CHA-type titanosilicate separation membrane and production method therefor and gas separation method

Abstract

The present invention provides a titanosilicate separation membrane which can also be used for separating a mixed gas containing a molecule having a relatively small size, has high durability in a high temperature environment, and has a high permeation rate and a high selectivity for a mixed gas containing water vapor. A titanosilicate separation membrane has a CHA-type titanosilicate crystal structure formed on a porous support, wherein aluminum is not substantially contained in the backbone of the titanosilicate crystal structure, and the titanosilicate crystal structure is constituted by silicon, oxygen, and titanium.

Claims

1. A method for producing a titanosilicate separation membrane having a CHA-type titanosilicate crystal structure formed on a porous support, comprising: a step of producing a seed crystal by performing hydrothermal synthesis in a first synthetic gel obtained by mixing a first raw material containing a silicon source, a titanium source, a fluoride, N,N,N-trialkyl-1-adamantaneammonium cations, and water; a step of carrying the seed crystal on the porous support; and a step of applying a second synthetic gel obtained by mixing a second raw material containing a silicon source, a titanium source, a fluoride, N,N,N-trialkyl-1-adamantaneammonium cations, and water to the porous support having the seed crystal carried thereon, and performing hydrothermal synthesis.

2. A gas separation method for separating a gas of a specific component from a mixed gas by using the titanosilicate separation membrane having a CHA-type titanosilicate crystal structure produced according to claim 1 formed on a porous support, and bringing the mixed gas into contact with the titanosilicate separation membrane, wherein the mixed gas contains one type or two or more types of first gases having a molecular diameter not smaller than the pore diameter of the CHA-type titanosilicate crystal and one type or two or more types of second gases having a molecular diameter smaller than the pore diameter of the CHA-type titanosilicate crystal, and one type or two or more types of gases selected from the second gases are separated by allowing the gasses to permeate through the titanosilicate separation membrane and the porous support.

3. The gas separation method according to claim 2, wherein the first gas is one or more types of gasses selected from the group consisting of methane, ethane, acetic acid, ethanol, 2-propanol, sulfur hexafluoride, benzene, o-xylene, m-xylene, p-xylene, and toluene, and the second gas is one or more types of gasses selected from the group consisting of hydrogen, helium, water, methanol, carbon dioxide, argon, oxygen, and nitrogen.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a schematic view of gas separation with a titanosilicate separation membrane of the present invention.

(2) FIG. 2 is a view schematically showing a crystal structure of a conventional zeolite membrane containing an aluminum element and a CHA-type titanosilicate crystal structure in which an aluminum element is replaced with a titanium element of the present invention.

(3) FIG. 3 is an electron micrograph of a surface of a titanosilicate separation membrane having a CHA-type titanosilicate crystal structure in which the Si/Ti ratio is 347 formed on an alumina support in Example.

(4) FIG. 4 shows X-ray diffraction patterns of a surface of the titanosilicate separation membrane having a CHA-type titanosilicate crystal structure in which the Si/Ti ratio is 347 formed on an alumina support in Example.

(5) FIG. 5 shows the results of a permeance test for various gasses with respect to various temperatures of the titanosilicate separation membrane having a CHA-type titanosilicate crystal structure in which the Si/Ti ratio is 347 formed on an alumina support in Example.

(6) FIG. 6 shows X-ray diffraction patterns before and after a heat resistance test at 1000° C. of an Al-containing CHA-type (Al-CHA) zeolite.

(7) FIG. 7 shows X-ray diffraction patterns before and after a heat resistance test at 1150° C. of an all-silica CHA-type (all-Si-CHA) zeolite.

(8) FIG. 8 shows X-ray diffraction patterns before and after a heat resistance test at 1150° C. when changing the Si/Ti ratio of the CHA-type titanosilicate according to the present invention.

(9) FIG. 9 shows comparison of the water vapor adsorption isotherms of an Al-containing CHA-type zeolite (Al-CHA) powder and a CHA-type titanosilicate powder according to the present invention.

(10) FIG. 10 shows comparison of the water vapor adsorption isotherms of an all-silica CHA-type (all-Si-CHA) zeolite powder and the CHA-type titanosilicate powder according to the present invention.

(11) FIG. 11 shows the CO.sub.2 permeance ratios and the CO.sub.2/CH.sub.4 selectivity ratios of an Al-CHA-type zeolite separation membrane and a CHA-type titanosilicate separation membrane according to the present invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

(12) Next, embodiments of the present invention will be described along with Example, however, the present invention is not limited thereto.

