An Optically and Electrochemically Decoupled ... - ACS Publications

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Communication

An Optically and Electrochemically Decoupled Monolithic Photoelectrochemical Cell for High Performance Solar-driven Water Splitting Seungtaeg Oh, Hakhyeon Song, and Jihun Oh Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b02023 • Publication Date (Web): 11 Aug 2017 Downloaded from http://pubs.acs.org on August 12, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Nano Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

An Optically and Electrochemically Decoupled Monolithic Photoelectrochemical Cell for High Performance Solar-driven Water Splitting Seungtaeg Oh1, Hakhyeon Song1, and Jihun Oh*1,2 1

Graduate School of Energy, Environment, Water, and Sustainability (EEWS), Korea Advanced

Institute of Science and Technology (KAIST), 291 Daehak-ro, Yuseong-gu, Daejeon 350-701, Republic of Korea 2

KI Institute for NanoCentury, Korea Advanced Institute of Science and Technology (KAIST),

291 Daehak-ro, Yuseong-gu, Daejeon 350-701, Republic of Korea

ABSTRACT Photoelectrochemical (PEC) cells have attracted much attention as a viable route for storing solar energy and producing value-added chemicals and fuels. However, the competition between light absorption and electrocatalysis at a restrained cocatalyst area on conventional planar-type photoelectrodes could limit their conversion efficiency. Here, we demonstrate a new monolithic photoelectrode architecture that eliminate the optical-electrochemical coupling by forming locally nanostructured cocatalysts on a photoelectrode. As a model study, Ni inverse opal (IO),

ACS Paragon Plus Environment

1

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 23

an ordered three-dimensional porous nanostructure, was used as a surface-area-controlled electrocatalyst locally formed on Si photoanodes. The optical-electrochemical decoupling of our monolithic photoanodes significantly enhances the PEC performance for the oxygen evolution reaction (OER) by increasing light absorption and by providing more electrochemically active sites. Our Si photoanode with local Ni IOs maintains an identical photo-limiting current density but reduces the overpotential by about 120 mV compared to a Si photoanode with planar Ni cocatalysts with the same footprint under 1 sun illumination. Finally, a highly efficient Si photoanode with an onset potential of 0.94 V vs. reversible hydrogen electrode (RHE) and a photocurrent density of 31.2 mA/cm2 at 1.23 V vs RHE in 1 M KOH under 1 sun illumination is achieved with local NiFe alloy IOs.

KEYWORDS: Photoelectrochemical water splitting, Photoelectrode, Device modeling, Inverse opal, Si photoanode.


2

Page 3 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Photosynthetic cells based on a semiconductor/liquid interface, also known as photoelectrochemical (PEC) cells, produce value-added chemicals and fuels such as hydrogen and various hydrocarbons by photo-driven electrolysis such as water electrolysis and CO2 reduction.1-4 In a photosynthetic cell, photo-excited electron-hole pairs in a semiconductor drive a desired (photo)electrolysis at a semiconductor/liquid interface. Because most of the semiconductors for photosynthetic cells have poor catalytic properties for photo-driven electrolysis, the semiconductors are usually enhanced with cocatalysts to reduce the overpotential for a high solar-driven electrolysis efficiency.5-7 For instance, the metal particles such as Pt, Ir and Ru, and Au and Cu are used to enhance PEC reactions on semiconductors for the hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) and for the CO2 reduction reaction (CO2RR), respectively.8-17 However, those metals are opaque and therefore, can block light absorption in the underlying photoelectrode if the size and distribution of the metal particles are poorly controlled. This limits the surface area of a cocatalyst on a photoelectrode and, thus, limits cocatalysts to using costly novel metals with excellent electrocatalytic performances.

Structured semiconductors, such as nano- and micro-wires and pores, can relax the coupling between light absorption and electrolysis of photoelectrodes by increasing the cocatalyst loading on the large surface area of structured semiconductors while maintaining or even enhancing light absorption in the semiconductors through the anti-reflection and/or light-trapping properties of the structured surface.18-24 In addition, the nano- and micro-structures can enhance the separation and injection of photoexcited carriers through a radial junction so that semiconductors with poor electronic properties, such as Fe2O3, WO3, and BiVO4, can be used.25-28 These enable the use of low cost and earth abundant cocatalysts and semiconductors for efficient photosynthetic cells.


