Hydrogen production via photocatalytic oxidation of aqueous ...
WHEC 16 / 13-16 June 2006 – Lyon France
Hydrogen production via photocatalytic oxidation of aqueous ammonium sulfite solutions Cunping Huanga, Olwale Odebiyib, Nazim Muradovc, Ali T-Raissid Florida Solar Energy Center, University of Central Florida, Cocoa, Florida, USA [email protected], [email protected], [email protected], [email protected]
a
ABSTRACT: A new hybrid photo-thermochemical sulfur-ammonia (S-NH3) cycle is proposed. The cycle is a modification of the well-known Westinghouse hybrid cycle wherein the electrochemical step is replaced by a photocatalytic step. The new cycle utilizes both IR and visible light portions of the solar spectrum to achieve higher overall solar-to-hydrogen energy conversion efficiencies than those utilizing heat only. The main process, unique to the S-NH3 cycle, is the visible light-induced photocatalytic oxidation of ammonium sulfite (NH4)2SO3) with production of hydrogen and ammonium sulfate ((NH4)2SO4). The (NH4)2SO4 product is processed to generate oxygen and recover NH3 and SO2, which are then recycled and combined with water to regenerate (NH4)2SO3 forming a closed cycle. In this paper, we report our experimental findings with regard to the photocatalytic oxidation of aqueous (NH4)2SO3 solutions containing suspended platinum-doped cadmium sulfide (Pt/CdS) particles irradiated by a xenon arc lamp. We achieved energy conversion efficiencies (absorbed light to chemical energy of hydrogen) of about 12%. The photocatalyst remained active for more than 70 hrs. The experimental results for the photocatalytic step of the S-NH3 cycle describing the hydrogen generation kinetics and effects of temperatures, pH, catalyst preparation and platinum loading on the H2 production efficiencies are also discussed. KEYWORDS: Solar, Hydrogen production, Thermochemical cycles, sulfur ammonia, photocatalysis
INTRODUCTION: Thermochemical water splitting cycles (TCWSCs) can achieve high overall heat-to-hydrogen energy conversion efficiencies [1,2]. Many TCWSCs have been conceived in the past 40 years or so but only few have been extensively studied. For example, the General Atomics Corp. has been developing the sulfuriodine (S-I) cycle [3,4] which belongs to a group of TCWSCs known as the "sulfur family cycles." Most TCWSCs, among them sulfur family cycles, include at least one endothermic reaction/process step that operates at high temperatures and requires a high temperature heat source. Prospective high temperature heat sources suitable for thermochemical process interface include solar thermal concentrator and central receiver systems, and nuclear power plants (i.e. high temperature gas-cooled reactors, HTGR). Solar energy can provide the high temperature heat source needed to drive TCWSCs. The high temperature heat source can be a solar thermal concentrator system. Solar heat source is unique in the sense that it can be concentrated and directed into a solar receiver/reactor system to generate very high reactor temperatures far exceeding those available from the advanced nuclear reactors. Furthermore, solar spectrum contains high energy photons unavailable in conventional and other heat sources. In the paper a solar-powered TCWSC that can utilize both the thermal (i.e. high temperature heat) and the light components (i.e. photonic energy) of the solar resource is discussed. The overall solar-to-hydrogen energy conversion efficiency can potentially be greater than those with heat-only input. The sulfur-ammonia (S-NH3) cycle under development at the Florida Solar Energy Center (FSEC) is a solar powered TCWSC with the following four basic steps [5,6]: SO2(g) + 2NH3(g) + H2O(l) → (NH4)2SO3(aq) (NH4)2SO3(aq) + H2O → (NH4)2SO4(aq) + H2(g) (NH4)2SO4(aq) →2NH3(g) + H2SO4(l) H2SO4(l) → SO2(g) + H2O(g) + 1/2O2(g)
300 K 300 K 525 K 1125 K
1/9
(chemical absorption step) (photocatalytic step) (thermochemical step) (thermochemical step)
(1) (2) (3) (4)
WHEC 16 / 13-16 June 2006 – Lyon France Solar thermal energy is utilized in reactions (3) and (4) for the production of oxygen via decomposition of ammonium sulfate and sulfuric acid. Step (2) is a photocatalytic step in which sulfite ions are oxidized in the presence of VIS-photons and a photocatalyst to sulfate ions and at the same time water is reduced to hydrogen. FSEC’s hybrid photo-thermochemical sulfur-ammonia cycle is a modification of the Westinghouse hybrid cycle [7] wherein the electrochemical step is substituted by a photocatalytic process. Figure 1(A and B) shows a schematic diagram of the S-NH3 cycle.