(13) The porous support is not particularly limited, but is preferably a support made of an alumina ceramic. A step of producing a titanosilicate separation membrane according to the present invention can be divided into two steps: a step of preparing a seed crystal and a step of synthesizing a CHA-type titanosilicate crystal structure. In the step of preparing a seed crystal and also in the step of synthesizing a CHA-type titanosilicate crystal structure, a raw material contains at least a silicon source, a titanium source, water, and a fluoride as a mineralizer, and contains N,N,N-trialkyl-1-adamantaneammonium cations as a structure directing agent.

(14) <Step of Preparing Seed Crystal>

(15) Colloidal silica as the silicon source, TiO.sub.2 as the titanium source, N,N,N-trialkyl-1-adamantaneammoniumcations as the structure directing agent, and a fluoride such as hydrofluoric acid, sodium fluoride, or calcium fluoride as the mineralizer are used. TiO.sub.2 is preferably anatase type. As the structure directing agent containing N,N,N-trialkyl-1-adamantaneammonium cations, a hydroxide is preferred, and in Example, N,N,N-trimethyl-1-adamant ammonium hydroxide (TMAdaOH) was used.

(16) Colloidal silica (40 wt %), TiO.sub.2 (anatase type), and TMAdaOH (25 wt %, manufactured by SACHEM, Inc.) were mixed, and HF (46 wt %) was added thereto so that the mixed solution became neutral. Thereafter, the mixed solution was heated to 200° C. and stirred at 250 rpm, whereby H.sub.2O was evaporated. The resulting product was crushed with an agate mortar, and H.sub.2O was added thereto, whereby a synthetic gel was prepared. The molar ratio of the gel is as follows: SiO.sub.2:TiO.sub.2:TMAdaOH:HF:H.sub.2O=1:x:1.4:1.4:6.0 (In Example, x was determined so that Si/Ti was 15, 30, and 57). The prepared gel was transferred to an autoclave, and a hydrothermal treatment was performed at 150° C. for 24 hours using an oven. The autoclave was taken out from the oven and cooled, and then, the resulting product was recovered by filtration. The product was washed with ion exchanged water and dried under reduced pressure for 24 hours. Finally, the product was fired at 700° C. for 10 hours in a firing furnace, whereby CHA-type titanosilicate particles were obtained.

(17) <Step of Synthesizing CHA-Type Titanosilicate Crystal Structure>

(18) Subsequently, the seed crystal is carried on the above-mentioned porous support. The carrying method is not particularly limited, but is preferably a rubbing method. A raw material containing the silicon source, the titanium source, the fluoride, N,N,N-trialkyl-1-adamantaneammonium cations, and water is mixed, whereby a synthetic gel is produced, and then, the gel is applied to the porous support having the seed crystal carried thereon, and hydrothermal synthesis is performed. The raw material of the synthetic gel is the same as the raw material used in the step of preparing the seed crystal. In Example, the synthesized CHA-type titanosilicate particles were used as the seed crystal, and carried on an alumina porous support (manufactured by Hitachi Zosen Corporation, outer diameter: 16 mm, inner diameter 12 mm) by a rubbing method. In the synthetic gel material, colloidal silica (40 wt %) as the silicon source, N,N,N-trimethyl-1-adamant ammonium hydroxide (TMAdaOH) (25 wt %, manufactured by SACHEM, Inc.) as the structure directing agent, and TiO.sub.2 (anatase type) as the titanium source were used. Colloidal silica, TiO.sub.2, and TMAdaOH were mixed, and HF (46 wt %) was added thereto so that the mixed solution became neutral. Thereafter, the mixed solution was heated to 200° C. and stirred at 250 rpm, whereby H.sub.2O was evaporated. The resulting product was crushed with an agate mortar, and H.sub.2O was added thereto so as to obtain the following composition. The final molar composition of the synthetic gel to be prepared is as follows: SiO.sub.2:TiO.sub.2:TMAdaOH:HF:H.sub.2O=1:x:1.4:1.4:6.0 (In Example, × was determined so that Si/Ti was 15, 30, and 57). The prepared synthetic gel was applied to the surface of the support after rubbing, and the entire face was sealed with a PTFE tape so that the synthetic gel did not flow down. The support having the synthetic gel applied thereto was transferred to an autoclave, and hydrothermal synthesis was performed at 150° C. for 72 hours. After the synthesis, the support was washed with water and dried under reduced pressure for 24 hours. Thereafter, the support was fired at 580° C. for 12 hours in a firing furnace.

