Substrate is regarded as one of the key impact factors of hydrogen production in MEC. The type, concentration and updating rate of substrates are directly related to the hydrogen production yield and the reaction rate in MEC . HER can occur in MEC with various substrates including acetic acid, butyric acid, lactic acid, glucose, cellulose, glycerol, methanol, milk, starch, phenol and different types of wastewater [21, 27]. The specific types and classification details of substrates used for MEC are presented in Table Ⅰ.
Table Ⅰ. Classification of different substrates used in MEC.
In lab experimental research, the anode of MECs is cultivated with wastewater that contains both microorganisms and some complex substrates and some additional nutriments. Sodium acetate is the most widely used substrate in lab at present, not only because of its low cost and sufficient supply in the market but also because of its strong electron-giving capacity. According to the reaction equation, one acetate can produce eight electrons. It was previously reported the MEC using acetate as a substrate could achieve a hydrogen production rate of 50 m3-H2·m-3·d-1 in a double-chamber MEC with an applied voltage of 1.0 V . Adding additional carbon source to MEC system with wastewater as substrate can promote the decomposition of complex organic matter in wastewater by microorganisms, improve the wastewater treatment efficiency, accelerate the decomposition rate and electron transfer, and ultimately promote the hydrogen production rate in the cathode . In this regard, MECs as a promising technology show potential in producing sustainable new energy (hydrogen) and remove complex compounds from wastewater. Shao et al. compared the hydrogen production properties of MECs by using sodium acetate, sodium butyrate, sodium propionate, glucose and starch as substrates under the same experimental conditions, which demonstrated that the MEC cultured with sodium acetate as a substrate was the best one while the starch was the worst . It is also discovered that the anode of MEC which was fed with glucose and starch exhibited high biodiversity, indicating mixing sodium acetate with glucose or starch in a certain proportion as a substrate for MEC may yield higher hydrogen production.
Glycerol is a common by-product of biodiesel fuel production and as well as a common carbon source for MEC. Since 1.1 L of glycerol is produced as a by-product when producing 10 L of biodiesel, the mass production of biodiesel fuel has contributed to glycerol production exceeding demand . Therefore, hydrogen production in MECs can be coupled with the valorization of glycerol. However, the maximum yield of hydrogen was only 0.77 mol-H2/mol-glycerol when glycerol was used as substrate in a double-chamber MECs . Selembo et al. compared the performance of MEC when P-glycerol (ultrapure), B-glycerol (the side product of biodiesel fuel production) and glucose were served as substrates with applied voltage of 0.5 V and 0.9 V . Results revealed that both hydrogen yield and hydrogen production rates for P-glycerol were better than that for B-glycerol. However, the B-glycerol can be converted, without purification, into hydrogen with a higher production rate and yield in MEC than fermentation. Another research put forward by Speers et al. showed that glycerol could be fermented into ethanol with yield of 90% by the exoelectrogen Geobacter sulfurreducens and the bacterium Clostridium cellobioparum in MEC and simultaneously offered the fermentative byproducts as electron donors for Geobacter sulfurreducens , such a strategy provided merits of glycerol conversion into ethanol production and the conversion of fermentation byproducts into H2, significantly improved the utilization efficiency of glycerol. In order to further improve the conversion rate of glycerol to H2, a two-stage process linking dark fermentation with a MFC or MEC was also reported .
Glucose is another widely used substrate for MEC, which can produce hydrogen in different condition. Glucose can produce hydrogen by fermentation at mesophilic temperature in dark or decompose its fermented products (organic acids) in MECs at a low temperature. Lu et al. reported that hydrogen could be produced with glucose as a substrate in a single-chamber MEC at a low temperature of 4 ℃, with a yield of about 6 mol-H2/mol-glucose and rate of (0.25±0.03)/(0.37±0.04) m3-H2·m-3·d-1 . In another study, hydrogen was obtained from glucose with a yield of 6.4 mol-H2/mol-glucose and rate of 0.83±0.3 m3-H2·m-3·d-1 in a single-chamber MEC operating under a temperature of 30 ℃, when a voltage of 0.5 V was applied. Moreover, when the applied voltage was added to 0.9 V, a higher hydrogen production yield of 7.2 mol-H2/mol-glucose and rate of 1.87±0.3 m3-H2·m-3·d-1 can be achieved .
Protein was also reported as a substrate of MECs with achieving high hydrogen production rates and yield. Bovine serum albumin (BSA, 700 mg/L) was used in a single-chamber MEC generating hydrogen at a rate of 0.42±0.07 m3-H2·m-3·d-1 and a yield of 21.0±5.0 mmol-H2/g-COD (chemical oxygen demand), with an energy recovery of 75%±12% at an applied voltage of 0.6 V . When complex protein (peptone) was served as a substrate under the same condition, hydrogen was generated at the rate of 0.05±0.01 m3-H2·m-3·d-1, the yield of 2.6±0.1 mmol-H2/g-COD, and the energy recovery of (14±3)%. These studies suggested that it may be a good method for both renewable energy evolution and treatment of wastewater containing high concentrations of proteins by using protein as a substrate in MECs.
In previous studies, lignocellulose has been regarded as a promising feedstock for hydrogen production in MECs because of two reasons: (i) It is rich in complex carbohydrates that can be used as substrates for microorganisms to produce renewable energy sources, like hydrogen, methane, ethanol and electricity. (ii) It is abundant in nature and renewable. However, lignocellulose cannot be directly used as substrates for MEC and must be converted into mono-saccharides or other low-molecular-weight compounds (organic acids) that are easier to be degraded by electrochemically active bacteria in MECs. The conversion of the refractory lignocellulosic materials into hydrogen gas was achieved at high yields and rates in a two-stage dark-fermentation and electrohydrogenesis process . In the first fermentation stage, fermentation using Clostridium thermocellum produced 1.67 mol-H2/mol-glucose at a rate of 0.25 L-H2·L-1·d-1 with a corn stover lignocellulose feed, and 1.64 mol-H2/mol-glucose and 1.65 L-H2·L-1·d-1 with a cellobiose feed. In the MEC stage, hydrogen yields and production rates with the actual fermentation effluents (synthetic effluents) and single substrate were 980±110 mL/g-COD and 1.11±0.13 L-H2·L-1·d-1 (synthetic), 900±140 mL/g-COD and 0.96±0.16 L-H2·L-1·d-1 (cellobiose, and 750±180 mL/g-COD and 1.00±0.19 L-H2·L-1·d-1 (lignocellulose). The results illustrated that high hydrogen yield and gas production rates can be achieved from lignocellulosic using a two-stage fermentation and MEC process.
