Microfluidic Electrochemical Desalination Systems: A Review

Abstract

Microfluidic techniques have emerged as promising, efficient, cost-effective, and environmentally friendly desalination solutions. By utilizing fluid dynamics at the microscale, these techniques offer precise control over chemical, biological, and physical processes, presenting advantages such as reduced energy consumption, miniaturization, portability, and enhanced process control. A significant challenge in scaling microfluidic desalination for macro applications is the disparity in flow rates. Current devices operate at microliters per minute, while practical applications require liters daily. Solutions involve integrating multiple units on a single chip and developing stackable chip designs. Innovative designs, such as 3D microfluidic chips, have shown promise in enhancing scalability. Fouling, particularly in seawater environments, presents another major challenge. Addressing fouling through advanced materials, including graphene and nanomaterials, is critical to improving the efficiency and longevity of devices. Advances in microfluidic device fabrication, such as photo-patterned hydrogel membranes and 3D printing, have increased device complexity and affordability. Hybrid fabrication approaches could further enhance membrane quality and efficiency. Energy consumption remains a concern, necessitating research into more energy-efficient designs and integration with renewable energy sources. This paper explores various electrochemical-based microfluidic desalination methods, including dialysis/electrodialysis, capacitive deionization (CDI)/electrochemical capacitive deionization (ECDI), ion concentration polarization (ICP), and electrochemical desalination (ECD).

1. Introduction

Population growth and climate change have led to an ever-increasing demand for clean water, placing immense pressure on water resources. Most water on Earth is seawater, which needs to be desalinated before it can be used as drinking water or industrially [1,2]. This growing demand for clean water is accelerating the development of desalination technologies as a viable solution to the global water scarcity problem [3]. Existing desalination technologies can be classified into thermal-based and membrane-based processes [4,5]. Thermal-based methods have been the backbone of the desalination industry for decades, characterized by their robustness and scalability; however, they are considered energy-intensive processes. On the other hand, membrane-based processes operate by applying a selective synthetic membrane across which one or more constituent transports is under the influence of a certain driving force. Such driving forces can be pressure gradients (e.g., reverse osmosis (RO)), temperature gradients (e.g., membrane desalination (MD)), chemical potential gradients (e.g., pervaporation), or electric potential gradients [6,7]. However, membrane-based technologies are limited by membrane fouling and require high-quality feed water to improve operational efficiency [8].

In recent years, electrochemical desalination techniques have gained significant attention as promising alternatives to conventional methods, offering potential advantages in energy efficiency, selective ion removal, and modularity. An important advantage of electrochemical methods is their ability to overcome the flow resistance limitations associated with pressure-driven systems, such as those governed by Hagen–Poiseuille flow. Microfluidic techniques have emerged as a promising technology for more efficient, cost-effective, and environmentally friendly solutions in specific or small-scale contexts. These techniques utilize the movement of fluids at the microscale to precisely control chemical, biological, and physical processes [9]. While they are not positioned to replace large-scale technologies such as RO, microfluidic devices offer distinct advantages in research settings and for specialized applications, including reduced energy consumption, miniaturization, portability, and enhanced process control [10].

Embedded membrane microfluidic devices integrate small membrane sections within the microfluidic chips, enabling precise control over the mass transport, real-time observation of the filtration process, enhanced ion separation, and improved antifouling characteristics. In addition, their compact design enables the development of portable desalination units with lower water and energy requirements [10,11,12]. Furthermore, they have been employed in studies on particle sorting, membrane fouling, and flow dynamics in membrane microreactors [10,11]. When electrochemical desalination processes are implemented within microfluidic systems, they enable fine control over electrochemical processes and ion transport, making them valuable platforms for investigating desalination at the microscale. Their primary value lies in allowing better mechanistic understanding and supporting the development of lab-on-chip systems tailored for niche, low-volume, or portable purification applications.

It is also important to recognize that many of the fundamental limitations in water purification, such as thermodynamic barriers (e.g., osmotic pressure, evaporation and condensation enthalpies, and electrolysis energy), long-term material stability, and fouling, remain difficult to overcome regardless of the scale or novelty of the approach. Microfluidic systems are no exception. Although often praised for enhanced control and miniaturization, they still face operational stability issues. In addition, many of the material compositions explored in microfluidic desalination are not yet stable for long-term use in aqueous environments, and their manufacturing cost remains high.

This review explores electrochemical-based microfluidic desalination techniques, examining their underlying principles, technological implementations, materials, recent advancements, and future challenges. It begins with the fundamentals of microfluidics, then presents an overview of the most widely used electrochemical techniques and their microfluidic applications. The next section discusses desalination materials, focusing on membranes, their synthesis techniques, and methods to enhance their selectivity, permeability, and long-term stability. While several prior reviews have examined desalination mechanisms or materials, this work offers a unique perspective that bridges microfluidic and electrochemical techniques and contextualizes recent innovations within broader water treatment goals. In doing so, it contributes a system-level understanding that complements and extends the scope of the existing literature. Finally, the review addresses key challenges.

2. Fundamentals of Microfluidics

Microfluidic desalination refers to using microscale fluidic channels and devices to remove salt and impurities from water. These systems take advantage of the unique fluid dynamics at the micron scale to improve separation efficiency, reduce energy consumption, and enable precise control over desalination processes. Unlike traditional desalination technologies, microfluidic approaches offer advantages such as lower operating costs, faster processing times, and enhanced scalability for portable applications [11,13].

Microfluidics involve the manipulation of fluids on the microscale where viscous forces dominate over inertial force effects. This has profound implications for heat and mass transfer processes. Transport physical properties such as surface tension, viscosity, and fluid resistance become more significant on the microscale level [9,14]. The laminar flow nature of fluids in microfluidic channels, and consequently, the dominance of viscous forces as compared to inertial forces, makes the mass transport mechanism in these channels easily characterized by diffusion with highly predictable kinetics and less intensive mathematical modelling [15,16]. This enables precise control and manipulation of fluid streams, which is particularly beneficial for water desalination processes, where controlled mixing can improve membrane-based separation or facilitate selective ionic transport [15,17]

Surface tension gives rise to capillarity and wetting phenomena, impacting fluid behavior, droplet formation, and fluid transport in microfluidic systems [18]. In microfluidic desalination, surface tension and capillary effects can be exploited to design microfluidic devices that transport water through microchannels without external pumps, thereby reducing energy consumption. Capillary-driven flow benefits portable desalination devices with limited power resources [18,19]. Fluidic resistance is another crucial kinetic factor in the design of microfluidic systems since it determines the pressure drop required to drive fluid through microchannels, which, in turn, affects the overall energy efficiency of the device [20]. In desalination applications, minimizing fluidic resistance while maintaining effective salt removal is a key design challenge [20]. This often involves optimizing channel geometry, length, and cross-sectional area to balance flow resistance with process performance.

The design and fabrication of microfluidic devices are crucial for their functionality. These devices consist of microchannels with dimensions that range from several tens to hundreds of microns, chambers, and valves capable of controlling fluid flow. Figure 1 provides a conceptual illustration of fluid separation within a microfluidic desalination device, highlighting the separation of feedwater into distinct outlet streams. However, the actual desalination process may involve additional mechanisms not depicted in this schematic. Material choice is a critical factor; microfluidic devices are typically made from silicon, quartz, glass, or polymers, with polymers becoming increasingly popular due to their cost-effectiveness and ease of manufacture. Polydimethylsiloxane (PDMS) has become a popular polymeric material in this regard due to its transparency, flexibility, and biocompatibility [21,22].

Various techniques are available for the fabrication of microfluidic devices. The choice of fabrication method depends on the material from which the device will be made, the end product requirements, accessibility, scalability, and the cost, especially since microfluidic devices are challenging to clean [22]. Fabrication techniques can be classified in several ways, but generally, they fall into the following categories:

• Direct manufacturing: this approach can be further classified into mechanical methods that use mechanical forces to remove excess material to shape the device and energy-assisted methods that use beams of energy (e.g., lasers, electron beams, or focused ion beams) to either add or remove layers directly, often guided by masks to create precise micro- or nanostructures [24].
• Replication manufacturing involves the creation of a master mold that features the desired micro- or nano-structures [25]. This mold is then replicated to produce multiple copies of the structure through processes such as casting, hot embossing, or micro-injection molding [21,22,26,27].

Table 1. Classification of the direct microfluidic device fabrication techniques (adapted from [21,22,26,27]).

Technique	Mechanical	Chemical	Energy-Assisted
Material removing methods	Micro-milling/grinding Ultrasonic machining Micro-abrasive air/water-jet machining	Dry/wet etching	Photothermal process Laser direct machining Focused ion beam
Material depositing methods	Injection molding Hot embossing	Lithography Inkjet 3D printing Direct writing	Selective laser sintering Stereolithography

summarizes the classification of the direct microfluidic device fabrication techniques.

In water desalination, microfluidic desalination approaches stem from techniques such as CDI and ED. CDI utilizes electrodes to adsorb and desorb ions from water as they flow through a microchannel, offering a low-energy alternative to conventional methods. Microfluidic desalination provides precise control over fluid flow, and the high area-to-volume ratio at the microscale level enhances mass and heat transfer efficiencies, making desalination processes potentially more effective and less energy-intensive. However, scaling up microfluidic desalination to meet industrial water demands presents significant challenges. These include fabricating large-scale microfluidic systems that maintain efficiency and cost-effectiveness, and managing the fouling of membranes and channels, which can reduce performance over time [15,28].

3. Overview of Dialysis and Electrodialysis

Dialysis is a selective separation process that utilizes a membrane with pores of specific sizes, defined by the molecular weight cut-off (MWCO). The membrane separates two streams: the influent stream, which contains the molecules or ions to be separated, and the receiving stream on the opposite side, which has a low or zero concentration of these constituents. Molecules larger than the membrane’s MWCO are retained on the influent side, while smaller molecules and ions diffuse across the membrane from the side of higher concentration (influent stream) to the side of lower concentration (receiving stream), driven by spontaneous diffusion (Figure 2).

Typically, influent and receiving streams flow in a counter-current arrangement. This setup maximizes the concentration gradient by continuously exposing the most concentrated portion of the influent stream to the least concentrated portion of the receiving stream, enhancing separation efficiency [28,29]. Unlike dialysis, which relies on a concentration gradient, ED uses an electric field to selectively remove or concentrate charged species (ions) from a solution. In an ED system, alternating cation-exchange membranes (CEMs) and anion-exchange membranes (AEMs) are arranged between two electrodes (anode and cathode), dividing the system into alternating dilute and concentrated compartments. The dilute compartments receive the feed solution, while the concentrate compartments collect the ions removed from the dilute compartments (Figure 3) [30].

When an electric potential is applied, electrochemical reactions occur at the electrodes. At the negative electrode (i.e., cathode), a reduction reaction reduces water to hydrogen gas and hydroxide ions. This generates a negative charge in the nearby concentrate compartment. To maintain charge neutrality, cations migrate from the dilute compartments through the CEMs toward the cathode, accumulating in the concentrate compartment. The CEM near the cathode blocks the hydroxide ions from entering the dilute compartments [30,31]. Under sufficiently applied potential, water undergoes an anode electrode reaction that produces oxygen gas, protons, and electrons (2H₂O = O₂ + 4H⁺ + 4e⁻); while the electrons travel through the external circuit, protons create a positive charge in the nearby concentrate compartment, causing anions to migrate from the dilute compartments through AEMs toward the anode. The AEM near the anode prevents the protons from moving into the dilute compartments. This process separates cations and anions from the feed solution, concentrating them in the concentrate compartments and producing a diluted stream with a reduced ion concentration [30,31].

Microfluidic Dialysis and Electrodialysis

Microfluidic applications are not only relevant to water desalination but also important in other applications. For example, integrating microdialysis with electrospray ionization mass spectrometry (ESI-MS) systems facilitates sample preparation. The analysis of ESI-MS spectra, particularly from protein or DNA samples laden with high concentrations of buffers and salts, is often hindered by ion suppression, leading to poor signal-to-noise ratios [30].

To address these challenges, advancements in miniaturized dialysis focus on developing membrane fabrication techniques directly on chips and enhancing process speed to support real-time desalination alongside analytical procedures. Advances in microfluidic technology have significantly impacted the development of miniaturized dialysis systems by enabling innovative on-chip membrane fabrication methods, which enhance processing speed and support applications such as online desalination in conjunction with analytical techniques [31]. The equation $t_D\approx {\displaystyle \frac{x^2}{2D}}$ models the diffusion time, which is significantly reduced in microfluidic systems relative to traditional larger dialysis devices. Specifically, narrowing the channel width by a factor of ten results in a hundredfold decrease in diffusion time. Nonetheless, the time required for solutes to traverse the membrane remains a critical limitation in these systems [32].