(19) FIG. 1 schematically shows a state where a gas is separated with the CHA-type titanosilicate separation membrane of the present invention. The porous support is composed of a porous portion, a cap-shaped dense portion, and a tube-shaped dense portion. On the surface of the porous portion, a CHA-type titanosilicate crystal structure is formed. A supply gas containing a gas to be separated is supplied from the left side, and a gas having a molecular diameter that allows permeation through the pore of the CHA-type titanosilicate crystal structure flows to the right inside the tube through the CHA-type titanosilicate crystal structure and the porous portion. A gas having a molecular diameter that does not allow permeation through the pore of the CHA-type titanosilicate crystal structure flows outside the tube, and therefore, the supply gas can be separated. By setting the pressure outside the tube higher than the pressure inside the tube, the permeating gas flows to the inside from the outside of the tube.

(20) The porous support is not particularly limited, but is preferably an alumina support (manufactured by Hitachi Zosen Corporation). The porous support preferably has a diameter of 10 to 50 mm and a length of 500 to 1500 mm. The thickness of the CHA-type titanosilicate crystal structure is not particularly limited as long as the crystal structure can be formed thin, and the thickness thereof is preferably from about 1.0 to 10.0 μm.

(21) Hereinafter, the CHA-type titanosilicate crystal structure which does not substantially contain Al and is composed of Si, Ti, and O according to the present invention is also abbreviated as “Ti-CHA”, a conventional zeolite membrane containing Al is also abbreviated as “Al-CHA”, and an all-silica (pure silica) zeolite membrane is also abbreviated as “All Silica” or “all-Si-CHA” in some cases.

(22) FIG. 2 schematically shows a zeolite crystal structure. The drawing on the left side of the arrow shows a conventional zeolite crystal structure containing an Al atom in the crystal structure, and the drawing on the right side of the arrow shows a CHA-type titanosilicate crystal structure of the present invention. The crystal structure does not substantially contain an Al atom and is constituted by Si, Ti, and O.

(23) FIG. 3 is a scanning electron microscopical image (SEM) of a surface of a titanosilicate separation membrane having a CHA-type titanosilicate crystal structure in which the Si/Ti ratio is 347 formed on an alumina support in Example. According to FIG. 3, it is found that crystals characteristic of a CHA-type titanosilicate are formed without gaps. The Si/Ti ratio of the raw material for synthesis to make the Si/Ti ratio of the CHA-type titanosilicate crystal structure after synthesis 347 was 30. Incidentally, the Si/Ti ratio of the raw material for synthesis to make the Si/Ti ratio of the CHA-type titanosilicate crystal structure after synthesis 45 was 15, and the Si/Ti ratio of the raw material for synthesis to make the Si/Ti ratio of the CHA-type titanosilicate crystal structure after synthesis 578 was 57.

(24) FIG. 4 shows X-ray diffraction patterns obtained at four sites (about 90° intervals) in the circumferential direction of a surface of the titanosilicate separation membrane having a CHA-type titanosilicate crystal structure in which the Si/Ti ratio is 347 formed on an alumina support in Example. According to these patterns, the synthesized CHA-type titanosilicate crystal structure could be identified as CHA type.

(25) FIG. 5 shows the results of a permeation rate test for various gasses with respect to various temperatures of the titanosilicate separation membrane having a CHA-type titanosilicate crystal structure in which the Si/Ti ratio is 347 formed on an alumina support in Example.

(26) The gas permeation characteristics of the prepared titanosilicate separation membrane were evaluated using H.sub.2 (0.29 nm), CO.sub.2 (0.33 nm), N.sub.2 (0.364 nm), CH.sub.4 (0.38 nm), and SF.sub.6 (0.55 nm). The inter-membrane differential pressure was set to 0.1 MPa, and the measurement temperature was set to 40, 80, 120, and 160° C. The permeation rate (mL/s) was measured using a bubble film flow meter. From the obtained results, the permeation rate P.sub.i [mol/m.sup.2 sPa] of each gas and the selectivity α.sub.single [−] were calculated using the following formulae (1) and (2), respectively.
P.sub.i=q.sub.i/SΔp.sub.i  (1)
α.sub.single=P.sub.x/P.sub.y  (2)

(27) Here, P.sub.i, q.sub.i, and Δp.sub.i denote the permeation rate (permeance) [mol/m.sup.2 sPa], the permeation amount [mol/s], and the partial pressure difference [Pa] of an i component, respectively, and S denotes the membrane area [m.sup.2]. P.sub.x and P.sub.y denote the permeance [mol/m.sup.2 sPa] of x component and the permeance [mol/m.sup.2 sPa] of y component, respectively.