Wastewater generally contains a great deal of complex organic pollutants that must be removed before discharge into environment. Activated sludge process is a global public sewage treatment method, but the treatment cost of this method is relatively high. Microbial electrolysis cells (MECs) hold great potential as a technology for wastewater treatment in parallel to H2 production . In recent years, the research of using MFC/MEC technology to treat various kinds of wastewater has attracted much attention. Studies have demonstrated that wastewater can be used as substrate in a MEC including: domestic wastewater (DWW), swine wastewater (SWW), refinery wastewater (RWW), potato processing wastewater (PPWW), dairy manure wastewater (DMWW), landfill wastewater (LWW), sewage sludge (SS) and industrial wastewater (IWW), but the hydrogen production rates and yields were much lower than that of the MECs with acetate as substrate.
As a large amount of domestic wastewater is produced every day, a non-secondary pollution treatment process must be selected to treat it. It is expectable to realize the treatment of domestic wastewater in MECs as the electricity can be generated from wastewater in MFC. Ditzig et al.  investigated hydrogen production from a MEC using domestic wastewater as substrates and evaluated the system performance in terms of hydrogen recovery, coulombic efficiency and the removal rate of COD, BOD and DOC. Though the hydrogen recovery, coulombic efficiency and overall hydrogen yield needed to be improved, the method to treat domestic wastewater by MEC was testified to be feasible. A pilot-scale MECs, addressing DWW, produced an equivalent of 0.015 L-H2·L-1·d-1, and recovered around 70% of the electrical energy input with a coulombic efficiency of 55% . Although the reactor did not reach a 100% electrical energy recovery and COD removal was limited, the performance levels of the system could be potentially promoted substantially by improving hydrogen collection and reactor design. Escapa et al. illustrated that low coulombic efficiencies and the occurrence of hydrogen recycling significantly limited the hydrogen production performance of MEC system, but MEC can be successfully used for DWW treatment .
Addressing swine wastewater with a generation of valuable products such as hydrogen can reduce the cost of treatment to a certain extent. H2 was produced in a single-chamber MEC with a graphite-fiber brush anode at the rate of 0.9-1.0 m3-H2·m-3·d-1 when high-concentration SWW and diluted SWW were used as substrates . A noble wastewater treatment method combining MFCs and flocculation provides an effective way to treat swine wastewater to low pollutant levels in the effluent at low cost (a net gain) . Therefore, combining the MEC with other wastewater treatment method could be a promising way to address the swine wastewater.
Industrial wastewater usually contains many refractory organic pollutants and heavy metals, which seriously harms environment. A methanol-rich industrial wastewater and a food processing wastewater used in MECs were examined . Molybdenum disulfide (MoS2) and stainless steel (SS) were used as alternative cathode catalysts to Pt in terms of treatment efficiency and energy recovery using actual wastewaters in this study. The industrial wastewater had higher biogas production rates of 0.8-1.8 m3-H2·m-3·d-1 and COD removal rates of 1.8-2.8 kg-COD·m-3·d-1 than the food processing wastewater. The overall energy recoveries were positive for the industrial wastewater (3.1-3.8 kWh/kg-COD removed), while the food processing wastewater required a net energy input of -0.7 to -1.2 kWh/kg-COD using MoS2 or Pt cathodes, and -3.1 kWh/kg-COD with SS, respectively.
A. Acetate and other non-fermentable substrates
B. Fermentable and complex substrates
1. Domestic wastewater (DWW)
2. Swine wastewater
3. Industrial wastewater
The HER in MECs takes place at the cathode, it is necessary to load electrocatalyst on the carbon material to reduce the overpotential and promote the reaction rate. While MEC has attracted wide attention in recent years due to its superior performance for hydrogen evolution, there are still issues in practical application. However, there is still a lot of room for catalyst development in MEC performance improvement, which needs to prepare more efficient and stable catalysts with high catalytic activity to promote the production of hydrogen. It is found that the catalytic activity of ternary or multi-component alloy is higher than that of binary and single metal, and the performance of nano crystalline electrode is better than that of ordinary alloy. If the multi-component nano crystalline electrode materials can be prepared, the catalytic activity of the electrode may be greatly improved. In MEC, platinum is the most commonly used catalyst, but its application is limited by its high cost and tendency to cause environmental pollution in the mining process. For a long time, researchers have been committed to find a non-noble metal catalyst to replace Pt and use the non-noble metal-based electrocatalyst to carry out hydrogen evolution reaction (HER) at the cathode of MEC. It is necessary to create and develop low-cost and efficient alternative catalysts, which can provide a promising avenue for further development of hydrogen evolution in MEC.
Noble metals have the best catalytic performance for MEC, almost all precious metals can be used as catalysts, but the most commonly used ones are platinum, palladium, rhodium, etc., especially platinum. In the past, it has been found that platinum group metals show high catalytic activity for MEC hydrogen evolution. Platinum group metals are easy to adsorb reactants and accelerate the reaction because their d-electron orbits are not full. Platinum shows good catalytic activity in MEC because of its large surface area and strong adsorption performance. However, platinum is easily poisoned by carbon monoxide and sulfide . Carbon monoxide can be preferentially adsorbed on the active sites of Pt catalysis, so as to prevent the acquisition of hydrogen and reduce catalyst activity. Sulfur pollutants also have an irreversible effect on the catalyst and have a negative impact on MEC.
Although platinum is the state-of-the-art electrocatalysts for HER, its high cost and vulnerability by chemicals limit its application. Therefore, the development of non-noble metal electrocatalysts to promote the release of MEC hydrogen has become a hot research topic. Transition metal (TMs) compounds have attracted great attentions in MEC applications due to their high conductivity and unique electronic structure. The outermost layer of transition metal element is not fully filled with electrons, so it tends to easily adsorb the electrons of reactant with generation of intermediate products to activate the adsorbed molecules. In addition, transition metal as solid catalysts usually exists in the form of crystalline and abundance of active sites tend to appear on the surface, which can facilitate the activity of reaction for reactant. TMs have gradually become the most popular catalytic material for HER in MEC. The structures, preparation methods of some typical transition metal compounds and their effect on MEC hydrogen evolution will be summarized in the following sections.
