JP2006274298A - Diplophase alloy for separating/refining hydrogen and manufacturing method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for manufacturing such a diplophase alloy for separating/refining hydrogen as to have high rolling-formability and a high hydrogen-permeation coefficient. <P>SOLUTION: This diplophase alloy is a Nb-Ti-Ni alloy formed of a diplophase comprising a phase having hydrogen permeability and a phase having hydrogen embrittlement resistance. The method for manufacturing the diplophase alloy includes heat-treating the Nb-Ti-Ni alloy at a temperature higher than 1,000°C for 100 hours or longer. The manufacturing method also includes plastic-working the Nb-Ti-Ni alloy and subsequently heat-treating it at a temperature higher than 1,000°C. Thus manufactured diplophase alloy can obtain the hydrogen permeation coefficient almost equal to that of the as-cast alloy. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a method for producing a hydrogen separation / purification double-phase alloy used for producing high-purity hydrogen, and particularly to a hydrogen separation / purification double-phase alloy comprising a novel alloy composition of Nb-Ti-Ni system. It relates to the manufacturing method.

Since hydrogen, which is a fuel for fuel cells, does not exist in nature alone, it must be artificially produced. Ideally, hydrogen is produced by decomposing water using electricity made from renewable natural energy such as solar heat, but the current technical level is expensive and difficult to apply. For the time being, it is considered realistic to produce hydrogen by steam reforming natural gas (methane) or the like. Steam reforming is to obtain hydrogen using the following chemical reaction. If only H ₂ is removed in this reaction, the equilibrium shifts to the right side according to Le Chatelier's law, a higher conversion can be obtained, and energy loss can be suppressed by lowering the reaction temperature, resulting in hydrogen production costs. Can be reduced. A hydrogen permeable alloy membrane is used to selectively remove hydrogen. Only hydrogen passes through the hydrogen permeable alloy membrane.

  Hydrogen penetrates into the metal crystal lattice in the form of interstitial impurities, and has a feature that its diffusion rate is ten orders of magnitude higher than that of nitrogen, carbon and oxygen. The hydrogen permeable alloy film utilizes such characteristics of hydrogen. In order to separate and purify the impure hydrogen, the impure hydrogen on the supply side is set to a high pressure across the hydrogen permeable alloy membrane, while the pressure collecting side is set to a low pressure to generate a pressure difference. Hydrogen molecules on the high pressure side dissociate into atomic hydrogen on the surface of the alloy film and dissolve in the metal. Hydrogen atoms are diffused from the high pressure side to the low pressure side using the hydrogen concentration gradient in the metal film caused by the pressure difference as a driving force, recombine on the low pressure side and become hydrogen molecules and flow to the low pressure side. At this time, impurity gases other than hydrogen cannot be dissociated in an atomic form on the high pressure side, and cannot be transmitted to the low pressure side because the diffusion speed in the metal is much slower than that of hydrogen atoms. According to the method using a metal film, theoretically, hydrogen having a purity of 100% can be obtained, and hydrogen having a purity of 99.999999% can be actually obtained. Currently, the hydrogen permeable alloy film in practical use is an alloy based on Pd. However, since Pd is a very rare and expensive metal, development of an inexpensive alloy to replace it is required.

As another system, for example, Nb-Ni system as described in Patent Document 1, Nb- (Ni, Co, Mo)-(V, Ti, Zr, Ta, Hf) based hydrogen permeable alloys have been studied.
JP 2001-170460 A (0026) JP 2004-42017 A (0007 to 0011, Table 1)

  Hydrogen permeable alloys are required to have a large hydrogen permeability coefficient and high hydrogen embrittlement resistance. Here, when a large amount of hydrogen is dissolved, the hydrogen permeability coefficient is improved, but at the same time, hydrogen embrittlement becomes remarkable. That is, the increase in the hydrogen permeability coefficient and the hydrogen embrittlement resistance are contradictory, and it is generally very difficult to achieve both with a single-phase (solid solution) alloy, and there is still room for studying the combination of compositions.

