Introduction

Nanomaterials like metal oxide nanoparticles have been shown to be an efficient alternative for the remediation of several recalcitrant organochlorinated pollutants such as perchloroethylene (PCE) in municipal wastewater and groundwaters (Albergaria et al. 2010; USEPA 2006, 2017; Hennebel et al. 2009; Khin et al. 2012; Bretón-Deval and Poggi-Varaldo 2017; Bretón-Deval et al. 2013).

Indeed, in order to achieve higher efficiencies in the remediation of chlorinated compounds, several nanomaterials based on iron, silver, nickel, cobalt, and titanium, inter alia, have been used (Lohner and Spormann 2017; Machado et al. 2014). These nanomaterials have been shown to improve/stabilize the process of dehalogenation and often have allowed the recovery of nanoparticles for their reuse (Seabra et al. 2013).

On the other hand, when oxygen is not available in water, several bacterial strains grown in anaerobic and anoxic conditions can degrade chlorinated chemical compounds, usually by reductive dehalogenation (Staniland et al. 2009; Garibay-Orijel et al. 2005). Regarding degradation of chlorinated aliphatics, most of such bacteria are related to Dehalobacter and Dehalococcoides genuses (Bretón-Deval et al. 2016; Judger et al. 2016; Marzorati et al. 2007; Mayer-Blackwell et al. 2017). Several anaerobic bacteria use the organic chlorine(s) present in the organochlorinated compounds as final electron acceptor in a process known as reductive dehalogenation (Jacob and Suthindhiran 2016; Staniland et al. 2009; Sundaram et al. 2012; Yan et al. 2017).

Magnetite Fe3O4 is an iron oxide commonly used in environmental remediation of chlorinated compounds and is typically obtained by a coprecipitation technique based on a mixture of ferric and ferrous salts in the presence of NaBH4 as a typical reductant (Laurent et al. 2008). Most of the research reported about iron nanoparticles (NP) by chemical synthesis is relatively well established (Chen et al. 2014; Corr et al. 2008). More recently, the efforts have been focused on implementing alternative methods in order to minimize or avoid the use of chemical agents and the discharges of toxic effluents (WHO 2000), as well as standardize the optimal concentration of nanoparticles in environmental processes (Thunugunta et al. 2015; Jiménez-Pérez et al. 2012).

Thus, biogenic methods are becoming increasingly used for the synthesis of iron NP (Bharde et al. 2006; Kaufmann and Lovley 2001). In the biosynthesis of magnetite is commonly found species of iron reducing bacteria such as Geobacter metallireducens (Lovley et al. 1993), Shewanella oneidensis (Fredrickson et al. 1998; Ross et al. 2007), Thermoanaerobacter ethanolicus (Roh et al. 2002), and Magnetospirillum magnetotacticum (Noguchi et al. 1999).

Iron reducing bacteria can use Fe(III) in oxidized state under anaerobic conditions as well as poorly secreted crystal Fe(II) in culture medium. Fe(II) in excess can be adsorbed as granular iron hydroxide and transformed into magnetite (Yeary et al. 2005). Magnetite synthesis can be favored by alkaline conditions (Eq. 1)

$$ \left.{}_{2{\mathrm{O}\mathrm{H}}^{-}+{\mathrm{Fe}}^{+2}+2\mathrm{Fe}\ {\left(\mathrm{OH}\right)}_3\to {\mathrm{Fe}}_3{\mathrm{O}}_4+4{\mathrm{H}}_2\mathrm{O}}^{{\mathrm{CH}}_3{\mathrm{COO}}^{-}+8\mathrm{Fe}\ {\left(\mathrm{OH}\right)}_3\to 8{\mathrm{Fe}}^{+2}+2{{\mathrm{H}\mathrm{CO}}_3}^{-}+15{\mathrm{O}\mathrm{H}}^{-}+5{\mathrm{H}}_2\mathrm{O}}\right\} $$
(1)

Typically, bacterial cytochromes are known to be involved in the nanosynthesis of magnetic NP. Gaspard et al. (1998) elucidated the reduction of Fe(III) by Geobacter sulfurreducens. This strain can generate electrons from oxidized NADH which are placed into 89 kDa cytochrome. Afterwards, they are transferred to the periplasmic space of 9 kDa cytochrome from which the electrons are used to reduce Fe(III) to Fe(II) through the 41 kDa cytochrome (Magnuson et al. 2001).

In recent years, Byrne et al. (2015) scaled up the biogenic process for magnetic nanoparticles by Geobacter sulfurreducens; they could produce up to 120 g of magnetite in 24 h. However, this process was carried out under optimal conditions using pure bacterial culture and electron donor and acceptor acetate and fumarate, respectively.

The emerging interest in developing sustainable and ecofriendly platforms are turning to the exploration of wastes that could substitute expensive raw materials and polluting nanosynthesis processes. About this approach, a wide variety of residues such as food, grapefruit, and rice husk wastes have been used (Piccinno et al. 2018; Satari and Karimi 2018; Vilaplana et al. 2010). Also, glass and plastic bag wastes were pretreated and recycled to produce engineered nanomaterials (Samaddar et al. 2018).

In the case of application of magnetite-based NPs in biorefinery processes, there are important potential benefits such as in situ recovery of the NPs which allows a continuous fermentation and separation of products and to reduce the washing time of production cycles, among others (Safarik et al. 2016). Also, hybrid iron nanoparticles such as NiFe exhibit another interesting application that consists of the sustainable hydrogenation of molecules from compounds such as glucose and xylose, which can be transformed to attractive products sorbitol and xylitol (Chieffi et al. 2014).

The objective of this work was to synthesize and characterize iron embedded nanobioparticles (ENBP) by biological means, using saccharified waste liquors (hydrolysates of the organic fraction of municipal solid waste, OFMSW) as well as anaerobic consortia anchored to bioparticles of fluidized bed bioreactors, in a waste-based biorefinery. We hypothesized that the hydrolysate would provide reducing power to the microbes in the bioparticles to effect iron reduction to magnetite. The experiment in our research evaluated the effects of “origin of bioparticles” (either BP from a methanogenic and hydrogenogenic fluidized bed reactors) and “type of precursor iron salt” (iron chloride and iron citrate) on the characteristics of Fe-NP.

Incidentally, but important, the synthesized ENBP could be used for treating groundwaters and wastewaters polluted with organochlorinated compounds and could become, hopefully, an additional source of revenues of waste-based biorefineries (Poggi-Varaldo et al. 2014; Escamilla-Alvarado et al. 2017).