Song et al. differentiated microdialysis into three configurations: tubular, flat chip-like devices with sandwiched membranes, and microdialysis probes [33]. Early implementations by Xu et al. [34] in 1998 introduced dialysis on a chip using flow rates as low as 2–5 μL/min, demonstrating the separation of buffer and analyte by a cellulose membrane in a counterflow setup. While the desalination percentage was not explicitly reported, the modifications significantly enhanced the ESI-MS signal-to-noise ratio by a factor of 40. Moving on, Eijkel et al. successfully integrated a compact polyimide (2.3 mm) membrane into a membrane-based microfluidic device for water removal by combining spinning and thin film deposition following the fabrication of the microchannels [35]. The microfluidic device was operated by removing water through evaporation or osmosis-based separation. During osmosis, a concentrated salt solution accumulates on the membrane’s surface, allowing water to pass through due to osmotic pressure, as the polyimide membrane was impermeable to salts. Additionally, the device demonstrated the capability for water removal via evaporation. It featured a large area-to-volume reservoir at the channel’s end, where water was passively pumped in and evaporated, facilitating efficient flow from the channel to the reservoir. This dual functionality highlighted the device’s potential for effective desalination applications.

Xiang et al. showed that integrating a dual microdialysis system with different MWCO membranes has further enhanced dialysis efficiency. Tibavinsky et al. [32] refined a microscale dialysis configuration, achieving a 95% desalination efficiency within 1 s. The operational parameters in this study involved a sample channel flow rate ranging from 30 to 150 μL/h and a buffer channel flow rate of 50 mL/h. Despite these rapid dynamics, the recovery rate remained below 0.3%. The experimental program utilized membranes with an estimated pore size of approximately 50 nm. The results posited that the diffusion across the membrane represented the most significant bottleneck in the process. To address this, they replaced the conventional cellulose membrane material with ultrathin alumina, which they hypothesized would enhance the diffusional flux of the microfluidic dialysis system.

Although the process achieved high desalination efficiency in a short time, a significant loss of analyte (~55% protein) was observed. To minimize analyte loss, Wu et al.[36] fabricated a flexible silica isoporous membrane (SIM) with ultrathin thickness (90 nm), uniform nanopores (~2.3 nm), and high pore density (16.7% porosity) on a PET support. This SIM-PET membrane was integrated into a three-layer microfluidic chip, enabling protein desalting via size-exclusion dialysis. This device retained nearly all proteins with only 5% loss, while achieving ~99% desalting efficiency, leveraging the membrane’s precise molecular sieving properties to overcome analyte loss limitations.

Recent efforts have also explored strategies to eliminate physical membranes entirely, thereby addressing challenges such as clogging and analyte loss. D’Amico et al. [37] developed a microfluidic device integrating membraneless dialysis and dielectrophoresis (DEP) to isolate and concentrate bacteria directly from whole blood. By flowing blood alongside a low-conductivity buffer in adjacent microchannels, ions rapidly diffused from the blood across the liquid–liquid interface. This ion depletion lowered the electrical conductivity of blood, which is essential for stabilizing the electric fields required for subsequent DEP. In the DEP stage, a non-uniform electric field was applied to selectively trap intact bacteria near the electrodes. At the same time, permeabilized blood cell debris was removed, achieving a bacterial recovery of around 73%.

Dialysis, however, is not without its limitations [28]. Some target compounds may unintentionally diffuse across the membrane into the buffer side. As a result, their concentration in the original sample decreases, which can diminish the accuracy and sensitivity of subsequent tests and analyses performed on the diluted sample [28]. Additionally, the membrane’s mechanical stability limits the permissible pressure differences and flow rates, potentially impacting system efficiency and throughput. Strathmann [38] adapted the electrodialysis technique to microscale applications, claiming a significant advancement in on-chip desalination technologies [38]. This study employed a miniaturized electrodialysis system incorporating two electrodes that sandwich parallel cation-exchange and anion-exchange membranes and their spacers. After applying a potential sufficient to generate a faradaic current, ions from the incoming electrolyte are driven through these ion-selective membranes. Specifically, cations traverse only through the CEMs, while anions pass exclusively through the AEMs. This configuration facilitates the formation of alternating streams of diluted and concentrated solutions.

The efficiency of ion separation and subsequent desalination, quantified in terms of desalination percentage, is influenced by several factors: the magnitude of the applied potential, the initial concentration of ions in the electrolyte, and the system’s flow rate [38,39]. While it does not match the energy efficiency of RO, it offers scalability advantages and operates without the need for high-pressure pumps. However, it is important to note that the use of high-pressure pumps in RO is not inherently disadvantageous; in fact, RO systems often operate closer to the thermodynamic minimum energy requirements than current microfluidic or electrochemical systems. Nonetheless, electrodialysis remains advantageous for specific use cases, particularly where selective ion removal or compact, low-pressure operation is desirable [1,38].

Considering the operational dynamics of an electrodialysis system, three distinct regimes can be identified based on the applied potential: Ohmic, limiting, and over-limiting. In the Ohmic regime, which occurs between 0–2 V, there is a linear relationship between the applied potential and the resultant faradaic current. As the applied voltage increases, the ion diffusion rate through the membrane becomes faster than the rate at which ions diffuse from the bulk solution to the membrane surface. In other words, ions diffuse through the membrane faster than they are being replenished, forming a region near the membrane where the ion concentration becomes very low. This is known as the depletion layer. Even if voltage continues to increase, the current stops growing significantly, and the system transitions into the limiting current regime. The system enters the over-limiting regime if the voltage exceeds the limiting current. In this regime, electroconvection, water splitting, and charge carrier movement at the AEM help overcome the limiting current. Among these, the dominant process is electroconvection, where the applied electric field moves the fluid, bringing ions from the bulk solution to the membrane surface [40]. The theoretical and experimental research on desalination at over-limiting currents have been reviewed by Nikonenko et al. [41]. Rubinstein et al.[42] have deduced from numerical analyses and membrane modification experiments that electroosmotic flow also significantly contributes to the over-limiting current.

Kwak et al. [39] explored the ion transport in a PDMS electrodialysis cell using a 10 mM NaCl solution. They visualized local salt concentrations and fluid dynamics by adding Rhodamine 6G, a positively charged dye. Experiments covering the three electrodialysis regimes showed that the asymmetrical vortices observed at the AEM and CEM could be attributed to differing Stokes radii and ion transport characteristics. The AEM and CEM’s limiting regimes were achieved at different electrical potentials. Their findings suggest that starting the over-limiting regime might be the most energy-efficient mode of operation, as it also achieved 90% desalination.

Kim et al. explored the superior Na⁺ ion-selective and water-impermeable properties of sodium superionic conductor ceramic (NASICON) membranes [43]. By incorporating a NASICON membrane instead of the conventional CEM, the system achieved enhanced selectivity for Na⁺ ions, removing 98% of Na⁺ while retaining water and other cationic minerals. This modification resulted in a 1.36-fold increase in the volume of desalted water compared to traditional electrodialysis systems, alongside a reduction in specific energy consumption for NaCl extraction by approximately 13%. Integrating a two-sided NASICON-structured rechargeable seawater battery into the system allowed for further energy savings of around 20% by coupling selective desalination with energy storage.

Similarly, Jashni et al. [44] explored the development of a mixed matrix ED CEM enhanced with CNFs for improved desalination performance. Incorporating CNFs into the membrane matrix significantly enhanced the membrane’s surface hydrophilicity, potential transport number, permselectivity, and ionic conductivity while reducing water uptake. The optimal performance in terms of sodium flux was observed with membranes containing 0.5 wt% and 2–8 wt% CNF ratios, demonstrating a smoother surface and improved electrochemical properties for efficient desalination processes. The study provided insights into ion transport mechanisms through these novel membranes, considering factors such as ionic and hydrated radius, hydration potential, hydration-free energy, the entropy of hydration, and the Jones–Dole coefficient. The Jones–Dole coefficient is a measure of the strength of solute–solvent interactions and reflects the solute’s impact on the solvent’s microstructure. Its sign provides insight into whether the solute is a structure maker or a structure breaker. A positive Jones–Dole coefficient indicates that the solute is a structure maker, enhancing the hydrogen-bonding network of the solvent and increasing the solution viscosity. On the other hand, a negative coefficient suggests that the solute is a structure breaker, disrupting the hydrogen-bonding network and decreasing the solution viscosity [45,46]. Jashni et al.[44] showed that potassium ions had higher permeability and flux through the membrane than sodium ions. This behavior was attributed to potassium ions’ lower hydration potential, lower hydration-free energy, higher entropy of hydration, and negative Jones–Dole coefficient relative to sodium ions. The implications of this work can be extended and applied to membranes integrated in miniaturized microfluidic systems.

Smith [47] evaluated different designs of cation intercalation desalination cells, which ranged from using two intercalation electrodes to desalinate water from a single dilute stream to ED stacks that used two intercalation electrodes to desalinate water from multiple dilute streams simultaneously. The evaluation included three cell designs that desalinate one stream at a time with a given pair of electrodes: (1) a flow-through (FT) design where the influent is pumped through microscopic pores of electrodes separated by an AEM; (2) a flow-by (FB) design where the influent is pumped through an open channel between the electrode and an AEM; and (3) a membrane flow-by (MFB) design similar to the FB design but with a CEM placed at the interfaces between the electrodes and the influent streams. The findings showed that using two open flow channels (OFCs) with CEMs at the flow channel/electrode interfaces and an AEM arranged between flow channels increased the salt removal efficiency compared to the original FT electrode design. The stacking sequence of ion exchange membranes within an MFB cell is the fundamental repeat unit for ED stacks handling multiple dilute streams. Simulations of such ED stacks using nickel hexacyanoferrate (NiHCF) intercalation compounds (IHCs) predicted a twenty-fold increase in salt adsorption capacity per unit mass of NiHCF relative to MFB and FT cells while effectively reducing 0.7 M NaCl feed water to 0.2–0.3 M in the dilute streams. Thus, stack performance in electrodialysis chips could be enhanced by implementing the proposed counterflow arrangement.

Dydek and colleagues [48] developed a variant of electrodialysis called “shock electrodialysis”, which, unlike traditional electrodialysis, is not constrained by diffusion. This method utilizes a porous frit with an average pore size of 500 nm, positioned above a CEM made of Nafion, with a liquid reservoir on top. An electrical potential ranging from 0 to 2 V is applied across these components, enabling operation through ohmic, limiting, and over-limiting regimes. The phenomenon of over-limiting current in microchannels is attributed to either electroosmotic flow or surface conduction, with electroosmotic flow becoming predominant in larger pores. The word “shock” in “shock electrodialysis” refers to the distinct boundary formed between the depleted region and the surrounding electrolyte in the frit. Experiments confirmed the capability of this system to extract desalted water from the reservoir [48].

Recent advances in microfluidic ED have highlighted the critical role of interface design in optimizing ion transport. Benneker et al. [49] developed a microfluidic ED system that integrates geometrically structured charged hydrogels to investigate the effect of interface heterogeneity on ion transport dynamics. Their design featured alternating anion- and cation-exchange hydrogels in both homogeneous and heterogeneous geometries, allowing for real-time visualization of electric field distributions and ion depletion zones. The study revealed that heterogeneous hydrogel configurations significantly enhance electroosmotic flow and increase ion currents, achieving up to 30% higher current densities at equivalent applied potentials. This improvement was attributed to the non-uniform electric fields generated by the heterogeneous configuration arrangement.

4. Overview of Capacitive Deionization

CDI is an electrochemical desalination method to remove salts from brackish water and industrial wastewater, offering a potentially energy-efficient solution. The CDI desalination process can proceed via two main mechanisms: (1) non-faradaic ion capture within the electric double layer (EDL) created by the employed electrostatic force field and (2) faradaic ion capture, which relies on electrochemical reactions involving electron transfer between the electrodes and the electrolyte [50].

