Electrochemical triggering of lipid bilayer lift-off oscillation at the electrode interface

Abstract

In situ studies of transmembrane channels often require a model bioinspired artificial lipid bilayer (LB) decoupled from its underlaying support. Obtaining free-standing lipid membranes is still a challenge. In this study, we suggest an electrochemical approach for LB separation from its solid support via hydroquinone oxidation. Layer-by-layer deposition of polyethylenimine (PEI) and polystyrene sulfonate (PSS) on the gold electrode was performed to obtain a polymeric nanocushion of [PEI/PSS]₃/PEI. The LB was deposited on top of an underlaying polymer support from the dispersion of small unilamellar vesicles due to their electrostatic attraction to the polymer support. Since lipid zwitterions demonstrate pH-dependent charge shifting, the separation distance between the polyelectrolyte support and LB can be adjusted by changing the environmental pH, leading to lipid molecules recharge. The proton generation associated with hydroquinone oxidation was studied using scanning vibrating electrode and scanning ion-selective electrode techniques. Electrochemical impedance spectroscopy is suggested to be a powerful instrument for the in situ observation of processes associated with the LB–solid support interface. Electrochemical spectroscopy highlighted the reversible disappearance of the LB impact on impedance in acidic conditions set by dilute acid addition as well as by electrochemical proton release on the gold electrode due to hydroquinone oxidation.

1. Introduction

A vital parameter necessary for the completion of various chemical processes in solution chemistry [1,2] and biochemistry [3] is the acidity at which the reactions take place. Fine-tuning the environmental acidity serves as a powerful regulatory method, influencing the outcome of these processes [4,5] and allowing for the desired effects to be achieved. pH-sensitive systems play a significant role in the design of those smart materials that respond to external stimuli [6] such as soft robots [7], self-organizing structures [8,9] and nanocapsules [10,11]. The acidity of the environment greatly affects those systems held together by electrostatic interactions [12].

Numerous strategies have been developed for modifying the pH of a solution on both large and small scales. Recent reports have included photochemical [13], enzymatic [14] and electrochemical [8,9] approaches to pH adjustment. Faraday's laws of electrolysis show that electrochemically adjusting the pH of a system enables quantitative control over the generation of pH-determining ions. Precise spatiotemporal control can be performed due to the high localization of reactions on the electrode surface.

Owing to their proton-coupled electron transfer [15], low redox potential and relative chemical stability quinones are widely used as the electroactive species for the controlled generation/consumption of protons [8,9]. Naturally occurring hydroquinone compounds play a major role in electron/proton transfer of many biological processes [16,17]. Thus, the quinone/hydroquinone transition serves as an important electrochemical model for the development of biomimetic systems. Its changing molecular structure allows for its electrochemical properties to be tuned [18].

Designing artificial composites that mimic living matter is of utmost importance to modern science. Many attempts have been made to imitate the structure and properties of biological membranes by use of artificial polyelectrolyte–lipid composites [19,20]. A lipid bilayer (LB) supported by a polyelectrolyte cushion provides a platform for the investigation of many cellular processes [21,22].

Typically, polyelectrolyte multilayers (PEM) are assembled by various adaptations of the classical layer-by-layer (LbL) technique. The basis of the LbL technique is the cyclical alternating deposition of oppositely charged components. Many variations of this method exist [23]. Two of the most developed techniques for LB formation are the Langmuir–Blodgett/Langmuir–Schaefer method [24] and deposition from the dispersion of vesicles in water [25,26]. The latter, technically simple and versatile, involves the adsorption of the dispersed vesicles to an oppositely charged surface, rupture of said vesicles and spreading to form a continuous bilayer [27,28].