Figure 1. Schematic flow diagram of solar powered S-NH3 thermochemical water splitting cycle. With reference to Figure (1), the thermal portion of the solar resource (i.e., near infrared NIR, and infrared, IR) is resolved by the receiver/photoreactor units (see Figure 1(B)) in the mirror field and concentrated into a high temperature thermocatalytic reactor located within the tower for the decomposition of (NH4)2SO4 to produce oxygen. The photonic (UV and visible light) portion of the solar resource is utilized for the hydrogen production via photocatalytic oxidation of (NH4)2SO3 into (NH4)2SO4 while water is reduced into hydrogen. The utilization of both thermal heat and photonic energy increase the solar to hydrogen energy conversion efficiency of the cycle. The resolution of the UV and visible light from NIR and IR portions of the solar spectrum can be accomplished by application of the broad-band antireflection coatings which provide very low reflectance over a broad range of wavelengths within the UV and visible spectra [8]. With this approach, photonic portion of the solar radiation can be made available for conducting photocatalytic hydrogen production step of the S-NH3 cycle. In the reminder of this paper, we primarily focus on the photocatalytic hydrogen production step of the S-NH3 cycle (Reaction (2)) described above. EXPERIMENTAL: Figure 2 depicts the experimental set-up employed in this work. Photocatalytic experiments were conducted using a 1000-Watt xenon arc lamp (Spectral Energy) as the light source. A water tank was used to filter out infrared radiation generated by the lamp. A 250 mL, custom-made borosilicate glass reactor was used as the reaction vessel. The reactor temperature was controlled by a water circulation jacket connected to a Polyscience MDL 612 analog recirculator. The reaction solution was stirred by a magnetic stirrer. A PyrexTM glass window at the front of the photoreactor facing the light source allowed absorption of the near-UV and visible light radiated from the xenon arc lamp only. Hydrogen was collected in a 1000 ml Pyrex glassmeasuring cylinder. On line pH measurements were carried out with a calibrated Oakton 1100 pH meter. Prior to each photolytic experiment, the reactor was purged with ultra-pure grade argon gas for two hours to flush out air/oxygen dissolved in the reacting solution and trapped gas inside the photoreactor. Spectral 2/9
WHEC 16 / 13-16 June 2006 – Lyon France power output from the Xe lamp was measured using a SHIMADZU UV-2401 spectrophotometer. Reagent grade ammonium sulfite ((NH4)2SO3.H2O, Aldrich) was used as received without further purification. Reagent grade cadmium sulfide (CdS) (99.999% purity, Aldrich) was used as photocatalyst. Commercial chloroplatinic acid (H2PtCl6) (Aldrich) was used for the preparation of Pt –doped photocatalyst particles (Pt/CdS).
Glass Window
Figure 2. Experimental set-up for H2 production via photocatalytic oxidation of (NH4)2SO3 aqueous solution. 20.00
Spectral Irradiance (W/m2/nm)
18.00
Arc Arc/Glass
16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 300
400
500
600
700
800
900
1000
1100
Wavelength (nm)
Figure 3. Spectral power output for xenon short arc lamp (1000 W). Light intensity measurement: Spectral irradiance from the xenon arc lamp was measured using a calibrated LI-1800 portable spectroradiometer. The measured values of the spectral irradiance for the xenon light source used are given in Figure 3. Light measurements were carried out both with and without a Pyrex glass window in front of the detector. Pyrex window served to filter out most of the UV light emitted by the lamp so that an emission spectrum closer to that of solar can be simulated. The UV content of solar radiation is only about 4%, which is less than that produced from a xenon arc lamp. Table 1 shows the measured spectral irradiance within various wavelength ranges for the Xe light source used. Table 1. Irradiance from xenon lamp at various wavelength ranges. Wavelength Light Intensity Intensity Power (W) (nm) (W/m2) Average (W/m2) 300-350 97 60 0.8 350-400 213 131 1.8 400-450 316 195 2.6 450-500 413 255 3.4 500-600 730 451 6.1 600-700 703 435 5.8 700-800 546 337 4.5 Total 3018 1865 25.1
3/9
WHEC 16 / 13-16 June 2006 – Lyon France The light intensity was also measured by using a pyranometer (LICOR LI-200) was placed at 12 inches and 25 inches away from the xenon arc lamp. The irradiance was measured at 9 different locations on the elliptically lit image formed by the collimated light. The total radiant power incident on the reactor aperture was determined to be about 25.07 W. The photocatalyst employed in this experiment is cadmium sulfide with a band gap of 2.4 eV, which is activated with light having wavelengths of 516 nm or shorter. We used the total radiant power with wavelengths in the range of 300 nm – 500 nm that falls on the photoreactor was measured to be 8.60 watts as the value for calculating both the process energy efficiency (incoming light energy to chemical energy stored in hydrogen) and the quantum efficiency. UV-Vis spectral analysis: A UV-VIS spectrophotometric analysis was conducted using a SHIMADZU UV2401) spectrometer on six different samples to determine the composition of the reacting solution and the concentration of ammonium dithionate (Na2S2O6), a dimer of ammonium sulfite. Samples used for the analysis were, 0.5M Na2SO3, 0.5M Na2SO4, 0.5M Na2S2O6 (sodium dithionate), 0.5M (NH4)2SO3, 0.5M (NH4)2SO4, and a reacted solution (after 64 hrs of irradiation). It was observed that there was virtually no cation effect on the absorbance of the solutions. Both ammonium sulfite and sodium sulfite solutions had almost identical UV-Vis spectra. Figure 4 depicts that spectra for the reacted solution falls between those of the sulfite and dithionate solutions. This suggests that after 64 hrs of light exposure, the reacted solution contains a mixture of sulfate, sulfite and dithionate ions. HPLC measurements were used to quantify concentration of the sulfate, sulfite and dithionate ions formed. 5.0 4.5
0.5M solutions
4.0 SO32-
Absorbance
3.5 3.0 Reacted Solution SO32- S2O62- SO32-
2.5 2.0 1.5 1.0
S2O62SO42-
0.5 0.0 200
225
250
275
300
325
350
Wavelength (nm)
Figure 4. UV-Visible spectra of sulfate, sulfite and dithionate ions. RESULTS AND DISCUSSION: 1. Effect of temperature: We have previously shown that temperature plays a significant role in the UV photolysis of aqueous (NH4)2SO3 solutions for hydrogen production [9]. Similar effect was found for the photocatalytic (Pt/CdS) process. Figure 5 depicts that the rate of hydrogen production increases in concert with the increase in temperature from 50oC to 70oC. However, the rate of hydrogen production drops as temperature is further increased from 70oC to 90oC. It should also be noted that pH of the solution is affected by the temperature. Increasing photolyte temperature to 90oC lowers solution pH, which in turn changes the rate of hydrogen evolution. At a given temperature, there is an optimum solution pH for which the rate of hydrogen production is the highest. 2. Effect of CdS platinum loading: Platinum metal loading of the CdS catalyst improves the rate of (NH4)2SO3 photocatalytic oxidation. However, if too much platinum is deposited on the photocatalyst surface, it can block the light from reaching the CdS surface. Experiments were conducted at four different CdS photocatalyst platinum loadings. Figure 6 depicts H2 production results for various Pt loadings. Figure 6 also shows that the highest rate of hydrogen production was achieved at a platinum loading of 1.0 wt.% (with respect to CdS). Decreasing the amount of Pt loading to 0.5%wt or increasing Pt loading to 2%wt (or higher) lowered hydrogen production rates. 4/9
WHEC 16 / 13-16 June 2006 – Lyon France
1400 1200
50C 70C 90C
Hydr ogen (mL
1000 800 600 400 200 0 0
50
100
150
200
250
300
350
400
450
T ime (min)
Figure 5. Effect of temperature on the rate of photocatalytic hydrogen production from (NH4)2SO3. Figure 7 shows the relationship between the rate of hydrogen production and Pt loading, indicating that the optimal Pt loading is approximately 1 wt. %. 3. Effect of (NH4)2SO3 concentration: Figure 8 depicts the effect of ammonium sulfite concentration on the rate of hydrogen production. It can be seen that as the solution concentration of (NH4)2SO3 increases from 0.432M to 0.866M, the rate of hydrogen also increases. But, as [(NH4)2SO3] is increased beyond 0.866M to 1.731M, the rate of H2 production remains unchanged and equal to that of 0.866M solution. This result implies that there is an optimum [(NH4)2SO3] for which photocatalytic hydrogen production rate is maximum. 250 0.50% 1%
Hydrogen (mL)
200
2% 3%
150
100
50
0 0
50
100
150
200
250
300
350
Tim e (m in)
Figure 6. Effect of platinum loading (percentage by weight of CdS) on the hydrogen production rate.