A CDI cell consists of two parallel electrodes separated by a small gap of a few millimeters, through which the electrolyte (e.g., feed water) flows. When an electric potential is applied across the electrodes, one becomes positively charged (the anode) while the other becomes negatively charged (the cathode). This creates an electrostatic field that drives the deionization process. The dissolved ions in the feed water are attracted to the electrode with the opposite charge: cations move toward the cathode, while anions migrate to the anode. An EDL is formed at the interface between each electrode and the feed water (electrolyte). The EDL consists of a layer of adsorbed counter-ions on the electrode surface (compact layer) and a layer of counter-ions in the electrolyte (diffuse layer). This structure enables the temporary storage of ions on the electrode surface within the EDL, resulting in deionized water exiting the system (Figure 4). The process described represents the non-faradaic mechanism of CDI [50,51].

A voltage of about 1 V is typically applied across the electrodes to extract ions from the solution and store them at the electrode of opposite charge within the electrical double layer [52,53]. The capacity of this system to store ions is directly linked to the electrodes’ effective surface area and the voltage applied [54]. During the cell’s charging phase, the applied voltage facilitates ion removal from the solution while simultaneously repelling co-ions from the electrodes, leading to some efficiency loss. The effectiveness of ion removal in relation to the charge stored is termed the charge efficiency, which is influenced by the solution’s ion concentration and the applied voltage. A charge efficiency exceeding 50% is necessary for effective desalination, with typical efficiencies for CDI systems using porous activated carbon (AC) electrodes ranging between 65% and 70% [55].

When all adsorption sites at the electrodes are saturated, the system reaches equilibrium, and water deionization ceases. At this point, the electrodes must either be replaced or regenerated. Regeneration involves either reversing the voltage or short-circuiting the electrodes, which causes the adsorbed ions to desorb from the electrode surface. A flushing solution is pumped between the electrodes to remove the desorbed ions, producing an ion-rich waste stream. However, during voltage reversal, there is a risk of desorbed ions being re-adsorbed by the opposite electrode. AEM and CEM can be placed between the electrodes to address this issue. These membranes guide the desorbed ions into the bulk solution and prevent them from being re-adsorbed by the opposing electrode. For example, when cations are desorbed from the anode, they pass through the adjacent CEM into the bulk solution. Simultaneously, the AEM near the cathode prevents these cations from being re-adsorbed by the cathode. The same principle applies to anions desorbed from the cathode [51].

The second working mechanism of CDI relies on faradaic reactions. Applying these reactions requires careful consideration, as they can positively and negatively affect CDI performance. On the positive side, certain reactions can enhance desalination efficiency by forming charged species or through pseudocapacitive and intercalation effects. On the other hand, faradaic reactions can also introduce challenges, including reduced electrode lifespan, performance degradation of the electrode, lower energy efficiency, the formation of chemical by-products, and undesirable pH fluctuations in the treated water [56].

Faradaic reactions in CDI are categorized into three main types:

• The first type involves anodic oxidation reactions. These reactions include oxidizing carbon electrodes, chloride ions, water, and other contaminants, such as inorganic ions and organic matter. Among these, carbon oxidation has been a primary focus due to its detrimental impact on the CDI system. Carbon oxidation can damage the electrode’s porous structure, reduce its mass, and significantly decrease both the electrode’s lifespan and the overall performance of the CDI system [51,56].
• The second type of faradaic reaction is cathodic reduction reactions. These reactions mainly include the reduction of oxygen, leading to the production of by-products such as hydrogen peroxide (H₂O₂) [51]. While cathodic reduction can cause an asymmetric potential distribution between the anode and cathode, which accelerates the unwanted carbon oxidation at the anode, it also offers potential benefits. For instance, the hydrogen peroxide generated can be utilized for water disinfection and the degradation of organic contaminants. Additionally, cathodic reduction reactions contribute to the removal of heavy metals from water by depositing them on the electrode surface [57].
• The third type is the faradaic ion storage, where reversible redox reactions enable ion storage through pseudocapacitive/intercalation effects. Unlike traditional electrostatic storage, which confines ions to the electrode surface, this process allows ions to penetrate and be stored within the electrode’s internal structure. For instance, in electrode materials like sodium transition metal oxides, cations intercalate into the electrode’s layered structure during reduction and deintercalate during oxidation when the voltage is reversed in the regeneration process. Similarly, conductive polymers and metal halides can reversibly bind anions. This approach significantly enhances the ion storage capacity by utilizing the electrode’s bulk rather than just its surface. Moreover, the reversibility of the redox reactions allows for repeated cycling, enabling efficient desalination while maintaining the structural stability of the electrodes [56,58].

The faradaic-based CDI includes various active materials that electrochemically facilitate different deionization mechanisms. These materials include EDL materials, pseudo-capacitive materials, and battery-type materials. EDL materials, such as activated carbon, store charge via the electrostatic accumulation of ions at the electrode–electrolyte interface without any faradaic reactions. This results in excellent cycling stability but relatively limited capacitance. In contrast, pseudo-capacitive materials, such as RuO₂ and MnO₂, store charge through surface or near-surface redox reactions, which are faradaic in origin but exhibit capacitive behavior due to fast and reversible kinetics. These materials typically offer higher capacitance than EDL materials but experience stability issues during prolonged cycling. Battery-type materials, such as LiCoO₂, store charge through bulk ion insertion/extraction (intercalation) into the electrode’s crystal lattice, involving phase transformations and distinct redox peaks. While they provide high energy density, their slower ion transport and diffusion-based mechanism result in lower power density. However, nanosizing these materials can enhance surface-dominated charge storage and even induce pseudo-capacitive-like behavior [59,60].

Suss and Presser [61] addressed CDI processes using energy storage and faradaic electrode materials. They discussed the use of various faradaic materials, including MXenes, MoS₂, layered titanium disulfide, sodium manganese oxide, and bismuth oxychloride, which represent a significant advancement in desalination technology, offering increased efficiency and capacity for brackish and seawater treatment. They highlighted the enhanced desalination performance offered by faradaic electrodes, such as ion intercalation materials, which have proven to have significantly higher salt storage capacities and potentially lower energy requirements than the traditionally used nanoporous carbon electrodes. This shift towards faradaic materials was particularly promising for treating higher salinity waters where conventional CDI techniques face limitations.

Chen et al. [62] investigated the performance of flow-electrode capacitive deionization (FCDI) by using carbon nanotubes (CNTs) mixed with AC in isolated closed-cycle (ICC) mode. The system included modified flow electrodes, acrylic endplates, graphite collectors, and ion exchange membranes. The introduction of CNTs notably increased the average salt removal rate, with a 1.7-fold improvement observed when 0.25 wt% CNTs were incorporated compared to the electrode without CNTs. This enhancement in salt removal was attributed to the improved electrical conductivity and the bridging effect on CNTs, facilitating more rapid charge transfer. The study also demonstrated that a higher initial salt concentration resulted in a better average salt removal rate due to the reduced electrical resistance in the cell. However, the efficiency slightly decreased for salt concentrations higher than 0.30 M due to co-ion leakage. Remarkably, the application of FCDI for real wastewater desalination achieved a high decrease in conductivity (94%), alongside high charge efficiency (>99%) and low energy consumption (0.034 kWh/mole), underscoring the potential of this method for sustainable wastewater treatment.

Likewise, Chang et al. [63] examined the preparation and electrosorption (a type of CDI desalination) performance of AC electrodes modified with titania (TiO₂) to enhance desalination efficiency. AC electrodes are known for their high surface area and electrical conductivity, and they attract and remove ions from water without membranes. Titania was loaded onto AC electrodes via a sol-gel method [64], thus improving the desalination performance. The electrodes were characterized using scanning electron microscopy, X-ray fluorescence, X-ray diffraction, and electrochemical workstation analysis. Results indicated that TiO₂ modification increased the formation rate of the electrical double layer and significantly enhanced electrosorption capacity, achieving a 62.7% increase in desalination ratio compared to unmodified AC electrodes.

Anderson et al. [65] and AlMarzooqi et al. [66] carried out macroscale comparisons of CDI with other desalination technologies. Additionally, Porada et al. [50] provided a literature review of CDI’s theoretical foundations, enhancing the understanding of the involved operational dynamics.

Microfluidic CDI

Typical investigations into CDI systems predominantly target the regeneration process of electrode materials to enhance storage capabilities, energy efficiency, and ion specificity, generally on a macroscale. These studies usually assess ion concentrations in the inflow and outflow processed streams to evaluate the overall system’s performance.

Various researchers have examined the behavior of ions within chip-based systems to extend the use of CDI technologies into microfluidic environments. Incorporating CDI into an optically transparent microfluidic chip has allowed direct observation of ionic movement using fluorescence microscopy. Suss et al. [67] explored the spatial and temporal changes in salt concentration during the charging process of a CDI cell. They integrated a neutral dye into the electrolyte, which experienced a decrease in fluorescence intensity when it encountered a chloride ion. This approach highlighted two distinct phases: a rapid charging phase and a slower desalination phase of the bulk solution.

Alternatively, Demirer et al. [68] adopted a different technique, utilizing laser-induced fluorescence within a CDI cell with semi-porous electrodes. Their research focused on monitoring the movement of charged fluorescent dyes at concentrations within the micro-molar range. By observing the quick charging of the cell and the slower desalination rate, the kinetic barriers and efficiency parameters of the CDI process were better understood, which are essential to enhancing the technology for practical desalination applications.

Roelofs et al. [69] focused on identifying the development of acidic and alkaline zones, or pH waves, visualized using fluorescence microscopy. Their study employed a chip-sized apparatus facilitating localized fluorescent imaging of ion concentration profiles in the flow channel, operated under static conditions with non-porous electrodes. The microfluidic chip featured a 97.5-μm-wide channel and 6 mm-long electrodes designed to minimize edge effects like uneven electric fields. A notable observation was the distinct intensity gradient across the channel for the dyes BODIPY and fluorescein, displaying a counterintuitive gradient. This suggested the formation of pH waves, even though the system functioned in a non-faradaic regime with an applied potential of 0.5 V. Computational analyses indicated that these pH variations were responsible for the unusual patterns observed. A leakage current of approximately 2 nA pointed to the dissociation of water, leading to a flow of H⁺ and OH⁻ ions at the electrodes, which altered local pH levels. The fluorescent dyes within the channel functioned as pH buffers, with their fluorescence intensity changing in response to shifts in valence, which allowed for the generation of the detailed intensity profiles observed. Understanding these pH variations is crucial for optimizing desalination processes in CDI systems.

Roelofs et al. [70] investigated CDI on a PDMS microfluidic chip as a pretreatment method for sample preparation, aiming to remove salts from solutions containing larger biomolecules. To validate their concept, the performance of this CDI chip was initially tested with a model sample containing 10 mM NaCl and fluorescein isothiocyanatedextran (FITC-dextran) at flow rates ranging from 1 to 10 μL/min. The chip is claimed to successfully desalinate the solution by removing 23% of Na⁺ and Cl⁻ ions, while the concentration of the larger molecule, FITC-dextran, remained unchanged. This study also introduced impedance spectroscopy to monitor salt concentration in situ and in real time through impedance measurements at a single frequency, using the same electrodes for desalination. Several modifications can further enhance the desalination efficiency of the system across various flow rates and concentration ranges. One approach is to increase the surface-to-volume ratio to improve the relative storage capacitance of the electrodes, achievable by narrowing the microfluidic channel, which is currently 1.5 mm wide. Another method involves increasing the residence time of the liquid between the electrodes, which can be achieved by increasing the lengths of the electrodes and the channel. The study found a correlation between the applied potential and the amount of charge stored at the electrode. While this proof-of-concept study demonstrated the chip’s potential for desalination, the authors noted that varying the pH and ionic strength can optimize the separation efficiency between salts and actual proteins. The device’s performance is claimed to be further improved by flushing samples multiple times through the device or multiple stages of identical devices, enabling the achievement of even lower salt concentrations, as required for applications such as ESI-MS.

Moreover, Tu et al. [71] developed a microfluidic electrochemical capacitive deionization (MFECDI) system featuring electrodes coated with electrochemically fabricated MnO₂ and polypyrrole (PPy). The microfluidic chip was constructed by stacking seven layers of different materials, with the top, bottom, and channel layers composed of 1.5 mm acrylic plates. The electrodes were made from titanium foil covered with electrochemically active materials, while polyethylene terephthalate (PET) layers were used to balance electrode thickness and prevent solution leakage. The desalination mechanism of the MFECDI cell involved pumping the solution through the channel layer, which is clamped between two electrodes. During the desalination step, a cell voltage of 0 V is applied, enabling electron transfer flow from PPy to MnO₂. This electron transfer is driven by the natural redox potential difference between the two materials. The electron flow triggers faradaic reactions at both electrodes, capturing ions from the solution by the electrodes’ surface. Conversely, during the salt-concentration step, a charging voltage of 1.2 V is applied to reverse electron flow and redox reactions. This applied voltage forces the system to release the captured ions into the solution. Desalination tests were conducted using various NaCl solutions, ranging from brackish water (8 mM) to seawater (600 mM). The MFECDI system demonstrated significant desalination capability, achieving up to 132 mg/g in a single-pass operation with 600 mM NaCl solution.