Transition metal phosphides (TMPs) are a new type of catalytic materials for hydrogen evolution in MEC following transition metal carbide (TMC) and transition metal nitride (TMN) . TMPs have been widely concerned by researchers, because of its high catalytic activity, high thermochemical stability and high conductivity in MEC. Therefore, the research on TMPs has important theoretical significance and practical value. CoP is widely concerned in TMPs because of its low cost and excellent catalytic activity, but its poor conductivity affects its application in electrocatalysis. Wu et al. combined CoP with carbon nanotubes (CoP/CNTs), obtaining excellent HER performance, because the high specific surface area of carbon nanotubes can accelerate the electron transfer and improve the conductivity, reaching the current density of 10 mA/cm2 at an overpotential of 139 mV . When Cu3P was used as a catalyst for hydrogen evolution, a high over-potential was required to achieve a current density similar to that of CoP. Du et al. synthesized Cu3P-CoP/CC by phosphating on carbon cloth , the synergistic effect between metals increased the catalytic activity of Cu3P. Compared with Cu3P/CC and CoP/CC, Cu3P-CoP/CC achieved a current density of 10 mA/cm2 at overpotential of 59 mV, showing great improvement toward electrocatalysis of HER. Du et al. also tried to improve the catalytic activity by further metal doping into TMPs , they deposited aluminum on Ni2P nanowire array (Al-Ni2P), which achieved a current density of 10 mA/cm2 at overpotential of 129 mV and long-term chemical stability .
There are many ways to prepare TMPs. Liu et al. prepared the CoP nanowire array by phosphating strategy at low temperature, the CoP nanowire shows an onset potential as low as 40 mV with a Tafel slope of 54 mV/dec and decent stability over 18-h operation . Logan et al. reported a liquid-phase synthesis method to prepare Ni2P catalyst on supported carbon (Ni2P/C) which was studied as an electrocatalyst of MEC for H2 production (FIG. 3(a)). The MEC with the Ni2P/C cathode could release a current density of 14 times of that at the Ni/C electrode, and showed good stability upon 11 day cycle running, as shown in FIG. 3(b). FIG. 3(c) revealed the effect of different pH values on the catalytic activities of Ni2P/C and Pt/C, the results showed that lower pH was likely more beneficial to HER, because the concentration of protons in acidic solution was high, which was conducive to the adsorption of hydrogen ions by the active center. Compared with Pt/C and Ni/C in MEC under 0.9 V applied voltage, it is found that both the amount of hydrogen produced and energy yields of the MECs by Ni2P are quite similar to Pt/C (FIG. 3(d, e)). Therefore, Ni2P is a good substitute for noble metal catalyst in MEC reaction .
Figure 3. (a) Schematic diagram of Ni2P cathode MEC. (b) Current generation of MECs with Ni2P/C, Ni/C, and Pt/C cathodes over 11 days (0.9 V applied voltage). (c) Chronopotentiometry (CP) test results from Ni2P/C and Pt/C catalysts at different pH (pH = 2 and pH = 7). (d) Hydrogen production rates (L-H2·L-1·d-1) of the MECs over a 24 h cycle with Ni2P/C, Ni/C, and Pt/C catalyst cathodes (over 11 days, 0.9 V applied voltage). (e) Cathodic hydrogen recoveries (rcat) and energy yields of the MECs with Ni2P/C, Ni/C, and Pt/C cathodes (0.9 V applied voltage). Reproduced from Ref. with the permission of Elsevier.
The transition metal carbides (TMCs) are formed by the fusion of carbon atoms into the transition metal lattice . The addition of metal makes the function of TMCs greatly different from that of metal, with expanded metal lattice and specific surface area, so the catalytic performance is expected to be better than that of parent metal . In addition, TMCs is a kind of material with high hardness and strength, high conductivity and thermochemical stability, so it is widely used in industrial production. But carbon pollution has become a non-ignorable problem that will block the active center needed for catalysis . Besides, the formed oxide film will be wrapped on the catalyst surface, blocking the close contact between the reactants and the catalyst surface, and affecting the reaction, resulting in the reduction of catalytic activity. Reducing carbon pollution, expanding catalyst active center, and preventing carbon particles from agglomerating are the key points of future research.
Molybdenum carbide (Mo2C) and tungsten carbide (WC) are two typical transition metal carbides with high activity and stability. WC has similar chemical stability to platinum-based catalyst and can also resist catalyst poisoning, so it is considered to be a candidate material to replace platinum-based catalyst . Fan et al. synthesized carbonization dock on vertically arranged carbon nanotubes to support the growth of WC. The prepared WC catalyst had good HER performance with the overpotential of 137 mV to reach a current density of 10 mA/cm2 inalkaline medium . However, the currently prepared WC particles have a low surface area, and it is easy to form thick oxide film to prevent reactants from entering the active center, thereby reducing the catalytic activity of WC. Compared with WC, Mo2C has a better catalytic effect on MEC cathode and has been widely used. Lin et al.  doped Co into the Mo2C crystal structure with a Co/Mo ratio of 0.020, increasing the electronic density of state of Fermi level and improving the catalytic activity. In 1.0 mol/L KOH solution, the current density at Co-Mo2C electrode was 10 mA/cm2 at overpotential of 118 mV, providing a feasible strategy to explore efficient electrocatalysts for hydrogen evolution in MEC by engineering on composition and nanostructure.
Although TMC shows favorable catalytic activity toward HER, but there are still some issues in the catalytic process, including the difficulty in preparing catalyst with high specific surface area and the optimization of catalytic conditions.
Like TMCs, transition metal nitrides (TMNs) are filled with nitrogen atoms in the voids of metal lattice structure, which leads to the change of the electronic structure of the d-band . These nitrides have the advantages of high hardness, chemical corrosion resistance, high conductivity and thermal stability. There are many TMNs catalysts, such as molybdenum nitride (MoN), tungsten nitride (WN), nickel nitride (Ni3N) and cobalt nitride (Co4N). Among them, MoN is the most widely used because of its excellent catalytic performance. Ramesh et al.  synthesized MoNx/NF catalyst on nickel foam by atomic layer electrodeposition (ALD). In acidic medium, the MoNx/NF only needed 148 mV overpotential to approach current density of 10 mA/cm2, likely because the Mo-N covalent bond enhanced the electron density of d-band, which contributed to higher catalytic activity than bare NF. Mo5N6 nanoflakes were synthesized by salt template method  with Mo/N ratio of 5:6, which increased nitrogen content and showed better catalytic activity than that of TMN in different electrolytes. The high valence state of Mo element and a large amount of nitrogen doping make Mo5N6 have the electronic structure similar to platinum, so it has high catalytic activity.