3

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 23

However, the large surface area of a structured semiconductor leads to an increased surface recombination of photoexcited carriers at the semiconductor/electrolyte and/or increased semiconductor/cocatalysts interfaces which can lead to a poor photo-conversion efficiency.20, 2931

Herein, we present a new PEC cell architecture in which high-surface area nanostructured cocatalysts are locally defined on a planar semiconductor so that the optical absorption and electrochemical performance of the photoelectrode are decoupled without increasing the recombination of photocarriers. As a model system, we fabricated a Si photoanode with micropatterned Ni inverse opal (IO) nanostructures for the PEC oxygen evolution reaction (OER) in an alkaline solution. The role of the micro-patterned Ni IO nanostructures on a Si photoanode in the PEC OER was investigated by independently controlling the surface area and surface coverage of the local Ni IO nanostructures on the Si photoanode. Our Si photoanode with the micropatterned Ni IOs had an overpotential for the OER that was 120 mV less than that of a Si photoanode with a planar Ni patch with the same geometrical footprint in 1 M KOH under a simulated 1 sun illumination. In addition, when a micro-patterned NiFe alloy IO was formed on a Si photoanode, a superior PEC OER performance was achieved with a potential onset of 0.94 V and a photocurrent density of 31.2 mA/cm2 at 1.23 V vs RHE (reversible hydrogen electrode) in 1 M KOH under a simulated 1 sun illumination. Steady-state device modeling was also performed to provide the design principles of our photoelectrode architecture for efficient PEC photolysis.

Figure 1a shows a conventional planar photoelectrode structure for a PEC electrolysis reaction, in our case, an OER. When sunlight is incident to an n-type semiconductor, electrons


4

Page 5 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

and holes are generated in the semiconductor, and holes are transported to cocatalysts to oxidize water. Then, OER performance of a PEC cell is determined by the photo-response of a semiconductor and electrochemical properties of a cocatalyst. Figure 1b shows the simulated current-potential (j-V) curves of a photoelectrode having various cocatalyst footprints. Details of the simulated j-V curves of a PEC photoelectrode are described in the supporting information (SI). The half-cell solar-to-hydrogen efficiency (ηhalf-STH) which indicates the PEC OER performance of a photoelectrode is given by

ηhalf-STH =

.

×

× 100,

(1)

where Vapp and jpec represent the external bias applied to a photoelectrode and the PEC current density of a photoelectrode under a 1 sun illumination, respectively. Pin is the power density of the incident sunlight: Pin = 100 mW/cm2 for AM 1.5 G illumination. Note that ηhalf-STH is used to diagnose the performance of the photoelectrode and does not represent the STH efficiency under the zero-bias condition. Increasing the footprint of a cocatalyst can reduce the electrocatalytic overpotential by providing more reaction sites and enhance the PEC OER rate for the given semiconductor and cocatalyst materials. However, optical shadowing coupled with the cocatalyst footprint decreases the photocurrent and limits the PEC OER performance of a conventional photoelectrode (Figure 1b).32

Nano- or micro-structured electrocatalysts are well-known to enhance the performance of catalysts by increasing electrochemically active sites.33-35 Therefore, if structured electrocatalysts are formed locally on a semiconductor, as shown in Figure 1c, it can significantly reduce the electrochemical overpotential without increasing the footprint of the structured electrocatalysts


5

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 23

on the photoelectrode surface. This leads to a dramatic improvement in the PEC performance because the photocurrent by light absorption in the semiconductor is now decoupled from the surface area of the electrocatalysts. In fact, the maximum PEC OER rate for a given semiconductor can be achieved as the current density operating at ηhalf-STH approaches the photolimiting current density (Figure 1d).

In order to assess the impact of decoupled light absorption and electrocatalysis on the PEC performance, we conducted steady-state PEC device modeling of a buried junction photoanode with local three-dimensionally (3D) structured cocatalysts. The detailed information of the modeling can be found in the SI. In this device modeling, we used the simple photovoltaic characteristics of a 14.6%-efficient solar cell with a photovoltage of 570 mV without nonidealities such as parasitic resistances and electrocatalytic characteristics following the ButlerVolmer equation. The exchange coefficient (α) of the cocatalyst was chosen as 0.5 which represents a moderate electrochemical property of a cocatalyst. Figure 2 shows the simulated half-cell solar-to-hydrogen (STH) efficiency (ηhalf-STH) of a buried pn junction photoanode with local 3D-cocatalysts with various footprints (Acat), surface areas (A3D), and exchange current densities (jo).

