MEMBRANE-ELECTRODE ASSEMBLY, REACTOR COMPRISING THE MEMBRANE-ELECTRODE ASSEMBLY AND PROCESS FOR SEPARATING OFF HYDROGEN

Abstract

Disclosed herein is a membrane-electrode assembly containing: a gastight, selectively proton-conducting membrane which has a retentate side having an anode and a permeate side having a cathode; a voltage source for generating a potential difference between the anode and the cathode; an anode catalyst having a catalytically active material on the retentate side; and a cathode catalyst having a catalytically active material on the permeate side, in which the cathode catalyst has a smaller amount of catalytically active material than the anode catalyst. The present disclosure also includes a reactor containing the membrane-electrode assembly, and a process for separating off hydrogen using the membrane-electrode assembly.

Claims

1. A membrane-electrode assembly, comprising: a gastight, selectively proton-conducting membrane which has a retentate side having an anode and a permeate side having a cathode; a voltage source for generating a potential difference between the anode and the cathode; an anode catalyst having a catalytically active material on the retentate side; and a cathode catalyst having a catalytically active material on the permeate side; wherein: the anode is a gas diffusion electrode having a gas diffusion layer adjoining the proton-conducting membrane and a microporous layer applied thereto, on top of which the anode catalyst has been applied, and/or the cathode is a gas diffusion electrode having a gas diffusion layer adjoining the proton-conducting membrane and a microporous layer applied thereto, on top of which the cathode catalyst has been applied; and the cathode catalyst has a smaller amount of catalytically active material than the anode catalyst.

2. The membrane-electrode assembly according to claim 1, wherein the amount of the catalytically active material of the cathode catalyst is from 0.001 mg/cm.sup.2 to 1.00 mg/cm.sup.2, based on the total area of the cathode (11).

3. The membrane-electrode assembly according to claim 1, wherein the ratio of the amount of the catalytically active material of the cathode catalyst to the amount of the catalytically active material of the anode catalyst (15) is from 1:100 to 1:1.25.

4. The membrane-electrode assembly according to claim 11, wherein the anode catalyst and/or the cathode catalyst has platinum as catalytically active material.

5. The membrane-electrode assembly according to claim 1, wherein the anode catalyst and the cathode catalyst have the same catalytically active material.

6. The membrane-electrode assembly according to claim 1, wherein the retentate side and/or the permeate side each have an active area of the membrane-electrode assembly which comprises the catalytically active material, where the active area is from 5 cm.sup.2 to 20 000 cm.sup.2.

7. (canceled)

8. The membrane-electrode assembly according to claim 1, wherein the gas diffusion layers comprise an electronically conductive and open-pored material.

9. The membrane-electrode assembly according to claim 1, wherein the gas diffusion layers and/or the microporous layers comprise carbon.

10. A reactor, comprising: at least one apparatus for carrying out a chemical reaction in which a mixed product gas comprising hydrogen is formed; and at least one membrane-electrode assembly according to claim 1, wherein the membrane-electrode assembly is joined to the apparatus in such a way that at least part of the mixed product gas can be conveyed from the apparatus onto the retentate side of the membrane-electrode assembly.

11. A process for separating off hydrogen, the process comprising: a) carrying out of a chemical reaction in which a mixed product gas comprising hydrogen is formed in the reactor according to claim 10; b) feeding the mixed product gas to the retentate side of a membrane-electrode assembly: and (1) c) electrochemically removing at least part of the hydrogen contained in the mixed product gas with the membrane-electrode assembly, wherein at least part of the hydrogen is oxidized to protons over the anode catalyst on the retentate side of the membrane and the protons are, after traversing the membrane, reduced to hydrogen over the cathode catalyst on the permeate side.

12. The process according to claim 11, wherein the chemical reaction in step a) is a dehydroaromatization of C.sub.1-C.sub.4-alkanes.

13. The process according to claim 11, wherein at least 30of the hydrogen contained in the mixed product gas is separated off from the mixed product gas.

14. The process according to claim 11, wherein the removal of the hydrogen occurs at temperatures of from 20 C. to 200 C.