  Further, when the hydrogen permeable alloy is used as a thin plate (membrane), more hydrogen can be produced efficiently, and the cost can be reduced. As a method for producing a thin plate, (1) an alloy is sliced thinly. (2) Preparation of an amorphous film by liquid quenching. (3) Rolling can be considered. Of these, slicing takes time and costs, and it is not easy to produce a film with a larger area. Although liquid quenching can form a thin film in a short time, it is technically difficult to produce a wide film and a thin plate having a different thickness. On the other hand, rolling is simple, easy, low cost, can produce a film with a large area, is widely used industrially, and has developed technically. If a thin plate can be produced by a simple method of rolling, it is expected that an alloy film having excellent hydrogen permeation characteristics can be mass-produced at a low cost. Therefore, it can be said that the rolling workability of the hydrogen permeable alloy is an important item.

  Accordingly, an object of the present invention is to provide a method for producing a hydrogen separation / purification double phase alloy having a large rolling workability and a large hydrogen permeability coefficient.

  The present inventor uses an Nb-Ti-Ni-based multiphase alloy for hydrogen separation / purification as the alloy composition, and heat-treats the alloy at a temperature exceeding 1000 ° C for 100 hours or more, thereby achieving the above-described hydrogen permeability coefficient. It has been found that a hydrogen separation / purification double phase alloy is obtained which has a high temperature and excellent cold workability. Even if plastic processing is performed, if heat treatment is performed at a temperature exceeding 1100 ° C., not only the TiNi phase structure but also the Nb—Ti phase structure is recrystallized as described later, and the as-cast Nb— Almost the same hydrogen permeability coefficient as Ti-Ni based hydrogen separation / purification double phase alloy is obtained. The resulting hydrogen separation / purification dual phase alloy is an Nb-Ti-Ni alloy characterized by comprising a composite phase of a phase responsible for hydrogen permeability and a phase responsible for hydrogen embrittlement resistance. The average crystal grain size (dc) of the phase (primary crystal) mainly composed of Nb mainly responsible for hydrogen permeability is 10 to 30 μm.

  In the present invention, since an Nb—Ti—Ni-based alloy is used, rolling can be adopted as a plastic working means for forming a thin film. The rolling rate can be 10% or more, and further 30% or more. Thereby, the thickness of the Nb—Ti—Ni-based double phase alloy for hydrogen separation / purification can be made 0.05 to 3 mm, and high hydrogen permeation performance can be obtained.

  The Nb—Ti—Ni-based alloy is, for example, an arc melting method in an inert gas atmosphere, a high-frequency induction heating melting method in an inert gas atmosphere or in a vacuum, an electron beam melting method in a vacuum, or a laser heating melting method. It can be prepared by dissolving by, for example. It is also possible to form a Pd film or a Pd alloy film on both sides of the surface of the multiphase alloy for hydrogen separation / purification on the side where the raw material to be treated is flowed and on the side where the purified hydrogen is taken out to obtain a final form.

The composition of the Nb-Ti-Ni alloy for use in the present invention, in _{atomic%, Nb 100-xy Ti y} Ni x ( however, x = 5 to 45, a y = 15 to 55, Nb 10 to 75 ). If the amount of Ni is less than 5%, the hydrogen permeation coefficient is low, and if it exceeds 45%, it becomes brittle and plastic processing such as rolling becomes difficult. When Ti exceeds 55%, the hydrogen permeation coefficient is low, and when Ti is less than 15%, the material becomes brittle and plastic processing such as rolling becomes difficult. It is also possible to replace part of the Ni element with an element such as Ag, Al, Cr, Cu, Ga, Zn, Fe, Mn, etc. with an upper limit of 50% with respect to the Ni element. Preferably it is 20% or less. It is also possible to replace a part of Ti with another 4A group element with an upper limit of 10% with respect to the Ti element. It is also possible to replace a part of Nb with another 5A group element with an upper limit of 10% with respect to the Nb element. This alloy mainly consists of two phases of bcc- (Nb, Ti) solid solution and B ₂ -NiTi compound. The Nb-Ti phase is responsible for hydrogen permeation properties by dissolving and diffusing hydrogen. On the other hand, the NiTi phase is difficult to be hydrogen embrittled and bears mechanical properties in hydrogen. In other words, a new alloy that has both excellent hydrogen permeation characteristics and resistance to hydrogen embrittlement has been realized. It is also attractive that Nb-Ti-Ni double phase alloys are much cheaper than Pd based alloys.