It is well known that the phospholipid membrane is attached to the underlaying polyelectrolyte cushion by a combination of electrostatic forces and hydrophobic interactions [24]. Polyelectrolyte multilayer assemblies are held together by electrostatic forces as well [29]. Therefore, biomimetic lipid–polyelectrolyte composites are strongly affected by the pH of the surrounding medium. Repelling of the components can be achieved when the isoelectric point of some component is reached and the interaction switches from attraction to repulsion. It has been shown that by making the environmental pH acidic, the distance between the LB and the underlaying support can be varied [30]. Structural characterization by neutron reflectometry revealed that, due to the pH-sensitivity of lipids, the distance between the polymeric cushion and the LB can be reversibly adjusted by as much as 300 nm [31] by simply varying the pH of aqueous environment. Molecular dynamic simulations later confirmed this. Such manipulation of the intermolecular interactions between the lipid membrane and its underlaying support and the subsequent distance between them may be useful in obtaining free-standing lipid membranes for the study of transmembrane proteins. This allows for the denaturation of these proteins to be avoided while simultaneously affording the use of specific analytical techniques not available for black lipid membranes [32]. Since mechanic properties of lipid membrane are highly dependent on fabrication approach, free-standing lipid membranes detached from underlaying polymer support could be useful model for the understanding of cell mechanics [33,34]. Electrotriggered lipid lift-off may also be considered as an alternative to phospholipid bilayer arrays patterning [24] and study of the lipid membrane's flexoelectricity [35].

Here we report the electrochemically triggered reversible LB lift-off from the underlying conducting substrates covered by polyelectrolytes via hydroquinone oxidation coupled with proton release (figure 1). We have conducted electrochemical studies on a phosphatidylcholine bilayer supported by a polyelectrolyte multilayer of polyethylenimine (PEI) and poly(sodium 4- styrenesulfonate) (PSS) on top of gold, obtained by the vesicle rupture and LbL techniques respectively. The anodic activity and proton concentration in close proximity to the gold electrode during the oxidation of hydroquinone to quinone was studied by use of the scanning vibrating electrode technique (SVET) and scanning ion-selective electrode technique (SIET). It has been demonstrated that the polyelectrolyte layers show a decrease in conductivity after LB assembly [36]. Electrochemical impedance spectroscopy (EIS) has become a widely used technique in the study of lipid layer formation and characterization [37]. Electrochemical spectra have revealed that a dense LB atop a polyelectrolyte multilayer can attain extremely high resistance and capacitance values [38].

Figure 1.

Local acidification of medium in close proximity to the electrode surface during the electrochemical oxidation of hydroquinone to quinone and the resulting lift-off of the LB from its underlying support due to electrostatic repulsion after lipid recharge due to protonation. (Online version in colour.)

Here we suggest EIS to be a powerful instrument for the routine analysis of the LB. It serves as a viable alternative to neutron reflectivity—a technique that is not always readily available and limited to flat, smooth specimens. Electrochemical spectroscopy revealed that in the acidic conditions generated by the quinone/hydroquinone redox reaction, the LB impact in impedance of gold/PEM/LB composites disappears. Its lift-off can be attributed to its electrostatic repulsion from its underlaying support [31]. The process described above is observed to be reversible. After a short relaxation time, the proton concentration decreases, returning to near the starting concentration due to diffusion, causing the LB to go back, followed by impedance growth.

2. Experiment

2.1. Chemicals

Branched PEI (Mw 70 kDa) 30% water solution purchased from Alfa Aesar; poly(sodium 4-styrenesulfonate) (PSS, Mw 500 kDa) purchased from Polysciences Inc.; liquid soya lecithin Lecisoy 400 containing mostly phosphatidylcholines obtained from Cargill, USA; reagent grade NaCl (99.5%) obtained from Merck and hexane from Ekos-1, Russia. All chemicals were used as received without any purification. Gold wire, 0.2 mm in diameter, was purchased from Agar (UK). Pieces of gold wire of 2–3 cm length were embedded in epoxy resin so that circular cross-section of gold wire exposed to outside media on the flat surface of the obtained holder. Prior to the assembly process, the gold substrates were rinsed with acetone then ethanol and, subsequently, in an excess of distilled water.

2.2. Polyelectrolyte multilayered film formation

The deposition of the polyelectrolyte cushion onto the gold embedded in the epoxy resin was performed using the classical LbL technique 2 mg ml⁻¹ each PEI and PSS were dissolved in 0.5 M aqueous NaCl to make polycation and polyanion solutions, respectively. Each layer took 20 min to be deposited after which it was rinsed with excess distilled water and then steam-dried. A total of seven layers were deposited, resulting in a [PEI/PSS]₃/PEI polyelectrolyte multilayer.

