Biorefinery perspectives of microbial electrolysis cells (MECs) for hydrogen and valuable chemicals production through wastewater treatment

The degradation of waste organics through microbial electrolysis cell (MEC) generates hydrogen (H2) gas in an economically efficient way. MEC is known as the advanced concept of the microbial fuel cell (MFC) but requires a minor amount of supplementary electrical energy to produce H2 in the cathode microenvironment. Different bio/processes could be integrated to generate additional energy from the substrate used in MECs, which would make the whole process more sustainable. On the other hand, the energy required to drive the MEC mechanism could be harvested from renewable energy sources. These integrations could advance the efficiency and economic feasibility of the whole process. The present review critically discusses all the integrations investigated to date with MECs such as MFCs, anaerobic digestion, microbial desalination cells, membrane bioreactors, solar energy harvesting systems, etc. Energy generating non-biological and eco-friendly processes (such as dye-sensitized solar cells and thermoelectric microconverters) which could also be integrated with MECs, are also presented and reviewed. Achieving a comprehensive understanding about MEC integration could help with developing advanced biorefineries towards more sustainable energy management. Finally, the challenges related to the scaling up of these processes are also scrutinized with the aim to identify the practical hurdles faced in the MEC processes.


Introduction
One of the biggest global challenges that needs to be addressed is to meet the growing energy demands. Due to the greenhouse gas (GHG) emissions dominantly from the usage of fossil fuels during past centuries, the global temperature has risen unfavorably. In the year 2015, the Paris Agreement set the target to maintain the global average surface temperature below 2 °C with reference to the pre-industrial period (Schleussner et al., 2016). This "safe" level requires to retain over 30%, 50% and 80% of oil, gas, and coal reserves, respectively, unused before 2050 (McGlade and Ekins, 2015). Therefore, developing renewable and eco-friendly alternatives to fossil fuels is essential (Kadier et al., 2016a).
Hydrogen (H2) is considered as a clean and sustainable energy carrier, and has a net calorific value of 119.9 kJ/g which is much higher than those of the other fuels like coal (29.0 kJ/g), petroleum (43.4 kJ/g), ethanol (26.7 kJ/g), etc. (Bartels et al., 2010;Njenga et al., 2014). Moreover, during ignition, H2 generates only water which is not harmful to the environment. The potential applications of H2 as future energy are tremendous but still challenging. One of the major issues is the low content of H2 in nature but the demand will be billions of tons per year to build a full hydrogen economy (Turner, 2004).
Today, H2 is mainly produced from unsustainable natural gas and coal Zhen et al., 2017). Other sources include thermolysis or electrolysis of biomass and water (Turner, 2004). These resources are renewable, but require high energy due to the high temperature or usage of electricity (Kadier et al., 2016a). Therefore, biological H2 production from renewable sources, which requires lower energy inputs, offers great potentials to meet future demands (Azman et al., 2016). To date, the most promising biological process for H2 production is dark fermentation (Khanna and Das, 2013). This process can be catalyzed, either by pure or mixed microbial culture (Rittmann and Herwig, 2012), and the substrate can be either pure carbohydrate or organics from diverse wastes (Ren et al., 2011). Furthermore, the specific and volumetric H2 production rate (HPR) can reach over 40 mmolH2/g/h and 40 LH2/L/d, respectively (Kumar and Das, 2001; Rittmann and Herwig, 2012). However, this process has some bottlenecks that restrict the upper boundary of H2 yield. In this process, the yield is relatively low (< 4 mol/mol of glucose) and the obtained H2 gas is always mixed with the other anaerobic digestion (AD) gas (e.g., CO2), necessitating further purification. These factors increase the overall cost of dark fermentation and thus, limit its industrial applications (Venkata Mohan et al., 2008; Mohanakrishna et al., 2010).
In the past decade, a novel bio-technique named microbial electrolysis cells (MECs) were developed for sustainably producing H2 from wastes (Liu et al., 2005;Rozendal et al., 2006). It Varanasi et al., 2019). In principle, MECs could overcome the fermentative barrier observed in the fermentation-based processes and exhibit a high theoretical yield of H2 (Kadier et al., 2016a). Moreover, MECs show advantages in terms of vast substrate diversity and low energy input, compared to the other H2 production routes . The future of MECs is promising but many issues need to be addressed before their commercialization. System optimization need to be conducted to increase the rate and reduce the cost. Moreover, due to the high modularity of the MEC system (anode, cathode, reactor structure, etc.), integrating with other existing techniques could also bring additional benefits beside the value of H2, which could potentially improve the economic feasibility of MECs and facilitate their commercialization. In this review, we briefly introduce the status quo of knowledge and discuss the pros and cons of MECs. Subsequently, the focus is placed on addressing the potentials of integrating MECs with other existing technologies for fuels and chemicals production.
on the Nernst equations, the cathode potential will be affected by the concentration of free protons (i.e., pH of electrolyte), temperature, and partial pressure of H2. As presented in the right side of Figure 1, the cathode potential decreases with increases in temperature, pH value, and partial pressure of H2. With respect to influence on cathode potential, pH is the most critical parameter (Rozendal et al., 2006). The energy input is essential for MECs because of the thermodynamic limitations (Liu et al., 2005;Call and Logan, 2008). Under standard conditions (pH of 7, H2 partial pressure (pH2) of 1 bar, and temperature of 25°C), the cathode potential for H2 evolution reaction is about -0.41 V (Rozendal, 2006) which is more negative than the anode potentials using most of the carbon sources listed in Figure 1. Theoretically, it should be possible to make the reaction happen spontaneously with glucose as substrate, but glucose in that case has to be fully oxidized to CO2 which does not happen in anaerobic fermentation (Wünschiers and Lindblad, 2002). Another approach to reduce or remove the energy requirement of MECs is to increase the cathode potential by changing the pH or pH2. This appears feasible since the cathode potential would be high enough if the cathodic electrolyte pH could be reduced to lower than 5 or maintain the pH2 at lower than 0.001 bar (see Fig. 1). However, maintaining a low pH in cathode would block the proton migration through the membrane (Rozendal et al., 2006), subsequently resulting in decreased pH in anode which is harmful to the microorganisms. On the other hand, maintaining an extremely low pH2 is impractical during the operations.