  Even if the thickness is reduced by plastic working such as cold rolling, a hydrogen separation / purification double phase alloy having the same hydrogen permeability coefficient (Φ) as that of the cast material can be obtained by applying the present invention. In addition, since the Nb-Ti-Ni double-phase hydrogen permeable alloy is used, it has excellent plastic working performance, and a high-performance thin plate can be produced as a double-phase alloy for hydrogen separation and purification by a process that combines heat treatment. It is possible to produce a thin plate with a thickness of 150 μm or less by this method, and furthermore, an extremely excellent hydrogen embrittlement-resistant double phase alloy that can withstand a pressure difference of 0.7 MPa or more even at a thickness of 120 μm. Can provide.

The investigation method for the recrystallization temperature performed in the examples will be extended. The alloy sample was cold worked to r = 50% under the same conditions as the rolling rate measurement, placed in an opaque quartz tube, and after replacing with Ar, the quartz tube was evacuated to about 6 × 10 ⁻³ Pa and sealed. The quartz tube was annealed in an electric furnace at 573 K to 1373 K for 1 hour and then water quenched. The surface was polished, and the recrystallization temperature was measured by applying a rolling process using a micro Vickers hardness tester and using a rolling mill until cracking occurred.

The hydrogen permeation test method is described. First, for the purpose of preventing oxidation and facilitating dissociation and recombination of hydrogen, the produced disk was coated with Pd using an RF / DC radio frequency magnetron sputtering apparatus (SPF-430H) manufactured by ANELVA. The furnace is evacuated to 4.0 × 10 ⁻³ Pa using a rotary pump and cryopump, and then reverse sputtering is performed for 10 minutes at an RF power of 50 W, then the substrate is heated to 523 K, and this sputtering is performed at a DC power of 0.05 W. It was performed for 5 minutes, and Pd was coated about 190 nm. Next, the Pd-coated disc was sandwiched between copper gaskets and set in a transmission device. After replacing the inside of the pipe with Ar, vacuuming was performed with an oil diffusion pump until the pressure became 2.7 × 10 ⁻³ Pa or less, and the furnace was heated to 673 K and held for 40 minutes. Thereafter, hydrogen was introduced, the supply-side hydrogen pressure was adjusted to 0.20 MPa, the permeation-side pressure was adjusted to 0.10 MPa, and the hydrogen permeation measurement was started after holding for 60 minutes. For the measurement, high-purity hydrogen having a purity of 7N was used and the hydrogen pressure on the supply side was in the range of 0.20 MPa to 0.8 MPa. This operation was similarly performed at 623K, 573K, and 523K. The flow rate method was used for measurement. The flowmeter used was MODEL3300 manufactured by KOFLOK. The transmission area is 2.46 × 10 ⁻⁵ m ² (true circle with a diameter of 5.6 mm).

  The method of tissue observation and composition analysis in the present invention will be described. A small piece of the prepared sample was embedded in acrylic pressure molding resin, polished to a mirror finish, and observed with a scanning electron microscope (SEM, JSM5300) manufactured by JEOL Ltd. did. The composition analysis was performed by X-ray spectrum collection using an energy dispersive X-ray spectrometer (EDS) manufactured by Oxford Instruments.

In hydrogen permeation using a metal film, the hydrogen flow rate J can be obtained from J = Φ (ΔP ^0.5 ) / L. However, since this equation is derived from Fick's first law and the Siebels law, the sample must satisfy the Siebels law. Therefore, J × L vs ΔP ^0.5 was plotted on a graph, and it was verified from the linearity whether the Siebelz rule was satisfied. Further, the hydrogen permeation coefficient (Φ) was determined from the slope of this straight line.

