Molten salt device resistant to corrosion

Abstract

The invention relates to a device (10) comprising: - a component (20) configured to contain or circulate a solution, the component (20) having a wall (21) made of steel or a nickel-based alloy comprising chromium, - a solution (30), in contact with the wall (21), comprising a mixture of molten chloride salts, the solution also comprising Ti2+ ions and Ti3+ ions, the Ti3+ / Ti2+ ratio being between 20 / 80 and 70 / 30.

Description

[0001] CORROSION-RESISTANT MOLTEN SALT DEVICE

[0002] DESCRIPTION

[0003] TECHNICAL FIELD

[0004] The present invention relates to the general field of nuclear and concentrated solar power.

[0005] The invention relates to a device comprising an element having a wall made of steel or a nickel-based alloy comprising chromium in contact with a solution of molten salts (molten chlorides).

[0006] The invention also relates to a method for preparing such a device as well as a method for measuring the potential of a molten salt solution of such a device.

[0007] The invention finds applications in many industrial fields, and in particular for molten salt nuclear reactors including their primary and secondary circuits or for concentrated solar power plants.

[0008] The invention is particularly interesting since it makes it possible to considerably reduce the corrosion phenomena occurring at the interface between a metal alloy wall and a solution of molten salts, even for devices operating at high temperatures (typically above 500°C).

[0009] STATE OF THE PRIOR ART

[0010] From an industrial point of view, molten salts can be used for the production of metals by electrolysis (Al, Na, Mg, etc.), or as a heat transfer medium for concentrated solar thermal applications and for nuclear applications, as fuel or as a heat transfer fluid, for example, for nuclear reactors and including actinide converters.

[0011] For example, in the case of solar applications, thanks to the thermal inertia of molten salts, it is possible to maintain high temperatures for several hours even after sunset. For applications at very high temperatures (typically above 600°C), chloride salts (such as KCl, NaCl and MgCb) are preferred over nitrate salts which decompose at lower temperatures.

[0012] However, the high chemical reactivity of these environments creates very aggressive conditions conducive to corrosion. This phenomenon is all the more accentuated when the temperatures used in solar and nuclear systems are very high. Thus, the materials in contact with these environments are heavily corroded, which poses problems for the operation and longevity of the installations. In addition, the dissolution of metallic materials also modifies the composition and thermophysical properties of the heat transfer fluid by adding CrC and FeC for example.

[0013] Even though the corrosive aspect of molten chlorides is known, the results of corrosion rates or material consumption (thickness per unit of time) are disparate and reflect different and poorly controlled experimental conditions.

[0014] Nevertheless, corrosion in molten chlorides continues to be the subject of numerous academic and industrial studies. Several solutions have been considered, such as:

[0015] - use steels with good corrosion resistance (nickel-rich steels),

[0016] - limit the presence of impurities such as hydroxides and oxides, HCl and C which participate in the oxidation or complexation mechanisms of steel constituents; for example by using a controlled inerting atmosphere (argon in particular) to reduce corrosion,

[0017] - modify the chemical potential of the molten salt solution.

[0018] While controlling the chemical potential of the molten salt solution is an effective way to limit corrosion, its value can be affected by fission and transmutation reactions and by the introduction of impurities into the system (H2O, O2, etc.). This is particularly likely in the case of a nuclear reactor, the presence of EuCb, for example, has, due to the high potential of the Eu couple 3+ / Eu 2+ , an oxidizing effect on chromium, iron and also nickel. The use of a redox buffer has historically been used in molten salt reactors to control the chemical potential of the fluoride salt and limit corrosion. For example, it has been shown that the U couple 3+ / U 4+ in molten fluorides helps mitigate corrosion by controlling the proportions of U 3+ and U 4+ dissolved in salt (Journal of Nuclear Materials 440 (2013) 243-249). A concentration ratio in salt U 4+ / U3+ < 100 is recommended to limit steel corrosion and intergranular corrosion by tellurium up to 750°C. Adjustment of uranium oxidation states can be achieved by using a reducing metal, such as uranium or beryllium.

[0019] US 2017 / 0294241 Al also proposes adding uranium to control the redox potential and mitigate corrosion in a molten salt reactor. Many elements that can act as a buffer pair are cited without giving precise experimental conditions.

[0020] This concept has also been applied for solar salts with NaCI-KCI-MgC concentration as described in WO 2019 / 075177 Al. To limit corrosion, it is possible to add magnesium metal to alkali and alkaline earth chloride salts containing MgC. The addition of Mg metal provides the electrochemical couple Mg 2+ / Mg° in solution which from a thermodynamic point of view is favorable for inhibiting the dissolution of chromium, iron and nickel. The use of excess Mg metal makes it possible to maintain the potential of the solution at a potential lower than that of dissolution of the major constituents of steels.

[0021] However, the presence of excess dissolved Mg in the salt can lead to the formation of Mg-Ni alloys on the surface of materials such as 316H steel (Corrosion Science 194 (2022) 109914). In addition, magnesium metal is a strong reducing agent capable of reducing actinide chlorides to metallic actinide. This solution therefore seems unsuitable for a molten salt nuclear reactor, since it would lead to the formation of metallic actinides that would potentially make the fuel salt inhomogeneous. In addition, metallic actinides could alloy with metallic materials in contact with the molten salt. For example, plutonium leads to the formation of defined compounds with iron (Fe2Pu and FeePu) that are liquid at low temperatures depending on the composition.

[0022] In document W02017 / 060741 Al, to reduce the corrosion of steels in contact with molten halogen salts (fluorides or chlorides), it is indicated that more reactive sacrificial metals, such as zirconium, vanadium or titanium, can be used. This document focuses mainly on the Pauling scale of metals. This document mentions the following pairs of species: ZrF2 and ZrF4, TiF2 and TiF4, VF2 and VF3. However, the existence of ZrF2 in molten salts has never been demonstrated, nor even that of TiF2-

[0023] Furthermore, this document does not specify the experimental conditions to be implemented. In particular, this document does not give precise indications on the concentrations or the redox potential to target, on how to find the appropriate buffer pair depending on the molten salts and the material to be protected, or even on the method to control and maintain the stability of the species in the molten salt solution.

[0024] STATEMENT OF THE INVENTION

[0025] An aim of the present invention is to propose a device making it possible to limit the corrosion phenomena of metal alloys in contact with molten chlorides.

