- Email: [email protected]
Preformance analysis of a water splitting reactor with hybrid photochemical conversion of solar energy
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/he
Preformance analysis of a water splitting reactor with hybrid photochemical conversion of solar energy E. Baniasadi*, I. Dincer, G.F. Naterer Faculty of Engineering and Applied Science, University of Ontario Institute of Technology (UOIT), 2000 Simcoe Street North, Oshawa, ON, Canada L1H 7K4
article info
abstract
Article history:
In this paper, a new hybrid system for hydrogen production via solar energy is developed
Received 23 September 2011
and analyzed. In order to decompose water into hydrogen and oxygen without the net
Received in revised form
consumption of additional reactants, a steady stream of reacting materials must be
25 January 2012
maintained in consecutive reaction processes, to avoid reactant replenishment or addi-
Accepted 26 January 2012
tional energy input to facilitate the reaction. The system comprises two reactors, which are
Available online 27 February 2012
connected through a proton conducting membrane. Oxidative and reductive quenching pathways are developed for the water reduction and oxidation reactions. Supramolecular
Keywords:
complexes [{(bpy)2Ru(dpp)}2RhBr2] (PF6)5 are employed as the photo-catalysts, and an
Hydrogen production
external electric power supply is used to enhance the photochemical reaction. A light
Solar energy
driven proton pump is used to increase the photochemical efficiency of both O2 and H2
Photocatalytic method
production reactions. The energy and exergy efficiencies at a system level are analyzed and
Hybridization
discussed. The maximum energy conversion of the system can be improved up to 14% by
Exergy
incorporating design modification that yield a corresponding 25% improvement in the
Efficiency
exergy efficiency. Copyright ª 2012, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Solar driven water splitting combines several attractive features for sustainable energy utilization. The energy source of the sun and the reactive media of water are readily available and renewable, and the resultant fuel of hydrogen and product of water are each environmentally clean. Fujishima and Honda [1] reported on hydrogen production from photocatalytic water decomposition by using a TiO2 single crystal electrode. The authors examined the processes of directly transforming solar energy into chemical energy. Numerous other past studies have been conducted on the photodecomposition of water as a clean solar energy conversion process leading to a non-polluting fuel.
Solar energy cannot be directly stored or continuously supplied. Therefore, the conversion of solar energy to a type of storable energy has crucial importance. To utilize a low density solar flux as effectively as possible, all wavelengths of light should be used, and the efficiency of each step of the energy conversion steps should be improved. Another alternative for hydrogen production by solar energy without consumption of additional reactants is a hybrid system. A hybrid system which combines photochemical, thermochemical and electrochemical reactions has been reported previously [2]. Compared to hydrogen production methods based on fossil fuels, the high investment cost of solar hydrogen generation is a challenging issue. Production of hydrogen in a regenerative
* Corresponding author. E-mail addresses: [email protected] (E. Baniasadi), [email protected] (I. Dincer), [email protected] (G.F. Naterer). 0360-3199/$ e see front matter Copyright ª 2012, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2012.01.128
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
fashion such as photochemical splitting of water has been studied [3]. The reductive side of this process requires the development of catalysts that promote the reduction of protons to molecular hydrogen, facilitated by direct excitation by a photosensitizer. Recently, molecular platinum- and palladium-based systems have been developed as heterogeneous catalysts [4,5]. Supramolecular complexes in a photocatalytic hydrogen production scheme that results in high turnover rates and numbers are another area of interest [6]. The long-lived metal-to-ligand charge-transfer (MLCT) excited state of [Ru (bpy)3]2þ has motivated photochemical and photophysical studies leading to light to energy conversion processes [7,8]. The MLCT excited state of [Ru (bpy)3]2þ and its analogs have the required energy to split water into hydrogen and oxygen, but it requires complicated multi component systems for operation. One of the most promising areas is development of photochemical water splitting using supramolecular devices, which are able to capture the incident solar radiation, and generate electrons or holes at the active center where water reduction or oxidization occurs, respectively. Such systems mimic natural photosynthesis and mainly consist of supramolecular complexes of organic molecules which possess active metallic centers [9]. The water photo-oxidizing reaction occurs as follows: hv
2H2 Oð1Þ / 4Hþ þ 4e þ O2 ðgÞ; Vox ¼ þ0:82 V; NHE:
(1)
Following this reaction under the influence of photonic radiation, the active center absorbs two electrons from one water molecule. The water reduction proceeds according to: hv
2H2 O þ 2e / H2 þ 2OH ; Vred ¼ 0:41 V; NHE:
(2)
For a complete cycle, two water molecules are required. Thus, the total cell potential, E, is written as E ¼ E1 þ E2 ¼ 0.83 V to 0.40 V ¼ 1.23 V. Numerous studies on photochemical production of hydrogen have focused on the development of photocatalytic materials. Relatively few have examined thermodynamic studies of photon-to-electron conversion, theoretical power conversion efficiency, or energy and exergy efficiencies. This paper extends the analysis of a newly proposed photochemical water splitting system by C. Zamfirescu et al. [10] by implementing new modifications to improve the system performance. These modifications mainly consist of hybridization of a photocatalytic system by an external electric source, implementation of a light driven proton pump, and efficient photosensitizers. This paper investigates the performance of a photochemical hydrogen generation reactor through energy and exergy methods, and also comparisons with other water splitting methods.
2.
System description
A photochemical reactor comprises a radiation source, a reactor transparent for the light, a gas product outlet and a photocatalytic reactor. As shown in Fig. 1, the vessel comprises two photochemical reactors separated by a light driven proton pump membrane. Similar to natural photo-
7465
Fig. 1 e Schematic of water splitting photochemical reactor with external power source.
systems, an artificial water splitting photo-system will require a number of essential molecular components that must be organized in supramolecular complexes such that the photo-generated electron-hole pair is quickly and efficiently separated and their potential energy is delivered to functional co-catalysts. The system contains a minimum number of necessary components such as a photosensitizer, oxygen and hydrogen evolution reaction catalysts. An efficient and robust sensitizer absorbs photons and generates long-lived electronhole pairs that are coupled to appropriate multi-electron cocatalysts for oxygen and hydrogen evolution. Photolysis solutions consist of [{(bpy)2 Ru (dpp)}2 RhBr2] (PF6)5 (65 mM), water (0.62 M) acidified to pH ¼ 2 with triflic acid, and dimethylaniline (1.5 M) in a solution of acetonitrile (4.46 mL). In the water photo-oxidation reactor, on the left, selected supramolecular photo-catalysts are used to capture light energy and generate electrical charges at reaction sites to oxidize water and produce oxygen gas and protons. Fresh water is continuously supplied to this reactor, and the flow rate of fresh water is adjusted such that the water level in the vessel remains constant. In the water reduction reactor, selected supramolecular complexes for photocatalytic reduction of water to hydrogen are dissolved in a proper concentration. These catalysts generate photoelectrons, which makes hydrogen evolution reactions occur. An auxiliary electric circuit has been considered to provide an external electric potential and improve the photocatalytic reaction performance. Negative charges, generated by a power supply, are donated at the electrode surface to the supramolecular devices, and these are transmitted to the reaction sites under the influence of photonic radiation. Two extraction fans are used above the liquid level, to extract oxygen and hydrogen, continuously. As mentioned previously, this research will focus on the use of Ru(II)- bipyridyl complexes as sensitizers to drive the water splitting reaction. As seen in the modified Latimer diagram, shown in Fig. 2, each cycle starts with photoreactions to capture the light energy (at w 450 nm or 2.75 eV) to form the excited single molecule [Ru (bpy)3]2þ**. This molecule
7466
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Fig. 2 e Photoreactions based on reductive and oxidative quenching of the photo-excited.
