Electropatterning—Contemporary developments for selective particle arrangements employing electrokinetics

3.1 |. Electrophoretic-based patterning of colloids

The formation of highly structured colloidal arrays using EP has garnered significant attention as a result of its high controllability, which is achieved by varying the size of the colloidal particles [58]. One of the most rapid and cost-effective EP-based techniques for creating dense and highly ordered colloidal assemblies is electrophoretic deposition (EPD) [59]. This technique refers to the controlled aggregation of particles moving under the influence of an EF [60]. In these systems, the aggregation rate depended on particle properties (zeta potential), experimental parameters (EF amplitude), and on the suspending media composition [61]. A recent study that employed EPD to obtain highly structured silver colloids was reported by Bi and Qi [59]. They used laser ablation on a silver plate that was submerged in pure water and observed the formation of well-defined sub-microscale silver peroxide (Ag₂O₂) cuboids on the anode surface. The Ag₂O₂ cuboid formation mechanism was attributed to two critical steps. First, the movement of the charged Ag ions to the cathode surface provoked an electrochemical oxidation reaction of Ag₂O₂ nanoparticles to occur. Second, the fusion of neighboring Ag₂O₂ nanoparticles created sub-microscale Ag₂O₂ particles through oxidization and oriented attachment. An image of the cuboids of Ag₂O₂ deposited on the electrode surface is shown in Figure 2A, which was obtained at a current density of 100 μA/cm². This study proposed a new growth mechanism using EPD that allowed for a constant deposition of metal oxide nanoparticles. However, this deposition was only achievable for 2D structures [59]. This limitation was addressed in a study by Muto et al. [58] where EPD was used for the formation of 2D and 3D metallic oxide films. Composite particles of SiO₂–SiO₂ with two distinct particle sizes of 200 nm and 16 μm were electrostatically assembled by employing a simple EP cell chamber where two conductive pieces of tape were attached to a glass slide as electrodes. The suspension of SiO₂ particles was introduced in the chamber, and by combining alternating current (AC) and direct current (DC) EFs, particle oscillations were induced. The use of AC fields proved to be essential as employing only a DC EF resulted in the formation of a disordered structure with voids. By fine-tuning the amplitude and frequency of the applied AC potential, the 3D hexagonal close-packed structure seen in Figure 2B was obtained at a potential of 30 V amplitude at 1 Hz. This is the first published work that reports the formation of 2D and 3D closely packed arrays of SiO₂ using DC and AC EFs. These studies illustrate that well-ordered colloidal arrays can be obtained by means of EK phenomena.

Another patterning study that used AC EFs to induce EP was reported by Fernández-Mateo et al. [57]. They observed that AC EFs had a predominant effect on the wall repulsion of charged colloidal particles in microfluidic channels. This work aimed to understand the effect of varying the frequency of the applied field in the 50 Hz to 10 kHz range on the repulsion of carboxylate particles. Figure 2C shows the superimposed images of 3 μm carboxylate particles at four distinct frequencies at 80 kV/m and a control image. As observed, the extent of the particle depletion region is a strong function of EF frequency, and the most significant wall repulsion was exhibited at lower frequencies. It was demonstrated that DEP, the mechanism believed to be responsible for this repulsion, was only valid at high frequencies. Their results provided a conceptual understanding that hydrodynamic wall repulsion is the major contributing force to wall repulsion of charged colloidal particles during EP induced by AC EFs. This work represented a new potential strategy for the design of microfluidic technologies that use EFs for particle manipulation and separation.

3.2 |. Dielectrophoretic-based patterning of colloids

Colloids have numerous drug delivery applications as customized carriers for loading and transporting specific targets using DEP phenomena [62, 66]. The frequency tunability in DEP phenomena enables the unique flexibility to vary the choice of target in situ. A notable advancement in active colloid target selection was reported by Boymelgreen et al., where dynamic particle control was achieved with pDEP/nDEP [62]. They unified the tasks of loading and transporting multiple nano- and micro-sized targets. The first advantage of their approach was combining the loading and transport capabilities as a unified singular task instead of loading a target by one mechanism and then driving the transport with another mechanism. The second advantage was enabling the on-demand selection of multiple micro- and nano-sized targets without the need for pre-labeling or surface functionalization. Demonstrations of target accumulation and repulsion under the effect of pDEP and nDEP, respectively, can be seen in Figure 2D when applied EFs range from 10 to 14 V. By fine-tuning the frequency of the applied EF, the configuration of the colloidal particles can be selectively loaded/unloaded and dynamically modulated. This study illustrated the potential of combining dynamic selectivity with direct motion to develop fabrication techniques such as additive manufacturing and soft robotics.