Description

[0073] Further objectives, features, advantages and possible uses can be derived from the following description of examples of the invention with the aid of the figures. Here, all features described and/or depicted form, individually or in any combination, the subject matter of the invention, also as a function of their summary in the claims or back-reference. In the figures:

[0074] FIG. (1) schematically shows a membrane-electrode assembly 1 in order to indicate the processes in the removal of hydrogen,

[0075] FIG. 2a shows a schematic sectional view of a membrane-electrode assembly 1 according to one embodiment of the invention,

[0076] FIG. 2b shows a schematic sectional view of a membrane-electrode assembly 1 according to another embodiment of the invention,

[0077] FIG. 3 shows a graph depicting the degree of removal of hydrogen over time for a membrane-electrode assembly 1 according to the first embodiment of the invention in example 1,

[0078] FIG. 4 shows a graph depicting the degree of removal of hydrogen over time for a membrane-electrode assembly 1 according to the first embodiment of the invention in example 3 and

[0079] FIG. 5 shows a graph depicting the degree of removal of hydrogen over time for a membrane-electrode assembly 1 according to the second embodiment of the invention in example 4.

[0080] FIG. 1 shows the basic concept of an electron membrane. A mixed gas comprising, in the present example, methane (CH.sub.4), hydrogen (H.sub.2) and impurities (X) is introduced on the anode side (retentate side 5) of the membrane-electrode assembly 1. At the anode 7, the hydrogen H.sub.2 comprised in the mixed product gas is oxidized to protons H.sup.+ and passes through the proton-conducting membrane 3 as a result of application of a potential difference between the anode 7 and the cathode 11 by means of the voltage source 13. On the cathode side (permeate side 9), the protons are reduced back to hydrogen in the gaseous state with addition of electrons.

[0081] FIG. 2a schematically shows the construction of a membrane-electrode assembly 1 according to the invention in a preferred embodiment. The core is formed by the membrane 3 which in this embodiment consists of polybenzimidazole doped with phosphoric acid. The membrane 3 is purely proton-conducting and gastight. A gas diffusion electrode 7, 11 is arranged on each side of the membrane 3. In the present embodiment, the gas diffusion electrodes 7, 11 have a symmetrical structure.

[0082] FIG. 2b schematically shows the construction of a membrane-electrode assembly 1 according to the invention in another preferred embodiment having an unsymmetrical structure.

[0083] To form the gas diffusion electrode, an electronically conductive and gas-permeable woven carbon fiber fabric is firstly used as gas diffusion layer 19, 23. This is coated with a microporous layer 21, 25 which in the present invention consists of carbon (e.g. industrial carbon black). The catalyst 15, 17 (optionally a supported catalyst, e.g. catalyst on a carbon support) is applied to the microporous layer. An ink comprising the carbon particles loaded with platinum (catalyst), water, dispersants (for example Nafion or EFKA) and thickeners, (for example Xanthan) can, for example, be produced for this purpose and this can then be printed on. This gas diffusion electrode 7, 11 is joined to the membrane 3, with the catalyst-coated side being in direct contact with the membrane.

[0084] Unsymmetrical membranes (FIG. 2b) differ from symmetrical members (FIG. 2a) in that a smaller amount of catalyst is applied to the cathode 11 compared to the anode 7.

EXAMPLES

First Embodiment

[0085] A first example of the use of the membrane-electrode assembly 1 of the invention is a novel and cost-effective route for preparing benzene (and also naphthalene) from natural gas, preferably by dehydroaromatization. This endothermic reaction according to the equation

6CH.sub.4.fwdarw.C.sub.6H.sub.6+9H.sub.2

[0086] is a reaction which is limited by the thermodynamic equilibrium and gives a conversion of less than 20% at from 700 C. to 800 C. Consequently, to increase the efficiency, the unreacted methane has to be recycled and the hydrogen produced (always 9 mol per 1 mol of benzene) has to be removed beforehand.

[0087] Construction of the Membrane-Electrode Assembly 1

[0088] The membrane-electrode assembly 1 had a polybenzimidazole membrane 3 doped with phosphoric acid, which is marketed by BASF SE under the product name Celtec P. Gas diffusion electrodes 7, 11 composed of carbon/platinum were arranged on both sides of the membrane 3. The active area of the membrane 3 was 25 cm.sup.2 in the specific examples.

[0089] The Celtec P membrane 3 used according to the invention is a gel-type membrane which conducts protons at a high temperature (from 120 C. to 180 C.) and has a high tolerance to impurities such as CO or sulfur in the mixed product gas.

[0090] The gas diffusion electrodes 7, 11 consisted of an electronically conductive woven carbon fiber fabric; the gas diffusion layer 9, 23 was covered with a microporous layer 21, 25 composed of industrial carbon black. As catalyst 15, 17, use was made of platinum supported on carbon (Vulcan XC72) with loadings in the range from about 10% to about 30% of Pt (based on the total mass of the catalyst consisting of platinum and carbon). The amount of platinum on the gas diffusion electrodes 7, 11 is indicated in the following tables.