Materials and methods

Experimental design

The experiment was arranged as a 22 factorial. The first factor was the “origin or source of the BP” (either from a hydrogenogenic fluidized bed bioreactor, AnFBR-H, or a methanogenic one, AnFBR-M). The second factor was the “type of precursor salts,” either (i) ferric chloride 0.25 M or (ii) ferric citrate 0.1 M. The effects of the above cited factors on ENBP characteristics were explored. A summary of the experimental conditions in the four treatments of the experiments is shown in Table 1.

Table 1 Experimental conditions for nanodecoration of bioparticles previously grown in anaerobic bioreactors using iron salts and hydrolysates of the organic fraction of municipal solid wastes
Full size table

BPs from anaerobic fluidized bed reactors were sampled and transferred to serum bottles loaded either with (i) ferric chloride 0.25 M or (ii) ferric citrate 0.1 M, additionally supplemented with hydrolysate enough for initial 10 g L−1 or 16 g L−1 reducing sugars, and incubated at 37 °C for 7 days, leading to generation of nanodecorated bioparticles (herein after called embedded nanobioparticles, ENBP). All treatments received hydrolysates of OFMSW enough to give initial 10 and 16 g L−1 reducing sugars for the M and H ENBP, respectively.

Response variables of the characterization experiment were the size distribution, specific surface area, shape, iron content and speciation in NP of the ENBP, and the presence of magnetite and siderite. The experiment was run in two replicates. The control was granular activated carbon (GAC) uncolonized by anaerobic bacteria. The BPs withdrawn from AnFBR-H and AnFBR-M (10 gwb) were added to the mixture of corresponding salt precursor and reducing sugars from hydrolysates in serum bottles (Table 1) for the biosynthesis of ENBP and incubated for 37 °C for 7 days.

Afterwards, the ENBPs were recovered from the liquor by centrifugation at 6000 rpm for 10 min and washed 5 times with ethanol. The solids were dried under N2 inert atmosphere and then pulverized in a mortar for microscopic analysis.

Two lab-scale AnFBRs were supplemented with hydrolysates from OFMSW and operated in mesophilic regime. Both AnFBRs were loaded with GAC as a support material. Anaerobic consortia were grown on the GAC particles, leading to the formation of colonized bioparticles in the reactors. One AnFBR was targeted to bioH2 generation (AnFBR-H), whereas the other one (AnFBR-M) was maintained in methanogenic regime.

Start-up and operation of the anaerobic reactors used in biological hydrogen and methane production fed with hydrolysate from the organic fraction of municipal solid waste

Two lab scales (AnFBR) were used for bioH2 (AnFBR-H) and CH4 production (AnFBR-M). The reactors were made with glass columns with an internal diameter of 4.5 cm and 185 cm height; their working volume was 2.8 L. Each reactor was loaded with 1 kg of granular activated carbon (GAC) with average diameter of 1 mm.

Both reactors were loaded with (at the start-up) and subsequently fed synthetic wastewater during operation. The composition of this wastewater was as follows (in mg L−1): urea (125), FeSO4·7H2O (5), CaCl2 anhydrous (47), FeCl3·6H2O (0.5), SeO2 (0.07), CoCl2 (0.08), KH2PO4 (85), Na2HPO4 (21.7), K2HPO4 (33.4), and NaHCO3 (1000).

The pH for feedwater of the AnFBR-H was adjusted to 6.5 with HCl and supplemented with hydrolysates to give 16 g L−1 in reducing sugars; the pH of feedwater of the AnFBR-M was adjusted to 8.0 using NaOH and supplemented with hydrolysates to attain 10 g L−1 reducing sugars.

The inoculum for the AnFBR-H consisted of 500 g of the contents from a methanogenic reactor (MRI); this inoculum was subjected to heat shock pretreatment (HSP) (Escamilla-Alvarado et al. 2012, 2013) with the purpose of eliminating methanogenic archaea as well as favoring sporulation (and survival) of selected fermentative hydrogen bacteria (like those of the genus Clostridium). Afterwards, the AnFBR-H was treated similarly to the AnFBR-M described below.

The inoculum for AnFBR-M (500 g) was sampled from a methanogenic reactor (MRI) according to previous works (Moreno-Medina et al. 2017; Poggi-Varaldo et al. 2012). After the AnFBR-M received the first load of feedwater and inoculum, it was flushed with N2 for 10 min in order to expel the atmospheric O2 and generate anoxic conditions in the reactor. The recirculation pump was set at a flowrate of 240 mL min−1.

Both reactors were kept in recirculation mode for 1 week without addition of feedwaters to foster the colonization of GAC particles.

Substrate preparation and detoxification

Hydrolysate from OFMSW was obtained by the method of acid hydrolysis (sulfuric acid 3%, 10:1 ratio) reported in a previous work (Hernández-Correa et al. 2017). In order to reduce the concentration of toxicants (phenolic compounds and furfural), the hydrolysate was subject to adsorption with powdered activated carbon (PAC) at 2% (w/v) for 2 h at 150 rpm. After this time, the hydrolysate was settled and further filtered using a Whatman paper filter no. 1 to separate the PAC.

Analyses

Parameters and variables of bioreactor performance

Gas flowrates were measured by acidified water displacement method (Escamilla-Alvarado et al. 2012). Biogas volumes and flowrates were standardized (273 K/101.18 kPa, Avogadro conditions). Concentrations of H2 and CH4 in bioreactors biogas were determined in a GOW-MAC-580 gas chromatograph equipped with a column packed with Molecular Sieve 5A and thermal conductivity detector as reported elsewhere (Robledo-Narváez et al. 2013).

The analysis of VFA (volatile fatty acids) was performed by gas chromatography in a Varian Star 3400 equipped with FID according to Muñoz-Páez et al. (2013). Prior to analysis, the effluents from the AnFBRs were centrifuged at 6000 rpm for 10 min, and the liquid was filtered through a 0.22-μm microfilter. The collected volume 1 mL was kept at − 4 °C until analysis.

COD, VSS, pH, and alkalinity were analyzed according to the Standard Methods (APHA-AWWA-WPCF 1985).

Reducing sugars, total phenolic compounds, and furfural were determined according to the techniques described in previous works (Hernández-Correa et al. 2017; Miller 1959; Graham 1992; Ghanavati et al. 2015).

Characterization of bioparticles and embedded nanobioparticles by scanning electron microscopy coupled to energy-dispersive spectroscopy

Once the AnFBR-H and AnFBR-M exhibited a stabilized performance in hydrogen or methane production (15% H2 and > 60% CH4 contents in biogas, respectively), 10 gwb samples of bioparticles were withdrawn from each bioreactor. Samples were dried under inert N2 atmosphere and stored in serum bottles in anoxic conditions. Then, biofilms developed in AnFBR-H and AnFBR-M were observed in a Dual Beam FIB microscope.