rapidly loses some energy upon intersystem crossing to the triplet state, [Ru (bpy)3]2þ* .The photo-excited complex, denoted [Ru (bpy)3]2þ* or more simply Ru2þ*, is both a good oxidant and a good reductant. Electron transfer either to or from the excited state complex traps the ruthenium complex as [Ru (bpy)3]þ or [Ru (bpy)3]3þ, respectively. These photoproducts are potent reductants ([Ru (bpy)3]þ) and oxidants ([Ru (bpy)3]3þ which can drive subsequent redox reactions. These various species can be used to drive a catalytic reaction by either an oxidative or reductive quenching pathway. The energetic of the water splitting reaction within these two manifolds is also shown in Fig. 2. The reactions are normalized to a 4-photon, 4-electron process, and a pH of 7.0 is assumed. EA and ED refer to generic electron acceptor and donor molecules. One of the challenges regarding oxidation and reduction mechanisms involves replenishment of electron acceptors and donors. Consumption of EDs will require replenishment of this material either by a sustainable process reaction that caused the ED to reform, or by extraction and replacement of the ED. Our new proposed approach will apply an external power source and two electrodes immersed in the catalyst solution to supply and transfer electrons inside two reactors. The equivalent circuit of the scheme shown in Fig. 3 should correspond to that of the schematic in Fig. 1. The electric potential generated by the photosensitizers of both oxygen and hydrogen production reactors is a minimum of 2.62 V NHE. The generated energy is used to drive the half reactions, which consume 0.82 V and 0.41 V, respectively. The remaining electrical energy is used to overcome various losses including ohmic, concentration and activation loses at the electrodes. In order to have a self-sustained system, the total loss should be less than 1.39 V. The presence of ionic molecular devices at the electrode interface, where they exchange electrons (donate or accept) with high activity, dispel the need for coating the electrode surface by expensive catalysts like platinum group materials. This research uses porous graphite electrodes [11]. The calculated cell potential (1.23 V) is the minimum voltage needed for electrolysis. In practice, however,
electrolysis generally starts at a somewhat higher potential to overcome the overpotential and ohmic losses needed to form hydrogen and oxygen [12]. In past experimental data [11], a stream of fine gas bubbles started to form at the anode and the size of these oxygen bubbles gradually increased with potential until w1.8 V. The first activity at the cathode was a stream of fine bubbles (formed at w2.3 V). Although the decomposition potential for oxygen and hydrogen are slightly different, the practical decomposition potential is nearly the same (w2.3 V), regardless of the immersed area of the electrodes. Another approach to extend the conceptual design in Ref. [10] is use of light driven proton pumps instead of conventional proton exchange membranes. This paper uses a slightly modified version of a system in Ref. [13], shown in Fig. 4. The proposed proton pump comprises a reaction center,
Fig. 3 e Equivalent schematic of the electric circuit for water splitting reactors.
7467
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
form hydrogen or oxygen. This translates to two important efficiencies regarding photochemical and electrochemical performances of both water oxidation and reduction reactors. Define the photochemical efficiency (hPC) of quenching pathways as the energy efficiency of this process, which is part of the transmitted irradiation contributed to the water oxidation/reduction reaction divided by the photonic energy. Referring to Fig. 2, the efficiencies of the catalytic cycle for water oxidization and reduction at 450 nm are as follows:
which is a molecular triad containing an electron donor and an electron acceptor, both linked to a photosensitive porphyrin group. The triad molecule is inside the bilayer of a liposome. The molecular triad absorbs a photon and translocates a negative charge near the outer surface and a positive charge near the inner surface of the liposome by generating charge separated species. Following these interactions, protons will be transported across the membrane from the lower proton potential to the higher proton potential side of the membrane, e.g. oxygen and hydrogen production reactors, respectively. The transport of protons in this system is mainly due to combined effects of photo-induced energy transduction within the molecular structure of membrane and electroosmotic forces applied at the two porous electrodes. The anode acts as a proton source and the cathode as a proton sink. The conversion of light energy into the electrochemical gradient of protons across the membrane can be quantitatively characterized by the quantum yield (or quantum efficiency) B of proton translocation. This parameter is defined as follows: B¼
number of protons pumped number of photons absorbed
(3)
Based on a case study in Ref. [13], in 1 ms, the shuttle makes nearly 16 trips and transfers ten protons through the membrane, provided that the light intensity I ¼ 0.133 mWcm2. Considering the number of photons absorbed in 1 ms is w18, the approximate quantum yield of the pumping process is w55%. The effect of this module on system performance will be discussed in the following section.
1:53 ev ¼ 55:5% 2:75 ev
(4)
hEC ¼
1:08 ev ¼ 39:3% 2:75 ev
(5)
The electrochemical efficiency (hEC) of water reduction has a similar definition, as the input energy is supplied by external electric power. There are several uncertainties related to overpotential values, or in other words, an effective contribution of the external electric potential in the decomposition potential. Therefore, this study conducted the energy and exergy analysis for different electrochemical efficiencies. The maximum energy conversion efficiency of the hybrid photochemical water splitting system is defined, based on the higher heating value (HHV) of hydrogen, as follows: hmax H2 ¼
n_ H2 ¼
Results and discussion
As shown in Fig. 2, only part of the energy captured by light will be used by an excited singlet molecule [Ru (bpy)3]2þ* to
(6)
WL hPC þ WEL hEC DEH2
(7)
and DEH2 ¼ 1060 kJ/mol, equivalent to the minimum required energy captured from light (WL at w 450 nm or 2.75 eV) or external electric power (WEL) to form the excited singlet molecule [Ru (bpy)3]2þ**. Fig. 5 shows the effect of the delivered potential at the electrode interface to electrolyte on the maximum energy conversion efficiency. The photochemical efficiency is assumed to be hpc ¼ 0.55 and this study considered different electrochemical efficiencies. The photolysis is conducted
0.2 0.18 0.16 0.14 0.12 0.1 0.08
3.
HHVH2 n_ H2 WL þ WEL
where n_ H2 is the maximum possible molar flow rate of produced H2,
Maximum Energy Conversion Efficiency
Fig. 4 e Schematic diagram of the light-induced proton pump across the lipid bilayer in a liposomic membrane. A molecular triad A-BC-D is symmetrically inserted in the lipid bilayer. “A” refers to acceptor, “D” for donor, and “S” for shuttle.
hPC ¼
0
1
2
3
4
5
External Voltage (V)
Fig. 5 e Effect of external electric potential on maximum energy conversion efficiency.
6
7468
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
using a 5W LED having an excitation wavelength of 470 nm. It has been observed that a hybrid configuration of water splitting system may improve the efficiency if the electrochemical efficiency is higher that the photochemical efficiency. The reason is a very low quantum efficiency of LEDs, which was taken as 1.3% in this analysis. In other words, the auxiliary power supply will be helpful if it delivers excitation energy more efficiently than the photonic source. A maximum 30% solar conversion efficiency improvement is estimated by applying 5 V electric potential. Evaluating the energy and exergy efficiencies of the hydrogen production reactor requires the hydrogen production rate from a correlation ðn_ H2 ¼ 2:725mmol=h; 0 < t < 5hÞ based on a recent study by Brewer et al. on a new photocatalyst as [{(bpy)2 Ru(dpp)}2 RhBr2] (PF6)5 [9]. Our investigation involved the determination of the energy and exergy efficiency by varying DC current density at a low voltage (0e6 V) using graphite electrodes and electrolyte concentration of 65 mM of aforementioned molecular complexes in 0.62 M water. The definition of energy efficiency is similar to the maximum energy conversion efficiency, Eq. (6), but the molar production rate of hydrogen is substituted by the value according to the aforementioned correlation. Fig. 6 shows the effect of external voltage on the energy efficiency. Compared to the electrolytic load, the resistive load of the cell is relatively high, yielding a very small current density which slowly changes linearly with increasing potentials (<2 V). Its effect on input electric power is shown in Fig. 6. At voltages higher than 2.3 V, the current density increases very rapidly due to the redox reactions. Thus, above the decomposition potential, the current density consists of the ohmic current density plus the current density produced by the redox reaction. Up to certain energy efficiency it doesn’t differ, but as the applied potential goes above 2.3 V, the efficiency increases rapidly and then decreases following higher concentration and ohmic losses. Almost a 27% improvement in energy efficiency is achieved over 6 V external electric potential. The energy efficiency of photochemical water splitting reactors without any external power is about 0.33%. The absorbed photonic radiation plus electric energy drives an endothermic chemical reaction. The measure of how these
0.44
sources of energy are converted into chemical energy stored in H2 for a given process is the exergy efficiency is defined as hexergy ¼
n_ H2 DGH2 þ0:5O2 /H2 O WL þ WEL
(8)
where n_ H2 is the actual molar flow rate of produced H2, based on a correlation extracted from experimental data [9], and DG is the standard Gibbs free energy change of the reaction at 298 K (237 kJ/mol), i.e., the maximum possible amount of work that may be extracted from H2 at 298 K, when both H2 and O2 are available at 1 bar. Fig. 7 shows the calculated exergy efficiencies over a range of applied electric potential, and it shows similar trends to energy efficiency. About 30% enhancement in exergy efficiency is calculated. If the auxiliary power supply is unplugged, this efficiency will be about 0.55%. The solution pH should is a major factor that can be tuned to favor hydrogen reduction or water oxidation. Appling the Nernst equation for the two manifolds yields: Reductive quenching pathway: DGOER ¼ 84 22:85ðpHÞðkJ=molÞ
(9)
DGHER ¼ 420 þ 22:85ðpHÞðkJ=molÞ
(10)
Oxidative quenching pathway: DGOER ¼ 116 22:85ðpHÞðkJ=molÞ
(11)
DGHER ¼ 220 þ 22:85ðpHÞðkJ=molÞ
(12)
From these expressions, an acidic solution oxidative quenching pathway is more exothermic during the hydrogen evolution process than a reductive pathway, which implies higher photochemical efficiency according to Eq. (4). Similarly, the oxygen evolution energy in the oxidative pathway is higher in a basic solution. Hence, the oxidative quenching pathway is selected in our design in order to take advantage of the light driven proton pump. Also, the pH range in which water splitting can occur should be determined, assuming the two half reactions are not energetically coupled. The free energy of the Hydrogen Evolution Reaction (HER) is not used to
200
0.75
200
0.7
160
120
0.36 80
0.34 0.32
40
0.3 0.28
0 0
1
2 3 4 External Voltage (V)
5
6
Fig. 6 e Energy efficiency and supplied electric energy versus external voltage.