  EXAMPLES The present invention will be described below with reference to examples, but the present invention is not limited to these examples.

Example 1
The purity (mass%) of the metal used in this experiment is Nb = 99.9%, Ti = 99.5%, and Ni = 99.9%. Each metal was weighed so as to obtain the desired composition, and then melted in an Ar atmosphere using an arc melting furnace (ACM-DS-01S) manufactured by Daiya Vacuum to produce an ingot. The alloy production procedure is as follows. After substituting the inside of the furnace with vacuum using an oil diffusion pump to 2.7 × 10 ⁻³ Pa or less, Ar gas (purity 99.99%) was introduced to about 5 × 10 ⁴ Pa, and a tungsten electrode rod was used. Arc discharge. In order to remove the impurity gas in the Ar gas, getter Ti was dissolved for about 2 minutes before the sample was dissolved. The alloy sample was then melted at an arc current of 400 A or more for about 2 minutes. Thereafter, in order to make the composition uniform, the operation of melting the alloy ingot upside down for about 2 minutes was performed about 10 times. The alloy compositions shown in this experiment are all mol%.
A 2.65 × 2.65 × 7 mm rectangular parallelepiped rolling test piece was cut out from an alloy produced by arc melting using a BROTHER HS-300 wire electric discharge machine. The prepared sample is polished with water-resistant abrasive paper made by Buramet, and the roll interval is gradually narrowed at room temperature using a cold two-high rolling mill (100Φ x 100W) made by Nihon Cross Rolling. Until cold rolling. When the thickness of the sample before rolling was T ₀ and the thickness of the sample after rolling was T, the rolling rate r (%) was obtained by the following equation.

First, the alloy composition dependency of the rolling rate of Nb-Ti-Ni alloy was investigated. Table 1 shows the various compositions (nominal composition) and the cold rolling rate (r) of the (Nb-Ti-Ni) alloy. FIG. 1 illustrates this. It can be seen that an alloy having a composition slightly on the Ti excess side has a high rolling reduction in the vicinity of the straight line connecting the intermetallic compound TiNi and pure Nb. The alloy of this composition consists of two phases of bcc- (Nb, Ti) solid solution phase and B2-TiNi compound, and does not contain brittle intermetallic compounds.
As the Ti / Ni ratio departs from 1/1, the rolling rate tends to decrease and become brittle.

Fig. 2 shows that (a) Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy with a Ti / Ni ratio of 1/1, (b) Nb ₄₀ Ti ₄₀ Ni ₂₀ and (c) Nb ₄₀ Ti ₅₀ Ni _{10 with} a Ti excess. And XRD diffraction patterns of (d) Nb ₄₀ Ti ₂₅ Ni ₃₅ and (e) Nb ₄₀ Ti ₂₀ Ni _{40 with} an excess of Ni. FIG. 3 shows SEM photographs of the respective alloys. As can be seen from XRD patterns and SEM photographs, alloys with a Ti-rich and brittle composition such as Nb ₄₀ Ti ₄₀ Ni ₂₀ and Nb ₄₀ Ti ₅₀ Ni ₁₀ contain brittle Ti ₂ Ni intermetallic compounds. On the other hand, Ni-rich Nb ₄₀ Ti ₂₅ Ni ₃₅ and Nb ₄₀ Ti ₂₀ Ni ₄₀ contain brittle intermetallic compounds such as Nb ₈ Ti ₃ Ni ₉ . Intermetallic compounds are generally poor in ductility, but if a brittle intermetallic compound is included as the third phase, it is considered that the ductility is lowered and the rolling rate is also lowered. Here, it can be said that the Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy excellent in hydrogen permeation ability reported by Hashi et al. Can be rolled by 70% or more by cold rolling, and has very good cold workability.

