Metallic bionanocatalysts: potential applications as green catalysts and energy materials

Summary Microbially generated or supported nanocatalysts have potential applications in green chemistry and environmental application. However, precious (and base) metals biorefined from wastes may be useful for making cheap, low‐grade catalysts for clean energy production. The concept of bionanomaterials for energy applications is reviewed with respect to potential fuel cell applications, bio‐catalytic upgrading of oils and manufacturing ‘drop‐in fuel’ precursors. Cheap, effective biomaterials would facilitate progress towards dual development goals of sustainable consumption and production patterns and help to ensure access to affordable, reliable, sustainable and modern energy.


Introduction
In the late 1990s, bacteria were reported to recover soluble palladium (II) via reduction into cell-bound precious metal (PM) nanoparticles (NPs) (Lloyd et al., 1998) with high catalytic activity (Baxter-Plant et al., 2003). Many authors have reiterated the scope for bio-PM catalysts (e.g. reviews by Deplanche et al., 2011;De Corte et al., 2012;Castro et al., 2014;Kulkarni and Maddapur, 2014;Rai et al., 2015;Singh, 2015;Ashok, 2016). A consultancy report (Catalytic Technology Management, unpublished, 2009) concluded that a 'me too' catalyst must be more active than those currently available, cheaper or both. The paradigm bacterial 'bio-Pd(0)' has significant potential applications in 'green chemistry' and environmental nanotechnology, but the criteria cannot yet be met in full due to high costs of (i) growing dedicated bacteria and (ii) precious metals; (iii) retention of the catalyst for re-use and (iv) potential catalyst poisoning at high reaction temperatures (e.g. by sulfur via degradation of the biomaterial). Waste yeast and bacteria have been successfully reused in 'second life' following primary fermentations (Dimitriadis et al., 2007;Orozco et al., 2010;Zhu et al., 2016) while waste precious metals have been bio-reprocessed into active neo-catalysts (e.g. Mabbett et al., 2006;Deplanche et al., 2007;Yong et al., 2010Yong et al., , 2015Murray et al., 2017a). Metal attrition from cells was negligible, enabling catalyst re-use [e.g. 6 cycles (Bennett et al., 2013)] and also as an immobilized catalyst (Beauregard et al., 2010). The highest 'green' potential probably lies in 'tandem' one-pot reactions which (e.g.) combine a biotransformation with a bio-Pd-catalysed step [e.g. in an enantioselective deracemization reaction (Foulkes et al., 2011)] although the low Pd loading necessary to permit continued physiological activity may not be optimal for chemical catalysis.
Concerns about nanoparticles in the environment (Valsami-Jones and Lynch, 2015) (biomass will eventually decompose) may restrict pollutant remediation to ex-situ applications. Catalyst poisoning is less relevant in a 'dirty' process, but this requires cheap, disposable catalyst. A life cycle analysis is needed to determine where bio-metallic catalysts may outcompete traditional comparators, taking into account socio-environmental as well as economic factors.
Bio-precious metal materials are emerging in energy applications (Table 1). Large-scale oil production may justify a once-through catalyst if a low-grade mixed metal 'dirty' catalyst can be used. On the other hand, in synthesis of fuel precursors from waste CO 2, some bionanoparticles [e.g. structured Pd/Au core shells (Deplanche et al., 2012)] could potentially be used for making (e.g.) formic acid electrochemically (Humphrey et al., 2016). Liu et al. (2016) reported a bio-Pd/Au alloy with electrocatalytic activity, but the thickness of the Pdshell is critical for product selectivity [e.g. for formate production a Pd-shell of 10 nm is optimal (Humphrey et al., 2016)]; such fine structure control is probably beyond the reach of biosynthetic capability. Moreover, Bio-Pt E. coli cleaned in NaOH gave enhanced cyclic voltammetry response, cf. sintered material. Bio-Pt D. desulfuricans was better when cleaned in phenol-chloroform. Max. current was 8.5 lA and 25 lA, respectively (glassy carbon rotating disc electrode) (see Table 2) 10 Application Comments Ref (b) Upgrading of 5-hydroxymethyl furfural (5-HMF) into 2,5, dimethyl furan (drop-in fuel) DMF production Pd-based catalyst supported on Bacillus benzeovorans (Table 3). The conversion was the same as Pd/C catalyst, but selectivity was higher 11 (c) Catalytic upgrading of pyrolysis oil Pyrolysis oil algal (Chlorella) Oil was from algae (Chlorella) slurry. Bio-Pd D. desulfuricans (5 wt% Pd:biomass; 5 wt% catalyst per reaction) and Pd/C (10% wt% catalyst per reaction: 15 g oil; 4 h; 325°C) had similar activities, O 2 and N 2 contents were reduced by 65% and 35% respectively. Bacterial component of catalyst contributed to bio-oil yield. Nanoparticle size increased in bio-catalyst but not Pd/C. Catalysts were reclaimed but not evaluated in repeated reactions.