Target selection is not the only application DEP has in the field of colloid manipulation. The assembly, transport, and deposition of colloidal particles can also be achieved by employing DEP, as can be seen in the work done by Shen et al. [63]. They reported an electro-microfluidic assembly platform (eMAP) for constructing and manipulating colloidal assemblies within water droplets. The main phenomena in their study for patterning these waterin-oil emulsions were DEP-induced droplet positioning, dielectrowetting-induced droplet deformation, and DEP colloidal assembly. They observed that upon application of AC voltage ranging from 0 to 130 V_pp, DEP and dielectrowetting phenomena started to act on the droplets. The authors state that, under these experimental conditions, there is a strong nDEP force acting upon the particle chains, resulting in a slight alteration to the chain orientation. Though there is a slight effect of the chain–chain interactions resulting from this change in orientation, the previously described wing-like structures are mainly governed by DEP in this assembly. This resulted in particle motion that enabled DEP-based colloidal assembly. These phenomena were used to manipulate motion and active colloid assembly. Various superstructures within water droplets were observed, which were further optimized by fine-tuning the inner fluidic and EFs along with the colloid properties. Figure 2E illustrates a “spreading wing” like 3D structure, where clusters of the polystyrene particle chains in the water droplets showed a wing-like pattern. This wing-like formation occurred as a response to increasing EF at a fixed frequency of 300 Hz as the direction of the DEP force exerted on particles can be adjusted by modifying the frequency of the applied EF. Overall, four distinct 3D colloidal assemblies were illustrated—the bottom base, wing-like, rising-chain-like, and top-cap-like structures. This controllable strategy was further extended to demonstrate the formation of anisotropic micro-gels from binary colloidal assembly and electrically driven optofluidic display devices.

3.3 |. Hybrid systems employed for the patterning of colloids

Although EK techniques such as EP and DEP are able to form densely packed layers of colloidal particles, there is little control in the particle packing order within these layers, highlighting a major area of improvement [67–72]. The combination of multiple EK techniques within the same system has allowed higher control, resulting in the patterning of colloidal particles into advanced structures [51, 58, 64]. An investigation of the robustness of a coating process was done by Gong and Wu, where both AC and DC EFs were used for the manipulation of the colloidal arrays [64]. Their study showed that the application of low-frequency AC EFs induced dipolar repulsion among particles and resulted in an ordered, non-closed-packed colloidal array. Application of a simultaneous sequence of DC pulses combined with the low-frequency AC field–induced EP motion of the particles toward the substrate. An image of a 2D non-closed-packed colloidal arrays of 700 nm particles obtained with an AC EF of 10 V_pp at 1 kHz is shown in Figure 2F. This method is versatile, as it can also be used to create colloidal arrays with smaller particles, as long as the EF strength is strong enough to generate sufficient dipolar repulsion to overcome Brownian motion. The application of a sequence of DC pulses is also a commonly used technique to avoid solvent evaporation problems. This technique has the potential to be adopted in applications requiring a continuous convective method for scalable production of non-closed-packed 2D particle arrays.

As previously stated in Section 3.1, AC EFs have the added advantage of high-frequency tunability of particle pattern formation when compared to DC. This was further proved by the work of Goel et al., where the authors combined AC-dielectrophoresis (DEP) alongside AC-electrohydrodynamic (EHD) effects in a microwell for the manipulation of gold nanoparticles [51]. This work demonstrated, in addition to the dependence that AC-DEP and AC-EHD forces have on the amplitude of the EF, the first scales with both particle size and field frequency, whereas the latter depends only on the field frequency. The results indicated that the colloidal structures transition from films to microwires as the particle diameters increase from nanometers to micrometers, with no observable assembly at intermediate particle sizes. The main finding of this work was the identification of the processes that governed the film and microwire formations. Film formation was purely governed by EHD forces occurring at the material’s surface depositing submicron particles, whereas the microwire synthesis was a product of EHD-assisted-DEP forces present in larger particles. The two images in Figure 2G illustrate the size and frequency-dependent phase chart for film to microwire transition when an AC EF of 20 V_pp was applied, with maximum assembly obtained in the frequency range of 100–500 Hz. This study illustrated a simple and robust method for creating biosensing platforms with applications in chemical studies, nanoelectronics, and pathogen capture and analysis. Another study in which AC EFs were used for the formation of patterns in colloids was reported by Katzmeier et al. [65]. Here, a combination of EK flow and hydrodynamic interactions resulted in the formation of patterns in a wide range of colloidal particles. Figure 2H contains two example images illustrating both band formations and a zigzag pattern of coacervates and coffee colloids obtained by employing an AC EF of 17 mV/μm at a frequency of 500 Hz, and the image acquisition time being 100 s. This study provided a better understanding of the physical mechanisms behind the pattern forming process.