Table 2 shows the rolling rate of the alloy with Nb = 40 mol% and the hydrogen permeability coefficient (Φ) at 673K in the cast state. The rolling rate of Nb ₄₀ Ti ₃₀ Ni ₃₀ is about 70%, while that of Nb ₄₀ Ti ₃₁ Ni ₂₉ , Nb ₄₀ Ti ₃₂ Ni ₂₈ , and Nb ₄₀ Ti ₃₃ Ni ₂₇ exceeds 80%. This becomes more noticeable as the Nb content increases. FIG. 4 is a photograph after rolling the original sample piece used in the rolling test and an alloy with poor ductility and an excellent ductility alloy. As shown in Table 2, Nb ₄₀ Ti ₂₈ Ni ₃₂ and Nb ₄₀ Ti ₃₆ Ni ₂₄ have a slightly lower rolling ratio than Nb ₄₀ Ti ₃₀ Ni ₃₀ , Nb ₄₀ Ti ₃₂ Ni ₂₈ , and Nb ₄₀ Ti ₃₄ Ni ₂₆ Although it is difficult to identify from the XRD pattern shown in FIG. 5, it is presumed that a trace amount of intermetallic compounds other than TiNi phase is contained as the third phase. The hydrogen permeability coefficient in the cast state is also higher in Nb ₄₀ Ti ₂₈ Ni ₃₂ alloy and Nb ₄₀ Ti ₃₆ Ni ₂₄ alloy than in Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy, Nb ₄₀ Ti ₃₂ Ni ₂₈ alloy and Nb ₄₀ Ti ₃₄ Ni ₂₆ alloy. Low. This means that in Nb-Ti-Ni multiphase alloys, the mechanical properties and hydrogen permeation coefficient are low even if there are so few intermetallic compound phases other than Nb-Ti and TiNi phases that are difficult to identify from XRD patterns. It is considered to have an adverse effect on

(Example 2)
The effect of cold rolling on the hydrogen permeability coefficient (Φ) of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy was investigated. FIG. 6 shows the relationship between the rolling rate and the hydrogen permeability coefficient, and FIG. 7 shows the temperature dependence of the hydrogen permeability coefficient (Φ) of the alloy at each rolling rate. The disc thickness after processing is 0.5 to 0.8 mm. As can be seen in the figure, the hydrogen permeability coefficient decreases as the degree of processing increases. The hydrogen permeability coefficient Φ ₆₇₃ of 673K is Φ ₆₇₃ = 1.78 × 10 ^-8 [mol H ₂ m ^-1 s ^-1 Pa ^-0.5 ] in the cast state, but the sample with a rolling rate (r) of 50% is Φ ₆₇₃ = It decreases to 5.66 × 10 ^-9 [mol H ₂ m ^-1 s ^-1 Pa ^-0.5 ].
As shown in the SEM picture of Fig. 8, the Nb-Ti primary crystal is crushed by rolling, and the distortion of this structure, and defects such as vacancies and dislocations grown by processing, affect the diffusion and solid solution of hydrogen. As a result, it is presumed that Φ was lowered by rolling. In addition, the sample after rolling did not generate cracks during the hydrogen permeation test. Therefore, it can be said that the sample after rolling is excellent in resistance to hydrogen embrittlement as the sample in the cast state.

(Example 3)
The recrystallization temperature of the Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy was measured. FIGS. 9 and 10 show the changes in hardness and rolling rate of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy that was rolled to 573 K to 1373 K for 1 hour after rolling to a rolling rate (r) of 50%. Hardness decreases rapidly with annealing above 873K. In addition, the rolling rate was consistent with the improvement of annealing at 1073K or higher, and recrystallization occurred near 1073K or higher, and reworking became possible. A 0.8 mm thick Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy was rolled to 0.3 mm, heat treated at 1273 K for 1 hour, and then further rolled to successfully produce a disk having a thickness of 150 μm or less.
FIG. 11 shows (a) Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy before rolling, (b) Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy rolled to a rolling rate of 70%, and (c) 1273K after rolling to 70%. This is a photograph of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy that was annealed for 1 hour and then rolled to the limit of the rolling mill.