TMN materials with excellent catalytic performance need more active centers and nanoparticles with uniform distribution. The methods to prepare TMN include temperature programmed reaction method (TPR) , synthesis of metal nitrides by the pulsed laser deposition (PLD) , reactive magnetron sputtering method , TPR is a method of mixing metals, metal oxides, metal organics with ammonia and then through nitriding to obtain metal nitrides. Different precursor and nitrogen will produce nitride products with different shape. The particle size and pore structure of metal nitride will be affected by precursor material, preparation temperature and nitriding conditions. Therefore, reaction conditions shall be reasonably controlled during preparation to ensure that materials with high catalytic activity are obtained.
The excellent catalytic performance of TMN makes it attract much attention in the field of catalysis, but its metastable characteristics often form impurities such as metal, oxide, hydroxide, etc. It is unclear what effect these impurities have on the catalytic performance and should be paid attention to in future research. Among the metal nitrides, MoN has good catalytic activity and plays an important role in the MEC hydrogen production process. The application of other new nitrides (such as cobalt nitride and tungsten nitride) in MEC hydrogen evolution is also a new research direction.
Stainless steel (SS) has good corrosion resistance and low price. In MEC, the cathode made of stainless steel with a high specific surface area can achieve catalytic performance similar to that of platinum-based catalysts. Furthermore, Munoz et al. reported that stainless-steel cathode showed good catalysis performance with the hydrogen flow rate of 4.9 L-H2·m-2·h-1 (11.2 A/m2) under the applied voltage of 0.8 V in weak alkaline solutions (pH = 8) . It has been found that the stainless-steel cathode material with high specific surface area can replace the noble metal catalyst to obtain a good hydrogen yield. Under the voltage of 0.6 V, using a stainless-steel brush cathode with a specific surface area of 810 m2/m3, the hydrogen production rate can reach (1.7±0.1) m3-H2·m-3·d-1. This study showed that although the specific surface area of reactor horizontal brush (HB) was larger than that of reactor vertical brush (VB), the produced current density of HB was lower than that of VB. It meant that the electrode spacing can avoid ohmic loss, which was also the reason why the performance of reactor VB was better than HB . The most studied three types of stainless steel as catalysts are SS304, SS316 and SS430 containing 9.25%, 12%, and 0.75% nickel respectively. Stainless steel 316 had the highest nickel content, it was found that it was the best cathode material in alkaline medium by electrochemical test . As a low-cost catalytic material, stainless steel mesh has been used in MEC research recently. When the stainless-steel mesh is used in MEC, the hydrogen production catalysis performance can be affected by its structure. Under low current density and small bubble coverage, the mesh diameter is the main factor affecting the hydrogen production rate, while under high current density and high bubble coverage, the mesh size is the main factor affecting the hydrogen production rate .
Nickel-based catalytic materials have been concerned to be comparable to platinum-based catalysts in the catalytic activity and efficiency of hydrogen production and the high conductivity, so they are most likely to replace platinum-based catalysts . The comparison of parameters of different catalysts in MEC hydrogen evolution is shown in Table Ⅱ . It shows that Ni/C is not as good as Pt/C in catalytic performance. Whether nickel is used alone or in combination with other metals, its catalytic performance is superior to other metal-based materials. The energy efficiency of Ni/C is 197%, and the highest conversion is 99%. The coulombic efficiency and energy efficiency of carbon nanotubes are lower than those of other catalysts [87, 88]. Although carbon nanomaterials have a large surface area, the carbon nanomaterials without metal doping still have great limitations in catalysis. Another study  used Pt wrapped nickel foam (NF) as a catalytic material. Compared with Pt/CC and NF, the energy efficiency of Pt/NF under 0.8 V applied voltage was 127%, higher than that of Pt/CC and NF, and the highest hydrogen production rate of the cathode was 0.71±0.03 m3-H2·m-3·d-1, which exceeded the 0.67±0.02 m3-H2·m-3·d-1 of Pt/CC, The reason is that the porous structure of nickel foam provides more active sites than planar carbon materials and has greater catalytic activity. However, during MEC operation, the porous structure of nickel foam causes nickel corrosion and increase cathodic overpotential, which will bring energy loss. Because of the low overpotential, electroformed nickel mesh is used as a substitute for platinum catalyst. When the applied voltage was 1.1 V, the hydrogen production rate of MEC was 4.18±1 m3-H2·m-3·d-1 at the current density of 312±9 A/m3, which was slightly lower than that of MECs with Pt catalyst (4.25±1 m3-H2·m-3·d-1, 314±5 A/m3) . The cost of electroformed nickel mesh is low, so it is promising to be widely used in MEC.
Table Ⅱ. The comparison of hydrogen evolution parameters of different catalysts in MEC.
The porous structure of nickel foam can cause corrosion of nickel. Plating tungsten and molybdenum on nickel foam can improve corrosion resistance. In a single chamber membrane free MEC reactor, the hydrogen evolution performance of the electrodes has been tested in neutral phosphate buffer by adding the voltage of 0.6 V. The result showed that the current density (250 and 125 A/m2), hydrogen production rate (0.14±0.01 and 0.13±0.01 m3-H2·m-3·d-1), cathodic hydrogen recovery rate ((78.9±1.7)% and (88.6±2.3)%) of NiW and NiMo cathodes were both higher than that of bare nickel foam . In order to improve the practical feasibility of hydrogen production in MEC, the performance of catalyst should be further optimized.
In recent years, great progress has been made in using nano materials as cathode catalysts. Nano materials include nano particles, nano films, nano crystals, etc. The tiny structure and large specific surface area of nanomaterials greatly increase the contact area with the reactants, which is why it is used as a new MEC hydrogen production catalyst.
Dai et al.  prepared nano-Mg(OH)2/graphene composite by using hydrothermal synthesis with MgSO4·7H2O and graphene oxide (GO) as precursors. The results of LSV showed that the best hydrogen production was obtained when the ratio of MgSO4·7H2O to GO was 1:1. The results showed that the hydrogen production of nano-Mg(OH)2/graphene was better than that of Pt/C, but the catalytic activity and energy efficiency of the former were similar to that of the latter. Dai also used the best raw material ratio of (NH4)2MoS4 to GO of 1:1, MoS2/graphene composite was also prepared by hydrothermal synthesis method . The results were similar to nano-Mg(OH)2/graphene. In the experiment of MEC hydrogen production, the current density, hydrogen production rate and coulombic efficiency of MoS2/graphene cathode were 9.96±0.65 A/m2, 0.424±0.04 m3-H2·m-3·d-1 and (89.11±5.87)% respectively, which were higher than those of Pt/C. The good hydrogen evolution performance of MoS2 in MEC has been confirmed, but its electrochemical behavior needs to be further explored. The catalytic activity of MEC was tested under two different electrode structures. The first electrode structure was to deposit MoS2 on stainless steel. The SS electrode polished with MoS2 had more catalytic activity than that uncoated MoS2, for the active sites of the polished MoS2 exposed, resulting in higher catalytic activity . The second electrode structure was to combine MoS2 with carbon black. By changing the optimal load of MoS2 composite cathode, the surface density of catalyst is affected. After LSV scanning in sodium perchlorate solution (NaClO4), the optimal load was 54 mg MoS2 and 60 mg carbon black, and the surface density was 45 g/m2 . This composite nano material has advantages due to its low cost and simple synthesis process.