Figure 2a shows the ηhalf-STH contour plot of a photoanode with a 3D-structured cocatalyst with a 10-fold increased surface area, i.e., A3D = 10, as a function of Acat and jo. In comparison, the ηhalf-STH of a photoanode with a planar cocatalyst is shown in Figure S1. As clearly seen in Figure 2a and S1, the local 3D cocatalyst significantly improves the ηhalf-STH of the photoanode compared to the planar cocatalyst because the independent electrochemical overpotential reduction is decoupled from the light absorption. For example, a photoanode with a planar


6

Page 7 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

cocatalyst with a jo of 0.1 mA/cm2 and an Acat of 26% has a maximum ηhalf-STH of 3.39% (Figure S1). Note that we assume the cocatalysts are completely opaque in this device simulation. However, if metal nanoparticles or a thin film of transparent materials such as metal-oxides are used as cocatalysts, they allow light transmission to the underlying semiconductors and can affect overall PEC performance (Figure S2).36-39 As shown in Figure S2, when the cocatalysts become transparent, Acat for maximum ηhalf-STH of a photoelectrode increases since more catalytically active sites are provided with reduced shadowing. When 3D cocatalysts with a high A3D are applied onto a photoelectrode, the overpotential reduction from the high surface area cocatalysts enables a smaller Acat so that higher light absorption can be used. Both the overpotential reduction and the enhanced light absorption cause a significant increase in the theoretical maximum ηhalf-STH of the photoanode with a local 3D cocatalyst. Indeed, a photoanode with a 3D cocatalyst with an A3D of 10 and an Acat of 18% has a maximum ηhalf-STH of 6% (see black dots in Figure 2a and S1).

The surface area of local 3D cocatalysts is also an imperative factor for the ηhalf-STH of a photoelectrode, as shown in Figure 2b. As expected, the ηhalf-STH of a photoanode with a local 3D cocatalyst is enhanced with the A3D for all jos. Interestingly, the enhancement in the ηhalf-STH predominantly appears when the electrocatalytic performance of 3D cocatalysts is poor. The poor catalytic performance results from a low jo and/or low surface area. For example, when the jo of a 3D cocatalyst is 10-3 mA/cm2, the ηhalf-STH increases by about 10.4 times as the A3D increases from 1 to 15. In contrast, when the jo is 1 mA/cm2, the ηhalf-STH increases only 1.6 times for the same increment of the A3D (Figure S3). Note that the former case, i.e., jo = 10-3 mA/cm2 and A3D = 15, corresponds to a cocatalyst with a jo of 1 mA/cm2 and an A3D of 0.015 as ηhalf-STH


7

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 23

∝ ∙ (see Eq. 1 in the SI). Nevertheless, the highest ηhalf-STH can be achieved when a high surface area of nanostructured cocatalysts with a high jo is formed locally on a photoelectrode.

To implement a photoelectrode structure with a local 3D cocatalyst experimentally, we constructed our model device using Si and Ni IO. Figure 3a shows the schematics of our model PEC device in which local Ni IOs are used as high surface area cocatalysts with a controlled Acat on an oxide-passivated p+n-Si photoanode. A Ni IO is an ordered 3D nanoporous network with a precisely controlled surface area and can be used as a highly efficient electrocatalyst for the OER in alkaline solutions. For example, we confirmed that the Ni IOs can reduce kinetic overpotentials for all ranges of OER current densities with an increasing surface area in a 1 M KOH solution (Figure S4). In addition, Si is used as a semiconductor due to its suitable bandgap (1.12 eV) for solar absorption with a large carrier lifetime: the long carrier lifetime in Si ensures the collection of photoexcited carriers to local Ni IO electrocatalyst patches separated by few tens of micrometers.40, 41 In Figure 3a, the buried p+n junction provides a photovoltage to drive the PEC OER, and the thermal SiO2 layer passivates the Si surface electrically and chemically for high performance and a durable PEC OER. For instance, we demonstrated that the micropatterned oxide passivated Si photoanode with planar cocatalysts operates for 24 hours without degradation during the PEC OER.32