Furthermore, the system provided an ultra-high salt removal rate of 30 mg/g/min, resulting in efficient removal of 88% of ions in a single-pass operation for an 8 mM NaCl solution, meeting drinkable water standards and comparable to RO systems under certain conditions. The MFECDI system also exhibited excellent stability, maintaining over 90% and 75% of salt removal capacity in 100-cycle and 180-cycle operations, respectively. The salt-removal and salt-concentration tests demonstrated the MFECDI system’s potential for valuable ion recovery applications. The semi-automatic MFECDI design was also tested with real samples containing ions such as Na⁺, K⁺, Mg²⁺, and Ca²⁺. The system achieved salt removal capacity values of 199 mg/g and 615 mg/g for salinized underground water and seawater. The low energy consumption of 1.83 kWh/m³ for salinized underground water and 4.32 kWh/m³ for seawater further supported the MFECDI technique’s potential as a future desalination technology.

Similarly, Kim et al. [72] presented a new hybrid ECDI system that works alongside an oxidation process to address the challenges of anion adsorption/intercalation. This system’s cations are intercalated into a cation-selective battery material through an applied electrical potential in the desalination part. At the same time, reactions occur on the anode to produce oxidants, such as reactive chlorine species, in the oxidation segment. Simultaneously, anions from the desalination part diffuse into the oxidation part through an AEM to maintain neutrality in each part. This allows the desalination process to be carried out using only the cation-selective battery material, resulting in increased desalination capacity. Additionally, the oxidants generated on the anode can be utilized in electrochemical water treatment. In this study, the cation intercalation/deintercalation electrode was sodium manganese oxide (Na_0.44MnO₂), a representative material from sodium-ion batteries. As for the anode for reactive chlorine species and reactive oxygen species, cathodic polarized TiO₂ nanotubes were used, which can be easily fabricated through an anodizing process and electrochemical reduction, without the need for precious metals like iridium, ruthenium, and platinum.

The desalination part of the hybrid system operated in batch mode, while the oxidation part operated in a single pass mode with a flow rate of around 2 mL min⁻¹ under a constant current of approximately 20 mAg⁻¹ based on the total mass of Na_0.44MnO₂. The hybrid ECDI system exhibited a coulombic efficiency of over 96% based on the feed conductivity change observed during desalination, which suggested that deionization is achieved through the intercalation of Na⁺ (approximately 28 mg deionized per gram of Na_0.44MnO₂) and the simultaneous diffusion of Cl⁻ (approximately 44 mg) into the oxidation part via the AEM. It also indicated that the Na_0.44MnO₂ specific capacity of around 33 mAhg⁻¹ was effectively utilized for a desalination capacity of approximately 32 mAhg⁻¹ (due to the 96% Coulombic efficiency).

5. Overview of Ion Concentration Polarization

Ion concentration polarization (ICP) is an electrokinetic phenomenon that arises at the interface of ion-selective membranes or nanochannels, where selective ion transport leads to asymmetric concentration distributions. This effect is crucial in microfluidic desalination by enabling ion removal through controlled electrochemical processes [73]. When an external field is applied across an ion-exchange membrane, the membrane selectively permits the passage of counterions while rejecting co-ions. For example, in the case of CEM, positively charged cations migrate through the membrane toward the cathode. In contrast, negatively charged anions are repelled due to Donnan exclusion (an electrostatic effect that prevents the entry of ions with the same charge as the membrane) [74,75].

This selective ion transport forms two distinct regions on either side of the membrane: an ion enrichment zone and an ion depletion zone. In the ion enrichment zone, ions accumulate through continuous ion transport through the membrane. On the other hand, the ion depletion zone is a region where ion concentration is significantly reduced due to the removal of counterions and the rejection of co-ions [74,75]. In nanochannels or nanoporous membranes, the scale of ion transport is influenced by the Debye length, which defines the length of the EDL. When the nanochannel size approaches or becomes smaller than the Debye length, the EDLs formed along the channel inner walls overlap, further suppressing co-ion transport and enhancing the ICP effect [74].

Microfluidic Ion Concentration Polarization Desalination

In microfluidic desalination devices, the ion depletion region is central in removing salt from water (Figure 5). For instance, applying a sufficient electric field across a CEM linking two microchannels in an H-shaped microfluidic system moves cations from the upper to the lower channel. The resulting strong electric field and space charge layer intensify this effect, initiating the electroosmotic flow of the second kind (EOF2) [76]. Additional methods to propel fluid movement in the upper microchannel might involve pressure or electroosmotic flow of the first kind (EOF1), which arises from the impact of a tangential electric field on the electric double layer adjacent to the charged channel walls. The expansion of the depletion zone downstream by the fluid flow in the upper channel leads to the collection of desalted fluid at the outlet, thereby facilitating the desalination process [77].

One of the primary benefits of the ICP system is its ability to remove all types of charged particles: positively charged species can traverse the membrane, whereas negatively charged species are impeded by the robust electric forces at the beginning of the ion depletion zone in the upstream channel [78]. Kim et al. [79] demonstrated that ICP desalination avoids the fouling issues typically associated with AEMs, as only CEMs are employed. The accumulation of compounds like Ca(OH)₂ and Mg(OH)₂ on the surfaces of CEMs is prevented because OH^- the strong electric field repels ions in the upper microchannel.

Pandurangappa and Raghu [80] examined the progress in applying CNT-modified CEMs for ICP electrochemical processes. They highlighted the significant improvements in CNT membranes over traditional polymer membranes in terms of performance, including enhanced water flow, stability, flexibility, and specific surface area. The hydrophobic nature of CNT membranes necessitates modification with various additives such as chitosan, gelatin gums, and surfactants to enhance selectivity and ion transport in electrochemical desalination methods. The CNT membrane surface can also become hydrophilic using covalent bonds by adsorbing various functional groups, such as carboxyl, amines, and alcohols. Similarly, Shen et al. [81] explored the advancements in polymeric membranes, mainly focusing on integrating zinc oxide (ZnO) nanoparticles. By embedding ZnO nanoparticles into polymeric bases such as PVDF and PES, the membranes acquire enhanced hydrophilicity, photocatalytic self-cleaning capabilities, and antimicrobial activities, which make them more suitable for ICP applications.

Furthermore, Jashni et al. [82] explored the integration of Fe₃O₄/PVP composite nanoparticles into heterogeneous CEMs to assess their impact on the membranes’ electrochemical properties and desalination performance. The study revealed a significant enhancement in the physicochemical properties of the membranes upon incorporating Fe₃O₄/PVP nanoparticles. This enhancement manifested in reduced electrical resistance, improved hydrophilicity, membrane potential transport number, permselectivity, and ion flux. The observed improvements were attributed to the heightened dispersion of ionic clusters and the inherently hydrophilic nature of the Fe₃O₄/PVP nanoparticles, facilitating ion transport and enhancing desalination efficiency through electrodialysis. Similarly, Wang et al. [83] explored the development of electrically conductive polyaniline (PANI) membranes through a non-solvent-induced phase separation method. Incorporating dodecylbenzene sulfonic acid (DBSA) as a dopant enhanced the membranes’ conductivity and hydrophilicity. This modification led to a notable improvement in antifouling performance within an electrofiltration cell, particularly when an external voltage was applied. It effectively reduced fouling by bovine serum albumin (BSA). An increase in voltage from 0 to 1 V markedly enhanced the flux, which is attributed to the heightened electrostatic repulsive force between the foulants and the membrane. Scanning electron microscopy (SEM) verified that this interaction facilitated a looser fouling layer, diminishing hydraulic resistance. While these advancements (

Table 2. Strategies for enhancing CEMS performance: targeted indicators, outcomes, practical implications, and limitations.

Enhancement Strategy	Targeted Performance Indicator	Outcome	Practical Implications	Limitations	Ref.
Incorporation of Fe3O4/PVP composite nanoparticles	Electrochemical properties and desalination performance	Significant enhancement in electrochemical properties, including reduced electrical resistance, improved hydrophilicity, increased membrane potential transport number and permselectivity, and enhanced ion flux.	Fe3O4/PVP composite nanoparticles can significantly improve the efficiency of cation-exchange membranes for desalination through electrodialysis	The study does not extensively discuss the modified membranes’ long-term stability and performance under varied operational conditions.	[82]
Modification with GO via phase inversion method	Hydrophilicity, water flux, and mechanical properties	Hydrophilicity and water flux significantly increased; pure water flux increased to 430.0 L/(m²·h·bar), and tensile strength to 305.3 cN with 0.1 wt% GO.	Demonstrates the effectiveness of GO in enhancing the performance of PVC ultrafiltration membranes for wastewater treatment, offering a cost-effective solution with improved efficiency.	The study primarily focuses on improving hydrophilicity and mechanical properties without extensive discussion on long-term stability and fouling resistance under various water qualities.	[84]
Incorporation of DBSA into PANI membranes via non-solvent-induced phase separation	Antifouling performance and conductivity	The application of external voltage significantly mitigated BSA fouling, demonstrating enhanced antifouling performance with increased voltage.	This approach offers a promising strategy for fabricating conductive membranes capable of electrically enhanced fouling mitigation, potentially improving the longevity and efficiency of membrane filtration systems.	The study primarily focuses on BSA as a model foulant, suggesting further research to evaluate performance with diverse foulants and in real wastewater treatment scenarios.	[83]
Incorporation of ZnO nanoparticles into polymeric membranes (e.g., PVDF, PES)	Hydrophilicity, photocatalytic self-cleaning capabilities, and antimicrobial activities	Enhanced antifouling performance, improved water permeability, and stability; introduction of photocatalytic self-cleaning and antimicrobial properties	Offers a promising approach for reducing membrane fouling in water treatment applications, potentially extending membrane lifespan and efficiency	The review calls for further research on overcoming ZnO nanoparticles’ aggregation and leakage and understanding long-term performance in practical applications.	[81]
Radiation grafting of HEA onto PVDF membrane	Hydrophilicity, pH-dependent water flux, and antifouling performance	Enhanced hydrophilicity and pH-responsive water flux; improved antifouling performance with a novel observation of higher BSA flux than water flux	Presents a novel approach for creating pH-sensitive and antifouling PVDF membranes, with potential applications in controlled drug release and water treatment	The study mainly focuses on the characterization and performance evaluation of modified membranes at the laboratory scale, with limited discussion on large-scale applicability and long-term operational stability.	[85]
Incorporation of CNFs into membrane matrix	Surface hydrophilicity, potential transport number, permselectivity, ionic conductivity, and sodium flux	Enhanced electrochemical properties and sodium flux, with optimal performance observed at 0.5wt% and 2–8wt%. CNF ratios.	Demonstrates the effectiveness of CNFs in enhancing the performance of CEMs for electrodialysis, offering a scalable approach for improved desalination processes.	The study focuses on a specific range of CNF concentrations, suggesting further research is needed to explore the full potential and practicality of CNF incorporation across a broader range of membrane formulations.	[44]

) were tested on the macro scale, they can be applied to miniaturized membranes for microfluidic settings.

A typical performance metric of ED and ICP systems is the efficiency of current utilization (CU), which determines how effectively current is used for salt reduction in the feed [79]. Ideally, the CU value would be unity, indicating no current loss. Noteworthy, CU values exceeding unity can be achieved in ICP systems, which can be attributed to three primary factors. First, the depletion regions of AEMs and CEMs are independent. In other words, ion depletion zones at CEMs and AEMs are interdependent in conventional ED because cations and anions must be transported balanced to remain electroneutral. However, in ICP, the depletion zone formed at CEMs and AEMs are independent. This means a depletion zone can form at a CEM due to selective cation removal without requiring simultaneous anion transport through an AEM. This independence allows for more localized and intensified depletion effects, contributing to higher CU efficiency. Second, under the same current as NaCl, CEMs generate a stronger depletion zone compared to AEMs due to the separate maintenance of electroneutrality and current conservation in each depletion region. Finally, as a result, ICP desalination utilizing only CEMs can achieve a higher salt removal in the dilute flow compared to ED [79].