4.1 |. Electrophoretic patterning of droplets

Electrophoretic control of discrete droplets (EPDD) is a promising technique used for merging aqueous droplets through electrical interaction, also known as “electrocoalescence” [73]. In this technique, rapid charging of conductive droplets is used to achieve subsequent EP-induced motion. Recent studies have used EPDD for merging aqueous droplets and minimizing direct liquid-solid contact. A recent study by Kim et al. [74] reported the electrocoalescence and merging of droplets by employing a programmable digital microfluidics device. Their digital device consisted of two parallel electrode modules that enabled 3D programmable droplet handling. Figure 3A shows the horizontal and vertical transport of two microdroplets, electrocoalescence, and reciprocating droplet motions observed at 400 V applied between two pin electrodes separated by 2 mm. This study illustrates a versatile strategy for droplet manipulation that can enable biochemical and biomedical assessments that require the simultaneous handling of microdroplets.

Droplet electrocoalescence using EPDD was also employed by Peng et al. [73] to develop a digital microfluidic device powered by piezoelectric deflection caused by human finger movement. Their device had an array of piezoelectric elements, connected in parallel to metal electrodes, immersed in dielectric fluids. Finger motion caused deflection that was used to charge the piezoelectric elements and actuate the droplets across each electrode pair through EP force. This study demonstrated the electrocoalescence of aqueous droplets actuated at 200 V, as seen in Figure 3B, which shows the merging of a clear and a dyed droplet with and without a sustained electrode bias. Enhanced mixing was achieved by inducing internal flows in the droplet through continuous transportation of the droplet back and forth between the electrodes. In addition to droplet manipulation and mixing, EPDD can also be used for droplet transport. An important challenge in many microfluidic biomedical applications lies in the usage of bodily fluids that possess unpredictable and challenging sample characteristics such as high viscosities and extreme pH values. A potential solution to this limitation lies in the device by Peng et al. [73] that demonstrated the transport of both saliva and urine droplets. Figure 3C illustrates the transport of droplets of human urine (pH = 5.8) and saliva (pH = 6.4) by employing an actuation voltage of 200 V. This work provides an example of a portable, piezoelectrically powered EPDD device with potential applications in medical diagnostics.

4.2 |. Dielectrophoretic patterning of droplets

Microfluidic systems are often prone to undesirable air bubbles being nucleated inside microchannels or inadvertently introduced when adding tubing connections to the devices [76, 79]. Air bubbles are highly undesirable, as they can affect device operation and generate parasitic heat transfer processes. Considerable efforts have been focused on developing DEP-based methods for the removal and subsequent transport of bubbles from surfaces. One of the most popular DEP-based techniques for this purpose is dielectrowetting or liquid dielectrophoresis (L-DEP), where the wetting behavior of a liquid on a solid surface is reversibly modified by DEP generated with a nonuniform EF. In a recent study, Vo and Tran [77] used DEP to deform the liquid surface when a droplet is sufficiently close to a solid surface. This process effectively induced controlled ruptures to the air film and created escaping tunnels for air. Entrapped air bubbles formed on the surface at applied voltages ranging from 0 to 400 V, as seen in Figure 3D. The entrapment of bubbles was observed to be manipulated by varying the applied voltage. At higher voltages, a large bubble (radius ~45 μm) experiences splitting of the air film and forms noticeably smaller bubbles along the electrode gaps. Bubble detachment, levitation, and reattachment using L-DEP were also demonstrated in the work by Brown et al. [76]. Successful bubble detachment and bubble levitation controlled by voltages ranging from 0 to 270 V are shown in Figure 3E. Captive bubbles were held indefinitely away from a solid surface with a programmable separation distance. These studies demonstrate that the effectiveness of DEP has for precise bubble elimination in microfluidic devices.

