Catalyst-free hydrogen evolution of Si photocathode ...

Journal of Power Sources 279 (2015) 151e156

Contents lists available at ScienceDirect

Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour

Catalyst-free hydrogen evolution of Si photocathode by thermovoltage-driven solar water splitting Sun-Mi Shin 1, 2, Jin-Young Jung 2, Min-Joon Park, Jae-Won Song, Jung-Ho Lee* Department of Materials and Chemical Engineering, Hanyang University, Ansan 426-791, Republic of Korea

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

A photoelectrochemical cell was electrically coupled with a thermoelectric device. Thermodynamic overpotentials for hydrogen evolution were offset by thermovoltage. A photon-to-current efficiency of ~20% was achieved at 56 C temperature gradient.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 8 October 2014 Received in revised form 31 December 2014 Accepted 4 January 2015 Available online 5 January 2015

An externally biased overpotential is normally required for photoelectrochemically cleaving water molecules. Moreover, very few semiconductors exhibit the necessary performance for the efficient transfer of photon energy to the binding electrons of water molecules unless a suitable catalyst is present. Here, we present a catalyst-free photoelectrochemical (PEC) cell electrically coupled in series with a thermoelectric device, which is capable of utilizing the full solar spectrum by synergistically collecting photon and phonon energies. Thermodynamic overpotentials originally required for the PEC reaction were spontaneously offset by the thermovoltage, which adjusts the Fermi level of a counter-electrode. Using a catalyst-free Si photocathode of unbiased conditions, we achieved a photon-to-current efficiency of ~20% at an 56 C temperature gradient by harnessing only solar energy. © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Photoelectrochemical cell Water splitting Hydrogen Thermovoltage Hybrid device

1. Introduction Photoelectrochemical (PEC) conversion is a photothermal process which is chemically driven by solar photons and heat supply. Extra heat from sunlight is used to reduce the thermodynamic barrier for water splitting. Solar photons are a type of free energy source analogous to chemical or electrical energy, but are readily degraded to heat whenever they are absorbed into substances [1,2].

* Corresponding author. E-mail address: [email protected] (J.-H. Lee). 1 Present address: Materials Science and Engineering Program, University of California, San Diego, La Jolla, CA 92093, USA. 2 These people contributed equally to this work.

Thus, the hybrid utilization of photon and heat energy for the effective evolution of solar-driven water splitting is justified. Here, we attempt to electrically couple the PEC and thermoelectric (TE) systems (Fig. 1a) in order to utilize a full solar spectrum by collecting phonon energy (by TE) as well as photon energy (by PEC). Moreover, the thermovoltage (VTE) is determined by the temperature gradient (DT) required to remove the necessity for overpotentials in the PEC reaction (Fig. 1b). This approach can be optimized for the TE system by employing high-temperature water heated by concentrated photovoltaics in which the photon energy degrades first to heating water, then also converts to phonon energy in the TE system. In our experiment, conducted at one-sun illumination, these two different contributions by temperature (in liquid) and phonons

http://dx.doi.org/10.1016/j.jpowsour.2015.01.020 0378-7753/© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

152

S.-M. Shin et al. / Journal of Power Sources 279 (2015) 151e156

Fig. 1. PEC-TE hybrid device. (a) Schematic illustration of a hybrid device consisting of PEC and TE devices. (b) Equivalent block diagram of electric circuit for PEC-TE hybrid device. The photoelectrochemical cell is a photon-active current source combined with an electrochemical cell, in which RS is the series resistance of the PEC cell, RCT is the charge transfer resistance, and CSC is the space charge region capacitance. The thermoelectric device is operating as a voltage source (depending upon temperature gradient) with an internal series resistance of RTE.

