ [NaCl ] e+e ψ0 / kT )(6)

σ2=√8ϵw ϵ 0 kT sinh (+eψ0

2kT )√[NaCl]

(7)

Both ψ0 and σ could be determined numerically by optimizing the expression σ1 – σ2

= 0 (Equations (6) and (7), σ ≤ 10-9 C/m2) using a nonlinear reduced gradient optimizationprogram (“Solver” in Microsoft Excel) with σ0 =

0.014 C/m2, and by assuming the same value for the dissociation constant as for the structurally similar molecule, hexane-MIM(+): Kd = 0.275 M.23

Figure 4 shows (a) the degree of neutralization(1- α) of the charged MIM(+) sites, (b) themembrane surface potential, ψ0 , (c) themembrane surface charge density, σ, and (d)the lateral oligomer separation interval, D, allas functions of the ionic strength ([NaCl]) inthe aqueous solution of the droplets. Thedashed lines in Figure 4 (d) correspond to theclosest possible distances between oligomers(D0 = 3.75 nm), and highlights the fact that theoligomer separation distance dependence onionic strength was a convolution of theinteractions between the tails in the oil phase,goverened by steric, osmotic and dispersiveinteractions, and the effective charge density ofthe ionic liquid MIM(+) headgroups in theaqueous phase, which were dependent on thestructure of the electric double layer in theaqueous phase.

The value of the separation distance at an ionicstrength of I = 1.0 M, D = 4.21 nm, was closeto D0, but not equal, because not all of thepositively charged headgroups were screenedby the counterions, even at the highest appliedionic strength. This resulted in a range of lengthscales for D as a function of ionic strength thatwere much larger than the range of Debyescreening lengths. Figure 4 also shows theeffects of decreasing the concentration of theoligomer in the oil phase from c = 2.0 mg/mL(blue trace) to c = 0.2 mg/mL (orange trace).While the fractions of neutralized sites versusionic strength look almost identical for the twoconcentrations in Figure 4 (a), there wereprofound changes in the membrane surfacepotential, effective interface charge density,and the oligomer separation distance asfunctions of oligomer concentration in the oilphase.This was the result of a decrease in themaximum number density at steady-state by afactor of approximately four due to thedecreased concentration of inverse micelles inthe oil.

Decreasing the bilayer thickness and oligomerseparation as functions of increasing ionicstrength in the aqueous phase resulted in theappearance of repulsive disjoining pressures forthe grafted siloxane oligomers when I ≥ 400mM. Below that value, bilayer formation wasspontaneous, but the bilayer was not as stable.The disjoining pressure consisted of both along-range repulsive term related to increasedosmotic stress in the hydrophobic tail region asoligomer mushrooms start to overlap, and ashort-range attraction due to dispersive van derWaals forces: Π = Pmush + PvdW.16 The long-rangerepulsive component of the disjoining pressurecan be deduced from a confinement energyterm acting on the oligomers at the two oil-

aqueous interfaces in the DIB due to the loss ofconformational entropy as the gap between thedroplets shrinks enough for the mushrooms tooverlap at h ≤ hc:24,25

U conf=kT ( hch )5/3

(8)

The associated force per unit area as a functionof membrane thickness, h, is given by26

Figure 4. Values of (a) (1- α), the fraction of neutralized (undissociated) [MIM-Cl] ion-paired sites, (b)membrane surface potential, ψ0 , (c) membrane surface charge density, σ, and (d) oligomer separation

interval, D, computed using a nonlinear reduced gradient optimization program as functions of theionic strength in the aqueous solution of the droplets. The dashed lines in (d) correspond to the closestdistances possible between adjacent oligomers at the interface, D0, due to repulsive interactions betweenthe hydrophobic ODMS tails in the organic phase. These quantities were determined by setting Equations(6) and (7) in the text equal, and then numerically solving for ψ0 .

