Anaerobic digestion is a commercial technology applied to convert municipal and industrial organic wastewater streams into renewable energy in the form of methane-rich biogas (Batstone and Virdis, 2014). Despite energy recovery, AD present other important advantages such as high organic matter removal efficiency, low excess sludge production and low space requirements (van Lier et al., 2015). Today, AD infrastructure is used to treat a wide variety of organic wastes including (i) sewage sludge, (ii) animal manures, (iii) food and paper industry wastes, including slaughterhouse waste, (iv) energy crops and harvesting residues, including microalgae, and (v) organic fraction of municipal solid waste (MSW) (Romero-Güiza et al., 2016). Nonetheless, digesters configuration is less diverse since most AD plants are either continuous stirred tank reactor or high-rate bed reactors (e.g., upflow anaerobic sludge blanket and expanded granular sludge bed reactors) used for highly particulate and highly soluble wastewater streams, respectively (van Lier et al., 2015; Romero-Güiza et al., 2016). However, the successful and quick development of anaerobic membrane bioreactors (AnMBR) will further expand the application of AD to a range of new substrates within a short period time (e.g., pharmaceutical, municipal sewage, petrochemical, and winery, among others) (Dereli et al., 2012). AnMBR, which combines the advantages of AD and membrane filtration, represent a sound alternative to high-rate bed reactors for intensive AD (Smith et al., 2012). The main advantage of AnMBR over high-rate bed systems is the total retention of particles. Thus, AnMBR (i) produce high quality effluents (free of solids and pathogens) and (ii) retain special microbial communities able to degrade specific pollutants and/or tolerate higher concentration of an inhibitor regardless of its aggregation or sedimentation properties (Dereli et al., 2012). The latter process advantage is key to treat heavy polluted wastewater streams from a variety of industries.

The implementation of AnMBR as a mainline process for domestic wastewater treatment plants appears as a promising technology to improve the economic feasibility of these plants (Smith et al., 2014). The main advantage of this configuration is its capacity to recover most of the energy potential in the wastewater rather than the fraction currently recovered by the aerobic-anaerobic treatment, where the energy potential of soluble organic matter is not recovered but removed by energy-intensive aerobic processes (Mccarty et al., 2011). According to Mccarty et al. (2011) the full anaerobic treatment of municipal sewage by AnMBR will double the energy production of DWWTP, and energy production will exceed the DWWTP energy needs. However, the net energy balance done by Smith et al. (2014) shows that the energy recovery will largely depend on the municipal sewage strength, the membrane flux, and the energy spent on fouling control. Nevertheless, AnMBR feasibility is expected to increase as the technology matures (Mccarty et al., 2011; Dereli et al., 2012; Smith et al., 2014). Beside the energy recovery through biogas production, AnMBR advantages over conventional aerobic systems include lower production of excess sludge and higher effluent quality in terms of solids and pathogens (Dereli et al., 2012; Smith et al., 2014). The latter is especially relevant as reclaimed wastewater reuse (e.g., landscape and crop irrigation and domestic/industrial consumption) becomes a common practice (Mccarty et al., 2011). The main drawback of AnMBR for municipal sewage treatment is that psychrophilic temperature (<20°C) is the only economically feasible option in temperate and cold climates (Smith et al., 2012; Gouveia et al., 2015). The operation of AnMBR at low temperature not only slows down the kinetics of all biological process (organic compounds degradation and biomass growth) but also increases the dissolved methane in the effluent as methane is approximately 1.5 times more soluble at 15°C compared to 35°C (Lettinga et al., 2001; Lin et al., 2013; Ozgun et al., 2013). The recovery of the dissolved methane is key to reach an energy-neutral operation for domestic wastewater but also to reduce greenhouse gas emissions (Smith et al., 2014). Several technologies have been developed to minimize and recover the dissolved methane in the effluent. Giménez et al. (2012) reported that biogas-assisted mixing avoids oversaturation and guarantees a minimum dissolved methane concentration in the AnMBR effluent (the average oversaturation value for AnMBR effluents is 1.5). Stripping the dissolved methane with air seems the simplest option to reduce the methane concentration below the saturation concentration as well as to add oxygen to the effluent stream (Mccarty et al., 2011). The use of degassing membranes has also been suggested (Bandara et al., 2011); nonetheless the energy requirements can be higher than the energy recovered. Another alternative is to use the dissolved methane as carbon source for methanotrophs, which can be combined with other biological process such as denitrification (Strong et al., 2015) and bioplastics production (Strong et al., 2016).