(in solid) were practically observed. The contribution of phonons was observed to be even larger because the VTE, induced by a slight increase (~18 C) in temperature, effectively offset the overpotential for the PEC reaction. Since the boiling water heated to 100 C is able to relieve only 5% of DG (the change of Gibbs free energy) required at room temperature (i.e., ~237.1 to 225.2 kJ mol1) [3], the direct impact of heated water on the reaction kinetics for water splitting is very limited at temperatures below 100 C. Thus, the practical impact of heat needs to be evaluated by DT determining the VTE, rather than being determined by the operation temperatures of the electrolytes. Smaller bandgap (Eg) semiconductors are normally better suited to enhancing light absorbance, but Eg also needs to be higher than the sum of DG and the thermodynamic/kinetic losses (i.e., 0.8 eV, losses from polarization, recombination and resistance) in order to decompose water molecules in practical PEC systems [2,4e6]. This trade-off relationship for determining the optimal Eg theoretically limits conversion efficiency to ~17%, which is optimized at the ideal Eg of 2.03 eV for single-absorber operation [5,7]. In our hybrid PECTE system, however, we added the VTE (converted from DT) in addition to the photovoltage originally determined by the Eg of a semiconductor photoelectrode. Further lowering the overpotentials improves the PEC conversion efficiency by adopting smaller Eg materials (for higher photocurrent). 2. Experimental 2.1. Setup of a hybrid cell PEC-TE hybrid systems for solar energy conversion were fabricated by electrically connecting the PEC with a TE device whose topside was heated by solar light, while the opposite side was connected to a heat sink for cooling (Fig. S1). To evaluate the influence of a TE cell on the performance of a hybrid circuit, two different TE cells with internal resistances of 1.2 and 2.1 U were employed for operating each hybrid circuit. For both TE142 and TE254 devices, different structural designs for integrating TE elements were adopted for varying the internal resistances. These devices have different numbers of legs, 142 and 254 for TE142 and TE254, respectively. The TE elements had cross-sectional areas of 12 103 cm2 and were 0.11 cm long. The TE modules used in this study are commercially available from Kryotherm (Saint-Petersburg, Russia) and Laird Technologies (St. Louis, MO). 2.2. Electrical measurement and characterization A 150-W Xe arc lamp with AM 1.5-G filters (Peccel, PEC-L11, Yokohama, Japan) was used as the light and heat source for

hybrid operations. Temperatures of a TE device were measured using K-type thermocouples (Center 306 Data Logger, New Taipei City Taiwan) attached to each device. To understand the characteristics of the hybrid circuits using various DT, the TE side of the hybrid device was mounted onto a passive heat sink and temperature controller (PEC-T10, Peccell Technologies). J-V characteristics of the PEC and hybrid circuit were investigated using a solar simulator (PEC-L11, Peccell technologies) and a potentiostat (Iviumstat, Eindhoven, Netherlands) under a 1-sun light intensity (100 mW/cm2). Conventional p-type Si(100) wafers (500 mm in thickness, 1e10 U cm in resistivity) were used for PEC samples (photoactive area: 0.28 cm2). Before detecting PEC performance, the wafer was dipped into a HF solution to remove the native oxide formed on the surface. The PEC properties were investigated in 0.5 M sulfuric acid with a Pt counter electrode and a silver/silver chloride reference electrode (Ag/AgCl). All electrode potentials were converted to reversible hydrogen electrode (RHE) potentials. 2.3. Measurement of hydrogen evolution P-type Si (working) and a Pt electrode (counter) were adopted as photocathode and anode for generating hydrogen and oxygen, respectively. Since our work focused only on the enhanced reaction performances of Si photocathode driven by thermovoltages, we have measured the hydrogen volumes generated by the photocathode. To collect hydrogen gas only without contamination by oxygen, we set the system in which a fritted glass tube isolated a Pt counter electrode (AFCTR5, Pine research instrumentation, Durham, NC). As shown in Fig. S5, a water-filled flask was placed upside-down during immersing in a water bath, then a gas tube connected to a quartz cell was installed into a flask. The atmospheric pressure prevented water from flowing from the flask to the bath. When hydrogen is collected inside the flask, water solution is able to flow out from the flask owing to a pressure exerted by the hydrogen produced, which results in level of the water meniscus in the flask. The variation of water menisci was read precisely using a graduated flask as a function of time for measuring hydrogen evolution, and the hydrogen amounts are also listed as a function of time in Table S2. 3. Results and discussion A DT of 10 C was observed to decrease the overpotential of ~300 mV, which corresponded closely to the voltage achievable by employing Pt catalysts [8e10]. Other crucial benefits of the catalyst-free PEC-TE hybrid system are the ability to avoid both light loss due to shade from metal catalysts and the recombination loss of charge carriers at the metal/semiconductor interface [8,11].