Fundamentals of microbial electrolysis cells
The first study on MECs was published by Liu et al. (2005). Since then, exponentially growing numbers of papers were published (Kadier et al., 2016b;Lu and Ren, 2016). Basically, the exoelectrogens (i.e., the bacteria which can transfer electrons through extracellular mechanism) in MECs use anode as an electron sink to oxidize organics and produce protons and electrons. The protons diffuse via a proton exchange membrane (PEM) to cathode and are reduced by the electrons transferred via the external electric circuit to produce H2. The MEC system cannot work spontaneously due to the thermodynamic barrier and additional energy is required to drive the reduction reaction (Kadier et al., 2015;Kumar et al., 2017). A typical schematic representation of MECs is shown as Figure 1.
In theory, any carbon compounds that can be digested by the exoelectrogens, can be used in anode of MECs. This has led to a large variety of organics sources studied in MECs, ranging from defined pure chemicals to mixture of real wastewaters Escapa et al., 2016). With different carbon sources, the anode potential can range from 0.2-0.5 V. The typical values of some commonly used carbon sources are shown in left side of Figure  1. The half-reaction at the cathode is the reduction of protons to H2 gas. Based

Advantages and disadvantages of MEC technology
Compared with other H2 producing techniques, one of the advantages of MECs is mild operating conditions. In principle, only 0.11 V input is needed to drive the H2 production from acetate (Liu et al., 2005;Kadier et al., 2016a), and this is less than 10% of the typical power (typically 1.23-2.0 V) required for water electrolysis (Kadier et al., 2015). Thus, the energy cost of H2 production in MEC is about 1-3 kWh/m 3 (Call and Logan, 2008), while the conventional industrial electrolysers would cost up to 4.5-5.0 kWh/m 3 . In addition to the cost, the yield of H2 is much higher in the MEC process, compared with the fermentation-based processes (Kadier et al., 2016a). Due to the energy barrier discussed above, the stoichiometric yield of H2 is the dark fermentation process is only 4 mol H2/molglucose (Venkata Mohan et al., 2007 and, and the typical values reported in real studies are 2.5-3 mol H2/molglucose (Ren et al., 2011). Many by-products (acetate, butyrate, etc.) are also produced during this process and their conversion into H2 gas is not thermodynamically feasible. Generation of these by-products during dark fermentation also indicates the incomplete degradation of organics during this process. However, if those substances (e.g., acetate, butyrate) can be completely oxidized, the H2 yield could then be dramatically improved . Stoichiometrically, this would add up 4 mol H2/molacetate and could bring up the total yield of H2 from glucose close to its theoretical limit of 12 mol H2/molglucose. Moreover, H2 production in dark fermentation is significantly affected by pH2 and would become thermodynamically unfavorable if H2 is highly accumulated in the gas phase (Khanna and Das, 2013). On the contrary, the effect of partial pressure on cathode potential in MEC is found to be minor and this factor will not alter the performance of MEC system. Furthermore, MECs are superior in terms of the purity of H2 as well. Due to the spatial separation of organic degradation and H2 production in MECs, it is likely to have other gases such as carbon dioxide and other fermentation gases in minor proportions (Sravan et al., 2019). The purity of H2 is essential for its various downstream applications, for example, it has to be sulfur-free to be used in proton exchange membrane fuel cell systems.
Overall, MECs present a bright future for superior H2 production and yield, with high purity at a low energy cost. However, its commercialization is strongly limited by HPR. More specifically, the HPR in MECs at laboratory scale could only reach about 3 m 3 H2/ m 3  . The lower hydrogen production and lower current density recorded with pilot reactors, compared to the laboratory reactors were due to several differences between these systems. These differences can be listed as reactor geometry, electrode materials, inclusion of glass fiber separators, possible connection resistances, and microbiological factors resulting in relatively slow startup of the reactor. Methane production is also one of the possibilities for the lower HPR at pilot scale (Cusick et al., 2012). Maximum volumetric HPR in small-scale (volume < 100 mL) MECs could reach 50 m 3 H2/m 3 /d (Lu and Ren, 2016). Therefore, MECs seem to be more suitable as decentralized systems for H2 production at domestic sites. Integrating MECs with other existing techniques to generate extra benefits could also potentially be an efficient approach to improve their industrial feasibility in the future.