While complete anaerobic treatment of municipal sewage has possibly the highest potential for recovering wastewaters organic energy content, revamping existing aerobic-anaerobic DWWTP to anaerobic facilities could be costly and therefore it may only be possible for new infrastructure (Mccarty et al., 2011).

Another opportunity to reduce the energy needs in DWWTP is anaerobic co-digestion (AcoD) (Mata-Alvarez et al., 2011). AcoD, the simultaneous AD of two or more substrates, is a proven approach to overcome the drawbacks of single digestion, boosting the energy production in AD plants to redirect waste away from landfill toward reuse while utilizing existing infrastructure (Jensen et al., 2014; Mata-Alvarez et al., 2014). AcoD is especially useful for AD systems that are operating under capacity and therefore able to receive onsite or external waste for combined treatment. Although literature values are scarce, up to 30% spare capacity has been reported for DWWTP sewage sludge digesters (Fonoll et al., 2015). MSW and fruit and vegetable waste from food processing industries are the most studied and applied co-substrates in sewage sludge digesters; and several successful full-scale AcoD experience at DWWTP have already been reported (Mata-Alvarez et al., 2014). Zupančič et al. (2008) increased by 45 and 130% the heat and electricity energy, respectively, when increasing the loading rate of the Velenje DWWTP sewage sludge digesters (Slovenia, 50,000 p.e.) a 40% (COD-based) with MSW. Similarly, Koch et al. (2016) reported that the self-generated energy at the DWWTP Garching/Alz (Germany, 30,000 p.e.) increased from 25% up to 78% when adding 10% food waste (mass-based) and substituting the old combustion heat and power units. The authors also estimated that the DWWTP could be self-sufficient if the food dose was increased to 16% (Koch et al., 2016). However, the feasibility of using MSW (even sorted) as co-substrate is dependent on the implementation and operational costs of the MSW conditioning (removal of undesired materials and particle size reduction). In this matter, Krupp et al. (2005) and Bolzonella et al. (2006) reported an industrial cost of 50 and 40 € per ton for MSW co-digestion at the Wiesbaden (Germany, 130.000 p.e.) and Treviso (Italy, 70.000 p.e.) DWWTP, respectively. AcoD between fat, oil and greases (FOG) from the DWWTP grit chamber and sewage sludge have also been trialed in several DWWTP (Long et al., 2012). The use of this onsite waste not only allows improving the performance of the digesters but also represents saving the cost of treating the residue outside the plant. However, FOG dosage is limited to an extra loading rate of 1.0–1.5 kgVS m^-3 d^-1 due to long chain fatty acids inhibition (Mata-Alvarez et al., 2014).

It is evident that the need to make DWWTP energy-neutral is making AcoD an emerging practice; in fact, it is likely that most medium to large size DWWTP will shortly practice AcoD (Arnell et al., 2016). Despite higher biogas yields, AcoD implementation has an impact on DWWTP performance such as supernatant nutrient content, sludge dewaterability, biosolid quality, and biogas composition (i.a. H₂S); all of them directly impacting the DWWTP economic balance. In this sense, Arnell et al. (2016), who modeled AcoD using the Benchmark Simulation Model no. 2, observed that while AcoD had a positive effect on the methane production it negatively affected the effluent water quality and the aeration indexes as well as increased sludge production. Therefore, co-substrate selection and dose should be carefully evaluated since random or heuristic decisions on the co-substrate proportion can negatively affect DWWTP performance (Mata-Alvarez et al., 2011).