Moreover, the magnitude of VTE is also controllable by adjusting DT. Fig. 2a and b compare current density and electrochemical potential (J-V) characteristics for the PEC systems employing two different TE modules (TE142 and TE254), in which devices with 142 and 254 thermoelectric legs had 1.2 and 2.1 U values for RTE (resistance of TE module), respectively. As DT increases, the J-V curves likely shift in the positive direction in voltage for decreasing overpotentials. To clearly differentiate the amount of the shifts, the overpotentials required for a photocurrent density (Jph) of 20 mA/cm2 were plotted as a function of DT along with the VTE of the TE modules. The Seebeck coefficient (S) represents the amount of VTE built in response to DT from the relation VTE ¼ SDT, which is determined by the slopes of the fitted lines in Fig. 2c. Values of 0.019 V/K (blue line) (in the web version) and 0.025 V/K (green line) (in the web version) were extracted for TE142 and TE254, respectively. Likewise, the line slopes shown in Fig. 2d represent a reduction rate of overpotential as a function of DT, corresponding to 0.0186 V/K and 0.023 V/K for TE142 and TE254, respectively. The reduction rate of overpotentials was strongly related to the Seebeck coefficients of TE devices. For instance, externally biased overpotentials were no longer necessary for the hybrid system employing a high S-device (TE254) because it is able to spontaneously generate 20 mA/cm2 when DT reaches ~20 C (Fig. 2d); in contrast, the low-S device (TE142) still required an overpotential of ~0.2 V. For full PEC operation without the need for external bias, bandedge potentials at the surfaces must straddle the hydrogen and oxygen redox potentials [12,13]. Silicon is not a suitable semiconductor for spontaneous water splitting because the position of the valence-band-minimum is not sufficiently positive to oxidize water (Fig. 3a) [12]. When a TE device is serially connected to a PEC cell, however, the applied VTE is capable of adjusting the Fermi level of the counter electrodes. Since the working cathode (p-Si) and its

153

counter electrode (Pt) are connected to the positive and negative terminals of a TE module, respectively, electrons injected from the Pt counter electrode flow into the anode of a TE module via electrical wiring. Note that the Fermi level shifts towards a more positive potential (downward) until the Fermi level of Pt lines up with that of the TE device (Fig. 3b). As a result of the applied VTE, water oxidation spontaneously occurs due to the Fermi level of metal being lower than the oxidation potential of water. This distinctive feature of PEC-TE hybrids is able to expand the availability of lightabsorbing semiconductors that optimize the three key factors (i.e., efficiency, robustness, and scalability) required for viable solar water splitting [14]. According to the ButlereVolmer relationship [15], which describes the exponential dependence between electrical current and electrode potential, we observed an exponential increase in current density as a function of VTE. The conversion efficiency of the PEC cell was greatly improved from 0% to ~20% by employing a TE device at a DT of 56 C (Fig. 3c). If all photogenerated electrons and holes are assumed to be used to generate hydrogen gas, i.e., a Faraday efficiency of 100%, the STH (solar-to-hydrogen) efficiency is given by

hSTH ¼

Pout Pin Jph ðVredox Vbias Þ ¼ Plight Plight

(1)

where Pout and Pin are the output and input electrical power, Jph is the current density at an applied bias (Vbias), Vredox is the Gibbs free energy change per electron for the water-splitting reaction (1.23 V at room temperature), and Plight is the illuminated solar irradiance (100 mW/cm2) [7,13,16]. Hydrogen evolution has been measured as a function of time at DT ¼ 18 C (Fig. 4). Measured volumes of hydrogen were also shown to match well with the calculated values, implying a Faraday efficiency of unity from the following Faraday's law of electrolysis;