12
Pyrolysis oil beechwood Oil was from beechwood (fast-pyrolized 500°C; < 2 h). Catalysts: 5 wt% Pd E. coli , 5 wt% Pd/C and commercial Pd/Al 2 O 3 (0.6 g catalyst to 20 g oil). Minimum viscosity was~0.-035 Pas (250°C, 3 h), increasing thereafter. Bio-Pd gave similar deoxygenation (~20%) at 160°C but performed less well at higher temperatures, attributed to higher coking than with Pd/C at > 259°C (see Fig. S1, As an alternative, chemical upgrading of CO 2 into hydrocarbon fuels is scalable. Recent advances employ the reverse water gas shift reaction (reduction of CO 2 to CO) in tandem with Fischer-Tropsch chemistry to convert the more reactive CO into hydrocarbons. Consuming CO (a catalyst poison) shifts the equilibrium towards products in the reverse water gas shift reaction, promoting CO 2 consumption. An efficient, abundant, low-cost catalyst(s) must give product selectivity in the desired range (~C5-8 for gasoline;~C9-16 for diesel fuel). Precious metals (Pd, Pt, Ru, Rh) on SiO 2 have been used in the reverse water gas shift reaction, while catalysts for the Fischer-Tropsch process are usually Fe or Cobased, with recent innovations towards one-pot reactions (see e.g. Mattia et al., 2015;Owen et al., 2016;Prieto, 2017) following early work (Dorner et al., 2010) that showed conversion of CO 2 to hydrocarbons (41% conversion and a C2-C5+ selectivity of 62%) using a doped Fe-based system. The potential for using biogenic catalysts has not been explored although a paradigm hybrid bio-magnetite/Pd(0) catalyst has been reported . When made from waste, biogenic precious metal catalysts (e.g. Pd/Pt mixtures~30% Pd) also contained other metals, for example from Degussa processing waste: Al (42%), Ag (6%) and Mg (3%) and from spent automotive catalyst leachate: Fe (14%), Mg (12%) and Al, 27%), i.e. metals that were present in the original solid material (Mabbett et al., 2006;Macaskie et al., 2010). Potentiation of catalytic activity occurred (e.g. more than 10-fold in the case of reduction of Cr(VI) to Cr (III)) using these waste-derived mixed metal catalysts . Hence, exploration of bio-catalyst made from automotive leachate may be warranted for CO 2 valorization, particularly as waste bacteria for use as catalyst support (Orozco et al., 2010;Zhu et al., 2016; Stephen et al., this volume) are readily available from other scalable biotechnology processes. However, selectivity towards specific hydrocarbon products may not be compatible with such economies.
In contrast to CO 2 -valorization, 'proof-of-principle' application of biogenic catalysts has been shown in four key areas of energy and fuels (Table 1).

