Proton conducting ceramic membrane

Abstract

A proton conducting ceramic membrane comprising a conducting layer, wherein said conducting layer comprises a mixture of a rare-earth tungstate as herein defined and a mixed metal oxide as herein defined. The invention also relates to a reactor comprising said membrane and the use of said membrane in a dehydrogenation process.

Claims

1. A proton conducting ceramic membrane comprising a conducting layer, wherein said conducting layer comprises a mixture of: (i) a rare-earth tungstate of formula (I)
(Ln.sub.z1Dp.sub.z2).sub.aW.sub.b-cM1.sub.cO.sub.12-y(I) wherein Ln is Y, an element numbered 57 to 71, or a mixture thereof; Dp is Y or an element numbered 57 to 71 of the periodic table different from Ln; Z1 is 0.5 to 1; Z2 is 0.5 to 0; M1 is a metal selected from the group consisting of Mo, Re, V, Cr, Nb, U and Mn, or a mixture thereof; the molar ratio of a:b is 4.8 to 6; c is 0 to (0.5*b); and y is 0?y?1.8 and is a number such that formula (I) is uncharged; and (ii) a mixed metal oxide of formula (II)
Ln.sub.e-dM3.sub.dCr.sub.1-fM2.sub.fO.sub.3-x(II) wherein Ln is Y, an element numbered 57 to 71, or a mixture thereof; M3 is a metal selected from Ca, Sr or Ba; M2 is a metal selected from the group consisting of Al, Ga, Co, Ti, Mg, Mn, Fe, Ni, Y, Sc, Yb and Lu, or a mixture thereof; e is 0.95 to 1; d is 0.4 to 0.01; f is 0 to 0.5; and x is 0?x?0.5 and is a number such that formula (II) is uncharged.

2. A membrane as claimed in claim 1, wherein the rare-earth tungstate (i) is of formula (I)
La.sub.aW.sub.b-cMo.sub.cO.sub.12-y(I) wherein the molar ratio of a:b is 4.8 to 6; c is 0 to (0.5*b); and y is a number such that formula (I) is uncharged and y is 0?y?1.8.

3. A membrane as claimed in claim 1, wherein the mixed metal oxide (ii) is of formula (II)
Ln.sub.e-dSr.sub.dCrO.sub.3-x(II) wherein Ln is Y, an element numbered 57 to 71, or a mixture thereof; e is 0.95 to 1; d is 0.4 to 0.01; and x is a number such that formula (II) is uncharged and x is 0?x?0.5.

4. A membrane as claimed in claim 1, wherein the membrane is coated on both sides with a porous electron conducting coating or a dense hydrogen permeable coating.

5. A membrane as claimed in claim 4, wherein the porous electron conducting coating is a single phase material selected from a metal, metal-based alloy or a ceramic compound.

6. A membrane as claimed in claim 4, wherein the porous electron conducting coating is Ni or Pt.

7. A membrane as claimed in claim 4, wherein the porous conducting coating has a thickness of less than 10 microns.

8. A membrane as claimed in claim 1, wherein the conducting layer is less than 100 microns in thickness.

9. A membrane as claimed in claim 1, wherein the ratio of component (i) to component (ii) is about 1:1.

10. A membrane as claimed in claim 1, wherein the particle size of components (i) and (ii) is less than 5 microns.

11. A proton conducting membrane reactor comprising a dehydrogenation catalyst and a proton conducting membrane as defined in claim 1.

12. A reactor comprising a first zone comprising a dehydrogenation catalyst and a second zone separated from said first zone by a proton conducting membrane as defined in claim 1.

13. A process for the dehydrogenation of substance, comprising introducing said substance into the first zone of a reactor as defined in claim 12 thereby dehydrogenate said substance; allowing hydrogen formed during said dehydrogenation to pass through said proton conducting membrane into said second zone; introducing a purge gas into said second zone, to react with the hydrogen; or applying reduced pressure in said second zone to thus remove hydrogen from said second zone.

