WO2014125217A2

WO2014125217A2 - Method of biologically-assisted mineral denitrification in a liquid medium

Info

Publication number: WO2014125217A2
Application number: PCT/FR2014/050286
Authority: WO
Inventors: Delphine GUERBOIS; Guillaume Morin; Georges ONA-NGUEMA
Original assignee: Universite Pierre Et Marie Curie (Paris 6); Centre National De La Recherche Scientifique
Priority date: 2013-02-13
Filing date: 2014-02-13
Publication date: 2014-08-21
Also published as: EP2956414A2; FR3001962B1; FR3001962A1; WO2014125217A3

Abstract

The present invention concerns a method for processing nitrogenous materials, and in particular a continuous method of biologically-assisted mineral denitrification in a liquid medium, without producing ammonium, implementing a biogenic carbonated green rust, and the use of such a green rust for reducing nitrite ions into nitrogenous gas.

Description

BIOLOGICALLY ASSISTED MINERAL DENITRIFICATION PROCESS IN A LIQUID ENVIRONMENT

The present invention relates to a process for the treatment of nitrogenous materials, and in particular to a continuous process of biologically assisted inorganic denitrification in a liquid medium, without the production of ammonium, using a biogenic carbonated green rust and the use of such a green rust for the reduction of nitrite ions into nitrogenous gases.

Nitrate ions exist in natural waters in the absence of pollution. Their concentration, then, does not exceed 3 to 5 mg / 1 in the surface waters of rivers, canals, lakes, ponds and some mg / 1 in groundwater. However, the levels of nitrate ions in these waters have been steadily increasing in recent years and reflect a pollution of the resource linked to human activities (industrial and domestic discharges from agriculture and livestock, effluents from livestock, food surface waters by polluted groundwater, "leakage" from industrial and inhabited areas) and in particular surface activities (dense urban and industrial occupancy and / or intensive agriculture, traditional medium intensity agriculture, dispersed rural housing). The concentrations of nitrate ions in the aquifers are generally all the higher as they are sensitive, poorly protected, influenced by urbanization, industrial sites and intensive agriculture or livestock farming. These concentrations also vary with the speed of circulation-renewal of water in the aquifers. The maximum permitted limit is 50 mg / l of nitrate ions in drinking water supplies. However, the upper limit of a "good ecological status" of water was set at 25 mg / 1 in the Water Framework Directive (WFD) of 23 October 2000.

Nitrate ions are not toxic in themselves. It is their transformation into nitrite ions and nitroso compounds (nitrosamines and nitrosamides) which can cause characteristic disorders such as methemoglobinemia or cyanosis of the infant, reproductive disorders, endocrine or vitamin effects in animals and finally the occurrence of digestive cancers in humans. In addition, the high levels of nitrate ions are at the origin of the proliferation of macroscopic green algae throughout the Brittany region.

The approved techniques used today to clean up nitrate ions consist of a process for the treatment of biologically or physicochemically polluted waters using ion exchange resins. These techniques are very effective in the short term, but expensive in investments and in operation. In particular, purely biological denitrification (ie bacterial denitrification), defined as the reduction of nitrate ions to gaseous nitrogen by means of denitrifying bacteria, has been massively studied in the context of wastewater treatment.

Existing purely biological denitrification processes do not involve pure bacterial strains but a bacterial consortium (that is to say a mixture of different bacterial strains) which develops according to the physicochemical parameters of the water to be treated. The main disadvantage of such a process is the lack of control of the reduction products formed, namely a majority of NH ₄ ⁺ ammonium (Cruz-Garcia et al., Journal of Bacteriology, 2007, 189 (2), 656- 662). In view of the water quality criteria set out above, ammonium production is not acceptable in the development of a large-scale denitrification process. In addition, when high concentrations of nitrate ions are present, an accumulation of toxic nitrite ions is observed during the implementation of such a method.

More recently, research has been carried out on processes for the oxidation of nitrate ions by Fe (II) or Fe (O). In particular, these processes can involve, as a source of Fe (II), green rust synthesized chemically. Green rust is a Fe (II) -Fe (III) structure with flake structure having a low oxy-reduction potential giving them a strong reducing power. Synthetic green rust, such as GR _S04 , GRci and GRco ₃ , generally have a ratio of Fe (II) / Fe (III) molar concentrations ranging from 2 to 3 depending on the pennant used and it has been shown that the reduction of nitrate ions in The presence of such synthetic green rust leads to the exclusive formation of ammonium and magnetite (Hansen et al., Environ Science Sci., 1996, 30, 2053-2056 and Applied Clay Science, 2001, 18, 81-91).

In addition, lower molar concentration Fe (II) / Fe (III) ratios can be obtained when green rust is prepared biogeneically from lepidocrocite or green ferric rust and ferri-reducing type bacteria. Shewanella putrefaciens. The ratio of Fe (II) / Fe (III) molar concentrations obtained is then of the order of 1 (Ona-Nguema et al., Environ Sci., Technol., 2002, 36, 16-20). To date, however, no process of the prior art has made it possible to isolate and characterize biogenic green rust with a Fe (II) / Fe (III) molar concentration ratio that is strictly less than 1, regardless of the flag used. Recently, it has been proposed in the international application WO 2008/110689 to use a carbonated Fe (II) -Fe (III) hydroxide, also called ferrous-ferric hydroxycarbonate or ferrous-ferric oxyhydroxycarbonate or carbonated green rust, corresponding to Fe formula ₆ ⁿ (i _-X) ^m _6x Oi2H2 Fe _(7- 3 _X) CO _3, for the implementation of an oxidation-reduction process of the nitrate ions in the presence of ferric-reducing bacteria. In this formula, x is defined as the ratio of Fe (III) / [Fe (III) + Fe (II)] molar concentrations. The ferrous-ferric hydroxycarbonate has a ratio of Fe (II) / Fe (III) molar concentrations ranging from 0.5 to 2 (corresponding to a value of x varying from 0.33 to 0.66) and oxide by reducing nitrate ions to gaseous nitrogen. The oxidized phase obtained, having a ratio of Fe (II) / Fe (III) molar concentrations of less than 0.5 (x> 0.66), is in turn reduced in the presence of ferri-reducing bacteria to regenerate the hydroxycarbonate Ferrous-ferric starting material having a ratio of Fe (II) / Fe (III) molar concentrations ranging from 0.5 to 2 without substantial structural changes. Examples 1 and 2 describe, on the one hand, the preparation of a biogenic carbonaceous green rust (the ratio of Fe (II) / Fe (III) molar concentrations ranging from 0.5 to 2) by bacterial reduction of iron ( III) to iron (II) in a ferric oxyhydroxycarbonate, and secondly, the catalytic reduction of nitrate ions to gaseous nitrogen in the presence of this carbonated green rust. However, comparative tests have been carried out by the inventors of the present application and demonstrate that biogenic green carbonate rust obtained by bioreduction of ferric green rust by Shewanella putrefaciens (molar concentration ratio Fe (II) / Fe (III) = 1, corresponding to x = 0.5) or by bioreduction of Lepidocrocite by Shewanella putrefaciens (molar concentration ratio

Fe (II) / Fe (III) = 1.25, corresponding to x = 0.44), are incapable of reducing nitrate ions regardless of the pH of the reaction medium or after at least 5 months of interaction (see comparative example 1 illustrating the present application). Furthermore, Example 3 of the international application WO 2008/110689 describes the catalytic reduction of nitrate ions to gaseous nitrogen in the presence of a synthetic carbonated green rust of Fe (II) / Fe (III) = 2 molar concentration ratio. , corresponding to x = 0.33. These results are however in perfect contradiction with the results described by Hansen [December 10, 2004, French Academy of Agriculture, "Green rusts and reduction of pollutants" Abstract In Colloquium of the Academy of Sciences "Biogeochemistry of the iron-rust cycle green and fougerite]. Thus, contrary to what is described in the international application WO 2008/110689, a biogenic carbonate green rust with a Fe (II) / Fe (III) molar concentration ratio of 1 or 1.25 is not able to reduce nitrate ions in nitrogen gas or very slowly. And when a synthetic carbonated green rust of molar Fe (II) / Fe (III) molar ratio greater than or equal to 2 is used, ammonium is produced.