5/9
WHEC 16 / 13-16 June 2006 – Lyon France
1.6 1.4
Rate (mL/m in)
1.2 1 0.8 0.6 0.4 0.2 0 0
0.5
1
1.5
2
2.5
3
3.5
Cd S Platin um Lo adin g %
Figure 7. Optimum level of the CdS platinum loading on the rate of hydrogen production from (NH4)2SO3. 800 700 0.432 M 0.866 M 1.731
Hydrogen (mL)
600 500 400 300 200 100 0 0
50
100
150
200
250
300
350
400
Reaction Time (min)
Figure 8. Effect of ammonium sulfite concentration on the rate of photocatalytic hydrogen production. 4. Process efficiency: Thermocatalytic decomposition of SO3 is an extensively studied process with the thermal efficiency as high as 70%.[3,4,10]. Also shown is the total solar to hydrogen efficiency of the S-NH3 thermochemical cycle being primarily dependent on the efficiency of the photocatalytic step (Reaction 2). Here, we define the process efficiency for the hydrogen generating Reaction (2) as the ratio of the higher heating value (HHV= 285.9 kJ/mol) of H2 to the total photonic energy (absorbed) input to the process in the wavelength range of 300 to 500 nm. The rationale for using this wavelength range is dictated by the availability of photons in solar spectrum (in the lower limit) and absorption of VIS photons by CdS (in the upper limit). . IR region of the radiation is used to carry out the high temperature oxygen production step. Based on the highest hydrogen production rate achieved so far (i.e., 3.2 mL/min over an 8-hr period), the overall photon-to-H2 efficiency was determined to be about 12.0%.
6/9
WHEC 16 / 13-16 June 2006 – Lyon France
5. Some mechanistic considerations: The kinetic measurements were carried out using 0.500 g of Pt/CdS (1.0 wt%) suspended in a 0.866 M (NH4)2SO3 solution (initial pH of 7.9) under isothermal conditions at 35oC. It took 32 hrs for the photocatalytic reaction to reach completion at which time H2 evolution had seized. Figure 9 depicts the amount of H2 generated and solution pH as a function of time during the photocatalytic oxidation of (NH4)2SO3. It can be seen that the rate of H2 production from photocatalytic conversion of (NH4)2SO3 for the first 25 hrs is a linear function of time yielding a constant hydrogen production rate of about 1.79 mL/min. The H2 production rate begins to diminish after 1500 minutes and levels off at about 2000 minutes of reaction time. Data for the online pH measurement showed that the solution pH increased slightly from 7.9 to 8.6 and then maintained constant until H2 production stopped. The increase in the solution pH can be attributed to the formation of dithionate ions in the solution [9]. To ascertain formation of the S2O62- ions in the solution and determine the initial and end products of the reaction, a high performance liquid chromatograph (Dionex DX-500 HPLC) was employed and used to measure the concentration of the sulfur species in the aqueous solution before and after the photocatalytic reaction. The results showed that before the photocatalytic reaction began, both (NH4)2SO4 and (NH4)2SO3 were present in the solution. This indicates that some of the as received (NH4)2SO3 material had already been oxidized to (NH4)2SO4. The proposed mechanism of photocatalytic oxidation of (NH4)2SO3 was found to be similar to that of direct UV photolysis as described in [9]. The difference between these two processes being, at the end of reaction, the solution pH for the photocatalytic process does not drop precipitously (see Figure 9) - as a result of S2O62- being converted to SO42- ions. This implies that S2O62- ions are more stable at the conditions of the photocatalytic experiment (at 35oC). In contrary, we have shown that during UV photolysis of (NH4)2SO3 at 50oC, intermediately formed S2O62 ions that are converted into SO42-. It should be noted that at elevated temperatures, photocatalytically formed S2O62- can also be converted to SO42- ions.