Description

DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows XRD patterns of the starting powders and the final composite membrane;

(2) FIG. 2 is a SEM image and EDX analysis of a 50/50 composite membrane;

(3) FIG. 3 is a SEM images of a 50/50 composite membrane and a LSC membrane;

(4) FIG. 4 is a SEM image of a 50/50 composite membrane with the top electronic coating;

(5) FIG. 5 is a schematic illustration of membrane operation of the composite membrane coated with a top porous electronic layer;

(6) FIG. 6 shows a total conductivity of LWO, LSC and different LWO-LSC composites as function of inverse temperature measured under four different reducing atmospheres;

(7) FIG. 7 shows the hydrogen separation flow obtained with three different kinds of membranes;

(8) FIG. 8 shows hydrogen flux as a function of temperature using wet Ar and wet 1% O.sub.2 diluted with Ar as sweep gas;

(9) FIG. 9 shows a Low temperature permeation test using a 0.4 mm-thick membrane, best composite composition;

(10) FIG. 10 shows the stability on stream in CO.sub.2 environment (0.7 mm-thick 50/50 composite membrane): Hydrogen flow as function of time using 15% CO2-85% Ar as sweep gas and 50% He-50% H.sub.2 as feed gas at 800? C.; both sides of the membrane were humidified;

(11) FIG. 11 shows TG measurements of LWO+LSC cer-cer in 5% CO.sub.2 in Ar;

(12) FIG. 12 shows the hydrogen separation flow obtained with three different kinds of membranes.

EXAMPLES

(13) Preparation of Materials

(14) Preparation by Sol-Gel (Pechini Method)

(15) Rare-Earth Tungstate [LWO](La.sub.5.55WO.sub.12-?)

(16) The preparation method employed here is based on the citrate-complexation route. The lanthanum oxide was dissolved in concentrated hot nitric acid (65% vol.) in stoichiometric proportion and the resulting nitrate was complexed using citric acid at a molar ratio 1:2 cation charge to citric acid. Another solution was prepared for the B cations (purity>99%), using ammonium tungstate, and complexing it with citric acid (Fluka, 99.5%) at the same ratio. Metal complexation in both cases was promoted by heat treatment at 120? C. for 1 hour. Both solutions were neutralized by controlled addition of ammonium hydroxide (32% wt.) and mixed at room temperature. The resulting solution was gradually concentrated by stepwise heating under stirring up to 150? C. and followed by foaming. The resulting product was subsequently calcined in air to eliminate carbonaceous matter and to promote mixed oxide crystallization. The final materials were annealed at 800? C.

(17) Mixed Metal Oxide [LSC](La.sub.0.85Sr.sub.0.15CrO.sub.3-?)

(18) The preparation method was similar to that employed for LWO. In this case Sr carbonate and Cr(VI) oxide were used as starting compounds. The final sintering temperature was 800? C.

(19) For the composite preparation, LSC and LWO powder were ball milled together in a 1:1 wt ratio in ethanol for 8 hours and then the dried powder was pressed into pellets.

(20) X-Ray Diffraction and SEM Technique

(21) XRD was carried out on a Philips X'Pert Pro equipped with a X'celerator detector using monochromatic Cu K.sub.? radiation. XRD patterns were recorded in the 20 range from 20? to 90? and analyzed using X'Pert Highscore Plus software (PANalytical).

(22) The microstructure of the composite membranes was analyzed by scanning electron microscopy (SEM-EDS) in a JEOL JSM6300 electron microscope.

(23) Hydrogen Permeation Test Procedure

(24) The composite membrane used in hydrogen permeation measurements consisted of a gastight 700-400 ?m thick disc with diameter 15 mm sintered at 1550? C. Both disk sides were coated by screen printing with a 20 ?m layer of a Pt ink (Mateck, Germany) in order to improve the catalytic activity of the sample.