MEC-microbial fuel cell (MFC) coupled system
Microbial fuel cells (MFCs) are systems used to produce power by degrading various wastes (Deval et al., 2017). In recent studies, MFCs were proposed as renewable power sources to operate MEC. The concept was demonstrated using a novel system integrating a single chambered MFC and a dual chambered MEC (MEC-MFC-coupled system) and was successfully used for H2 production without external power supplementation. Peak systemic H2 yield of 1.21 mol H2/molacetate and HPR of 14.9 ± 0.4 mL/L/d were achieved using acetate as substrate with the developed MEC-MFC-coupled system ( Fig.  2) (Sun et al., 2008). Later, a bio-photo-electrochemical cell (BPEC) design was proposed by Wan et al. (2015), also comprising MFC and MEC. Illumination of BPEC photocathode with the visible light led to H2 production, and the MFC supplied the voltage for electrolysis. The process produced H2 at the rate of 1.35 ± 0.15 mL/h and registered a current density of 0.68 A/m 2 ( Table 1) (Wan et al., 2015).
Apart from H2 generation, MFCs can be used to power virtually any MEC systems (Wan et al., 2015). Accordingly, a group of researchers proposed the use of MFC to supply power to an MEC used to convert carbon dioxide to formic acid and obtained a significant production rate (21.0 ± 0.2 mg/L/h) (Zhao et al., 2012). The energy generated from the wastewater treatment or acetate oxidation in MFC can also be integrated with MEC to drive H2 production or CO2 reduction to value-added products (termed as microbial

Submersible microbial electrolysis cell (SMEC)
Submersible microbial electrolysis cells (SMECs) were designed in an effort to remove the external power supply required for the operation of a typical MEC. These cells can be easily fixed into existing anaerobic digesters, where two jointed chambers are inserted functioning as two cathode chambers. There is no separate anode chamber and one of the cathode chambers is used to produce electricity, while the other chamber is designed to produce H2. In an innovative study, Zhang and Angelidaki (2012) introduced a self-powered SMEC to produce H2 in-situ from anaerobic digesters. The highest systemic yield of H2 achieved was 1.43 mol H2/molacetate with 20 mM acetate, where the maximum current density and coulombic efficiency were registered at 1778 mA/m 2 and 28%, respectively (Zhang and Angelidaki, 2012) ( Table 1). This same group of researchers also developed a similar technology called submersible MFC (SMFC) as a biosensor to monitor microbial activity and biological oxygen demand (BOD) in groundwater (Zhang and Angelidaki, 2011). A mature anodic biofilm was used in the MFC to measure BOD, wherein the current production was correlated with BOD of the water sample. Since, type of the inoculum and its concentration influences the bacterial cell adhesion to the electrode surface, a fresh anode was used to evaluate the microbial activity (Zhang and Angelidaki, 2011). In another study, ammonia inhibition in AD was overcome by a novel hybrid system consisting of a submersible microbial desalination cell (SMDC) and a continuous stirred tank reactor (CSTR). This integration not only resulted in in-situ ammonia recovery and electricity production but also led to 112% additional biogas production (Zhang and Angelidaki, 2015). Another variation of submersible cells was also proposed later, called submerged microbial desalination-denitrification cell (SMDDC). It was used for in-situ removal of nitrate from groundwater and produced electric energy. Additionally, added value was created in electrosynthesis (MES) (Mohanakrishna et al., 2016 and2018;Roy et al., 2016). As the integration is possible in-situ, energy losses can be minimized significantly. However, as MEC requires a constant reduction potential, therefore, more stable MFC systems need to be developed for a balanced integration.   . 3a) (Zhang and Angelidaki, 2013). Bioelectricity was produced by the anodic bacterial activity. Both NO3and Na + were ion-transported to the anode and cathode by anion (AEM) and cation exchange membranes (CEM), respectively. The effluent from the anode was directed to the cathode, where NO3was reduced to nitrogen gas by autotrophic denitrification. Subsequently, SMDDC removed 90.5% of the nitrate from groundwater with wastewater hydraulic retention time (HRT) of 12 h (Zhang and Angelidaki, 2013). Submerged electrolysis cells configuration demonstrated the potentiality of the integration of the MECs with anaerobic digesters and MDCs. The proof of concept studies also proved that submersible configurations could also offer simultaneous economic bioremediation of wastewater and energy recovery.