2.3. Lipid vesicles preparation and characterization

For the preparation of the lipid small unilamellar vesicles (SUV) solution following procedure was performed. A soy lecithin solution in hexane (10 mg ml⁻¹) was kept under vacuum for solvent evaporation for at least 3 h. After removing any hexane traces, a thin lipid film on the bottom of the vessel was obtained. The resulting lipid film was rehydrated using distilled water for a final 10 mg ml⁻¹ concentration with simultaneous sonication in an ultrasonic bath for 15 min. In order to determine the size and charge of the prepared vesicles, dynamic light scattering (DLS) and ζ-potential measurements were conducted with a Photocor Compact-Z analyser. Centrifugation at 14 000 r.p.m. for 20 min was performed to separate the different sized fractions in lipid vesicles suspension.

2.4. Lipid bilayer deposition

To obtain the model LB atop a polyelectrolyte cushion, modified substrates were immersed in 10 mg ml⁻¹ dispersion of SUV for 1 h. Vesicles attached themselves to the oppositely charged PEM surface then ruptured, fused and spread on the surface, forming a continuous bilayer. The resulting structure was then rinsed with a large quantity of distilled water. LB formation as well as PEI/PSS multilayer formation were followed by QCM-D monitoring (Q-Sense AB).

2.5. Acidity adjusting

Two approaches to adjusting the pH were tested. By adding dilute solutions of NaOH and HCl, pH 3 and 7 were gained. The alternative method was the electrogeneration of protons via the oxidation of 60 mM hydroquinone solutions in 150 mM KNO₃ at a constant electric current of 5 µA. Owing to the proton-coupled electron transfer of the quinone/hydroquinone redox system, pH gradient occurs in close proximity to the electrodes during oxidation and reduction processes (figure 2a).

Figure 2.

(a) Cyclic voltammetry curves of hydroquinone on a pure gold electrode and a gold electrode covered by a polyelectrolyte multilayer and polymer cushioned LB, (b) scheme of SVET measurement, (c) Pt-Ir vibrating probe, (d) top view of gold wire embedded in epoxy holder, red arrows are representative of ionic currents measured by SVET, the platinum vibrating electrode is seen at the right bottom of the image, (e) schematic of SIET measurement, (f) glass capillary microelectrode with liquid ion-selective membrane, (g) schematic of EIS measurement, (h) equivalent electrical scheme representing parts of electrochemical system with gold electrode covered by a polyelectrolyte multilayer and LB. (Online version in colour.)

2.6. SVET-SIET measurements

The anodic activity of gold immersed in a hydroquinone solution under an applied positive potential was studied by SVET and generated pH gradients by SIET which enables voltage measurement with nV precision level, electrochemical activity of material to be studied with micrometre spatial resolution and the measurement of extremely low ion concentration gradients [39]. To perform the SVET and SIET measurements, a system from Applicable Electronics (USA) modulated by an ASET program (Sciencewares, USA) was used. As a vibrating probe for SVET (figure 2b) experiments, an insulated Pt-Ir microprobe (Microprobe Inc., USA) with a platinum black spherical tip 30 µm in diameter was used (figure 2c). The probe was made to vibrate both parallel and perpendicular to the specimen surface at a height of 150 µm. The amplitude of vibration was 30 µm, while the probe vibrated at frequencies of 136 Hz (perpendicular to surface) and 225 Hz (parallel to surface). Only the perpendicular component was used in the treatment and presentation of the data (figure 2d). The environmental pH measurements by SIET (figure 2e) were carried out using glass capillary microelectrodes filled with Hydrogen Ionophore Cocktail I (Sigma) based liquid pH-selective membrane and KCl + KH₂PO₄ internal solution (figure 2f). Ag/AgCl/KCl (sat) was used as the external reference electrode. The pH-selective microelectrodes were calibrated using commercially available pH buffers and demonstrated a linear Nernstian response –55 to −58 mV pH⁻¹—in a pH range from 3 to 8. The local activity of H⁺ was detected 25 µm above the surface. The local pH and ionic current density were mapped on a 11 × 11 grid in a 150 mM KNO₃ solution containing 60 mM hydroquinone.

2.7. Electrochemical impedance spectroscopy

Electrochemical measurements were performed using potentiostat Compactstat.e (Ivium, Netherlands). The polyelectrolyte multilayer and LB covered gold substrates were used as working electrodes, Pt wire as a counter electrode and an Ag/AgCl/3 M KCl reference electrode in a quartz electrochemical cell with a volume of approximately 50 ml filled by 150 mM KNO₃ containing 60 mM hydroquinone (figure 2g). The distance between the working and counter electrode was approximately 15 mm. Deposition of the PEM and lipid were performed as described above. Impedance spectra for frequencies between 0.1 and 1000 Hz with an AC modulation amplitude of 5 mV were recorded at a bias potential set equal to the open circuit potential of the working electrode. Data analysis was performed using standard Ivium software (figure 2h).