Pmush=( kTD2hc )(hch )

8 /3

(9)

Here d = 3.95 nm, the equilibrium bilayerthickness determined from the specificcapacitance, is used for hc. The short-rangeattractive interactions are given by:14

PvdW=−( H6π )( 1h )3

(10)

where H is the estimated Hamaker coefficient,H = 5.7 x 10-20 J, a value consistent with thatreported in the literature for similarimidazolium-based ionic liquids.27

The application of an electrical potentialapplied a long-range attractive electricalcompressive stress to the membrane thatovercame the net repulsive disjoining pressure,

and enabled the formation of the ODMS-MIM(+)

bilayer at high ionic strength.9,16,28 The dielectricstress is given by

Pec=−12ϵ rϵ 0(Vh )

2

(11)

where h is the membrane thickness as above,and V is the applied potential. This expressionwas found by minimizing the free energy in thebilayer, including an electrical work term as aresult of a voltage-dependent capacitiveenergy. The bilayer became thinner in theelectric field, most likely as a result of thesqueezing out of residual oil solvent from themembrane.

Plots of the pressure components acting on theoligomer bilayer given by equations 9 through11 are included in Figure 5 (a) through (d) andincluded in Tables S4-S6 of the SupportingInformation. Figure 5(a) plots the dielectricstress as a function of voltage at the criticalmembrane thickness of hc = 3.95 nm,calculated from Equation (11). It can be seenfrom the plot that the dielectric stressexceeded the threshold disjoining pressureabove 122 mV, corresponding to the voltage-dependent transition from repulsive toattractive interactions at constant membranethickness. Figure 5(b) plots the differences, P= Π + Pec, between the disjoining pressures andthe dielectric stresses as functions of themembrane thickness, h, for varying appliedvoltages ranging from 0 to 160 mV, at an ionicstrength of I = 1.0 M (and an oligomerconcentration of c = 2.0 mg/mL). The signconvention for P is positive values representnet repulsive interactions and negative valuesrepresent net attractive interactions. Thedashed line at P = 0 is the demarcationbetween these two regimes. One can see inFigure 5(b) that P was net repulsive formembrane thickness greater than hc = 3.95 nmand voltages less than 122 mV but becameattractive (P < 0)

Figure 5. Calculated values of (a) the voltage-dependent dielectric stress, Pec, at a membrane thicknessof 3.95 nm with the disjoining pressure threshold at that thickness, (b) P = Π + Pec, at an ODMS- MIM(+)

oligomer concentration in the oil phase of 2.0 mg/mL, and an ionic strength in the aqueous phase of 1.0M. (c) P at an oligomer concentration of 0.2 mg/mL in the oil phase, and an ionic strength in the aqueousphase of 1.0 M. (d) P at an oligomer concentration of 2.0 mg/mL in the oil phase, and an ionic strength inthe aqueous phase.

below hc = 3.95 nm at V ≥ 122 mV. At highervoltages, the interaction was completelyattractive at all values of the membranethickness. Of course, at even lower values ofthe membrane thickness, short-range stericrepulsion ultimately becomes the predominateinteraction. The pressure components thatmake up P as functions of membranethickness, Pmush, PvdW, and Pec, are plotted inFigure S6 in the Supporting Information.

Figure 5 also contains plots of P = Π + Pec for5(c), where the ionic strength decreased from I= 1.0 M to I = 0.1 M, and 5(d), the case wherethe concentration of oligomer added to the oilphase was diluted from c = 2.0 mg/mL to c =0.2 mg/mL. In both cases, P was negative atall values of the membrane thickness, even at0 mV. These results are consistent with theexperimental measurements depicted in Figure2, which showed the bilayer membraneformation was spontaneous for these twocases. For Figure 5 (c), where the concentrationof the oligomer in the oil phase was c = 2.0mg/mL, the same as Figure 5 (a), but with theionic strength decreased from I = 1.0 M to I =

0.1 M, the value of the separation interval hadchanged to D = 6.15 nm, depicted in Figure 4(d), and the critical membrane thickness hadincreased from hc = 3.95 nm to hc = 4.41 nm,as determined by the specific capacitancemeasurements computed from Equation 1. ForFigure 5 (d), where the ionic strength was keptthe same as Figure 5 (a), but the concentrationof the oligomer in the oil phase was reduced bya factor of 10, the critical membrane thicknesswas still hc = 3.95 nm, but the value of theaverage separation interval betweenneighboring oligomers had increased by afactor of 101/3, from D = 4.21 nm to D = 9.11nm, shown in Figure 4 (d).