Finally, Hansen et al. (Geochimica and Cosmochimica Acta, 1994, 58, 12, 2599-2608) have described the reduction of nitrite ions in the presence of a synthetic green rust GR _S04 with a molar Fe (II) / Fe (III) concentration ratio of 2 and indicate that the reduction may lead to several nitrogen compounds, especially ammonium.

Tests carried out by the inventors of the present application confirm, moreover, on the one hand that a GRQ synthetic green rust and on the other hand, that a synthetic green rust GRc ₀₃ , both being obtained chemically, ( molar concentration Fe (II) / Fe (III) ratios of 3, corresponding to x = 0.25 and of 2, corresponding to x = 0.33) reduce the nitrite ions to ammonium (see Comparative Examples 2 and 3 illustrating this application).

Therefore, all known abiotic denitrification reactions involving Fe (II) or Fe (0) have shown the production of ammonium as a nitrate / nitrite ion reduction product or very slow kinetics to obtain nitrogen gas.

Simple, environmentally friendly and less costly remedial solutions must therefore be proposed to the populations concerned in order to drastically reduce the nitrogen concentrations in the water by avoiding the production of ammonium.

Thus, the inventors have set themselves the goal of providing a process for the treatment of nitrogenous materials, in particular for the treatment of nitrate / nitrite ions present in a liquid medium, which is simple and inexpensive to implement, usable at the scale industrial, for example in sewage treatment plants, without the production of ammonium.

The subject of the present invention is therefore a biologically assisted inorganic denitrification method in a liquid medium, said process being characterized in that it comprises the following steps: i) a step of preparing a carbonated green rust by bioreduction of an Fe (III) oxyhydroxide under anaerobic conditions in the presence of a culture of at least one bacterium having a ferri-reducing activity until obtaining a carbonate green rust with a ratio of Fe (II) / Fe (III) molar concentrations ranging from 1 to 1.5;

ii) a step of reducing the nitrite ions present in a liquid medium by contacting said carbonated green rust with a ratio of Fe (II) / Fe (III) molar concentrations ranging from 1 to 1.5 and a medium liquid comprising nitrite ions, said step leading to the formation of one or more nitrogenous gases, and to the production of Fe (III) associated with a solid phase comprising one or more oxyhydroxides of Fe (III) and / or one or more carbonated Fe (II) -Fe (III) hydroxides;

iii) a step of bioreduction of said Fe (III) produced in step ii), in the presence of a culture of at least one bacterium having a ferri-reducing activity to obtain a carbonated green rust with a molar concentration ratio Fe (II) / Fe (III) ranging from 1 to 1.5.

Indeed, it has surprisingly been found that the nitrite ions could be reduced and converted into a gaseous mixture comprising N ₂ O and N ₂ and without generating ammonium, in the presence of a biogenic carbonate green rust having a ratio Fe (II) / Fe (III) molar concentrations typically vary from 1 to 1.5. Outside the value range [1-1,5] of the Fe (II) / Fe (III) molar concentration ratio, either the reduction of nitrite ions can lead to ammonium or the reduction kinetics is very low. slow and therefore unsuited to the use of this process on an industrial scale.

Thus, said step ii) leads to the formation of one or more nitrogen gases, without production of ammonium, and to the production of Fe (III) associated with a solid phase comprising one or more Fe oxyhydroxides (III). and / or one or more carbonated Fe (II) -Fe (III) hydroxides.

Carbonated Fe (II) -Fe (III) hydroxides means biogenic carbonate green rust obtained in step i) and which has not reacted in the following step ii), especially when this step is introduced in excess relative to the nitrite ions in said step ii).

According to a preferred embodiment, the Fe (III) oxyhydroxide used in the step i) of preparing a carbonated green rust having a ratio Fe (II) / Fe (III) molar concentrations ranging from 1 to 1.5, is chosen from lepidocrocite (γ-FeOOH), ferric green rust (Fe ^m ₆ O ₂ H ₈ CO ₃ ), ferrihydrite ( 5FeOOH.nH ₂ O), in order to favor a complete bioreduction.

The steps of preparation and regeneration of a carbonate green rust with a ratio of Fe (II) / Fe (III) molar concentrations ranging from 1 to 1.5 according to steps i) and iii) can be independently carried out in the presence a pure culture of at least one bacterium with a ferri-reductive activity.

The bacteria having a ferri-reducing activity that can be used during steps i) and iii) of the process according to the invention may be independently selected from the bacterial species of aquatic environments belonging to the genera Shewanella, Geobacter, Pseudomonas, Desulfovibrio, Geothrix and Pelobacter. . It is preferred to use the bacteria of the genus Shewanella and even more preferably the bacteria chosen from S. putrefaciens CIP 59.28, S. putrefaciens CIP 80.40 and S. oneidensis MR1.

The bioreduction time during step i) is preferably at least about 20 hours.

Moreover, the bioreduction time during step iii) preferably varies from about 5 hours to about 5 days depending on the oxidation state of the system.

The concentration of ferri-reducing bacterium during steps i) and iii) is preferably greater than 10 CFU / ml. Then, the concentration of Fe (III) during steps i) and iii) is preferably less than 400 mM.

The contact time during step ii) is preferably less than 5 days, so as to avoid the formation of undesirable side products, such as for example magnetite.

In an even more preferred embodiment, the contact time between the nitrite ions and the carbonated green rust during step ii) preferably varies from 1 hour to 48 hours approximately.

The amount of green rust used in step ii) is advantageously such that the ratio of molar concentrations [Fe (II) ions present in the green carbonate rust] / [nitrite ions] is at least 3, in order to to favor a complete reduction of nitrite ions as well as the production of nitrogen gas as final product. According to a preferred embodiment of the method according to the invention, steps ii) and iii) can be carried out simultaneously in one and the same step. According to an even more preferred embodiment, the process is carried out continuously and said steps ii) and iii) can then be repeated until the total possible depletion of the nitrite ions in the starting liquid medium.

According to the process according to the invention, the reduction of nitrite ions is rapid, of the order of a few hours, and reproducible with the assurance of never leading to the production of ammonium. In addition, the oxidation of biogenic carbonated green rust by nitrite ions leads to the production of one or more Fe (III) oxyhydroxides and / or one or more carbonated Fe (II) -Fe (III) hydroxides. , which in turn are new reducible by ferri-reductive bacteria.

The process according to the invention operates from nitrite ions present in a liquid medium. Consequently, the nitrogenous materials present in said liquid medium, such as, for example, nitrate ions, must first be reduced to nitrite ions.