0.65 120 0.6 80 0.55 40
0.5
0
0.45 0
1
2
3
4
5
6
External Voltage (V)
Fig. 7 e Exergy efficiency and supplied electric energy versus external voltage.
Electrical Energy (kJ)
0.38
Exergetic Efficiency
160
0.4
Electrical Energy (kJ)
Energetic Efficiency
0.42
7469
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Fig. 9 e Effect of external power supply on exergy efficiency.
energy for all four cases is assumed to be equal and working conditions are consistent. Although it couldn’t reach the efficiency of a photoelectrochemical system, but hybridization of photochemical processes is promising. Hybridization made an acceptable improvement in the photochemical performance, and it shows almost a 30% increase in efficiencies. Compared to conventional electrolysis, solar based methods are relatively deficient, and they need more efforts for commercialization purposes. Low quantum efficiency of current photo-catalysts is a crucial challenge to be addressed in order to decrease energy and exergy destructions and losses. The effects of [Ru (bpy)3]2þ, Eosin Y, and Cuphthalocyanine, as different photosensitizers types for hydrogen production photoreactions, on energy and exergy efficiencies, are shown in Fig. 11. Their effectiveness occurs in the following order: Cu-phthalocyanine > [Ru (bpy)3]2þ > Eosin Y. The difference in the hydrogen production rate is based on the structures and properties of these dyes and the differences in their electron injection characteristics. The energy conversion efficiency calculated for the proposed system of this study is compared and qualitatively validated with some experimental data on photo- or solarelectrolysis efficiencies, as contained in Table 1. Although some of the numbers claimed have been questioned by others,
Efficiency (%)
help drive the Oxygen Evolution Reaction (OER) or vice versa. These energies and pH ranges are specific for [Ru (bpy)3]2þ, assuming a four-photon process but they are tunable by modifying the complex structure. Figs. 8 and 9 represent the maximum energy conversion and exergy efficiencies of oxygen and hydrogen production reactors versus electrical potential at the electrode interface, considering the effect of the solution pH difference between reactors. The case DpH ¼ 0 implies that a photolysis solution of pH ¼ 7 and increasing the pH difference makes hydrogen and oxygen quenches acidic and basic, respectively. It was found that the hydrogen production volume in the acidic medium exceeds the basic medium. At an acidic pH, more Hþ ions would be absorbed by the [Ru (bpy)3]2þ*, and hence the possibility for reduction of Hþ to H2 increases. Increasing the proton concentration in the catalyst solution of the hydrogen evolution manifold could lead to about 10% improvement in maximum energy conversion efficiency, and 13% in exergy efficiency. All graphs meet at a point corresponding to 2.65 V where efficiencies have an opposite trend before and afterward. In other words, increasing the external voltage up to about 2.65 V doesn’t make any difference, and a higher pH difference leads to efficiency reduction with an opposite trend after the coincidence point. Due to high ohmic losses at low voltages that makes a poor current density, the input electrical power decreases and it coincides with the photochemical deficiency due to low proton concentration in the oxygen evolution manifold. This subsequently makes the light energy absorption deficient. This provides completely the opposite trend for voltages over 2.65 and the current density at the electrode interface increases exponentially. Also, the electrochemical efficiency of hydrogen production quench rises with higher proton concentration, for a more acidic catalytic solution. This study compared the proposed hybrid photochemical method with three other water splitting techniques in Fig. 10. This is helpful for better understating of the current status of the state-of-the-art hydrogen production scenarios from energy and exergy aspects of view. Consider photochemical (PC), photoelectrochemical (PEC), and conventional electrolysis (EC) methods for comparison, based on experimental data published in references [9,11,14], respectively. The input
Hybrid-PC
Fig. 8 e Effect of external power supply on maximum energy conversion efficiency.
PC PEC Water Splitting Method
Fig. 10 e Comparison between energy and exergy efficiencies for different water splitting methods.
EC
7470
Efficiency (%)
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Cu-phthalocyannine
Ru (bpy)3 2+ photosensitizer
Eosin Y
Fig. 11 e Comparison between energy and exergy efficiencies for different photosensitizers.
but independent verification is undoubtedly warranted, and the losses associated with process scale-up are, as yet, unknown. The energy efficiencies calculated in this study are almost of the same order of magnitude with respect to similar systems, whereas photoelectrochemical water splitting systems [15,17] with higher efficiencies are not exactly utilizing the same phenomenological approach as photocatalytic method of this study.
3.1. Conversion efficiency of light energy based on energy balance The light energy conversion efficiency is defined as the conversion efficiency of the system when the hydrogen energy output is equal to an equivalent energy input of the system, which is denoted as “hc”. Fig. 12 shows the energy flow of the hydrogen production system, proposed in this study for hybrid photocatalytic water decomposition. Assuming the energy input comprises only the solar energy, and operation energy. The former can be considered as a free energy. The
latter is consumed to drive the supplying pump, extraction fans and stirrer, which is usually electricity and should be invested. The energy output of the system is contributed to the chemical energy of hydrogen output flow. The minimum conversion efficiency of light energy of a certain system would be the case when the hydrogen energy output of the system will be equal to the equivalent input energy, so as to compensate the expenditure of the system. Therefore, only if the conversion efficiency of light energy is greater than this minimum value, the energy profit can be achieved. The concept of conversion efficiency of light energy, by the aforementioned definition, sets a target of commercialscale production for the current experiments of the hydrogen production from photocatalytic water decomposition. It can be concluded that if the energy efficiency of light energy for an experimental system is greater than this value, the system could be considered for a commercial-scale production. Neglecting the factors such as dissolvability, viscosity, and pressure loss in pipeline, the energy balance of the system is: WSolar þ WPump þ WFan þ WStirrer ¼ HHVH2 n_ H2 þ Wloss
(13)
where Wloss denote energy loss. In order to find the minimum value for the conversion efficiency of light energy, the equation of energy balance of the system could be written as: WPump þ WFan þ WStirrer ¼ HHVH2 n_ H2
(14)
Assuming a range of 100e500 tons annual hydrogen production capacity, and 3000 h annual duration of sunshine in Canada, the conversion efficiency of light based on energy balance is calculated as follows; Considering the conversion efficiency of light energy “hc”, the minimum imposed area of the reactor needed for solar energy harvesting is: SR ¼
HHVH2 n_ H2 ¼ 351:19 h1 c W0 hcr
(15)
Selecting the tri-lamina propeller standard formant stirrer, and assuming the diameter of the impeller is one third of reactor diameter, then:
Table 1 e Comparison between experimentally and theoretically obtained efficiencies for the photoelectrolysis of water. Entry number
Efficiency (%)
1
0.6e2.2
2
1e2.8
3
1.84
4
0.05
5
0.33e0.43
Comments A tandem monolithic configuration used with WO3 films being biased by a PV junction. Bipolar CdSe/CdS panels used under 52 mW/cm2 effective solar flux. Upper limit after correction for light absorption by the electrolyte. Measured for a cell with n-Fe2O3 photoanode under 50.0 m W/cm2 Xe arc lamp irradiation and at a bias potential of 0.2 V/SCE at pH 14. Measured for a polycrystalline p/n diode assembly based on Fe2O3. Poor efficiency attributed to the non-optimal charge-transfer properties of the oxide. Calculated for a hybrid system based on Supramolecular complexes [{(bpy)2Ru(dpp)}22RhBr2](PF6)5 as the photo-catalysts, and an external electric power supply to enhance the photochemical reaction.
Reference [15] [16,17]
[18]
[19]
This Study
7471
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
7
Conversion Efficiency
6
Fig. 12 e Energy flow of the photocatalytic system.