3. Results and discussion

Owing to the proton-coupled electron transfer of the quinone/hydroquinone redox system, a pH gradient occurs in close proximity to the electrodes during oxidation and reduction processes. A two-step hydroquinone oxidation takes place (figure 2a) where each step and each peak on cyclic voltammetry corresponds to one electronic process—the release of one proton. The oxidation of hydroquinone happens at an oxidative potential of about 0.5–0.6 V while the electrolysis of water requires a minimum of 1.5 V. According to the rules of stoichiometry, the same acidification can be achieved via both methods: water electrolysis and the oxidation of hydroquinone to 1,4-benzoquinone. The latter is preferable because of its lower oxidative potential. It was calculated that hydroquinone oxidation leads to a pH of 3.6 at the surface of electrode and gradually increases to neutral (pH ∼ 7) away from it [40]. It is important that CVA curves are not affected by further deposition of polyelectrolyte layers and LB formation (figure 2a). Since molecules of hydroquinone and quinone are rather small and uncharged they are able to pass freely through the LB and polyelectrolytes, a feat impossible for large molecules and ions.

The anodic activity of the gold electrode and the pH gradients were visualized using scanning microelectrode techniques SVET and SIET (figure 3). The microelectrode measurements demonstrated that the oxidation of quinone to hydroquinone on a bare gold electrode (figure 3a) leads to the local acidification, pH decrease to approximately 4 from initial 7 (figure 3b) in accordance with theoretical modelling [41]. An ionic current of more than 500 µA cm⁻² was observed over the surface of the electrode (figure 3c).

Figure 3.

Schematic of electrode structure, pH map of adjusted to electrode media after passing 5 µA current through electrode immersed in a 60 mM hydroquinone solution in 150 mM KNO₃ measured by SIET, anodic activity of electrode immersed in a 60 mM hydroquinone solution in 150 mM KNO₃ during 5 µA current passing measured by SVET, respectively, for bare gold (a–c), gold/PEM (d–f), gold/PEM/LB (g–i). (Online version in colour.)

Further assembly of a biomimetic nanocomposite simulating a living cell membrane was done using the LbL technique (electronic supplementary material, figure S1). The electrode was first covered by branched PEI to provide strong anchoring to the surface and to act as a positively charged terminating layer. Deposition of PSS was then carried out, making a surface negative. Eventually, a gold/(PEI/PSS)₃PEI multilayered composition was obtained (figure 3d). The choice of terminating layer (positive PEI) is predetermined by the necessity of further LB formation from lipid vesicles suspension. It is of utmost importance that the lipid be oppositely charged to the capping layer of the polymer cushion for strong attachment of lipid vesicles, their rupture and complete coverage of the surface when forming a defect-free self-organized LB on top of polymer support. Lipid vesicles obtained from soy lecithin consisting mostly of phosphatidylcholine were observed to be negative (ζ-potential approx. −35 mV) at neutral pH. Owing to their short radius of curvature, small vesicles are more likely to rupture. Thus size distribution of lipid vesicles significantly affects their stability. As prepared vesicles have bimodal size distribution representing small unilamellar vesicles (SUV) with 100 ± 11 nm and large aggregates of vesicles 5.0 ± 0.5 µm and large fractions can be separated by centrifugation. Negative small vesicles are stable for weeks when stored at 4°C in a fridge.

Although polymer multilayer assembly leads to no change in redox processes at the electrode/hydroquinone solution interface, anodic activity measured by SVET is significantly lower for polyelectrolyte coated gold than for a bare one (figure 3f). During the oxidation of hydroquinone to 1,4-benzoquinone two electrons are transferred to the electrode and release of two protons occurs. Positive protons are repelled by and attracted to the uncompensated positive and negative charges of PEI and PSS, respectively. Since the polyelectrolyte multilayer consists of alternating PEI and PSS layers, the diffusion of charged ions is difficult. The ionic current 250 µm above the surface where it was measured is therefore lower than for a bare gold electrode. The polyethylenimine layer serving as a proton sponge [42] is responsible for the higher pH during hydroquinone oxidation on the gold/PEM electrode in comparison to pure gold (figure 3e).