Thus, according to a particular embodiment of the invention, the method further comprises a step ii ₀ ) which is carried out prior to step ii) and during which the nitrate ions present in the liquid medium are reduced nitrite ions in the presence of a culture of at least one bacterium capable of reducing nitrate ions to nitrite ions.

When the bacterium capable of reducing nitrate ions to nitrite ions only leads to the reduction of nitrate ions to nitrite ions, step ii ₀ ) leads to nitrite ions only, without producing ammonium.

When the bacterium capable of reducing nitrate ions to nitrite ions causes the reduction of nitrate ions to gaseous nitrogen and / or gaseous nitrogen oxide (ie bacterial denitrification), step ii ₀ ) leads to one or more nitrogen gases, without ammonium production.

When the bacterium capable of reducing the nitrate ions to nitrite ions causes at least a portion dissimilatory nitrate reduction to ammonium ions, said step ii ₀₎ can lead to the production of ammonium.

This step ii ₀ ) of prior reduction of nitrate ions to nitrite ions can preferably be carried out using bacteria capable of reducing nitrate ions to nitrite ions chosen from the genera Alcaligenes, Paracoccus, Thiobacillus, Vibrio, Desulfovibrio, Shewanella, Micrococcus, Geobacter, Pseudomonas and Bacillus. However, it is preferable to use the bacteria of the genus Shewanella and even more preferably the bacteria selected from S. putrefaciens CIP 59.28, S. putrefaciens CIP 80.40 and S. oneidensis MR1.

In a particular embodiment, step ii ₀ ) of prior reduction of nitrate ions to nitrite ions can be carried out in the presence of a culture of at least one strictly denitrifying bacterium, that is to say capable of reducing nitrate ions into nitrite ions without ammonium production. In other words, step ii ₀ ) leads to one or more nitrogen gases without producing ammonium.

Such a strictly denitrifying bacterium may be advantageously Paracoccus denitrificans.

Moreover, the reduction time during step ii ₀ ) preferably varies from approximately 2 hours to 24 hours.

The concentration of bacteria capable of reducing the nitrate ions to nitrite ions during stage ii ₀ ) of the process according to the invention is preferably at least 10 'CFU / ml in order to favor a reaction time included in the range. of time mentioned above.

When the liquid medium to be treated comprises nitrate ions, the amount of green rust used during stage ii) is advantageously such that the ratio of the molar concentrations [Fe (II) ions present in green carbonate rust] / [ions nitrates] is at least 3, and this to favor a complete reduction as well as the production of nitrogen gas as final product.

According to a preferred embodiment of the process according to the invention, the steps ii ₀ ) and ii) can be carried out simultaneously in one and the same step, that is to say that the nitrite ions generated in situ during step ii ₀ ) are simultaneously reduced according to step ii). As a result, carbonated green rust with a ratio of Fe (II) / Fe (III) molar concentrations ranging from 1 to 1.5 obtained according to step i) is directly brought into contact with a liquid medium comprising nitrate ions. and culturing at least one bacterium capable of reducing nitrate ions to nitrite ions.

According to this embodiment, the nitrite ions formed in situ in the liquid medium by contacting in step ii) with a green carbonate rust having a Fe (II) / Fe (III) molar concentration ratio varying from 1 at 1.5, lead to one or more nitrogen gases without ammonium production, in particular when step ii ₀ ) uses one or more bacteria which allow bacterial denitrification and / or that they catalyze only the transformation of nitrate ions into nitrite ions.

Indeed, the inventors of the present application have surprisingly discovered that when steps ii ₀ ) and ii) are carried out simultaneously in one and the same step and that they use indigenous bacteria of a water when purified, ammonium production is completely avoided.

When step ii ₀ ) employs one or more bacteria resulting at least in part in a concealing reduction of ammonium nitrate ions, said step ii ₀ ) can lead to the production of ammonium. However, the presence of green carbonate rust with a ratio of Fe (II) / Fe (III) molar concentrations varying from 1 to 1.5 in the liquid medium makes it possible to considerably reduce its production. Step ii ₀ ) is then a concurrent step of step ii).

In a particular embodiment, identical bacteria may be used in steps ii ₀ ) and iii) and the method according to the invention. They are selected from the genera Shewanella, Geobacter. In this case, the steps ii ₀ ), ii) and iii) can then be advantageously performed simultaneously during a single step. According to an even more preferred embodiment, the process is carried out continuously and said steps ii ₀ ), ii) and iii) can then be repeated until any nitrate ions and nitrite ions in the liquid medium have been completely exhausted.

According to a preferred embodiment, it is preferable, however, to obtain a better yield, to decouple the step iii) of regeneration of the biogenic carbonated green rust, step ii ₀ ) of reduction of the nitrate and nitrite ions. This makes it possible to prevent the presence of the nitrate / nitrite pair partially inhibiting the bioreduction step by the ferri-reducing bacteria.

In order to better understand the invention, the appended FIG. 1 lists the various steps that can be implemented when nitrate ions and nitrite ions are present in the liquid medium to be treated. Step ii ( ₀ ) represents bacterial denitrification and / or dissimilar reduction of ammonium nitrate ions. Since nitrite ions are systematic intermediates of microbial metabolisms of denitrification and dissimilar reduction of ammonium nitrate ions, step ii ₀ ) can be divided into two sub-steps, a first sub-step ii ₀ a) of ion transformation. nitrates to nitrite ions and a second sub-step ii ₀ b) of transformation of the nitrite ions into reduction products N ₂ O, N ₂ and NH ₄ ⁺ . Step i) represents the initial production of a carbonate green rust with a Fe (II) / Fe (III) molar ratio ratio varying from 1 to 1.5 of the process according to the invention, by bioreduction of an oxyhydroxide of Fe (III). Step ii) represents the abiotic denitrification by a carbonate green rust of Fe (II) / Fe (III) molar concentration ratio varying from 1 to 1.5 of the process according to the invention. Step iii) represents the dissimilatory reduction step of the Fe (III) of the process according to the invention allowing the regeneration of the green carbonate rust with a Fe (II) / Fe (III) molar concentration ratio ranging from 1 to 1.5.

The inventors of the present application have discovered that the rate of abiotic reduction of nitrite ions by biogenic carbon green rust according to step ii) could compete with the bacterial reduction of nitrite ions corresponding to step ii ₀ b), and thus take charge of this intermediary. It is then conceivable to couple the biological denitrification involving bacteria capable of reducing nitrate ions to nitrite ions (step ii ₀ a)) and abiotic denitrification involving biogenic carbonated green rust capable of reducing nitrite ions to nitrogenous gases (step ii). )).

The method according to the invention therefore has the advantage of being able to be carried out continuously and to lead to a denitrification of liquid media much faster than the known processes of the prior art (of the order of a few hours instead of Several weeks).

The process is inexpensive, with autonomous regeneration involving low maintenance, ecological because based on natural processes, and adaptable to the concentration of nitrogenous material to be treated.

Furthermore, it is known that an accumulation of nitrite ions occurs in the treatment plants whose operation is based on principles of biological nitrification / denitrification but also when mixed cultures are contained in the activated sludge. Thus, the intervention of biogenic carbon green rust on nitrite ions, whatever the metabolism present, prevents their accumulation and improve yields of nitrogen gas. Indeed, according to the process according to the invention, the intervention of green rust during the reduction of nitrate ions can support about 50% of the nitrite ions present in the liquid medium. Ammonium production is not totally avoided but it is significantly reduced. The process according to the invention can be used in domestic and industrial wastewater treatment plants, in water purification plants and in manure treatment or sludge recycling plants.