Fuel cell electrocatalysts
Fuel cells comprise an anode [where fuel, e.g. H 2, is split catalytically to give electrons (current) and protons], a cathode (where protons combine with O 2 in air to give water) (Kraytsberg and Ein-Eli, 2014) and an electrolyte that allows passage of positive ions between them (Kirubakaran et al., 2009). The polymer electrolyte fuel cell (PEM fuel cell) uses purified hydrogen at low temperatures (80°C) with rapid start up (Mehta and Cooper, 2003). PEM fuel cells are applicable for use in (e.g.) Noble metals are used to split water to make H 2 . (Bio-PMs are not yet tested) 20 Optically active ZnS was made using mine drainage treatment bioprocess off-gas (H 2 S) 21 CuS has been separated from other metals in biogenic metal recovery from waste 22 CuS/ZnS hybrid (chemically made) was used in photocatalytic water splitting 23 CdS quantum dots were made with bacteria and cysteine (Cd is more toxic than Zn; an energy application was not reported).

24
Biogenic metal selenides were made ex-situ vehicles (Hannan et al., 2014) or larger 'stacks' for domestic power (Staffell et al., 2015). Durability targets (internationally) are 5000 h and 40 000 h of operation for automotive and stationary fuel cells, respectively (Rice et al., 2015). The US Department of Defence installed 5 kW PEM fuel cell systems in~40 military bases (cost of > $100 000 per system) but, with an operational lifespan of only 500 h, the systems required overhauling annually (Staffell et al., 2015). The cost of precious metal catalysts is restrictive; other catalysts are under development but the power to weight ratio is key, especially for portable and aerospace applications. Substitution of fuel cell Pt-electrocatalysts (0.2-0.8 mg cm À2 ) would use lighter, equivalently performing (but robust against the high local acidity) metals like Pd (e.g. Meng et al., 2015;G omez et al., 2016), particularly in the cathodic reduction of O 2 (He et al., 2005) which is rate-limiting. Reducing Pt costs/loadings could also be achieved by optimizing Pt nanoparticles (via size control and increased uniformity), or by developing alloys (Zhu et al., 2015). Developments towards microbially derived fuel cell catalysts are shown in Table 1. The anodic reaction is well reported, and the challenge is now to develop an efficient cathodic oxygen reduction catalyst; electrochemical test data are shown in Table 2.
Various catalytic applications have been developed to achieve good yield and selectivity to 2, 5-dimethyl furan. Most have focused on commercial heterogeneous mono and bimetallic catalysts based on (e.g.) Ru, Pd, Pt, Au, Cu (Tong et al., 2011;Hansen et al., 2012;Nishimura et al., 2014;Zu et al., 2014;Luo et al., 2015) which are costly. A bacterial platform would provide a cheaper, sustainable source of supported precious metal catalyst, possibly using metals biorefined from waste sources. As the first step, conversion of 5-HMF to DMF using a bio-Pd-based catalyst and a Pd/carbon catalyst was compared (Omajali, 2015), with better selectivity to DMF observed using the biomaterial (Table 3). Nishimura et al. (2014) showed application of a Pd/Au bimetallic/C catalyst and hence the use of bio-Pd/Au (Deplanche et al., 2012;Hosseinkhani et al., 2012;Liu et al., 2016) is worth evaluating in this application.

Upgrading of pyrolysis oils
Hydrothermal liquefaction (HTL) and fast pyrolysis are thermal treatments of wet organic feedstocks (e.g. agrifood wastes, manure, algae) which produce a highly viscous biofuel (pyrolysis oil). This is unsuitable as an alternative to fossil fuels without further processing. In addition to high viscosity, crude bio-oil contains large quantities of oxygenated molecules which are unsuitable for use directly in vehicle engines and are incompatible for blending with fossil fuels, without further upgrading to remove oxygen (hydrodeoxygenation, HDO).