Solar powered MEC
Solar energy is renewable and is freely available all over the world and thus, is considered as a favorable option to meet the world energy demands. Therefore, the energy required to run MECs could also be potentially derived from solar energy in the form of a dye-sensitized solar cell (DSSC). Solar cell-MEC-coupled system converts solar energy to liquid or gas transportation fuels such as H2, CH4, and ethanol, which could be subsequently stored for future use ( (Table 1). Also, as reported by other researchers, H2 production in DSSCs was similar to that obtained from MECs with a conventional power supply (Fig. 3b). The substrate to product and that it can be met with light of moderate intensity, the microbially catalyzed anode becomes the limiting factor for such systems. Fischer et al. (2018) argued that the application of solar powered MECs would be a cheaper way to produce hydrogen than most other comparable processes. Different photocatalytic materials enable the hydrogen evolution reaction. Materials like TiO2 nanorods, Cu2O, and Cu2O/NiOx composite were evaluated for the effective light supported hydrogen evolution reaction in MECs. However, more research is required to transform the currently available solar powered MECs to commercial hydrogen production platforms (Fischer et al., 2018).

Dark fermentation and MFC-MEC coupled system
Biological H2 production from dissolved organic materials in wastewaters provides an opportunity to utilize this untapped resource via dark fermentation. However, there are many thermodynamic barriers which pose technical challenges to obtain favorable yields through the dark fermentation process (Rozendal et al., 2006). Due to these thermodynamic limitations, several byproducts like acetate and butyrate are formed instead of H2, and external energy needs to be supplied for a thermodynamically feasible reaction ( Table 1). The overall H2 production for the integrated system comprising a dark fermenter, MFC and MEC (in series) resulted in 41% higher H2 production than the fermentation alone. Thus, without using an external electrical supply, higher H2 yield was achieved using this combined fermentation and MFC-MEC.

Integration of pyrolysis-MEC
Fuel sources produced from the biomass origin by different biological and thermochemical processes are named renewable hydrocarbon biofuels. These fuels are also termed as green-hydrocarbons, bio-hydrocarbons, drop-in biofuels, and sustainable or advanced hydrocarbon biofuels (Fig.  4). Pyrolysis is a process, where thermochemical decomposition of organic materials occurs at high temperatures in the absence of O2 and is used for the production of these bio-hydrocarbons (Borole, 2015). The high O2 content in biomass (> 40%), lowers the yield of bio-hydrocarbons from them. This problem is often handled by supplying the fuel finishing step with a source of H2. Due to the current unavailability of commercial renewable sources of H2, fossil resources are required for its production, which in turn, impact its sustainability. Alternatively, H2 can be derived from the liquid/aqueous waste streams produced during pyrolysis via microbial electrolysis, thus offering a sustainable source of H2 (Borole, 2015). The blend of organic and aqueous phases generated through pyrolysis is known as bio-oil. The organic content of liquid phase bio-oil includes organic acids, furan aldehydes, phenolic compounds, and sugar derivatives. These organic compounds are reported as potential bioanode substrates for H2 production in MECs. The aqueous stream generated during pyrolysis of switchgrass was used as substrate for H2 production in MEC, reaching an overall energy efficiency of 48-63% (Lewis et al., 2015; Table 1). Maximum HPR of 4.3 L H2/Lanode/d (loading of 10 g COD/Lanode/d) was recorded through (approx.) complete conversion of acetic acid, propionic acid, levoglucosan, and furfural. Thus, the H2 produced from the integrated process could be utilized to hydrodeoxygenate bio-oil to produce fuel (