Finally, the invention also relates to the use of at least one carbonate green rust, of the formula ^II ₆ Fe _(i-X) Fe ₂ ^m _6x Oi2H _(7- _3x) CO3, wherein the ratio of Fe (II) / Fe (III) molar concentrations vary from 1 to 1.5 in order to catalyze, in a liquid medium, the nitrite ion reduction reaction without producing ammonium.

The present invention is illustrated by the following examples, to which it is however not limited.

EXAMPLES

The raw materials used in the examples that follow are listed below:

- Sodium nitrate: Carlo-Erba Reagents 99%

- Sodium nitrite: Sigma-Aldrich> 97%

- Hydrochloric acid (HC1) 37%: Sigma-Aldrich 37%

- Ultra-pure and oxygen-free distilled water (millipore water): Millipore, previously degassed water under N ₂ (Alphagaz, impurities: H ₂ O <3 ppm, O ₂ <2 ppm C _n H _n <0.5 ppm) and filtered using a sterile filter (filter pore diameter: 0.22 μηι)

- Ferrous hydroxide: Merck 99%

- Ferrous chloride: Merck 99%

The experiments of contacting the biogenic carbonate green rust with the nitrite ions according to step ii) of the process according to the invention or with the nitrate ions and the nitrite ions when the steps ii ₀ ) and ii) are carried out from Simultaneously, were conducted in a glove box sold under the trade name Jacomex BS531 by the company Jacomex, under an argon atmosphere to prevent the oxidation of green rust by oxygen present in the air.

The reactions were carried out in 100 ml glass bottles and with rotary stirring (10 rotations per minute). The abiotic conditions were obtained as follows: 0.1 g of biogenic carbonate green rust (hereinafter referred to as GR _L or GR _F ) obtained by bioreduction of lepidocrocite (Lp) of formula γ-FeOOH or by bioreduction of ferric green rust (RVF) of formula Fe ^m ₆ (OH) ₄ (OH) ₄ CO ₃ ) was ultrasonically dispersed in 59 mL of milliQ water and the pH was adjusted to 7.5 ± 0.2 with 1M HCl solution. Two stock solutions of nitrite ions and Nitrate ions of 0.3 M concentration, oxygen-free were prepared. 1 mL of the mother solution of nitrite ions or nitrate ions according to the study carried out, was added to the solution containing biogenic green carbonate rust GR _L or GR _F as prepared above. Thus, the theoretical initial concentration of nitrite ions or nitrate ions is about 5 mM at the start of the reaction.

However, the initial concentration of nitrite ions or nitrate ions determined experimentally varies from about 4 mM to 7 mM. Indeed, in order to preserve the anaerobic conditions of the system, inaccuracy exists on the volume of the mother solution in nitrite ions actually injected into the closed bottles. This inaccuracy induces an initial concentration of experimentally calculated nitrite ions which may be slightly different from the theoretical initial concentration of nitrite ions.

Samples were collected during the reaction time using a syringe with a needle to analyze Fe (II) concentrations, nitrite ions, nitrate ions (when the liquid medium contains nitrate ions) and ammonium during the reaction. On the other hand, the oxidized solid phases (Fe (III) oxyhydroxides) were centrifuged, dried and characterized by X-ray diffraction analysis (XRD).

The concentration of Fe (II) _to tai was determined after mixing 0.5 ml of a collected sample and 0.5 mL of a solution of HCl concentration of 1 M. The concentration of Fe (II) _SO i _u bi _e was determined after filtering a sample on a Minisart brand cellulose membrane filter (filter pore diameter: 0.22 μηι) and mixing 0.5 mL of the filtrate obtained and 0.5 mL of 1M HCl solution. The concentrations of Fe (II) _to tai and Fe (II) _n _e i _u bi were determined seconds after collection of each sample by the modified method of 1,10-phenanthroline (Fadrus and Maly, Analyst, 1975 , 100, 549-554).

The concentrations of nitrite ions, nitrate ions and ammonium were determined from the same filtered samples as those used for determining the concentration of Fe (II) _solu bi _e by ion chromatography with an apparatus sold under the trade name DIONEX ISC 3000.

The pH values in solution were determined using a pH meter sold under the trade name Consort C830 by Fisher Bioblock Scientific and calibrated at pH 4, 7 and 10 with Fisher Scientific standard solutions.

The green carbonate rust obtained according to stage i) of the process according to the invention and the oxyhydroxides of Fe (III) (oxidized solid phases) obtained according to stage ii) of the process according to the invention were subjected to XRD analysis. To avoid oxidation by air, the samples were taken under an argon atmosphere in a glove box (O ₂ content of about 40 ppm) with a syringe and then placed directly on a silicon wafer. and dried in a vacuum desiccator. The pellet containing the dry sample was fixed in an anaerobic chamber. The X-ray diffractometry measurements were carried out with a Co Ka radiation (λ = 0.17889 nm) using a diffractometer sold under the trade name Panalytical X'Pert Pro MPD mounted according to the Debye-Scherrer configuration. using an elliptical mirror to obtain a high flux, a parallel incident beam and a detector sold under the trade name X'Celerator to collect the diffracted beams. The data was recorded in continuous scanning mode with an angle θ between 5 ° and 80 ° and a pitch of 0.0167 °.

The green carbonate rust obtained according to step i) of the process according to the invention were subjected to analysis by transmission electron microscopy (TEM). A Philips CM 20 electronic microscope operating at 200 kV equipped with a dispersive energy spectrometer was used to observe the solid phases. A suspension of the green carbonate rust to be analyzed was rapidly dispersed in air on a grid covered with amorphous carbon and loaded into the microscope analysis support. The products were identified from selected area diffraction models and dispersive energy analysis.

The green carbonate rust obtained according to stage i) of the process according to the invention was also subjected to transmission Mössbauer spectroscopy (SMT) analysis. The samples were prepared under a nitrogen atmosphere in a glove box and rapidly transferred to a cryostat under an inert helium atmosphere prior to measurements at 77 K or 8 K. Môssbauer spectroscopy in transmission was carried out from 8 K to 295 K with a Mössbauer cryostat with variable temperature close to the helium cycle, equipped with vibratory insulation manufactured by Cryo Industries of America, using a spectrometer sold under the trade name Môssbauer constant acceleration with a source of cobalt 57 embedded in a rhodium matrix and an activity of 50 mCi (millicuries) and calibrated with an iron sheet 25 μηι thick at room temperature. The spectra were adjusted using Lorentzian profile lines.

EXAMPLE 1

Preparation and characterization of biogenic carbonate green rust according to step i) of the process according to the invention

Two oxyhydroxides of Fe (III), lepidocrocite (Lp) and ferric green rust (RVF), were used to prepare respectively bioreduction according to step i) of the process according to the invention, two biogenic carbonaceous green rust denominated GR _L and GR _F.

Firstly, inocula of strains of Shewanella putrefaciens bacteria CIP 59.28 equivalent to the Collection of American Type Cultures (ATCC) 12099 were prepared according to the method described in Ona-Nguema et al., Geochim. Cosmochim. Acta, 2009, 73, 1359-1381. The cell density of each inoculum was determined by the number of colony forming units (CFU). A ferri-reducing bacterium concentration of 5.75 × 10 ⁹ CFU / ml was used in the following two bioreductions.