Fig. 2. Electrical characteristics as a function of DT. Light J-V characteristics of PEC-TE hybrid circuits (a: TE142, b: TE254) with increasing DT. (c) Thermovoltage as a function of DT. (d) Overpotential values (recorded at 20 mA/cm2) as a function of DT.

154


Fig. 3. Adjusting the Fermi level of the counter electrode with DT. Energy band diagrams of (a) sole PEC and (b) a PEC-TE hybrid circuit under illumination depicting the influence of qVTE upon device hybridization. (c) STH efficiencies (assuming Faraday efficiency of 100%) as a function of DT.

Z Mole of H2 ¼

1 2

t

Idt 0

F

(2)

where F is the Faraday constant (quantity of charge in Coulomb carried by one mole of electrons), I is the measured current, and t is time. Moles of hydrogen convert into the gas volumes at 1 atm,

298 K. At the initial stage (~300 s) of reaction, the current was observed to drastically decrease due to the hydrogen bubbles adhering to a silicon surface, which caused a loss of effective surface area. To quantitatively analyze the effectiveness of our hybrid system, the theoretical efficiency of a PEC-TE device was calculated for comparison with that of a PEC using Pt catalysts. The Pt-coated Si photocathode was prepared by immersing into the solution containing 0.4 M HF and 1 mM K2PtCl6 during the optimal deposition time (~10 min) which was reported previously [17]. Our experimental results were measured at AM 1.5 global solar distributions. Since the Pt-catalyzed PEC lowered the overpotential of 270 mV, spontaneous water splitting was able to occur at an Eg of 1.76 eV without external bias (Fig. 5a). If we define the threshold bandgap energy (Eth) as a minimum Eg for spontaneous water splitting, the Eth of a hybrid device can be lower than 2.03 eV according to the DT built across a TE device (Fig. 5b). When DT is larger than 10 C, we observed crossover behavior to reduce the overpotential, in which the hybrid system (TE254) was more effective than the Pt-catalyzed PEC. The Vbias can be explained by the following relation because the VTE further lowers the Eth,

Vbias

Fig. 4. Hydrogen evolution by Si photocathode measured as a function of time at DT¼ 18 C. Red circles (measured) and a dotted line (calculated) denote the volumes of collected hydrogen. Desorption of hydrogen bubbles from the Si surface caused sudden peaks at the current measurement (blue line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

8 > ð2:03 eV qVTE Þ Eg : q

if Eg Eth if Eg < Eth

(3)

where q is electron charge. This analysis assumes the standard potential of the reaction at 1 atm and at 298 K. In addition to the Vbias, the Jph is another crucial factor to determine the STH efficiency. Note that the Pt-catalyzed PEC cell recorded a lower Jph than the PEC cells without metal catalysts


155

Fig. 5. Numerical calculations for estimating solar-to-hydrogen efficiencies. (a) A comparison of the light J-V characteristics of a sole PEC, PEC-TE at DT ¼ 0 and Pt-catalyzed PEC (Pt/ PEC) devices. (b) Threshold bandgap energy of hybrid systems with TE142 (blue line) and TE254 (green line) as a function of DT. The black and red dotted-line represents the practical limit of bandgap energy for sole PEC (2.03 eV) and PEC with Pt catalyst (1.76 eV), respectively. (c) External biases and achievable photocurrent densities are plotted as a function of Eg at AM 1.5 global solar distributions. (d) STH efficiencies of PEC systems as a function of Eg. Black symbols denote the highest STH efficiencies recorded in previous research, i.e., , for Refs. [9,18], B for Ref. [19], △ for Ref. [20], - for Ref. [21], C for Ref. [22], and : for Ref. [23]. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