Example 4
The change of the structure by heat treatment was investigated. 12, 13 and 14 are SEM photographs of samples heat-treated at 1073K, 1273K, and 1373K, respectively. Focusing on the structural change, it can be seen that the eutectic structure of Nb-Ti and TiNi is gradually lost as the heat treatment time elapses. The samples heat-treated at 1073K and 1273K consist of Nb-Ti primary crystals, fine Nb-Ti phase with a grain size of about 1 μm, and TiNi phase. On the other hand, the sample heat-treated at 1373K is composed of an Nb—Ti primary crystal, an Nb—Ti phase having a small particle size, and a TiNi phase at the time when heat-treated for 1 hour. However, the grain size of the Nb—Ti phase grows as time passes, and the grain size increases to about 10 to 30 μm at the time of heat treatment for 168 hours. For the measurement example of the average crystal grain size (dc) of the primary crystal of the hydrogen permeable alloy, a total of five cross-sectional photographs taken with SEM photographs were taken in different fields of view as in FIGS. 12 to 14, and a diagonal line was drawn for each. Thus, the average crystal grain size (dc) was determined by dividing the line segment length occupied by the crystal grains present on each diagonal line by the number of the crystal grains.
The temperature required for grain growth and diffusion is said to be about _Tm / 2 ( _Tm : melting point) or higher, and since the melting point of Nb is 2742K, the Nb-Ti phase diffuses a long distance separating the TiNi phase. It can be estimated that about 1371K is necessary to do this. At 1073K and 1273K with T _m / 2 or less, the atomic diffusion distance is short and the temperature is insufficient for grain growth, so a fine Nb-Ti phase is formed, while the sample heat-treated at 1373K has a long atomic distance. Since diffusion and grain growth are possible, the small Nb-Ti phase grows large in a form that bites each other, and it is thought that the structure is composed of a large grain-shaped Nb-Ti phase and TiNi phase.

(Example 5)
The effect of heat treatment on the permeability coefficient Φ was investigated. FIGS. 15, 16, and 17 show the relationship between the heat treatment time and the hydrogen permeation coefficient of samples heat treated at 1073K, 1273K, and 1373K, respectively. The hydrogen permeation coefficient of the samples heat treated at 1073K and 1273K decreased as the eutectic structure was lost, and decreased to Φ ₆₇₃ = 1.0 × 10 ^-8 [mol H ₂ m ^-1 s ^-1 Pa ^-0.5 ]. Thereafter, the hydrogen permeation coefficient was almost constant even when heat treatment was continued for a long time. On the other hand, the eutectic structure disappeared in the samples heat-treated at 1373K as in the samples heat-treated at 1073K and 1273K. However, the hydrogen permeation coefficient (Φ) decreases after heat treatment at 1373K for 100 hours, but the hydrogen permeation coefficient (Φ) reaches a value close to that of an alloy with a eutectic structure before heat treatment at 673K. Recovered. Therefore, it can be said that the eutectic structure is not necessarily essential for hydrogen embrittlement resistance. 18, 19, and 20 show Arrhenius plots of the hydrogen permeation coefficient (Φ) of the samples heat treated at 1073K, 1273K, and 1373K.

Table 3 shows the hydrogen permeability coefficient at 673K of the alloys of each structure. From the experimental results, the following can be inferred.
(1) When comparing Nb ₆ Ti ₄₂ Ni ₄₂ alloy and Nb ₂₀ Ti ₄₀ Ni ₄₀ alloy, the eutectic structure of Nb-Ti and TiNi is more advantageous for hydrogen permeation than TiNi single phase. (2) By comparing the Nb ₂₀ Ti ₄₀ Ni ₄₀ alloy and the Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy, the hydrogen permeation coefficient is improved when the primary Nb—Ti phase is present rather than the eutectic structure alone. (3) The hydrogen diffusion coefficient of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy heat-treated at 1073K and 1272K for 168 hours has decreased, so the loss of the eutectic structure has reduced the hydrogen diffusion coefficient, and it has a fine Nb-Ti phase. This suggests that the hydrogen permeation coefficient does not improve much. (4) The hydrogen permeation coefficient of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy heat treated at 1373 K for 168 hours showed almost the same value as the cast state at 673 K. It can be inferred that the presence of the -Ti phase has some influence on the temperature dependence of the hydrogen diffusion coefficient D and the hydrogen solubility K. Note that no cracking occurred in the sample after the heat treatment during the hydrogen permeation test.