[0026] For this, the present invention proposes a device comprising:

[0027] - an element configured to contain or circulate a solution, the element having a wall made of steel or a nickel-based alloy comprising chromium,

[0028] - a solution, in contact with the wall, comprising a mixture of molten chloride salts, the solution further comprising Ti ions 2+ and Ti ions 3+ , the concentration ratio Ti 3+ / Ti 2+ being between 20 / 80 and 70 / 30.

[0029] The invention is fundamentally distinguished from the prior art by the use of titanium to reduce the corrosion of metal parts in contact with molten chloride salts.

[0030] Titanium is much less reducing than magnesium and therefore more suitable for application in a molten salt reactor (MSR) while being negative enough to prevent corrosion of metal alloys, for example steel. In addition, titanium is transparent enough to neutrons to limit its transmutation in the reactor.

[0031] In addition, the control of the Ti ion concentration 2+ and in Ti ions 3+ is essential to place the potential domain where effective redox control is operated and therefore to limit or even prevent the dissolution of metal alloys. Outside this immunity domain, elements such as chromium and iron are dissolved in the molten salt solution.

[0032] Such a device can be used for concentrated solar power applications and for molten salt nuclear reactor (MSR) applications.

[0033] According to an alternative embodiment, the device is a nuclear device and the element is a primary circuit of a nuclear fission reactor or a secondary circuit of a nuclear fission reactor. The primary circuit of a nuclear fission reactor and the secondary circuit of a nuclear fission reactor allow the mixture of chloride salts to be circulated.

[0034] The reactor can be a molten salt reactor in fast neutron spectrum.

[0035] According to another embodiment, the device is a solar device, for example a concentrated solar power plant, and the element is a fluid circuit of molten chlorides or a reservoir of molten chlorides.

[0036] The solution includes a mixture of molten chlorides. This can be, for example:

[0037] - a binary mixture or a ternary mixture of alkali and alkaline earth metal chlorides, generally used as a heat transfer fluid, or

[0038] - a mixture containing one or more actinide chlorides generally used for nuclear applications.

[0039] According to a first advantageous embodiment, the mixture of molten chloride salts is a binary or ternary mixture of alkali and alkaline earth metal chlorides. Preferably, the mixture is obtained from molten chlorides chosen from LiCI, NaCI, KCI, MgC, CaC and BaC. For a binary mixture, it is possible to choose, for example, a NaCI-MgC mixture or a KCI-MgCb mixture, whereas for a ternary mixture it is possible to choose, for example, a NaCI-KCI-MgC mixture. These mixtures form a particularly interesting heat transfer fluid when the element is a secondary circuit of a nuclear fission reactor, a reservoir or a fluid circuit of a concentrated solar power plant.

[0040] According to a second advantageous embodiment, the mixture of molten chloride salts comprises at least one actinide chloride. By at least, it is meant that the mixture of molten chlorides comprises one or more actinide chlorides. The actinide chlorides are compounds of formula AcCb with Ac an actinide. Preferably, Ac is Pu, Am, U or Cm, even more preferably Pu, Am or Cm. More preferably, Pu or Am will be chosen. The actinide chloride(s) will preferably be added to one of the binary mixtures described above.

[0041] In other words, the mixture of molten chloride salts is a binary mixture of alkali and alkaline earth metal chlorides and one or more actinide chlorides (preferably one or two actinide chlorides). The actinide chlorides are preferably chosen from UCh, PuCb and AmCb, and even more preferably from PuCb and AmCb.

[0042] Preferably, the mixture of molten chloride salts is a NaCI-MgC -PuCb or NaCI-MgC -PuCb-AmCb mixture. These mixtures form the fuel of the nuclear reactor. This embodiment is particularly advantageous when the element is a primary circuit of a nuclear fission reactor.

[0043] As a chemical analogue of NaCI-MgC -PuCb, it is possible to use the NaCI-MgC -CeCb mixture.

[0044] Advantageously, the solution is at a temperature between 450°C and 700°C and preferably between 550°C and 650°C.

[0045] The wall is made of steel or a nickel-based alloy. Advantageously, the wall is made of a material that is mechanically resistant to high temperatures. Preferably, the wall is made of stainless steel, for example 316L steel. According to another preferred embodiment, the wall is made of a nickel-based alloy preferably comprising at least 40% by mass of nickel. For example, it is a nickel-chromium-molybdenum alloy, in particular NiCr22Mo9Nb.

[0046] Advantageously, the Ti ratio 3+ / Ti 2+ is between 30 / 70 and 50 / 50.

[0047] The device has many advantages. For example, it allows effective control of the potential to a value low enough to be placed in the immunity range. Moreover, in the case of a molten salt nuclear reactor (MSR), such a device prevents the formation of metallic actinide and in particular the formation of metallic plutonium unlike metallic Mg.

[0048] Moreover, such a device is simple to prepare.

[0049] The invention also relates to a solution comprising a mixture of molten chloride salts, intended to be in contact with a wall of an element such as a primary circuit of a nuclear fission reactor, a secondary circuit of a nuclear fission reactor, a tank of a concentrated solar power plant or a fluid circuit of a concentrated solar power plant, the wall being made of steel or a nickel-based alloy comprising chromium. The solution further comprises Ti ions. 2+ and Ti ions 3+ , the concentration ratio Ti 3+ / Ti 2+ being between 20 / 80 and 70 / 30.

[0050] Preferably, the mixture of molten chloride salts is a binary or ternary mixture of alkali and alkaline earth metal chlorides or a binary mixture of alkali and alkaline earth metal chlorides and one or more actinide chlorides. The actinide chlorides are preferably chosen from UCI3, PuCh and AmCIs, and even more preferably from PuCIs and AmCIs.

[0051] The invention also relates to a method for preparing a device as defined above, the method comprising the following steps:

[0052] - providing a device comprising an element configured to contain or circulate a solution, the element having a wall made of steel or a nickel-based alloy comprising chromium, and a solution, in contact with the wall, comprising a mixture of molten chloride salts,

[0053] - immerse titanium metal in the solution, for a sufficient time to have Ti ions2+ and Ti ions 3+ , the Ti report 3+ / Ti 2+ being between 20 / 80 and - possibly, add ions of a metal having a potential lower than the dissolution of the chromium of the iron metal alloy in the solution, whereby the formation of Ti ions is accelerated 2+ and Ti ions 3+ by oxidation-reduction reactions.