Figure 3b shows the representative tilted-view scanning electron microscope (SEM) image of an oxide-passivated Si photoanode with Ni IO micropatches with an Acat of approximately 13%. The fabrication process is detailed in Figure S5 in the SI. Briefly, an oxide-passivated p+nSi photoanode with Au micro-patches was prepared as reported previously.32 Polystyrene (PS) nanoparticles, 600 nm in diameter, were electrophoretically deposited to cover the entire Si


8

Page 9 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

surface. Then, Ni was electrodeposited onto the PS template. Finally, the PS templates were dissolved in toluene, and the Ni IO micro-patches are selectively formed on the micro-patterned Au patches. Various layer thicknesses of Ni IOs were grown by adjusting the electrodeposition times to control the surface area of the Ni cocatalysts. Figure 3c-f show the cross-section SEM images of a micro-patterned planar Ni patch and Ni IO patches with 3, 7 and 9 layers on the Si photoanodes with the same 13 % footprint, respectively. The enhanced surface areas of the corresponding Ni IOs compared to the planar Ni are about 4.29, 9.73, and 12.45.42, 43 In addition, the Acats of the Ni IO cocatalysts were independently varied from 7 to 32 % (Figure 3g-i). Figure 4a shows the PEC OER performance of the oxide-passivated p+n-Si photoanodes with the planar and high surface area Ni micro-patches as a function of the increasing height of the Ni IOs in 1 M KOH under a 1 sun illumination. As clearly shown in Figure 4a, our photoelectrode with local cocatalyst nanostructures enhances the PEC OER performance of a Si photoanode by decoupling light absorption and electrolysis. First, the photo-limiting current densities of the Si photoelectrodes are nearly identical to about 30 mA/cm2 because the Acats of the Ni cocatalysts are about 11 – 13 % for all the photoelectrodes (Figure S6). Second, the overpotential of the Si photoanodes is dramatically reduced with the increasing height of the Ni IOs: the electrochemically active surface area (ECSA) of the Ni cocatalysts, determined from the area of the Ni oxidation peak near 0.9 V, increases about 6.4 times for the Ni IO with 9 layers compared to the planar Ni (inset of Figure 4a). Note that the ECSA ratio of the Ni IO/planar is lower than the estimated surface area ratio. We attribute it, in part, to roughening of the micropatterned Ni cocatalyst shown in Figs. 3 and S7. In particular, an overpotential reduction of 128 mV for producing 25 mA/cm2 is achieved for the Si photoanode with 7 layers of the Ni IO (Figure 3e). We also prepared a Si photoanode with local Ni IOs with similar height and


9

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 23

footprint in Figure 3e but with distinctively different surface area. As shown in Figure S8, this photoanode has 4 layers of Ni IO micro-patches with 1 µm of spherical pore diameter and exhibits about half surface area of Ni IO micro-patches in Figure 3e. Reduced surface area of this Si photoanode with local Ni IOs with 1 µm pores increases overpotential about ~ 75 mV compared to Si photoanodes with Ni micro-patches with 600 nm pores in Figure 3e during PEC OER (Figure S9). The overpotential reduction at other photocurrent densities for the Si photoanodes with various A3Ds for the Ni cocatalysts with a 13% footprint is shown in Figure S10 and Table 1. Interestingly, when the height of the Ni IO further increases to 9 layers, the overpotential reduction of the Si photoanode slightly decreases at a current density higher than 2.5 mA/cm2. In addition, turnover frequency (TOF) of Ni IO micro-patches on a Si photoanode is continuously decreasing with increasing height of IOs in spite of the high ECSA (see the detailed information about TOF calculation in the SI). We attribute this, mainly, to the concentration and/or bubble overpotentials associated with the transport of the reactants and products in/near the Ni IOs at a high OER current density. For example, the size of the channel which connects a spherical pore to the pore in the Ni IOs is about 130 nm; thus, diffusion of the reactants and products can be hindered if the thickness of the Ni IO increases. Alternatively, O2 bubbles could be trapped inside the NI IO and could effectively reduce the ECSA of the Ni IOs.