More recently, another interesting DEP-based actuation and manipulation of droplets to control their wetting behavior were reported by Frozanpoor et al. [75]. They developed a droplet-actuating platform by exploiting high EFs at the droplet solid–liquid interface using L-DEP. The actuation of deionized (DI) water droplets with different volumes (2.5–9.0 μL) at an actuation voltage of 85 V at 40 kHz using the iterative DEP approach are shown in Figure 3F. The main novelty of this work is the development of a droplet-actuating platform to control the wetting behavior of droplets and the properties at the droplet solid–liquid interface by introducing a lubricant layer and by controlling the L-DEP force. The actuation behavior was directed toward a cleaning application, resembling a scrubbing motion, while also transporting the droplets for disposal. This study offers an example of an energyefficient droplet actuating platform to clean a variety of surfaces, a process with potential in the cleaning processes of automobile parts such as sensors, cameras, door mirrors, front side glass, and rear screen. It is expected that the next decade will bring programmable droplet actuating platforms for tackling cleaning issues, especially surface contamination in automobiles [80–82].

5.1 |. Electrophoretic-based patterning of synthetic particles

EPD is a commonly used technique for the patterning of synthetic particles, as it offers superior control of the thickness, stoichiometry, and microstructure of the deposited layers than more traditionally approaches, such as hydrodynamic and magnetic forces [61]. This level of control is attained by adjusting both the magnitude of the applied voltage and treatment time [89]. Despite being a little over a century old, EPD is still a dynamic field, and exciting new developments are being made. In 2018, Mora et al. [90] reported a novel improvement called light-directed EPD (LDEPD). This technique involves the use of ultraviolet (UV) light that enables a localized control of an applied EF by using photo conductive electrodes [90]. In their work, the authors performed a three-step deposition of 70 nm tungsten, 300 nm alumina, and 150–190 nm polystyrene particles in three different patterns without using any sort of physical mask. Instead, a cell formed by titanium oxide and indium–titanium oxide (ITO) glass electrodes was energized with a +2 V DC bias voltage and UV light at a 100 mm distance from the emitting source for 90 s. This allowed for the patterning of the particles into simple geometrical shapes (a square, triangle, and circle) without requiring any hard tooling. Images of the patterns formed with this technique are shown in Figure 4A, where particle arrangement of into three distinct geometrical formations is observed. Although LDEPD can fully be automated, the patterning deposition was only possible when done at different times due to how the materials interact with the UV light. This limitation makes the simultaneous formation of distinct patterns a challenge to address in future LDEPD studies.

A possible solution to asynchronous treatment limitation in EPD was given in 2021 by Sopubekova et al. [91] with an EPD methodology capable of binary and successive patterning of silver, latex, and silica microparticles. Using chromium–gold electrodes fabricated by photolithography, the authors were able to deposit all three kinds of particles into the patterned structures using different voltages and treatment times. Silver particles, with sizes ranging from 80 to 100 nm using a 20 V DC bias voltage for 10 min, result in a deposition with 90% efficiency. Latex particles of 30 μm in diameter were also pattered alongside the electrode’s structure with the same efficiency by applying a 15 V DC bias for 5 min. This decrease in the treatment conditions is attributed to the fact that latex particles presented a larger particle zeta potential magnitude (~70 mV) compared to silver particles (~60 mV). This is consistent with Equation (1); as latex particles have a greater zeta potential, they also possess greater υ_EP, meaning that they are able to cover the same distance than the silver particles in a shorter time. Silica particles of 30-μm diameter were also treated with a 20 V DC bias for 5 min, resulting in the same deposition efficiency. A property unique to the silica particles was their noticeable tendency to deposit at the edges of the electrodes rather than at the center. This tendency is attributed to fringing field effects on the electrode edges. This unique property also allowed generating of silica/latex particle composites, with the former being deposited alongside the electrode edges and the latter occupying the middle regions. Successive patterning was done by submerging the electrode in an electrolyte solution containing silica particles, applying 15 V DC bias for 5 min, and then replacing the electrolyte solution to one containing latex particles without interrupting the voltage bias. This successive patterning of two different kinds of particles is shown in Figure 4B. The authors explicitly state that the methodology being used is applicable to any kind of conductive electrode surface and any kind of material as long as it is electrically charged. The authors also stated that patterns of any shape can be formed and that the resolution is directly dependent on the dimensions of the electrode [91].