[0054] The ions can be iron ions or chromium ions for example.

[0055] The invention also relates to a method for measuring the potential of a solution of a device as defined previously.

[0056] Indeed, it is particularly difficult to use conventional electrodes in molten salts, such as Ag type reference electrodes. + / Ag° whose stability over time in these aggressive environments is questionable. There is therefore a need to be able to control the potential of a molten salt solution.

[0057] According to the invention, the method for measuring the potential of a solution of a device as defined previously comprises the following successive steps:

[0058] - providing a device comprising an element configured to contain or circulate a solution, the element having a wall made of steel or a nickel-based alloy comprising chromium, and a solution, in contact with the wall, the solution comprising a mixture of molten chloride salts, Ti ions 2+ and Ti ions 3+ , the Ti report 3+ / Ti 2+ being between 20 / 80 and 70 / 30, the mixture of molten salts comprising at least MgC, and optionally PuCIs,

[0059] - immersing an anode, a cathode and a reference electrode in the solution; the anode and the cathode being made of the same material, preferably tungsten, and the reference electrode being, for example, made of molybdenum or preferably tungsten,

[0060] - measure the potential Ei between the cathode and the reference at zero or almost zero current (typically the potential is less than 50mV, for example of the order of a few mV),

[0061] - impose a current between the anode and the cathode, to deposit metallic magnesium on the cathode, or to deposit metallic plutonium on the cathode if necessary, - measure the potential E2 between the cathode and the reference at zero current, whereby the value of the Mg couple is obtained 2+ / Mg° or Pu 3+ / Pu°, if applicable,

[0062] - calculate the potential of solution E soi according to E soi = E2-E1.

[0063] When the value of Esol differs from a reference value, titanium can be added to the solution to have a Ti concentration ratio 3+ / Ti 2+ between 20 / 80 and 70 / 30.

[0064] This measurement method is particularly advantageous because it allows the formation in situ of a so-called dynamic (or ephemeral) reference electrode and thus the reliable measurement of the potential of the solution.

[0065] The invention also relates to a method for controlling a solution of a device as defined above, the control method comprising the following successive steps:

[0066] - provide the device comprising the element configured to contain or circulate the solution, the element having a wall made of steel or a nickel-based alloy comprising chromium, and the solution, in contact with the wall, comprising a mixture of molten chloride salts, Ti ions 2+ and Ti ions 3+ ,

[0067] - take a sample of the solution and analyze it to determine the presence or absence of iron ions and / or chromium ions,

[0068] - if iron ions and / or chromium ions are present, immerse titanium metal in the solution for a sufficient time for the solution to contain Ti ions 2+ and Ti ions 3+ , according to a Ti report 3+ / Ti 2+ between 20 / 80 and 70 / 30.

[0069] Iron ions and / or chromium ions present in solution come from corrosion of the element wall. They are present when the concentration ratio of Ti ions 2+ and Ti ions 3+ is such that it no longer corresponds to the immunity domain (i.e. it no longer corresponds to the potential domain where redox control is effective). It is then necessary to immerse titanium metal in the solution for cathodic protection.

[0070] Advantageously, the formation of Ti ions 2+ and Ti ions 3+from titanium metal is accelerated by the presence of iron ions and / or chromium ions in solution. Other characteristics and advantages of the invention will emerge from the additional description which follows.

[0071] It goes without saying that this additional description is given only as an illustration of the subject of the invention and must in no case be interpreted as a limitation of this subject.

[0072] BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The present invention will be better understood by reading the description of exemplary embodiments given purely for informational purposes and in no way limiting, with reference to the appended drawings in which:

[0074] The figure schematically represents in section a device according to a particular embodiment of the invention.

[0075] Figure 2 is a graph representing the dissolution voltammograms of pure metals, according to particular embodiments of the invention.

[0076] Figure 3 is a voltammogram of a NaCI-MgC -CeCb solution containing a titanium salt at 600°C, according to a particular embodiment of the invention.

[0077] Figure 4 is a graph representing the evolution of the equilibrium potential of a NaCI-MgC -CeCb solution at 600°C as a function of the immersion time of a metallic titanium plate.

[0078] Figure 5 is a graph representing the equilibrium potential of a NaCI-MgC -CeCb solution at 600°C after removing a titanium plate, according to a particular embodiment of the invention.

[0079] Figure 6 is a graph of square wave voltammetry in a 600°C NaCl-MgCb-CeCb solution containing Ti ions 3+ and Ti 2+, according to a particular embodiment of the invention.

[0080] Figure 7 represents the potential of a NaCl-MgCb-CeC solution at

[0081] 600°C depending on the Ti concentration ratio 3+ and Ti 2+ , according to a particular embodiment of the invention. Figure 8 is a snapshot obtained using a scanning electron microscope of a sample of 316L steel after immersion in a NaCI-MgC -CeCIs solution at 600°C for 168 hours, without redox control.

[0082] Figure 9 is a scanning electron microscope image of a 316L steel sample after immersion in a NaCI-MgC-CeCIs solution containing Ti ions. 3+ and Ti 2+ having a potential of approximately -2V vs C / CI", at a temperature of 600°C for 168h, according to a particular embodiment of the invention.

[0083] Figure 10 is an EDX analysis performed on a 316L steel sample after immersion in a NaCl-MgC - CeCls solution containing Ti ions 3+ and Ti 2+ having a potential of approximately -2V vs C / CI", at a temperature of 600°C for 168h, according to a particular embodiment of the invention.

[0084] Figure 11 is a scanning electron microscope image of an Inconel® 625 sample after immersion in a NaCI-MgC-CeCIs solution containing Ti ions. 3+ and Ti 2+ having a potential of approximately -2V vs CI2 / CE, at a temperature of 600°C for 168h, according to a particular embodiment of the invention.

[0085] Figure 12 is an EDX analysis performed on an Inconel® 625 sample after immersion in a NaCl-MgC - CeCls solution containing Ti ions 3+ and Ti 2+having a potential of approximately -2V vs CI2 / CE, at a temperature of 600°C for 168h, according to a particular embodiment of the invention.

[0086] Figure 13 is a graph representing the potential of a NaCI-MgCI2-CeCI3 solution at 600°C.

[0087] Figure 14 is a graph representing the potential of a NaCI-MgC -CeCIs solution containing Ti ions 3+ and Ti 2+ at 600°C.