Hydrodeoxygenation is well studied, but a mechanistic understanding for disassembly of biopolymers and their subsequent deoxygenation is incomplete. New catalytic routes are required to make cost effective drop-in fuels (Huber et al., 2006;Rinaldi and Schuth, 2009). Precious metal-based catalysts feature prominently (e.g. Bouxin Taken from Williams (2015). Pt loading was 5 wt% of the biomass. Bio-Pt E. coli was cleaned using NaOH. Bio-Pt D. desulfuricans was cleaned using phenol-chloroform. As with the anodic and cathodic tests in the PEM fuel cell (Table 1), the E. coli biomaterial was~25% as active (mass activity) as that from D. desulfuricans and had~half the specific activity (mA cm À2 ). However, growth of E. coli is readily scalable and it makes active bio-metal catalyst when used in 'second life' following an independent primary fermentation (Orozco et al., 2010;Zhu et al., 2016). In contrast to E. coli, D. desulfuricans cells are obligately anaerobic, growth is less readily scalable, and they produce H 2 S, a powerful catalyst poison that requires more extensive washing of the cells prior to use. However, a metal bioremediation process that couples excess biogenic H 2 S (used for minewater clean-up with respect to heavy metals (Hedrich and Johnson, 2014)) also produces waste biomass of a sulfate-reducing bacterial consortium which may find a 'second life' use as a bio-metallic catalyst for fuel cell application, mitigating waste disposal costs.  (111) faces (Liu et al., 2017). However commercial precious metal catalysts would be uneconomic; moreover, fuel cell electrocatalysts (above) will compete in parallel for limited global resources. The energy demand of winning precious metals from primary ores (e.g. 14 t of CO 2 is emitted per kg of Pt produced: Anon 2008) is a major consideration; carbon-neutral fuel needs, in itself, to produce a low carbon footprint. Catalytic deoxygenation reactions include dehydration, hydrogenolysis, hydrogenation, decarbonylation and decarboxylation (Fig. S1, Table S1). For fuels in the diesel range, C-C coupling reactions can be achieved through routes such as aldol-condensation, ketonization, oligomerization and hydroxyalkylation, but hydrodeoxygenation is currently considered the most effective method for bio-oil upgrading, improving the effective H/C ratio and leading to hydrocarbons.
To date, several classes of catalysts are reported for hydrodeoxygenation. In addition to precious metal catalysts, other metals (Fe, Ni and Cu) have shown good selectivities in hydrogenation and hydrogenolysis reactions. However, high hydrogen pressures can lead to complete hydrogenation of double bonds (Bykova et al., 2012). Industrial catalysts based on Co-Mo and Ni can provide good hydrodeoxygenation performance, but these deactivate rapidly due to coke formation and water poisoning (Badawi et al., 2011).
The acidic, corrosive biofuels also limit catalyst lifetime and compromise the process economics. Work has focused on upgrading of biocrude oils using various catalysts (e.g. Co-Mo, Ni-Mo Pd/C (e.g. Biller et al., 2015;Si et al., 2017)). Of these, Pd (which is a noble metal and hence is dissolution tolerant) is promising. 5-hydroxymethylfurfural is made within in the product oil (Dang et al., 2016), enabling possible catalytic conversion into 2,5-dimethyl furan into a second fuel stream.
Despite the high costs, precious metal-based catalysts are favoured for bio-oil upgrading (see reviews : Watson, 2014;Pandey et al., 2015;Cheng et al., 2016;Lee et al., 2016;Nam et al., 2017;Fermoso et al., 2017); bio-manufactured catalyst should now be added to the portfolio (Table 1). Bio-reprocessed precious metal waste has not yet been tested as a cheap metal source (c.f. below). The upgrading efficiency, product spectrum and cost savings are being factored into ongoing research via life cycle analysis, towards the dual development goals of sustainable consumption and production patterns for affordable, reliable, sustainable (non-fossil) energy.