(28) FIG. 5 shows the permeation rate of a single gas with respect to the measurement temperature of the titanosilicate separation membrane. When the measurement temperature was 40° C., the CO.sub.2 permeation rate was 1.5×10.sup.−6 molm.sup.−2s.sup.−1 Pa.sup.−1, and a very high value was obtained.

(29) Next, the selectivity of each gas is shown in Table 1. A high value could be obtained for the selectivity of H.sub.2 and CO.sub.2 with respect to CH.sub.4 having the same size as the pore diameter (0.38 nm) of the CHA-type titanosilicate crystal structure and SF.sub.6 having a larger size than the pore diameter. This is considered to be due to a molecular sieve effect, and it was confirmed that a dense membrane can be synthesized.

(30) TABLE-US-00001 TABLE 1 Selectivity of each molecule Temperature (° C.) H.sub.2/CH.sub.4 CO.sub.2/CH.sub.4 H.sub.2/SF.sub.6 40 16.3 37.7 43.3 80 15.5 23.1 44.6 120 14.7 14.8 43.6 160 13.9 13.9 42.9

(31) FIG. 6 shows an X-ray diffraction pattern before a heat resistance test of the Al-CHA-type zeolite and an X-ray diffraction pattern after heat load at 1000° C. After the heat resistance test, the diffraction pattern characteristic of the CHA-type zeolite disappeared, and therefore, it is found that the zeolite containing Al in the backbone has low heat resistance. Incidentally, the Si/Al ratio of the Al-CHA-type zeolite subjected to this test is 10.

(32) FIG. 7 shows an X-ray diffraction pattern before a heat resistance test of an all-Si-CHA-type zeolite and an X-ray diffraction pattern after heat load at 1150° C. After the heat resistance test, the diffraction pattern characteristic of the CHA-type zeolite remained, and therefore, it is found that the all-Si-CHA-type zeolite which does not contain Al in the backbone has high heat resistance.

(33) FIG. 8 shows an X-ray diffraction pattern before a heat resistance test of the CHA-type titanosilicate and an X-ray diffraction pattern after heat load at 1150° C. After the heat resistance test, the diffraction pattern characteristic of the CHA-type titanosilicate remained, and therefore, it is found that the CHA-type titanosilicate in which the Al atoms are replaced with Ti atoms has high heat resistance in the same manner as the all-Si-CHA-type zeolite. Here, even when the Si/Ti ratio was changed to 45, 347, and 578, no change was observed in the heat resistance. As a result, it was found that the CHA-type titanosilicate in which the Al atoms are replaced with Ti atoms has high heat resistance without being affected by the Si/Ti ratio.

(34) In the X-ray diffraction patterns in FIGS. 7 and 8, the comparison results of the peak height ratio at a 2 θ angle of around 21° after the heat resistance test with respect to before the heat resistance test are shown in the following Table 2.

(35) TABLE-US-00002 TABLE 2 Peak height ratio between before Zeolite type and after heat resistance test (%) FIG. 7 all-Si-CHA 27 left in FIG. 8 Ti-CHA (Si/Ti = 45) 60 middle in Ti-CHA (Si/Ti = 347) 79 FIG. 8 right in FIG. 8 Ti-CHA (Si/Ti = 578) 60

(36) According to the results of Table 2, it is found that all the CHA-type titanosilicates according to the present invention shows even higher heat resistance performance than the all-Si-CHA-type zeolite.

(37) Table 3 shows the measurement results of the BET specific surface area (a.sub.sBET) and the micropore volume V.sub.t-plot before and after a heat resistance test for the Al-CHA-type zeolite, the all-Si-CHA-type zeolite, and the CHA-type titanosilicate. The conditions for the heat resistance test are the same as the conditions in the test whose results are shown in FIGS. 6 to 9. The measurement of the BET specific surface area was performed with a measuring device, BELSORP according to the BET method (Brunauer-Emmett-Teller method) manufactured by MicrotracBEL Corp. The measurement of the micropore volume was performed with a measuring device, BELSORP according to the t-plot method manufactured by MicrotracBEL Corp.

(38) TABLE-US-00003 TABLE 3 Al-CHA All-Si-CHA Ti-CHA (Si/Ti = 347) Before test After test Before test After test Before test After test a.sub.sBET (m.sup.2g.sup.−1) 776 238 714 462 813 819 V.sub.t-plot (cm.sup.3g.sup.−1) 0.277 4.58 × 10.sup.−3 0.25 0.16 0.295 0.286

(39) The ratio (%) of the value after the test to the value before the test obtained from the test results in Table 3 is shown in the following Table 4.