It is known that different electrode structures have influence on the catalytic activity of MEC cathode. Chen et al.  synthesized an excellent nano material, which had good hydrogen evolution performance in a wide range of pH. Ni-N-MoCx doped with Ni-N was synthesized by the mixture of NiMoO4 and C2H4N4. In the alkaline medium, the initial potential was -37 mV, the overpotential was 124 mV, and the current density was 10 mA/cm2. In the acid medium, the initial potential was -74 mV, the overpotential was 163 mV, and the current density was 10 mA/cm2. The Co doping of Ni and N increased the HER activity in MEC. Another new type of catalyst N-Fe/Fe3C@C with nitrogen core-shell structure, also showed good hydrogen evolution performance in MEC. The new catalyst took iron-based composite (Fe/Fe3C) nanorods as the core and graphite carbon as the shell. It showed good hydrogen evolution activity and stability in MEC. Under 0.8 V applied voltage, the peak current density of N-Fe/Fe3C@C was 2.60±0.07 A/m2, higher than CC cathode (1.36±0.01 A/m2) and carbon nanotube cathode (1.30±0.09 A/m2) . The low cost and simple synthesis process of the composite nanomaterials make them have advantages and can be used as catalysts to replace precious metals.
Catalyst plays an important role in improving hydrogen production of microbial cell. Table 3 summarizes the influence parameters of different kinds of catalyst on MEC hydrogen evolution performance, providing reference for finding efficient and stable catalyst.
Table Ⅲ. Summary of the H2 production performance in MECs with difference cathodic catalysts.
The typical Pt-free catalysts developed in the recent years for HER in MEC are summarized in FIG. 4. As it depicts transition metal-based catalysts (especially nickel) and nanomaterials attracted more attention in the recent few years. With the rapid development of platinum alternative catalysts, it is believed that application of MEC technology for hydrogen production becomes more and more prospective.
The bio-electrode MEC can be divided into semi-bio-electrode MEC and full-bio-electrode MEC according to the location of microorganisms. Semi-biological cathodes are divided into anode microorganism MEC and cathode microorganism MEC according to the role of microorganisms in the hydrogen evolution process of MEC (FIG. 5).
The MEC cathode is where hydrogen gas is generated, so the performance of the MEC cathode will directly affect the hydrogen production to a large extent. In addition to the use of metal catalysts, the study of microorganisms as cathode catalysts has also been included in the list of MEC cathode catalyst research, because its cost is lower than chemical catalysts, it can generate itself, and does not produce secondary pollution and other advantages. Although the current density of biocathode MEC is lower than that of traditional electrolytic cell, the advantage of low cost makes it have the potential for further exploration [100, 101]. Biocathode is a welcome advancement in the practical application of wastewater treatment, metal ion recovery, and preparation of chemical products such as hydrogen.
In 2008, Rozendal et al.  for the first time reported that the anode used acetate and the cathode used microorganisms to catalyze hydrogen evolution. By reversing the polarity of the electrode, a bioanode of acetate and hydroxide was converted into a biocathode that produced hydrogen. In this way, the MEC can achieve a current density of -1.2 A/m2 at an applied voltage of 0.7 V, which was 3.6 times that of the blank control electrode. In addition, the microbial biocathode electrode has a cathode hydrogen efficiency of 49% and a hydrogen generation rate of 0.63 m3-H2·m-3·d-1, while the control electrode cathode hydrogen efficiency was 25% and a hydrogen generation rate of 0.08 m3-H2·m-3·d-1. SEM images of both microbial electrodes showed a well-developed biofilm on the electrode surface.
In order to compare the difference between the performance of microbial cathode MEC and ordinary chemical catalysts for hydrogen evolution of MEC, explore the reason why the rate of hydrogen evolution of biocathode MEC was lower than that of chemical catalysts, Jeremiasse et al.  found that biocathode could also realize hydrogen production under suitable conditions, and the more microorganisms on the electrode, the more helpful it was to improve the current density and hydrogen production effect of MEC. However, when using a metal catalyst relative to MEC, the hydrogen production rate of the biocathode was low, only 0.04 L-H2·L-1·d-1, which was 2 orders of magnitude lower than the metal catalyst. Although the speed of hydrogen preparation by using biological cathode was improved, it was only 0.30 L-H2·L-1·d-1 . Cathode microorganism growth rate is too slow and the negative impact of cathode precipitates on hydrogen production efficiency may be the main reason for the low rate of hydrogen production of biocathode MEC.
Temperature is one of the important factors that affect the hydrogen evolution performance of MEC, mainly because the activity of microorganisms is greatly affected by temperature. In 2013, Fu et al.  reported for the first time a dual-chamber high-temperature biocathode that did not need to change the polarity of the hydrogen evolution MEC. The anode was a carbon cloth that was not inoculated with microorganisms. The CV curve showed that the reducing activity of the biocathode was significantly higher than that of the control electrode, which verified the catalytic activity of the microorganism. And under the applied voltage of 0.8 V, the current density and hydrogen yield produced by the thermophilic biocathode were 10 times that of the uninoculated control cathode. And the recovery rate of cathode hydrogen reached 72%.
In 2017, Jafary et al.  used sulfate-reducing bacteria as catalysts for the MEC cathode hydrogen evolution reaction, by changing the cathode electrolyte pH from 3.5 to 6.5, the applied voltage from 0.6 V to 1.3 V, and the sodium sulfate concentration from 0.1 g/L to 1.0 g/L, explored the effect of pH, applied voltage and catholyte concentration on MEC hydrogen evolution performance. The results showed that a medium rich in H2, acidic (pH = 4), and low sulfate content (0.2 g/L) accelerated the formation of a biocathode and increased the rate of hydrogen generation. Compared with the uninoculated graphite felt cathode, the biocathode MEC's HER potential was reduced by 430 mV, and the hydrogen production was increased by a factor of six, and the energy required for the HER reaction of about 1 kWh/m3 was saved compared to the uninoculated graphite felt cathode. This report provides a new way for the clean production of hydrogen in biocathode MECs.