For the membrane in Figure 5(b), with I = 1.0 Mand c = 2.0 mg/mL, the system had toovercome a significant disjoining pressuredominated by osmotic stress of the ODMS withoil before a stable bilayer could be formed. Thiswas achieved by applying a voltage-dependentelectrocompressive stress (above V = 122 mV)that thinned the membrane sufficiently forshort-range, dispersive interactions to becomeimportant. The bilayer formed this way was the

most stable, consistent with it having thelargest membrane rupture voltage at I = 1.0 M(359 mV) compared to I = 0.1 M (156 mV), andthe fact that the transition at I = 1.0 M wasirreversible (i.e., it was nonvolatile: onceformed the bilayer was stable, even at 0 mV).

Corroborating evidence was provided by in situmonolayer and bilayer interfacial tensionmeasurements of the DIBs, which providedestimates for membrane adhesive forces (freeenergy per unit area of interface):−ΔF=2 γm−γ b .17 For I = 1.0 M, the

adhesive force was −ΔF = 0.27 J/m2 [γm=0.70 mN/m; γ b=1.13 mN/m], while

for I = 0.1 M, it decreased to −ΔF = 0.14

J/m2 [ γm=1.72 mN/m; γ b=3.31 mN/m].Increased molecular packing at higher ionicstrength resulted in lower monolayer andbilayer tensions, and an increased energy ofadhesion and hence, stability.

CONCLUSIONS

In this paper, we emulated lipid bilayers withmonodisperse, ionic, amphiphilic oligomersconsisting of N = 13 dimethylsiloxanehydrophobic units covalently bound to cationicmethylimidazolium ionic liquid headgroups.These oligomers allowed for a differentapproach than the traditional use of blockcopolymers, which can phase separate to formmembranes with nanoscopic features but aretypically much larger and bulkier than lipidbilayers. A novel feature of this approach wasthe inclusion of positively charged MIM(+)

headgroups, which were free to diffuse in twodimensions at the aqueous/oil interface butwere also covalently bound to the hydrophobicODMS tails. This diffusion profile resulted in theformation of electric double layers in theaqueous phases of the DIB droplet pair, andequilibrium oligomer separation intervals thatwere functions of both the aqueous phase ionicstrength, and of excluded volume interactionsbetween oligomers in the dilute mushroom limitin the organic phase.

MIM-Cl ion-pair associations were essential tomechanistically describe the appearance of avoltage dependent threshold for bilayerformation at higher values of the ionic strength.The driving force for ion-pairing was notcoulombic; instead, it originated from structuraldisturbances of neighboring water molecules bythe MIM(+) backbone, which was hydrophobicdue to a delocalized π bond network alongthe imidazole ring. This oriented watermolecules in the vicinity of the headgroupbeyond levels found at a neat oil-aqueousinterface, resulting in a loss of entropy.29,30

Water-water interactions were also affected bythe headgroup. Ion-pairing alleviated thesestructural disruptions of water.31 This pairingscreened the interfacial potential seen by bulk

