Method for storage and release of hydrogen

Abstract

The invention provides a process for the production of hydrogen, comprising catalytically decomposing a concentrated aqueous solution of potassium formate in a reaction vessel to form bicarbonate slurry and hydrogen, discharging the hydrogen from said reaction vessel, and treating a mixture comprising the bicarbonate slurry and the catalyst with an oxidizer, thereby regenerating the catalyst. Pd/C catalysts useful in the process are also described.

Claims

1. A process for the production of hydrogen, comprising catalytically decomposing a concentrated aqueous solution of potassium formate in a reaction vessel to form bicarbonate slurry and hydrogen, discharging the hydrogen from said reaction vessel, and treating a mixture comprising the bicarbonate slurry and the catalyst with an oxidizer, thereby regenerating the catalyst.

2. A process according to claim 1, wherein the decomposition of the formate takes place at a temperature above 50? C.

3. A process according to claim 1, wherein the decomposition of the formate takes place at a temperature below 50? C., in the presence of an acid.

4. A process according to claim 3, wherein the acid is formic acid.

5. A process according to claim 3, wherein the decomposition of the formate takes place at ambient temperature.

6. A process according to claim 5, wherein the molar ratio between potassium formate and formic acid is in the range from 10:1 to 10:10.

7. A process according to claim 1, wherein the decomposition of the formate is started at a first temperature T.sub.1 below 50? C. in the presence of an acid, and on consumption of the acid, the reaction vessel is heated to a second temperature T.sub.2 above 50? C.

8. A process according to claim 7, wherein T.sub.1 is ambient temperature.

9. A process according to claim 1, wherein the concentration of the potassium formate solution is not less than 8M.

10. A process according to claim 1, wherein the potassium bicrbonate slurry is treated with air or oxygen, to regenerate the catalyst.

11. A process according to claim 1, wherein the concentrated aqueous solution of potassium formate is prepared by catalytically hydrogenating a potassium bicarbonate slurry.

12. A process for storage and subsequent release of hydrogen, comprising: (i) catalytically hydrogenating in a reaction vessel potassium bicarbonate slurry to form a concentrated aqueous solution of potassium formate; (ii) catalytically decomposing said formate solution in a reaction vessel to form bicarbonate slurry and hydrogen; (iii) discharging the hydrogen from said reaction vessel; and (iv) treating a mixture comprising the bicarbonate slurry and the catalyst with an oxidizer, thereby regenerating said catalyst.

13. A process according to claim 1, wherein the catalyst is palladium on carbon support (Pd/C), with palladium loading from 0.15 to 1.0%, characterized in that at least a portion of the palladium is in the form of sub-nanometer particles.

14. A process according to claim 13, wherein the Pd/C catalyst is prepared by the reduction of a palladium salt, with potassium formate as reductant, over activated carbon.

15. A process for the storage of hydrogen, comprising treating a mixture of potassium bicarbonate slurry and a catalyst with an oxidizer, and catalytically reducing said bicarbonate slurry in a reaction vessel to form a concentrated aqueous solution of potassium formate.

16. A process comprising subjecting a potassium bicarbonate slurry, in admixture with catalyst particles, to heterogeneous catalytic hydrogenation of potassium bicarbonate to form potassium formate.

17. A power system comprising at least one fuel cell and a hydrogen-generating unit for delivering hydrogen to the anodic compartment of said fuel cell, said hydrogen-generating unit comprising a catalyst-containing composition capable of reversible catalytic hydrogenation and dehydrogenation, which composition is in the form of potassium bicarbonate slurry and a concentrated aqueous potassium formate solution, respectively, wherein said hydrogen-generating unit is provided with a first feed line for introducing an incoming hydrogen stream for hydrogenating said poatassium carbonate slurry and a second feed line for introducing an oxidizer into said reaction vessel for regenerating said catalyst, and a discharge line for directing a hydrogen stream generated on dehydrogenating said potassium formate solution to said fuel cell.