5 4 3 2 1
d D ¼ ¼ 21:15 h0:5 ðmÞ c 3
(16)
where D is the diameter of the impeller and d is the reactor diameter. The rotational speed is set to n ¼ 10 rpm. Assuming that the density and viscosity of the solution are equal to those of water, that is rH2 O ¼ 996:9 kg=m3 , mH2 O ¼ 0:8973 cp, under 298 K. Therefore, the Reynolds number of flow based on the stirrer diameter is 2 D2 n rH2 O 10 996:9 21:15 h0:5 cr ¼ mH2 O 60 0:8973 103 ¼ 8:3 107 h1 c
Re ¼
(17)
Knowing that 0 < hc < 1, the Reynolds number is greater than 100,000, and hence, the solution in the reactor is in sufficient turbulence; therefore, the power factor of the stirrer, B, can be set as 6.1. The power consumption of the stirrer becomes 3 5 _ stirrer ¼ 6:1rH2 O n D ¼ 6:14 105 h2:5 ðkWÞ W c 0:8
(18)
The annual energy consumption of the stirrer is written as 3000 3600 Wstirrer ¼ ð6:14 105 Þ þ h2:5 c ¼ ð6:63 1012 Þ h2:5 ðkJÞ c
(19)
Assuming the theoretical water decomposition rate, the maximum energy consumption of pump and extraction fans are calculated. Substituting the energy consumption values for each utility into Eq. (14), the conversion efficiency calculation for 100 tons annual hydrogen production follows: 12; 893:5 þ 6130 105 h2:5 ¼ 142; 915 105 c
0 0
100
200
300
400
500
600
Annual Production Hydrogen (Tons) Fig. 13 e Light energy conversion efficiency of photocatalytic system.
parallel units with the same capacities is recommended instead of large scaled of the system. Another approach to address this issue would be exergy efficiency evaluation of this system. The following definition is applied to find the exergy efficiency of such systems: hI ¼ 1 I_total =E_ in
(22)
where I_total and E_ in are the irreversibility, and input exergy rate of the hydrogen production system, respectively. Fig. 13 shows exergy efficiency of this system for 100e500 tons hydrogen production capacities. These results imply that exergy efficiency at higher capacities improves significantly. This matter together with the previous conclusion introduce a tradeoff between scaled up and parallelized production units. A detailed optimization is required to distinguish the most beneficial option. Due to production rate limiting nature of photocatalytic systems, the results of Fig. 13 or Fig. 14 may not be extended to very high capacity ranges.
(20)
8
hc ¼ 0:113AT0:6
(21)
where AT is the annual hydrogen production capacity of photocatalytic system. According to the definition, it can be concluded that by doubling the production capacity, the minimum desired conversion efficiency of the system should be improved by almost 50% to be commercially justifiable. The satisfaction of this requirement is quite difficult in practice, and utilizing
7
Exergy Efficiency
Therefore, hc ¼ 1.79%. Fig. 13 shows the conversion efficiency of light energy for the range of 100e500 tons annual hydrogen production capacity. A correlation for conversion efficiency of such system could be developed by curve fitting of these data with a power function, as follows;
6 5 4 3 2 1 0 0
100
200
300
40 0
500
600
Annual Production Hydrogen (Tons) Fig. 14 e Exergy efficiency of photocatalytic system.
7472
4.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Conclusions [5]
The following concluding remarks can be extracted from this study. The hybridization of photochemical water splitting based on supramolecular complexes [{(bpy)2 Ru(dpp)}2 RhBr2] (PF6)5, using external electric power, results in the enhancement of system performance. The maximum energy conversion efficiency is improved almost 27% over 6 V external electric potential, and about 30% enhancement in exergy efficiency is expected. It was concluded that the hydrogen production process in the acidic medium was more efficient than a basic medium. Providing a 6 unit pH difference between hydrogen and oxygen reactor catalytic solutions could lead to about a 10% increase in energy conversion and exergy efficiency. The ranking of photosensitizers in terms of the degree of enhancement of energy and exergy efficiencies occurred in the following order: Cu-phthalocyanine > [Ru(bpy)3]2þ > Eosin Y. Considering the overall effect of all performance improvements including auxiliary electric power, a light driven proton pump, and efficient photosensitizer, the photochemical production of hydrogen is predicted to achieve about a 42% increase in the energy and exergy efficiencies. Investigation of conversion efficiency of light together with exergy efficiency analysis for scaled up photocatalytic system shows a tradeoff between profits of parallel network of low capacity reactors and a high capacity reactor.
[6]
[7] [8]
[9]
[10]
[11]
[12]
[13]
Acknowledgments [14]
Financial support of Phoenix Canada Oil Company Ltd. and the Natural Sciences and Engineering Research Council of Canada (NSERC) is appreciated.
references
[1] Fujishima A, Honda K. Electrochemical photolysis of water at a semiconductor electrode. Nature 1972;238:37e8. [2] Ohta T, Kamiya N, Yama Gucm M, Gotoh N, Otagawa T, Asakura S. System efficiency of a water splitting system synthesized by photochemical and thermoelectric conversion of solar energy. Int J Hydrogen Energy 1977;3:203e8. [3] Lewis NS, Nocera DG. Powering the planet: chemical challenges in solar energy utilization. PANS 2006;103(43): 15729e35. [4] Du P, Schneider J, Li F, Zhao W, Patel U, Castellano F, et al. Biand Terpyridyl Platinum(II) chloro complexes: molecular catalysts for the photogeneration of hydrogen from water or
[15]
[16]
[17]
[18]
[19]
simply precursors for colloidal platinum? J Am Chem Soc 2008;130:5056e8. Lei P, Hedlund M, Lomoth R, Rensmo H, Johansson O, Hammarstrc¸m L. The role of colloid formation in the photoinduced H2 production with a RuII-PdII supramolecular complex: a study by GC, XPS, and TEM. J Am Chem Soc 2008; 130:26e7. Streich D, Astuti Y, Orlandi M, Schwartz L, Hammarstrc¸m L. High-turnover photochemical hydrogen production catalyzed by a model complex of the [FeFe] - hydrogenase active site. Chem Eur J 2010;16:60e3. Balzani V, Juris A. Supramolecular photochemistry, NATO ASI Series; 2001 [Chem. 97: Rev 211]. Brauns E, Jones SW, Clark JA, Molnar SM, Kawanishi Y, Brewer KJ. Electrochemical, spectroscopic and spectroelectrochemical properties of synthetically useful supramolecular light absorbers with mixed-polyazine ligands. Inorg Chem 1997;36:2861e7. Arachchige SM, Brown J, Brewer KJ. Photochemical hydrogen production from water using the new photocatalyst [{(bpy)2 Ru(dpp)}2 2RhBr2] (PF6)5. J Photochem Photobiol A: Chem 2008;197:13e7. Zamfirescu C, Dincer I, Naterer GF. Analysis of a photochemical water splitting reactor with supramolecular catalysts and a proton exchange membrane. Int J Hydrogen Energy 2011;36(17):11273e81. Senftle FE, Grant JR, Senftle FP. Low-voltage DC/AC electrolysis of water using porous graphite electrodes. Electrochim Acta 2010;55:5148e53. Bagotsky V. Fundamentals of electrochemistry. 2nd ed. NYC, NY: Wiley-Interscience, John Wiley and Sons; 2006 [Chapter 17]. Ghosh PK, Smirnov A Yu, Nori1 F. Modeling light-driven proton pumps in artificial photosynthetic reaction centers. J Chem Phys 2009;131:035102. Huang CW, Liao CH, Wu JC. Hydrogen generation from photocatalytic water splitting over TiO2 thin film prepared by electron beam-induced deposition. Int J Hydrogen Energy 2010;35:12005e120010. Paluselli D, Marsen B, Miller EL, Rocheleau RE. Nitrogen doping of reactively sputtered tungsten oxide films. Electrochem Solid-State Lett 2005;8(11):G301e3. Smotkin ES, Cerveramarch S, Bard AJ, Campion A, Fox MA, Mallouk T, et al. Bipolar CDSE/COS semiconductor photoelectrode arrays for unassisted photolytic water splitting. J Phys Chem 1987;91(1):6e8. Cervera-March S, Smotk ES. A photoelectrode array system for hydrogen production from solar water splitting. Int J Hydrogen Energy 1991;16(4):243e7. Khan S, Akikusa J. Photoelectrochemical splitting of water at nanocrystalline n-Fe2O3 thin-film electrodes. J Phys Chem 1999;103(34):7184e9. Leygraf C, Hendewerk M, Somorjai GA. Photo-dissociation of water by paper-type and normal-type polycrystalline ironoxides by using visible-light (less than or equal to 2.7 e.v) in the absence of external potential. Proc Natl Acad Sci USA Phys Sci 1982;79(18):5739e41.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/he
Preformance analysis of a water splitting reactor with hybrid photochemical conversion of solar energy E. Baniasadi*, I. Dincer, G.F. Naterer Faculty of Engineering and Applied Science, University of Ontario Institute of Technology (UOIT), 2000 Simcoe Street North, Oshawa, ON, Canada L1H 7K4
article info
abstract
Article history:
In this paper, a new hybrid system for hydrogen production via solar energy is developed
Received 23 September 2011
and analyzed. In order to decompose water into hydrogen and oxygen without the net
Received in revised form
consumption of additional reactants, a steady stream of reacting materials must be
25 January 2012
maintained in consecutive reaction processes, to avoid reactant replenishment or addi-
Accepted 26 January 2012
tional energy input to facilitate the reaction. The system comprises two reactors, which are
Available online 27 February 2012
connected through a proton conducting membrane. Oxidative and reductive quenching pathways are developed for the water reduction and oxidation reactions. Supramolecular
Keywords:
complexes [{(bpy)2Ru(dpp)}2RhBr2] (PF6)5 are employed as the photo-catalysts, and an
Hydrogen production
external electric power supply is used to enhance the photochemical reaction. A light
Solar energy
driven proton pump is used to increase the photochemical efficiency of both O2 and H2
Photocatalytic method
production reactions. The energy and exergy efficiencies at a system level are analyzed and
Hybridization
discussed. The maximum energy conversion of the system can be improved up to 14% by
Exergy
incorporating design modification that yield a corresponding 25% improvement in the
Efficiency
exergy efficiency. Copyright ª 2012, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Solar driven water splitting combines several attractive features for sustainable energy utilization. The energy source of the sun and the reactive media of water are readily available and renewable, and the resultant fuel of hydrogen and product of water are each environmentally clean. Fujishima and Honda [1] reported on hydrogen production from photocatalytic water decomposition by using a TiO2 single crystal electrode. The authors examined the processes of directly transforming solar energy into chemical energy. Numerous other past studies have been conducted on the photodecomposition of water as a clean solar energy conversion process leading to a non-polluting fuel.