water molecules analogously to Debyescreening, but with a different charge screeningmechanism. Unlike the regulation of chargedensity at the interface due to ion-pairing,there is no explicit dependence of the Debyescreening length on oligomer concentration(Equation 2). In addition, if only electrostaticconsiderations were important, the separationinterval of oligomers in the plane of theinterface would depend solely on the overlap ofthe hydrophobic tails in the oil phase (D0 = 3.75nm), which was several times greater than theDebye screening lengths for all ionic strengthvalues studied ( κ−1 < 1 nm for [NaCl] ≥ 0.1M). In that case, there would not have been anionic strength dependent switch in oligomermembrane assembly behavior. On the otherhand, the water structure-enforced ion-pairingof MIM(+) and Cl (-) resulted in oligomerseparation intervals that exceeded D0 at lowvalues of ionic strength or low concentrations ofoligomers in the organic phase (Figure 5c andd). This separation prevented long-rangerepulsion due to excluded volume interactionsbetween neighboring ODMS tails in the oilphase, allowing the bilayer to formspontaneously, behavior that switched to arequirement for voltage activation at highervalues of the ionic strength.

These results are consistent with our previousreport on the structure and dynamics of ODMS-MIM(+) monolayers at the oil-aqueous interface.There we showed that at low ionic strength, thesiloxane linkages appeared to initially orientthemselves parallel to the aqueous-oil interfacebut extended out into the oil phase to formdilute “mushrooms” in order to accommodatemore oligomers at the interface at elevatedionic strength.11 We also captured vSFG spectrawith NaI replacing NaCl in the aqueous phase.The more surface active and polarizable I (-)

anion resulted in more compact ODMSmushrooms in the oil phase, indicative ofstronger iodide interactions with theheadgroups and the hydrogen-bonded networkwith water at the interface. One would expectthat for the bilayer, the shift in membraneassembly behavior from spontaneous tovoltage-activated would occur at lower valuesof the ionic strength for I (-) compared to Cl (-), ifit were to occur at all.

The structural simplicity of the charged ODMS-MIM(+) molecules gave them many of theproperties of small-molecule ionic surfactants,which allowed for more thorough andquantitative analyses of their structure andordering at the interface. This analysis will aidthe search for the design rules necessary forthe assembly of novel, bio-inspiredneuromorphic devices. The magnitude of theionic strength in the droplets effectively actedas a “switch” between two very different typesof behavior: the spontaneous formation of a

“volatile” bilayer membrane of lower stability atmoderate values of the ionic strength (≤ 400mM), or a “nonvolatile” bilayer at higher ionicstrength that, although requiring a voltagethreshold to be exceeded for assembly, wasessentially irreversible. In this way, aparameter as easily adjustable as theconcentration of a simple salt (NaCl) in theaqueous droplets of a DIB could lead to facilechemical tunability of the interface and newavenues for controlling synaptic plasticity inmem-elements composed of soft materials.

METHODS

ODMS-MIM(+) Synthesis. Detailed proceduresfor the synthesis of the ODMS-MIM(+), mesylateanion, OMs (-), ionic oligomers, including NMRcharacterization, can be found in theSupporting Information of Reference 11.

Ionic Oligomer DIB Formation andCharacterization. Droplet interface bilayerswere formed as described previously in studiesof lipid-based DIBs.2 The aqueous dropletsconsisted of deionized water (18.2 MΩ cm)⸱containing 10 mM 3-(N-morpholino)propanesulfonic acid (MOPS, Sigma), with a pHof 7.4, and varying amounts of sodium chloride(NaCl, 0.1 to 1.0 M) for electrical conductivityand to control the ionic strength, which is given

by I ¿ 12∑i

n

ci z i2 , where ci is the molar

concentration of the ion i, zi is its valence, andthe sum is taken over all the ions in thesolution. Here we assume that the solutionconcentration of the initial mesylate counterionto the MIM(+) head group in the inverse micellesin the oil phase was negligible compared to thechloride anion at all NaCl concentrations in theaqueous phase of the droplets. The dropletswere anchored to silver/silver chloride(Ag/AgCl) wires (Goodfellow) by coating theirball-ended tips with a 1% agarose gel solution.The droplets, anchored on the electrodes, weresubmerged in hexadecane (≥ 99%, Sigma)which contained 2 mg/mL ionic oligomerdissolved in it (unless otherwise indicated indilution experiments). Droplet were placed intocontact after allowing 10 minutes formonolayers to assemble and attain equilibriumpacking.