Description

(1) In the drawings:

(2) FIG. 1 shows the solubility curves of potassium bicarbonate and potassium formate.

(3) FIG. 2 is a graph showing the effect of catalyst regeneration on the hydrogenation-dehydrogenation cycle.

(4) FIG. 3 schematically illustrates an apparatus for carrying out the process of the invention.

(5) FIGS. 4A and 4B are graphs showing the effect of catalyst regeneration on the hydrogenation-dehydrogenation cycle.

(6) FIG. 5 is a bar diagram showing TON and TOF measured in 4M potassium formate solution with different amounts of formic acid, at 70? C.

(7) FIG. 6 is a bar diagram showing TON measured in 4M potassium formate solution with different amounts of formic acid, at room temperature.

(8) FIG. 7 is a plot of TON against time on dehydrogenating different concentrated potassium formate solutions in the presence of formic acid, at room temperature.

(9) FIG. 8 is a curve showing the TON (left ordinate) and percentage of formate conversion (right ordinate) as function of reaction time, measured at two different temperatures.

(10) FIG. 9 is a bar diagram showing the dehydrogenation of potassium formate solution in the presence of formic acid at different temperatures.

(11) FIG. 10 is a graph showing TON measured on dehydrogenating potassium formate solution with different Pd/C catalysts.

(12) FIG. 11 is a STEM picture of Pd/C 0.2% prepared by formate reduction.

(13) FIG. 12 is a STEM picture of Pd/C 0.2% prepared by formate reduction.

EXAMPLES

Example 1

Reversible Hydrogen Absorption Over Potassium Bicarbonate Slurry

(14) 23.8 g of Pd/C 5% (51.3% wet, Engelhardt 5.47 mmol) were placed in an autoclave vessel along with 28.9 g of potassium bicarbonate (0.29 mol) and 12.23 g of water. The autoclave was sealed and washed with nitrogen gas 3 times.

(15) Hydrogen was added to the autoclave at 35? C. to 9.5 atmospheres and mixed for at least 2 hours. Then the initial pressure was released and gas flow from the autoclave was recorded while heating to 70? C. This cyclic procedure (formate synthesis and formate decomposition) was repeated 7 times without opening the autoclave. Then air was added to the autoclave to 10 atmospheres and heated to 70? C. for 2 hours with stirring in order to refresh the catalyst. After air addition, the autoclave was washed with nitrogen followed by hydrogen. Again hydrogen was loaded and its release was recorded for 5 more rounds, followed by catalyst regeneration under the conditions set forth above. The cyclic procedure was repeated again up to a total of 21 rounds, with the catalyst regeneration step taking place after the seventh, twelfth and seventeenth rounds. The results are graphically presented in FIG. 2, for the 2.sup.nd, 8.sup.th, 12.sup.th and 18.sup.th cycles, showing that following the catalyst regeneration according to the treatment of the invention (i.e., the 8.sup.th and 18.sup.th cycles), the rate of the reaction is significantly increased.

Example 2

Reversible Hydrogen Absorption Over Potassium Bicarbonate Slurry Using Pd/C 0.4% Catalyst

(16) Potassium Bicarbonate Slurry.Math.10M Potassium Formate Solution

(17) A 300 ml autoclave was fed with 4 g of Pd/C 0.4% of Preparation 6 (60% wet, 0.06 mmol), 3 g of potassium bicarbonate (0.03 mol, Sigma 23705) and 1.125 g of deionized water (0.0625 mol). The molar ratio between palladium and bicarbonate is 500:1. The molar ratio between water and bicarbonate is 2.1:1 and it fits the molar ratio between water and formate in a 10M potassium formate solution (the bicarbonate at these conditions is in the state of slurry). The autoclave was purged 3 times with nitrogen gas before hydrogen was allowed to flow into it to a pressure of 9.6 bar. The temperature was set to 35? C. and mechanical stirring (cross impeller) was activated at 400 rpm for 2 hours. After 2 hours the pressure was 8.3 bar. Then the autoclave's faucet was opened to reach rapidly to atmospheric pressure. The autoclave was connected to a water based flow-meter through a Ba(OH).sub.2 trap and heated to 70? C. to release hydrogen.