(Fig. 5a) due to the light shadowing and carrier recombination losses. Since a total loss in Jph by Pt covering is estimated to be ~10% of Jph, the quantum efficiency (QE) for Pt-catalyzed PEC cells was calculated, using the following equation, to be ~90% of the values obtained by PEC cells without Pt:

Z∞ Jph ¼ q

QEðEÞIs ðEÞdE

(4)

Eg

where Is is spectral irradiance of solar light as a function of photon energy. These calculated Jph values are plotted as a function of Eg in Fig. 5c. From Eq. (1) we can theoretically estimate the ultimate STH efficiencies of various PEC devices (Fig. 5d) using Vbias and Jph. Since the conversion efficiency shows the tradeoff between Vbias and Jph as a function of Eg, the STH efficiency initially increases until Eg is larger than Eth because Jph increases according to Eg. However, the efficiencies must decrease according to Eq. (3) because the available Gibbs energy (per photon) decreases for Eg values which are less than Eth. The PEC cells coupled with TE modules show a higher efficiency than sole (or Pt-catalyzed) PEC cells; in particular, PECTE254 reaches a STH efficiency of 32.2% at an Eg of 1.58 eV at Vbias ¼ 0. The maximum efficiency of the hybrid device was estimated to be 42.8% at an Eg of 1.36 eV. This result implies that a solution to overcome the limit of conventional PEC systems exists, i.e., the overpotentials are lowered via temperature gradient, while the band-edge alignments are adjusted accordingly. 4. Conclusion To obtain a reasonable solar-to-hydrogen efficiency, a great deal of research has been conducted on dual photosystems using

effective photocatalysts. Although conventional dual photosystems achieve higher photovoltage, total current density is limited by a current matching point formed by crossing between cathodic and anodic curves (see Supplementary Fig. S6) [5,6,12]. Typically, the quantum yield for conversion of absorbed photons into products would be 0.5 for dual photosystems and 1.0 for single photosystems [5,24]; however, the PEC-TE-hybrid system employing a single-semiconductor gains additional voltage from a thermal gradient while the overall current is determined by the Jph of the PEC cell (i.e., the TE is only used as a voltage source) [25]. This feature illustrates that the catalyst-free PEC performance in our hybrid system can be greatly improved by decoupling the photovoltage control from the photocurrent. As a result, a photon-to-current efficiency of ~20% was achieved at an 56 C temperature gradient by synergistically collecting photon and phonon energies, thereby utilizing the full solar spectrum. This approach can effectively make use of low-grade heat, and has the possibility of revolutionizing the industrial impact of water splitting. This might be necessary for regenerative operation of fuel-cell vehicle producing waste heat and clean water at the same time. Acknowledgments This research was supported by the Pioneer Research Center Program (No. 2009-0083007) and the grant (No. 2011-0028604) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning. This work was also supported by the New & Renewable Energy of the Korea Institute of Energy Technology Evaluation and Planning(KETEP) grant (No. 20123010010160) funded by the Korea government Ministry of Trade, Industry and Energy.