Electrical Measurements and Imaging.Ionic oligomer membrane formation wasdetected as an increase in membranecapacitance by supplying a 10 Hz, 10 mVtriangular waveform from a function generatorclamped at zero bias voltage (Agilent). Due tothe highly resistive and capacitive nature of thebilayer membrane, the resulting currentresponse was square-like. To determine voltagethresholds for DIB formation, a custom LabViewcode was used to overlay the triangularwaveform on a slowly increasing bias ramp.

These input signals were routed through anAxopatch 200B or an Axopatch 1D patch-clampamplifier (Molecular Devices). For convention,positive current refers to current flowing fromthe positive terminal of the Axopatchheadstage. The capacitance of the polymerinterface was extracted from sections of thesquare-wave current response using a MATLABscript (available upon request). In parallel,changes in the minor axis of the membrane, R,were acquired from bright field images of thedroplets viewed from below through a 4xobjective lens on a Nikon TE-300 invertedoptical microscope. The images were post-processed using custom scripts in MATLAB toextract values of R. The bilayer areas fromthese measurements were then used tocalculate the specific capacitance, cm, fromwhich the membrane thickness d was obtained.Simultaneously, images of contact angles ofthe DIB as functions of applied voltages werecaptured and used to help determine changesto both the monolayer and bilayer interfacialtensions.17 All current recordings were madeusing the patch clamp amplifiers listed aboveand a Digidata 1440 data acquisition system(Molecular Devices). For all measurements,droplets and measurement probes were placedunder a custom-made Faraday cage tominimize ambient electrical noise.

ASSOCIATED CONTENT Supporting Information. Data for Figure 4 andFigure 5 in table format, plot of the disjoiningpressure, the electrocompressive stress at 122mV, and the total pressure acting on the oligomerbilayer, which is the sum of these twocontributions. This material is available free ofcharge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

Corresponding Author*colliercp@ornl.gov

ORCIDGraham J. Taylor - 0000-0001-8833-7705Yingdong Luo - 0000-0002-1945-0680Kunlun Hong - 0000-0002-2852-5111Stephen A. Sarles - 0000-0002-6694-6451Robert L. Sacci - 0000-0002-0073-5221Benjamin Doughty - 0000-0001-6429-9329C. Patrick Collier - 0000-0002-8198-793X

Present Address† T&T Scientific Corporation, 7140 Regal Lane,Knoxville, TN 37918, United States‡ Applied Materials, Inc., 3050 Bowers Avenue, Santa Clara, CA 95054, United States

Author ContributionsG.J.T and S.A.S. carried out electricalmeasurements on DIBs, R.L.S., B.D. and C.P.C.carried out data analysis, K.H. and Y.L. preparedthe ODMS-MIM(+) sample. The manuscript waswritten through contributions of all authors. All

authors have given approval to the final version ofthe manuscript.

Funding SourcesY.L. and K.H. were sponsored by the LaboratoryDirected Research and Development Program atOak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U.S. Department of Energy.R.L.S. was supported by the U.S. Department ofEnergy, Office of Science, Basic Energy Sciences,Materials Sciences and Engineering Division. B.D.was supported by the U.S. Department of Energy,Office of Science, Basic Energy Sciences, ChemicalSciences, Geosciences, and Biosciences Division.S.A.S. was supported by the Air Force Office ofScientific Research Multi-University ResearchInitiative Grant FA9550-19-1-0213.

ACKNOWLEDGMENT A portion of this research was carried out at theCenter for Nanophase Materials Sciences, which isa DOE Office of Science User Facility. Researchdata analysis was also supported by the U.S.Department of Energy, Office of Science, BasicEnergy Sciences, Chemical Sciences, Geosciences,and Biosciences Division (B.D.), and by the U.S.Department of Energy Office of Science, BasicEnergy Sciences, Materials Science andEngineering Division (R.L.S.).