5.2 |. Dielectrophoretic-based patterning of synthetic particles

Particle patterning and arrangement by rapid and simple manipulation have been demonstrated extensively using DEP [46, 95, 96]. Due to the unique ability of DEP to exert a force on both uncharged and charged particles, it has numerous applications in a wide array of fields, such as manipulation and assembly of micro components, the self-assembly of electronic devices, and, relevant to this review, microfluidic devices [97–100]. DEP-based manipulation has been extensively employed in the patterning, concentration, and separation of bioparticles [101, 102]. Although it was demonstrated that DEP can achieve higher order particle organization in porous membranes, there is still a research gap where particle arrangement at specific position on the porous membrane was not possible [84]. As a potential solution to address this limitation, Yasukawa et al. [84] demonstrated the positioning and patterning of polystyrene particles at a localized area on a silicon nitride (SiN) membrane with an array of 2-μm microholes by pDEP. As seen in Equations (3) and (4), when the f_CM has a positive value, the resulting V_DEP leads to pDEP. When an AC voltage was applied to a needle-type microelectrode, strong EFs were formed at the microholes of the SiN membrane in the region directly under the microelectrode, resulting in the particles moving toward the lower surface of the region. The movement of the microelectrode at 10 μm/s obtained with AC voltage resulted in microparticles patterning into a thick line formation along the trail of microelectrode movement. This was attributed to the shift of microholes by the strong EF, as seen in Figure 4C, which shows a line pattern of particles formed on the lower surface of the membrane at 400 V_pp with microelectrode motion at 10 μm/s. This method enables precise arranging particles at a desired position without requiring any templates or special electrode configurations. Through the unique capability to position particles at any location on membranes, this study provides a promising methodology for novel cocultivation systems and sensing platforms.

Another DEP-based study performed on latex beads was reported by Kale et al. [92, 103] where particle manipulation was enabled by EF gradients generated at the reservoir-microchannel junction. Their approach exploited the size difference between the inlet liquid reservoir and a microchannel, thereby eliminating the microfabrication of mechanical or electrical parts within the microchannel. Using latex beads as a proof of concept, they showed that a 3D reservoir–microchannel junction is crucial in enhancing the dielectrophoretic preconcentration potential of the devices. Their system, called reservoir-based dielectrophoresis (rDEP), resulted in a six-fold increase in the enrichment of latex beads and a significant reduction in the power consumption, which allows for a higher device integration [103]. Successful particle focusing was observed with an increase in AC voltage, and the beads were deflected to the channel centerline, as observed in Figure 4D obtained by applying a three voltage setting of 25 V DC, 25 V DC/150 V AC, and 25 V DC/300 V AC. Their work demonstrated that Joule heating effects can reduce rDEP particle focusing and could also potentially disable rDEP particle trapping. The authors attributed this effect to fluid temperature rise, which was indicated by electric current increase in the channel, which impacted particle velocity within the system. These findings provide means for integrating particle enrichment with downstream processes, for example, a detection system or a cell culture chamber in which the enriched target species would already be in a sedimented state to facilitate culturing.

5.3 |. Hybrid systems employed for the patterning of particles

Though EP and DEP have been extensively used in particle patterning, deposition, and sorting, there are some drawbacks in employing these techniques such as the requirement of complex electrode designs and limited flexibility [104]. However, efficient particle patterning has been achieved by hybrid approaches that combine active external force fields with passive methods that depend on the microchannel structure and intrinsic hydrodynamic forces [93]. Hybrid systems that combine, for example, DEP with inertial forces [93, 94, 104], DEP with optically induced effects [104], and linear and non-linear EK [105] have shown promising results for electropatterning of micro/nanoparticles.

A major advantage of hybrid systems is their ability to tune and adapt to different particle samples once the microchannel structure and dimensions are fixed. Zhang et al. [93] demonstrated this capability using an innovative hybrid DEP-inertial microfluidic platform with three different sizes (4.8, 8, and 13 μm) of fluorescently-labeled polystyrene particles. The microfluidic device featured a sheath liquid flow to generate inertial forces, which combined with the DEP force induced by the bottom microelectrodes, resulted in effective microparticle focusing within the channel cross section. A unique aspect of this platform is the capability for real-time tuning of the dimensions of the target binary particle mixture by adjustments to the electric voltage, without having redesigned the channel structure. Figure 4E illustrates the particle focusing streaks of 13 μm polystyrene particles obtained at an AC EF of 30 V_pp and frequencies ranging from 0.1 to 30 MHz. Their findings provided a superior alternative over other traditional inertial-based microfluidic platforms. Due to the high dependency on the sheath flow speed, the traditional microfluidic platforms can only be effective for specific range of target particles. However, when combined with DEP, as shown in this study, the DEP-inertial platform can provide real-time target tunability by simply adjusting the electric voltage. Another study that used the combination of EP and inertial forces was performed by Li et al. [94] with the development of a new cell sorting and separation method that combined 3D sidewall electrodes and contraction/expansion (CEA) structures. This system was used for the continuous flow separation of three types of particles: 4 μm polystyrene particles, 20 μm polystyrene particles, and 4 μm silica particles. The strategic interaction between electrophoretic forces and inertia forces enabled the separation. Due to inertial forces, sorting particles with different sizes was achieved in the CEA channel, as demonstrated in Figure 4F where it can be observed that the separation of 4 and 20 μm polystyrene particles was achieved at 14 μL/min at a potential of 5 V. The nonuniform EF generated in the 3D electrodes at the same height as the channels increased the action range of the DEP force. Their novel method of combining DEP and inertial forces achieved high precision sorting and separation of particles and cells.