Cathode reaction catalysts of MEC usually use a certain chemical catalyst or a certain type of microbial catalyst alone to attach a conductive carrier to catalyze the hydrogen evolution reaction, but there is little research on the chemical catalyst and microorganisms catalyzing the hydrogen evolution of MEC cathode together. Chen et al.  prepared PANI/MWCNT composite materials and used the synergistic coupling effect between the materials and cathode microorganisms to improve the hydrogen evolution performance of single-chamber membrane-less MEC. Compared with pure biocathode without modification of composite materials, the modified biocathode had higher current density. And the research showed that the types of microorganisms and the homogeneity of the microorganisms in the modified biocathode were different from those in the unmodified biocathode. The Pearson correlation coefficient of the two cathodes was only 32.09%, indicating the difference in the types and numbers of microorganism. When the applied voltage was 0.9 V, the hydrogen production rate of the composite material modified microbial cathode was 0.67 m3-H2·m-3·d-1, the coulombic efficiency was 72%, the cathode hydrogen recovery rate was 42%, and the input power efficiency was 81%. This research provides a new idea for the study of the combination of common chemical catalysts and microbial biocathode.
The common dual-chamber MEC hydrogen evolution reactor uses a microbial anode for the anode, and a non-biological common chemical catalyst cathode for the cathode [97, 107, 108]. However, because the cost of preparing such a cathode is too high, researchers began to develop MEC hydrogen production reactors in which both the cathode and anode are microorganisms. Table 4 is a categorical display of the effect on the hydrogen evolution performance of the MEC using biocathode.
Table Ⅳ. Effects of cathode and/or anode microorganisms on MEC hydrogen evolution performance.
In 2008, the microbial biocathode MEC developed by Rozendal was a semi-biological MEC hydrogen evolution reactor . Therefore, in 2010, Jeremiasse et al. began to study the full bioelectrode hydrogen evolution MEC whose cathode and anode were both microorganisms. The principle of microbial catalytic cathode and anodic oxidation and reduction reactions in MEC were investigated. At the same cathode potential (-0.7 V), the microbial biocathode MEC (3.3 A/cm2) had a higher current density than the non-biocathode MEC (0.3 A/m2) using graphite felt. However, compared with the Pt cathode MEC, the start-up time of the full biocathode MEC was longer. This study showed that the current density of the full biocathode MEC was greater than the semi-biocathode MEC .
Recently, in order to compare the difference between full bioelectrode hydrogen evolution MEC and semi-biological hydrogen evolution MEC. Jafary et al.  developed a MEC with both cathode and anode microorganisms. After cultivating the microbes of the anode and cathode with semi-bio anode MFC (HB-MFC) and semi-bio cathode MEC (HB-MEC), respectively, using their anodes and cathodes to assemble a full bio MEC hydrogen evolution reactor (FB- MEC). When the oxidation reaction of the MEC anode and the reduction reaction of the cathode were biocatalyzed by the biofilm attached to the electrode in the FB-MEC, the initial potential of the HER of the non-biological control system was reduced by 500 mV, and the maximum current density was also increased by 6.5 times. But compared with HB-MEC, the hydrogen evolution rate of FB-MEC was reduced. However, the cathode hydrogen recovery rate increased from 42% to 65%, indicating that the effective oxidation and reduction efficiency of the full bio-MEC reactor were higher than that of the semi-bioreactor.
A. Noble metal based electrocatalysts
B. Transition metals (TMs) as catalysts
1. Transition metal phosphides
2. Transition metal carbides
3. Transition metal nitrides
4. Stainless steel alloy cathodes
5. Nickel and Ni-based alloys
C. Nanostructured materials as catalysts
D. Biocathode in MEC
1. Single microbial biocathode MEC
2. Chemical catalyst synergistically coupled with microbial biocathode MEC for hydrogen evolution
3. Full biocathode MECs
The hydrogen evolution performance of MEC was affected by many factors, such as the structure of the electrolytic cell, the activity of the anode microorganisms, the materials that constitute the electrolytic cell, and the environment in which the battery operates. These should affect the hydrogen evolution efficiency of the MEC. Additionally, MEC is a technology developed on the basis of microbial fuel cell. Therefore, the factors that affect the activity of MFC anode are also parameters affecting the MEC anode microorganism electricity generating ability. Understanding these influencing factors of MEC hydrogen evolution performance can help to reduce or eliminate the influence of these factors as much as possible, thereby improving the MEC hydrogen evolution rate, and providing help for the practical promotion and application of MEC hydrogen production technology in the future. In this section, the influence of temperature, catholyte pH, applied voltage, and catalyst activity on MEC hydrogen evolution performance are discussed.
During the operation of the MEC, temperature is one of the important parameters that affect microbial activity, dominant strains, and reactor performance. Appropriate culture temperature has a positive effect on microbial reaction kinetics, accelerating the reaction rate on the cathode and increasing the rate of proton transfer in the reaction solution. The performance of MEC and the speed and structure of anode biofilm formation were affected by the temperature. Previous studies have shown that temperature affects the anode potential of the MEC, which actually represents the life metabolism activity of anode microorganisms . Although the living temperature range of microorganisms is very wide, the operating range of temperature in MFC and MEC is very narrow, generally 20-45℃, because most microorganisms show higher activity in this temperature range [98, 118, 119]. The operation of the laboratory MEC is generally at room temperature (25 ℃). The optimal temperature of the continuous MEC reactor is 30 ℃ to 35 ℃ [100, 120, 121]. If the temperature is too high, it will be detrimental to the activity of electricity-producing microorganisms [120, 122].
Kyazze et al.  used a two-chamber concentric tubular microbial electrolysis cell, in which the anode substrate was sodium acetate, the pH of the anolyte was controlled to be around 7.4 by adding 1.2 mol/L HCl, and the catholyte was a 50 mmol/L phosphate buffer (PBS) with pH 7.0. The applied voltage was about 0.92 V (vs. Ag/AgCl). They studied the current density variation of MEC and hydrogen production with temperature. Current density and hydrogen production both increased in the range of 20-30 ℃, and then decreased as the temperature rised to 53 ℃. The optimal operating temperature of MEC was about 30 ℃ (173 mL·L-1·d-1, current density was 1.69 A/m2), and the average hydrogen generation rate was 56.5 mL/d. At room temperature (23±1.4 ℃), the average hydrogen generation rate of MEC is 42.2 mL/d (current density is 1.33 A/m2). After the study temperature was raised to 53 ℃, the anode microorganism recovered its activity after dropping to room temperature, indicating that the MEC anode electricity-producing microorganism had a higher elasticity to temperature. In addition, the optimal temperature of MEC is 30 ℃, which also shows that the practical application of MEC is suitable for countries with higher temperatures.