The bioreduction of Lp in biogenic green biogenic rust GR _L was carried out according to the procedure described in Ona-Nguema et al., Environ. Sci. Technol., 2002, 36, 16-20: a solution of Lp of concentration 80 mM and pH 7.5 ± 0.2 was prepared. The medium obtained was then sterilized by autoclave for 20 minutes at a temperature of 120 ° C, and then a suspension of Shewanella putrefaciens CIP 59.28 bacteria was added.

The bioreduction of the RVF in biogenic green carbon rust GR _F was carried out according to the procedure described as follows: an RVF suspension of concentration 80 mM and pH 7.5 ± 0.2 was prepared in glove box under atmosphere N ₂ in water and containers previously sterilized by autoclave for 20 minutes at a temperature of 120 ° C. Only RVF was not sterilized because heating at 120 ° C alters its structure. Sterility of the preparation environment was ensured with a Hofmann electric burner sold by Horo Dr. Hofmann GmbH. Then, a suspension of Shewanella putrefaciens bacteria CIP 59.28 was added.

Samples were collected during the reaction to monitor the Fe (II) production and to determine the bioreduction rates of both Fe (III) oxyhydroxides. The experiments were carried out over 12 days for RVF and 26 days for Lp. Immediately after the suspension of ferri-reducing bacteria was added Fe production is observed (II) _to tai The attached Figure 2 shows the evolution of the concentrations of Fe (II) _to tai (Fe (II) _to t , in mM) as a function of time (in days) during the bioreduction of RVF (solid circles) and during bioreduction of Lp (solid squares). The value n indicates the number of identical replicates. From Figure 2a, 38.6 mM Fe (II) _to tai was produced after 12 days of RVF bioreduction, while the same concentration of Fe (II) _to tai was produced after 24 days of bioreduction. of the Lp. As shown in FIG. 2b, the initial bioreduction rate could be calculated over the first 5 days and according to a second-order kinetics. It is then 5.29 mM / h for RVF and 3.21 mM / h for Lp.

These results suggest an almost twice as fast bioreduction of RVF compared with Lp. In both experiments, the kinetic profiles are similar with rapid production of Fe (II) _to tai on the first day, followed by a deceleration phase until the 7 ^th day and finally a linear production after 7 days. These observations suggest that Fe (III) is more available for the ferri-reducting Shewanella putrefaciens CIP 59.28 in RVF than in Lp.

During the production of Fe (II) _to tai, the appearance of green rust was followed by mineralogical characterization of the precipitates formed and the results are reported in the appended FIG. Thus, FIG. 3a shows a DRX analysis of the RVF (lower part) and of the biogenic carbonate green rust obtained GR _F after 14 days of RVF bioreduction (characteristic peaks labeled GR, upper part). FIG. 3b shows a DRX analysis of the Lp (lower part), the biogenic carbonate green rust obtained GR _L after 14 days of bioreduction of the Lp (intermediate part) and after 33 days of bioreduction of the Lp (characteristic peaks denoted GR , upper part). In these figures, the intensity (in arbitrary units, ua) is a function of the angle 2 Theta (2Θ). These XRD analyzes indicate that all RVF is reduced after 14 days in a single phase corresponding to green carbonate rust whereas the complete bioreduction of Lp is observed after 33 days of incubation. The bioreduction of RVF is therefore faster than that of Lp.

Moreover, green rust GR _L and GR _F respectively obtained after 33 days of bioreduction from Lp and 12 days of bioreduction from RVF were centrifuged in a glove box under nitrogen flow, washed twice. with milliQ water, and dried to be characterized by SMT and MET analyzes.

The attached FIGS. 4a and 4b respectively show the hexagonal form of the crystallites isolated from green carbonate rusts GR _F and GR _L observed by transmission electron microscopy (TEM). The average diameters were respectively measured on 8 and 10 crystallites of GR _F and GR _L and are respectively 2.29 ± 0.42 μηι and 4.96 ± 0.44 μηι.

The Môssbauer spectra are shown in FIGS. 4c and 4d on which the transmittance (in%) is a function of the speed (in mm.s ^-1 ). The Môssbauer spectra of the samples of GR _F (FIG. 4c) and GR _L (FIG. 4d) measured at 77K show three quadrupole doublets The ratio of Fe (II) / Fe (III) molar concentrations in the GR _F and GR _L green carbonate rust can be determined from the D1 doublets (attributed to Fe (II) carbonate green rust), D2 (also attributed to Fe (II) of carbonate green rust) and D3 (attributed to Fe (III) of green carbonate rust) characteristic of a green rust spectrum and the ratio of relative areas corresponding [AR (D1) + AR (D2)] / [AR (D3)] The ratios of Fe (II) / Fe (III) molar concentrations of GR _F and GR _L are respectively 1 and 1.25. Similar ratios can be obtained by directly calculating molar concentrations of [Fe (II) + Fe (III)] _to tai - Fe (II) _to tai in the green carbonate rust GR _F and GR _L. From a solution of Lp of 80 mM concentration, 38.6 mM Fe (II) _to tai are produced at 26 days of incubation while there remains 41.4 mM Fe (III). From an RVF solution of 80 mM concentration, 43.2 mM Fe (II) _to tai are produced at 12 days of incubation while 36.8 mM Fe (III) remains. The ratios of Fe (II) / Fe (III) molar concentrations of GR _F and GR _L estimated according to this method are thus respectively 0.93 and 1.17. These theoretical ratios are slightly underestimated since they are calculated from the concentration values of Fe (II) _to tai during the analysis of the last samples at 12 days for bioreduction of Lp and at 27 days for bioreduction. of the RVF and are not calculated from the actual Fe (II) _to tai concentration values at the end of the experiments, ie 14 days for RVF bioreduction and 33 days for bioreduction of RVF. the Lp. EXAMPLE 2

Reduction of the nitrite ions present in a liquid medium by contact with a biogenic green carbon rust GRg or GRr according to step ii) of the process according to the invention

The GR _F and GR _L biogenic green carbonate rusts obtained above in Example 1 (0.1 g) were incubated with 5.4 mM nitrite ions at pH 7.5.

It has been observed that these carbonate green rust particles are in equilibrium with 0.59 mM Fe (II) _solved two _E - 5 The figure attached shows the evolution of the concentrations (in mM) nitrite ions (solid triangles) ammonium (solid diamonds), and Fe (II) _to tai (Fe (II) _to t, solid circles) as a function of time (in hours).

In the two experiments corresponding to FIGS. 5a and 5b, after addition of the sodium nitrite solution, the Fe (II) _to tai is immediately oxidized and the nitrite ions reduced. In addition, no ammonium formation is observed during the use of such GR _F and GR _L biogenic green rusts, while the use of GRQ synthetic green rust mainly leads to ammonium (see example Comparative 2 below). When GR _F is used, the reaction is rapid in the first thirty minutes with the rapid oxidation of Fe (II) _to tai and then the reaction follows a second kinetics slower to 48 hours of reaction (Figure 5a). The reaction is stopped after 48 hours by total depletion of total Fe (II). With GR _L , the same observations can be made, except that the total oxidation of Fe (II) _to tai is slower and the reaction It is not over after 6 days (ImM of Fe (II) _to tai is still present after 6 days) (Figure 5b). For both reactions, the ratio of Fe (II) _O 2 reduced _O 2 reduced molar concentrations was calculated after 48 hours of reaction; it is 2.1 for GR _F and GR _L. This suggests the establishment of the same type of reaction between the nitrite ions and the two green biogenic carbonate rusts GR _F and GR _L. However, the kinetics of reduction of nitrite ions and oxidation of Fe (II) _to tai is higher when GR _F is used. This can be explained by the fact that the GR _F particles are smaller than those of GR _L and thus have more sites of reactive surfaces and therefore an ability to reduce nitrite ions faster than GR _L.