To obtain polyelectrolyte/lipid composites, lipid vesicles which are negative in neutral media were deposited at pH 7 on the positive PEI capping layer of the polymer cushion obtained by LbL deposition on gold (figure 3g). The LB, which is impermeable to charged species, blocks the ion current from the gold surface (figure 3i), causing all protons released during hydroquinone oxidation to be concentrated beneath the LB and proton concentration above it to be rather low (figure 3h). Whereas strong polyelectrolyte molecules are not pH-sensitive, PSS is one which dissociates in an extremely wide pH range. With its isoelectric point being much lower than that of the lipid [43], PEI is always protonated in aqueous solution [42]. Zwitterionic lipid molecules are able to adsorb protons. Phospholipid molecules forming soy lecithin contain positively and negatively charged moieties, –PO⁻ and N⁺R₃, which can take part in equilibrium processes H⁺ and OH⁻.

$$-\mathrm{P}{\mathrm{O}}^{-}+{\mathrm{H}}^{+}=-\mathrm{POH}$$

$${\mathrm{N}}^{+}{\mathrm{R}}_3+\mathrm{O}{\mathrm{H}}^{-}=\mathrm{N}{\mathrm{R}}_3\mathrm{OH}.$$

Thus -PO⁻, -POH, N⁺R₃ and NR₃OH groups are in equilibrium and the concentration of each is determined by pH and acid–base constants:

$$K_a={\displaystyle \frac{a_{\mathrm{H}}+a_{\mathrm{P}{\mathrm{O}}^{-}}}{a_{\mathrm{POH}}}\mathrm{and}{K}_b={\displaystyle \frac{a_{{\mathrm{N}}^{+}{\mathrm{R}}_3}{a}_{\mathrm{O}{\mathrm{H}}^{-}}}{a_{\mathrm{NR}}{{\hspace{0.17em}}_{{\hspace{0.17em}}_3}}_{\mathrm{OH}}}.}}$$

Because of the constant value, the amine group of phospholipids always remains in a positively charged state. The pKa of the phosphate group for phosphatidylcholine was reported to be approximately 3.5 [44]. Therefore, at pH 7 $a_{\mathrm{P}{\mathrm{O}}^{-}}$ in 10³–10⁴ higher than a_POH, and at pH 3 the protonated state starts to prevail. Thus, at neutral conditions phosphate groups exist in the deprotonated state. As some components of soy lecityhin in this case carry only negative charge, which is not compensated by the positive amine group (e.g. phosphatidylinositol and phosphatidic acid) [45], there is negative net charge of lipid structures. In acidic media, due to protonated phosphate groups, positive net charge occurs.

The pH-dependence of lipid vesicles was demonstrated by ζ-potential measurements (electronic supplementary material, figure S2). Phospholipid vesicles adjusted by diluted HCl to pH 5, 3 and 2 demonstrate exponential growth of charge, becoming positive in pH lower than 4 and keeping constant size distribution. Thus, the protonation of lipids followed by their recharge from initially being strongly negative to only slightly negative and even positive at low pH occurs, whereas PSS is still negative and PEI, which is on top of the polymer cushion, is still positive. Such charge-shifting behaviour leads to differing strengths of electrostatic interactions between the LB and its underlaying support, causing LB lift-off from the polymeric cushion due to Coulombic repulsion in strong acidic pH [31]. Neutron reflectometry is suggested to be a powerful instrument in the investigation of the dynamic behaviour of polymer–lipid composites and the separation distance can be determined within nanometre precision [31]—but only for extremely flat and smooth species. EIS can serve as a more accessible and less expensive alternative for the qualitative evaluation of the process because continuous LBs are characterized by high resistivity. Experiments were performed on the gold/PEM and gold/PEM/LB working electrodes at pH 7 and pH 3, adjusted by dilute HCl (figure 4) and later at pH 7 before and after running a galvanostatic 5 µA current (figures 5 and 6).

Figure 4.

Bode plots for impedance spectra of (a) gold/[PEI/PSS]₃/LB in neutral conditions, (b) gold/[PEI/PSS]₃/LB in acidic conditions set by dilute HCl addition to solution, (c) gold/[PEI/PSS]₃ in neutral conditions, (d) gold/[PEI/PSS]₃ in acidic conditions set by addition of dilute HCl to the solution, top insets are schematics of processes occurring on the electrode, bottom insets are equivalent electrical schematics used to fit experimental data. (Online version in colour.)

Figure 5.