Figure 9. Amount of hydrogen produced and solution pH as a function of time during photocatalytic hydrolysis of ammonium sulfite. Based on the extensive literature data on the CdS photocatalysis and our experimental observations, we tentatively propose the following mechanism for the photocatalytic conversion of ammonium sulfite Upon UV-VIS illumination, electron-hole pairs are generated on the surface of CdS particles: CdS + hν → h+ + e-
(5)
Electrons (e-) reduce protons to generate hydrogen, while SO32- ions are oxidized by the holes (h+) forming SO42- ions. These processes are shown in Reactions 6 to 8 below: SO32- + H2O + 2h+ = SO42- + 2H+ 2H+ + 2e- = H2 Overall: SO32- + H2O = SO42- + H2
(6) (7) (8) 7/9
WHEC 16 / 13-16 June 2006 – Lyon France
As noted earlier, the increase in the solution pH can be associated with the formation of S2O62- ions. HPLC measurement of S2O62- ion concentration in the photocatalyzed sample of (NH4)2SO3 solution taken after 32 hrs showed an S2O62- ion peak. Formation of S2O62- ions was also observed during UV photolysis of (NH4)2SO3 as a reaction intermediate as shown in Figure 4. The following reactions could possibly lead to the formation of S2O62- : 2SO32- = S2O62- + 2e2H2O + 2e- = H2 + 2OH2SO32- + 2H2O = S2O62- + H2 + 2OH-
(oxidation) (reduction) (overall reaction)
(9) (10) (11)
Table 2 depicts the material balance for the 32-hr photocatalytic hydrolysis of (NH4)2SO3. Total of 123.4 mmol of sulfite ions were consumed during the reaction. In solution, 47 mmol of sulfate ions and 69 mmol of dithionate ions were generated. The difference between the amount of SO32- ions consumed and the sum of SO42- and S2O62- ions produced was 5.2%. In the gas phase, 2,651 mL of hydrogen was produced over the 32 hr period. The amount of hydrogen measured was 7.3% less than the theoretical value predicted based on Reactions 8 and 11. These discrepancies are well within the experimental margin of error. Table 2. Material balance for the photocatalytic process of aqueous (NH4)2SO3 solution. Species Before reaction After reaction Produced Consumed (mmol) (mmol) (mmol) (mmol) 123.4 0.0 0.0 123.4 SO3239.6 87.0 47.4 0.0 SO420.0 70.0 69.6 0.0 S2O62Theoretical H2 Measured Difference Difference H2 Production (mL) (mL) (mL) (%) 2,859 2,651 208 7.3 CONCLUSIONS: A new hybrid photo-thermochemical cycle - sulfur-ammonia (S-NH3) cycle has been proposed. The cycle is a modification of the Westinghouse hybrid cycle but with the electrochemical step replaced by a photocatalytic process. The step, unique to the S-NH3 cycle, is the light-induced photocatalytic oxidation of ammonium sulfite ((NH4)2SO3) with production of hydrogen. In this work, we report the experimental findings with regard to the Pt doped CdS photocatalytic oxidation of aqueous (NH4)2SO3 solutions irradiated by a xenon arc lamp. Results show that the concentration of (NH4)2SO3 in the solution and reaction temperatures had a major influence on the rate of hydrogen production. It was found that the rate of H2 production can be maximized with respect to both (NH4)2SO3 concentration and reaction temperature. The effect of Pt loading on photocatalytic activity of CdS was also determined. The data indicate that CdS with 1.0 wt% of Pt loading yields the highest rate of hydrogen production. From the kinetic and material balance experiments conducted, it was found that SO32- ions can be completely converted into SO42- and S2O62- ions. The S2O62ions formed during the photoreaction can be transformed to SO42- ions at elevated temperatures. The overall light-to-hydrogen conversion efficiency of the photocatalytic conversion of (NH4)2SO3 was determined to be about 12%. ACKNOWLEDGMENT: Authors are grateful for the financial support of this work to the National Aeronautics and Space Administration (NASA) through Glenn Research Center (GRC) under Grant NAG3-2751 and to Mr. Timothy Smith, the Grant Program Manager at NASA-GRC. REFERENCES: 1. Funk, J.E., "Thermochemical Hydrogen Production: Past and Present," Int. J. of Hydrogen Energy, 26(3), 185-190, 2001. 2. Funk, J.E., & England, C., "Solar Thermochemical Hydrogen: The Heat Source-Process Interface," Int. J. of Hydrogen Energy, 10(6), 359-64, 1985. 3. Besenbruch, G.E., “General Atomic Sulfur-Iodine Thermochemical Water Splitting Process,” Am. Chem. Soc., Div, Pet, Chem. Preprint, 271(48), 1982.
8/9
WHEC 16 / 13-16 June 2006 – Lyon France 4. Brown, L. C., Funk, J.F., Showalter, S.K., “High Efficiency Generation of Hydrogen Fuels Using Nuclear Power,” Annual Report to the U.S. Department of Energy GA-A3451, July 2000. 5. Huang, C., T-Raissi, A., "Analysis of Sulfur–Iodine Thermochemical Cycle for Solar Hydrogen Production. Part I: Decomposition of Sulfuric Acid," Hydrogen Energy Special Issue, Solar Energy, A. T-Raissi & S.A. Sherif (Editors), 78(5), 632-46, 2005. 6. Huang, C., T-Raissi, A., "A New Sulfur Ammonia Thermochemical Water Splitting Cycle," Proc. of the 15th World Hydrogen Energy Conf., Yokohama, Japan, 2004. 7. Brecher, L.E., S. Spewock, et al., “Westinghouse Sulfur Cycle for the Thermochemical Decompoistion of Water,” Int. J. Hydrogen Energy, 21(7), 1977. 8. Denton Vacuum, Online at: http://www.dentonvacuum.com/coatings/dielectrics.html#six. 9. Huang, C., Adebiy, O., Muradov, N., T-Raissi, A., “Hydrogen Production via UV Photolysis of Aqueous Ammonium Sulfite Solutions,” 16th World Hydrogen Energy Conf., Lyon, France, June 1316, 2006. 10. T-Raissi, A., Muradov, N., Huang, C., Adebiy, O., ”Hydrogen from Solar via Light-Assisted HighTemperature Water Splitting Cycles,” J. of Solar Energy Engineering, in press, 2006.