The final oxidation products of biogenic green carbonate green rusts GR _F and GR _L were characterized by XRD. The appended FIG. 6 shows the intensity (in arbitrary units, ua) as a function of the angle 2-Theta (in degrees). The oxidation of biogenic carbonate green rust by nitrite ions leads to the production of Fe (III) oxyhydroxides with no trace of magnetite. Indeed, the oxidation of GR _F and GR _L leads to a mixture of goethite (characteristic peaks Gt) and lepidocrocite (characteristic peaks noted Lp) (Figure 6, upper part: oxidation of GR _F and lower part: oxidation of the GR _L ).

Larger amounts of biogenic GR _F and GR _L biogenic green rust obtained above in Example 1 (0.3 g) were then incubated with 7 mM nitrite ions at pH 7.5. As a result, the ratio of the molar concentrations [Fe (II) ions present in green carbonate rust] / [nitrite ions] is 3.

In the two experiments corresponding to FIGS. 5c and 5d, it is observed that this ratio of the molar concentrations [Fe (II) ions present in the green carbonate rust] / [nitrite ions] = 3, is sufficient for the complete reduction of the nitrite ions after 5 days of reaction with GR _F (Figure 5c) and still without production of ammonium. With GR _L (Figure 5d) _; the reaction is slower, probably because of the larger size of these particles which thus have fewer surface reactive sites. For both reactions, the ratio of Fe (II) _O 2 _O 2 reduced molar concentrations was calculated after 48 hours of reaction. It is 2.7 when GR _F and 3.2 when GR _L is used. The presence of an excess of GR _F or GR _L with respect to the nitrite ions implies that after 48 hours of reaction, the amount of nitrite ions reduced is similar to that reduced in the experiments where 3 times less ions Fe (II) _to tai were used. In contrast, more Fe (II) _to tai ions have been oxidized, suggesting a more complete reduction of nitrite ions. Fe (II) _to tai being introduced in excess, only a part of this Fe (II) _to tai is oxidized.

EXAMPLE 3

Bacterial denitrification and accumulation of nitrite ions in a liquid medium according to step iin) of the process according to the invention

Bacterial nitrate ion reduction experiments involving model bacteria were conducted with 3 distinct strains of Shewanella species (S. putrefaciens CIP 59.28, S. putrefaciens CIP 80.40 and S. oneidensis MR1). Each strain (2.5 × 10 ⁹ CFU / ml) was incubated with 5 to 6.5 mM nitrate ions or with 4 to 5 mM nitrite ions at pH 7.5.

The attached FIG. 7 shows the evolution of the concentrations of nitrate ions (in mM) (solid circles), nitrite ions (solid squares) and ammonium ions (solid triangles) as a function of time (in minutes) during the contacting of the ions nitrates (diagrams on the left) or nitrite ions (diagrams on the right) with the S. putrefaciens bacteria CIP 59.28 (Figure 7a), with S. putrefaciens CIP 80.40 bacteria (Figure 7b) and with bacteria S. oneidensis MR1 (Figure 7c).

The strains S. putrefaciens CIP 59.28 and S. putrefaciens CIP 80.40 accumulate the nitrite ions in about 90 minutes and the production of ammonium appears after about 120 minutes. The strain S. oneidensis MR1 accumulates nitrite ions in only about 30 minutes and ammonium production appears more rapidly after about 90 minutes.

Experiments have shown that these three strains accumulate enough nitrite ions to consider their possible management by biogenic green carbonate rust.

EXAMPLE 4

Coupling of bacterial denitrification and abiotic denitrification (steps iin) and ii) of the process according to the invention carried out simultaneously)

The attached FIG. 8 shows the evolution of the concentrations of nitrate ions (in mM) (solid circles), nitrite ions (solid squares), ammonium (solid triangles) and Fe (II) _to tai (Fe (II) tot, solid diamonds ) versus time (in minutes) during contacting of 6 mM of nitrate ions with 0.1 g of GR _F biogenic bacteria and S. putrefaciens CIP 59.28 (8a), S. putrefaciens CIP 80.40 ( FIG. 8b) and S. oneidensis MR1 (FIG. 8c) (3.75 × 10 ⁹ CFU / ml).

The nitrate ions are rapidly reduced by the bacteria in 1 to 2 hours depending on the strain used. Nitrite ions are observed as intermediates. The results show that the bacteria first convert the nitrate ions to nitrite ions, and then the bacteria reduce some of these nitrite ions leading to the production of ammonium.

In addition, the Fe (II) ions of biogenic carbonate green rust are oxidized, demonstrating an intervention of green rust particles in the system. As soon as all the nitrate ions have been reduced, the ferri-reductive activity is put in question to produce Fe (II) again.

In FIGS. 8a, 8b and 8c, it is observed that the concentration of Fe (II) _at initial tai (at t = 0) is substantially identical to the concentration of Fe (II) _{at the} final tai (at t = 300 minutes). Moreover, unrepresented XRD and SEM (scanning electron microscopy) analyzes show that a green carbonate rust is regenerated well and that the green rust particles are of similar sizes to those of the green rust starting particles. So, we have regeneration according to step iii) of a carbonate green rust having a ratio of Fe (II) / Fe (III) molar concentrations ranging from 1 to 1.5.

The attached FIG. 9 shows the accumulation of nitrite ions (in mM) when 6 mM nitrate ions are brought into contact with bacteria alone (2.5 × 10 ⁹ CFU / ml) (solid line) or with bacteria ( 3.75 x 10 ⁹ CFU / ml) and 0.1 g of GRp (dotted line) as a function of time (in minutes). Different strains of bacteria are used: S. putrefaciens CIP 59.28 (solid circles), S. putrefaciens CIP 80.40 (solid squares) and S. oneidensis MR1 (solid triangles).

From FIG. 9, it appears that in the presence of green rust, the accumulation of nitrite ions is reduced by 50% when the incubations are carried out with the bacteria of strains S. putrefaciens CIP 80.40 and S. oneidensis MR1. This shows the involvement of biogenic green carbonate rust in the presence of bacteria capable of reducing nitrate ions.

COMPARATIVE EXAMPLE 1

Reduction of nitrate ions in liquid medium by green rust

biogenic carbonates GRF and GRr

Comparative nitrate reduction tests were conducted with the green rust GR _F prepared according to Example 1 and having a Fe (II) / Fe (III) molar concentration ratio of 1. The attached FIG. evolution of concentrations (in mM) in Fe (II) _to tai ( ⁼ Fe (II) _G RF + Fe (II) _S oi _u bie) (Fe (II) _to t, solid circles), in Fe (II) _{n / a} i _u bie (Fe (II) _n i, filled squares) and nitrate ions (filled triangles) versus time (in days) during contacting of 0.1 g of carbonate green rust GR _F with 5 mM nitrate ions for 13 days at pH 6.5 (Fig. 10a), pH 7 (Fig. 10b), and pH 10 (Fig. 10c).