Bode plots for impedance spectra of gold/PEM in 60 mM hydroquinone solution (a) in neutral conditions before passing electrical current and (b) after 15 min of passing galvanostatic current of 5 µA followed by local acidification due to hydroquinone oxidation coupled with protons release. (Online version in colour.)

Figure 6.

Plots of impedance spectra of gold/PEM in 60 mM hydroquinone solution (a) before passing electrical current, (b) after 15 min of passing galvanostatic current of 5 µA followed by local acidification due to hydroquinone oxidation coupled with protons release and relaxation to the initial state (a) after 15 min at open circuit potential. (Online version in colour.)

Under an open circuit potential, no Faradaic processes are involved and the impedance of the electrochemical system may be described in terms of Ohmic resistance and capacitance. An equivalent electrical circuit describing gold/PEM/LB composition is given in figure 4a (bottom inset). R_E corresponds here to the Ohmic resistance of the electrolyte solution, R_S for the Ohmic resistance of the gold/PEM solid support, CPE_S (CPE_S = 1/4 exp(αjπ/2)/QωN, where j is the unit imaginary number, ω is frequency, while Q and N are frequency independent parameters) is for the gold/PEM imperfect capacitance representing its significantly rough surface, R_LB and C_LB are related to the Ohmic resistance and capacitance LB, respectively. The contribution of the counter electrode is also considered (R_CE and C_CE) [36]. When the pH of the surrounding medium was adjusted to 3 by adding dilute HCl dropwise, the impedance characteristics of the gold/PEM/lipid composite change dramatically. The equivalent electrical circuit described above does not suitably describe the impedance of the system under these new conditions and changes that have occurred cannot be attributed to R_S. Another equivalent electrical circuit that does not take into account the contributions of the LB (R_LB and C_LB) was therefore employed. The plots of the impedance spectra of gold/PEM/LB composite in neutral and acidic conditions are both shown in figure 4a,b. These results demonstrate that in acidic conditions the LB drifts away from the gold/PEM support due to lipid protonation and recharge from negative to positive, resulting in the repulsion of the positive protonated LB from the positive underlaying PEI. To demonstrate the stability of the polyelectrolyte multilayer, impedance measurements for the gold/PEM composite were also taken. It was observed that neither the impedance spectrum nor equivalent cirquit and best-fitting parameters undergo any changes when the environmental acidity changes from pH 7 to pH 3. Gold/PEM impedance may be described by a combination of R_E, R_S and CPE_S (figure 4c,d, bottom insets). Any minor differences in impedance between neutral and acidic conditions can be related to a decrease in the Ohmic resistance of the solution due to an increase in the concentration of the conducting protons. It is important to emphasize that the S-shaped phase curve and parabolic impedance curves for gold/PEM electrodes are similar to those obtained for gold/PEM/lipid electrodes in acidic conditions. Passing a 5 µA electrical current through the solution and the resulting acidification of the surrounding medium does not influence the impedance characteristics of the gold/PEM electrode much (figure 5). Experimental impedance of gold substrates coated by the PEI/PSS polymer cushion before passing the current (figure 5a) and immediately after in open circuit conditions (figure 5b) are almost identical as well as fitting parameters according to equivalent circuit described above for gold/PEM composite (figure 5a,b, bottom insets, table 1). As for hydroquinone oxidation and release of protons, measuring the impedance of the gold/PEM/LB before and after passing an electrical current show significant differences (figure 6a,b). Local acidification using electrochemical proton-coupled electron transfer have the same effect on pH-sensitive species as the acidification of bulk solution (figure 4a,b). Local acidification leads to the ‘disappearance’ of the contribution of the LB on the total impedance, presented by R_LB and C_LB elements of the equivalent cirquit (figure 6a). After current passing and 15 min relaxation followed by the diffusion of excess protons from adjoined to electrode media and pH retrieval to bulk value lipid molecules come back to the initial not protonated state and are attracted back to the oppositely charged underlying support (figure 6a). The best-fitting parameters are presented in table 2. Reversibility is a distinguishing feature of LB lift-off triggered by electrochemical protonation. Apart from the solution exchange method of pH adjustment, hydroquinone oxidation is not associated with any significant fluxes able to remove protonated lipid molecules loosely bonded to a positive polymer cushion. Figure 7 demonstrates the oscillatory behaviour of impedance during direct current on–off cycles. As is evident, the LB on the top of the gold/PEM substrate demonstrates resistive behaviour (greater than 5000 Ω at 1 Hz oscillation of open circuit potential with amplitude of 5 mV). Immediately after 15 min of passing current, the impedance was observed to be approximately 2000 Ω under open circuit potential with the same conditions. After 15 min of relaxation, the impedance increases to 3000–3500 Ω but never returns to its starting value (figure 7a). During further cycles of passing current and relaxation, the oscillating behaviour of impedance was observed. Owing to proton release from hydroquinone oxidation during passing current followed by reversible lipid protonation and its recharge, oscillation of electrostatic forces between the LB and its underlaying support occurs. Sustainability of the LB and reversibility of the process was demonstrated for at least six cycles during a period of 3 h (figure 7b).