[0091] General Test Conditions

[0092] Based on the specific use of the present membrane-electrode assembly 1 for benzene production, a test gas mixture (i.e. the gas supplied to the anode) comprised 88.1% of CH.sub.4 and 11.4% of H.sub.2 together with traces of impurities (5000 ppm of C.sub.2H.sub.4, 100 ppm of C.sub.6H.sub.6 and 50 ppm of C.sub.2H.sub.6). During the experiments, the cathode 11 was in each case flushed with N.sub.2. The typical flow rates for both the anode gas stream and the cathode gas stream was 500 ml per minute. As potential difference, 150 mV were applied. The membrane-electrode assemblies 1 were humidified (2 g/h of water on the cathode side). The test temperature was 160 C. For the specific membrane-electrode assembly 1, a gauge pressure of 3 bar was set on the anode side and the cathode side.

[0093] Measurement Data and Analysis

[0094] During the experiments, the current (I), the voltage (V), the time (t), the cell temperature (T), the gauge pressure (p.sub.an, p.sub.cath) at the anode 7 and at the cathode 11 and the gas composition at the outlet from the anode 7 were determined by gas chromatography. Proceeding from the data determined in this way, the degree of removal of hydrogen can be calculated in two different ways. On the one hand, the measured current corresponds according to Faraday's law to the protons transported through the membrane 3. The ratio of the hydrogen transported through the membrane 3 to the known amount of hydrogen at the anode 7 leads to the value of the current-based removal of hydrogen. On the other hand, the known hydrogen content at the anode 7 and the hydrogen content measured on the side of the cathode 11 can be used to calculate a gas chromatography-based removal of hydrogen. The current-based and gas chromatography-based degrees of removal do not display any significant differences, for which reason only the gas chromatography-based degrees of removal are shown in FIGS. 2 and 4.

Example 1

According to the Invention

[0095] The total platinum loading was 0.50 mg/cm.sup.2, with 0.34 mg/cm.sup.2 of Pt being present on the anode side and only 0.16 mg/cm.sup.2 being present on the cathode side (unsymmetrical structure). The catalyst used was loaded with about 30% of Pt.

[0096] Compared to a symmetrical cell having 1 mg/cm.sup.2 of Pt on each side (prior art), the amount of Pt has been reduced by 75%. Using this cell, from 97% to 98% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions. FIG. 3 shows the degrees of removal of hydrogen for example 1 and comparative example 1.

Example 2

According to the Invention

[0097] The total platinum loading was 1.26 mg/cm.sup.2, with 0.79 mg/cm.sup.2 of Pt being present on the anode side and only 0.47 mg/cm.sup.2 being present on the cathode side (unsymmetrical structure). The catalyst used was loaded with about 15% of Pt. Compared to a symmetrical cell having 1 mg/cm.sup.2 of Pt on each side (prior art), the amount of Pt has been reduced by 37%. Using this cell, more than 99% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions.

Example 3

According to the Invention

[0098] The total platinum loading was 0.30 mg/cm.sup.2, with 0.20 mg/cm.sup.2 of Pt being present on the anode side and only 0.10 mg/cm.sup.2 being present on the cathode side (unsymmetrical structure). The catalyst used was loaded with about 10% of Pt. Compared to a symmetrical cell having 1 mg/cm.sup.2 of Pt on each side (prior art), the amount of Pt has been reduced by 85%. Using this cell, more than 99% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions.

Comparative Example 1

Not According to the Invention

[0099] The total platinum loading was 0.49 mg/cm.sup.2; this loading is virtually identical to that in example 1. As in example 1, a catalyst loaded with about 30% of Pt was used. In contrast to example 1, a symmetrical cell having 0.24 mg/cm.sup.2 of Pt on the anode and 0.25 mg/cm.sup.2 of Pt on the cathode was tested in the comparative example (symmetrical structure). Using this cell, from 93% to 96% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions, i.e. significantly less than in example 1 (from 97% to 98%). FIG. 3 shows the degrees of removal of hydrogen for example 1 and comparative example 1.

Comparative Example 2

Not According to the Invention

[0100] The total platinum loading was 1.26 mg/cm.sup.2; this loading is identical to that in example 2. As in example 2, a catalyst loaded with about 15% of Pt was used. In contrast to example 2, a symmetrical cell having 0.64 mg/cm.sup.2 of Pt on the anode and 0.62 mg/cm.sup.2 of Pt on the cathode was tested in the comparative example (symmetrical structure). Using this cell, less than 96% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions, i.e. significantly less than in example 2 (>99%).