It is thus shown that the nitrate ions are not reduced by carbonated green rust with a ratio of Fe (II) / Fe (III) molar concentrations of 1.

The same reaction carried out with GR _L carbonate green rust having a Fe (II) / Fe (III) molar concentration ratio of 1.25 and prepared according to Example 1, shows the reduction of nitrate ions but only after 5 months of 'interaction. COMPARATIVE EXAMPLE 2

Preparation of GRRI synthetic green rust and reduction of nitrite ions in the presence of this green rust

A synthetic green rust GR _C i corresponding to the formula [Fe ₄ ⁿ (i _-X) ^m Fe _4x OH ₈ Cl.nH ₂ O] was prepared by air oxidation of ferrous hydroxide slurry in the presence a slight excess of dissolved ferrous chloride (Refait et al., Corrosion Science, 1997, 39, 539-553).

The appended FIG. 11 (bottom part) shows the DRX analysis of the GR _C i synthetic green rust thus obtained (characteristic peaks denoted RV). In this figure, the intensity (in arbitrary units) is a function of the angle 2Θ (in degrees). According to the literature, the ratio of Fe (II) / Fe (III) molar concentrations in GRci synthetic green rust is close to 3 (Genin et al., Solid State Sciences, 2004, 39, 705-718).

40 ml of an aqueous suspension of green rust GRQ (0.3 g) containing 2 mM Fe ions corresponding to Fe (II) _solu bi _e naturally in equilibrium with the _GRci. Figure 12a attached shows the changes in concentrations (in mM) nitrite ions (closed triangles), ammonium (closed diamonds), Fe (II) _solu bi _e (filled squares) and Fe (II) _to tai (round full) as a function of time (in hours). The injection of 5 mM sodium nitrite results in a decrease in the concentration of Fe (II) _to tai and Fe (II) _so i _u bi _e and rapid consumption of nitrite ions. The Fe (II) of green rust and the soluble Fe (II) participate in the reaction and ammonium is produced. The consumed soluble Fe (II) reappears rapidly after 40 minutes by equilibration with green rust only when the nitrite ions have been completely reduced. It is assumed that the absence of Fe (II) _solved bi _e during the first 40 minutes is explained by his release and simultaneous consumption in the presence of nitrite ions.

In order to verify which type of Fe (II) is involved in ammonium production, the denitrification in the presence of a solution of Fe ions from a FeCl ₂ solution of 2 mM concentration without green carbonate rust was tested. . Four starting mixtures I ₀ , II ₀ , IIIo and IV ₀ containing varying concentrations of 0.05 mM to 0.5 mM nitrite ions, 2 mM nitrate ions, 1 mM ammonium ions and 0 or 2 mM Fe ions were prepared. The IVo control solution does not contain Fe ²⁺ ions (see Table 1 below). TABLE 1

Figure 12b attached shows the concentrations (in mM) nitrite ions ⁽¹ black rectangles), nitrate ion (2 ^th gray rectangles), ammonium (3 ^rd rectangles having bias bars) and ions Fe ²⁺ (4 ^th rectangles comprising horizontal bars) in the four final mixtures I, II, III and IV respectively obtained after preparation of the starting mixtures I ₀ , II ₀ , IIIo and IV ₀ and reaction for 48 hours.

Under all conditions, Fe reduces nitrite ions without producing ammonium, involving the formation of nitrogen gas. No reduction of nitrite ions was observed in the control solution without Fe ions and no reactivity of Fe ions with nitrate ions was observed. Thus, ammonium production in the NQI nitrite reduction experiment should be attributed solely to the portion of nitrite ions reduced by GRQ green green rust.

After 1:30 (Figure 12a), the reaction was terminated, the nitrite ions were completely reduced. Specifically, after 20 minutes, 0.67 mM ammonium was formed and was assigned to 0.67 mM nitrite reduced by green rust on a total of 2.94 mM reduced at this time. Thus, green rust contributes only 20% of the reaction when Fe ions coexist. Then, all the Fe ions are consumed between 20 and 40 minutes and 54% of the reduced nitrite ions are converted to ammonium. Thus, the involvement of green rust increased to 54% after the disappearance of Fe ions. This implies, however, that 46% of nitrite ions are always reduced to nitrogen gas. This might suggest that Fe ions are gradually re-dissolved from green rust as early as 20 minutes and that their absence between 20 and 40 minutes would imply that they are immediately oxidized by nitrite ions at the onset of their release. The total balance at a concentration of 2 mM reappears only after 1:30, when nitrite ions are absent.

The appended FIG. 11 (upper part) also shows the DRX analysis of the product obtained after oxidation of the green green rust GR _C i by nitrite ions. The product obtained comprises a mixture of magnetite (peaks characteristic Mt and residual green rust GRQ residual (characteristic peaks noted RV) .Thus, the difference in reactivity between synthetic green rust GRci and green biogenic carbonaceous rust GR _F and GR _L (Example 2) with respect to nitrite ions is also observed in mineralogy since the oxidized mineral phases obtained during the reduction of nitrite ions are different.

COMPARATIVE EXAMPLE 3

Preparation of GRrm synthetic green rust and reduction of nitrite ions in the presence of this green rust

A GRC ₀₃ synthetic carbonyl green rust having the formula Fe ⁿ ₄ Fe ^m ₂ (OH) 12 CO ₃ .3H ₂ O was prepared by co-precipitation of FeSO ₄ .7H ₂ O and Fe ₂ SO ₄ .5H ₂ O in condition anoxia. The ratio of Fe (II) / Fe (III) molar concentrations in GRc ₀₃ synthetic green carbon rust obtained was 2 (Bocher et al, Solid State Sciences, 2004, 6, 117-124). The green rust was then centrifuged, washed with degassed water and dried in a vacuum desiccator under anoxic conditions.

The synthetic GRco ₃ thus obtained (0.1 g of powder) was resuspended in the presence of 6.5 mM nitrite ions at pH 8 in a volume of 60 ml. The appended FIG. 13 shows the evolution of the concentrations (in mM) in nitrite ions (solid triangles) and in ammonium (solid diamonds) as a function of time (in hours).

It is observed that after 48 hours of contacting the nitrite ions with GRCO ₃ , 3.05 mM nitrite ions are reduced and 0.70 mM ammonium is produced. This leads to the conclusion that at least a significant part of the nitrite ions are reduced to ammonium. EXAMPLE 5

Coupling of bacterial denitrification and abiotic denitrification (step iin) and ii) of the process according to the invention carried out simultaneously involving an aboriginal biological consortium of purified water comprising nitrite ions or nitrate ions

Two experiments were carried out to determine the capacity of native bacteria of purified wastewater to reduce nitrate and nitrite ions in the presence or absence of green biogenic carbon rusts GR _L and GR _F as obtained in the example 1.

It is important to note that this purified wastewater included approximately 60 mg / 1 ammonium prior to the experiments, likely due to a malfunction in the nitrification treatment used in the Rhysostep ^® SAUR wastewater treatment plant in Douy-la- Ramée (77), from where it was taken. This station is of type filters planted with reeds.