For SEM analysis, the ENBP were analyzed with a field emission scanning electronic microscopy JEOL JJSM-7401F equipped with an EDS microprobe. The conditions were an accelerating voltage from 3 to 20 keV, with takeoff angle of 30° in a magnification range from 500 to 10,000.

Dynamic light scattering

The samples were dried under inert N2 atmosphere. One milligram of pulverized sample was resuspended in 1 mL of deionized water. Afterwards, samples were sonicated for 40 min and immediately placed into a zeta sizer under the following conditions: angle 173 °C, refractive index 2.42, and an absorption index of 0.001 reported for magnetite (Lim et al. 2013).

High-resolution transmission electron microscopy

The ENBPs were analyzed by high-resolution transmission electron microscopy (HRTEM) using a JEOL Model JEM-2010 microscope equipped with an LaB6 cathode, at 200 keV and current of 102 μA (Ramírez-Nuñez et al. 2018). The samples were diluted in methanol and sonicated for 10 min. Once dispersed, the ENBP samples were placed on a copper grid mesh no. 200 for further analysis.

X-ray photoelectron spectroscopy

The surface of bionanoparticles was examined by XPS using a Thermo Scientific spectrophotometer equipped with a monochromatic Al K Alpha X-ray source and excitation line at 1487 eV. The samples were degassed previously under N2 atmosphere for 24 h and then loaded in the analysis chamber at a pressure of 1 × 10−9 Torr (Pérez-González et al. 2018).

X-ray diffraction analyses

The diffraction pattern of the crystalline structure of ENBP was observed in a Bruker D8 Advance Eco diffractometer. The position angle 2θ varied from 5 to 100°, the step degree was 0.02°, and discrimination ranges from 0.18 to 0.25, with rotation of 10 rpm and copper source without monochromator. The X-ray structure was compared to reference crystallographic structure #998-002-0596 from de ICDD database.

Infrared spectroscopy

The ENBPs were pressed previously using KBr and examined in a FTIR Nicolet 6700 spectrometer. The transmittance was determined at room temperature in the interval 400–4000 cm−1.

Specific surface area by Brunauer-Emmett-Teller adsorption method

The samples were dried under nitrogen atmosphere and degassed at 250 °C overnight and then were measured using a Gemini 2360 Surface Area Analyzer according to the instructions in the Operator’s manual V5.01.

Results and discussion

Characteristics of hydrolysates

The efficiency in volume recovery of hydrolysate was 85%; the concentration of reducing sugars was 28 g L−1. Regarding the toxicants, the average concentration of total phenolic compounds was 700 mg L−1, whereas the average concentrations of furfural and 5-hydroxymethylfurfural were 690 and 2600 mg L−1, respectively. At the end of the detoxification with PAC, the toxic compounds decreased at least by 90% and reducing sugar concentration remained close to 25 g L−1. The detoxified hydrolysate was kept frozen and thawed before use for feeding bioreactors and the biogenic synthesis of ENBPs.

Performance of anaerobic fluidized bed reactors using hydrolysates from the organic fraction of municipal solid wastes

In general, the performance of AnFBR-H regarding bioH2 production was poor to moderate. H2 concentration in biogas was equal or less than 15% over time. At the end of S3 (stage 3), there was a slight increase in H2 production, which could result from a greater proportion in the volume of hydrolysate, compared to the sucrose proportion added to the feed wastewater (Fig. S1, in the Supplementary Material document). In the early stages S1 and S2 when sucrose was the main carbon source, the low molecular weight metabolites were predominantly acetic, propionic, and butyric acid likewise ethanol (Table S1). When the feed was switched to 100% hydrolysates, VFA and ethanol concentrations significantly decreased, although lactic acid increased 18-fold.

It has been reported that under certain conditions the anaerobic metabolic pathways lead to ethanol production and acetate, which causes the reduction of theoretical hydrogen yield to 2 mol H2/mol glucose (Eq. 2) (Ghimire et al. 2015).

$$ {\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_6+2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{C}\mathrm{H}}_3{\mathrm{C}\mathrm{H}}_2\mathrm{O}\mathrm{H}+{\mathrm{C}\mathrm{H}}_3\mathrm{COOH}+2{\mathrm{C}\mathrm{O}}_2+2{\mathrm{H}}_2 $$
(2)

In our case, in the early stages of operation the AnFBR-H when the carbon source was sucrose, there was an outstanding accumulation of propionic acid (16 g L−1, S1). This could be explained by the exceptional behavior of selected strains of Clostridium, i.e., Clostridium articum, that is a H2 consumer (instead of H2 producers, such as most known Clostridium strains) and produces propionic acid (Kumar et al. 2006). A decrease in VFA concentration was observed in S3, when AnFBR-H was fed with a greater volume proportion of hydrolysates (60%). In this case, the metabolic deviation favored the accumulation of lactic acid presumably by shifts in the composition of the anaerobic consortium, related to the presence of lactic acid bacteria (Eqs. 3a and 3b) (Muñoz-Páez et al. 2013).

$$ {\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_6\to 2{\mathrm{C}\mathrm{H}}_3\mathrm{CHOHCOOH} $$
(3a)
$$ {\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_6\to {\mathrm{C}\mathrm{H}}_3\mathrm{CHOHCOOH}+{\mathrm{C}\mathrm{H}}_3{\mathrm{C}\mathrm{H}}_2\mathrm{OH}+{\mathrm{C}\mathrm{O}}_2 $$
(3b)

Low hydrogen production in AnFBR-H has been associated to the accumulation of lactic acid which is converted from glucose by species such as Bacillus spp. or by heterolactic fermentation from pentoses and hexoses by strains Lactobacillus spp. and Lactobacillus oligofermentans (Andreevskaya et al. 2016; Ghimire et al. 2015). Interestingly, several works on biological hydrogen production have observed that some strains of Clostridium can enhance hydrogen yields from lactic acid (Chojnacka et al. 2011; Hung et al. 2007; Sikora et al. 2013; Yang et al. 2007). This could explain that despite having a high cumulative concentration of lactic acid (≈ 5305 mg COD L−1) in S5, our AnFBR-H was able to increase its bioH2 production (Fig. S1).

On the other hand, AnFBR-M showed a more regular performance using either sucrose or hydrolysates as carbon sources (Fig. S1). Particularly at the start of S4 (day 60), there was a gradual increase of CH4 production and relatively steady performance until the bioreactor received 100% of hydrolysates in its feed. At this time, CH4 content in biogas reached 70 to 80% v/v.

The soluble metabolites in the effluent of AnFBR-M are presented in Table S1. It was found that once the feed wastewater was turned to a regime based in hydrolysates (S4), the average concentration in VFA decreased in an outstanding way, which was consistent with increased methane production (Fig. S1).