The PEC performance of our photoelectrodes can be further enhanced by optimizing the footprint of the catalysts to increase solar absorption while maintaining the overpotential reduction from the Ni IOs. Figure 4b shows the PEC current density-potential (j-V) curves of the oxide-passivated p+n-Si photoanode with the micro-patterned Ni IOs with different footprints (Figure 3g-i). These micro-patterned Ni IOs have about 6 layers of pores (Figure S11). The cocatalyst footprints on the Si photoanodes were varied from about 7 to 32% by adjusting the


10

Page 11 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

distance between the patches with fixed patch sizes (Figure 3g-i): the longest patch distance on the Si photoanodes was about 56 µm for a 7% footprint (Figure 3g). Note that the long distance does not affect carrier collection in the Si photoanodes with the micro-patterned catalyst due to the long lifetime of the photoexcited carriers in Si, while a high Acat over 40% degrades the photovoltage from shading (Figure S12). Figure 4b clearly shows that lowering the Acat on a photoelectrode surface can improve light absorption and drive a greater photocurrent density. For example, by decreasing the Acat of the Ni IOs from about 31% to 7%, the photo-limiting current density increases from about 22 to 32 mA/cm2 under a 1 sun illumination. Surprisingly, the Si photoelectrode with the 7% Ni IO footprint exhibits less overpotential when the photocurrent density is higher than 8 mA/cm2, despite the fact that the total surface area of the Ni IO with the 7% footprint is only half of that with the 14% footprint.

To examine the PEC performances of the buried Si photoelectrodes with the local Ni IOs, the ηhalf-STHs of our photoelectrodes were compared based on the PEC j-V curves in Figure 4. Obviously, the Ni nanostructure drives about a 3 times higher ηhalf-STH of our photoelectrodes by the improved OER kinetics from the high surface area of the Ni IOs for a given Acat (Figure S13a). This ηhalf-STH enhancement by the local Ni nanostructures is extended to all Acats of the Ni cocatalysts (Figure S13b). Strikingly, the ηhalf-STH does not decline and even slightly increases when the Acat is lower than 10%. This is in contrast with our device modeling where the ηhalf-STH decreases sharply when the Acat is lower than 15% (Figure 2a). We believe that this unexpected high ηhalf-STH of our photoelectrodes at a low Acat originates from the enhanced mass transport of the reactants/products due to the micro-patterned cocatalysts. Recently, it was reported that a micro-patterned cocatalyst structure on a photoelectrode can act as an ultra-microelectrode (UME) which is capable of reducing the concentration overpotential because of its easy-to-


11

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 23

diffuse structures.44 Indeed, our buried junction Si PEC cell with micro-patterned Ni catalysts has a nearly identical j-V curve to that of a planar Ni electrolyzer externally connected with a Si photovoltaic cell, despite the fact that the Ni coverage of our PEC cell is only 12% (Figure S14). Therefore, the combination of the UME-like 3D-structured cocatalyst on a photoelectrode results in a high ηhalf-STH for our photoelectrode with a lower Acat by enabling more light absorption.

As discussed in the device modeling in Figure 2, using a cocatalyst material with a high jo is critical for developing a highly efficient water splitting device. Hence, we formed NiFe IOs locally on our Si photoanode to further enhance the PEC OER performance (Figure S15). Alloying Fe in Ni dramatically improves the OER performance.33, 45-47 The NiFe IO is formed by electroplating, and the Ni/Fe ratio of our micro-patterned NiFe IO on a Si photoanode is about 4. Electrocatalytic performance of electroplated Ni and NiFe planar films is compared in Figure S16. Figure 5a shows the PEC j-V curves of the oxide-passivated p+n-Si photoanode with the micro-patterned NiFe IO. As shown in Figure 5a, the local NiFe IO reduces an additional 80 mV overpotential at 25 mA/cm2 compared to the local Ni IO on the Si photoanodes and produces about a 31.2 mA/cm2 photocurrent density at the water oxidation potential. The ηhalf-STH of a Si photoanode with the local NiFe IO is about 2.7%, an absolute 1.27% enhancement over the Ni IO counterpart. Moreover, the onset potential is about 0.94 V vs RHE which is among the best for the Si based photoanodes for the OER. In addition, our oxide-passivated p+n-Si photoanode with the micro-patterned NiFe IO operated continually at 1.5 V vs RHE for 11 hours without current degradation.