Hydrogen (H₂) has emerged as a valuable energy carrier since it does not produce CO₂ during combustion, and it has a high energy density per unit mass (Roy and Das, 2015). The dominant technologies for H₂ production use fossil fuels, consume a lot of energy and have a high carbon footprint; these include natural gas steam reforming (50% world’s production), oil reforming (30%), and coal gasification (18%) (Dincer and Acar, 2015; Roy and Das, 2015). Sustainable hydrogen production needs to rely on environmentally friendly and cost-effective technologies (Dincer and Acar, 2015; Kumar et al., 2015). Biological processes, both autotrophic (e.g., biophotolysis) and heterotrophic [e.g., photo-fermentation and dark fermentation (DF)], are among the more environmentally benign methods for H₂ production (Das and Veziroglu, 2008; Dincer and Acar, 2015). However, DF is the only technology that accomplishes the dual goal of waste treatment and energy recovery as it utilizes organic waste and wastewater as feedstock (Han and Shin, 2004; Das and Veziroglu, 2008).

Dark fermentation is a process (part of the full AD process) where anaerobic and facultative bacteria degrade carbohydrate-rich substrates into simpler organic compounds [mainly volatile fatty acids (VFA)] with simultaneous production of H₂ (Turon et al., 2016). DF feasibility is largely limited by its low hydrogen yield, maximally 4 mol of H₂ per mol of glucose (i.e., it can only recover up to 33% of the biomass energy content). Nonetheless, literature values using mixed cultures and real waste streams (e.g., food waste, lignocellulosic feedstock, agro-industrial waste) rarely exceed 2 mol of H₂ per mol of glucose (Ghimire et al., 2015; Zhang et al., 2016). Besides the low H₂ yields, DF commercial feasibility is also limited by several other factors, including (i) controlling the process end product and H₂ yield, (ii) reducing the presence of H₂-consumers microorganisms such as methanogens and homoacetogens, and (iii) using DF “biogas” as combustible in hydrogen fuel cells (Levin and Chahine, 2010; Ghimire et al., 2015; Roy and Das, 2015). Furthermore, poorly biodegradable substrates such as waste activated sludge and lignocellulosic residues need to be pre-treated (e.g., ultrasonic, acid, alkaline, and thermal) to reach acceptable H₂ yields (Noike and Mizuno, 2000; Cai et al., 2004; Levin et al., 2006; Panagiotopoulos et al., 2009).

It does not seem likely that DF will displace medium-term AD as the main technology to convert organic-rich waste and wastewater streams into bioenergy. However, the capability of DF to procure an effluent rich in VFAs (mainly acetate and butyrate) makes it very attractive to be combined with other process for energy and chemical production as well as to further stabilize the DF effluent. Biogenic process that can be combined with DF comprise photofermentation (H₂ production), microbial electrolysis cells (H₂ production), AD (methane production), microalgae cultivation (add-value products production), bioplastics (PHA production), and sulfate reduction to H₂S (metals precipitation) (Muyzer and Stams, 2008; Ghimire et al., 2015; Roy and Das, 2015; Turon et al., 2016). Compared to undeveloped alternatives, the combination of DF and AD in a two-phase anaerobic system represents a plausible modification for existent AD infrastructure (Cavinato et al., 2011; Giuliano et al., 2014). The DF followed by AD process (also known as biohythane) produces a H₂-rich biogas which improves the thermal efficiency and power output as well as reduces the pollutant emissions of the combustion engine (Porpatham et al., 2007; Moreno et al., 2012). Although the higher efficiency of the two-stage AD over the traditional one-stage AD configuration has been largely proven (Liu et al., 2006; Ge et al., 2010; Riau et al., 2012; Schievano et al., 2014), its full-scale implementation only represents a small fraction of the existing AD infrastructure for municipal waste and wastewater treatment. For instance, in Europe, two-phase systems only represent 7% of the AD infrastructure treating MSW (∼250 plants) (De Baere and Mattheeuws, 2015). According to De Baere and Mattheeuws (2000, 2010), the advantages of the two-stage systems are not enough to compensate the higher investment cost and operation complexity. On the other hand, a multi-stage system (two or more reactors) is the most applied configuration in Germany for the treatment of animal manures, energy crops and other co-substrates; however, the treatment conditions are set to maximize the methane recovery and mitigate emissions rather than producing H₂ (Weiland, 2006; Lebuhn et al., 2014).