ABBREVIATIONS

DIB, droplet interface bilayer; ODMS-MIM(+), oligodimthylsiloxane - methylimidazolium cation; OMs (-), mesylate anion; vSFG, vibrational sum frequency generation; PDT, pendant drop tensiometry.

REFERENCES1. Najem, J.S.; Taylor, G.J.; Weiss, R.J.; Hasan,

M.S.; Rose, G.; Schuman, C.D.; Belianinov, A.; Collier, C.P.; Sarles, S.A. Memristive Ion Channel-Doped Biomembranes as Synaptic Mimcs. ACS Nano 12, 4702-4711 (2018).

2. Bayley, H.; Cronin, B.; Heron A.; Holden, M.A.; Hwang, W.L.; Syeda, R.; Thompson, J.; Wallace, M. Molecular BioSystems 4, 1191-1208 (2008).

3. Najem, J.S.; Hasan, M.S.; Williams, R.S.; Weiss, R.J.; Rose, G.S.; Taylor, G.J.; Sarles, S.A.; Collier, C.P. Dynamical Nonlinear Memory Capacitance in Biomimetic Membranes. Nat. Commun. 2019, 10, 3239.

4. Nardin, C.; Winterhalter, M.; Meier, W. Giant Free-Standing ABA Triblock Copolymer Membranes. Langmuir 2000, 16, 7708-7712.

5. Nardin, C.; Meier, W. Hybrid Materials from Amphiphilic Block Copolymers and Membrane Proteins. Rev. Mol. Biotechnol. 2002, 90, 17-26.

6. Sustina, B.; Polymeropoulos, G.; Musteata, V.; Peinemann, K.-V.; Avgeropoulos, A.; Smilgies, D.-M.; Hadjichristidis, N.; Nunes, S.P. Design of Block Copolymer Membranes Using Segregation Strength Trend Lines. Mol.Syst. Des. Eng. 2016, 1, 278-289.

7. Stoenescu, R.; Graff, A.; Meier, W. Asymmetric ABA-Triblock Copolymer Membranes Induce a Directed Insertion of Membrane Proteins. Macromol. Biosci. 2004,4, 930-935.

8. Morton, D.; Mortezaei, S.; Yemenicioglu, S.; Isaacman, M.J.; Nova, I.C.; Gundlach, J.H.; Theogarajan, L. Tailored Polymeric Membranes for Mycobacterium smegmatis porin A (MspA) Based Biosensors. J. Mater. Chem. B 2015, 3, 5080-5086.

9. Tamaddoni, N.; Taylor, G.; Hepburn, T.; Kilbey, S.M.; Sarles, S.A. Reversible, Voltage-Activated Formation of Biomimetic Membranes between Triblock Copolymer-Coated Aqueous Droplets in Good Solvents. Soft Matter 2016, 12, 5096-5109 (2016).

10. M. Pitsikalis, M.; Hadjichristidis, N.; Mays, J.W.Model Mono-, Di-, Tri-ω-Functionalized Three-Arm Star Polybutadienes. Association Behavior in Dilute Solution by Dynamic Light Scattering and Viscometry. Macromolecules 1996, 29, 179-184.

11. Chowdhury, A.U.; Taylor, G.J.; Bocharova, V.; Sacci, R.L.; Luo, Y.; McClintic, W.T.; Ma, Y.-Z.; Sarles, S.A.; Hong, K.; Collier, C.P.; Doughty, B. Insight into the Mechanisms Driving the Self-Assembly of Functional Interfaces: Moving from Lipids to Charged Amphiphilic Oligomers J. Am. Chem. Soc. 2020, 142, 290-299.

12. Everitt, C.T.; Haydon, D.A. Electrical Capacitance of a Lipid Membrane SeparatingTwo Aqueous Phases. J. Theoret. Biol. 1968, 18, 371-379.