[0088] The various parts shown in the figures are not necessarily on a uniform scale, in order to make the figures more readable. DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS

[0089] Although this is in no way limiting, the invention particularly finds applications in the field of nuclear power or solar energy (concentrated solar thermal).

[0090] The invention is applicable to any device 10 comprising:

[0091] - an element 20 configured to be in contact with a solution 30 and having a wall 21 made of steel or a nickel-based alloy comprising chromium, and

[0092] - a solution 30 of molten chlorides, in contact with the wall 21.

[0093] In particular, the element 20 can be used to contain the solution 30 of molten salts. The device 10 can be, for example, a tank, a reservoir or an enclosure.

[0094] The element 20 can also be used to circulate the solution 30 of molten salts. The element 20 can be, for example, a tube.

[0095] In particular, the element 20 may be a primary circuit of a molten chloride nuclear fission reactor, a secondary circuit of a nuclear fission reactor comprising molten chlorides or even a molten chloride reservoir or a molten chloride circulation circuit of a concentrated solar power plant.

[0096] The element 20 has at least one wall 21. It may have several. For example, in the case of a tube, it is a side wall. In the case of a tank, a vat or an enclosure, the element may comprise a side wall, a bottom wall (or base) and a top wall (or cover). Each of these walls may be in contact with the solution 30.

[0097] The material forming the wall (steel or nickel-based alloy) comprises chromium. It also comprises iron. It may also comprise one or more other elements. By way of non-limiting example, the other element(s) may be chosen from nickel, molybdenum, tungsten, silicon, cobalt, aluminium, titanium, niobium and manganese.

[0098] Chromium is the least stable element in the metal alloy, in other words, it is the most negative element. Then come iron, nickel, and molybdenum. Tungsten is the material that resists oxidation the most; its anodic limit is the oxidation of chlorides to chlorine, and its cathodic limit is the reduction of Mg. 2+ in magnesium metal. There is a potential range between the reduction potential of Mg 2+ in magnesium metal and the oxidation potential of the most negative element (i.e. here chromium) for which it is possible to limit the dissolution of steels.

[0099] According to a first embodiment, the wall 21 is made of stainless steel. It may, for example, be 316L steel.

[0100] According to another embodiment, the wall 21 is made of a superalloy. For example, a nickel-chromium superalloy will be chosen, and preferably a nickel-chromium-molybdenum superalloy. In particular, the superalloy NiCr22Mo9Nb, marketed under the name Inconel® 625 by the company Special Metals Corporation, may be chosen.

[0101] The molten salt solution 30 (also called molten salt bath) is in contact with the wall 21 of the element 20.

[0102] Solution 30 comprises a mixture of molten chlorides. This may be, for example, a binary mixture or a ternary mixture of alkali and alkaline earth metal chlorides, generally used as a heat transfer fluid, or a mixture containing actinide chlorides generally used for nuclear applications

[0103] The molten salts to form the mixture of the first type are preferably chosen from LiCI, NaCI, KCI, MgC, CaC and BaC. For a binary mixture, one could choose, for example, a NaCI-MgCb mixture or a KCI-MgCb mixture, while for a ternary mixture one could choose, for example, a NaCI-KCI-MgCb mixture.

[0104] In the second case, for application in an RSF, one or more compounds of formula AcC with Ac an actinide will preferably be added to one of the binary mixtures described above. Preferably, Ac is Pu, Am, U or Cm. Even more preferably, Pu, Cm or Am will be chosen, or even more preferably Pu or Am. For example, a NaCI-MgCb-PuC mixture or a NaCI-MgCb-AmC mixture or a NaCI-MgCb-PuC -AmC mixture may be chosen.

[0105] As a chemical analogue of NaCI-MgCb-PuCb, it is possible to use the mixture NaCI-MgCb-CeC . The compound AcCb preferably represents between 4.5% and 20% mol of the molten salt mixture.

[0106] For example, a mixture of NaCI-MgCb-CeCb (57 / 30 / 13 or 61 / 34 / 5) can be used as a simulant for the actinide converter (NaCI-MgC -PuCb). The ratios are molar ratios given relative to the solution.

[0107] Solution 30 further comprises Ti ions 2+ and Ti ions 3+ . The choice of the ratio of Ti ion concentrations 2+ and Ti ions 3+ is essential to protect the wall material 21 in contact with the solution 30. The Ti ratio 3+ / Ti 2+ is between 20 / 80 and 70 / 30, for example between 30 / 70 and 70 / 30 or between 20 / 80 and 50 / 50, and preferably between 30 / 70 and 50 / 50.

[0108] The potential of solution 30 depends on both the ratio Ti 3+ / Ti 2+ and temperature according to Nernst's law (equation (1)):

[0109] With :

[0110] E the potential in Volts,

[0111] E° the standard potential of the couple in Volts,

[0112] R the constancy of ideal gases,

[0113] T the temperature in Kelvin, n the number of electrons exchanged,

[0114] F the Faraday constant, aTiC the chemical activity of the oxidant, aTiCb the chemical activity of the reducer.

[0115] Preferably, in the solution, the titanium concentration is 0.1 to 3 mol% relative to the total concentration of the solution.

[0116] During operation of the device 10, the solution 30 is preferably at a temperature between 450°C and 700°C and preferably between 550°C and 650°C.

[0117] For example, for a temperature of 600°C, with a ratio Ti 3+ / Ti 2+ of 20 / 80, the potential of solution 30 is about -2V vs Cb / Ch Such a potential is sufficiently negative to prevent the dissolution of chromium observed around -1.75V vs Ch / Cl'. Such a potential also prevents the dissolution of other elements such as iron, nickel, molybdenum, for example, which have potentials higher than that of chromium.

[0118] For a temperature of 600°C, with a Ti ratio 3+ / Ti 2+ of 50 / 50, the potential of solution 30 is approximately -1.9V vs Ch / CI', which is also sufficient to prevent the dissolution of the chromium present in the metal alloy of the wall.

[0119] With a ratio of 70 / 30, the calculated potential of the solution is -1.83V vs Ch / CI' at 600°C, which is also sufficient to protect the chromium.

[0120] So, for higher temperatures (for example for a temperature of 700°C), the potentials will be higher.

[0121] It is possible to use this device 10 in the presence of an inert atmosphere (argon or helium in particular).