Biodiesel is a carbon-neutral energy source to partly replace fossil fuels, especially in the transport sector which is responsible of 23% of the world’s greenhouse gasses emissions (Muniraj et al., 2015; Zhang et al., 2016). Although H₂ and methane (after biogas upgrading) can also be used as vehicle fuels, biodiesel represents a smoother alternative since it can be used in existing engines as well as distribution and supply infrastructure without major modifications (Yusuf et al., 2011; Zhang et al., 2016). Today, most biodiesel (>95%) is produced from the transesterification of edible vegetable oil (e.g., canola, palm, rapessed, and soybean); the so-called first-generation biofuels. However, due to the associated food-versus-fuel competition for land and water the production of biodiesel from non-edible oils (second-generation biofuels) is gaining attention (Ashraful et al., 2014; Jin et al., 2015).

Oleaginous microorganisms (microbes able to accumulate more than 20% of their dry weigh as oil) including microalgae, fungi, yeast, and bacteria, are a promising alternative to vegetable oil since they have faster growth rates than plants (Jin et al., 2015; Muniraj et al., 2015). Feedstock cost is one of the main challenges to make microbial biodiesel profitable; therefore, the combination of microbial lipids production and waste and wastewater treatment has been carefully examined (Ashraful et al., 2014; Jin et al., 2015; Muniraj et al., 2015). Among them, the combination of municipal wastewater treatment and microalgae-based biofuels by phototrophic microalgae in high rate ponds algal has been largely investigated (Craggs et al., 2011; Park et al., 2011; Pittman et al., 2011; Mehrabadi et al., 2015). However, this approach is still limited by several drawbacks linked to microalgae growth rates, lipids yield, and lipid extraction. While lipid extraction challenges are shared among all oleaginous microorganisms (Milledge and Heaven, 2014; Jin et al., 2015), limitations affecting phototrophic microalgae growth rates and lipid yield include (i) light penetration and cells mutual shading, (ii) the supply of nutrients and CO₂, (iii) avoiding contamination by microbes and toxic compounds from the wastewater, and (iv) achieve high lipid yields at high growing rates rather than at under stress treatment conditions (Scott et al., 2010; Liang, 2013; Ward et al., 2014; Mehrabadi et al., 2015). Heterotrophic microalgae can overcome most of the limitations linked to phototrophic microalgae cultivation together with faster growing rates and higher lipid yield (Miao and Wu, 2006; Lowrey et al., 2015). Nonetheless, the dual goal of municipal sewage treatment and lipid production is currently unfeasible (Liang, 2013). Heterotrophic microalgae have also been successfully cultivated using several waste and wastewater streams including molasses (cane, sorghum), crude glycerol (by-product of biodiesel production), sugars from lignocellulosic feedstocks, and VFA-rich effluents from DF (Liang, 2013; Turon et al., 2016). The combination of bioH₂ production and biodiesel by heterotrophic microalgae seems an interesting approach to achieve waste treatment and energy recovery. However, the main challenges are tuning DF (pH, HRT, and temperature) toward a desired VFA distribution and concentration, as well as avoiding bacterial contamination (Turon et al., 2016). Finally, it is worth highlighting that fungi, yeast, and bacteria are emerging as microbial oil producers. Although this approach is still on its early stages their high growth rates, productivities and yields while using a wide variety of carbon source makes them a worth considering alternative to microalgae-base microbial lipids (Muniraj et al., 2015; Whiffin et al., 2016; Zhang et al., 2016).