(Example 6)
The hydrogen permeability coefficient of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy after rolling-heat treatment (r = 50%) was investigated. First, after rolling (r = 50%), samples heat-treated at 1073K and 1273K were used. FIGS. 21 and 22 show the relationship between the heat treatment time and the hydrogen permeation coefficient (Φ) of the sample heat-treated at 1073 K after rolling to a rolling rate of 50%, and the temperature dependence of the hydrogen permeation coefficient. FIGS. This is the relationship between the heat treatment time and the hydrogen permeation coefficient of a sample heat-treated at 1273 K after rolling to 50% and the temperature dependence of the hydrogen permeation coefficient. The hydrogen permeability coefficient (Φ) of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy rolled to a rolling rate of 50% decreases to Φ = 5.66 × 10 ^-9 [mol H ₂ m ^-1 s ^-1 Pa ^-0.5 ] at ^673K , and heat treatment As can be seen from the graph of the relationship between time and hydrogen permeation coefficient, the hydrogen permeation coefficient remained lowered after heat treatment at 1073K and 1273K. As can be seen from the SEM photograph of FIG. 25, the structure of the samples heat-treated at 1073K and 1273K after cold rolling is composed of distorted Nb—Ti primary crystals, fine Nb—Ti phases, and TiNi phases. This suggests that the TiNi phase structure is recovered by heat treatment at 1073K and 1273K, but is insufficient for recrystallization of the Nb-Ti phase structure.

(Example 7)
FIG. 26 and FIG. 27 show the relationship between the hydrogen permeation coefficient and heat treatment time of a sample heat-treated at 1373 K after rolling (r = 50%), and the temperature dependence of the hydrogen permeation coefficient. The hydrogen permeability coefficient improved with the heat treatment time, and after one week of heat treatment, the hydrogen permeability coefficient (Φ) was almost the same as in the cast state.
Looking at the SEM photograph of the structure in FIG. 28, although it has anisotropy with respect to the rolling direction, it consists of a Nb—Ti phase and a TiNi phase having a large particle size as in the case of heat treatment without rolling. 1073 K, the effect is greatly different due on the transmission coefficient at the heat treatment in the heat treatment and 1373K at 1273K, in Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy 1073 K, 1273K is a T _m / 2 or less, 1373K is T _m / 2 The above is considered to be greatly related.
FIG. 29 shows the relationship between the hydrogen permeation flow rate J and the hydrogen permeation coefficient Φ when the film thickness and the pressure difference are 0.2 MPa. As the film thickness decreases, the hydrogen permeation flow rate J increases in proportion to 1 / L even when Φ is constant, as can be seen from the equation of J. That is, reducing the film thickness is effective for increasing the hydrogen permeation flow rate J, as is the improvement of the hydrogen permeation coefficient Φ. Note that no cracking occurred during the hydrogen permeation test in the sample after the processing and heat treatment.