However, other studies have shown that when glucose was used as a substrate, hydrogen can also be generated by MEC at low temperature (4 ℃). Lu et al.  used a single-chamber membrane-less MEC  to prove that under low temperature (4 ℃) conditions, glucose-fermenting bacteria can survive in large quantities, and decompose glucose to generate electrons and obtain hydrogen. And the hydrogen yield at low temperature (6 mol-H2/mol-glucose, the rate was (0.25±0.03)-(0.37±0.04) m3-H2·m-3· d-1) was equivalent to that under medium temperature (25 ℃, 5.8 mol-H2/mol-glucose, the rate was 1.01 m3-H2·m-3·d-1). In addition, this study shows that another isotropic interaction in the MEC at 25 ℃, such as methane production and homolactic acid production, was slightly undetectable at 4 ℃ since Dysgonomonas (36.6%) is the main microorganisms instead of methanogens or others at low temperature. And at 25 ℃, bacteroides (accounting for 21.5% of the community) is the main genus in MEC isolates. This research has laid a certain foundation for the development of MEC using only carbohydrates to produce hydrogen at low temperatures. Furthermore, Wang et al.  studied influence of temperature and electrolyte pH on hydrogen production through simultaneous saccharification and fermentation (SSF) of lignocellulose in MEC. The influence of temperature and electrolyte pH on hydrogen production is shown in FIG. 6. It indicates that a moderate temperature (35-40 ℃) of initial anolyte is suitable for the growth and metabolism of exoelectrogenic microbes and can improve metabolic activities of these microbes and cellulase during SSF. While a near neutral pH condition is appropriate for the growth and activity of the microbials in MEC to promote hydrogen production performance.
Figure 6. Influence of temperature and initial anolyte pH on hydrogen production. (a) Current change of the MEC system at different initial anolyte pH. (b) Effect of the initial anolyte pH value on hydrogen production of the MEC system. (c) Current change of the MEC system at different temperatures. (d) Effects of temperature on hydrogen production. Reproduced from Ref. with the permission of Elsevier.
The nature of catholyte in MEC is another important parameter that affects its hydrogen evolution efficiency. There are many electrolytes used to study the hydrogen evolution of MEC. The pH of the electrolyte will affect the activity of electrochemically active bacteria and be used to control the redox reaction potential on the electrode. In addition, it also affects the activity of methanogenic bacteria [43, 125, 126]. Therefore, the following work has been done to study the effect of MEC electrolyte pH on hydrogen evolution performance. Kyazze et al.  have shown that running at a low pH cathode can increase the rate of hydrogen production and reduce the total electrical energy added to the system. When a voltage of 0.6 V was applied, the order of cathode overvoltage was pH = 5 (152 mV)>pH = 7 (132 mV)>pH = 9 (116 mV). In addition, Yossan et al.  studied the effects of 12 electrolyte solutions with different pH values on the hydrogen yield of the dual-chamber MEC at 20 ℃, 0.8 V applied voltage, and 10 Ω resistance. The results of the study are presented in Table 5. The maximum hydrogen yield (QH2) of 100 mmol/L PBS (pH = 9.2) can reach 0.237 m3-H2·m-3·d-1, and the hydrogen yield of electrolytes with different concentrations of NaCl was also relatively high. 134 mmol/L NaCl (pH = 12.2) can reach 0.171 m3-H2·m-3·d-1, in addition, the acidic electrolyte (pH = 9.0) adjusted with sulfuric acid also showed excellent hydrogen production effect (QH2 = 0.171 m3-H2·m-3·d-1). However, the acidic electrolyte with pH adjusted by sulfuric acid (pH = 11.9) had the lowest hydrogen yield, even the same as that of deionized water. Rago et al.  learned through research that compared with the neutral electrolyte (pH = 7.3), alkaline (pH = 9.3) bioelectrochemical hydrogen production performance was better than conventional neutral electrolyte, which can reach hydrogen production efficiency of 2.6 vs. 1.2 LH2·LReactor-1·d-1 and current density of 50 mA/m2.
Table Ⅴ. Coulombic efficiency and hydrogen production of different catholytes in MEC .
Recently, Wang et al.  used a double-chamber glass tube MEC hydrogen evolution reactor with a bipolar membrane (BPM) as a diaphragm, with an external resistance of 10 Ω, an external voltage of 0.9 V and sulfuric acid as its catholyte. In order to explore the effect of pH on the hydrogen evolution performance of MEC, the catholyte H concentration was changed. The results showed that the maximum current density of MEC decreased almost linearly with the increase of pH in the pH range of 0.5-3.5, and the maximum current density was obtained when pH = 0.3. However, when the electrolyte was changed to a phosphate buffer solution with pH = 7.5, the jmax of MEC was similar to that at pH = 3.5, reflecting the importance of the H concentration (pH) in the electrolyte and the catholyte's proton concentration had a great limitation on the performance of MEC when pH was low. It also shows that the acidic electrolyte used by MEC can effectively increase its current density compared with the traditional neutral electrolyte.
From a thermodynamic point of view, MEC hydrogen evolution is an endothermic reaction and does not occur under spontaneous conditions. Under standard conditions, if the anode uses acetate as a substrate, the standard potentials for anodic acetate oxidation and the reduction of cathode proton hydrogen to hydrogen are 0.290 V and 0.414 V, respectively . However, due to the existence of over-potential and internal resistance, it is necessary to apply an external voltage of 0.3-1.0 V [129-131]. Moreover, different applied voltages have different effects on the hydrogen evolution performance of MEC. So far, the actual applied voltage of hydrogen production by MECs reported is at least 0.22 V . Higher current density will increase the rate of H2 production and recovery, because the high external voltage will produce higher performance of the substrate used, the increased voltage promotes proton transfer and reduction. However, higher voltage will also make the applied cost higher, resulting in a lower energy recovery rate. In order to cut down the cost of hydrogen production, the applied voltage of MEC for hydrogen production should be as low as possible .