Single cell protein can be produced by microalgae, fungi, and bacteria and can be used to substitute for conventional agricultural products such as fishmeal and soy, which are major components in feed formula for aquaculture and livestock (Van Huis et al., 2013). Various sources of SCP were trialed as feed additives for cattle, sheep, swine, poultry, and fish (Hintz et al., 1966). The main advantages of microorganisms for protein production are rapid growth, high protein content and the ability to grow on a wide range of substrates (Tusé and Miller, 1984). These alternatives are less dependent (or not at all) of climate, weather, soil characteristics and available land (Moraine et al., 1979).

However, the applicability of SCP as feed additive depends on the composition. Most microalgae can be feed directly to, e.g., cattle but the algal cell walls need to be digested when feed to monogastric organisms (FAO, 2010). Although microalgae are generally less likely to produce toxins and significant amounts of different species are used for animal as well as human nutrition (e.g., Spirulina, Scenedesmus, Chlorella, Dunaliella) (Becker, 2004), the risk of contamination with toxin producing species can only be excluded in pure cultures whereby mixed culture, especially in open systems, have to be closely monitored. Problems might arise from certain species of cyanobacteria which can be part of the consortium and are known to produce toxins (Aráoz et al., 2010).

Single cell protein from fungal biomass, specifically yeast, has been used for a century as animal and human feed additive (Ugalde and Castrillo, 2002). However, depending on the species, fungal SCP might contain elevated levels of nucleic acids and mycotoxin which have to be removed prior of application as feed additive. Both are known to cause severe health effects in humans as well as animals (Speijers and Speijers, 2004). The same applies to bacterial biomass. Depending on the species, a whole variety of endo- and exotoxins can be produced (Anupama and Ravindra, 2000) and the nucleic acid content is generally high (up to 16% of dry weight) which limits the daily intake to a couple of grams per day (Hedenskog and Ebbinghaus, 1972). PPB (Shipman et al., 1975) generally do not produce toxins and have nucleic acids contents comparable to algal biomass and seem suitable as feed additive (Ponsano et al., 2004). A summary of the SCP composition of the different types is given in Table 1.

Besides the general biomass characterisation, the essential amino acid content, required for animal and human nutrition, plays an important role in the SCP evaluation. Table 2 shows the essential amino acid content of the relevant SCP source in perspective to egg albumin, a well-balanced reference source for human nutrition. The table shows that the amino acid profiles of each SCP source compare well with egg albumin whereby microalgae and fungi are deficient in sulfur containing amino acids, specifically cysteine. The data of cysteine are missing for PPBs.

After agricultural products outcompeted SCP, the cost of SCP today are still higher compared to conventional products and production costs cannot fully be covered by the product itself (Ugalde and Castrillo, 2002). Although resource limitation is expected to change this picture in the future, the overall SCP production costs have to be reduced. SCP production can be combined with liquid and/or solid waste treatment. As such, algal ponds can effectively treat wastewater and removed constituents are utilized for algal SCP production (Zittelli et al., 2013). The same has been reported for PPB whereby considerable less research has focussed on this source (Hülsen et al., 2014). In this context, the production of fungal SCP from lignocellulose with solid state fermentation, e.g., white rot fungi does not seems to be profitable, mainly due to transport and fermentation costs (FAO, 2010). However, when the raw materials are low-cost and optimum culture conditions are achieved liquid fermentation and SCP from fungal biomass seems feasible (Begea et al., 2012).