13. Hanaor, D.A.; Ghadiri, M.; Chrzanowski, W.; Gan, Y. Scalable Surface Area Characterization by Electrokinetic Analysis ofComplex Anion Adsorption. Langmuir 2014, 30, 15143-15152.

14. Israelachvili, J.N. Intermolecular and Surface Forces; Academic Press: San Diego, 1992; pp 233-237.

15. Helm, C.A.; Laxhuber, L.; Lösche, M.; Möhwald, H. Electrostatic Interactions in Phospholipid Membranes 1: Influence of Monovalent Ions. Colloid Polym. Sci. 1986, 264, 46-55.

16. Anklam, M.R.; Saville, D.A.; Prud’homme, R.K. Disjoining Pressure and Film Tension in Comb-Graft Copolymer Stabilized Oil Films. Langmuir 1999, 15, 7299-7307.

17. Taylor, G.J.; Venkatesan, G.A.; Collier, C.P.; Sarles, S.A. Direct In-Situ Measurement of Specific Capacitance, Monolayer Tension, and Bilayer Tension in a Droplet Interface Bilayer. Soft Matter 2015, 11, 7592-7605.

18. Verbeeck, A.; Voortmans, G.; Jackers, C.; De Schryver, F.C. Characterization and Stabilization of Inverse Micelles. Langmuir 1989, 5, 766-776.

19. Pileni, M.-P. Water in Oil Colliodal Droplets Used as Microreactors. Adv. Colloid Interface Sci. 1993, 46, 139-163.

20. Dukhin, A.; Parlia, S. Ions, Ion Pairs and Inverse Micelles in Non-Polar Media. Curr. Op. Coll. Interface Sci. 2013, 18, 93-115.

21. Flory, P.J. The Configuration of Real Polymer Chains J. Chem. Phys. 1948, 17, 303-310.

22. Rubinstein, M.; Colby, R.H. Polymer Physics; Oxford University Press: Oxford, 2003; pp 102-104.

23. R. Sadeghi, R.; Ebrahimi, N. Ionic Associationand Solvation of the Ionic Liquid 1-Hexyl-3-methylimidazolium Chloride in Molecular Solvents Revealed by Vapor Pressure Osmometry, Conductometry, Volumetry, andAcoustic Measurements J. Chem. Phys. B 2011, 115, 13227-13240.

24. de Gennes, P.G.; Scaling Concepts in Polymer Physics; Cornell University Press: Ithaca, 1979; pp 38-43.

25. Casassa, E.F. Equilibrium Distribution of Flexible Polymer Chains between a Macroscopic Solution Phase and Small Voids. Polym. Lett. 1967, 5, 773-778.

26. de Gennes, P.G. Polymers at an Interface; A Simplified View. Adv. Colloid Interface Sci. 1987, 27, 189-209.

27. Ueno, K.; Inaba, A.; Kondoh, M.; Watanabe, M. Colloidal Stability of Bare and Polymer-Grafted Silica Nanoparticles in Ionic Liquids. Langmuir 2008, 24, 5253-5259.

28. El-Beyrouthy, J.; Makhoul-Mansour, M.M.; Taylor, G.; Sarles, S.A.; Freeman, E.C. A New Approach for Investigating the Response of Lipid Membranes to Electrocompression by Coupling Droplet Mechanics and Membrane Biophysics. J. R. Soc. Interface 2019, 16, 20190652.

29. Diamond, R.M. The Aqueous Solution Behavior of Large Univalent Ions. A New Typeof Ion Pairing. J. Phys. Chem. 1963, 67, 2513-2517.

30. Marcus, Y.; Hefter, G. Ion Pairing. Chem. Rev. 2006, 11, 4585-4621.

31. Ramanathan, P.S.; Krishnan, C.V.; Friedman, H.L. Models Having the Thermodynamic Properties of Aqueous Solutions of Tetraalkylammonium Halides. J. Solution Chem. 1972, 1, 237-262.

15

download fileview on ChemRxivMIM-ODMS oligomer bilayer membranes Collier_final.docx (2.72 MiB)