Table 1.

Best-fitted parameters of impedance data of gold/PEM composite before and after passing 5 µA current.

	before	after
RS (Ω)	15 ± 6	80 ± 17
RCE (Ω)	640 ± 20	640 ± 45
CCE (F)	2.3 ± 0.1 × 10−3	2.3 ± 0.1 × 10−3
RWE (Ω)	910 ± 10	2200 ± 70
QWE (Siemens × second)	9.8 ± 0.5 × 10−6	9.7 ± 0.5 × 10−6
N	0.85	0.84

Table 2.

Best-fitted parameters of impedance data of gold/PEM/LB composite before and after passing 5 μA current.

	before current	after current	after relaxation
RS (Ω)	50 ± 10	60 ± 30	38 ± 9
RCE (Ω)	70 ± 30	60 ± 10	110 ± 50
CCE (F)	9 ± 3 × 10−5	7 ± 2 × 10−4	3.7 ± 1 × 10−3
RWE (Ω)	260 ± 60	700 ± 120	1000 ± 100
QWE (Siemens × second)	1.5 ± 0.4 × 10−3	4 ± 2 × 10−5	9 ± 1 × 10−5
N	0.90	0.74	0.87
RLB (Ω)	470 ± 40		160 ± 20
CLB (F)	1.1 ± 0.1 × 10−4		1.56 × 10−1

Figure 7.

(a) Impedance curves of Bode plots of impedance spectra of gold/PEM/LB composite before passing 5 µA current, after 15 min of current passing and after 15 min of relaxation, (b) cycles of passing current and relaxation at open circuit potential associated with impedance oscillations at 1 Hz of gold/PEM/LB composite due to LB repulsion in acidified media via hydroquinone oxidation. (Online version in colour.)

4. Conclusion

Phospholipids are known to be charge-shifting molecules highly sensitive to pH. Thus, it is possible to control the electrostatic interaction of lipids with a charged underlaying support. A phosphatidylcholine bilayer, which is negative in neutral media, was deposited on a positive PEI capping layer of a polyelectrolyte cushion on a gold electrode. Lipids, which themselves are negatively charged in neutral media, strongly attached to underlaying support, become positive due to protonation of the phosphate group compensating the negative charge of the zwitterion, lowering electrostatic attraction. As a result, the separation distance between the LB and its underlaying support increases. Impedance spectroscopy measurements have demonstrated their effectiveness for supported LB characterization. When the lipid layer is deposited on a thin polymer cushion, the resulting composite demonstrates much more resistive behaviour. It was shown that the LB impact in impedance disappears with an increase in proton concentration.

Presented here is a new electrochemical approach for a free-standing lipid membrane fabrication. Owing to the proton-coupled oxidation of hydroquinone, the pH of the surrounding medium decreased to 3–4 from an initial 7. Reversible LB lift-off was observed during electrotriggered acidification of the surrounding medium and relaxation cycles whereas the polymer cushion was proven not to undergo any changes during hydroquinone oxidation and local acidification.

Data accessibility

The authors confirm that the data supporting the findings of this study are available within the results section of this paper and its electronic supplementary material.

Authors' contributions

N.V.R. carried out the laboratory work and data analysis, participated in the design of the study and drafted the manuscript; N.A.M. carried out the laboratory work; E.V.S. designed and coordinated the study, and helped draft the manuscript. All authors gave final approval for publication.

Competing interests

We have no competing interest.

Funding

E.V.S. acknowledges RSF grant no. 17-79-20186 and ITMO Fellowship Professorship Program for Infrastructural Support. N.V.R. thanks RFBR for support of electrochemical experiments according to the research project no. 18-38-00640.