(18) The foregoing hydrogenation-dehydrogenation procedure was carried out two times, and then the catalyst was reactivated.

(19) Catalyst reactivation: at the end of dehydrogenation the autoclave was purged 3 times with nitrogen at 10 bar and then air was allowed to flow into it to a pressure 10 bar. The autoclave was heated to 70? C. for 2 hours and mechanical stirring (cross impeller) was set to 400 rpm. Then the autoclave's faucet was opened to reach rapidly to atmospheric pressure. The autoclave was purged 3 times with nitrogen at 10 bar before it was charged with hydrogen to a pressure of 9.6 bar and heated to 35? C. for 2 hours. Mechanical stirring (cross impeller) was activated at 400 rpm. Then the autoclave's faucet was opened to reach rapidly to atmospheric pressure. The autoclave was connected to a water based flow-meter through a Ba(OH).sub.2 trap and heated to 70? C. to release hydrogen.

(20) After catalyst reactivation, the cyclic hydrogenation-dehydrogenation procedure was repeated four times, and then the step of catalyst reactivation took place again. Thus, a total of seven cycles were run, with the catalyst regeneration step taking place after the second and sixth rounds. The results are presented in FIG. 4A. The upper line represents bicarbonate concentration and the lower line represents formate concentration.

(21) Potassium Bicarbonate Slurry.Math.15.7M Potassium Formate Solution

(22) A 300 ml autoclave was fed with 26.5 g of Pd/C 0.4% of Preparation 6 (60% wet, 0.4 mmol) and 5 g of potassium bicarbonate (0.05 mol, Sigma 23705). The molar ratio between palladium and bicarbonate is 125:1. Hydrogenation of solid bicarbonate to formate without addition of water can theoretically produce 15.7M potassium formate solution (in case all the bicarbonate is hydrogenated to formate and water). The autoclave was purged 3 times with nitrogen gas before hydrogen was allowed to flow into it to a pressure of 9.6 bar. The temperature was set to 35? C. and mechanic stirring (cross impeller) was activated at 400 rpm for 2 hours. After 2 hours the pressure was 8.5 bar. Then the autoclave's faucet was opened to reach rapidly to atmospheric pressure. The autoclave was connected to a water based flow-meter through a Ba(OH).sub.2 trap and heated to 70? C. to release hydrogen.

(23) The foregoing hydrogenation-dehydrogenation procedure was carried out four times, and then the catalyst was reactivated.

(24) Catalyst reactivation: at the end of dehydrogenation the autoclave was purged 3 times with nitrogen at 10 bar and then air was allowed to flow into it to a pressure 10 bar. The autoclave was heated to 70? C. for 2 hours and mechanical stirring (cross impeller) was set to 400 rpm. Then the autoclave's faucet was opened to reach rapidly to atmospheric pressure. The autoclave was purged 3 times with nitrogen at 10 bar before it was charged with hydrogen to a pressure of 9.6 bar and heated to 35? C. for 2 hours. Mechanical stirring (cross impeller) was activated at 400 rpm. Then the autoclave's faucet was opened to reach rapidly to atmospheric pressure. The autoclave was connected to a water based flow-meter through a Ba(OH).sub.2 trap and heated to 70? C. to release hydrogen.

(25) After catalyst reactivation, the cyclic hydrogenation dehydrogenation procedure was repeated three times. Thus, a total of seven cycles were run, with the catalyst regeneration step taking place after the fourth round. The results are presented in FIG. 4B. The upper line represents bicarbonate concentration and the lower line represents formate concentration.