Microbial reverse-electrodialysis electrolysis cells (MRECs)
Reverse electrodialysis (RED) generates power from the salinity gradient that develops between sea water and fresh water using a series of CEM and AEM. The chemical potential difference prevailing between salt and fresh water generates a voltage over individual membranes and the total potential of the system is the summation of the potential differences from all membranes. Typically, RED systems use many stacked cells to have a substantial energy recovery. The process ensuing high capital expenses for the large number of membranes, and increases energy losses from pumping water through a large number of cells designed in the process. Integrating MECs with the RED system forms MREC, in which high overpotentials can be overcome through the oxidation of organic matters by anodic biocatalyst while the low voltage prevailed in MFCs can be increased due to the salinity driven potential with the RED stack (Zhang and Angelidaki, 2014). Cusick et al. (2012) reported the use of MRECs to capture salinity-gradient energy from thermolytic ammonium bicarbonate solutions generating low waste heat (> 40°C). Capturing salinitygradient energy from such thermolytic solutions removes the dependency of this process on seawater and freshwater availability. However, the limitation associated with MREC stack arrangement with NH4HCO3 is nitrogen crossover from the stack into the anode chamber resulting in contamination of anodic solution with ammonia and thus, loss of the salt solution. Designing future MRECs with bipolar membranes or a low-salt solution in the membrane stack nearest to the anode can help to minimize the above stated losses. In a study, maximum energy recovery with acetate reached 30 ± 0.5% with a power density of 5.6 W/m 2 (with respect to cathode surface area), which was five times that produced without the dialysis stack (Cusick et al., 2012). Furthermore, as a solution to the nitrogen crossover problem, Wallack et al. (2015) proposed the use of additional low concentration chamber before the anode using an additional AEM adjacent to similar AEM, and filled with varied amounts of both anion or cation ion exchange resins. N2 crossover to the anodic chamber was reduced by up to 97% using 50% of the chamber filled with an anion exchange resin than control, in which no additional chamber was present. Moreover, loss of power in the MREC due to this additional chamber near the anode could be overcome by placing a pair of additional membranes to enhance the stack voltage (Wallack et al., 2015). Another variation of MREC, called microbial reverse-electrodialysis chemical-production cell (MRCC) was designed by Zhu et al. (2013). MRCC produced acid and alkali by utilizing the energy produced from organic matter (acetate) and salinity gradients (seawater and river water simulated with different concentrations of sodium chloride). No external power supply was required because this system itself produced the stipulated electricity (908 mW/m 2 ). Fed-batch cycle operation resulted in the production of 1.35 mmol of acid (pH, 1.65 ± 0.04) and 0.59 mmol of alkali (pH, 11.98 ± 0.10) (Zhu et al., 2013).

Integration of MEC with anaerobic digestion (AD)
AD is carried out through the anaerobic bacterial/archaeal metabolisms breaking down biodegradable materials and generates methane gas. This is a breaking down biodegradable materials and generates methane gas. This is a well-known technology and has been commercially used for simultaneous methane generation and waste/wastewater treatment (Batstone and Virdis, 2014). Electrochemical technologies and AD can be integrated for increased efficiency with the stillage released from the latter process used in the former as feedstock and energy resource recovered from the former used in the latter (Sadhukhan et al., 2016). Bo et al. (2014) inserted a pair of MEC electrodes in an anaerobic digester and generated H2 at the cathode which subsequently reacted with carbon dioxide to produce methane in-situ by the action of hydrogenotrophic methanogens. Compared to the conventional anaerobic digesters, the MEC-anaerobic digester system achieved higher methane content (up to 98%). In this system, along with 24-230% improvement in methane yield, substrate (COD) degradation and carbon recovery were also enhanced by 130-300% and 55-56%, respectively, compared with the conventional AD reactor (Bo et al., 2014) ( Table 1). In another study, an up-flow anaerobic bioelectrochemical (UABE) reactor was compared with an up-flow anaerobic sludge blanket (UASB) reactor for the treatment of acidic distillery wastewater to produce methane (Feng et al., 2016). The UABE was poised at 300 mV and both UASB and UABE were operated in continuous mode. The integration, i.e., UABE, resulted in an improved CH4 yield of 407 mL/g CODr at 4.0 g COD/L.d, which was significantly higher than that of UASB (282 mL/g CODr) (Feng et al., 2016). De Vrieze et al. (2014) tried to evaluate the plausible mechanism behind the improved performance of AD by inserting an MEC system into the reactor. They reported increased stability of the AD treating molasses by inserting electrodes and poising them at 0.75 and 1.205 V. In the control reactors, CH4 production reduced to 50% of the initial rate (on day 91), while it remained stable in the MEC-AD reactors indicating a stabilizing effect. Interestingly, when the electrodes from these reactors were inserted in the control reactors, the methane production increased by 3-4 times. This revealed that the electroactive biofilm formed on the electrode surface must have enhanced the stability of the AD rather than the electrical current (De Vrieze et al., 2014). Similarly, Liu et al. (2016) combined MEC and AD to form an MEC-AD system to improve CH4 production rate (MPR) from waste activated sludge. MPR was enhanced to 91.8 g CH4/m 3 reactor/d in the integrated MEC-AD reactor, compared to 30.6 g CH4/m 3 reactor/d in the AD alone . A novel combination of AD and MEC was also designed by Cai et al. (2016) in which, two anaerobic digesters were separated by AEM, and each functioned as an electrode chamber, i.e., anode and cathode. With sludge fermentation liquid, 0.247 mL CH4/mLreactor/d was produced at the cathodic anaerobic digester (increased by 51.53% than control) . Thus, methane production was enhanced with improving reactor stability, while lower contamination was observed by combining the MEC with AD (Clauwaert et al., 2008;Tartakovsky et al., 2011). Combining MEC with AD is an excellent example of how MECs can be used in a modular way to enhance the productivity of an existing technology (Aryal et al., 2018).