Single cell protein can be additionally used as probiotics. Thousands of tons of microalgae, fungi, and bacteria are used as mixed probiotics for aquaculture worldwide every year, particularly in China [up to 50,000 ton per year for (Qi et al., 2009)]. However, probiotic additions to, e.g., fish ponds are not necessarily considered as SCP feed additives because probiotics are added to the water (with other natural occurring bacteria) rather than being the feed. However, the market for probiotics is massive (globally 19,600 million USD in 2013) and several products are commercially marketed but the production occurs mainly batchwise due to difficulties of industrial scale up (Martínez Cruz et al., 2012).

Large-scale installations are required to resynthesise protein from wastewater constituent in reasonable amounts. For microalgae the protein resynthesis depends on the wastewater, the loads and residence time but also the nitrogen removal efficiency which was reported to be between 36 and 87% for open high rate algal ponds (HRAP) (Shoener et al., 2014). Areal productivities vary between 1.6 and 23.5 g_biomass m^-2d^-1(National-Research-Council, 2012) whereby carbon is partly provided from the wastewater but extra CO₂ is added in most of the cases to drive autotrophic algal growth (Park et al., 2011). Assuming abundant sun energy and carbon supply, nitrogen becomes limiting for the protein synthesis. In this perspective, the main mechanism for nitrogen removal from HRAP is ammonium stripping (Picot et al., 1991; El Halouani et al., 1993) and up to 50% of the nitrogen is dissipated and around 25% are assimilated by the microalgae (García et al., 2000). Additionally, treating non-sterile wastewater is likely to diversify the HRAP community which be comprised of microalgae, plankton, detritus, and terrestrial plant debris, referred to as “Albazod” (Maazouzi et al., 2008) which is likely to impact typical microalgal SCP composition. Since current investment costs exceed the price for economic wastewater treatment with microalgae in closed photo bioreactors by far (Posten, 2009) the SCP production in open systems is limited and losses are only slightly lower compared to agricultural losses [70% loss, (Matassa et al., 2015)].

In this context, the anaerobic growth of PPB as SCP in a closed photobioreactor seems promising as nitrogen removal is non-destructive and losses were reported to be less than 10% with 90% being incorporated into biomass (Hülsen et al., 2016b). Complete assimilation of nutrients by PPB depends on the available organics which have to be added for domestic wastewater but are abundant in industrial sources. Additionally, PPB dominance in a treatment system can be up to 90% (Hülsen et al., 2016a) which would enhance SCP composition. Other bacteria with high nutritional potential such as Cellulomonas sp. and Alcaligenes faecalis were applied to degrade cellulose containing substrates by solid fermentation of, e.g., wheat straw to enrich proteins (Han, 1975). Complete wastewater treatment systems based on these organisms are not reported in literature although Cellulomonas sp. was applied for bio-augmentation in wastewater treatment (Chin et al., 1996).

Also fungal biomass is predominantly applied for solid fermentation whereby wastewater treatment is limited to specific applications such as color and heavy metal removal (Kapoor and Viraraghavan, 1995; Pokhrel and Viraraghavan, 2004). Literature about main line organic and nutrient removal with focus on SCP production is limited (Pant and Adholeya, 2007). Fungal biomass can be applied for sugar and starch containing wastewaters although commercially available technologies would utilize bacteria (anaerobic or aerobic). However, several solid wastes such as corn and sorghum have been used to produce SCP (Anupama and Ravindra, 2000). Other substrates include cellulose and lignin but the pre-treatments to produce accessible substrate is often prohibitive for SCP production (FAO, 2010).

The knowledge of SCP production from the last century did not result in major industrial scale production of SCP after the 1980s due to high costs. Despite the nutritional value of SCP from each source looks promising, microalgae are the only source readily used as animal as well as human feed additive. Fungal biomass used for baking is discontinued today. The increasing resource limitations are expected to drive SCP production and improve the economic feasibility in the future. The application for wastewater treatment with simultaneous production of SCP in large scale is only applied for microalgae. Wastewater treatment with fungi is very limited whereby solid waste treatment seems economically prohibitive. PPB seems to offer another option, with effective wastewater treatment capacities, acceptable nutritional value and high yields but without current full-scale installation.