[0122] It is also advantageous to control the purity of the inputs associated with the sealing of the heat transfer circuit.

[0123] Solution 30, and in particular the mixture of molten salts, are first purified before being used.

[0124] The method for preparing a device 10 as defined previously comprises the following steps:

[0125] - providing a device 10 comprising a wall 21 made of steel or a nickel-based alloy comprising chromium, the wall 21 being in contact with a solution 30 comprising a mixture of molten chloride salts,

[0126] - immerse a piece of titanium metal in the solution, for a sufficient time to obtain Ti ions 2+ and Ti ions 3+ , the Ti report 3+ / Ti 2+ being between 20 / 80 and 70 / 30,

[0127] - possibly, add to the solution ions of a metal having a potential lower than the dissolution of chromium from the metal alloy, whereby the formation of Ti ions is accelerated 2+ and Ti ions 3+ .

[0128] The addition of titanium metal to a solution comprising a mixture of molten chloride salts leads to the oxidation of titanium and the formation of Ti ions 2+ and Ti 3+. Indeed, for example, the solution of molten chlorides NaCI-MgC -CeCh or NaCI- MgC -PuCIs in the absence of titanium has an equilibrium potential between -1.2V and -IV vs Ch / Ch Thus, when titanium is added, it is oxidized and the equilibrium potential of the solution decreases until it stabilizes around -2V vs Ch / CI' for a Ti ratio 3+ / Ti 2+ of 20 / 80.

[0129] Adding ions to solution accelerates the formation of Ti ions 2+ and Ti ions 3+ via oxidation-reduction reactions. Any element with a potential lower than the dissolution of chromium from the metal alloy can be used. Preferably, chromium, aluminum, or zirconium can be used.

[0130] The duration for which the titanium part is immersed in the solution depends on several parameters, including the surface area of ​​the immersed part of the titanium part, the volume of solution, agitation, circulation speeds, etc. The person skilled in the art will choose the duration according to the sizing of the device and in such a way as to have the Ti ratio 3+ / Ti 2+ sought. For example, a sufficient duration might be a duration of 1h to 10h.

[0131] We will now describe in more detail a method for measuring the Ti ratio 3+ / Ti 2+ in a device 10 as defined previously.

[0132] The measurement process includes the following steps:

[0133] - provide a device 10 comprising a wall 21 made of steel or a nickel-based alloy comprising chromium, the wall 21 being in contact with a solution 30 comprising a mixture of molten chloride salts, Ti ions 2+and Ti ions 3+ ,

[0134] - perform square wave voltammetry in the solution to obtain reduction and oxidation curves representing the differential current as a function of the potential,

[0135] - determine the proportion of Ti ions 2+ and Ti ions 3+ , by integrating the area under the reduction and oxidation curves.

[0136] Knowing the Ti ratio 3+ / Ti 2+ , and from Nernst's law, it is possible to calculate the potential of solution 30. The apparent standard potential integrating the activity coefficients was evaluated at -1.89V vs CI / CI2. This measurement method is particularly advantageous, because from the shape of the curves, it is possible to quickly and simply determine in which case we are. Indeed, when the oxidation curve shows an almost complete Gaussian, the Ti content 2+is well above 50%. At a 50 / 50 ratio, the oxidation curve starts at the peak and only shows the right-hand part of the Gaussian. The same is true for the reduction part.

[0137] We will now describe a method for measuring the potential of a solution of molten salts.

[0138] The measuring method comprises the following successive steps: a) immersing an anode, a cathode and a reference electrode in the solution 30 comprising a mixture of molten chloride salts, and possibly Ti ions 2+ and Ti ions 3+ , the mixture of molten salts comprising at least MgCh, b) measuring the potential Ei between the cathode and the reference at zero or almost zero current, c) imposing a current between the anode and the cathode, to deposit metallic magnesium on the cathode, d) measuring the potential E2 between the reference and the cathode at zero current, whereby the value of the couple Mg is obtained 2+ / Mg°, e) calculate the potential of solution E soi according to E soi = E2-E1.

[0139] The electrodes are advantageously made of the same material which does not oxidize in the solution. The electrodes (anodes and cathodes) are advantageously made of tungsten.

[0140] The reference electrode is advantageously a metal wire that does not oxidize in the solution: this electrode is called a dynamic electrode or ephemeral electrode. Preferably, a molybdenum reference electrode or even more preferably a tungsten electrode will be chosen.

[0141] In step b), the potential between the cathode and the reference is measured in solution 30 at zero current. The potential is almost zero (typically less than 50 mV, especially less than 5 mV) since the electrodes are made of the same metallic conductor. Before step c), it is also possible to measure the potential between the anode and the reference at zero current. The potential will also be almost zero.

[0142] In step c), a constant current pulse is applied, for example for a duration between 5 s and 20 s, for example 10 s. During this step, the cathode potential is fixed on that of the reduction of Mg 2+ in magnesium metal, while that of the anode evolves towards the oxidation potential of chlorides into gaseous chlorine. Preferably, the anode / cathode surface ratio is between 1 and 3, ideally 2, and Jcat is between 0.05 A / cm 2 and 0.8 A / cm 2 .

[0143] In step d), the anode and cathode potentials are measured in solution 30 at zero current relative to the reference. The cathode potential temporarily takes the value of the electrochemical couple Mg 2+ / Mg°.

[0144] The cathode potential after the pulse does not depend on the applied current. According to Nernst's law, it only depends on the Mg concentration. 2+ . However, due to the high concentration of Mg 2+ in the solution (logarithmic term), the cathode potential is also little dependent on variations in the Mg concentration 2+ .

[0145] The time during which the potential of the Mg couple 2+ / Mg° can be observed is dependent on the applied current: the greater the current, the greater the quantity of Mg metal deposited and the longer the potential of the Mg couple 2+ / Mg° can be used as a reference value.

[0146] In step e), the potential of solution 30 is obtained by comparing the values ​​of the potentials between step b) and step d) when the magnesium is still present on the cathode.

[0147] In the case of a ternary mixture of molten salts including PuCls, plutonium will be formed instead of magnesium. In step c), metallic plutonium will be deposited on the cathode and in step d), the value of the Pu couple is obtained 3+ / Pu°. illustrative and non-limiting of an embodiment: Study of the anodic dissolution of chromium, iron and nickel:

[0148] The anodic dissolution of chromium, iron and nickel was first studied in a NaCI-MgC-CeCIs (57 / 30 / 13) solution at 600°C. This salt is used as a simulant for the actinide converter (NaCI-MgC-PuCIs).