(25) Permeation measurements were performed on a double chamber quartz reactor. Hydrogen (100 mL/min) was separated from a mixture of H.sub.2He (dry or saturated in water at 25? C.) using argon as sweep gas (150 mL/min). Feed and sweep humidification was accomplished by saturation at 20? C. using Milli-Q water. From the hydrogen content measured in the argon side (permeate side) and the argon flow rate, the total hydrogen permeation rate was calculated, assuming ideal gas law. The permeation fluxes (mL.Math.min.sup.?1.Math.cm.sup.?2) were calculated by dividing the permeation rates by the effective surface area of the membranes. The hydrogen content in the permeate side was analyzed using micro-GC Varian CP-4900 equipped with Molsieve5A, PoraPlot-Q glass capillary, and CP-Sil modules. Qualitative analysis of water concentration is done in the PoraPlot-Q channel. Sealing was done using gold rings and an acceptable sealing was achieved when the helium concentration was lower than 5% of the H.sub.2 permeated. Data reported were achieved at steady state after thirty minutes of stabilization and each test was repeated at least three times to minimize analysis error, obtaining an experimental standard deviation of 10.sup.?4.

(26) Conductivity Measurement Procedure

(27) Standard four-point DC conductivity measurements were carried out as a function of temperature for 4 different environments: dry H.sub.2, H.sub.2+H.sub.2O (2.5% vol.), dry D.sub.2 and D.sub.2+D.sub.2O (2.5% vol.), where hydrogen and deuterium are diluted in He (95%). A constant current was supplied by a programmable current source (Keithley 2601) while the voltage drop through the sample was detected by a multimeter (Keithley 3706).

(28) Results

(29) XRD

(30) The compatibility of both oxidic phases after sintering of the membrane was evaluated by XRD diffraction and the results are shown in FIG. 1. The XRD patterns confirm good mixing.

(31) SEM

(32) Mixing of the two metal oxides phases was also evaluated by SEM analysis performed on a 50/50 composite membrane (FIG. 2). FIG. 3 presents the SEM analysis of a LWO/LSC composite membrane compared with an all-LSC membrane, both sintered at 1550? C. The grains in the composite membrane have an average grain size around 2 ?m, while the grains are well-sintered and the remaining porosity is negligible. In the case of the all-LSC membrane (FIG. 3, right-hand image), the membrane porosity is still very high due to the low sintering activity of LSC.

(33) SEM analysis was then repeated following deposition of a porous platinum layer on both sides of the membrane. The SEM images are shown in FIG. 4. The layer was around 2-3 ?m-thick and continuous along the composite membrane surface. As a result, the entire surface of the LWO grains was connected/contacted to/with the porous Pt coating, which acts as catalyst and an electron collector/distributor. Indeed, the whole LWO surface is connected to the electron percolating LSC phase through the top porous coating. FIG. 5 is a schematic representation of the role of the electronic coating in the function of the membranes of the invention.

(34) Conductivity Results

(35) Total conductivity measurements for LWO, LWO+LSC (50/50 composite) and LSC are presented in FIG. 6 as a function of inverse temperature in H.sub.2, D.sub.2, H.sub.2+H.sub.2O and D.sub.2+D.sub.2O (where H.sub.2 and D.sub.2 are diluted (5%) in helium (95%) and pH.sub.2O and pD.sub.2O are 0.025 atm) atmospheres. In the case of LWO, proton transport prevails up to 800? C. as can be deduced from the hydration effect (?.sub.H.sub.2.sub.+H.sub.2.sub.O>?.sub.H.sub.2 and ?.sub.D.sub.2.sub.+D.sub.2.sub.O>?.sub.D.sub.2) and the isotopic effect (?.sub.H.sub.2.sub.+H.sub.2.sub.O>?.sub.D.sub.2.sub.+D.sub.2.sub.O). However, at higher temperatures, both n-type and oxygen-ion conduction prevail with respect to proton conduction (as discussed in Solis C., Escol?stico S., Haugsrud R., Serra J. M., J. Phys. Chem. C 115 (2011) 11124-11131). P-type conductivity is reported to prevail in LSC in the whole temperature range measured. When looking at the conductivity results, the expected p-type behavior is observed, which is related to the variation of pO2 when hydrogen is humidified, however this behavior is very similar to the one of pure proton conductors (hydration effect and the isotopic effectsee San Ping Jiang, Li Liu, Khuong P. Ong, Ping Wu, Jian Li, Jian Pu. Electrical conductivity and performance of doped LaCrO.sub.3 perovskite oxides for solid oxide fuel cells, Journal of Power Sources 176 (2008) 82-89) and no clear conclusions can be extracted with regard to the proton conducting character of LSC. The same behavior as for LSC can be observed in the 50/50 composite conductivity results, which appears to be principally a p-type electronic conductor; however the magnitude of the gas humidification and the isotopic effect are lower than in LSC. In summary, the total conductivity of the composite is 10 times lower than LSC but it is 2 orders of magnitude higher than LWO. The effect of the proportion between LWO and LSC phases on the total conductivity is shown in FIG. 6 (right-hand chart).