Table 3 shows the theoretical biomass production based on nitrogen content of the wastewater assuming 30% utilization for soybean production, 25% for microalgae and 90% yield for PPBs. Comparing microalgae and PPB with agricultural plant growth, e.g., soybeans shows that nitrogen fertilizer can be used most effectively by PPBs. The overall yields for biomass production from microalgae and PPB in terms of nitrogen are far higher compared to plants because nitrogen is effectively assimilated. However, due to stripping of ammonia, the nitrogen losses in open HRAP are almost comparable to runoff and volatilisation in agriculture.

Radionuclides metals have unstable atomic structure with excess nuclear energy, which emits ionizing radiation that can affect genetic structure of most living beings, including microorganisms that can suffers acute radiation hormesis phenomenon (Kudryasheva and Rozhko, 2015). Even such properties are not an impediment to have biotechnological applications for radionuclides recovery from mineral rock or radioactive waste and wastewater (Francis, 2012). Most of the recovery technologies entail immobilization of the radioactive metals and recovery after de-structuring of biological waste.

Biosorption plays a key role on biological recovery of radionuclides. Uranium has been previously immobilized by using organisms such as Rhodotorula glutinis (Bai et al., 2014) and Pseudomonas putida (Choi et al., 2009) with efficiencies higher than 70%. Similarly, a simulated metal refinery wastewater containing high ammonium and rhodium concentrations was treated by nitrification-denitrification for N removal followed by Rh recovery by using different microbial adsorbents, achieving >50% recovery efficiency (Manipura and Burgess, 2008). Other microorganisms can actively increase their radionuclides sorption capacity. For example, non-sterilized active anaerobic bacteria, originally used for the treatment of pulp and paper wastewater, accumulated considerable amounts of plutonium, actinium and neptunium, in all the cases increasing their bioaccumulation capacity compared to sterilized biomass, strongly suggesting biologically mediated process (Sasaki et al., 2001). Also, some microbes can enhance chemisorption of radionuclides by promoting their interference with the Fe-S-P nexus. This is the case with Citrobacter sp. that can regulate its phosphate biochemical mechanisms to entrap Np and Pu and co-precipitate as metallic phosphates. The transuranic elements removal by this technology is via a hybrid of bioaccumulative and chemisorptive mechanisms (Macaskie and Basnakova, 1998).

Redox processes can also interfere with immobilization of radionuclides. Anionic Rh can be biologically reduced to Rh(III) by SRB at low pH inside the cell, process mediated by the enzyme hydrogenase. Subsequently, Rh is excreted and precipitated outside the cell. SRB can immobilized up to 66 mg/g of Rh using this mechanism (Ngwenya and Whiteley, 2006). Likewise, Pu can be immobilized by the indirect actions of microorganisms resulting in changes in Eh and its reduction from a higher to lower oxidation state, with the precipitation of Pu, its bioaccumulation by biomass, and bioprecipitation reactions (Francis, 2001).

Generally, most of metallic effluents contain low concentrations of metals that convert the use of traditional chemical methods of recovery a chimera. Therefore, a step of pre-concentration of metals is essential but abiotic pre-concentration methods (e.g., nanofiltration, electrodialysis, or reverse osmosis) are costly technologies, especially for high flows as is the case with DWW and some industrial WW. Partitioning techniques involving microorganisms in novel DWW platforms (Verstraete et al., 2009; Batstone et al., 2015) can be combined with recent advances of metal recovery by immobilization (partition) followed by mobilization of metals from concentrated sludge (release). The metals can be subsequently recovered by chemical or even biological transformations into minerals, with the possibility of fractionation of the precipitation reactions by modulating the precipitant concentration as well as the pH of the medium. This multi-stage process would upgrade the current Partition-Release-Recover for metals recovery from wastewater, and is described in Figure 1.