9/9
Hydrogen production via photocatalytic oxidation of aqueous ammonium sulfite solutions Cunping Huanga, Olwale Odebiyib, Nazim Muradovc, Ali T-Raissid Florida Solar Energy Center, University of Central Florida, Cocoa, Florida, USA [email protected], [email protected], [email protected], [email protected]
a
ABSTRACT: A new hybrid photo-thermochemical sulfur-ammonia (S-NH3) cycle is proposed. The cycle is a modification of the well-known Westinghouse hybrid cycle wherein the electrochemical step is replaced by a photocatalytic step. The new cycle utilizes both IR and visible light portions of the solar spectrum to achieve higher overall solar-to-hydrogen energy conversion efficiencies than those utilizing heat only. The main process, unique to the S-NH3 cycle, is the visible light-induced photocatalytic oxidation of ammonium sulfite (NH4)2SO3) with production of hydrogen and ammonium sulfate ((NH4)2SO4). The (NH4)2SO4 product is processed to generate oxygen and recover NH3 and SO2, which are then recycled and combined with water to regenerate (NH4)2SO3 forming a closed cycle. In this paper, we report our experimental findings with regard to the photocatalytic oxidation of aqueous (NH4)2SO3 solutions containing suspended platinum-doped cadmium sulfide (Pt/CdS) particles irradiated by a xenon arc lamp. We achieved energy conversion efficiencies (absorbed light to chemical energy of hydrogen) of about 12%. The photocatalyst remained active for more than 70 hrs. The experimental results for the photocatalytic step of the S-NH3 cycle describing the hydrogen generation kinetics and effects of temperatures, pH, catalyst preparation and platinum loading on the H2 production efficiencies are also discussed. KEYWORDS: Solar, Hydrogen production, Thermochemical cycles, sulfur ammonia, photocatalysis
INTRODUCTION: Thermochemical water splitting cycles (TCWSCs) can achieve high overall heat-to-hydrogen energy conversion efficiencies [1,2]. Many TCWSCs have been conceived in the past 40 years or so but only few have been extensively studied. For example, the General Atomics Corp. has been developing the sulfuriodine (S-I) cycle [3,4] which belongs to a group of TCWSCs known as the "sulfur family cycles." Most TCWSCs, among them sulfur family cycles, include at least one endothermic reaction/process step that operates at high temperatures and requires a high temperature heat source. Prospective high temperature heat sources suitable for thermochemical process interface include solar thermal concentrator and central receiver systems, and nuclear power plants (i.e. high temperature gas-cooled reactors, HTGR). Solar energy can provide the high temperature heat source needed to drive TCWSCs. The high temperature heat source can be a solar thermal concentrator system. Solar heat source is unique in the sense that it can be concentrated and directed into a solar receiver/reactor system to generate very high reactor temperatures far exceeding those available from the advanced nuclear reactors. Furthermore, solar spectrum contains high energy photons unavailable in conventional and other heat sources. In the paper a solar-powered TCWSC that can utilize both the thermal (i.e. high temperature heat) and the light components (i.e. photonic energy) of the solar resource is discussed. The overall solar-to-hydrogen energy conversion efficiency can potentially be greater than those with heat-only input. The sulfur-ammonia (S-NH3) cycle under development at the Florida Solar Energy Center (FSEC) is a solar powered TCWSC with the following four basic steps [5,6]: SO2(g) + 2NH3(g) + H2O(l) → (NH4)2SO3(aq) (NH4)2SO3(aq) + H2O → (NH4)2SO4(aq) + H2(g) (NH4)2SO4(aq) →2NH3(g) + H2SO4(l) H2SO4(l) → SO2(g) + H2O(g) + 1/2O2(g)
300 K 300 K 525 K 1125 K
1/9
(chemical absorption step) (photocatalytic step) (thermochemical step) (thermochemical step)
(1) (2) (3) (4)
WHEC 16 / 13-16 June 2006 – Lyon France Solar thermal energy is utilized in reactions (3) and (4) for the production of oxygen via decomposition of ammonium sulfate and sulfuric acid. Step (2) is a photocatalytic step in which sulfite ions are oxidized in the presence of VIS-photons and a photocatalyst to sulfate ions and at the same time water is reduced to hydrogen. FSEC’s hybrid photo-thermochemical sulfur-ammonia cycle is a modification of the Westinghouse hybrid cycle [7] wherein the electrochemical step is substituted by a photocatalytic process. Figure 1(A and B) shows a schematic diagram of the S-NH3 cycle.