A drawback of DEP technology is its limited flexibility in practical applications, mainly attributed to the requirement of complex electrode designs in traditional DEP systems [93]. An alternative is optically induced DEP (ODEP) [104], which uses virtual electrodes formed by illuminating optical patterns onto photoconductive layers with DEP to carry out micro/nano-object manipulation. The complex electrode manufacturing process is simplified by replacing the physical electrodes with virtual ones. This approach led to a shorter and simpler chip design cycle and lower processing costs. ODEP has been actively used to assemble and pattern polystyrene beads in recent years. Shi et al. [104] reported a model that solved an ODEP-based transient numerical model that simulated microparticles under AC EF coupling with an open flow field. The numerical results showed that particles with different conductivity could be separated with pDEP and nDEP effects caused by light illumination. The experimental results showed that the two DEP actions differentiated the conductive particle trajectories, enabling particle separation and patterning. Parameters such as the width of the bright area, applied AC voltage, and inlet flow velocity were optimized for efficient separation performance. Their study explained the mechanism behind the continuous particle manipulation was the combined action of AC EF and flow field. These findings unveiled the theoretical support behind the principles of particles continuous patterning and separation, opening the possibility for future device designs based on AC-ODEP technology.

6.1 |. Electrophoretic-based patterning of cells

The superior degree of control that EPD offers over more traditional deposition methodologies makes its applications to biological materials, such as cells, highly adaptable. A recent study by Brisson and Tilton [112] presented a method capable of the selective patterning of bacterial cells in a 2D arrangement. For these experiments, Saccharomyces cerevisiae cells were deposited onto a metallic electrode surface comprised of sodium bicarbonate (NaHCO₃). A cell concentration of 0.07% (w/v) was empirically determined to be below the number needed to form a complete monolayer on the surface. As such, any cell aggregation would be considered a direct result of the EF applications. The results seemed to indicate that there was no direct relation between the field frequency and the aggregate density; however, aggregate density was determined to be strongly dependent on the strength of the EF. Stronger AC voltages (>50 V) resulted in what the authors referred to as densely packed arrangements. On the other hand, lower intensity AC voltages (<50 V) resulted in loosely packed arrangements. An example of these “densely” and “loosely” packaged arrangements obtained using 10 Hz, 100 V and 10 V, respectively, is shown in Figure 5A,B. This difference in packaging arrangement is solely attributed to the increased intensity of the EF. As can be seen in Equation (1), the electrophoretic velocity of particles is directly proportional to the EF magnitude. An increase in electrophoretic velocity would explain why cells would form denser arrangements under the same frequency and application time, as they would be moving faster at 100 V when compared to 10 V. The patterned arrays in this work were obtained without any chemical modification of the target surface, which was a limiting factor in the previous EPD research related to cells.

Yet another interesting development in the field of EPD with cells was reported by Neirinck et al. [113], where both living Staphylococcus aureus and Escherichia coli were deposited on a stainless steel surface. The purpose of this work was to measure cell viability of two industry-relevant bacterial cells in the context of continuously stirred tank reactors. The authors aimed to prove that EPD was an effective method to generate biofilms while maintaining cell viability within an acceptable margin. Deposition was done using two suspending solutions, a sucrose suspension and demineralized water, with the latter obtaining the most promising results. The deposition of both S. aureus and E. coli after 2400 s of applying 100V_pp at 25 Hz is shown in Figure 5C,D. As can be seen, the vast majority of cells remain alive after the treatment with the EF, suggesting that the treatment conditions have no significant impact on cell viability. It is also noticeable that S. aureus had better deposition than E. coli; the authors attribute this to the difference in conductivity between the two cell lines. S. aureus presented a conductivity of ~50 μS/cm, whereas E. coli only reached ~38 μS/cm. Regardless, the experiments appeared to have yielded positive results and both bacterial cell populations could be deposited into metallic surfaces without significantly affecting their viability.