In a first experiment, this purified wastewater was incubated with 76.4 mg / l N-NO ₂ ^~ approximately.

The terms "mg / 1 N-NO ₂ ^~ ", "mg / 1 N-NO ₃ ^" "and" mg / 1 N-NH ₄ ⁺ "respectively mean the nitrogen concentrations (in mg / l) contained in NO ₂ ^" , in NO ₃ ^" , and in NH ₄ ⁺ .

The attached FIG. 14 shows the evolution of concentrations (in mg-N / 1) of nitrite ions (solid triangles) and ammonium (solid diamonds) as a function of time (in days) in this purified wastewater comprising nitrite ions, in the absence of biogenic carbonate green rust (FIG. 14a), in the presence of GR _L (FIG. 14b) or in the presence of GR _F (FIG. 14c).

In purified waste water alone (ie without biogenic carbon green rust), 30.6 mg / 1 N-NO ₂ ^" are reduced in 5 days without ammonium production (Figure 14a). in the purified wastewater, the reduction of nitrite ions is strongly improved still without any production of ammonium (Figures 14b and 14c), after 4 days of incubation, 70.4 mg / 1 N-NO ₂ ^" and 76.4 mg / 1 N-NO ₂ ^" are reduced in the presence of GR _L and GR _{F respectively} . Thus, the reactivity of the green biogenic carbonaceous rust in this purified wastewater is not impaired and remains similar to that already observed on synthetic waters. These results demonstrate that the biogenic carbonated green rust has a high reactivity vis-à-vis the nitrite ions (step ii) of the process according to the invention) even in a natural liquid medium comprising a natural biological consortium (ie a purified wastewater including an indigenous biological consortium). Biogenic carbonate green rust reduces nitrite ions to nitrogen gas without producing ammonium.

This experiment shows the reactivity of biogenic carbonate green rust with respect to nitrite ions in purified wastewater.

In a second experiment, the purified wastewater as defined above was incubated with nitrate ions.

FIG. 15 shows the evolution of concentrations (in mg-N / 1) in nitrate (solid round), nitrite (full triangles) and ammonium (solid diamonds) ions as a function of time (in days) in this purified wastewater comprising nitrate ions, in the absence of biogenic carbonated green rust (FIG. 15a) or in the presence of GR _L (FIG. 15b).

Figure 15a shows that in purified wastewater incubated with 73 mg / l N-NO ₃ ^~ and in the absence of GR _L , 28.6 mg / 1 N-NO ₃ ^" are reduced on the first day with accumulation. maximum nitrite ion of 16.8 mg / 1 N-NO ₂ ^~ during the first hour and no ammonium production is observed.No further reduction of nitrate is observed the following days.

Figure 15b shows that in purified waste water incubated with 83.9 mg / 1 N-NO ₃ ^" in the presence of GR _L (Figure 2b), the nitrate ions are completely reduced in 4 days, still without ammonium production , with 47.1 mg / 1 N-NO ₃ ^" reduced on the first day and a maximum nitrite ion accumulation of 24.4 mg / 1 N-NO ₂ ^" .

Thus, the reduction of nitrate ions in the presence of biogenic carbonate green rust and of a natural biological consortium is markedly improved compared to that conducted in the presence of water comprising a single type of pure bacteria (see Example 4). The accumulation of nitrite ions is less prolonged over time, suggesting that biogenic carbonate green rust reacts with nitrite ions as they are produced.

These results can be explained by the fact that in the presence of a biogenic carbonate green rust according to the invention, the autochthonous biological consortium of the treated waste water seems to use the organic carbon "pool" only to catalyze the conversion of nitrate ions to nitrites (step ii ₀ a) (Figure 1), the nitrite ions are then in turn reduced by biogenic carbonated green rust to nitrogen gas (step ii); whereas when purified waste water undergoes purely biological denitrification, the indigenous biological consortium seems to use the organic carbon "pool" to catalyze the transformation of nitrate ions into nitrites and the transformation of nitrite ions into nitrogenous gases (steps ii ₀ a and ii ₀ b) (Figure 1).

Thus, in the presence of the biogenic carbonated green rust of the invention, the transformation of nitrite ions into nitrogenous gases is mainly abiotic and does not directly require the additional supply of organic carbon.

Claims

1. A method of mineral denitrification biologically assisted in a liquid medium, said method being characterized in that it comprises the following steps: i) a step of preparation of a carbonated green rust by bioreduction of an Fe (III) oxyhydroxide in anaerobic conditions in the presence of a culture of at least one bacterium having a ferri-reducing activity until a carbonated green rust with a ratio of Fe (II) / Fe (III) molar concentrations ranging from 1 to 1 is obtained , 5;

2. Method according to claim 1, characterized in that the Fe oxyhydroxide (III) used in step i), is selected from lepidocrocite, ferric green rust and ferrihydrite.

3. Method according to claim 1 or 2, characterized in that the bioreduction steps i) and iii) are independently carried out in the presence of a pure culture of at least one bacterium having a ferri-reducing activity.

4. Method according to any one of the preceding claims, characterized in that the bacteria having a ferri-reducing activity that can be used in steps i) and iii), are independently selected from the bacterial species of aquatic environments belonging to the genera Shewanella, Geobacter , Pseudomonas, Desulfovibrio, Geothrix and Pelobacter.

5. Method according to any one of the preceding claims, characterized in that the bioreduction time in step iii) varies from 5 hours to 5 days.

6. Method according to any one of the preceding claims, characterized in that the concentration of ferri-reducing bacterium in steps i) and iii) is greater than 10 CFU / ml.

7. Method according to any one of the preceding claims, characterized in that the contact time in step ii) is less than 5 days.

8. Method according to any one of the preceding claims, characterized in that the amount of green rust used in step ii) is such that the ratio of the molar concentrations [Fe (II) ions present in the green carbonate rust] / [nitrite ions] is at least 3.

9. Method according to any one of the preceding claims, characterized in that steps ii) and iii) are performed simultaneously in one and the same step.

10. Process according to claim 9, characterized in that it is carried out continuously and said steps ii) and iii) are repeated until the total possible exhaustion of the nitrite ions in the starting liquid medium.

11. Process according to any one of the preceding claims, characterized in that a step ii ₀ ) for reducing the nitrate ions present in the liquid medium to nitrite ions, in the presence of a culture of at least one bacterium capable of reduce the nitrate ions to nitrite ions, is carried out prior to step ii).

12. The method of claim 1 1, characterized in that the bacteria used in step ii ₀ ), capable of reducing nitrate ions to nitrite ions, are chosen from the genera Alcaligenes, Paracoccus, Thiobacillus, Vibrio, Desulfovibrio, Shewanella, Micrococcus, Geobacter, Pseudomonas and Bacillus.

13. The method of claim 1 1 or 12, characterized in that the reduction time in step ii ₀ ) varies from 2 hours to 24 hours.

14. Method according to any one of claims 11 to 13, characterized in that the concentration of bacteria capable of reducing nitrate ions to nitrite ions in step ii ₀ ) is at least 10 CFU / ml.

15. Method according to any one of claims 11 to 14, characterized in that the amount of green rust used in step ii) is such that the ratio of the molar concentrations [Fe (II) ions present in the rust carbonated green] / [nitrate ions] is at least 3.

16. Method according to any one of claims 11 to 15, characterized in that the steps ii ₀ ), ii) and iii) are performed simultaneously in one and the same step.