Figure 1. Schematic flow diagram of solar powered S-NH3 thermochemical water splitting cycle. With reference to Figure (1), the thermal portion of the solar resource (i.e., near infrared NIR, and infrared, IR) is resolved by the receiver/photoreactor units (see Figure 1(B)) in the mirror field and concentrated into a high temperature thermocatalytic reactor located within the tower for the decomposition of (NH4)2SO4 to produce oxygen. The photonic (UV and visible light) portion of the solar resource is utilized for the hydrogen production via photocatalytic oxidation of (NH4)2SO3 into (NH4)2SO4 while water is reduced into hydrogen. The utilization of both thermal heat and photonic energy increase the solar to hydrogen energy conversion efficiency of the cycle. The resolution of the UV and visible light from NIR and IR portions of the solar spectrum can be accomplished by application of the broad-band antireflection coatings which provide very low reflectance over a broad range of wavelengths within the UV and visible spectra [8]. With this approach, photonic portion of the solar radiation can be made available for conducting photocatalytic hydrogen production step of the S-NH3 cycle. In the reminder of this paper, we primarily focus on the photocatalytic hydrogen production step of the S-NH3 cycle (Reaction (2)) described above. EXPERIMENTAL: Figure 2 depicts the experimental set-up employed in this work. Photocatalytic experiments were conducted using a 1000-Watt xenon arc lamp (Spectral Energy) as the light source. A water tank was used to filter out infrared radiation generated by the lamp. A 250 mL, custom-made borosilicate glass reactor was used as the reaction vessel. The reactor temperature was controlled by a water circulation jacket connected to a Polyscience MDL 612 analog recirculator. The reaction solution was stirred by a magnetic stirrer. A PyrexTM glass window at the front of the photoreactor facing the light source allowed absorption of the near-UV and visible light radiated from the xenon arc lamp only. Hydrogen was collected in a 1000 ml Pyrex glassmeasuring cylinder. On line pH measurements were carried out with a calibrated Oakton 1100 pH meter. Prior to each photolytic experiment, the reactor was purged with ultra-pure grade argon gas for two hours to flush out air/oxygen dissolved in the reacting solution and trapped gas inside the photoreactor. Spectral 2/9
WHEC 16 / 13-16 June 2006 – Lyon France power output from the Xe lamp was measured using a SHIMADZU UV-2401 spectrophotometer. Reagent grade ammonium sulfite ((NH4)2SO3.H2O, Aldrich) was used as received without further purification. Reagent grade cadmium sulfide (CdS) (99.999% purity, Aldrich) was used as photocatalyst. Commercial chloroplatinic acid (H2PtCl6) (Aldrich) was used for the preparation of Pt –doped photocatalyst particles (Pt/CdS).
Glass Window
Figure 2. Experimental set-up for H2 production via photocatalytic oxidation of (NH4)2SO3 aqueous solution. 20.00
Spectral Irradiance (W/m2/nm)
18.00
Arc Arc/Glass
16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 300
400
500
600
700
800
900
1000
1100
Wavelength (nm)
Figure 3. Spectral power output for xenon short arc lamp (1000 W). Light intensity measurement: Spectral irradiance from the xenon arc lamp was measured using a calibrated LI-1800 portable spectroradiometer. The measured values of the spectral irradiance for the xenon light source used are given in Figure 3. Light measurements were carried out both with and without a Pyrex glass window in front of the detector. Pyrex window served to filter out most of the UV light emitted by the lamp so that an emission spectrum closer to that of solar can be simulated. The UV content of solar radiation is only about 4%, which is less than that produced from a xenon arc lamp. Table 1 shows the measured spectral irradiance within various wavelength ranges for the Xe light source used. Table 1. Irradiance from xenon lamp at various wavelength ranges. Wavelength Light Intensity Intensity Power (W) (nm) (W/m2) Average (W/m2) 300-350 97 60 0.8 350-400 213 131 1.8 400-450 316 195 2.6 450-500 413 255 3.4 500-600 730 451 6.1 600-700 703 435 5.8 700-800 546 337 4.5 Total 3018 1865 25.1
3/9
WHEC 16 / 13-16 June 2006 – Lyon France The light intensity was also measured by using a pyranometer (LICOR LI-200) was placed at 12 inches and 25 inches away from the xenon arc lamp. The irradiance was measured at 9 different locations on the elliptically lit image formed by the collimated light. The total radiant power incident on the reactor aperture was determined to be about 25.07 W. The photocatalyst employed in this experiment is cadmium sulfide with a band gap of 2.4 eV, which is activated with light having wavelengths of 516 nm or shorter. We used the total radiant power with wavelengths in the range of 300 nm – 500 nm that falls on the photoreactor was measured to be 8.60 watts as the value for calculating both the process energy efficiency (incoming light energy to chemical energy stored in hydrogen) and the quantum efficiency. UV-Vis spectral analysis: A UV-VIS spectrophotometric analysis was conducted using a SHIMADZU UV2401) spectrometer on six different samples to determine the composition of the reacting solution and the concentration of ammonium dithionate (Na2S2O6), a dimer of ammonium sulfite. Samples used for the analysis were, 0.5M Na2SO3, 0.5M Na2SO4, 0.5M Na2S2O6 (sodium dithionate), 0.5M (NH4)2SO3, 0.5M (NH4)2SO4, and a reacted solution (after 64 hrs of irradiation). It was observed that there was virtually no cation effect on the absorbance of the solutions. Both ammonium sulfite and sodium sulfite solutions had almost identical UV-Vis spectra. Figure 4 depicts that spectra for the reacted solution falls between those of the sulfite and dithionate solutions. This suggests that after 64 hrs of light exposure, the reacted solution contains a mixture of sulfate, sulfite and dithionate ions. HPLC measurements were used to quantify concentration of the sulfate, sulfite and dithionate ions formed. 5.0 4.5
0.5M solutions
4.0 SO32-
Absorbance
3.5 3.0 Reacted Solution SO32- S2O62- SO32-
2.5 2.0 1.5 1.0
S2O62SO42-
0.5 0.0 200
225
250
275
300
325
350
Wavelength (nm)
Figure 4. UV-Visible spectra of sulfate, sulfite and dithionate ions. RESULTS AND DISCUSSION: 1. Effect of temperature: We have previously shown that temperature plays a significant role in the UV photolysis of aqueous (NH4)2SO3 solutions for hydrogen production [9]. Similar effect was found for the photocatalytic (Pt/CdS) process. Figure 5 depicts that the rate of hydrogen production increases in concert with the increase in temperature from 50oC to 70oC. However, the rate of hydrogen production drops as temperature is further increased from 70oC to 90oC. It should also be noted that pH of the solution is affected by the temperature. Increasing photolyte temperature to 90oC lowers solution pH, which in turn changes the rate of hydrogen evolution. At a given temperature, there is an optimum solution pH for which the rate of hydrogen production is the highest. 2. Effect of CdS platinum loading: Platinum metal loading of the CdS catalyst improves the rate of (NH4)2SO3 photocatalytic oxidation. However, if too much platinum is deposited on the photocatalyst surface, it can block the light from reaching the CdS surface. Experiments were conducted at four different CdS photocatalyst platinum loadings. Figure 6 depicts H2 production results for various Pt loadings. Figure 6 also shows that the highest rate of hydrogen production was achieved at a platinum loading of 1.0 wt.% (with respect to CdS). Decreasing the amount of Pt loading to 0.5%wt or increasing Pt loading to 2%wt (or higher) lowered hydrogen production rates. 4/9
WHEC 16 / 13-16 June 2006 – Lyon France
1400 1200
50C 70C 90C
Hydr ogen (mL
1000 800 600 400 200 0 0
50
100
150
200
250
300
350
400
450
T ime (min)
Figure 5. Effect of temperature on the rate of photocatalytic hydrogen production from (NH4)2SO3. Figure 7 shows the relationship between the rate of hydrogen production and Pt loading, indicating that the optimal Pt loading is approximately 1 wt. %. 3. Effect of (NH4)2SO3 concentration: Figure 8 depicts the effect of ammonium sulfite concentration on the rate of hydrogen production. It can be seen that as the solution concentration of (NH4)2SO3 increases from 0.432M to 0.866M, the rate of hydrogen also increases. But, as [(NH4)2SO3] is increased beyond 0.866M to 1.731M, the rate of H2 production remains unchanged and equal to that of 0.866M solution. This result implies that there is an optimum [(NH4)2SO3] for which photocatalytic hydrogen production rate is maximum. 250 0.50% 1%
Hydrogen (mL)
200
2% 3%
150
100
50
0 0
50
100
150
200
250
300
350
Tim e (m in)
Figure 6. Effect of platinum loading (percentage by weight of CdS) on the hydrogen production rate.