In the early stage S1, when AnFBR-M was fed with only sucrose, there was an increase in VFA concentration that could be due to an imbalance between acidogenic and methanogenic populations reflected in low methane production (Habouzit et al. 2014). In course of time, the decrease in the organic loading rate from 10 to 3.33 g L−1 day−1 possibly allowed for attaining an equilibrium of populations that later promoted a stable performance. A more robust methanogenic performance was observed in S5; there was a decrease in VFA concentration as well as enhanced methane production as it was mentioned above.

Characterization of embedded nanobioparticles

Results of the characterization of ENBP are presented in a summarized form in Table S2 (Supplementary Material document) following the scheme of the 22 factorial experiment.

SEM microscopic characteristics of ENBPs are presented in Fig. 1. Table 2 exhibits results of the specific surface area of ENBPs by BET analysis. The elemental composition of ENBPs by EDS is shown in Fig. 2, with emphasis on total iron. Figures 3, 4, 5, and 6 exhibit the HRTEM micrographs of the ENBPs. The XPS results that provide evidence of iron speciation are displayed in Fig. 7. Further evidence of size distribution and agglomeration of NP in the ENBPs using DLS is shown in Fig. 8. Structural patterns of the oxides found in the ENBPs were determined by XRD and collected in Table 3 and Fig. 9. More information on chemical composition of the ENBPs was obtained by IR and displayed in Fig. 10 as well as Tables S3, S4, S5, and S6 (Supplementary Material document). Finally, results of pH of biosynthesis media and COD removal efficiencies were reported in the text of this subsection.

Fig. 1
figure 1

Scanning electron microscopy micrographs of biofilms developed in 90 days of operation in a AnFBR-H, hydrogenogenic and b AnFBR-M, methanogenic bioreactors fed with hydrolysates from the organic fraction of municipal solid wastes. Different nanodecorated bioparticles synthesized from FeCl3 0.25 M and bioparticles from AnFBR-H (c) and AnFBR-M (d). Frames e and f correspond to nanodecorated bioparticles synthesized from ferric citrate 0.1 M using bioparticles from AnFBR-H and AnFBR-M, respectively

Full size image
Table 2 Specific surface areas and Langmuir isotherms for nanoparticles in embedded nanobioparticles
Full size table
Fig. 2
figure 2

Energy-dispersive spectroscopy (EDS) of several ENBPs a originated in the hydrogenogenic bioreactor and synthesized from 0.25 M FeCl3 and hydrolysates (16 g L−1), b originated in the methanogenic bioreactor and using 0.25 M FeCl3 and hydrolysates (10 g L−1), c originated in the hydrogenogenic bioreactor and synthesized from 0.1 M ferric citrate and hydrolysate (16 g L−1), and d originated in the methanogenic bioreactor and using 0.1 M ferric and hydrolysate (10 g L−1)

Full size image
Fig. 3
figure 3

Images of bionanoparticles ENBP-H FeCl3 by HRTEM: a, b medium resolution, c high resolution, and d selected area electron diffraction pattern (SAED)

Full size image
Fig. 4
figure 4

Micrographs of bionanoparticles ENBP-M FeCl3 analyzed by HRTEM: a, b medium resolution, c high resolution, and d image corresponding to SAED pattern

Full size image
Fig. 5
figure 5

Micrographs by HRTEM corresponding to ENBP-H Ferric citrate 0.1 M: a, b medium and lower resolutions, c image in high resolution of nanoparticle lattice showing the crystal orientation, and d SAED pattern

Full size image
Fig. 6
figure 6

Images of ENBP-M citrate analyzed by HRTEM: a, b low and medium magnifications, c micrography of lattice showing the crystal orientation, and d SAED pattern

Full size image
Fig. 7
figure 7

Photoelectron spectra by XPS corresponding to Fe and O: a, b ENBP-H FeCl3, c, d ENBP-M FeCl3, e, f ENBP-H citrate, g, h ENBP-M citrate

Full size image
Fig. 8
figure 8

Average dispersion of nanoparticles in ENBP analyzed by DLS (dynamic light scattering): a bioparticles from hydrogenogenic bioreactor and using 0.25 M FeCl3, b bioparticles from the methanogenic bioreactor using 0.25 M FeCl3, c bioparticles from hydrogenogenic bioreactor and using 0.1 M ferric citrate, and d bioparticles from the methanogenic bioreactor using 0.1 M ferric citrate

Full size image
Table 3 Planes and positions for the embedded nanobioparticles in our work from X-ray diffraction analysis
Full size table
Fig. 9
figure 9

X-ray diffraction plots corresponding to magnetite, siderite, and hematite present in ENBPs from hydrogenogenic and methanogenic bioreactors, synthesized using either 0.25 ferric chloride or 0.1 M ferric citrate

Full size image
Fig. 10
figure 10

Infrared spectra for embedded nanobioparticles (ENBP): a ENBP-H and b ENBP-M, both synthesized with 0.25 M FeCl3; c ENBP-H and d ENBP-M, both produced from ferric citrate as salt precursor

Full size image

Microscopic characteristics of bioparticles and embedded nanobioparticles

The bioparticles withdrawn from de AnFBR-H and ANFBR-M were observed by SEM in order to evaluate the colonization of the GAC. Figure 1 a and b indicate that biofilm coverage of the GAC particle was incomplete; this is also known as “patchy growth.” Moreover, biomass growth on the BP withdrawn from the hydrogenogenic bioreactor was lower than that in the BP from the methanogenic bioreactor. This was possibly due to the selective pressures applied to the inoculum in the hydrogenogenic bioreactor (such as the heat shock pretreatment and the acidic regime), which could have limited the microbial diversity.

Presumably, it seems that there was some deposition of small mineral scales on BP of the hydrogenogenic bioreactor. Magnification of the micrographs was not sufficient to appreciate in detail the morphology of microbes that constituted the biofilms, although in Fig. 1b, it seems that there are some diplococcus-like and rod-like shapes. Several methanogenic archaea that can be found in methanogenic reactors devoted to waste treatment are typically short and long bacilli, cocci in several association patterns (sarcina), as well as filamentous shapes (Corrales et al. 2015).

Structures that resembled isolated NPs in the amorphous matrix of the ENBP were observed in frames c to f of Fig. 1. Regarding this, the asymmetric and unequal morphology of biomagnetite nanoparticles has been discussed; previous reviews have found that biomagnetite nanoparticles have amorphous distribution and size, in contrast with the chemical synthesized magnetite (Guo and Barnard 2013).