Analyte detection is challenging due to small sample volumes and low concentrations of target molecules or ions, which reduce the sensitivity of conventional detection methods. To address this limitation, a pre-concentration step is often required to locally increase the analyte concentration, thereby enhancing detection accuracy and signal strength. This process can be implemented in microfluidic devices. Therefore, Knust et al. [86] developed a microelectrochemical cell with two neighboring microchannels separated by a bipolar electrode (BPE). The BPE addressed this challenge by balancing incompressible fluids’ convective flow against electromigration and enriching cations and anions. This dual-channel configuration can be achieved using ICP, which establishes an ion depletion zone through the applied potential across the channels. This results in the selective transportation of ions through a material such as Nafion.

Kim et al. [87] developed a microfluidic desalination system that integrates a Y-shaped microchannel with a nanojunction to achieve ICP-based salt removal. The nanojunction, a nanoscale channel connecting two microfluidic branches, acts as a selective ion transport pathway. By introducing a salt solution into the feed channel and applying a potential across the nanochannel, an ion depletion zone was formed at the interface between the nano- and microfluidic channels. This setup resulted in one outlet producing a stream of fresh water and the other a concentrated brine, achieving a water recovery rate of 50% at a salt rejection rate of 99% and an energy efficiency of 3750 mWh/L. The principle of ICP applied here focused on ion removal rather than separation based on mobility, and it relied on the robust geometry where the separation was achieved as ions diverted from, rather than through, the nanopores. Moreover, using ICP with CEMs, the same group of researchers obtained an increased salt removal efficiency by up to 20% over traditional electrodialysis methods when a constant current is applied [88].

Motivated by these findings, Kim et al. [79] explored this method in a PDMS-based microfluidic device, passing an ion current through ion-selective nanopores for treating highly contaminated brine with total dissolved solids concentrations as high as 100,000 mg/L. The system exhibited a specific energy consumption of approximately 3.5 kWh/m³, with the added benefits of avoiding membrane fouling and scaling. The operational dynamics within the microfluidic device included monitoring pH changes near the AEM and CEM using pH indicator dyes in the feed water. Significant pH changes were predominantly observed near the AEM, along with some particle aggregation, suggesting localized chemical reactions or interactions at these sites. Despite its effectiveness, the application of this technology has been primarily limited to portable desalination units. This limitation emphasizes the need for further development to scale the technology for broader applications.

However, the throughput of an ICP-based microfluidic device is significantly low when operating as a single H-shaped microchannel system when compared to conventional desalination technologies. For example, a typical microchannel measuring 100 μm × 15 μm may produce only about 3.6 × 10⁻⁴ L per day if the fluid flow speed is 1 mm/s [89]. This output is insufficient for any industrial-scale water production. Increasing the height of the channel to boost throughput is not feasible because the desalination efficiency declines if the ion depletion zone does not adequately cover the channel height. Thus, parallelization of microchannels emerged as a practical solution to enhance throughput while maintaining the effectiveness of the desalination process.

In this context, Ko et al. [90] proposed an ICP-based preconcentration system comprising 128 parallel channels connected to a single reservoir at the inlet or outlet. While the system was initially designed for biomolecular detection, the underlying mechanism works well with desalination. However, scaling up this approach to an industrially viable level remains challenging. Even with such parallelization, the sheer number of microchannels required to produce sufficient water for large-scale applications presents significant logistical challenges, making it difficult to meet industrial demand through this method alone. Similarly, Kwon et al. [91] approached the parallelization of desalination systems by employing random microporous ion exchange membranes. Their design mirrors that of ED systems but omits the use of AEMs. Instead, desalted water is collected directly from the micropores within the membrane, providing a unique and potentially more efficient method for desalination. This approach leveraged the structural benefits of microporous membranes to enhance water extraction while simplifying the overall system layout.

Consequently, Tang et al. [89] introduced a method for the large-scale parallelization of microchannels in ICP-based seawater desalination. Unlike previous devices that directly parallelize multiple microchannels, their design features parallel microchannels (micropores) solely within the CEM, while merging other regions into single wide channels. This system facilitated selective ion transport and ion depletion within the membrane area, with the wide channels upstream and downstream managing fluid and ion transport. The microporous permselective membranes functioned as ion filters, with pores on the micrometer scale blocking the passage of Cl⁻ ions (and most Na⁺ ions required for neutrality) through a locally amplified electric field induced by ICP. They showed that the CEM allowed Na⁺ ions to pass through but not Cl⁻ ions, forming regions near the membrane surfaces with reduced ion concentration. This localized increase in electric field strength near the membranes caused the expulsion of Cl⁻ ions and the expansion of the ion depletion zone into the internal spaces of the microchannels and adjacent upstream areas. As a result, Na⁺ ions could move into the membranes immediately upon reaching the membrane surface, while Cl⁻ ions, initially migrating rightward, ultimately shifted leftward in the boundary region, resulting in almost no net Cl⁻ flux. This process ensured that the downstream channel maintained very low Na⁺ and Cl⁻ ions concentrations, ultimately producing desalinated water at the outlet [89].

Simulation results indicated that the system achieved a salt removal rate of 97.47%, an average velocity of 0.26 mm/s, an energy consumption of 7.1 kWh/m³, and an energy performance index ratio of 21 at a current density of 19,969 A/m². The average ion concentration at the outlet was 12.6 mM. Increasing the salinity of the produced water further reduced the energy consumption. A significant advantage of this method is the ease of manufacturing microporous membranes compared to systems with multiple full-length microchannels. The membrane’s micropores do not require high precision; as long as they are of microscale size, the system can function effectively regardless of the pores’ shape or structure.

However, it is important to note a few limitations. First, this study was based on numerical simulations and needs experimental validation. Second, the model only represented part of the desalination system responsible for the ICP effect and salt removal. Third, the high current density in the membrane system is significant, considering the small membrane area relative to the achieved desalination. Nevertheless, the system’s feasibility is supported by its relatively low power consumption and high fluid flow rates within the micropores, with experimental results indicating minimal adverse effects from Joule heating [89].

6. Overview of Electrochemical Desalination

Electrochemical desalination (ECD) is a membraneless desalination technology that utilizes faradaic (redox) reactions and electric-field-driven ion transport to separate saline water into desalinated water and brine. The process occurs within a microfluidic channel network, where feed water flows through a primary microchannel that branches into two outlets: desalinated water and concentrated brine. At this branching junction, a bipolar electrode is positioned. When a high driving potential (e.g., 3.0 V) is applied across the bipolar electrode, chloride ions (Cl⁻) in seawater are oxidized (neutralized) at the anode, creating a localized ion depletion zone at the branched junction. As a result, a localized electric field develops, which influences the movement of ions within the channel [92,93].

Ion transport in ECD is governed by two forces: convection and electromigration. Convection, driven by a pressure gradient between the inlet and outlet reservoirs, moves the bulk fluid forward, following a parabolic flow profile. However, the intensified electric field accelerates ion migration near the depletion zone, increasing their electrophoretic velocity. When the electrophoretic velocity exceeds the convective velocity, cations are redirected toward the brine stream, followed by anions to maintain electro-neutrality, while ion-depleted water flows into the desalinated outlet (Figure 6) [92,93]. This process introduces a different mechanism of salt removal compared to non-faradaic processes like ion exchange or CDI [9].

Moreover, the operational principle of ECD diverges significantly from those used in membrane and thermal desalination technologies. For instance, membrane-based desalination requires applying pressure exceeding seawater’s high osmotic pressure (approximately 30 atm) to achieve desalination. Conversely, thermal desalination requires substantial energy input to vaporize water [28]. Both technologies have well-defined baseline energy requirements. In contrast, ECD operates differently, requiring only enough energy to oxidize a minimal fraction—approximately 0.01%—of the total chloride ions in seawater to create a local electric field gradient. The proposed platform of ECD offers several significant advantages over conventional desalination techniques: (1) it obviates the need for membranes, thus eliminating a common source of operational challenges related to membrane fouling; (2) ECD operates on a simple 3.0 V power supply, which makes it a feasible option for resource-limited settings where only battery power or low-power renewable energy sources are available; and (3), the low initial capital investment and the potential for scaling up through parallel configurations enhance its applicability [93].

Microfluidic Electrochemical Desalination

Gude and colleagues [94] explored the functionality of the zinc ferricyanide desalination battery designed for both desalination and energy storage. The battery is configured with a zinc (Zn) electrode in a zinc chloride (ZnCl₂) aqueous solution at the anode and a graphite electrode in a mixed electrolyte solution of K₃[Fe(CN)₆] and K₄[Fe(CN)₆] at the cathode, separated by saline water in the central chamber. The setup was divided by both AEM and CEM. The team conducted several experiments to assess the effects of charge and discharge cycles on desalination rates, energy storage, and discharge performance at various charging levels. They also evaluated the power density and overall desalination efficacy under different conditions. The findings revealed that this electrochemical system stored energy and utilized the battery’s charge–discharge cycles to facilitate desalination without additional energy expenditure. Specifically, a desalination rate of 30.3% was recorded over a 12-h discharge period, reducing the salinity from 41.2 g/L to 28.7 g/L, with 74.2% of the used input energy being recuperated through the desalination and discharging process.

Energy consumption in electrochemical systems heavily depends on the overall ohmic resistance of the electrical circuit, which is significantly influenced by the design of the electrode cell. This ohmic drop, resulting from energy dissipation, adversely affects cell performance and is closely associated with the inter-electrode (IE) gap, namely, the distance between the electrodes. Recent microfluidic desalination devices employing the concept of ECD with less than 1000 μm IE gaps have been under investigation. This reduction in gap size has led to notably improved performance [93]. Microfluidic devices achieve higher current efficiencies in oxidation processes by enhancing mass transfer. Thus, the quest for efficiency and miniaturization drives the exploration of microfluidic-based ECD systems, which operate on similar faradaic principles but on a significantly smaller scale. These systems integrate the principles of electrochemistry within microfluidic channels, providing precise control over the desalination process and potentially reducing overall energy consumption. Scialdone et al. [95] highlighted several benefits of microfluidic devices compared to traditional macroscale systems. These include facilitating scaling up through parallelization, rapid optimization of operating conditions owing to smaller IE gaps and reduced residence times, and the feasibility of employing multistage systems.

Crooks et al. [93] investigated the performance of a microchannel equipped with a bipolar electrode (BPE). In the ECD process, seawater is divided into brine and desalted water streams at the junction of a microchannel equipped with a BPE. The experimental setup for ECD involved using a PDMS/quartz hybrid microfluidic device, designed with channels tall enough (22 μm) and sufficiently wide (100 μm inlet, 50 μm outlets) to facilitate efficient flow dynamics and prevent clogging by naturally occurring sediments in seawater. This preemptive measure of removing sand and debris through sedimentation before use allowed the system to operate without the extensive pre-treatment required by membrane-based systems, such as disinfection or the addition of anti-scaling chemicals. At the junction, seawater was divided into brine and desalted water streams, and the anodic pole of the BPE created an ion depletion zone that generated a local electric field gradient, directing ions toward the brine channel. The system operated with remarkable energy efficiency at 25 mWh/L, close to the theoretical minimum (~17 mWh/L), achieving 25 ± 5% salt rejection and 50% recovery.

Grygolowicz et al. [96] developed a cylindrical two-electrode ECD cell featuring a central silver/silver chloride (Ag/AgCl) electrode enclosed by a Nafion membrane, which is further surrounded by a solution. This setup was effective at potentials exceeding 400 mV in stationary and flow injection modes. Challenges were noted with substantial residual currents during electrolysis, attributed to NaCl’s back diffusion from the membrane’s external side due to the Nafion membrane’s limited permselectivity. The electrochemical process involved the removal of Cl⁻ ions by oxidizing the silver at the Ag/AgCl electrode, which resulted in the formation of silver chloride. The Nafion membrane, being cation-selective, permits only the passage of Na⁺ ions, blocking Cl⁻ ions. The cell removed 90% of the salt in flow-through mode within 90 s.