Integration of MEC with anaerobic membrane bioreactor (MBR)
The term membrane bioreactor (MBR) is used to define wastewater treatment processes where a perm-selective/semipermeable membrane (for example microfiltration or ultrafiltration) is integrated with a suspended growth bioreactor. Membrane filtration can be integrated into bio-electrochemical systems (BES) in the following ways: (a) as a separator between the electrodes, (b) an internal filtration component in the anode/cathode compartment, or (c) an external treatment process before or after the BES. More efficient treatment (more favorable quality of treated water), high energy efficiency, reduced investment, and mitigated fouling and/or sustainable desalination are the advantages of such integration systems ( Table 1) . Katuri et al. (2014) developed a novel anaerobic treatment system named as anaerobic electrochemical membrane bioreactor (AnEMBR); a combination of MEC with membrane filtration phenomenon employing electrically conductive, porous, nickel-based hollow-fiber membranes (Ni-HFMs) (Fig. 5). The AnEMBR was employed for the treatment of low organic strength wastewaters/solutions and to recover biogas (energy). The Ni-HFM served two different functions as cathode electrode for H2 production and as membrane to filter the treated water. Removal of COD (initial COD: 320 mg/L) was >95% and up to 71% of the substrate energy was recovered (CH4-rich biogas, 83%) at an applied potential of 700 mV. Additionally, there was less membrane fouling observed in the AnEMBR than in the control reactor (open circuit), which was due to H2 bubble formation, low cathode potential, and localized high pH at the cathode surface, (Katuri et al., 2014). In another study, a pilot scale submerged membrane electro-bioreactor (SMEBR) was operated based on the interaction between biological processes and membrane filtration for the treatment of wastewater, and electrokinetic processes were assessed (Hasan et al., 2012). A current density of 12 A/m 2 was maintained in the SMEBR which was operated in 235 L volume, at an HRT of 11 h for 7 weeks. In a single chamber, municipal wastewater was first treated biologically where most of COD and ammonia were removed followed by phosphorus, and flocs were coagulated in the electrical zone. Later the treated water (effluent) was filtered with a hollow fiber Microza microfiltration membrane module placed at the center of the SMEBR. The integrated system exhibited a superior function (phosphorus removal of 99%, ammonia removal of 99%, and COD removal of 92%), over MBR (phosphorus removal of 59%, ammonia removal of 97%, and COD removal of 87%) (Hasan et al., 2012). Although coupling an MBR with MEC has many potential benefits, more insights about of energy generation and consumption in the coupled system are needed.

MEC-pressure retarded osmosis (PRO) system
MEC systems require energy input in form of electricity to function and synthesize desired products. Combining MECs with renewable sources of energy can make the whole process more sustainable and eco-friendlier. Pressure retarded osmosis (PRO) is a technique to separate a solvent from a solution that is more concentrated and also pressurized through a semipermeable membrane. The salinity gradient generated during PRO can be used for electricity production, as also in the RED systems (see Section 4.6). Thus, the PRO unit can reduce the volume of wastewater and extract treated water. Further, the effluents from PRO unit can be treated by the MEC; and the osmotic energy obtained from the PRO unit can be integrated into the MEC for sustainable H2 production. A proof-of-concept for this system was demonstrated by . They first developed a time dependent PRO model and a batch model for MEC. Subsequently, using the predicted water flux obtained from the PRO model, the anolyte and the catholyte were prepared for an MEC operated experimentally using a power supply to mimic the energy supply process. The MEC system removed approx. 94% of the organics at 800 mV and produced 32.8 mL H2 at the expense of 470 J of energy after 46.9 h. The PRO unit produced 579 J of energy, demonstrating that such a system could be effectively used to power MECs. Thus, PRO-MEC system can help with organics removal, H2 production, and water recovery simultaneously (Yuan et al., 2015).

Integration of MEC with acidogenic bioreactor
Similar to combining AD with MEC (see Section 4.7), acidogenic bioreactors can also be combined with MECs. Biohydrogen production was evaluated using acetate, butyrate, and propionate as substrates in a single chamber MEC by applying different voltages by Babu et al. (2013b). Maximum HPR of 2.42 mmol/h was recorded at 600 mV along with about 53% removal of synthetic acids (Babu et al., 2013a). Furthermore, acidogenic bioreactor was also combined with MEC by the same group of researchers to improve product recovery and H2 yield (Babu et al., 2013b) ( Table 1) , 2014). A two-stage integrated or hybrid system with hydrolysis (first stage) followed by acidogenic fermentation for H2 production (second stage) was used. As a result of pretreatment, BEH showed a higher H2 production (29.12 mL/h) than the control (26.75 mL/h). Additionally, substrate degradation also improved with the BEH-pretreated substrate (COD removal of 52.42%) over the control (COD removal of 43.68%) (Chandrasekhar and Mohan, 2014). Modestra et al. (2015) also designed a single chambered MEC with acid pretreated biocatalyst, for electrofermentation of effluents towards additional H2 production with simultaneous treatment. The effect of VFA concentration (4000 mg/L and 8000 mg/L) on biohydrogen production with simultaneous remediation was studied at 200 mV and 600 mV applied potential. Maximum HPR of 0.057 mmol/h was observed at 600 mV along with a VFA utilization of 68% (Modestra et al., 2015). Thus, by combining MEC with acidogenic reactors, waste effluent from such reactors could be potentially converted into useful products like H2.