Solar energy cannot be directly stored or continuously supplied. Therefore, the conversion of solar energy to a type of storable energy has crucial importance. To utilize a low density solar flux as effectively as possible, all wavelengths of light should be used, and the efficiency of each step of the energy conversion steps should be improved. Another alternative for hydrogen production by solar energy without consumption of additional reactants is a hybrid system. A hybrid system which combines photochemical, thermochemical and electrochemical reactions has been reported previously [2]. Compared to hydrogen production methods based on fossil fuels, the high investment cost of solar hydrogen generation is a challenging issue. Production of hydrogen in a regenerative
* Corresponding author. E-mail addresses: [email protected] (E. Baniasadi), [email protected] (I. Dincer), [email protected] (G.F. Naterer). 0360-3199/$ e see front matter Copyright ª 2012, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2012.01.128
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
fashion such as photochemical splitting of water has been studied [3]. The reductive side of this process requires the development of catalysts that promote the reduction of protons to molecular hydrogen, facilitated by direct excitation by a photosensitizer. Recently, molecular platinum- and palladium-based systems have been developed as heterogeneous catalysts [4,5]. Supramolecular complexes in a photocatalytic hydrogen production scheme that results in high turnover rates and numbers are another area of interest [6]. The long-lived metal-to-ligand charge-transfer (MLCT) excited state of [Ru (bpy)3]2þ has motivated photochemical and photophysical studies leading to light to energy conversion processes [7,8]. The MLCT excited state of [Ru (bpy)3]2þ and its analogs have the required energy to split water into hydrogen and oxygen, but it requires complicated multi component systems for operation. One of the most promising areas is development of photochemical water splitting using supramolecular devices, which are able to capture the incident solar radiation, and generate electrons or holes at the active center where water reduction or oxidization occurs, respectively. Such systems mimic natural photosynthesis and mainly consist of supramolecular complexes of organic molecules which possess active metallic centers [9]. The water photo-oxidizing reaction occurs as follows: hv
2H2 Oð1Þ / 4Hþ þ 4e þ O2 ðgÞ; Vox ¼ þ0:82 V; NHE:
(1)
Following this reaction under the influence of photonic radiation, the active center absorbs two electrons from one water molecule. The water reduction proceeds according to: hv
2H2 O þ 2e / H2 þ 2OH ; Vred ¼ 0:41 V; NHE:
(2)
For a complete cycle, two water molecules are required. Thus, the total cell potential, E, is written as E ¼ E1 þ E2 ¼ 0.83 V to 0.40 V ¼ 1.23 V. Numerous studies on photochemical production of hydrogen have focused on the development of photocatalytic materials. Relatively few have examined thermodynamic studies of photon-to-electron conversion, theoretical power conversion efficiency, or energy and exergy efficiencies. This paper extends the analysis of a newly proposed photochemical water splitting system by C. Zamfirescu et al. [10] by implementing new modifications to improve the system performance. These modifications mainly consist of hybridization of a photocatalytic system by an external electric source, implementation of a light driven proton pump, and efficient photosensitizers. This paper investigates the performance of a photochemical hydrogen generation reactor through energy and exergy methods, and also comparisons with other water splitting methods.
2.
System description
A photochemical reactor comprises a radiation source, a reactor transparent for the light, a gas product outlet and a photocatalytic reactor. As shown in Fig. 1, the vessel comprises two photochemical reactors separated by a light driven proton pump membrane. Similar to natural photo-
7465
Fig. 1 e Schematic of water splitting photochemical reactor with external power source.
systems, an artificial water splitting photo-system will require a number of essential molecular components that must be organized in supramolecular complexes such that the photo-generated electron-hole pair is quickly and efficiently separated and their potential energy is delivered to functional co-catalysts. The system contains a minimum number of necessary components such as a photosensitizer, oxygen and hydrogen evolution reaction catalysts. An efficient and robust sensitizer absorbs photons and generates long-lived electronhole pairs that are coupled to appropriate multi-electron cocatalysts for oxygen and hydrogen evolution. Photolysis solutions consist of [{(bpy)2 Ru (dpp)}2 RhBr2] (PF6)5 (65 mM), water (0.62 M) acidified to pH ¼ 2 with triflic acid, and dimethylaniline (1.5 M) in a solution of acetonitrile (4.46 mL). In the water photo-oxidation reactor, on the left, selected supramolecular photo-catalysts are used to capture light energy and generate electrical charges at reaction sites to oxidize water and produce oxygen gas and protons. Fresh water is continuously supplied to this reactor, and the flow rate of fresh water is adjusted such that the water level in the vessel remains constant. In the water reduction reactor, selected supramolecular complexes for photocatalytic reduction of water to hydrogen are dissolved in a proper concentration. These catalysts generate photoelectrons, which makes hydrogen evolution reactions occur. An auxiliary electric circuit has been considered to provide an external electric potential and improve the photocatalytic reaction performance. Negative charges, generated by a power supply, are donated at the electrode surface to the supramolecular devices, and these are transmitted to the reaction sites under the influence of photonic radiation. Two extraction fans are used above the liquid level, to extract oxygen and hydrogen, continuously. As mentioned previously, this research will focus on the use of Ru(II)- bipyridyl complexes as sensitizers to drive the water splitting reaction. As seen in the modified Latimer diagram, shown in Fig. 2, each cycle starts with photoreactions to capture the light energy (at w 450 nm or 2.75 eV) to form the excited single molecule [Ru (bpy)3]2þ**. This molecule
7466
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Fig. 2 e Photoreactions based on reductive and oxidative quenching of the photo-excited.