Comparative Example 3

Not According to the Invention

[0101] The total platinum loading was 0.36 mg/cm.sup.2; this loading is higher than in example 3 (+0.06 mg/cm.sup.2 of Pt). As in example 3, a Pt particle catalyst supported on carbon and having a loading of 10% of Pt was used. In contrast to example 3, a symmetrical cell having 0.18 mg/cm.sup.2 of Pt on the anode and 0.18 mg/cm.sup.2 of Pt on the cathode was tested in the comparative example (symmetrical structure). Using this cell, less than 98.5% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions, i.e. significantly less than in example 3 (>99%) in the unsymmetrical structure with less platinum.

[0102] The following table 1 show examples 1 to 3 and comparative examples 1 to 3 to indicate the influence of the amount of catalytically active material on the side of the anode 7 and on the side of the cathode 11. At an identical amount of Pt, higher degrees of removal of hydrogen can be achieved when using unsymmetrical cells than when using symmetrical cells. Table 1 also shows that a significant cost reduction can be achieved when a smaller amount of catalytically active material is present on the side of the cathode 11 than on the side of the anode 7. As can be seen from table 1, reductions in the platinum content of up to 85% can be realized, with the degree of removal of hydrogen continuing to remain above 95%.

TABLE-US-00001 TABLE 1 Pt Pt catalyst loading Example/ in total Pt in Pt in on Comparative cell H.sub.2 anode cathode carbon Pt example and [mg/ removal [mg/ [mg/ support reduction * cell design cm.sup.2] [%] cm.sup.2] cm.sup.2] [%] [%] Example 1 0.50 97-98 0.34 0.16 30 75.0 (unsym.) Comparative 0.49 93-96 0.24 0.25 30 75.5 example 1 (sym.) Example 2 1.26 >99 0.79 0.47 15 37.0 (unsym.) Comparative 1.26 <96 0.64 0.62 15 37.0 example 2 (sym.) Example 3 0.3 >99 0.20 0.10 10 85 (unsym.) Comparative 0.36 <98.5 0.18 0.18 10 82 example 3 (sym.) * compared to a symmetrical cell having 2 mg/cm.sup.2 of Pt (1 mg/cm.sup.2 of Pt on each side).

[0103] Evaluation

[0104] FIG. 3 shows a graph depicting the removal of hydrogen when using the membrane-electrode assembly 1 as per example 1 (continuous line) and comparative example 1 (broken line) over 500 hours of operation. The unsymmetrical membrane-electrode assembly 1 as per example 1 showed good performance over 500 hours of operation with a degree of removal of hydrogen of from >97% to 98%. The course of the degree of removal of hydrogen for the symmetrical membrane-electrode assembly 1 as per comparative example 3 under identical experimental conditions, with the same catalyst (about 30% Pt loading) and comparable total amount of Pt (about 0.5 mg/cm.sup.2 of Pt) over 500 hours of operation is likewise shown. The degree of removal of hydrogen is in this case significantly lower at from 93% to 96%.

[0105] Example 1 shows that traces of other hydrocarbons (e.g. ethane, benzene) do not damage the membrane 3 or deactivate the electrodes 7, 11.

[0106] FIG. 4 shows a graph depicting the removal of hydrogen when using the membrane-electrode assembly 1 as per example 3 (continuous line) and comparative example 3 (broken line) over 20 hours of operation. The unsymmetrical membrane-electrode assembly 1 as per example 3 showed good performance with a degree of removal of hydrogen of greater than 99%. The course of the degree of removal of hydrogen for the symmetrical membrane-electrode assembly 1 as per comparative example 3 under identical experimental conditions is likewise shown. The degree of removal of hydrogen is in this case significantly lower at about 98%.

[0107] As can also be seen from the results, no degradation of the platinum catalysts 15, 17 was observed for the unsymmetrical structure over 500 hours of operation, while in the case of the symmetrical structure, a significant reduction in the degree of removal of hydrogen was observed toward the end; in particular, the degree of removal dropped below the desired 95%.

Second Embodiment

[0108] A second example of the use of the membrane-electrode assembly 1 of the invention is the separation of H.sub.2 a from a gas mixture containing CO.sub.2, H.sub.2 as well as traces of impurities (CH.sub.4, C.sub.2H.sub.4, C.sub.2H.sub.6, C.sub.6H.sub.6,).

[0109] Construction of the Membrane-Electrode Assembly 1

[0110] The construction of the membrane-electrode assembly 1 is identical to the membrane-electrode assembly 1 described above in the first embodiment (examples 1 to 3). The amount of platinum on the gas diffusion electrodes 7, 11 can be taken from the following table 2.