If the applied voltage is greater than 1.0 V, the purpose of saving electricity costs to produce hydrogen cannot be achieved. At the same time, the applied voltage should not be too low, because if the applied voltage is less than 0.3 V, there will be problems such as the instability of the MEC hydrogen production system and the low cathode hydrogen recovery rate (HPR) [12, 43, 133]. In addition, the applied voltage affects the type and ratio of anode microorganisms, thereby affecting the electrical activity of microorganisms [43, 132, 134, 135]. In addition to the applied voltage, there are many factors that affect the hydrogen evolution of MEC. When studying the impact of other factors on the hydrogen evolution of MEC, the researchers, when designing the applied voltage of the experiment is generally greater than or equal to 0.7 V [43, 136], mainly shorten the experimental cycle period and save time costs. Almatouq et al.  used a dual-chamber MEC to study the effect of applied voltage on the hydrogen evolution efficiency of MEC. The results showed that the gas produced by MEC was variable. Under the same conditions, increasing the voltage will increase the current density. When the applied voltage was increased from 0.4 V to 1.2 V, the hydrogen production increased by 4 times. Cheng et al.  used a dual-chamber MEC to study hydrogen evolution with acetate as a substrate and found that when the applied voltage was increased from 0.6 V to 0.8 V, the hydrogen evolution rate of MEC increased from 1.1 L-H2·L-1·d-1 to 1.5 L-H2·L-1·d-1, an increase of 36.4%.
Tartakovsky et al.  found that whether or not there was a proton exchange membrane (PEM), the coulombic efficiency generally increased with increasing voltage. The high coulombic efficiency indicated that more electrons were present in H2. However the study also pointed out that the presence or absence of PEM could affect the degree of coulombic efficiency increase. The coulombic efficiency of MECs that did not use PEM increased more significantly, from less than 20% at 0.4 V to 90% at 1.0 V, while MECs with PEM only increased to about 45%. However, when the voltage reaches a certain level, the hydrogen production effect and organic consumption will not continue to increase, and it will have a negative impact in severe cases. This is related to the irreversible impact of excessively high voltage on the metabolism of electricity-producing microorganisms and the oxygen released by the anode on anaerobic microorganisms. The results of Mohan et al.  showed that the change in applied potential affected the activity of dehydrogenase, which in turn affected the rate of substrate degradation, the production of volatile fatty acids, and biological hydrogen production. The dehydrogenase activity and the biological hydrogen production activity had maximum values at the applied potentials of 0.6 V and 1.0 V, respectively. Liang et al.  reported a new catalyst of phosphating cobalt (CoP) acicular nanoarray in-situ growing on a 3D commercial nickel foam forming CoP-NF. The schematic diagram and the synthesis process are shown in FIG. 7 (a, b). As is shown in FIG. 7(b), the CoP-NF demonstrates higher electrochemical active surface area of electrolyte interface for its acicular nano-structure. In MEC, the input energy and the energy consumed by the substrate were recovered in the form of hydrogen, while the total energy efficiency depended on the hydrogen evolution performance of the cathode. When the applied voltage varied between 0.5 V-0.8 V, the energy input efficiency, the substrate energy efficiency and the total energy efficiency of CoP-NF were much higher than those of other catalytic materials, which proved the superior hydrogen evolution performance of CoP-NF (FIG. 7(c-e)).
Figure 7. (a) Schematic diagram of MEC in this study. (b) Composite diagram of CoP-NF. (c) Energy efficiencies related with input electricity (ηE), (d) consumed substrate (ηs), and (e) overall energy efficiency (ηE+s) in MECs equipped with CoP-NF, Pt/C-NF, Pt/C-CC, and NF cathodes. Reproduced from Ref. with the permission of Elsevier.
There are many types of microorganisms in MEC anodes. The types and ratios of anode microorganisms are different, and their functions are also different. For example, the competition between methanogens and electricity-generating microorganisms in a single-chamber MEC seriously affects the hydrogen evolution performance of the MEC system . Since the anode conditions of the MEC used for hydrogen production and the microbial fuel cell (MFC) are the same, there is almost no difference between the electricity-producing microorganisms of the MEC and MFC.
There are several inoculation methods for anode microorganisms in MEC: (i) Use MFC to screen domesticated microorganisms until a stable current can be generated, and then directly connect MFC to an applied voltage, convert to MEC and start [96, 141]. The advantage of this inoculation method is that it does not change the operating conditions of the microorganisms, and the type and number of anode microorganisms do not change, so that the MEC can be started as quickly as possible. (ii) In some cases, the microbial-enriched anode that has been cultured and domesticated is used as the cathode of the new MEC by replacing the anode and cathode of the battery. At this time, the microorganisms in the new MEC system are not used to generate electricity, but as a biocatalyst to accelerate the generation of hydrogen gas [11, 110]. (iii) Some studies directly use municipal wastewater, brewery wastewater, pig farm wastewater, food processing wastewater and activated sludge as anode microbial seeds, because most wastewater contains a large number of mixed microorganisms, which is beneficial to MEC systematic research [27, 138, 142]. (iv) Use cultured pure bacteria [42, 143]. (v) Use the running MEC or MFC anolyte or anode biofilm as seeds, directly used for the newly constructed MEC hydrogen evolution research [101, 137, 144].
In recent years, many studies have been devoted to exploring the metabolic processes of microorganisms present in the MEC biocathode to catalyze the hydrogen evolution reaction, which provides guidance results for the development of more biocathode catalyzed hydrogen evolution reactions . Existing reports  used sulfate-reducing bacteria (SRB) as a catalyst for hydrogen evolution from MEC biocathode. The effects of different cathode seeding methods on MEC hydrogen evolution performance were compared. Inoculate the MEC cathode by the following two methods: (1) First add H2, CO2 and sulfate to the palm oil wastewater mixed cultured source SRB and ectopically enrich it, and then introduce this solution rich in sulfate reducing bacteria into the cathode chamber. After that supply the appropriate electrolyte to the system and apply the required voltage to complete in-situ enrichment of the cathode of the SRB. (2) The in-situ enrichment of the palm oil wastewater mixed cultured source at the anode of the MFC reverses the polarity. The MFC bioanode is converted into a MEC biocathode, and suitable medium and voltage are applied to run the MEC reactor. It is found by comparison that the main product of the biocathode (MEC-O) inoculated according to method (1) is hydrogen, which also contains a small amount of methane, while the product of the biocathode MFC-MEC inoculated according to method (2) is only methane. This result shows that we can obtain the cathode reduction products we need by changing the inoculation method of MEC cathode microorganisms, which is beneficial to the industrial practice of MEC.