rapidly loses some energy upon intersystem crossing to the triplet state, [Ru (bpy)3]2þ* .The photo-excited complex, denoted [Ru (bpy)3]2þ* or more simply Ru2þ*, is both a good oxidant and a good reductant. Electron transfer either to or from the excited state complex traps the ruthenium complex as [Ru (bpy)3]þ or [Ru (bpy)3]3þ, respectively. These photoproducts are potent reductants ([Ru (bpy)3]þ) and oxidants ([Ru (bpy)3]3þ which can drive subsequent redox reactions. These various species can be used to drive a catalytic reaction by either an oxidative or reductive quenching pathway. The energetic of the water splitting reaction within these two manifolds is also shown in Fig. 2. The reactions are normalized to a 4-photon, 4-electron process, and a pH of 7.0 is assumed. EA and ED refer to generic electron acceptor and donor molecules. One of the challenges regarding oxidation and reduction mechanisms involves replenishment of electron acceptors and donors. Consumption of EDs will require replenishment of this material either by a sustainable process reaction that caused the ED to reform, or by extraction and replacement of the ED. Our new proposed approach will apply an external power source and two electrodes immersed in the catalyst solution to supply and transfer electrons inside two reactors. The equivalent circuit of the scheme shown in Fig. 3 should correspond to that of the schematic in Fig. 1. The electric potential generated by the photosensitizers of both oxygen and hydrogen production reactors is a minimum of 2.62 V NHE. The generated energy is used to drive the half reactions, which consume 0.82 V and 0.41 V, respectively. The remaining electrical energy is used to overcome various losses including ohmic, concentration and activation loses at the electrodes. In order to have a self-sustained system, the total loss should be less than 1.39 V. The presence of ionic molecular devices at the electrode interface, where they exchange electrons (donate or accept) with high activity, dispel the need for coating the electrode surface by expensive catalysts like platinum group materials. This research uses porous graphite electrodes [11]. The calculated cell potential (1.23 V) is the minimum voltage needed for electrolysis. In practice, however,
electrolysis generally starts at a somewhat higher potential to overcome the overpotential and ohmic losses needed to form hydrogen and oxygen [12]. In past experimental data [11], a stream of fine gas bubbles started to form at the anode and the size of these oxygen bubbles gradually increased with potential until w1.8 V. The first activity at the cathode was a stream of fine bubbles (formed at w2.3 V). Although the decomposition potential for oxygen and hydrogen are slightly different, the practical decomposition potential is nearly the same (w2.3 V), regardless of the immersed area of the electrodes. Another approach to extend the conceptual design in Ref. [10] is use of light driven proton pumps instead of conventional proton exchange membranes. This paper uses a slightly modified version of a system in Ref. [13], shown in Fig. 4. The proposed proton pump comprises a reaction center,
Fig. 3 e Equivalent schematic of the electric circuit for water splitting reactors.
7467
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
form hydrogen or oxygen. This translates to two important efficiencies regarding photochemical and electrochemical performances of both water oxidation and reduction reactors. Define the photochemical efficiency (hPC) of quenching pathways as the energy efficiency of this process, which is part of the transmitted irradiation contributed to the water oxidation/reduction reaction divided by the photonic energy. Referring to Fig. 2, the efficiencies of the catalytic cycle for water oxidization and reduction at 450 nm are as follows:
which is a molecular triad containing an electron donor and an electron acceptor, both linked to a photosensitive porphyrin group. The triad molecule is inside the bilayer of a liposome. The molecular triad absorbs a photon and translocates a negative charge near the outer surface and a positive charge near the inner surface of the liposome by generating charge separated species. Following these interactions, protons will be transported across the membrane from the lower proton potential to the higher proton potential side of the membrane, e.g. oxygen and hydrogen production reactors, respectively. The transport of protons in this system is mainly due to combined effects of photo-induced energy transduction within the molecular structure of membrane and electroosmotic forces applied at the two porous electrodes. The anode acts as a proton source and the cathode as a proton sink. The conversion of light energy into the electrochemical gradient of protons across the membrane can be quantitatively characterized by the quantum yield (or quantum efficiency) B of proton translocation. This parameter is defined as follows: B¼
number of protons pumped number of photons absorbed
(3)
Based on a case study in Ref. [13], in 1 ms, the shuttle makes nearly 16 trips and transfers ten protons through the membrane, provided that the light intensity I ¼ 0.133 mWcm2. Considering the number of photons absorbed in 1 ms is w18, the approximate quantum yield of the pumping process is w55%. The effect of this module on system performance will be discussed in the following section.
1:53 ev ¼ 55:5% 2:75 ev
(4)
hEC ¼
1:08 ev ¼ 39:3% 2:75 ev
(5)
The electrochemical efficiency (hEC) of water reduction has a similar definition, as the input energy is supplied by external electric power. There are several uncertainties related to overpotential values, or in other words, an effective contribution of the external electric potential in the decomposition potential. Therefore, this study conducted the energy and exergy analysis for different electrochemical efficiencies. The maximum energy conversion efficiency of the hybrid photochemical water splitting system is defined, based on the higher heating value (HHV) of hydrogen, as follows: hmax H2 ¼
n_ H2 ¼
Results and discussion
As shown in Fig. 2, only part of the energy captured by light will be used by an excited singlet molecule [Ru (bpy)3]2þ* to
(6)
WL hPC þ WEL hEC DEH2
(7)
and DEH2 ¼ 1060 kJ/mol, equivalent to the minimum required energy captured from light (WL at w 450 nm or 2.75 eV) or external electric power (WEL) to form the excited singlet molecule [Ru (bpy)3]2þ**. Fig. 5 shows the effect of the delivered potential at the electrode interface to electrolyte on the maximum energy conversion efficiency. The photochemical efficiency is assumed to be hpc ¼ 0.55 and this study considered different electrochemical efficiencies. The photolysis is conducted
0.2 0.18 0.16 0.14 0.12 0.1 0.08
3.
HHVH2 n_ H2 WL þ WEL
where n_ H2 is the maximum possible molar flow rate of produced H2,
Maximum Energy Conversion Efficiency
Fig. 4 e Schematic diagram of the light-induced proton pump across the lipid bilayer in a liposomic membrane. A molecular triad A-BC-D is symmetrically inserted in the lipid bilayer. “A” refers to acceptor, “D” for donor, and “S” for shuttle.
hPC ¼
0
1
2
3
4
5
External Voltage (V)
Fig. 5 e Effect of external electric potential on maximum energy conversion efficiency.
6
7468
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
using a 5W LED having an excitation wavelength of 470 nm. It has been observed that a hybrid configuration of water splitting system may improve the efficiency if the electrochemical efficiency is higher that the photochemical efficiency. The reason is a very low quantum efficiency of LEDs, which was taken as 1.3% in this analysis. In other words, the auxiliary power supply will be helpful if it delivers excitation energy more efficiently than the photonic source. A maximum 30% solar conversion efficiency improvement is estimated by applying 5 V electric potential. Evaluating the energy and exergy efficiencies of the hydrogen production reactor requires the hydrogen production rate from a correlation ðn_ H2 ¼ 2:725mmol=h; 0 < t < 5hÞ based on a recent study by Brewer et al. on a new photocatalyst as [{(bpy)2 Ru(dpp)}2 RhBr2] (PF6)5 [9]. Our investigation involved the determination of the energy and exergy efficiency by varying DC current density at a low voltage (0e6 V) using graphite electrodes and electrolyte concentration of 65 mM of aforementioned molecular complexes in 0.62 M water. The definition of energy efficiency is similar to the maximum energy conversion efficiency, Eq. (6), but the molar production rate of hydrogen is substituted by the value according to the aforementioned correlation. Fig. 6 shows the effect of external voltage on the energy efficiency. Compared to the electrolytic load, the resistive load of the cell is relatively high, yielding a very small current density which slowly changes linearly with increasing potentials (<2 V). Its effect on input electric power is shown in Fig. 6. At voltages higher than 2.3 V, the current density increases very rapidly due to the redox reactions. Thus, above the decomposition potential, the current density consists of the ohmic current density plus the current density produced by the redox reaction. Up to certain energy efficiency it doesn’t differ, but as the applied potential goes above 2.3 V, the efficiency increases rapidly and then decreases following higher concentration and ohmic losses. Almost a 27% improvement in energy efficiency is achieved over 6 V external electric potential. The energy efficiency of photochemical water splitting reactors without any external power is about 0.33%. The absorbed photonic radiation plus electric energy drives an endothermic chemical reaction. The measure of how these
0.44
sources of energy are converted into chemical energy stored in H2 for a given process is the exergy efficiency is defined as hexergy ¼
n_ H2 DGH2 þ0:5O2 /H2 O WL þ WEL
(8)
where n_ H2 is the actual molar flow rate of produced H2, based on a correlation extracted from experimental data [9], and DG is the standard Gibbs free energy change of the reaction at 298 K (237 kJ/mol), i.e., the maximum possible amount of work that may be extracted from H2 at 298 K, when both H2 and O2 are available at 1 bar. Fig. 7 shows the calculated exergy efficiencies over a range of applied electric potential, and it shows similar trends to energy efficiency. About 30% enhancement in exergy efficiency is calculated. If the auxiliary power supply is unplugged, this efficiency will be about 0.55%. The solution pH should is a major factor that can be tuned to favor hydrogen reduction or water oxidation. Appling the Nernst equation for the two manifolds yields: Reductive quenching pathway: DGOER ¼ 84 22:85ðpHÞðkJ=molÞ
(9)
DGHER ¼ 420 þ 22:85ðpHÞðkJ=molÞ
(10)
Oxidative quenching pathway: DGOER ¼ 116 22:85ðpHÞðkJ=molÞ
(11)
DGHER ¼ 220 þ 22:85ðpHÞðkJ=molÞ
(12)
From these expressions, an acidic solution oxidative quenching pathway is more exothermic during the hydrogen evolution process than a reductive pathway, which implies higher photochemical efficiency according to Eq. (4). Similarly, the oxygen evolution energy in the oxidative pathway is higher in a basic solution. Hence, the oxidative quenching pathway is selected in our design in order to take advantage of the light driven proton pump. Also, the pH range in which water splitting can occur should be determined, assuming the two half reactions are not energetically coupled. The free energy of the Hydrogen Evolution Reaction (HER) is not used to
200
0.75
200
0.7
160
120
0.36 80
0.34 0.32
40
0.3 0.28
0 0
1
2 3 4 External Voltage (V)
5
6
Fig. 6 e Energy efficiency and supplied electric energy versus external voltage.