In summary, we proposed a novel photoelectrode architecture that breaks the coupling between the optical absorption and electrolysis for the solar driven water splitting reaction. By


12

Page 13 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

locally forming high surface area 3D cocatalysts on a photoelectrode, the restrained footprint of a cocatalyst, arising from the optical-electrochemical coupling, can be relaxed and therefore, lead to a dramatic efficiency enhancement of the photoelectrode by enabling more light absorption and providing more reaction sites. Design principles that focused on the role of the electrochemical parameters of the local cocatalyst, such as the jo, A3D, and Acat, for the PEC performance were quantitatively investigated by device modeling. An experimental model study based on a Si photoanode and Ni and NiFe IOs with an independently controlled Acat and A3D, confirmed the PEC OER performance enhancement by the overpotential reduction and photocurrent enhancement by increasing the A3D and decreasing the Acat of the local Ni IO, respectively. Moreover, it was found that the mass transport of reactants and products should be facilitated for highly efficient and low-cost PEC cells because our UME type local cocatalyst with a 12% footprint provides a nearly comparable performance to full-coverage electrocatalysts. With these device design principles, our Si photoanode with the NiFe IO has an onset of 0.94 V vs RHE and 31.2 mA/cm2 at the water oxidation potential for 11 hours under a 1 sun illumination. Therefore, our photoelectrode structure is a suitable strategy for highly efficient and costeffective PEC water splitting cells and can be easily applied to various cocatalysts and semiconductors for other important electrochemical reactions such as the HER and CO2 reduction reaction (CO2RR).

ASSOCIATED CONTENT

Supporting Information


13

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 23

Supporting information contains fabrication of Si photoanode with Ni or NiFe IO, experimental details, device modeling details, additional SEM images, additional plots, and j-V comparison of PV-EZ and PEC. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Authors *E-mail: [email protected] Notes The authors declare no competing financial interest. ACKNOWLEDGMENT This work was supported by the Korea CCS R&D Center (KCRC) grant funded by the Korea government (Ministry of Science, ICT and Future Planning) (No. NRF-2014M1A8A1049303)

REFERENCES 1. Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 15729-35. 2. Bard, A. J.; Fox, M. A. Acc. Chem. Res. 1995, 28, 141-145. 3. Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Chem. Rev. 2010, 110, 6446-6473. 4. Murphy, A.; Barnes, P.; Randeniya, L.; Plumb, I.; Grey, I.; Horne, M.; Glasscock, J. Int. J. Hydrogen energy 2006, 31, 1999-2017. 5. Zhong, D. K.; Gamelin, D. R. J. Am. Chem. Soc. 2010, 132, 4202-4207. 6. Kim, J.; Minegishi, T.; Kobota, J.; Domen, K. Jpn. J. Appl. Phys. 2011, 51, 015802. 7. Oh, J.; Deutsch, T. G.; Yuan, H.-C.; Branz, H. M. Energy Envrion. Sci. 2011, 4, 16901694. 8. Trasatti, S. J. Electroanal. Chem. 1972, 39, 163-184. 9. Trasatti, S. J. Electroanal. Chem. 1980, 111, 125-131. 10. Walton, D.; Burke, L.; Murphy, M. Electrochim. Acta 1996, 41, 2747-2751.