Examples 3 to 9

Formate Decomposition in the Presence of an Acid

(26) To investigate the effect of acidic pH and type of acid on formate decomposition, various acids were added to potassium formate (KHCO.sub.2, abbreviated PF) 4M aqueous solutions in different acid: PF molar ratios and the so-formed acidic solutions went through dehydrogenation over Pd/C at 70? C. (commercial Pd/C 5%, Sigma 205680). The conditions of the reactions and the performance of the catalyst in the acidic environment are tabulated in Table 1.

(27) TABLE-US-00001 TABLE 1 Ratio TON at TOF Example Acid (KHCO.sub.2:acid:Pd) pH deactivation (min.sup.?1) 3 HCOOH 2000:2000:1 4 264 62 4 HCOOH 2000:200:1 5 1131 93 5 HCl 2000:1000:1 4 172 12 6 HNO.sub.3 2000:800:1 4 330 124 7 CH.sub.3COOH 2000:5000:1 4.5 288 64 8 HNO.sub.3 2000:200:1 6 475 91 9 H.sub.2SO.sub.4 2000:200:1 6 289 78

(28) In the absence of an acid, the TON and TOF were ?850 and ?30, respectively. The experimental results set out in Table 1 indicate that the addition of an acid leads to increased TOFs, but in some cases the TONs were lower than can be achieved at neutral media. Formic acid (abbreviated FA) emerges as especially useful acid for enhancing the decomposition of the formate to give hydrogen. In the experimental work to follow, formic acid was chosen for acidifying the formate solution.

Examples 10 to 16

Formate Decomposition in the Presence Varying Amounts of Formic Acid

(29) The following set of experiments illustrates the effect of the molar ratio between potassium formate and formic acid on formate decomposition. In the tested solutions, the concentration of potassium formate was 4M and the molar ratio potassium formate to palladium catalyst was constant (2000:1). Various amounts of FA were added to these PF 4M aquatic solutions, which went through dehydrogenation over Pd/C 5% (Sigma 205680) at 70? C. The conditions of the reactions and the performance of the catalyst in the presence of formic acid are tabulated in Table 2.

(30) TABLE-US-00002 TABLE 2 Ratio TON at TOF Example (KHCO.sub.2:HCOOH:Pd) deactivation (min.sup.?1) 10 (comparative) 2000:0:1 846 34 11 2000:20:1 914 47 12 2000:100:1 838 83 13 2000:200:1 1131 86 14 2000:350:1 716 102 15 2000:500:1 405 102 16 2000:2000:1 264 62

(31) The results are also presented in the form of a bar diagram in FIG. 5, where thick and narrow bars indicate the TONs and TOFs, respectively (the left and right ordinate correspond to the TON and TOF, respectively). At an elevated temperature (e.g., T=70?), the decomposition of formate in acidic environment runs most efficiently when the ratio potassium formate to formic acid lies in the range from 10:0.5 to 10:10, especially from 10:1 to 10:5, e.g. from 10:1 to 10:3.

Example 17

Formate Decomposition in a Highly Concentrated Solution in the Presence Formic Acid

(32) To illustrate that acidic environment generated by FA is capable of advancing formate decomposition also in highly concentrated PF solutions, FA was added to 16M PF aquatic solution at ratio FA:PF 1:10 (giving a solution of 16M PF, 1.6M FA). This solution went through dehydrogenation over Pd/C (Sigma 205680) at 70? C. Initial TOF=178 min.sup.?1, TON (at deactivation)=646.

Examples 18-24

Formate Decomposition in Acidic Environment at 25? C.

(33) The ability of formic acid to advance formate decomposition at 25? C. was studied. Various amounts of FA were added to PF 4M aquatic solutions. These solutions went through dehydrogenation over Pd/C (Sigma 205680) at 25? C. The conditions of the reactions and the performance of the catalyst in the presence of formic acid at 25? C. are tabulated in Table 3.