[0111] Der Aufbau der Membran-Elektroden-Anordnung 1 ist identisch zu dem vorstehend in der ersten Ausfuhrungsform beschriebenen Aufbau (Beispiele 1 bis 3). Die Menge des Platins auf den Gasdiffusionselektroden 7, 11 ist der nachstehenden Tabelle 2 zu entnehmen.

General Test Conditions

Example 4

[0112] The test gas mixture (i.e. the gas supplied to the anode 7) contained 5% of H.sub.2 in CO.sub.2, as well as the following traces of impurities: 881 ppm CH.sub.4, 5 ppm C.sub.2H.sub.4, 0,05 ppm C.sub.2H.sub.6 and 0.1 ppm C.sub.6H.sub.6. During the experiments, the cathode 11 was flushed with N.sub.2. The typical flow rates for both the anode gas stream and the cathode gas stream was 500 ml per minute. As potential difference, 150 mV were applied. The membrane-electrode assemblies 1 were humidified (1 g/h of water on the cathode side). The test temperature was 160 C. For the specific membrane-electrode assembly 1, a gauge pressure of 3 bar was set on the anode side and the cathode side.

[0113] Measurement Data and Analysis

[0114] Measurement data and analysis are identical to the measurement data and analysis mentioned above in the first embodiment (examples 1 to 3).

Example 4

According to the Invention

[0115] The total platinum loading was 0.68 mg/cm.sup.2, with 0.43 mg/cm.sup.2 of Pt being present on the anode side and only 0.25 mg/cm.sup.2 being present on the cathode side (unsymmetrical structure). The catalyst used was loaded with about 16% of Pt. Compared to a symmetrical cell having 1 mg/cm.sup.2 of Pt on each side (prior art), the amount of Pt has been reduced by 66%. Using this cell, appr. 95% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions (example 4). FIG. (5) shows the hydrogen removal rates for example 4 and comparative example 4.

Comparative example 4

Not According to the Invention

[0116] The total platinum loading was 0.58 mg/cm.sup.2; this loading is virtually identical to that in example 4. As in example 4, a catalyst loaded with about 16% of Pt was used. In contrast to example 4, a symmetrical cell having 0.29 mg/cm.sup.2 of Pt on the anode and 0.29 mg/cm.sup.2 of Pt on the cathode was tested in the comparative example (symmetrical structure). Using this cell, from appr. 93% of the hydrogen could be separated off from the gas mixture under the experimental conditions described in General test conditions (example 4), i.e. significantly less than in example 4 (95%). FIG. 5 shows the hydrogen removal rates for example 4 and comparative example 4.

[0117] The following table 2 show example 4 and comparative example 4 to indicate the influence of the amount of catalytically active material on the side of the anode 7 and on the side of the cathode 11. At an comparable amount of Pt, higher hydrogen removal rates can be achieved when using unsymmetrical cells than when using symmetrical cells. Table 2 also shows that a significant cost reduction can be achieved when a smaller amount of catalytically active material is present on the side of the cathode 11 than on the side of the anode 7.

TABLE-US-00002 TABLE 2 Pt Pt catalyst loading Example/ in total Pt in Pt in on Comparative cell H.sub.2 anode cathode carbon Pt example and [mg/ removal [mg/ [mg/ support reduction * cell design cm.sup.2] [%] cm.sup.2] cm.sup.2] [%] [%] Example 1 0.68 95 0.43 0.25 16 66.0 (unsym.) Comparative 0.58 93 0.29 0.29 16 71.0 example 1 (sym.) * compared to a symmetrical cell having 2 mg/cm.sup.2 of Pt (1 mg/cm.sup.2 of Pt on each side).

[0118] Evaluation

[0119] FIG. 3 shows a graph depicting the removal of hydrogen when using the membrane-electrode assembly 1 as per example 4 (continuous line) and comparative example 4 (broken line) in a short-time test over appr. 6 hours of operation.

[0120] The unsymmetrical membrane-electrode assembly 1 as per example 4 showed a remarkably higher hydrogen removal rate (95%) than the symmetrical membrane-electrode assembly 1 as per comparative example 4 (93%) under identical experimental conditions, with the same catalyst (about 16% Pt loading) and comparable total amount of Pt.

[0121] Together with examples 1 to 3 example 4 shows that better hydrogen removal rate can be obtained by using the unsymmetrical membrane-electrode assembly 1 in different gas mixtures containing H.sub.2 and impurities than by using a symmetrical membrane-electrode assembly with comparable or identical total platinum amount. In addition, the examples show that an unsymmetrical membrane-electrode assembly 1 is preferred with regard to the H.sub.2 removal rates in comparison to a symmetrical membrane-electrode assembly 1 irrespective of the side of humidifying (cathode side 11 in examples 1 to 3, anode side 7 in example 4.