6.2 |. Dielectrophoretic-based patterning of cells

Over the last few decades, DEP has garnered considerable interest in biological analysis. This interest is mainly due to its relatively short response time, precise and easy operation, and low degree of cell damage [47]. Another major advantage DEP presents as a methodology for studying cellular systems is that it is label-free, which trivializes the difficulty of finding an assay marker that does not interact with the model organism [116, 117]. The differing EK properties between the membranes of different cells allow DEP to be used as a highly tunable technique for the manipulation of bioparticles. Examples of this manipulation include the selective trapping of cells according to their species [118], the structured arrangement of different kinds of microbial cells to a surface [119, 120], the controlled formation of cell aggregates with defined size and shape [121, 122], among others.

Despite its attractive characteristics, the spatially controlled assembly of live bacterial cells employing DEP remains mostly unexplored. Due to this apparent gap in knowledge, Goel et al. [114] developed a methodology capable of patterning two different bacterial cells. In this research work, the authors demonstrated the dynamic patterning of two live, pathogenic Gram-negative bacteria, Salmonella typhi and E. coli, respectively. This work aimed to collect bacterial cell inside microwells to discern among different types of bacteria within two-population mixtures. Cell suspensions were subjected to varying nonuniform AC EFs applied through microelectrode arrays resulting in the slow collection of bacteria in the microwells. The process was fully reversible; the electrodes could be restored to their initial state within minutes of turning off the EF. Hence, each device was expected to withstand at least 50 trials of these kinds of experiments. The EF application time was empirically selected to be 15 min, as this allowed a high enough number of cells to draw sensible inferences on the results without clogging the device. The results yielded by these experiments showed that for assembly to occur, 10 V_pp and 50 kHz were the minimum requirements for cells to assemble. Lower frequencies and V_pp values resulted in the cells not showing any effect to the EF application. The best cell assembly results were obtained at a voltage of 20 V_pp and 5 MHz, experimental images of the collection of S. typhi cells under these conditions are shown in Figure 5E,F. The highest frequency at which cells experienced pDEP was determined to be 20 MHz, which lies above the experimental system limits of the described experimental setup. The overall reported cell collection efficiency for a bulk concentration of 10⁵ cells/mL was estimated to be around 90%. Of the collected cells, 96% were viable at the end of the 60 min experimental process [114].

It is worth noting that most bacterial cells are more resistant to external stresses than eukaryotic cells [123]. Thus, the assumption that a particular methodology would yield the same degree of viability using the two different types of cells is incorrect. To this end, Henslee et al. [115] used DEP-capable dot electrodes to form 3D cellular structures, with promising results using yeast, HL-1, K562, and HeLa cells. This work aimed to answer two questions: (1) Does DEP manipulation has a negative impact on cell viability? (2) Is there a significant difference in cell viability between 2D and 3D cell structures? To measure cell viability, the cell aggregates were subjected to two distinct drugs: Vinblastine and Amphotericin B. Dot electrodes were used to contain cells via nDEP, which caused repulsion from the edges of the electrode patterns and forced cells into aggregates. The resulting cell aggregates, obtained at 10 V_pp at 10 kHz, can be observed in Figure 5G–I. Given that the cell lines used in this work were of similar size (~10 μm in radius), the resulting aggregates also presented similarities in size. The authors pointed to these finding as evidence of successfully producing homogenous aggregates. To properly verify the effect of DEP-based 3D cell culture on cells, the authors investigated the properties of the studied cells after DEP treatment. The electrophysiological properties of cells grown in 2D were determined for 24 and 48 h and were compared to those produced in 3D for the same time periods. Data analysis revealed a significant change in the membrane capacitance for HeLa cells after 3D encapsulation compared to their 2D counterparts. Membrane conductance for yeast and K562 cells also demonstrated a significant change, which was assumed to be due to the 3D enclosing. It is worth mentioning that in the case of yeast cells, the changes in membrane conductance were only observable at 24 h. This fact seemingly implies that at least for that cell line, changes in electrophysiological properties were only temporary.

7.1 |. Electrophoretic-based patterning of gels

Since the early 2000s, there has been an increased interest in developing 2D surfaces with molecular arrangements by photolithography and microfluidic approaches [130–132]. However, their functionality and effectiveness are severely limited by the inability to create an accurate representation of the 3D natural cellular environment [133–135]. Although a vast number of fabrication methodologies exist for 3D gel environments, most possess notable disadvantages hindering the possibility of widespread usage. These disadvantages include the use of UV radiation [136], limitations in the capacity to pattern large or complex molecules [137], and expensive processes or chemical reactions [138]. In recent years, the usage of EFs to achieve 3D patterning of gels has increased, as these simple and affordable systems allow the localization of molecules within the gel network [139–141].