(40) TABLE-US-00004 TABLE 4 Ti-CHA Al-CHA All-Si-CHA (Si/Ti = 347) a.sub.sBET after test/before test (%) 30.3 64.7 100.7 V.sub.t-plot (cm.sup.3g.sup.−1) after test/before 1.65 64.0 96.9 test (%)

(41) According to the results of Table 4, a change between before and after the test is smaller and heat resistance is higher in the all-Si-CHA-type zeolite than in the Al-CHA-type zeolite. Further, the results show that a change between before and after the test is even smaller and heat resistance is even higher in the CHA-type titanosilicate than in the all-Si-CHA-type zeolite. Incidentally, the Si/Al ratio of the Al-CHA-type zeolite subjected to this test is 10.

(42) FIG. 9 shows the water vapor adsorption isotherms of the Al-CHA-type zeolite and the CHA-type titanosilicate. Here, P/P.sub.0 of the horizontal axis represents a relative pressure, and P.sub.0 denotes a saturated water vapor pressure at 313 K. The vertical axis represents a water adsorption amount. The solid line is a curve during adsorption, and the dotted line is a curve during desorption. This measurement was performed with a measuring device, BELSORP-max according to the constant volume method manufactured by MicrotracBEL Corp.

(43) It is found that as compared with the Al-CHA-type zeolite, the water adsorption amount of the CHA-type titanosilicate (Si/Ti ratio=347) is drastically reduced.

(44) FIG. 10 shows the water vapor adsorption isotherms of the all-Si-CHA-type zeolite and the CHA-type titanosilicate (Si/Ti ratio=578, 347, 45). It is found that as compared with the all-Si-CHA-type zeolite, the water adsorption amount of the CHA-type titanosilicate is reduced. It is found that the reduction amount is increased as the Si/Ti ratio is decreased (as the amount of Ti component is increased), and a smaller Si/Ti ratio is preferred.

(45) FIG. 11 is a graph showing comparison of the effect of a water vapor content on the CO.sub.2 permeance and the CO.sub.2/CH.sub.4 selectivity between the CHA-type titanosilicate membrane and the Al-CHA-type zeolite membrane in the case where water vapor is contained in a mixed gas (CO.sub.2/CH.sub.4 ratio=50/50).

(46) The measurement was performed by setting the inter-membrane differential pressure to 0.1 MPa, the temperature to 40° C., and the water vapor concentration on a volume basis to 0%, 0.1%, 0.2%, 1.0%, 2.0%, and 5.0% (a total of 6 cases) as the test conditions.

(47) In the graph of FIG. 11, the vertical axis on the left side represents the CO.sub.2 permeance ratio (%) when the content of water vapor was changed, and this is a ratio (%) when the CO.sub.2 permeance P.sub.wet in the case where water vapor is contained in the mixed gas is divided by the CO.sub.2 permeance P.sub.dry in the case where water vapor is not contained in the mixed gas. Incidentally, the Si/Al ratio of the Al-CHA-type zeolite membrane subjected to this test is 10.

(48) According to the graph of FIG. 11, the CO.sub.2 permeance ratio of the CHA-type titanosilicate membrane is dominantly higher than that of the Al-CHA-type zeolite membrane, and is maintained at about 80% even in the case where water vapor is contained at 1% in the mixed gas. On the other hand, the CO.sub.2 permeance ratio of the Al-CHA-type zeolite membrane decreases by 50% or more.

(49) In the graph of FIG. 11, the vertical axis on the right side represents the CO.sub.2/CH.sub.4 selectivity ratio (%) when the content of water vapor was changed, and this is a ratio (%) when the CO.sub.2/CH.sub.4 selectivity Q.sub.wet in the case where water vapor is contained in the mixed gas is divided by the CO.sub.2/CH.sub.4 selectivity Q.sub.dry in the case where water vapor is not contained in the mixed gas. It is found that also the CO.sub.2/CH.sub.4 selectivity ratio of the CHA-type titanosilicate membrane is dominantly higher than that of the Al-CHA-type zeolite membrane, and hardly decreases even in the case where water vapor is contained at 1% in the mixed gas. On the other hand, the CO.sub.2/CH.sub.4 selectivity ratio of the Al-CHA-type zeolite membrane decreases by about 50% or more.