Bio-electrochemical systems were firstly considered as a green technique to produce electricity from waste water, a so called microbial fuel cell (Rabaey et al., 2004; Franks and Nevin, 2010; Janicek et al., 2014). In such systems the bacteria oxidize organic carbon to CO₂ and use the anode as electron acceptor. BES technology experienced a dramatic development over the past decades with ever widening applications beyond being a power source (Rabaey and Rozendal, 2010; Logan and Rabaey, 2012; Tremblay and Zhang, 2015). Due to technical challenges (current densities, over-potential etc.) and due to the low value of electricity, bio-electricity still remains a niche application for remote deployment that lacks alternative power sources, for instance deep sea applications (Lovley and Nevin, 2011). In a recent development, the production of hydrogen or hydrogen peroxide have been considered as alternatives electricity (Foley et al., 2010) and these systems, termed microbial electrolysis cells, can be used in conjunction with wastewater treatment leading to recovery of phosphorous (Cusick and Logan, 2012) and methane production (Villano et al., 2013).

Originally the term microbial electrosynthesis (MES) referred to the assimilation of CO₂ by microbes using a cathode as the source of electrons, however, the term MES has constantly evolved and now considers more broadly microbial bioconversions powered by a BES and leading to production of organic chemicals.

On the side of the anode, electricity production dominated the focus so far, but in a recent study it has been shown that using a pure culture of a strict aerobe, breathing the anode allowed the high yield production of a chemical precursor for industrial antioxidant production (Lai et al., 2016). This is a key development, as it highlights the advantage of using a strict aerobe in such a system. The cells are unable to convert the substrate to undesired fermentation products and due to the constrained metabolism full oxidation to CO₂ is low. The use of a pure culture also has the advantage that the field is opened up to metabolic engineering for rate enhancement of product diversification.

On the cathode side, many more works have been published that achieved production of chemicals. First demonstrated by the Lovley group (Nevin et al., 2010) pure cultures of Sporomusa ovata could produce acetate from CO₂ and electricity. This could recently be extended to a mixed culture system (Jourdin et al., 2014), however, it is important to mention that acetate as a commercial product would struggle to be competitive and is now seen in this context as an interesting option for in situ COD production at waste water treatment plants. Methane can also be produced on a bio-cathode from CO₂ (Villano et al., 2010a) but the commercial value of this remains to be seen.

Bio-electrochemical systems technology can also be used to recover metals and nutrients. For instance as entioned above phosphorous can be recovered (Cusick and Logan, 2012) and also nitrogen removal and recovery either via electrodialysis (Thompson Brewster et al., 2016) or as ammonium bicarbonate from source-separated urine has been investigated (Tice and Kim, 2014). In addition, the BES also offers the possibility to recover metals fom wastewater, recently reviewed (Wang and Ren, 2014).

Apart from the mentioned challenges in maximizing the electron transfer rates and choosing a suitable mechanism, it is not trivial to decide which products could be made in a BES. There are three groups of targets that could in theory be pursued (i) the production of bulk chemicals such as biofuels, plastics or platform chemicals or (ii) the production of high value chemicals such as pharmaceutical precursors or hard to synthesize complex or chiral structures including antibiotics, pesticides, or herbicides and (iii) the production of inorganics such as metal complexes that can serve as fertilizer or a source of valuable metals. The problem here will be to find products with a real advantage over traditional production systems in the first group this means competing with the petrochemistry and highly efficient sugar based bioprocesses. In addition, downstream processing is a big challenge when producing chemicals or fuels from waste streams, especially when looking into chemical feedstocks, where the highest purity is needed. In the second group, the substrate cost is negligible, which makes it even harder to compete on a cost basis, but oxidative process on an anode for instance could avoid the formation of toxic oxygen adducts as a fermentation by-product (for instance epoxide formation in terpene production processes).