5/9
WHEC 16 / 13-16 June 2006 – Lyon France
1.6 1.4
Rate (mL/m in)
1.2 1 0.8 0.6 0.4 0.2 0 0
0.5
1
1.5
2
2.5
3
3.5
Cd S Platin um Lo adin g %
Figure 7. Optimum level of the CdS platinum loading on the rate of hydrogen production from (NH4)2SO3. 800 700 0.432 M 0.866 M 1.731
Hydrogen (mL)
600 500 400 300 200 100 0 0
50
100
150
200
250
300
350
400
Reaction Time (min)
Figure 8. Effect of ammonium sulfite concentration on the rate of photocatalytic hydrogen production. 4. Process efficiency: Thermocatalytic decomposition of SO3 is an extensively studied process with the thermal efficiency as high as 70%.[3,4,10]. Also shown is the total solar to hydrogen efficiency of the S-NH3 thermochemical cycle being primarily dependent on the efficiency of the photocatalytic step (Reaction 2). Here, we define the process efficiency for the hydrogen generating Reaction (2) as the ratio of the higher heating value (HHV= 285.9 kJ/mol) of H2 to the total photonic energy (absorbed) input to the process in the wavelength range of 300 to 500 nm. The rationale for using this wavelength range is dictated by the availability of photons in solar spectrum (in the lower limit) and absorption of VIS photons by CdS (in the upper limit). . IR region of the radiation is used to carry out the high temperature oxygen production step. Based on the highest hydrogen production rate achieved so far (i.e., 3.2 mL/min over an 8-hr period), the overall photon-to-H2 efficiency was determined to be about 12.0%.
6/9
WHEC 16 / 13-16 June 2006 – Lyon France
5. Some mechanistic considerations: The kinetic measurements were carried out using 0.500 g of Pt/CdS (1.0 wt%) suspended in a 0.866 M (NH4)2SO3 solution (initial pH of 7.9) under isothermal conditions at 35oC. It took 32 hrs for the photocatalytic reaction to reach completion at which time H2 evolution had seized. Figure 9 depicts the amount of H2 generated and solution pH as a function of time during the photocatalytic oxidation of (NH4)2SO3. It can be seen that the rate of H2 production from photocatalytic conversion of (NH4)2SO3 for the first 25 hrs is a linear function of time yielding a constant hydrogen production rate of about 1.79 mL/min. The H2 production rate begins to diminish after 1500 minutes and levels off at about 2000 minutes of reaction time. Data for the online pH measurement showed that the solution pH increased slightly from 7.9 to 8.6 and then maintained constant until H2 production stopped. The increase in the solution pH can be attributed to the formation of dithionate ions in the solution [9]. To ascertain formation of the S2O62- ions in the solution and determine the initial and end products of the reaction, a high performance liquid chromatograph (Dionex DX-500 HPLC) was employed and used to measure the concentration of the sulfur species in the aqueous solution before and after the photocatalytic reaction. The results showed that before the photocatalytic reaction began, both (NH4)2SO4 and (NH4)2SO3 were present in the solution. This indicates that some of the as received (NH4)2SO3 material had already been oxidized to (NH4)2SO4. The proposed mechanism of photocatalytic oxidation of (NH4)2SO3 was found to be similar to that of direct UV photolysis as described in [9]. The difference between these two processes being, at the end of reaction, the solution pH for the photocatalytic process does not drop precipitously (see Figure 9) - as a result of S2O62- being converted to SO42- ions. This implies that S2O62- ions are more stable at the conditions of the photocatalytic experiment (at 35oC). In contrary, we have shown that during UV photolysis of (NH4)2SO3 at 50oC, intermediately formed S2O62 ions that are converted into SO42-. It should be noted that at elevated temperatures, photocatalytically formed S2O62- can also be converted to SO42- ions.