Integration of MEC with microbial electrolysis and desalination cell (MEDC)
A BES system developed by combining microbial electrolysis and desalination cell (MEDC) can concurrently reduce the salinity of salt water and produce H2. The benefit of MEDC is that it can generate pure and collectable H2 gas without dealing with contamination or voltage fluctuation and can achieve improved desalination assisted by an external power supply (Saeed et al., 2015). Mehanna et al. (2010) developed a three chambered MEDC consisting of an anode, desalination and cathode chambers, by using a pair of ion exchange membranes. Maximum HPR of 0.16 m 3 H2/m3/d was obtained in this reactor at an applied voltage of 550 mV ( Table 1). Also, the conductivity of the water in the desalination chamber decreased by 68 ± 3% with the 5.0 g/L NaCl sample and by 37 ± 4% with the 20 g/L NaCl sample (Mehanna et al., 2010). Similarly, Luo et al. (2012) reported 98.8% removal of NaCl (initial concentration: 10 g/L) from the middle desalination chamber per day. Apart from NaCl removal, H2 production at the rate of 1.5 m 3 /m 3 /d (1.6 mL/h) was achieved at 0.8 V . Thus, MEDC systems can potentially treat saline wastewaters along with H2 production. However, due to the sharp fluctuations in pH, exoelectrogens at anode and HPR at cathode could suffer. A solution to this problem was reported by Chen et al. (2012) who integrated an acid-production chamber and a bipolar membrane (BPM) into MEDC to form the microbial electrolysis desalination and chemical-production cell (MEDCC), which could simultaneously desalinate seawater, produce hydrochloric acid and generate NaOH. This arrangement helped to alleviate the difficulties of pH fluctuations and chloride ion (Cl − ) accumulation observed in a conventional MEDC. They also compared the performance of an MEDC and an MEDCC in their experiments. With the applied voltage of 1.0 V, the coulombic efficiency values of the MEDCC and MEDC were 97±2% and 65±2%, respectively. Furthermore, 86±4% and 60±4% of the 10 g/L NaCl (initial concentration) was removed in the desalination chamber of the MEDCC and MEDC, respectively, within 18 h. Lastly, with the applied voltage of 1.0 V within 18 h, the MEDCC produced 0.10 g of NaOH with 7.46 × 10 −5 kWh electricity (Chen et al., 2012). Therefore, by using an MEDCC, salty water can be treated without large pH changes. However, more research is required to make the membranes leakproof while provisions for up-scaling also need to be explored (Zhang and Angelidaki, 2014).

Integration of MEC with lignocellulosic ethanol biorefinery process
Lignocellulosic ethanol production is considered as one of the important second-generation biofuels. The substrate that is generated from lignocelluloses such as wheat straw by hydrothermal treatment followed by enzymatic hydrolysis can be used to produce ethanol. This process also generates a wide range of products viz., VFAs, phenolics, xylose, and polysaccharides, exhibiting the potential to generate a considerable amount of energy. Accordingly, an MEC was coupled to harvest H2 from the effluent of lignocellulosic ethanol production which converted 60-70% of COD to H2 (Borole and Mielenz, 2011). Moreover, during the pretreatment stage, a large number of metabolic inhibitors like furanic and phenolic compounds, furfural and hydroxymethyl furfural were generated which are critical impediments for the downstream process. Rich bacterial diversity/community developed in the system was helpful for the utilization of phenolics and polysaccharides (Zeng et al., 2015). H2 was generated in the MEC using a mixture of furfural and 5-

Thermoelectric microconverter-MEC coupled system
Industrial processes viz., automobile industries, steel industries, etc., generate waste heat as by-product which can be tapped as an energy source by thermoelectric converters and this energy is called as thermoelectricity. This phenomenon is called as Seebeck effect and was discovered in 1821 (Elsheikh et al., 2014). Thermoelectricity converters function based on the temperature gradient that exists in the medium, which can be water or any solid phases. Industrial activities generate a large amount of heat and that could create a temperature gradient, which could help with thermoelectricity generation. Thermoelectricity can also be captured from low temperatures on which limited studies are available. Recovery of waste heat from such low temperatures is more environmentally friendly. MECs operation for the production of H2 requires low amounts of electrical energy. Thermoelectricity is renewable in nature. Thus, the integration of MECs with thermoelectricity can make the H2 production process more sustainable. Chen et al. (2016) evaluated the effects of different temperature ranges in a thermoelectric microconverter-MEC coupled system on H2 production from acetate as carbon source. HPR of the systems was found to depend on the generated electric potential by the thermoelectric microconverter. Based on the temperature (between 35 to 55°C), the voltage varied between 170 mV to 830 mV. Maximum H2 production of 0.16 m 3 /m 3 /d and yield of 2.7 mol/molacetate were recorded at 55°C of the hot side, where an average voltage of 700 mV was sustained with the current density in the range of 0.28 to 1.10 A/m 2 (Chen et al., 2016).