Analysis by EDS of nanoparticles embedded in the agglomerated material confirmed the presence of iron and oxygen in the mapped region and a size of approximately 40 nm for NP in the matrix (Fig. 1c).

Our results were close to those reported by Kim et al. (2015), where anaerobic bacteria biosynthesized siderite, akaganeite, and schwertmannite particles from acid mine drainage that contained iron oxyhydroxide at room temperature. Glucose was added as a source of reducing power (10 mM).

Specific surface area

On the other hand, according to BET analysis (Table 2), the specific surface area of the NP of our four ENBPs (480 to 625 m2 g−1) was superior to that reported in previous works for NP synthesized by co-precipitation method (286.9 m2 g−1; Ma et al. 2014). Iron oxide nanoparticles with a high specific surface area have been widely studied because of its physicochemical properties and potential applications in catalysts and adsorbents, for instance, in remediation of organochlorinated pollutants.

Furthermore, in previous studies, it was observed that small size of nanoparticles and high surface area play an important role in environmental processes, for instance, in mine tailing remediation as well as groundwater treatment (Kim et al. 2015).

Elemental composition of the embedded nanobioparticles analyzed by energy dispersive spectroscopy

The composition of iron ENBPs synthesized from 0.25 M FeCl3 and hydrolysates was determined by EDS (Fig. 2); the values obtained for elemental composition in ENBP-H were O (35.36%) and Fe (64.64%); and for the ENBP-M, the composition was O (39.45%) and Fe (60.55%).

Regarding the synthesis of ENBP using ferric citrate 0.1 M as salt precursor, the results in composition analysis by EDS were ENBP-H, O (91.41%) and Fe (8.59%); in the case of ENBP-M, O (37.46) and Fe (62.54%) were found.

There was a strong interaction between both factors for the iron content response variable (p < 0.05) (in this scenario, it is recommended to disregard the main effects and concentrate in the interaction; Montgomery 2017) (Fig. S2): when the salt was FeCl3, the Fe content slightly decreased from 64.64 to 60.55% for the ENBP-H and ENBP-M, respectively. When the synthesis was based on Fe citrate, the Fe content dramatically increased from a very low value 8.59% to a maximum of 62.54% for the ENBP-H and ENBP-M, respectively. Overall, Fe contents of nanoparticles in ENBP-M were less sensitive to the effect of the “type of iron salt.” These results were somewhat unexpected. The statistical interaction outcome deserves more research and interpretation in terms of physical, chemical, and biological arguments. Also, we should be cautious because the iron content can be contributed not only by magnetite but also by other compounds such as siderite and di-iron trioxide (hematite).

Assessment of bionanoparticles by high-resolution electron microscopy

Table S2 in the Supplementary Material document summarizes the results of HRTEM as well as others for the four “treatments” in our work. From analysis of the bionanoparticles by HRTEM, it was observed a spherical shape and agglomeration, particularly for samples H and O synthesized from ferric chloride. Particle size was ~ 40 nm (Fig. 3a–c) which was consistent with findings by SEM mentioned before. The electron diffraction confirmed the crystallinity of the ENBP-H.

Regarding ENBP-M FeCl3, bionanoparticles exhibited a spherical shape and a finer crystalline structure and particle size ~ 42 nm (Fig. 4a, b). High-resolution lattice image showed a heterogeneous structure that could be attributed to different orientations of crystals (Fig.4c). The diffracted electron pattern of ENBP-M FeCl3 can be observed in Fig. 4d by SAED.

In case of ENBP-H citrate, the shape of crystals was hexagonal whereas size was ca. ~ 62 nm; there was a certain degree of agglomeration (Fig. 5a and b). The image of lattice in Fig. 5c showed a contrast in different zones in which dark areas were attributed to organic matter, mainly in the form of carbon and precursors such as citrate. Also, in Fig. 5c, the same phenomenon of different crystallographic orientation was observed, possibly due to the presence of mixture of iron phases (as suggested by diffraction patterns by XRD; Fig. 9). Regarding the SAED pattern (Fig.5d), we found an increase in the presence of bright spots in the rings of diffracted electron.

Concerning ENBP-M citrate, the particles presented lower agglomeration, hexagonal shape, and size ~ 40 nm (Fig. 6). The lattice showed a more uniform orientation, and more fine crystalline structure was observed in SAED (Fig. 6d).

In the present work, according to SEM and HRTEM micrographs, the nanomaterial prepared from FeCl3 exhibited characteristic clusters of agglomerated particles, size dimensions in a range from 40 nm, and spherical shape (Figs. 1c, 3, and 4).

Regarding our ENBP, the use of ferric citrate as iron precursor promoted greater crystallinity (according to SAED pattern) as well as changes in morphology since presenting hexagonal shape of particles (HRTEM) (Table S2; Figs. 5 and 6). The average size was ca. 40 and 60 nm for ENBP-M and ENBP-H citrate and presented different orientation shown in the lattice image by high resolution, particularly in ENBP-M citrate (Fig. 6).

Shape and size of EBNP synthesized with FeCl3 in our work were consistent with previous research of Kastrinaki et al. (2018). In their experiments, the authors used two iron salts, Fe (NO3)3 and FeCl3, in order to compare the effect of iron source on the morphology of the NP. According to their TEM results, spherical morphology was observed in both cases. Their NP presented cavities when synthesized from Fe (NO3)3 and in case of FeCl3 NP, an additional oxide layer was formed. The authors discussed that these variations could be related to differences between water solubility of Fe (NO3)3 and FeCl3, 150 mg mL−1 and 92 mg mL−1, respectively. Higher solubility values for iron nitrate solution would have released more soluble ions, unlike ferric chloride. Thus, the spherical shape of Fe (NO3)3 NP resulted in homogeneous morphology, whereas in the opposite case, lower solubility of ferric chloride would lead to an amorphous surface.

In a work of synthesis of iron nanoparticles by extracts of Mucuna pruriens (a seed from Indonesia) and FeCl3 (Sardjono et al. 2018), the main material formed was magnetite, whereas the NP shape was spherical and exhibited a medium size (30 nm average). The NP were present as agglomerates.

Our results on crystallinity of the EBNP (Table S2) were in agreement with findings by Liu et al. (2017). They evaluated the effect of type of iron salts on crystallinity of the FeCo NP. They found higher crystallinity and more uniform distribution (better dispersion) in NP produced from ammonium iron citrate that in those prepared from iron nitrate salt. Using XRD analysis, the FeCo NP synthesized with ammonium iron citrate showed more crystalline domains (311, average FWHM value of 0.9°) than the NP samples prepared from iron nitrate solution (average 2.0° for the same peak). This effect was explained by the authors in terms of the formation of a complex ion from citrate and iron atoms and free oxygen in the hydroxyl and carboxyl group. According to the authors, this complex ion would have allowed the precipitation and greater dispersion of the NP.