0.65 120 0.6 80 0.55 40
0.5
0
0.45 0
1
2
3
4
5
6
External Voltage (V)
Fig. 7 e Exergy efficiency and supplied electric energy versus external voltage.
Electrical Energy (kJ)
0.38
Exergetic Efficiency
160
0.4
Electrical Energy (kJ)
Energetic Efficiency
0.42
7469
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Fig. 9 e Effect of external power supply on exergy efficiency.
energy for all four cases is assumed to be equal and working conditions are consistent. Although it couldn’t reach the efficiency of a photoelectrochemical system, but hybridization of photochemical processes is promising. Hybridization made an acceptable improvement in the photochemical performance, and it shows almost a 30% increase in efficiencies. Compared to conventional electrolysis, solar based methods are relatively deficient, and they need more efforts for commercialization purposes. Low quantum efficiency of current photo-catalysts is a crucial challenge to be addressed in order to decrease energy and exergy destructions and losses. The effects of [Ru (bpy)3]2þ, Eosin Y, and Cuphthalocyanine, as different photosensitizers types for hydrogen production photoreactions, on energy and exergy efficiencies, are shown in Fig. 11. Their effectiveness occurs in the following order: Cu-phthalocyanine > [Ru (bpy)3]2þ > Eosin Y. The difference in the hydrogen production rate is based on the structures and properties of these dyes and the differences in their electron injection characteristics. The energy conversion efficiency calculated for the proposed system of this study is compared and qualitatively validated with some experimental data on photo- or solarelectrolysis efficiencies, as contained in Table 1. Although some of the numbers claimed have been questioned by others,
Efficiency (%)
help drive the Oxygen Evolution Reaction (OER) or vice versa. These energies and pH ranges are specific for [Ru (bpy)3]2þ, assuming a four-photon process but they are tunable by modifying the complex structure. Figs. 8 and 9 represent the maximum energy conversion and exergy efficiencies of oxygen and hydrogen production reactors versus electrical potential at the electrode interface, considering the effect of the solution pH difference between reactors. The case DpH ¼ 0 implies that a photolysis solution of pH ¼ 7 and increasing the pH difference makes hydrogen and oxygen quenches acidic and basic, respectively. It was found that the hydrogen production volume in the acidic medium exceeds the basic medium. At an acidic pH, more Hþ ions would be absorbed by the [Ru (bpy)3]2þ*, and hence the possibility for reduction of Hþ to H2 increases. Increasing the proton concentration in the catalyst solution of the hydrogen evolution manifold could lead to about 10% improvement in maximum energy conversion efficiency, and 13% in exergy efficiency. All graphs meet at a point corresponding to 2.65 V where efficiencies have an opposite trend before and afterward. In other words, increasing the external voltage up to about 2.65 V doesn’t make any difference, and a higher pH difference leads to efficiency reduction with an opposite trend after the coincidence point. Due to high ohmic losses at low voltages that makes a poor current density, the input electrical power decreases and it coincides with the photochemical deficiency due to low proton concentration in the oxygen evolution manifold. This subsequently makes the light energy absorption deficient. This provides completely the opposite trend for voltages over 2.65 and the current density at the electrode interface increases exponentially. Also, the electrochemical efficiency of hydrogen production quench rises with higher proton concentration, for a more acidic catalytic solution. This study compared the proposed hybrid photochemical method with three other water splitting techniques in Fig. 10. This is helpful for better understating of the current status of the state-of-the-art hydrogen production scenarios from energy and exergy aspects of view. Consider photochemical (PC), photoelectrochemical (PEC), and conventional electrolysis (EC) methods for comparison, based on experimental data published in references [9,11,14], respectively. The input
Hybrid-PC
Fig. 8 e Effect of external power supply on maximum energy conversion efficiency.
PC PEC Water Splitting Method
Fig. 10 e Comparison between energy and exergy efficiencies for different water splitting methods.
EC
7470
Efficiency (%)
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Cu-phthalocyannine
Ru (bpy)3 2+ photosensitizer
Eosin Y
Fig. 11 e Comparison between energy and exergy efficiencies for different photosensitizers.
but independent verification is undoubtedly warranted, and the losses associated with process scale-up are, as yet, unknown. The energy efficiencies calculated in this study are almost of the same order of magnitude with respect to similar systems, whereas photoelectrochemical water splitting systems [15,17] with higher efficiencies are not exactly utilizing the same phenomenological approach as photocatalytic method of this study.
3.1. Conversion efficiency of light energy based on energy balance The light energy conversion efficiency is defined as the conversion efficiency of the system when the hydrogen energy output is equal to an equivalent energy input of the system, which is denoted as “hc”. Fig. 12 shows the energy flow of the hydrogen production system, proposed in this study for hybrid photocatalytic water decomposition. Assuming the energy input comprises only the solar energy, and operation energy. The former can be considered as a free energy. The
latter is consumed to drive the supplying pump, extraction fans and stirrer, which is usually electricity and should be invested. The energy output of the system is contributed to the chemical energy of hydrogen output flow. The minimum conversion efficiency of light energy of a certain system would be the case when the hydrogen energy output of the system will be equal to the equivalent input energy, so as to compensate the expenditure of the system. Therefore, only if the conversion efficiency of light energy is greater than this minimum value, the energy profit can be achieved. The concept of conversion efficiency of light energy, by the aforementioned definition, sets a target of commercialscale production for the current experiments of the hydrogen production from photocatalytic water decomposition. It can be concluded that if the energy efficiency of light energy for an experimental system is greater than this value, the system could be considered for a commercial-scale production. Neglecting the factors such as dissolvability, viscosity, and pressure loss in pipeline, the energy balance of the system is: WSolar þ WPump þ WFan þ WStirrer ¼ HHVH2 n_ H2 þ Wloss
(13)
where Wloss denote energy loss. In order to find the minimum value for the conversion efficiency of light energy, the equation of energy balance of the system could be written as: WPump þ WFan þ WStirrer ¼ HHVH2 n_ H2
(14)
Assuming a range of 100e500 tons annual hydrogen production capacity, and 3000 h annual duration of sunshine in Canada, the conversion efficiency of light based on energy balance is calculated as follows; Considering the conversion efficiency of light energy “hc”, the minimum imposed area of the reactor needed for solar energy harvesting is: SR ¼
HHVH2 n_ H2 ¼ 351:19 h1 c W0 hcr
(15)
Selecting the tri-lamina propeller standard formant stirrer, and assuming the diameter of the impeller is one third of reactor diameter, then:
Table 1 e Comparison between experimentally and theoretically obtained efficiencies for the photoelectrolysis of water. Entry number
Efficiency (%)
1
0.6e2.2
2
1e2.8
3
1.84
4
0.05
5
0.33e0.43
Comments A tandem monolithic configuration used with WO3 films being biased by a PV junction. Bipolar CdSe/CdS panels used under 52 mW/cm2 effective solar flux. Upper limit after correction for light absorption by the electrolyte. Measured for a cell with n-Fe2O3 photoanode under 50.0 m W/cm2 Xe arc lamp irradiation and at a bias potential of 0.2 V/SCE at pH 14. Measured for a polycrystalline p/n diode assembly based on Fe2O3. Poor efficiency attributed to the non-optimal charge-transfer properties of the oxide. Calculated for a hybrid system based on Supramolecular complexes [{(bpy)2Ru(dpp)}22RhBr2](PF6)5 as the photo-catalysts, and an external electric power supply to enhance the photochemical reaction.
Reference [15] [16,17]
[18]
[19]
This Study
7471
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
7
Conversion Efficiency
6
Fig. 12 e Energy flow of the photocatalytic system.