[0149] The voltammograms obtained for these metals are presented in Figure 2. Several remarks can be made by comparing the intensity-potential plots obtained with the different metals:

[0150] - tungsten is the material that best resists oxidation, its anodic limit is the oxidation of chlorides to dichlorine and its cathodic limit is the reduction of Mg 2+ in magnesium metal,

[0151] - the dissolution potentials of chromium, iron and nickel confirm that chromium is the least stable element (most negative potential) followed by iron and nickel, which is in agreement with the sequence predicted by the thermodynamics of pure bodies,

[0152] - there is a wide potential range (noted Z in figure 2) between - 1.75 and -2.75 V vs E(Cl2 / Cl) where effective redox control can be carried out to limit the dissolution of steels.

[0153] The difficulty is therefore not only to be able to obtain this potential domain in a highly corrosive solution and generally operating at high temperature (typically above 450°C, or even above 550°C and even more preferably between 600°C and 700°C), but also to be able to control and maintain it.

[0154] Study of the Ti buffer electrochemical couple 3+ / Ti 2+ :

[0155] Titanium was studied in a molten salt solution NaCI-MgC -CeCIs at 600°C.

[0156] Titanium can, depending on the potential of the solution, be metallic, or dissolved in the salt in the form of divalent, trivalent and tetravalent complexes (Figure 3).

[0157] The wave observed before the dichlorine wall corresponds to the oxidation of Ti 3+ in Ti 4+ The reduction wave towards -1.6 V corresponds to the reduction of Ti 3+ in Ti 2+ . The peak reduction of Ti2+ in metallic titanium is observed around -2.35 V, it is associated with its fine oxidation peak characteristic of the dissolution of the electrodeposited metal.

[0158] The equilibrium potential of a solution of molten alkali and alkaline earth chlorides in the absence of titanium is between -1 and -1.2 V.

[0159] In the presence of Ti metal, Ti 3+ is reduced to Ti 2+ .

[0160] Ti ions 3+ and Ti 2+ are stable and then coexist in the solution according to the equilibrium:

[0161] 2 Ti 3+ + Ti° <-> 3 Ti 2+

[0162] In this case, the potential control is ensured by the Ti couple 3+ / You 2+ .

[0163] The presence of metallic titanium leads to a decrease in the potential of the solution to a stable value around -2 V (figure 4).

[0164] Once the reduction of Ti 3+ in Ti 2+obtained, the titanium metal plate is removed from the solution and the equilibrium potential of the solution is measured as a function of time: it rises very slowly and linearly until reaching its original value before the reduction phase after several hundred hours (figure 5).

[0165] Nernst's law was then verified using the electroanalytical technique of square wave voltammetry (SWV) to determine the proportions of Ti 3+ and Ti 2+ as a function of time in the molten salt solution. The proportion of the two species is estimated by integrating the areas under the Ti reduction curves 3+ in Ti 2+ and oxidation of Ti 2+ in Ti 3+ measured regularly over time after removal of the titanium plate (Figure 6). The curve with crosses in Figure 6 represents a solution where titanium is present at more than 90% in the Ti form2+ , while the other curve shows a 50 / 50% mixture of Ti 2+ and Ti 3+ .

[0166] The evolution of the potential as a function of the logarithm of the ratio of Ti concentrations 3+ and Ti 2+ obtained by SWV is linear in accordance with Nernst's law (figure 7). The experimental slope (RT / nF) is 0.078, i.e. an electron number of 0.97, corresponding well to the Ti electronic transition 3+ / Ti 2+ . Verification of Nernst's law allows us to conclude that the potential of the solution is effectively controlled by the joint presence in the solution of molten salts of divalent and trivalent titanium. The most negative potential corresponds to a ratio Ti 3+ / Ti 2+ of 20% / 80%, the highest potential a ratio of 70% / 30%.

[0167] Immersion of 316L steel in NaCI-MgCh-CeCh without redox control for one week:

[0168] The immersion of a 316L steel plate was carried out in NaCI-MgC - CeCIs without redox control of the solution potential for one week (i.e. 168 hours) at 600°C. The plate is held in the solution by a molybdenum wire, the molybdenum-material contact is kept out of the molten salt. The tests are carried out in an inert glove box under argon, the molten salt is in an alumina crucible, itself placed inside a sealed quartz cell swept by high purity argon (Alphagaz2).

[0169] After one week of immersion, the submerged part of the sample is cut and analyzed by SEM-EDX (figure 8). The analysis of the section shows a classic corrosion profile in these environments:

[0170] - Intergranular corrosion,

[0171] - Attack depth of approximately 100pm,

[0172] - Beyond this depth the mass composition of 316 steel is found.

[0173] The corresponding EDX analysis confirms the loss of Cr and Fe.

[0174] Immersion of Inconel 625® in NaCI-MgCb-CeCh without redox control for one week:

[0175] The immersion of an Inconel 625® plate was carried out in NaCI- MgC -CeCIs without redox control of the solution potential for one week (i.e. 168 hours) at 600°C as in the previous example with 316L steel.

[0176] After a week of immersion, the submerged part of the sample was cut and analyzed by SEM-EDX. Six different sites were observed on the sample. All showed uniform corrosion at a depth of between 2 and 5 pm. Inconel 625 offers much greater corrosion resistance than 316 steel.

[0177] EDX analysis of the immersed and corroded part shows a loss of chromium and a surface enrichment in Ni and Mo. Immersion of Inconel 625® in NaCI-MgCb-CeCh without redox control for three weeks:

[0178] The immersion of an Inconel 625® plate was carried out in NaCI- MgC -CeCIs without redox control of the solution potential for three weeks (i.e. 504 hours) at 600°C.

[0179] After immersion for three weeks, the submerged part of the sample is cut and analyzed by SEM-EDX. A uniform corrosion with a thickness of approximately 8 μm is observed, confirming that the corroded thickness increases between 1 and 3 weeks of immersion. EDX analysis of the submerged and corroded part confirms that there is a loss of chromium and a surface enrichment in Ni and Mo.