X-ray photoelectron spectroscopy

The XPS spectra are shown in Fig. 7. Spectra were obtained with a constant energy step size of 1.0 eV; they can be used for organic and inorganic materials (Pérez-González et al. 2018). It should be noted that in the four ENBPs, two characteristic binding energies for iron oxides species at 2p3/2 for high bonding energies and 2p1/2 for lower energies were found, see Fig. 7a, c, g and e (and Fig. 9 for XRD pattern). According to literature, the binding energy for the spectra for Fe 2p1/2 was in the range of 717–725 eV (Kim and Park 2002).

In the case of Fe 2p3/2 for ENBP-M (using both salts, FeCl3 and ferric citrate), the spectra presented more signals that could be possibly attributed to FeIII tetrahedral (712–714 eV) or FeIII octahedral (710–712 eV) (Fig. 7c and g).

According to our results in XRD analysis, the ENBPs presented a mixture of iron oxides phases. This was also suggested in the XPS spectra by the occurrence of multipeaks at several binding energies whose values were near to those of the characteristic iron oxides reported in literature such as Fe3O4 (magnetite), hematite (α-Fe2O3), and maghemite (γ-Fe2O3) (Poulin et al. 2010; Lesiak et al. 2019; Moulder et al. 1992).

It has been previously reported that in the Fe2p line spectrum from Fe3O4 (magnetite), the satellite is usually at 714 eV (Fujii et al. 1999). For α-Fe2O3 (hematite), the satellite peak appears approximately at 719 eV (Radu et al. 2017; Zhong et al. 2017). In our spectra, we found satellite peaks at 713.3 and 719.5 eV for Fe+2 and Fe+3, respectively. Other satellite values were also found for Fe+2 in our samples (Table S2). This suggested the presence of magetite and hematite. Also, these findings were consistent with results obtained with XRD and IR (Table S2).

From the XPS analysis in the four ENBPs, there was a tendency to obtain characteristic peaks of 711 eV corresponding for hematite (α-Fe2O3) high spin Fe 2p3/2 and, for low spin 2p1/2, the binding energy was 724–725 eV.

Interestingly, it was found a slight effect of the precursor salt citrate or chloride on the peak intensities: values corresponding to nanomaterials synthesized with iron citrate were 4.9 to 6.5% higher than those of nanomaterials synthesized with iron chloride. Yet, this trend should be taken with caution because XPS determines the composition of the surface of the sample (Moulder et al. 1992; Zhong et al. 2017); the signals are thought to be more a qualitative indication than quantitative and also the small size of samples (Pérez-González et al. 2018).

In summary, XPS analysis showed the presence of both Fe(II) and Fe(III) in all the nanomaterials synthesized in our work. Satellite peaks revealed the likely presence of magnetite and hematite. These results agree with results obtained in this work using XRD and IR (Table S2; Figs. 9 and 10).

Size distribution of embedded nanobioparticles by dynamic light scattering

The size distribution in the ENBPs was also determined by DLS. It was observed that using 0.1 M ferric citrate as a salt precursor lead to smaller diameter of agglomerates than that obtained when using 0.25 M FeCl3. Figure 8 shows that the average dispersion for the agglomerated ENBP were in the range from 600 to 1000 nm (frames a and b), for ENBP-H and ENBP-M synthesized from 0.25 M FeCl3, whereas the average range of dispersion was much lower, in the range 400–700 nm (frames c and d) for ENBP-H and ENBP-M from ferric citrate. Thus, results from DLS confirmed the agglomeration found in some SEM and HRTEM micrographs of ENBP described before.

Structural pattern by X-ray diffraction studies

Table 3 shows the results obtained from the XRD patterns in the four different ENBPs. We obtained broad, relatively low relative intensity (RI) peaks (Fig. 9). The highest RI in were diffracted from the samples of ENBP synthesized with 0.1 M ferric citrate. Saranya et al. (2015) in a work about magnetite NP (Fe3O4) produced by hydrothermal method also reported very broad peaks that indicate a small crystallite size.

The diffraction patterns in the XRD analysis of ENBP showed a mixture of phases of iron oxides (Fig. 9; Table 3): magnetite (Fe3O4) (Anyika et al. 2017), siderite (FeCO3) (Oza and Joshi 2017), and hematite α-Fe2O3 (Jaafarzadeh et al. 2017).

The following positions and planes were found in ENBP-H FeCl3: magnetite Fe3O4 35° (3 1 1), 43° (4 0 0), and 62° (4 0 0), for siderite FeCO3 26° (0 1 2) and 38° (1 1 0), whereas for hematite α-Fe2O3 24° (0 1 2), 36° (1 1 0), and 50° (0 2 4).

On the other hand, in ENBP-M FeCl3, the pattern diffraction showed the following phases and positions: magnetite Fe3O4 35° (3 1 1), 43° (4 0 0), and 62° (4 0 0); for siderite FeCO3 26° (0 1 2) and 38° (1 1 0); moreover, hematite was found at 24° (0 1 2), 36° (1 1 0), 41° (1 1 3), 50° (0 2 4), and 54° (1 1 6).

XRD analysis of ENBP-H citrate revealed magnetite Fe3O4 35° (3 1 1) and 62° (4 0 0) and siderite FeCO3 26° (0 1 2), 38° (1 1 0), and 42° (1 1 3). Regarding the hematite, the positions found were at 24° (0 1 2), 36° (1 1 0), 50° (0 2 4), and 54° (4 0 0).

In the samples of ENBP-M citrate, the angles and positions were the following: magnetite Fe3O4 35° (3 1 1), 43° (4 0 0), and 62° (4 0 0) and for siderite FeCO3 26° (0 1 2), 38° (1 1 0), and 42° (1 1 3). For hematite α-Fe2O3, the positions were at 24° (0 1 2), 36° (1 1 0), 50° (0 2 4), and 54° (4 0 0).]

Analysis by infrared spectroscopy

Infrared spectra allowed to identify vibrational modes for magnetite (Fe3O4) and siderite in three of the ENBPs (Fig. 10a, b, and d). Regarding ENBP-H citrate (Fig. 10c), the peak for magnetite was not found; however, it was observed a signal for siderite (780 cm−1).

Concerning the potential application of the ENBP of this work, there is a wide research about dehalogenation of organochlorinated compounds by magnetite nanoparticles, as well as positive evidence concerning the involvement of siderite and hematite, which will be discussed below. In general, the IR results were consistent with XRD and XPS analysis (Fig. 9; Tables S3, S4, S5, and S6), which reinforces the evidence found in this work.