Motivated by this work, Shea et al. [97] miniaturized the system developed by Grygolowicz et al. [96] for in situ deployment, aiming to prevent any leakage during treatment and to control the thickness of the seawater fluidic compartment precisely. To carry out this miniaturization, the system transitioned from a coaxial arrangement to a planar configuration, which offered improved control over channel thickness. The transition allowed for better management of the fluidic compartment’s dimensions, enhancing the desalination performance. Specifically, the planar design facilitated precise regulation of the seawater layer, optimizing the system’s capacity to remove sodium chloride. This approach increased the system’s efficiency and made it more suitable for practical applications in various environmental and industrial settings. The desalination performance of the microfluidic device was evaluated through conductivity measurements of the treated samples, demonstrating a proof of concept. When a potential of +0.9 V was applied, the conductivity, and consequently the concentration, was significantly reduced from an initial value of 42.13 mS (0.6 M) to 20.36 mS (0.206 M) after 120 s, 17.37 mS (0.157 M) after 240 s, and 6.38 mS (0.0516 M) after 360 s of oxidation. These results indicated a substantial decrease in salt concentration over time. The operating conditions enabled the regeneration of the silver electrode, allowing for its use in six to ten cycles before the depletion of the silver working electrode.

Another study employed an inverted cation intercalation desalination cell architecture for the ECD of aqueous 600 mM NaCl [98]. This study demonstrated the effect of silver nanoparticles (AgNP) paired with silver chloride (AgCl) in a high-performance desalination battery featuring two flow channels separated by a 180 μm-thick CEM. The cell used AgNP as the positive electrode and AgCl as the negative electrode, operating by cycling between −0.1 V and +0.1 V, i.e., a net voltage difference of 200 mV. This voltage range was selected based on galvanostatic charge/discharge profiles. During operation, the AgNP was converted to AgCl, removing chlorine ions from one flow channel. Simultaneously, charge neutralization of the electrolyte was achieved by transferring sodium ions across the ion-exchange membrane into the other flow channel. This process enriched the latter with sodium and chlorine ions, which were liberated as AgCl was converted back to Ag. Once both electrodes were fully converted to AgCl or Ag, the operation could be inverted, allowing the flow channel with increased salt concentration to become the effluent stream of desalinated water and vice versa. The initial desalination capacity was recorded at 176–190 mg/g in 600 mM NaCl, corresponding with an initial charge capacity exceeding 200 mA h/g. However, the desalination capacity decreased after several cycles, measuring 125 mg/g after 8 cycles and stabilizing at approximately 115 mg/g after 15 cycles. This performance was achieved at a desalination rate of 23 mg/g/min with an applied current of 0.1 A/g. The initially fluctuating charge efficiency also stabilized at about 98% after 15 cycles. The decreasing desalination capacity is likely due to the partially irreversible formation of AgCl at the positive AgNP electrode, which correlates with the decreased charge capacity observed during cycling [99].

Table 3. Overview of Microfluidic Desalination Systems: Key Parameters, Performance, and Improvements.

	Applications	Feed	Feed Channel Width	Feed Channel Height	Flow Rate	Efficiency	Advantages	Potential Improvements	Ref
Dialysis	Cleanup of biological samples for ESI-MS	5 µM horse heart myoglobin in 500 mM NaCl, 100 mM Tris, and 10 mM EDTA	160 µm	60 µm	0.3 µL/min	Signal-to-noise ratio increased by a factor of 40	Increased (ESI)-MS sensitivity	Reduction of analyte loss	[34]
	Salt removal from biological samples prior to ESI-MS	40 μM cytochrome-c in 100 mM KCl	100 µm	6 µm	30–150 µL/h	95% salt removal in 1 s	Fast Increased (ESI)-MS sensitivity	Reduction of buffer mass transfer resistance Reduction of analyte loss Reduction of sample channel length for lower nano-ESI flow rates	[32]
	ESI-MS	Cytochrome dissolved in 154 mM NaCl	0.5 mm	89 µm	1 µL/min	~99%	Increased (ESI)-MS sensitivity Fast	Reduction of chip volume to improve salt transfer distance. Optimizing microfluidic device structure and SIM.	[36]
	Isolation and concentration of bacteria from blood	100 µL of blood and an E coli suspension in 500 µL dielectrophoresis buffer	250 µm	3 mm	0.6 mL/h	79 ± 3% of E. coli separated 78 ± 2% of Staphylococcus aureus separated	Increased signal-to-noise ratio Fast Minimal operator intervention	Achieving sufficient volumetric throughput.	[37]
ED	Investigating and optimizing ED systems	10 mM NaCl 10 µM Rhodamine 6G	1 mm	200 µm	10 µL/min	90% salt removal	Scalable High water purity High ion controllability	Improving energy efficiency	[39]
	Usage of Na super ionic conductor (NASICON) in ED	Seawater	-	-	500 mL/min	Achieved up to 98% of Na+ removal	Enhanced salt removal while retaining water and cationic minerals Reduction of energy consumption by around 13% Desalted water volumes were 1.36 times higher than traditional ED	Integration with renewable energy	[43]
	Investigating the effect of hydrogel geometry on ion transport in electrodialysis	0.1 mM NaCl	700 µm	75 µm	3 µm/min		Enhanced electroosmotic ion transport in heterogeneous hydrogel geometries Higher ionic currents for heterogeneous geometries at equal potentials Earlier onset of over-limiting current	Further exploration of hydrogel geometries to optimize ion transport efficiency	[49]
CDI	In situ concentration mapping	50 or 80 mM KCl 5mM 6-methoxy-N-(3-sulfopropyl)quinolinium (SPQ)	1.5 mm	5 mm	0	10% within 25 s	Provided spatially and temporally resolved salt mapping between charging porous electrodes Enhanced understanding of ion dynamics	Reduction of channel dimensions Development of a non-invasive, probe-free measurement method suitable for potable water	[67]
	Laser-induced fluorescence visualization of ion transport in a pseudo-porous CDI microstructure	0.7 mM solution of Sulforhodamine B (cationic dye) and Fluorescein (anionic dye)	0.2 mm	0.1 mm	0	60% within 60 s	Direct visualization of salt concentration gradients	Reduction of electrode pores’ size to allow for EDL overlap	[68]
	Visualization of pH waves	10 mM NaCl Dyes: BODIPY and fluorescein	97.5 µm	~7 µm	0		High-resolution, localized pH mapping	Adaptation for dynamic flow conditions	[69]
	Sample clean up to concentrate FITC-dextran	10 mM NaCl with FITC-dextran (model protein)	1.5 mm	0.4 mm	1–10 µL/min	23% salt removal within 30 min	Real-time salt monitoring via impedance spectroscopy Minimal analyte loss (FITC-dextran retained)	Optimizing flow rates to enhance desalination efficiency Mitigating pore blockage by larger molecule	[51]
	Water purification	Brackish water (8 mM) seawater (600 mM)	2 mm	1.5 mm	0.5–2 mL/min	88% ion removal in a single pass	High purification efficiency Low energy consumption	System scalability by connecting multiple MFECDI cells in series	[71]
	Hybrid electrochemical desalination with integrated oxidation	3–35 g/L NaCl			Desalination: batch mode Oxidation: 2 mL/min	28 mg Na+ deionized/g Na0.44MnO2 (electrode material)	Simultaneous desalination and oxidation Enhanced removal of contaminants Overcome the limitations of the materials used to capture anions	Scaling-up	[72]
ICP	Brine desalination	0.1–1.7 M NaCl	1.5–2.0 mm	0.2–0.6 mm	20 μL/min	Up to 70% salt rejection	Reduced membrane fouling/scaling	Scaling-up Optimizing multi-stage configurations	[79]
	Simultaneous separation and enrichment of anions and cations in a single microchannel	0.5 M TrisH+ and acetate buffers	50 µm	8.7 µm	Pressure-driven counter-flow		Enabling concurrent enrichment of cations and anions Utilizing local electric field gradients for separation Simple fabrication procedure	Reduction of voltage requirements	[86]
	Seawater desalination	Seawater (~500 mM salinity)	1 µm (simulation-based)	1 m (simulation-based)	0.56 m/s	Salt removal > 98%	High desalination for drinkable water (~10 mM salinity) Low energy consumption (12.82 kWh/m3)	Scaling up for industrial use	[89]
ECD	Electrochemical desalination with energy recovery and storage	Seawater			-	30.3% salt removal during a 12-h discharge	Dual function: desalination + energy storage	Exploring electrode/electrolyte combinations to enhance desalination efficiency Integrating with renewable energy grids	[94]
	Electrochemical removal of NaCl from seawater samples	0.6 mM NaCl	30 µm		40 µL/min	90% salt removal within 90 s	Enables rapid NaCl removal to facilitate nutrient analysis without affecting nitrite or nitrate content	Mitigation of Donnan failure (back-diffusion)	[96]

summarizes key parameters, performance metrics, and potential improvements discussed across the different microfluidic desalination systems reviewed in this work.

Table 4. Comparison of Electrochemical-based Microfluidic Desalination Techniques. Adapted from [99,100].

Method	Driving Mechanism	Salt Removal Efficiency	Scalability	Fouling Risk	Energy Requirement	Key Advantages	Limitations
Dialysis/ED	Concentration gradient/Electric field	Moderate to High (up to ~95%)	Moderate (parallelization needed)	Moderate (membrane-based)	Moderate	Established method; high selectivity; modular	Membrane fouling; limited flow rates; mechanical stability limits
CDI/ECDI	Electric double layer or redox ion storage	Moderate to High (~60–90%)	Moderate (surface area dependent)	Low to Moderate	Low to Moderate	Low energy consumption; modular; reusable electrodes	Electrode degradation (faradaic); co-ion effects; limited for high salinity
ICP	Electrokinetic ion transport at nano/membrane interface	High (up to ~99%)	Low (single chip = low throughput)	Low	Low to Moderate	No membrane fouling; high salt rejection; suitable for portable use	Limited throughput; needs extreme parallelization to scale; energy at high current
ECD	Redox reactions and local electric fields	Moderate (up to ~90%)	High (via parallelization)	Low	Very Low (~25 mWh/L reported)	Membraneless; simple setup; renewable integration possible	Electrode material consumption; slower desalination rate; stability issues

provides a comparison of the previously discussed electrochemical-based microfluidic techniques [99,100].

7. Desalination Membrane Materials

As mentioned above, membranes are an essential component of microfluidic desalination systems as they govern the desalination performance; therefore, the selection of membrane materials is necessary for achieving high throughput in microfluidic desalination systems [101,102,103]. Synthetic membranes, in terms of their materials, can be roughly categorized into the following:

• Polymeric membranes, including those made out of cellulose acetate (CA), polyethersulfone (PES), polypropylene (PP), poly(vinyl alcohol) (PVA), as well as polyamide (PA) and its derivatives. Early membranes made out of CA polymers suffered from low water permeance, a narrow pH operating range, and poor biodegradation resistance. Limitations of CA membranes led to their replacement by thin-film composite (TFC) polyamide membranes. TFC polyamide membranes offer higher water permeance, broader pH tolerance, and excellent salt rejection but face a trade-off between permeability and selectivity due to structural factors like pore size distribution and crosslinking density [104].
• Inorganic or ceramic membranes, typically made from silica, alumina, zirconia, or a mixture of these materials. Inorganic membranes, featuring a macro-porous support-layer and a meso- or micro-porous active layer, offer the advantage of durability in harsh environments, e.g., high temperatures, highly contaminated feeds, and/or corrosive environments where polymeric membranes might fail [105]. Zeolites, MXenes, and molybdenum disulfide (MoS₂) are additional examples of inorganic materials used to make desalination membranes. Zeolite membranes are made from natural or synthetic crystalline aluminosilicates with Group I and Group II cations, and the oxygen ring size defines their pore structure. In contrast, the Si/Al ratio controls properties like wettability and surface charge [106]. MXenes have a general formula of M_n+1X_n, where M denotes the transition metal and X denotes the carbon, nitrogen, or carbon–nitrogen that makes up the MXene material. The formula of a MXene could also be M_n+1X_nT_x, where T denotes a functional group such as -O, -OH, or -F. In addition, MoS₂ membranes offer high hydrophilicity, chemical stability, and stronger interactions between MoS₂ nanosheets, leading to a more stable laminar structure. Although these membranes have shown superior separation performance, they tend to be more expensive and more challenging to fabricate as a continuous, defect-free layer than polymeric membranes [11,101,106].
• Composite or mixed matrix membranes (MMMs), which consist of polymers embedded with nanoparticles “fillers” to enhance the physicochemical properties of the membrane and, in turn, its separation performance. A well-designed composite membrane should ensure the stability and homogeneous distribution of the filler material in the polymer-solution system, as well as the interfacial compatibility between the polymer and the filler material [103]. One of the main advantages of MMMs is their high surface-to-volume ratio, which can provide high porosity on the surface and in the membrane bulk. Moreover, incorporating the nanomaterials into the polymeric matrix can enhance desalination performance by modifying membrane properties, leading to improved permeability and selectivity, increased hydrophilicity, better mechanical strength, enhanced antifouling and antibacterial properties, and greater thermal stability [106].