5 4 3 2 1
d D ¼ ¼ 21:15 h0:5 ðmÞ c 3
(16)
where D is the diameter of the impeller and d is the reactor diameter. The rotational speed is set to n ¼ 10 rpm. Assuming that the density and viscosity of the solution are equal to those of water, that is rH2 O ¼ 996:9 kg=m3 , mH2 O ¼ 0:8973 cp, under 298 K. Therefore, the Reynolds number of flow based on the stirrer diameter is 2 D2 n rH2 O 10 996:9 21:15 h0:5 cr ¼ mH2 O 60 0:8973 103 ¼ 8:3 107 h1 c
Re ¼
(17)
Knowing that 0 < hc < 1, the Reynolds number is greater than 100,000, and hence, the solution in the reactor is in sufficient turbulence; therefore, the power factor of the stirrer, B, can be set as 6.1. The power consumption of the stirrer becomes 3 5 _ stirrer ¼ 6:1rH2 O n D ¼ 6:14 105 h2:5 ðkWÞ W c 0:8
(18)
The annual energy consumption of the stirrer is written as 3000 3600 Wstirrer ¼ ð6:14 105 Þ þ h2:5 c ¼ ð6:63 1012 Þ h2:5 ðkJÞ c
(19)
Assuming the theoretical water decomposition rate, the maximum energy consumption of pump and extraction fans are calculated. Substituting the energy consumption values for each utility into Eq. (14), the conversion efficiency calculation for 100 tons annual hydrogen production follows: 12; 893:5 þ 6130 105 h2:5 ¼ 142; 915 105 c
0 0
100
200
300
400
500
600
Annual Production Hydrogen (Tons) Fig. 13 e Light energy conversion efficiency of photocatalytic system.
parallel units with the same capacities is recommended instead of large scaled of the system. Another approach to address this issue would be exergy efficiency evaluation of this system. The following definition is applied to find the exergy efficiency of such systems: hI ¼ 1 I_total =E_ in
(22)
where I_total and E_ in are the irreversibility, and input exergy rate of the hydrogen production system, respectively. Fig. 13 shows exergy efficiency of this system for 100e500 tons hydrogen production capacities. These results imply that exergy efficiency at higher capacities improves significantly. This matter together with the previous conclusion introduce a tradeoff between scaled up and parallelized production units. A detailed optimization is required to distinguish the most beneficial option. Due to production rate limiting nature of photocatalytic systems, the results of Fig. 13 or Fig. 14 may not be extended to very high capacity ranges.
(20)
8
hc ¼ 0:113AT0:6
(21)
where AT is the annual hydrogen production capacity of photocatalytic system. According to the definition, it can be concluded that by doubling the production capacity, the minimum desired conversion efficiency of the system should be improved by almost 50% to be commercially justifiable. The satisfaction of this requirement is quite difficult in practice, and utilizing
7
Exergy Efficiency
Therefore, hc ¼ 1.79%. Fig. 13 shows the conversion efficiency of light energy for the range of 100e500 tons annual hydrogen production capacity. A correlation for conversion efficiency of such system could be developed by curve fitting of these data with a power function, as follows;
6 5 4 3 2 1 0 0
100
200
300
40 0
500
600
Annual Production Hydrogen (Tons) Fig. 14 e Exergy efficiency of photocatalytic system.
7472
4.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 7 ( 2 0 1 2 ) 7 4 6 4 e7 4 7 2
Conclusions [5]
The following concluding remarks can be extracted from this study. The hybridization of photochemical water splitting based on supramolecular complexes [{(bpy)2 Ru(dpp)}2 RhBr2] (PF6)5, using external electric power, results in the enhancement of system performance. The maximum energy conversion efficiency is improved almost 27% over 6 V external electric potential, and about 30% enhancement in exergy efficiency is expected. It was concluded that the hydrogen production process in the acidic medium was more efficient than a basic medium. Providing a 6 unit pH difference between hydrogen and oxygen reactor catalytic solutions could lead to about a 10% increase in energy conversion and exergy efficiency. The ranking of photosensitizers in terms of the degree of enhancement of energy and exergy efficiencies occurred in the following order: Cu-phthalocyanine > [Ru(bpy)3]2þ > Eosin Y. Considering the overall effect of all performance improvements including auxiliary electric power, a light driven proton pump, and efficient photosensitizer, the photochemical production of hydrogen is predicted to achieve about a 42% increase in the energy and exergy efficiencies. Investigation of conversion efficiency of light together with exergy efficiency analysis for scaled up photocatalytic system shows a tradeoff between profits of parallel network of low capacity reactors and a high capacity reactor.
[6]
[7] [8]
[9]
[10]
[11]
[12]
[13]
Acknowledgments [14]
Financial support of Phoenix Canada Oil Company Ltd. and the Natural Sciences and Engineering Research Council of Canada (NSERC) is appreciated.
references
[1] Fujishima A, Honda K. Electrochemical photolysis of water at a semiconductor electrode. Nature 1972;238:37e8. [2] Ohta T, Kamiya N, Yama Gucm M, Gotoh N, Otagawa T, Asakura S. System efficiency of a water splitting system synthesized by photochemical and thermoelectric conversion of solar energy. Int J Hydrogen Energy 1977;3:203e8. [3] Lewis NS, Nocera DG. Powering the planet: chemical challenges in solar energy utilization. PANS 2006;103(43): 15729e35. [4] Du P, Schneider J, Li F, Zhao W, Patel U, Castellano F, et al. Biand Terpyridyl Platinum(II) chloro complexes: molecular catalysts for the photogeneration of hydrogen from water or
[15]
[16]
[17]
[18]
[19]
simply precursors for colloidal platinum? J Am Chem Soc 2008;130:5056e8. Lei P, Hedlund M, Lomoth R, Rensmo H, Johansson O, Hammarstrc¸m L. The role of colloid formation in the photoinduced H2 production with a RuII-PdII supramolecular complex: a study by GC, XPS, and TEM. J Am Chem Soc 2008; 130:26e7. Streich D, Astuti Y, Orlandi M, Schwartz L, Hammarstrc¸m L. High-turnover photochemical hydrogen production catalyzed by a model complex of the [FeFe] - hydrogenase active site. Chem Eur J 2010;16:60e3. Balzani V, Juris A. Supramolecular photochemistry, NATO ASI Series; 2001 [Chem. 97: Rev 211]. Brauns E, Jones SW, Clark JA, Molnar SM, Kawanishi Y, Brewer KJ. Electrochemical, spectroscopic and spectroelectrochemical properties of synthetically useful supramolecular light absorbers with mixed-polyazine ligands. Inorg Chem 1997;36:2861e7. Arachchige SM, Brown J, Brewer KJ. Photochemical hydrogen production from water using the new photocatalyst [{(bpy)2 Ru(dpp)}2 2RhBr2] (PF6)5. J Photochem Photobiol A: Chem 2008;197:13e7. Zamfirescu C, Dincer I, Naterer GF. Analysis of a photochemical water splitting reactor with supramolecular catalysts and a proton exchange membrane. Int J Hydrogen Energy 2011;36(17):11273e81. Senftle FE, Grant JR, Senftle FP. Low-voltage DC/AC electrolysis of water using porous graphite electrodes. Electrochim Acta 2010;55:5148e53. Bagotsky V. Fundamentals of electrochemistry. 2nd ed. NYC, NY: Wiley-Interscience, John Wiley and Sons; 2006 [Chapter 17]. Ghosh PK, Smirnov A Yu, Nori1 F. Modeling light-driven proton pumps in artificial photosynthetic reaction centers. J Chem Phys 2009;131:035102. Huang CW, Liao CH, Wu JC. Hydrogen generation from photocatalytic water splitting over TiO2 thin film prepared by electron beam-induced deposition. Int J Hydrogen Energy 2010;35:12005e120010. Paluselli D, Marsen B, Miller EL, Rocheleau RE. Nitrogen doping of reactively sputtered tungsten oxide films. Electrochem Solid-State Lett 2005;8(11):G301e3. Smotkin ES, Cerveramarch S, Bard AJ, Campion A, Fox MA, Mallouk T, et al. Bipolar CDSE/COS semiconductor photoelectrode arrays for unassisted photolytic water splitting. J Phys Chem 1987;91(1):6e8. Cervera-March S, Smotk ES. A photoelectrode array system for hydrogen production from solar water splitting. Int J Hydrogen Energy 1991;16(4):243e7. Khan S, Akikusa J. Photoelectrochemical splitting of water at nanocrystalline n-Fe2O3 thin-film electrodes. J Phys Chem 1999;103(34):7184e9. Leygraf C, Hendewerk M, Somorjai GA. Photo-dissociation of water by paper-type and normal-type polycrystalline ironoxides by using visible-light (less than or equal to 2.7 e.v) in the absence of external potential. Proc Natl Acad Sci USA Phys Sci 1982;79(18):5739e41.