It is a figure which shows the relationship between a composition and a rolling rate. It is an XRD diffraction pattern of the double phase alloy for hydrogen separation / purification used in the present invention. It is a SEM photograph of the alloy of FIG. It is a figure which shows the state of rolling. 5 is an XRD diffraction pattern of the hydrogen separation / purification double phase alloy shown in FIG. It is a diagram showing a relationship between rolling rate and the hydrogen permeability coefficient of Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy. It is a figure which shows the temperature dependence of the hydrogen permeability coefficient ((PHI)) of the alloy of each rolling rate. It is a SEM photograph of FIG. After rolled to a rolling ratio (r) 50%, it is a diagram showing a hardness of 1 hour annealing treated Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy 573K～1373K. After rolling to 50% rolling ratio (r), it is a diagram illustrating a 1 hour rolling reduction annealing treated Nb ₄₀ Ti ₃₀ Ni ₃₀ alloy 573K～1373K. It is a figure which shows the state of rolling. SEM photograph of the sample heat-treated at 1073K. SEM photograph of the sample heat-treated at 1273K. SEM photograph of the sample heat-treated at 1373K. It is a figure which shows the relationship between the heat processing time of the sample heat-processed by 1073K, and a hydrogen permeation coefficient. It is a figure which shows the relationship between the heat processing time of the sample heat-processed at 1273K, and a hydrogen permeation coefficient. It is a figure which shows the relationship between the heat processing time of the sample heat-processed at 1373K, and a hydrogen permeation coefficient. It is a figure which shows the Arrhenius plot of the hydrogen permeation coefficient ((PHI)) of the sample heat-processed at 1073K. It is a figure which shows the Arrhenius plot of the hydrogen permeation coefficient ((PHI)) of the sample heat-processed at 1273K. It is a figure which shows the Arrhenius plot of the hydrogen permeation coefficient ((PHI)) of the sample heat-processed at 1373K. It is a figure which shows the relationship between the heat processing time and the hydrogen permeation coefficient ((PHI)) of the sample heat-processed by 1073K after rolling. It is a figure which shows the temperature dependence of the heat processing time and the hydrogen permeation coefficient of the sample heat-processed by 1073K after rolling. It is a figure which shows the relationship between the heat processing time of the sample heat-processed by 1273K after rolling, and a hydrogen permeation coefficient. It is a figure which shows the temperature dependence of the heat processing time and the hydrogen permeation coefficient of the sample heat-processed by 1273K after rolling. 24 is a SEM photograph of the alloy used in FIGS. It is a figure which shows the relationship between the hydrogen permeation coefficient of the sample heat-processed by 1373K after rolling, and heat processing time. FIG. 6 is a graph showing the hydrogen permeation coefficient and the temperature dependence of a sample heat-treated at 1373 K after rolling (r = 50%). It is a SEM photograph of FIG. It is a figure which shows the relationship between a film thickness, hydrogen permeation | transmission flow rate J, and hydrogen permeation coefficient (PHI).

Claims (8)

An Nb-Ti-Ni alloy comprising a composite phase of a phase responsible for hydrogen permeability and a phase responsible for hydrogen embrittlement resistance is subjected to heat treatment at over 1000 ° C for 100 hours or more. Of producing a multi-phase alloy for hydrogen separation and purification of the system. Hydrogen separation, characterized by subjecting an Nb-Ti-Ni-based alloy composed of a composite phase of a phase responsible for hydrogen permeability and a phase responsible for hydrogen embrittlement resistance to plastic processing, followed by heat treatment above 1000 ° C A method for producing a refining double phase alloy. The method for producing a multiphase alloy for hydrogen separation / purification according to claim 2, wherein the plastic working is rolling, and a rolling rate is 10% or more. The method for producing a multiphase alloy for hydrogen separation / purification according to claim 2 or 3, wherein the plastic working is performed to a thickness of 0.05 to 3 mm. The Nb-Ti-Ni alloy is melted by an arc melting method in an inert gas atmosphere, a high-frequency induction heating melting method in an inert gas atmosphere or in a vacuum, an electron beam melting method in a vacuum, or a laser heating melting method. The method for producing a multiphase alloy for hydrogen separation / purification according to any one of claims 1 to 4, wherein the method is produced. 6. The Pd film or the Pd alloy film is formed on both sides of the surface of the multiphase alloy for hydrogen separation / purification on the side where the raw material to be treated is passed and the side where the purified hydrogen is taken out. 2. A method for producing a double phase alloy for hydrogen separation / purification as described in 1. The Nb-Ti-Ni alloy is composed of Nb _100-xy Ti _y Ni _x (where x = 5 to 45 and y = 15 to 55) in atomic%. 6. A process for producing a multiphase alloy for hydrogen separation / purification according to any one of 6 above. In an Nb-Ti-Ni-based alloy comprising a composite phase of a phase responsible for hydrogen permeability and a phase responsible for hydrogen embrittlement resistance, a phase mainly composed of Nb mainly responsible for hydrogen permeability ( A hydrogen separation alloy having an average crystal grain size (dc) of primary crystal of 10 to 30 μm.