Figure 9. Amount of hydrogen produced and solution pH as a function of time during photocatalytic hydrolysis of ammonium sulfite. Based on the extensive literature data on the CdS photocatalysis and our experimental observations, we tentatively propose the following mechanism for the photocatalytic conversion of ammonium sulfite Upon UV-VIS illumination, electron-hole pairs are generated on the surface of CdS particles: CdS + hν → h+ + e-
(5)
Electrons (e-) reduce protons to generate hydrogen, while SO32- ions are oxidized by the holes (h+) forming SO42- ions. These processes are shown in Reactions 6 to 8 below: SO32- + H2O + 2h+ = SO42- + 2H+ 2H+ + 2e- = H2 Overall: SO32- + H2O = SO42- + H2
(6) (7) (8) 7/9
WHEC 16 / 13-16 June 2006 – Lyon France
As noted earlier, the increase in the solution pH can be associated with the formation of S2O62- ions. HPLC measurement of S2O62- ion concentration in the photocatalyzed sample of (NH4)2SO3 solution taken after 32 hrs showed an S2O62- ion peak. Formation of S2O62- ions was also observed during UV photolysis of (NH4)2SO3 as a reaction intermediate as shown in Figure 4. The following reactions could possibly lead to the formation of S2O62- : 2SO32- = S2O62- + 2e2H2O + 2e- = H2 + 2OH2SO32- + 2H2O = S2O62- + H2 + 2OH-
(oxidation) (reduction) (overall reaction)
(9) (10) (11)
Table 2 depicts the material balance for the 32-hr photocatalytic hydrolysis of (NH4)2SO3. Total of 123.4 mmol of sulfite ions were consumed during the reaction. In solution, 47 mmol of sulfate ions and 69 mmol of dithionate ions were generated. The difference between the amount of SO32- ions consumed and the sum of SO42- and S2O62- ions produced was 5.2%. In the gas phase, 2,651 mL of hydrogen was produced over the 32 hr period. The amount of hydrogen measured was 7.3% less than the theoretical value predicted based on Reactions 8 and 11. These discrepancies are well within the experimental margin of error. Table 2. Material balance for the photocatalytic process of aqueous (NH4)2SO3 solution. Species Before reaction After reaction Produced Consumed (mmol) (mmol) (mmol) (mmol) 123.4 0.0 0.0 123.4 SO3239.6 87.0 47.4 0.0 SO420.0 70.0 69.6 0.0 S2O62Theoretical H2 Measured Difference Difference H2 Production (mL) (mL) (mL) (%) 2,859 2,651 208 7.3 CONCLUSIONS: A new hybrid photo-thermochemical cycle - sulfur-ammonia (S-NH3) cycle has been proposed. The cycle is a modification of the Westinghouse hybrid cycle but with the electrochemical step replaced by a photocatalytic process. The step, unique to the S-NH3 cycle, is the light-induced photocatalytic oxidation of ammonium sulfite ((NH4)2SO3) with production of hydrogen. In this work, we report the experimental findings with regard to the Pt doped CdS photocatalytic oxidation of aqueous (NH4)2SO3 solutions irradiated by a xenon arc lamp. Results show that the concentration of (NH4)2SO3 in the solution and reaction temperatures had a major influence on the rate of hydrogen production. It was found that the rate of H2 production can be maximized with respect to both (NH4)2SO3 concentration and reaction temperature. The effect of Pt loading on photocatalytic activity of CdS was also determined. The data indicate that CdS with 1.0 wt% of Pt loading yields the highest rate of hydrogen production. From the kinetic and material balance experiments conducted, it was found that SO32- ions can be completely converted into SO42- and S2O62- ions. The S2O62ions formed during the photoreaction can be transformed to SO42- ions at elevated temperatures. The overall light-to-hydrogen conversion efficiency of the photocatalytic conversion of (NH4)2SO3 was determined to be about 12%. ACKNOWLEDGMENT: Authors are grateful for the financial support of this work to the National Aeronautics and Space Administration (NASA) through Glenn Research Center (GRC) under Grant NAG3-2751 and to Mr. Timothy Smith, the Grant Program Manager at NASA-GRC. REFERENCES: 1. Funk, J.E., "Thermochemical Hydrogen Production: Past and Present," Int. J. of Hydrogen Energy, 26(3), 185-190, 2001. 2. Funk, J.E., & England, C., "Solar Thermochemical Hydrogen: The Heat Source-Process Interface," Int. J. of Hydrogen Energy, 10(6), 359-64, 1985. 3. Besenbruch, G.E., “General Atomic Sulfur-Iodine Thermochemical Water Splitting Process,” Am. Chem. Soc., Div, Pet, Chem. Preprint, 271(48), 1982.
8/9
WHEC 16 / 13-16 June 2006 – Lyon France 4. Brown, L. C., Funk, J.F., Showalter, S.K., “High Efficiency Generation of Hydrogen Fuels Using Nuclear Power,” Annual Report to the U.S. Department of Energy GA-A3451, July 2000. 5. Huang, C., T-Raissi, A., "Analysis of Sulfur–Iodine Thermochemical Cycle for Solar Hydrogen Production. Part I: Decomposition of Sulfuric Acid," Hydrogen Energy Special Issue, Solar Energy, A. T-Raissi & S.A. Sherif (Editors), 78(5), 632-46, 2005. 6. Huang, C., T-Raissi, A., "A New Sulfur Ammonia Thermochemical Water Splitting Cycle," Proc. of the 15th World Hydrogen Energy Conf., Yokohama, Japan, 2004. 7. Brecher, L.E., S. Spewock, et al., “Westinghouse Sulfur Cycle for the Thermochemical Decompoistion of Water,” Int. J. Hydrogen Energy, 21(7), 1977. 8. Denton Vacuum, Online at: http://www.dentonvacuum.com/coatings/dielectrics.html#six. 9. Huang, C., Adebiy, O., Muradov, N., T-Raissi, A., “Hydrogen Production via UV Photolysis of Aqueous Ammonium Sulfite Solutions,” 16th World Hydrogen Energy Conf., Lyon, France, June 1316, 2006. 10. T-Raissi, A., Muradov, N., Huang, C., Adebiy, O., ”Hydrogen from Solar via Light-Assisted HighTemperature Water Splitting Cycles,” J. of Solar Energy Engineering, in press, 2006.
9/9