[0180] Immersion of 316L steel and Inconel® 625 in NaCI-MgCb-CeCh with redox control by the Ti couple 3+ / Ti 2+for a week:

[0181] Immersions of 316L steel plates or Inconel® 625 were carried out in NaCI-MgC -CeCh containing the Ti couple 3+ / Ti 2+ at 600°C for one week (168 hours) and one month (720 hours). The potential of the solution was previously adjusted by immersing a metallic titanium plate in order to obtain the titanium in solution mainly in the Ti form 2+ (~80% of Ti 2+ and 20% Ti 3+ ). The obtained potential of about -2 V vs CI2 / CI ) is sufficiently negative to prevent the dissolution of chromium observed around -1.75 V vs CI2 / CE.

[0182] The SEM image of the 316 L sample after immersion for 168 hours with redox control reveals that the 316L steel was protected during this test (Figure 9). The presence of deposits is observed on the surface of the sample.

[0183] EDX analysis indicates an attack of the material over 1 to 2 pm at the molten salt / metal interface (Figure 10). This layer is rich in Ti. Molybdenum is also found, certainly coming from contamination of the measuring electrodes.

[0184] Observations of the Inconel® 625 sample after immersion are similar to those made for 316L steel. The coupon is not corroded or only slightly corroded (Figure 11). EDX analysis (Figure 12) also shows a deposit of titanium and aluminum (contamination from the alumina crucible) alloyed with nickel. The thickness of the layer is slightly greater (~3-4 pm) than that observed on 316 steel (~2 pm).

[0185] Under the reducing conditions imposed by the Ti redox buffer 3+ / Ti 2+ , the deposition of metals on the surface seems to be facilitated with Inconel®. Indeed, this material being richer in nickel can form numerous alloys in these temperature ranges.

[0186] Immersion of 316L and Inconel® 625 in NaCI-MgCb-CeCh with redox control by the Ti couple 3+ / Ti 2+ for one month (720 hours):

[0187] A 316L steel plate was immersed at 600°C for one month. The sample was cut to characterize an immersed part and a part positioned above the salt. As before, the sample is little attacked. A thin layer of Ti at the extreme periphery of the sample can be observed. Comparison of the immersed and non-immersed parts of the sample does not show more pronounced corrosion in the immersed part and allows us to rule out uniform corrosion of the sample in the part in contact with the solution.

[0188] EDX analysis of the sample periphery confirms that, as before, the sample is attacked over ~2 pm with a loss of chromium, iron and nickel and an enrichment in titanium. This corrosion thickness, identical for the one-week and one-month immersions, seems to show that the corrosion layer does not grow over time.

[0189] After immersion of the Inconel® 625 plate at 600°C for one month, the sample is little corroded and a layer of Ti alloyed with nickel in the Inconel® 625 samples forms at the periphery of the sample. The thickness of the corrosion layer is about 3 μm as in the one-week test, which tends to show that the corrosion progresses little over time, as for 316 steel.

[0190] Measuring the potential of the solution:

[0191] Corrosion control based on the use of a suitable buffer couple is effective but requires a method for measuring the chemical potential of the solution. An electrochemical technique has been developed to monitor the chemical potential of the solution. It is based on the principle of a so-called dynamic reference electrode. The method consists of creating a metallic deposit on an inert electrical conductor (tungsten or molybdenum for example) using a current pulse. In a salt containing a large proportion of MgCI2 such as NaCI-MgCI2, KCI-MgCI2, NaCI-KCI-MgC, the passage of current leads to the production of metallic magnesium on the cathode. When the current is stopped, the cathode potential takes that of Nernst's law for the Mg couple 2+ / Mg° and then constitutes a reference electrode generated in situ in the saline environment. The potential value of this reference electrode is stable long enough to serve as a reference electrode, the value is also particularly reproducible due to the high MgCI2 content.

[0192] To perform the measurement, a three-electrode setup is used. The cathode and anode are made of tungsten and the quasi-reference electrode is made of molybdenum or preferably tungsten.

[0193] The method includes three distinct phases:

[0194] - phase 1: Measurement of the potential at zero current between the two electrodes (anode vs. reference and cathode vs. reference),

[0195] - phase 2: Imposition of a current over a short period (< 10 s)

[0196] - phase 3: Measurement of the potential at zero current between the two electrodes (anode vs. reference and cathode vs. reference).

[0197] These three phases are illustrated in Figure 13. The solution used is a NaCl-MgC -CeCls solution at 600 °C. Ea represents the anodic potential and Ec represents the cathodic potential. During phase 1 (marked 1 on the graph), the anode and cathode potentials are measured at zero current relative to the reference, the values ​​are close to zero (< 50 mV). During phase 2 (marked 2 on the graph), a 10s constant current pulse is applied, the cathode potential is fixed on that of the reduction of Mg2+ to magnesium metal, while that of the anode will be fixed on the oxidation of chlorides to gaseous chlorine. Several currents (20 mA, 40 mA and 60 mA) were tested. During phase 3 (marked 3 on the graph), the anode and cathode potentials are measured at zero current. The cathode potential temporarily takes the value of the electrochemical couple Mg 2+ / Mg° before returning to its original value. The anode potential after the current stops quickly returns to its starting potential.

[0198] The cathode potential after the pulse does not depend on the applied current as shown in Figure 13, it depends according to Nernst's law only on the Mg concentration 2+ , it is also little dependent on variations in Mg concentration 2+ due to its high concentration in the solution (logarithmic term). The time during which the potential of the Mg couple 2+ / Mg° can be observed is dependent on the applied current, the greater the current, the greater the quantity of Mg metal deposited and the longer the potential of the Mg couple 2+ / Mg° can be used as a reference value.

[0199] The measurement of the potential of the solution is obtained by comparing the values ​​of the potentials between phase 1 (close to 0) and phase 2 (when the Mg couple 2+ / Mg° is stable). In the case shown in Figure 13, the potential measurement is about -1.5 V, this indicates that the potential of the tungsten electrode in the molten salt is 1.5 V higher than that of the dynamic reference electrode Mg 2+ / Mg°.

[0200] This method was repeated in a NaCI-MgC -CeCIs solution including the Ti buffer couple 3+ / Ti 2+ at 600 °C (figure 14). The characteristics of the curves obtained are identical to those of figure 13, the major difference lies in the value of the measurement of the potential of the molten salt solution: the potential of the molten salts containing titanium is no longer higher than about 0.5 V than that of the dynamic Mg electrode 2+ / Mg°. The presence of titanium in the solution therefore lowers its potential by approximately 1 V, thus bringing the potential of the solution and therefore of the electrically conductive materials found there into an immunity range which greatly limits the corrosion of metals.