In samples of ENBP-H FeCl3, the typical vibrational modes for magnetite were observed at 565, 667, and 684 cm−1 (Fig. 10a; Table S3) (Ramírez-Nuñez et al. 2018; Saranya et al. 2015; Yadav 2018). Peaks attributed to siderite (FeCO3) at 776 cm−1 (Santillán and Williams 2004) and hematite (α-Fe2O3) were detected at 465 and 477 cm−1 (Fig. 10a; Table S3) (Ristic et al. 2017; Li et al. 2012).

Magnetite Fe3O4 is recognized to be a conducting mineral that fosters the electron transfer in microbial species and is involved in dechlorination of toxic organochlorinated compounds. In a recent work, Leitao et al. (2018) evaluated the effect of the addition of magnetite nanoparticles (Fe3O4) when studying the dehalogenation of 1,2-dichloroethane (1,2-DCA). They reported that dechlorination rate increased 3.3 times when Fe3O4 nanoparticles were added to the dehalogenating mix culture, mostly composed by Dehalococcoides mccartyi. The final removal rate of 1,2-DCA in the treatment containing magnetite was 2.3 μeq L−1 day−1, whereas in the experiment with no addition of nanoparticles, it was only 0.7 μeq L−1day−1.

It is interesting to remark that siderite FeCO3 is a mineral generated from metal iron by oxidation in the presence of HCO3−, and it is naturally found in groundwater and saline groundwater (Xin et al. 2018), although it can be formed in nanosynthesis processes. Moreover, there is some evidence that siderite could be related to dehalogenation of selected organochlorinated compounds, according to Wu et al. (2014). They examined the remediation of saline waters contaminated with 1,1,1-TCA (1,1,1-trichloroethane) using zero valent iron nanoparticles (nZVI). They found that degradation efficiencies reached a maximum at higher concentrations of bicarbonate (100 mM) in the first 6 h.

Xin et al. (2018) studied the removal of trichloroethylene (TCE) by varying the composition of mineral salts and TCE concentration in the feedwater. Precipitated minerals produced in the experiments containing nZVI and HCO3− were analyzed by XRD, and the diffraction pattern for siderite (FeCO3) was identified. In that case, the best efficiency for TCE removal was 99.6%.

There is also some evidence on the dehalogenating potential of hematite α-Fe2O3 on 2,4-dichlorophenoxy-acetic acid (2,4-D) (Jaafarzadeh et al. 2017). Nearly 30% removal of 2,4-D was obtained by using hematite NPs with the help of a redox mediator peroxymonosulfate. It seems that hematite NPs activated the peroxymonosulfate and the latter effected the dechlorination of 2,4-D. These results should be interpreted with caution because the experiment suggests that hematite NPs acted indirectly on the dechlorination of 2,4-D.

Choi et al. (2014) described the use of bimetallic NP Ni-Pd + hematite in experiments that allowed significant removal of perchloroethylene (PCE) (75% in 6 h) in batch tests initially loaded at 300 mM of PCE. On the other, the control without NPs exhibited a 5% removal efficiency.

In our work, the IR analysis of ENBP-M FeCl3 showed the vibrational mode signal for magnetite at 563 and 667 cm−1 and FeCO3 at 760, 825, and 1435 cm−1 (Fig. 10b). Additionally, a typical band for hematite α-Fe2O3 at 460 cm−1 was observed.

Regarding NP in ENBP-H citrate, weak bands for siderite (780 cm−1) and for hematite (466 cm−1) (Fig. 10c) were found.

In the case of ENBP-M citrate, the signals for magnetite were at 532, 559, and 629 cm−1. Also, the vibrational modes for siderite were found at 1110 and 1420 cm−1; moreover, the corresponding peaks for hematite were detected at 459 and 473 cm−1 (Fig. 10d).

Concerning changes in media pH after biosynthesis of NPs, a tendency to the acidification of the spent liquors was observed. In short, pH values in the “treatments” are mentioned below (where subscript i stands for initial, subscript f means final): ENBP-H FeCl3 0.25 M (pHi 5.55, pHf 4.61), ENBP-M FeCl3 0.25 M (pHi 8.28, pHf 6.89), ENBP-H Ferric citrate 0.1 M (pHi 5.74, pHf 4.87) and ENBP-M Ferric citrate 0.1 M (pHi 8.22, pHf 6.56).

Differences between initial pH of H and M synthesis are related to the pH required for dark fermentation H that has been reported to be between 4.5 and 5.8, whereas the typical pH of methanogenic bioreactors is from neutral to slightly alkaline (Estrada-Vázquez et al. 2003; Garibay-Orijel et al. 2005; Escamilla-Alvarado et al. 2013; Muñoz-Páez et al. 2013; Sotelo-Navarro et al. 2017).

Differences between pHf of H and M treatments (pHf of M > pHf of H) can be explained in terms of two issues: M treatments not only started at higher pHi than H ones, but also the buffering was higher (due to presence of bicarbonates and carbonates a pH 8.2; Ripley et al. 1986; Poggi-Varaldo and Oleszkiewicz 1992) than the buffer in the H vials. Therefore, final pHs in liquors of M biosynthesis are expected to be higher. Trends in final pH related to the type of precursor salt used were not so clear.

In our experiment, the association of vial pH and size of the ENBP (Table S2) corresponded strictly to a statistical interaction. Lowest sizes of NPs corresponded in general to the treatments EBNP-M (average size ca. 40 nm, independently of the type of precursor salt). Both treatments exhibited pHf around 6.6–6.8. Yet, size of the EBNP-H at low pH (ca. 4.6–4.9 pHf) had a divergent pattern depending upon the precursor salt: the one with FeCl3 also displayed an average size of ca. 40 nm, whereas the one with iron citrate exhibited an average size near to 60 nm.

Previous works about microbial nanosynthesis of iron NPs such as magnetite (Fe3O4) have presented evidence regarding influence of pH on controlling the size of nanoparticles. For instance, when pH increased from neutral values to slightly alkaline values, NP size decreased. Kim and Roh (2018) examined a culture of indigenous Iron reducing bacteria that contained Clostridium strains and akaganeite as a magnetite precursor. They observed a quick shift in color of media and formation of the magnetic phase at pH of 8.5 and reduction of the size of nanoparticles in alkaline conditions (5–10 nm), whereas for pH 7.5 and 6.5, the values were slightly larger (5–25 nm and 15 nm, respectively). Regarding influence of pH on the shape of nanocrystals, the authors could not find significant differences since in all ranges of pH, the nanoparticles were principally spherical.