156


Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2015.01.020. References [1] T. Ohta, Int. J. Hydrog. Energy 13 (1988) 333e339. [2] T. Ohta, J.E. Funk, J.D. Porter, B.V. Tilak, Int. J. Hydrog. Energy 10 (1985) 571e576. [3] M.W. Chase, J. Phys. Chem. Ref. Data Monogr. 9 (1998) 1e1951. [4] T. Bak, J. Nowotny, M. Rekas, C.C. Sorrell, Int. J. Hydrog. Energy 27 (2002) 991e1022. [5] J.R. Bolton, S.J. Strickler, J.S. Connolly, Nature 316 (1985) 495e500. [6] M.F. Weber, M.J. Dignam, J. Electrochem. Soc. 131 (1984) 1258e1265. [7] A.B. Murphy, P.R.F. Barnes, L.K. Randeniya, I.C. Plumb, I.E. Grey, M.D. Horne, J.A. Glasscock, Int. J. Hydrog. Energy 31 (2006) 1999e2017. [8] I. Lombardi, S. Marchionna, G. Zangari, S. Pizzini, Langmuir 23 (2009) 12413e12420. [9] S.W. Boettcher, E.L. Warren, M.C. Putnam, E.A. Santori, D. Turner-Evans, M.D. Kelzenberg, M.G. Walter, J.R. McKone, B.S. Brunschwig, H.A. Atwater, N.S. Lewis, J. Am. Chem. Soc. 133 (2011) 1216e1219. [10] J.R. Mckone, E.L. Warren, M.J. Bierman, S.W. Boettcher, B.S. Brunschwig, N.S. Lewis, H.B. Gray, Energy Environ. Sci. 4 (2011) 3573e3583.

[11] S. Yae, L. Nakanishi, Y. Nakato, N. Toshima, H. Mori, J. Electrochem. Soc. 141 (1994) 3077e3081. [12] M.G. Walter, E.L. Warren, J.R. McKone, S.W. Boettcher, Q. Mi, E.A. Santori, N.S. Lewis, Chem. Rev. 110 (2010) 6446e6473. [13] Z. Chen, T.F. Jaramillo, T.G. Deutsch, A. Kleiman-Shwarsctein, A.J. Forman, N. Caillard, R. Garland, K. Takanabe, C. Heske, M. Sunkara, E.W. McFarland, K. Domen, E.L. Miller, J.A. Turner, H.N. Dinh, J. Mater. Res. 25 (2010) 3e16. [14] J.R. Mckone, N.S. Lewis, H.B. Gray, Chem. Mater. 26 (2014) 407e414. [15] R. Memming, Semiconductor Electrochemistry, WILEY-VCH Verlag GmbH, Weinheim, Germany, 2001. [16] S. Choudhary, S. Upadhyay, P. Kumar, N. Singh, V.R. Satsangi, R. Shrivastav, S. Dass, Int. J. Hydrog. Energy 37 (2012) 18713e18730. [17] I. Oh, J. Kye, S. Hwang, Nano Lett. 12 (2012) 298. [18] Y. Nakato, Y. Egi, M. Hiramoto, H. Tsubomura, J. Phys. Chem. 88 (1984) 4218e4222. [19] J.A. Baglio, G.S. Calabrese, D.J. Harrison, E. Kamieniecki, A.J. Ricco, M.S. Wrighton, G.D. Zoski, J. Am. Chem. Soc. 105 (1983) 2246e2256. [20] A. Heller, Science 223 (1984) 1141e1148. [21] E. Aharon-Shalom, A. Heller, J. Electrochem. Soc. 129 (1982) 2865e2866. [22] O. Khaselev, A. Bansal, J.A. Turner, Int. J. Hydrog. Energy 26 (2001) 127e132. [23] O. Khaselev, J.A. Turner, Science 280 (1998) 425e427. [24] S. Hu, C.-Y. Chi, K.T. Fountaine, M. Yao, H.A. Atwater, P.D. Dapkus, N.S. Lewis, C. Zhou, Energy Environ. Sci. 6 (2013) 1879e1890. [25] K.-T. Park, S.-M. Shin, A.S. Tazebay, H.-D. Um, J.-Y. Jung, S.-W. Jee, M.-W. Oh, S.-D. Park, B. Yoo, C. Yu, J.-H. Lee, Sci. Rep. 3 (2013) 2123.