(34) TABLE-US-00003 TABLE 3 PF:FA TON at TOF Example ratio deactivation (min.sup.?1) 18 10:1 259 17 19 10:2 353 32 20 10:3 615 33 21 10:4 752 29 22 10:5 676 31 23 10:6 613 30 24 10:10 313 30

(35) The results show that the catalytically-driven decomposition of formate progresses satisfactorily even at room temperature, with the aid of formic acid. On graphically presenting the results in a bar diagram, where the abscissa and ordinate are MR.sub.PF:FA and TON, respectively, a curve resembling inverted parabola is seen (FIG. 6), with an axis of symmetry in the range of 10:3<MR.sub.PF:FA<10:5.

Examples 25-27

Formate Decomposition in Acidic Environment at 25? C.

(36) The experimental procedures set forth in the previous set of examples were repeated, but this time with higher concentration of PA, and correspondingly, with higher concentration of FA. The solutions went through dehydrogenation over Pd/C (Sigma 205680) at 25? C. The conditions of the reactions and the performance of the catalyst in the presence of formic acid at 25? C. are tabulated in Table 4.

(37) TABLE-US-00004 TABLE 4 PF concentration PF:FA TON at TOF Example (M) molar ratio deactivation (min.sup.?1) 25 12.0 10:2 519 50 26 12.0 10:3 563 88 27 14.5 10:1.2 344 47

(38) It can be seen that formic acid promotes the catalytically-driven decomposition of highly concentrated formate solutions at room temperature. The results set out in Table 4 are shown graphically in FIG. 7, where the TON is plotted as function of reaction time for each of the three experiments of Examples 25, 26 and 27, illustrating that TONs (at deactivation) higher than 300 are achievable in the very high concentration regimen of formate aqueous systems, with fairly reasonable PF:FA molar ratio.

Example 28

Formate Decomposition in Acidic Environment at 25? C. and in Neutral pH at 70? C.

(39) PF and FA were added to water to form an aqueous solution with PF and PA concentrations of 12.0M and 2.4M, respectively. The catalytically-driven reaction started at room temperature in the presence of Pd/C (Sigma 205680, the catalyst loading was 1:500 relative to the PF). After 90 minutes, the acid was essentially consumed (as indicated by cessation of hydrogen evolution) and the reaction mixture was heated to 70? C. and kept at that temperature for about additional 90 minutes to reach almost full decomposition of the formate.

(40) A curve showing the TON (left ordinate) and percentage of formate conversion (right ordinate) as function of reaction time is plotted in FIG. 8.

Examples 29-35

Formate Decomposition in Acidic Environment at Different Temperatures

(41) The effect of temperature variation on formate decomposition in an acidic environment induced by formic acid was tested. Aquatic solutions of potassium formate 4M an formic acid 0.8M went through dehydrogenation over Pd/C 5% (Sigma 205680) at a variety of temperatures, as described in Table 5:

(42) TABLE-US-00005 TABLE 5 Temperature TON at TOF Example (? C.) deactivation (min.sup.?1) 29 0? 519 9 30 10? 645 23 31 20? 858 41 32 30? 799 46 33 40? 1110 85 34 50? 632 63 35 60? 726 101

(43) A bar diagram showing the results of TON at deactivation as function of temperature is given in FIG. 9.

Examples 36-37

Hydrogen Release from Potassium Formate Solution over Pd/C 0.2% and Commercial Pd/C 5%

(44) 10 ml of 4M solution was prepared by addition of water to 3.3648 g (0.04 mol) of potassium formate. The solution was added to 0.59 g of Pd/C 0.2% (40% wet, 0.0067 mmol) of Preparation 7 or 0.014 g of Pd/C 5% (Sigma 205680, 0.0067 mmol) and went through dehydrogenation at 70? C. (formate: Pd molar ratio of 6000: 1). The profile of the reaction is illustrated in the graph of FIG. 10, where the turn over number (TON) is plotted against time. The results indicate that hydrogen can be released from potassium formate solution over Pd/C 0.2% with TON over 7 times better than when using Pd/C 5%.