The effects of pH, FeCl3, and temperature on the production of iron oxide nanoparticles synthesized by Penicillium waksmanii were analyzed by Honary et al. (2015). In that study, they found an inverse relationship between the pH and the dimensions of nanoparticles, since the size decreased from 131 to 77 nm when increasing values of the pH (7.0) in the medium.

Therefore, our results regarding the influence of pH on NP size generally agreed with trends presented in the open literature.

The COD (chemical oxygen demand) values were determined at the initial and final time of the bionanoparticle synthesis. The results were the following: ENBP-H FeCl3 0.25 M (CODi 17,920 mg O2 L−1; CODf 7424 mg O2 L−1), ENBP-M FeCl3 (CODi 10, 880 mg O2 L−1; CODf 3328 mg O2 L−1), ENBP-H citrate (CODi 17,600 mg O2 L−1; CODf 10,880 mg O2 L−1), and ENBP-M citrate (CODi 9984 mg O2 L−1; CODf 5376 mg O2 L−1). Consequently, the COD removal efficiencies (ƞCOD) were (in descending order) as follows:

$$ {\displaystyle \begin{array}{l}69.44\%\left(\mathrm{ENBP}\mathrm{M}\hbox{-} {\mathrm{FeCl}}_3\right)>51.46\%\left(\mathrm{ENBP}\hbox{-} \mathrm{M}\ \mathrm{Citrate}\right)>\\ {}>41.58\%\left(\mathrm{ENBP}\hbox{-} \mathrm{H}\ {\mathrm{FeCl}}_3\right)>34.62\%\left(\mathrm{ENBP}\hbox{-} \mathrm{H}\ \mathrm{Citrate}\right)\end{array}} $$
(4)

Standard deviations of ƞCOD were ≤ 5%.

First, the effect type of salt precursor indicated a main effect on ƞCOD of 55.51 and 43.04% for FeCl3 and iron citrate, respectively (pooling the sources of the ENBPs).

On the other hand, the effect of source of ENPBs on ƞCOD resulted in 60.45 and 38.10% for M and H, respectively (pooling the types of precursor salt).

From these trends, it could be inferred that there could be a higher bioreduction power in methanogenic environments used for the synthesis of ENBP-M (either FeCl3 or ferric citrate precursor salt) due to more efficient consumption of organic matter and its transformation to reduced methane, as compared to dark fermentation processes (H). Thus, it could be hypothesized that this effect could influence selected properties of the ENBPs. For instance, our results indicated that there was a higher degree of crystallinity and dispersion in ENBP-M than that in ENBP-H (Figs. 4, 6, and 8; Table S2 in the Supplementary Material document). According to the XRD patterns, XPS analyses, and HRTEM results, the ENBP-M presented higher number of iron oxides phases, and the signals in their spectra were stronger (Figs. 9 and 10). The latter was in agreement with data obtained by EDS analysis, where the concentration of iron in two types of ENBP-M was consistently high (60–65%), whereas in ENBP-H, there was a variation in iron concentration that included low values (8.5–62%) (Fig. 2; Table S2).

Also, it is likely that the ƞCOD could have also been associated to the smaller size of the ENBP-M, in addition to the possible influence of pH that was discussed above. Whether one or the other (ƞCOD or pH) could have contributed more to the size of the ENBP cannot be determined with the current experimental information in our work.

An analysis that takes into account the ƞCOD in terms of precursor salt shows that biosynthesis performed with FeCl3 exhibited higher ƞCOD than that with Fe citrate as it was mentioned above (Table S2). It seems that the type of precursor salt with the associated effect on ƞCOD might have influenced (or be associated with) the NP shape, crystallinity, and dispersion (Table S2). For instance, ENBPs synthesized with Fe citrate showed hexagonal shape (Figs. 5 and 6) (whereas the ones synthesized with FeCl3 were spherical; Figs. 3 and 4), exhibited a higher degree of crystallinity, and were more disperse than those synthesized with FeCl3.

So, the important issues are that we can associate a probable positive effect of ƞCOD on selected characteristics of ENBPs. This, in turn, could be related to the biological process that dominated the biosynthesis (H or M). On the other hand, it seems that the type of precursor salt used in the biosynthesis of ENBPs might have been related to another set of ENBPs features, such as crystallinity, shape, and agglomeration.

So far, in our research, we were able to implement a biological synthesis procedure for producing ENBPs using biocatalysts from bioreactors of a waste-based biorefinery as well as a hydrolysate of OFMSW. Characteristics of these nanomaterials seemed to be very attractive for possible application on remediation of wastewaters, groundwater, and soil, due to their high specific surface area, the presence of iron nanomaterials such as magnetite and siderite that have been reported to be successful for degradation of organochlorinated compounds, and the bioreductive power likely associated to the significant organic matter removal ηCOD during the biosynthesis stage.

Conclusion

Bioparticles from biological fluidized bed reactors from a biorefinery of OFMSW (that produces hydrolysates rich in reducing sugars) were nanodecorated by biological reduction of iron salts (ENBP). Factor origin of bioparticles (either from hydrogenogenic or methanogenic fluidized bed reactor) and “type of iron precursor salt” (iron chloride or iron citrate) were explored. X-ray diffraction patterns of ENBP showed the typical peaks for magnetite at 35° (3 1 1), 43° (4 0 0), and 62° (4 0 0); moreover, siderite diffraction pattern was found at 26° (0 1 2), 38° (1 1 0), and 42° (1 1 3). SEM and HRTEM showed amorphous distribution of NP on the bioparticles surface, although small structures that are nanoparticle-like could be seen in the SEM micrographs. Some agglomeration of NPs was confirmed by DLS. Average NP size was lower in general for NP in ENBP-M than ENBP-H according to HRTEM. The factors did not have a significant influence on the specific surface of NPs, which was high and in the range of 490 to 650 m2 g−1.

Analysis by EDS displayed consistent iron concentration 60–65% iron in nanoparticles present in ENBP-M (bioparticles previously grown in methanogenic bioreactor), whereas the iron concentration in NPs present in ENBP-H (bioparticles previously grown in methanogenic bioreactor) was more variable in a range from 8.5 to 62%, depending on the iron salt. These results are known as statistical interaction, so no main effects of the factors could be drawn.

Results of infrared analysis of ENBP in our work were congruent with the presence of magnetite and occasionally siderite determined by XRD analysis as well as the presence of both Fe+2 and F+3 (and selected satellite peaks) observed by XPS.

Our results on the ENBPs seemed to be attractive, since iron NPs are known to be useful materials for wastewater remediation of several contaminants, such as the removal of chlorinated organic compounds. Finally, the biosynthesis of ENBP from waste hydrolysates could become a sustainable platform for such nanomaterials as well as an additional source of revenues for a waste-based biorefinery.