A recent example of the formation of 3D gel environments lies in the work by Aguilar et al. [142], where they reported the combination of EP with hydrogel photopolymerization for the patterning of NIH-3T3 cells. They developed a platform called “3D electrophoresis-assisted lithography” (3DEAL), which created molecular patterns of different sizes, geometry, and composition. This platform used EP to manipulate fibronectin (FN) molecules independently of charge and size and localize them within the hydrogel. The described system is composed of five elements: power supply, platinum electrodes, buffer reservoirs, hydrogel compartment, and tubing connections. This process used EP to drag the “printing” molecules along the direction of the EF, which generated both the size and shape of the desired pattern within the gel network. Given the capability of the 3DEAL system to pattern multiple types of hydrogels with numerous types of proteins, it offers an alternative to generate hydrogels with anisotropic chemical compositions that can be selectively recognized by cells. To evaluate this possibility, a polyacrylamide–collagen (PA–Colla) gel with circular columns of FN that were 150 μm wide in diameter and 800 μm in depth was patterned. Immunofluorescent microscopy images of the resulting pattern obtained at 150 V are shown in Figure 6A. Although the 3DEAL methodology generated patterns that could theoretically allow cells to migrate up to 800 μm in depth into the hydrogel network, cells did not penetrate beyond ~100 μm. The authors theorized that this limitation in movement can be attributed to the low porosity of the PA–Colla gel, which may affect cell and oxygen diffusion. These results seemingly demonstrated the capacity of the 3DEAL system for the generation of hydrogel environments with functional macromolecules. This platform serves as an example of a simple and accessible method to create anisotropic environments using hydrogel pattern generation.

Another recent EP-based study by Gargava et al. [143] demonstrated the use of EFs to rapidly form alginate (Alg) gels in specific 3D patterns. Gel synthesis starts with an agarose, a thermoresponsive biopolymer, which was previously loaded with Ca²⁺ ions (0.1 wt% of CaCl₂) and placed in a beaker. The inner surface of the beaker was wrapped with aluminum foil, which acted as a cathode, and copper wire, which served as an anode. When a potential of ~10 V was applied with a DC power source, an Alg gel formed, having the shape of the mold. Gelation occurred due to the Ca²⁺ ions electrophoretically migrating away from the mold, at which point they crosslinked with the Alg chains. An overview of this methodology and a fluorescence micrograph of the obtained polymer are shown in Figure 6B. The main advantage of this technique was that it provided a framework for manufacturing 3D gel networks without using any additional instruments.

7.2 |. Dielectrophoretic-based patterning of gels

As DEP does not depend on electrical charge to achieve the manipulation of molecules or particles, it is the technique of choice when handling uncharged materials. An example of these important dielectric materials is polyethylene glycol (PEG). PEG-based hydrogels show high antifouling properties, mainly attributed to their neutral charge and absence of hydrogen bond donors [144]. These attractive characteristics have enabled the use of PEG-based hydrogels sensors in clinical trials [145–147]. Kim et al. [148] spatially immobilized beta amyloids 1–40 and 1–42 (Aβ_1–40 and Aβ_1–42) within a PEG gel network to create a sensor to diagnose Alzheimer’s disease. Their work aimed to demonstrate the feasibility of plasma-based Alzheimer’s diagnosis by combining the properties of PEG-based hydrogels with DEP forces. The sensor was based on a spiral microelectrode (SME) structure covered with a hydrogel with a specific pore size and thickness, onto which fragmented antibodies were immobilized. A visual representation of the immobilized Aβ_1–40 and Aβ_1–42 anti-bodies can be seen in Figure 6C. Both antibodies are mostly confined to the bounds set by the hydrogel network and concentrated to the bottom of the gel, where the sensor is located. The switching between pDEP and nDEP forces is achievable by a change in the relative polarizability, which is denoted as a change in the sign of the f_CM in Equation (4). This variation in the sign of the f_CM is the result of varying the applied field frequency. The frequency conditions for pDEP and nDEP were 1 kHz and 100 MHz, respectively. The dynamic capability of switching between pDEP and nDEP in tandem further concentrated the amyloids and removed the nonspecific plasma proteins, improving the density of immobilized fragmented antibodies within the sensor region. The SME sensor could quantify disease-related proteins of known sizes with high sensitivity, providing high utility for diagnosing multiple diseases besides Alzheimer’s at its earlier stages. This SME sensor could represent a significant leap in plasma-based disease diagnosis with high accessibility and affordability.