Existing challenges and limitations to scaling-up the MEC technology
Bringing the MEC technique into a pilot or industrial scale is still a challenge. MECs, as being a newly developed technique in the past decade, hydroxymethyl furfural and three phenolic compounds such as syringic acid, vanillic acid, and 4-hydroxybenzoic acid as substrate in an anodic oxidation reaction. The initial concentration of 8.7 mM of the mixture of the five compounds was used, which bio-transformed at a rate ranging from 0.85 to 2.34 mM/d. The H2 yield varied in the range of 0.26 to 0.42 g H2-COD/g COD removed through the anodic reaction (Zeng et al., 2015). The concept of biorefineries could also be beneficial for polyphenol purification as well as for targeted modification of fruit-based phenolics. As waste products can be directly considered as substrate input for an MES, the combination of an ethanol biorefinery process could allow the maximization of the energy output and simultaneous valorization of the ethanol waste stream (Fig. 6) (Thygesen et al., 2010).      Figure 7. Firstly, there is a major challenge to maintain the electron transfer efficiency from exoelectrogens to the anode while scaling up the MEC reactors. Even though the relationship between current density and anode surface area has not been systematically quantified, it is generally accepted that the current density would decrease with increases in anode size (Dewan et al., 2008). This will reduce energy efficiency and increase the energy input cost for H2 production. Establishing a highly electrochemically active biofilm and maintaining the sufficient mass/electron transfer within the biofilm is essential.
Another important issue affecting MEC efficiency is energy losses. The first factor that causes energy loss is the pH gradient, which is a common issue for membrane-based MEC reactors regardless of the membrane material used (Rozendal et al., 2006). This could cause a major shift in cathode potential and interfere with anode exoelectrogen metabolism. Rozendal et al. (2006) reported that about 0.38V out of 1V applied in the system was lost due to the proton gradient using a CEM. The ion selection of membrane as well as the ions presented in the electrolyte and its pH buffering capability determine the pH gradient between the two compartments. It should be noted that buffering the electrolyte by chemical dosing is also impractical at the industrial scale. Therefore, a practical solution to solve this issue is highly required.
The second type of energy loss is called the ohmic energy loss, mainly originating from the ohmic resistance of anode and cathode as well as the conductivity of electrolytes and membrane. Because of their relatively lower conductivity and lower ohmic resistance, carbon-based electrodes could serve more ideally at large scale applications. Moreover, the conductivity of electrolytes is also an issue, especially when an MEC is combined with wastewater treatment. Although some wastewaters have high conductivities, such as distillery wastewater (Mohanakrishna et al., 2010), wastewater from chemical industries (Venkata Mohan et al., 2008), and source-separated urine (Ledezma et al., 2015), the majority of domestic and industrial wastewaters typically show low conductivity that could cause large ohmic losses (Jeremiasse et al., 2010). This will require shortening the distance between the two electrodes (anode and cathode) to reduce the ohmic drop, but another issue of gas diffusion will become serious in this case: either the diffusion of H2 to anode to induce the methanogens growth or the diffusion of carbon dioxide to cathode to reduce the purity of H2 gas.
In addition to the issue solely related to MEC reactors, integrating MECs with other techniques on a large scale also still challenging. Integrating with other techniques could significantly expand the potentiality of MECs and could enhance their industrial competitiveness. The major challenge of this approach is that most of the options discussed above are still in the stage of infancy and sufficient insights are lacking. Firstly, advanced system structure designs are required to meet both requirements of MECs and the other processes to be integrated. This may lead to higher costs for system construction, which needs to be compensated for by additional benefits. Furthermore, the comparability between MECs and the technique to be integrated is unclear and needs to be systematically investigated. Since the two techniques are dependent of each other, the performance of one process could exert a significant influence on that of the other. This influence could be negative if the system is not well designed. Dynamic evaluation of the configuration is also essential, especially because of the relatively low rate of MECs. Different scaling up factors, required for MEC reactors and other techniques to achieve a dynamic equilibrium status, might be problematic and need to be solved during the scaling up process. Table 2 presents a comparison between the present review and the review articles previously published on MECs (2013-2019). As seen, to the best of our knowledge, this work is the most inclusive review on the topic in particular for its coverage of the various MFC integration options.

Conclusions and future prospects
In summary, the future of MECs seems promising in reducing the overall cost of wastewater treatment and in providing additional benefits through the production of H2 or other value-added fuels/chemicals. However, MEC technology is still in its infancy, facing some challenges which should be overcome prior to its scaling-up and commercialization. Issues like energy loss, mass transfer limitations, pH gradient, etc., need to be investigated systematically at pilot-scale using real-field wastewaters. Application of dynamic modeling and designing MECs having lower mass transfer limitations and energy loss are also equally important. The advantage of integrating MEC with different energy-generating processes will also improve the pace of biorefineries growth towards sustainable development.