Nanoparticles are particularly useful in desalination due to their unique structural and morphological features, which promote rapid water transport and high salt rejection. Nanoparticles can be integrated into polymeric membranes by blending them with the polymer solution or depositing them on the surface of the polymeric membrane. Various nanoparticles have been used to fabricate composite membranes for desalination, including metal/metal oxide nanoparticles (e.g.,TiO₂, Al₂O₃) silica (SiO₂), graphene oxide (GO), CNTs, and cellulose nanocrystals (CNCs) [103,106,107].

GO is a 2D carbon-based material that has recently garnered a lot of attention in membrane technology. GO is very hydrophilic due to its surface functional groups, such as the carboxyl and hydroxyl. Its impermeability stems from the delocalized electron clouds in the p-orbitals, which hinder the gaps in the aromatic rings of graphene and make it act as a barrier to the diffusion of gases and liquids. Stacked GO membranes’ improved permeability and selectivity are achieved through size exclusion and electrostatic interactions [103,107].

CNTs can also be used as filler materials in composite membranes due to their unique structural, mechanical, and chemical properties, which help enhance membrane mechanical strength, thermal stability, and resistance to fouling. CNTs’ surfaces can also be functionalized to improve their hydrophilicity, further increasing water permeability and selectivity in desalination applications [107].

Metal-organic frameworks (MOFs) have also proven to have great potential for desalination applications. MOFs are hybrid nanoporous materials consisting of metal ions and organic linkers. MOFs are excellent filler materials in composite membranes due to their highly porous structures, large surface areas, and tunable pore sizes. In addition, MOFs can be easily functionalized to enhance hydrophilicity, improving water permeability and antifouling properties [107,108].

Table 5. Comparison between desalination membrane materials (adapted from [101,105,106,109,110]).

Membrane Material	Polymeric	Inorganic	Composite
Characteristics	- Made from polymers - Based on the operation temperature, it can turn from rigid in the glassy state to flexible in the rubbery state - Can be TFC or asymmetric	- Made from ceramics, zeolites, metals, or a mixture of these materials - Typically, asymmetric with a porous support and a dense active layer	- Hybrid of polymeric and inorganic materials - Particulate fillers embedded in a polymeric matrix - Combine the best properties of polymeric and filler material
Advantages	- Readily available - Relatively low cost - Well-developed technology - Good selectivity - Flexible and processable - Easy fabrication	- Can withstand harsh environments - Chemical and thermal stability - Mechanical strength - Longer lifespan compared to polymeric membranes	- Enhanced permeability and selectivity - Better fouling resistance - Can be tailored to a specific application - Enhanced mechanical properties
Disadvantages	- Limited operating temperature and pressure - Prone to fouling - Tradeoff between permeability and selectivity - Limited chemical and thermal stability	- High fabrication cost - Brittle - Difficult to process - Limited availability for largescale applications	- Interfacial defects - Complex fabrication process - More expensive

below provides a summarized comparison between the different membrane materials.

Biomimetic materials are promising novel materials for desalination membranes. Biomimetic materials have a similar action mechanism to water channels in biological membranes, i.e., molecular sieving. In molecular sieving, highly uniform, rigid pores reject ions by size exclusion as the diameter of the pores is smaller than that of hydrated salt ions [111]. Aquaporins (AQPs) such as human AQP1 and bacterial AQPZ that transport only water have received particular interest in biomimetic desalination membrane applications. AQPs are transmembrane proteins that can transport water across the membrane with high selectivity. AQP1 and AQPZ have been successfully integrated into lipid bilayers and/or block copolymer matrices. Integrating AQPs ensured they would maintain their water selectivity and help mimic natural water transport processes in desalination membranes. Although AQPs can achieve high water flux compared to conventional membranes, difficulties in their large-scale production might limit their widespread application, as even at lab-scale, sufficiently high-salt rejection has not yet been achieved by biomimetic desalination membranes [101,102,111].

7.1. Membrane Fabrication and Synthesis Techniques

Various techniques are available for membrane fabrication; the selection criteria for the fabrication method are the membrane properties (mechanical strength, chemical stability, hydrophilicity, and selectivity), membrane efficiency, and fabrication cost. A balance must be struck between these factors to ensure that the chosen fabrication technique produces a high-performing and cost-effective membrane [110]. Various techniques are available for membrane fabrication, such as [103,106,110]:

• Solution casting: this technique involves dissolving the polymer and any necessary additives in a suitable solvent until homogeneity is attained. The polymer solution is then spread onto a flat surface (e.g., a flat glass plate), followed by evaporation or phase inversion to remove the solvent. Once dry, the membrane is peeled off from the support plate, and a flat-sheet membrane is formed.
• Stretching: this is a solvent-free technique used to fabricate porous membranes. The polymer is first heated to its melting point, followed by extrusion and stretching to achieve the desired thickness and structure. The membrane is first stretched under cooler conditions to create nucleate structures. This is then followed by stretching at high temperatures, e.g., 130–140 °C, to further expand and refine the pores.
• Solution coating: a method suitable for making composite membranes as it involves depositing a thin layer of the polymer solution on the microporous substrate.
• Spray coating: here, a polymer solution is atomized and deposited onto a substrate, forming a thin, uniform membrane layer after solvent evaporation.
• Hollow fiber spinning: In this technique a polymer solution is extruded through a spinneret with a coaxial flow of bore fluid, forming continuous hollow fibers with a porous or dense selective layer. The solvent is then removed through phase inversion or thermal processing, allowing the fibers to solidify and create high-surface-area membranes.
• Interfacial polymerization: a technique that involves dissolving two reactive monomers separately in immiscible phases, one in an aqueous phase and the other in an organic phase. When these two phases come into contact, polymerization occurs at the interface, forming an ultrathin selective membrane layer. This technique is commonly used in TFC membranes for desalination.
• Electrospinning: an electrodynamic process in which the polymer solution is expelled through a needle under a high-voltage electric field. The ejected droplets then stretch and elongate, forming ultrafine fibers that are collected as a porous nanofibrous membrane with high surface area and tunable properties.

Various approaches have been used to integrate membranes into microfluidic devices, including direct incorporation of the membrane, membrane preparation as part of the microfluidic device fabrication process, in situ preparation of the membrane, and microfluidic membrane mimic [11]. The process for selecting and incorporating the membrane in the microfluidic device is detailed in Figure 7.

7.2. Enhancing Membrane Selectivity, Permeability, and Stability

Extensive efforts have been made to enhance the performance and stability of desalination membranes. The developed strategies include the following:

• Optimization of fabrication methods, where adjusting monomer concentrations and reaction conditions can help create more uniform polymeric membranes and improve salt rejection.
• Modification and post-modification of membrane materials: As mentioned above, incorporating nanomaterials into polymeric membranes can help improve selectivity, thermal, chemical, and mechanical stability. In addition, adding or grafting functional groups (e.g., carboxylate) onto the polymer can help increase water affinity. Acid- and alcohol-based surface modifications have also improved hydrophilicity, surface charge, and water flux. Improving membrane hydrophilicity has also been achieved by coating the membrane surface with more hydrophilic compounds, e.g., PVA and poly(N,N-dimethylaminoethyl methacrylate). These coatings have proven useful because they help improve resistance to chlorine attacks and reduce fouling effects [105]. Other surface modification techniques used to graft useful monomers on the surfaces of membranes include free-radical-, photochemical-, radiation-, plasma-induced grafting, and gas plasma treatment [104,105]. For instance, oxygen plasma treatment can help improve water permeability by introducing the hydrophilic carboxylate group. In contrast, argon plasma treatment can improve resistance to chlorine attacks by increasing the cross-linking at the nitrogen sites [105].
• Post-treatment of the membranes with acids and/or organic solvents [112]: post-treatment with solvents, known as solvent activation, is a promising method for improving the performance of desalination membranes. Aliphatic and aromatic alcohols, polar aprotic solvents (e.g., NMP, DMF, DMSO), and ionic liquids have been tested as activating solvents. While weak solvents (e.g., aliphatic alcohols) were able to provide minor permeability improvements, strong solvents (e.g., NMP, DMF, DMSO) were found to increase water flux significantly but compromised salt rejection and damaged membrane supports [112,113].

8. Important Advances in Microfluidic Desalination

Water scarcity is a growing global challenge, necessitating the development of efficient and sustainable desalination technologies. Microfluidic desalination has emerged as a promising approach due to its ability to process small volumes of water with high precision and energy efficiency. Recent advances have focused on enhancing desalination performance through solar energy utilization and the application of nanofluids. These innovations can potentially revolutionize water purification, making it more accessible and environmentally friendly.

Harnessing solar energy for desalination has gained significant attention due to its sustainability and cost-effectiveness. Several studies have explored using advanced photothermal materials to improve solar-driven water evaporation and purification. One such advancement is developing a mussel-inspired photothermal system, as demonstrated by Zhu et al. [114]. They focused on enhancing solar energy absorption through a composite material, CuO@PDA/PB, synthesized by integrating marigold-like CuO nanoflowers with Prussian Blue nanoparticles. The material exhibited broad-spectrum solar absorption (300–2500 nm) and demonstrated an evaporation rate of 1.39 kg m^–2h⁻¹ with 87.10% efficiency under 1.0 sun illumination. The superhydrophilic variation of this material was particularly effective, displaying excellent salt tolerance and antifouling properties, making it suitable for both desalination and wastewater treatment.

Another breakthrough in solar desalination is the development of photocorrosion-based BiOCl photothermal materials, as explored by He et al. [115]. They introduced a novel BiOCl material capable of driving water evaporation and storing and releasing solar energy. The dual-function material solved a critical limitation of conventional photothermal systems by enabling desalination to continue beyond direct sunlight exposure. The BiOCl material exhibited strong solar absorption, high thermal efficiency, and stability against photo corrosion, ensuring long-term application viability. These findings revealed that integrating solar desalination with energy storage can significantly enhance water purification reliability in off-grid and remote locations.

Furthermore, nanofluids, which are engineered colloidal suspensions containing nanoparticles, have demonstrated significant potential in enhancing microfluidic desalination efficiency. The high thermal conductivity and superior heat transfer properties of nanofluids contribute to improved water evaporation rates and energy utilization [116,117,118]. Recent research has focused on incorporating metal oxide nanoparticles such as titanium dioxide (TiO₂), silicon dioxide (SiO₂), and graphene-based materials into microfluidic desalination systems. These nanoparticles enhance the photothermal conversion process, allowing for more effective solar-driven desalination. Additionally, nanofluids help mitigate scaling and fouling issues, prolonging the lifespan of desalination membranes and reducing maintenance requirements.

8.1. Evaluation of Solar-Energy-Powered Seawater Desalination Processes

Al-Obaidi et al. [119] assessed the feasibility of integrating solar energy into seawater desalination. The study examined different solar desalination technologies, including concentrated solar power (CSP) and photovoltaic (PV)-driven systems. The authors showed that CSP-based desalination offers high efficiency but requires significant capital investment and infrastructure, whereas PV-based systems are more modular and scalable. The study also explored emerging hybrid systems that combine solar thermal energy with reverse osmosis (RO) to improve overall energy efficiency. These findings identify the importance of selecting appropriate solar desalination methods based on regional energy availability, economic considerations, and technological feasibility. An emerging approach to desalination integrates hydrogen production through solar electricity. This dual-purpose strategy optimizes freshwater generation while utilizing waste heat from electrolysis to enhance thermal desalination efficiency.

A study by Arunachalam and Han [120] analyzing polymer electrolyte membrane (PEM) electrolyzers demonstrated that coupling desalination with hydrogen production significantly reduces operational costs. The system used low-cost solar and wind electricity to ensure high economic feasibility while producing deionized water essential for hydrogen generation. Their findings revealed that operating electrolyzers at high current densities maximized both hydrogen output and thermal energy recovery, showing the potential of this integrated system for sustainable clean energy and water production.

Similarly, Wani et al.[121] explored an innovative microfluidic salinity gradient system to generate freshwater and electricity simultaneously. Their design involved a two-legged paper-based system, where seawater and tap water flowed through separate channels, leveraging ion concentration gradients to continuously generate electricity. The study demonstrated an open-circuit voltage of 150 mV per channel, increasing to 250 mV with higher salt concentrations. A multi-channel setup achieved a maximum power density of 9.9 mW/m² with 88% evaporation efficiency while effectively harvesting salt at a 0.33 kg/m²/h rate. The findings revealed that microfluidic salinity gradient technology could provide a cost-effective and renewable solution for water purification, energy production, and salt recovery.