[0201] The oxidizing character of molten salt baths tends to convert over time the reduced species of the redox buffer couple Ti 2+ in Ti 3+ . As a result, the salt potential according to Nernst's law tends to increase over time. The proposed measurement allows detecting this increase in potential. The potential of the solution can be lowered within a few hours by soaking metallic titanium in the solution in order to convert the Ti 3+ in Ti 2+ .

Claims

DEMANDS 1. Device (10) comprising: - an element (20) configured to contain or circulate a solution, the element (20) having a wall (21) made of steel or a nickel-based alloy comprising chromium, - a solution (30), in contact with the wall (21), comprising a mixture of molten chloride salts, characterized in that the solution further comprises Ti ions 2+ and Ti ions 3+ , the concentration ratio Ti 3+ / Ti 2+ being between 20 / 80 and 70 / 30.

2. Device according to claim 1, characterized in that the mixture of molten chloride salts is a binary or ternary mixture of alkali and alkaline earth metal chlorides or a binary mixture of alkali and alkaline earth metal chlorides and one or more actinide chlorides.

3. Device according to any one of claims 1 and 2, characterized in that the element (20) is a primary circuit of a nuclear fission reactor or a secondary circuit of a nuclear fission reactor.

4. Device according to any one of claims 1 and 2, characterized in that the element (20) is a reservoir or fluidic circuit of a concentrating solar power plant.

5. Device according to any one of claims 1 to 4, characterized in that the element (20) is a secondary circuit of a nuclear fission reactor, a reservoir or a fluidic circuit of a concentrating solar power plant and in that the mixture of molten chloride salts is a binary or ternary mixture obtained from molten chlorides selected from LiCl, NaCl, KCl, MgC, CaC and BaC.

6. Device according to claim 5, characterized in that the mixture of molten chloride salts is a NaCl-MgC, KCl-MgC or NaCl-KCl-MgC mixture.

7. Device according to claim 3, characterized in that the element (20) is a primary circuit of a nuclear fission reactor and in that the mixture of molten chloride salts comprises at least one actinide chloride.

8. Device according to claim 7, characterized in that the mixture of molten chloride salts is a NaCl-MgC-PuCls or NaCl-MgC-PuCls-AmCh mixture.

9. Device according to any one of the preceding claims, characterized in that the solution (30) is at a temperature between 450°C and 700°C and preferably between 550°C and 650°C.

10. Device according to any one of the preceding claims, characterized in that the wall (21) is made of stainless steel, for example 316L steel, or of a nickel-chromium-molybdenum alloy, for example NiCr22Mo9Nb.

11. Device according to any one of the preceding claims, characterized in that the ratio Ti 3+ / Ti 2+is between 30 / 70 and 50 / 50.

12. Solution (30) comprising a mixture of molten chloride salts, intended to be in contact with a wall of an element such as a primary circuit of a nuclear fission reactor, a secondary circuit of a nuclear fission reactor, a tank of a concentrating solar power plant or a fluidic circuit of a concentrating solar power plant, the wall being made of steel or a nickel-based alloy comprising chromium, characterized in that the solution further comprises Ti ions 2+ and Ti ions 3+ , the concentration ratio Ti 3+ / Ti 2+ being between 20 / 80 and 70 / 30.

13. Solution according to claim 12, characterized in that the mixture of molten chloride salts is a binary or ternary mixture of alkali and alkaline earth metal chlorides or a binary mixture of alkali and alkaline earth metal chlorides and one or more actinide chlorides.

14. A method for preparing a device (10) as defined in any one of claims 1 to 11, comprising the following steps: - provide a device (10) comprising an element (20) configured to contain or circulate a solution, the element (21) having a wall (21) of steel or of a nickel-based alloy comprising chromium, and a solution (30), in contact with the wall (21), comprising a mixture of molten chloride salts, - Immerse titanium metal in solution (30) for a sufficient time to obtain Ti ions 2+ and Ti ions 3+ , the Ti report 3+ / Ti 2+ being between 20 / 80 and 70 / 30, - optionally, add to solution (30) ions of a metal having a potential lower than the dissolution potential of the chromium in the metal alloy, thereby accelerating the formation of Ti ions 2+ and Ti ions 3+ .

15. Method for measuring the potential of a solution (30) of a device (10) as defined in any one of claims 1 to 11, the measurement method comprising the following successive steps: - to provide a device (10) comprising an element (20) configured to contain or circulate a solution (30), the element (21) having a wall (21) of steel or of a nickel-based alloy comprising chromium, and a solution (30) in contact with the wall (21), the solution (30) comprising a mixture of molten chloride salts, Ti ions 2+ and Ti ions 3+ , the Ti report 3+ / Ti 2+ being between 20 / 80 and 70 / 30, the mixture of molten salts comprising at least MgC, and possibly PuCls, - immersing an anode, a cathode and a reference electrode in solution (30), the anode, the cathode and the reference electrode being made of the same material, advantageously tungsten, - measure the potential Ei between the cathode and the zero or near-zero current reference, - to impose a current between the anode and the cathode, to deposit metallic magnesium on the cathode, or to deposit metallic plutonium on the cathode as appropriate, - measure the potential E2 between the cathode and the zero-current reference, thereby obtaining the value of the Mg couple 2+ / Mg° or Pu 3+ / Pu°, if applicable, - calculate the potential of solution E soi according to E soi = E2-E1.

16. A method for testing a solution (30) of a device (10) as defined in any one of claims 1 to 11, the testing method comprising the following successive steps: - to provide the device (10) comprising the element (20) configured to contain or circulate the solution (30), the element (20) having a wall (21) of steel or of a nickel-based alloy comprising chromium, and the solution (30), in contact with the wall (21), comprising a mixture of molten chloride salts, Ti ions 2+ and Ti ions 3+ , - take a sample of the solution (30) and analyze it to determine the presence or absence of iron ions and / or chromium ions, - In the presence of iron ions and / or chromium ions, immerse titanium metal in solution (30) for a sufficient time so that solution (30) contains Ti ions 2+ and Ti ions 3+ , according to a Ti report 3+ / Ti 2+ between 20 / 80 and

Images