(36) Hydrogen Permeation

(37) Thickness of Membrane and Ratio of Components

(38) The effect of blending the LWO and LSC phases on hydrogen permeation/separation was investigated. Membranes with a porous electron conducting coating were used. Results are shown in FIG. 7. At 750? C., the 50/50 composite membrane (0.7 mm-thick) achieves hydrogen fluxes around 0.1 ml/min*cm.sup.2 while the all-LWO membrane (0.7 mm-thick) exhibits a flow around 0.015 ml/min*cm.sup.2. The enhancement in hydrogen permeation observed is considered to be due to the improvement in electron percolation by adding both (i) a mostly electronic conducting phase (LSC) and (ii) a current distributor coating on both membrane sides.

(39) The effect of the LWO/LSC ratio in the membrane on hydrogen flux was also investigated at 700? C., where proton transport is the prevailing mechanism. The 50/50 composite membrane (0.4 mm-thick) achieves hydrogen fluxes around 0.15 ml/min*cm.sup.2 while the 20/80 LWO/LSC composite membrane (0.4 mm-thick) exhibits a flow around 0.025 ml/min*cm.sup.2 despite the much higher total conductivity observed for this composite (FIG. 6). This result suggests that (1) the proton percolation in LSC is lower than in LWO; (2) there exists an optimum ratio between LWO and LSC, which would also depend on the microstructure.

(40) To investigate the effect of the chemical potential gradient, flux was monitored using 1% wet O.sub.2 as sweep gas. The results are given in FIG. 8 where it is also compared using wet Ar as sweep gas. The hydrogen flux increases approximately 5 times moving to oxidizing sweep conditions. This confirms that the electronic conductivity in the LWO/LSC is p-type in the whole pO.sub.2 range and that air can be used as sweep gas to obtain an increase in chemical potential gradient and a corresponding increase in hydrogen flux.

(41) An 100% LSC membrane was not tested since it was not possible to achieve high density membranes (leak-free) even after sintering at very high temperatures, which is thought to be due to the low LSC sintering activity and the associated chromium evaporation.

(42) Humidification and Temperature

(43) The best performing membrane (50/50 LWO/LSC) was further investigated for effects of humidification and temperature on hydrogen permeation. The results are shown in FIG. 9. The left-hand chart shows the effect of the humidification of sweep and feed stream. Having both sides of the membrane wet optimised hydrogen flux. Steam permeation is still relevant although proton transport is the prevailing mechanism. The right-hand chart illustrates the effect of the hydrogen concentration (in He) in the low temperature range. Flux was highest at 700? C.

(44) CO.sub.2 Atmosphere

(45) Hydrogen permeation measurements were performed using 15% CO.sub.2 in Ar as sweep gas. The measurements were carried out at 800? C. during 3 days using as feed gas 50% H.sub.2 in helium and both sides of the membrane were humidified. FIG. 10 shows hydrogen flow under these conditions as a function of time and it can be observed that hydrogen permeation is stable in CO.sub.2 atmospheres. The value of the hydrogen flow is slightly lower as compared with the case of using pure Ar as sweep gas, which is ascribed to competitive adsorption effects on the membrane/catalyst surface between CO.sub.2, H.sub.2 and H.sub.2O.

(46) Other stability tests were carried out and showed the high stability of LSC/LWO composites even in CO.sub.2 (FIG. 11) and CO.sub.2+H.sub.2S+HCN-containing wet environments.

(47) Porous Coating

(48) FIG. 12 shows the permeation results for 50/50 composite membrane with different coatings and thickness. The best result is obtained when the membrane is coated with a porous Pt coating. The lack of a porous electronic-conducting coating reduces substantially the permeation while the application of LSC porous coating improves slightly the permeation of the bare composite membrane.