Preparation 1

Supported Palladium Catalyst

(45) CNT (commercial multiwall carbon nano-tubes) or activated carbon was placed in a flask with isopropyl alcohol (IPA). The mixture was subjected to sonication for a total of 20 minutes (activation periods of one seconds each, with an intermission of one second between each activation period). Palladium acetate solution in IPA was prepared and added to the flask. The flask was heated to reflux (85? C.) for 3 hours followed by evaporation of the IPA. The content of the flask was dried for 1 hour under vacuum at 65? C.

Preparation 2

Supported Palladium Catalyst

(46) Montmorillonit k-10 and di-n-decyldimethylammonium bromide were placed in a flask (weight ratio ammonium salt:mineral 5:3). Ethanol was added as a solvent for the ammonium salt.

(47) The mixture was stirred at room temperature for 3 hours and then filtered and washed with ethanol. The ammonium salt absorbed-montmorillonit was placed in water with palladium (II) nitrate (weight ratio palladium: mineral 1: 9). The mixture was stirred for 18 hours, then filtered, washed with water and dried at vacuum at 60? C.

Preparation 3

Supported Palladium Catalyst

(48) Palladium (II) acetate was dissolved in acetone. CNT or activated carbon was added to the solution according to the desired percentage of palladium. The mixture was stirred while aquatic solution of hydrazine was added dropwise for 30 minutes. The mixtures were left over night and filtered by gravitation the next morning.

Preparation 4

Supported Palladium Catalyst

(49) CNT, activated carbon or PANI (polyaniline) was placed in water with palladium (II) nitrate. A reductive agent such as hydrazine solution or sodium borohydride aquatic solution was added dropwise for 30 minutes. The mixture was stirred for 18 hours, then filtered, washed with water and dried at vacuum at 60? C.

Preparation 5

Supported Palladium Catalyst

(50) CNT, activated carbon or PANI (polyaniline) was placed in water with palladium (II) nitrate. Hydrogen gas was added to 9.5 atmospheres for 2 hours at room temperature. Then the mixture was filtered, washed with water and dried at vacuum at 60? C.

Preparation 6

Supported Palladium Catalyst

(51) Palladium (II) nitrate dihydrate (0.096 mmol, Sigma 76070) was dissolved in water (1 L). Activated carbon (Sigma C-3345) was heated to 200? C. for 1 hour. The treated activated carbon (25 g in order to get 0.4% Pd/C) was added into the palladium solution and stirring was activated to 700 rpm for 1 hour. Then an aqueous solution of potassium formate (0.081 g in 200 ml of water) that was used as a reduction agent was added dropwise for 30 minutes at 25? C. (molar ratio between palladium and reduction agent is 10:1, total concentration of formate in the vessel was 0.008M). Following that the mixture was left while stirring continued at room temperature for 24 hours. After 24 hours the mixture was filtered, washed thoroughly with deionized water and left to dry at room temperature.

Preparation 7

Supported Palladium Catalyst

(52) Pd/C 0.2% was prepared using a procedure similar to that of Preparation 6, i.e., via formate reduction of Pd.sup.2+ under mild conditions, but this time 0.5 g of the treated activated carbon were added to the palladium salt solution to achieve the 0.2% loading.

(53) The Pd/C powder collected was subjected to STEM-EDS analysis [Tecnai F20 G2 (FEI company)]. FIG. 11 is a STEM image showing a region of the Pd/C sample which contains palladium (36.3%). The Pd particles in the region observed are invisible, indicating that the palladium present in that region consists of sub-nanometer Pd particles (below TEM resolution). There was no indication of the presence of potassium in the sub-nanometer Pd particles regions.

(54) On the other hand, in the STEM image of FIG. 12, another region of the Pd/C sample is shown, where Pd particles in the low nanometer range of size (from 1 nm to 5 nm) are identified. Compositionally, the palladium content is 65.59% and the potassium content is 0.55% (w/w).