Upgrading of fossil oils
The hydrogen economy and carbon-neutral biofuels lag behind the timeline for change articulated by the Stern Review (Stern, 2006). Hence, cleaner production of fossil fuels is vital as these will remain the predominant short-term sources of supply (~80% of global needs; Anon 2014). With globally declining light crude oil reserves, heavy oil and bitumen use will increase from 2 to 7 million barrels day À1 by 2030 (Anon 2015). Oil sands production emits more greenhouse gas and uses more water than conventional light oil production (Findlay, 2016) and additional refining is also required (Huc, 2011). Heavy oil exploitation is complicated by the high viscosity, low hydrogen content and high amounts of resins and asphaltenes. (Bio)geochemical processes evolved the hydrocarbons, leading to materials with high contents of heavy molecules rich in sulfur, nitrogen, oxygen and metals (e.g. Ni, V), with high viscosity and acidity (Head et al., 2003;Huc, 2011). Technologies for heavy oil upgrading have been reviewed (Heraud et al., 2011;Casta neda et al., 2014). Upgrading in situ gives a cleaner production of less viscous oil, which is more easily transported without the use of diluents (Shah et al., 2010). The THAI-CAPRI (Toe-to-Heel Air Injection coupled with Catalytic Upgrading Process In situ) technology combines thermally enhanced oil recovery with down-hole catalytic upgrading of heavy oil into light fractions (Greaves and Xia, 2004;Hart et al., 2014). This catalytic upgrading, using steam, hydrogen and methane, showed significant improvements over non-catalytic thermal processes (Hart et al., 2014). Conventional cracking catalysts such as supported noble metals are prohibitively expensive; one economic option could utilize regenerated catalysts from treated oils but these have lower activity (Hart et al., 2014).
The future is uncertain. Healing (2015) questioned the economics of the THAI process, Findlay (2016)   the wider issues of oil sands processes in terms of performances and economics including greenhouse gas emissions and other environmental impacts. Meanwhile, emerging researches include modelling of the THAI-CAPRI process (Ado et al., 2017) and the successful use of bio-nanoparticle catalysts sourced from wastes ( Table 1). Application of nanoparticles in enhanced oil recovery has been reviewed (Negin et al., 2016;Sun et al., 2017). While bio-nanoparticles are held immobilized on bacterial cells, thermal degradation would evolve them in association with biomass-carbon. Environmental concerns about nanoparticles have been expressed (see earlier), but it is also argued that naturally occurring nanoparticles are ubiquitous in the environment (Monta no et al., 2014) while evidence for natural biogeochemical cycling of platinum has been reported (Reith et al., 2016).

Conclusions and future scope
As far as we are aware, this is the first overview of the potential for biogenic catalysts in various energy applications within the 'whole energy mix picture'. The concept of using microbial technologies to make materials for application to sustainable energy-generating processes (as compared to energy and waste savings via use in 'green chemistry') is a new direction. Palladium occurs in spent nuclear fuels (one ton of spent nuclear fuel contains > 2 kg of Pd or 10% of global requirements: Bourg and Poinssot, 2017). Even though radiation-resistant bacteria are well known, and the ability of microorganisms to discriminate between isotopes of essential metals (Fe, Mo) has been reported (Wasylenki et al., 2007), separation of the active 107 Pd (half-life 6.5 m yrs; 15% of the Pd inventory) from the stable isotopes (85%) is probably beyond the reach of 21st-century biotechnology. Moreover, while biorecovery of precious metals from aggressive solutions (using prepalladized cells, via chemical catalysis using Pd-bionanoparticle 'seeds') has been shown (Murray et al., 2017a), the metal composition of the neo-catalyst reflects the metallic composition of the waste , and hence, selective biorecovery of Pd against higher active radionuclide contaminants would be prohibitively difficult. Photochemical water splitting to make clean hydrogen is a well-established solar technology, best achieved traditionally using noble metal catalysts (Ran et al., 2014). As potential alternatives, various biogenic, optically active, materials have been made (metal sulfides, selenides; Table 1) but as far as we are aware these have not yet been tested in this application. Biotechnologically, hydrogenase-metal selenide hybrids show potential (Ran et al., 2014). Economic attractiveness is boosted by the potential fabrication of such materials from metallic or H 2 S wastes (Nancucheo and Johnson, 2012;Murray et al., 2017b) but, given that the elements are abundant and cheap, the main driver may be waste valorization and mitigation of disposal costs, while incorporation of even a small amount of metal impurity may affect the optical property, and hence, neo-material from metallic waste may not be a useful option.
For biofuels, an unexpected 'biotechnology' has added value towards pine wood-derived pyrolysis oil. Very large areas of pine forest in the USA have been killed by beetles, producing very dry, porous wood. This enables the use of larger wood chips (Luo et al., 2017), reducing comminution costs and energy use. It would be interesting to apply novel biogenic catalysts to pyrolysis oil obtained from this material.