Additionally, Kaur and Nagao [122] investigated material-based approaches that integrate solar desalination with hydroelectric power generation. They focused on applying three-dimensional porous media, nanotubes, and carbonized wood structures to enhance water evaporation and electricity generation. One notable innovation was a wood-based hydroelectric generator (WHEG) utilizing carbonized wood infused with FeCl₃, achieving a voltage output of 96 mV and a short-circuit current of 10.5 mA. Increasing FeCl₃ concentrations further improved energy output, though performance declined with excessive water vapor accumulation. This research identifies the feasibility of using nanostructured materials to generate freshwater and electricity, paving the way for sustainable desalination solutions integrated with energy harvesting technologies.

These advancements in microfluidic desalination show the potential of combining solar energy, AI-driven optimization, and nanomaterial innovations to enhance water purification. Emerging technologies such as hydrogen production integration, salinity gradient systems, and hydroelectric desalination offer promising pathways to more sustainable and efficient desalination processes. In addition, by optimizing material properties, leveraging renewable energy, and applying intelligent control mechanisms, future desalination systems could significantly contribute to global water security and energy sustainability.

8.2. Nanofluids in Microfluidic Desalination

The recent research has focused on incorporating metal oxide nanoparticles such as titanium dioxide (TiO₂), silicon dioxide (SiO₂), and graphene-based materials into microfluidic desalination systems. Luo et al. [123] conducted a study to enhance the efficiency of desalination by modifying graphene oxide (GO) membranes. Traditional 2D lamellar membranes suffer from inconsistent channel sizes, leading to inefficient salt rejection. To solve this, the researchers incorporated graphitic carbon nitride (g-C₃N₄) sheets into GO membranes to form heterostructured nanofluidic channels, improving salt rejection and water permeability while ensuring long-term stability. The study employed a lamellar assembly process to synthesize the GO/g-C₃N₄ composite membranes, creating sub-nanometer-wide (0.42 nm) channels. Various experimental techniques were used to measure salt rejection and water permeability under controlled pressure conditions. Conductivity measurements helped analyze ion exclusion efficiency, while molecular dynamics simulations provided insights into the underlying desalination mechanisms, such as size exclusion and friction reduction. Long-term stability tests assessed the membranes’ resistance to oxidation and mechanical pressure. Findings from the study demonstrated that the introduction of g-C₃N₄ significantly enhanced desalination performance. The composite GO/ g-C₃N₄ membrane achieved a salt rejection rate of approximately 90%, a remarkable improvement compared to the pure GO membrane’s less than 30% efficiency. Moreover, water permeability increased dramatically to over 30 L h⁻¹ m⁻² bar⁻¹, while pure GO membranes only reached 4 L h⁻¹ m⁻² bar⁻¹. These enhancements were attributed to uniform and stable nanofluidic channel formation, facilitating ultralow-friction water flow and effective ion exclusion. Theoretical modeling corroborated these results, reinforcing that size exclusion was dominant in salt rejection. Additionally, the GO/g-C₃N₄ membrane exhibited strong oxidation and mechanical stress resistance, making it a promising candidate for long-term desalination applications.

Likewise, Iqbal et al. [124] examined the effectiveness of nanofluids in enhancing solar still desalination, aiming to improve efficiency for remote and underdeveloped communities. The study sought to optimize heat transfer and increase productivity in solar stills by leveraging the superior thermal properties of nanofluids. The authors conducted a comprehensive review of nanofluid-assisted desalination techniques employed over the past decade. They focused on nanoparticles’ thermal and plasmonic effects, analyzing their role in enhancing energy transfer and boosting water output. The study also compared different nanofluids and evaluated improvements in solar still designs. The findings revealed that nanofluids significantly improved the thermal conductivity of base fluids, leading to increased water evaporation and condensation rates. Among the tested nanoparticles, alumina emerged as the most cost-effective option, while carbon nanostructures exhibited superior thermal conductivity. The integration of nanofluids into solar stills was found to improve desalination efficiency substantially. However, challenges such as the long-term stability of nanoparticles and cost considerations pose potential limitations to widespread adoption. Furthermore, the study emphasized that nanofluid-assisted solar still desalination presents a promising solution to global water scarcity. By improving heat transfer efficiency, these advanced systems could make desalination more viable for resource-limited settings. However, future research should optimize nanofluid properties to ensure stability and cost-effectiveness. Additionally, integrating solar stills with other renewable energy sources could enhance sustainability. These findings show the importance of continued innovation in desalination technologies to solve the growing global demand for fresh water.

Singh et al. [125] reviewed the role of nanofluids and renewable energy in advancing sustainable desalination. The study aimed to solve the limitations of conventional desalination methods by exploring the potential of nanotechnology and renewable energy sources. The authors conducted an extensive literature review on integrating nanofluids into various desalination techniques, including membrane distillation, reverse osmosis, and solar desalination. Additionally, they examined the feasibility of combining these methods with renewable energy sources such as solar, wind, and geothermal power. The study evaluated the impact of nanofluids on key desalination parameters, including thermal efficiency, salt rejection rates, and energy consumption.

The findings indicated that using nanofluids significantly improved desalination performance across various technologies. Nanofluids enhanced heat transfer rates in membrane distillation, leading to increased water flux and energy efficiency. Nanomaterial coatings improved membrane durability and fouling resistance in reverse osmosis, extending operational lifespans. Solar desalination systems saw marked improvements in evaporation rates and water output when integrated with plasmonic nanofluids. In addition, the study showed that coupling desalination with renewable energy sources reduced overall energy consumption and operational costs, making the process more sustainable. Despite these advantages, the authors acknowledged several challenges, including the high cost of nanomaterials, potential environmental concerns related to nanoparticle disposal, and the need for further research on long-term stability, that remain key obstacles. Nonetheless, the findings identify the transformative potential of nanotechnology and renewable energy in desalination.

Furthermore, Parmar et al. [126] investigated the use of nanofluids to improve the energy efficiency of MD for thermal desalination. They aimed to enhance heat transfer properties in MD systems by incorporating nanomaterials into working fluids, solving the need for more efficient desalination methods. They examined the effects of different nanofluids in gap-based MD configurations, specifically copper oxide and carbon nanotube-based solutions. The methodology included nanofluid preparation through sonication and surfactant stabilization, experimental evaluation of micro-mixing effects via Brownian motion, and comparative analysis of different MD configurations such as air gap membrane distillation (AGMD), permeate gap membrane distillation (PGMD), and conductive gap membrane distillation (CGMD).

The findings showed the significant enhancement of heat transfer efficiency with nanofluid incorporation. Carbon nanotube nanofluids consistently demonstrated superior performance, while copper oxide nanofluids showed diminishing returns beyond a 0.7% concentration. Nanoparticles promoted micro-mixing through Brownian motion, improving thermal conductivity and overall system efficiency. Heat exchangers utilizing nanofluids exhibited efficiency improvements of 15–41% when nanofluid concentrations ranged from 0.1–2%. However, challenges such as high surfactant levels causing membrane fouling and hydrophobicity loss were identified, particularly in gap-based MD systems. The study concluded that while carbon nanotubes provide superior performance enhancements across various conditions, copper oxide nanofluids are beneficial at lower concentrations. Careful optimization of nanoparticle concentration is necessary to mitigate membrane fouling issues, ensuring practical applications of nanofluids in MD systems.

9. Challenges and Future Directions

One of the primary challenges in adapting microfluidic desalination technology for macroscale applications is the significant disparity in flow rates. Current microfluidic devices typically operate at flow rates of several microliters per minute or milliliters per day. However, a substantial scale-up is required to meet requirements ranging from liters to several hundred liters daily. Addressing this limitation involves integrating multiple desalination units on a single chip, utilizing a singular inlet and outlet, and designing these chips to be stackable [127]. For instance, achieving a daily water supply of 3 L per person at a production rate of 5 μL/min would necessitate approximately 416 chips, collectively occupying an estimated volume of 0.25 L. Recent advancements have introduced innovative design solutions to enhance the scalability of these systems. Lo et al. [128] have explored an “out-of-plane” design that expands the chip architecture into a third dimension, potentially increasing the production capacity per unit volume. Additionally, Zhang et al. [129] developed techniques for creating multi-layer PDMS structures for 3D microfluidic chips. The proposed methods were fast, repeatable, and scalable, facilitating the mass production necessary for larger-scale desalination. These innovative design considerations are essential for implementing microfluidic desalination techniques at a scale suitable for larger populations, bridging the gap between micro-scale experimentation and macro-scale water needs. Such developments promise to enhance the efficiency of microfluidic desalination systems and pave the way for their practical application in addressing global water scarcity.

Moreover, continuous operation in desalination environments, especially those involving seawater, exposes microfluidic devices to high fouling potential, which can significantly degrade their performance. Membrane fouling (Figure 8) refers to the deposition and accumulation of organic particles, inorganic particles, or microorganisms (biofoulants) on the surface of or between the pores of desalination membranes. Microfluidic channels and membranes are prone to fouling due to the accumulation of salts, biological materials, and other impurities. This fouling phenomenon is exacerbated by the high surface-area-to-volume ratio typical of microfluidic devices [10].

Membrane fouling remains a key challenge, as it reduces the effective lifetime of the system and necessitates frequent maintenance or replacement. It also causes flowrate reduction, decreasing the overall performance and increasing energy requirements and operational costs [130,131]. Although membrane fouling is unavoidable in membrane processes, several ways can be used to help reduce it. Developing materials that resist fouling and mechanisms to reverse or mitigate fouling effects are critical for innovation.

The use of nanomaterials, like graphene or advanced polymers, provides breakthroughs in membrane technology, enhancing desalination efficiency and device longevity[8,130,132]. Studies have also reported that the surface properties of microfluidic devices can be tuned to reduce fouling. Microfluidic devices made out of polymeric materials such as PDMS and PMMA can readily be modified by plasma oxidation, polyelectrolyte multilayer coating, or laser radiation-grafting [10].

For example, Olubunmi Ige et al. [133] deposited GO on a paper substrate embedded in a microfluidic device to reduce salt-ion concentration. When tested with a 0.6 N NaCl solution, the five-layer GO-coated membrane achieved the best salt-rejection capacity of 97%. The same device showed 83.3% salt rejection when tested with a real seawater sample. However, this tends to introduce extra steps during manufacturing, increasing costs. Another approach is to use hydrodynamic techniques, such as pulsatile or turbulent flows, to push foulants away from the membrane surface [12,13]. Periodic physical and chemical cleaning protocols are also essential in combating fouling. Despite the continuous efforts to combat fouling, it remains an unavoidable limitation in both membrane separation processes and microfluidic devices, requiring further research [10,12,13].

Recent advancements in microfluidic device fabrication have also significantly enhanced the complexity and affordability of such devices. Innovative fabrication techniques, including the utilization of photo-patterned hydrogel membranes and additive manufacturing (3D printing), have facilitated the production of devices with greater complexity and resolution [134].

Three-dimensional printing also improves traditional fabrication methods, e.g., soft lithography, which allows for more robust devices and intricate flow regulation components. These developments refine the manufacturing process, expand the functional capabilities of microfluidic devices in real-world settings, broaden the field of applications, and help develop new separation methods [10]. With respect to membrane fabrication, current 3D printing methods cannot achieve the required resolution; therefore, a hybrid approach combining 3D printing and electro-spraying could produce highly porous membranes more efficiently [135,136,137].

While microfluidic desalination devices offer reduced energy consumption compared to traditional methods, the total energy footprint, including the energy for manufacturing and maintaining these systems, remains a concern. To minimize their environmental impact, further research is needed to create more energy-efficient designs and possibly integrate renewable energy sources directly with desalination systems. Linking microfluidic desalination systems with solar, wind, or other renewable energy sources could dramatically increase their sustainability and deployment in off-grid locations. Finally, developing compact, integrated systems that efficiently convert and store energy would be particularly beneficial for remote or disaster-stricken areas where fresh water is urgently needed [137,138,139].

10. Conclusions

Microfluidic desalination presents a promising approach to address global water scarcity by offering compact, energy-efficient, and potentially scalable solutions. While several technical challenges remain, ongoing innovations in materials, device design, and energy integration continue to develop the field forward. With continuous research and development, microfluidic desalination systems can potentially become viable tools for sustainable and decentralized water purification in the near future.