14

Page 15 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

11. Tsuji, E.; Imanishi, A.; Fukui, K.-i.; Nakato, Y. Electrochim. Acta 2011, 56, 2009-2016. 12. Spurgeon, J. M.; Velazquez, J. M.; McDowell, M. T. Phys. Chem. Chem. Phys. 2014, 16, 3623-3631. 13. Hori, Y., Electrochemical CO2 reduction on metal electrodes. In Mod. Aspects Electrochem., Springer: 2008; pp 89-189. 14. Hori, Y.; Takahashi, I.; Koga, O.; Hoshi, N. J. Mol. Catal. A: Chem. 2003, 199, 39-47. 15. Hori, Y.; Wakebe, H.; Tsukamoto, T.; Koga, O. Electrochim. Acta 1994, 39, 1833-1839. 16. Song, J. T.; Ryoo, H.; Cho, M.; Kim, J.; Kim, J. G.; Chung, S. Y.; Oh, J. Adv. Energy Mater. 2016, 7, 1601103. 17. Kuhl, K. P.; Cave, E. R.; Abram, D. N.; Jaramillo, T. F. Energy Envrion. Sci. 2012, 5, 7050-7059. 18. Branz, H. M.; Yost, V. E.; Ward, S.; Jones, K. M.; To, B.; Stradins, P. Appl. Phys. Lett. 2009, 94, 231121. 19. Chattopadhyay, S.; Huang, Y.; Jen, Y.-J.; Ganguly, A.; Chen, K.; Chen, L. Mater. Sci. Eng., R 2010, 69, 1-35. 20. Garnett, E.; Yang, P. Nano Lett. 2010, 10, 1082-1087. 21. Garnett, E. C.; Yang, P. J. Am. Chem. Soc. 2008, 130, 9224-9225. 22. Boettcher, S. W.; Spurgeon, J. M.; Putnam, M. C.; Warren, E. L.; Turner-Evans, D. B.; Kelzenberg, M. D.; Maiolo, J. R.; Atwater, H. A.; Lewis, N. S. Science 2010, 327, 185-187. 23. Hwang, Y. J.; Boukai, A.; Yang, P. Nano Lett. 2008, 9, 410-415. 24. Zhang, L.; Liu, C.; Wong, A. B.; Resasco, J.; Yang, P. Nano Res. 2015, 8, 281-287. 25. Kim, J. Y.; Magesh, G.; Youn, D. H.; Jang, J.-W.; Kubota, J.; Domen, K.; Lee, J. S. Sci. Rep. 2013, 3, 2681. 26. Tacca, A.; Meda, L.; Marra, G.; Savoini, A.; Caramori, S.; Cristino, V.; Bignozzi, C. A.; Pedro, V. G.; Boix, P. P.; Gimenez, S. ChemPhysChem 2012, 13, 3025-3034. 27. Kim, T. W.; Choi, K.-S. Science 2014, 343, 990-994. 28. Su, J.; Guo, L.; Yoriya, S.; Grimes, C. A. Cryst. Growth Des. 2009, 10, 856-861. 29. Feldmann, F.; Bivour, M.; Reichel, C.; Hermle, M.; Glunz, S. W. Sol. Energy Mater. Sol. Cells 2014, 120, 270-274. 30. Konin, A. Semicond. Sci. Technol. 2014, 29, 095009. 31. Dan, Y.; Seo, K.; Takei, K.; Meza, J. H.; Javey, A.; Crozier, K. B. Nano Lett. 2011, 11, 2527-2532. 32. Oh, S.; Oh, J. J. Phys. Chem. C 2015, 120, 133-141. 33. Hoang, T. T.; Gewirth, A. A. ACS Catal. 2016, 6, 1159-1164. 34. Lu, X.; Zhao, C. Nat. Commun. 2015, 6, 6616. 35. Tseung, A.; Vassie, P. Electrochim. Acta 1976, 21, 315-318. 36. Kenney, M. J.; Gong, M.; Li, Y.; Wu, J. Z.; Feng, J.; Lanza, M.; Dai, H. Science 2013, 342, 836-840. 37. Chen, Y. W.; Prange, J. D.; Dühnen, S.; Park, Y.; Gunji, M.; Chidsey, C. E.; McIntyre, P. C. Nat. Mater. 2011, 10, 539-544. 38. Sun, K.; McDowell, M. T.; Nielander, A. C.; Hu, S.; Shaner, M. R.; Yang, F.; Brunschwig, B. S.; Lewis, N. S. J. Phys. Chem. Lett. 2015, 6, 592-598. 39. Chen, L.; Yang, J.; Klaus, S.; Lee, L. J.; Woods-Robinson, R.; Ma, J.; Lum, Y.; Cooper, J. K.; Toma, F. M.; Wang, L.-W. J. Am. Chem. Soc. 2015, 137, 9595-9603. 40. Jacoboni, C.; Canali, C.; Ottaviani, G.; Quaranta, A. A. Solid-State Electron. 1977, 20, 77-89.


15

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 23

41. Seitz, L. C.; Chen, Z.; Forman, A. J.; Pinaud, B. A.; Benck, J. D.; Jaramillo, T. F. ChemSusChem 2014, 7, 1372-1385. 42. Eagleton, T. S.; Searson, P. C. Chem. Mater. 2004, 16, 5027-5032. 43. Huang, Y.-J.; Lai, C.-H.; Wu, P.-W.; Chen, L.-Y. J. Electrochem. Soc. 2010, 157, P18P22. 44. Chen, Y.; Sun, K.; Audesirk, H.; Xiang, C.; Lewis, N. S. Energy Envrion. Sci. 2015, 8, 1736-1747. 45. Luo, J.; Im, J.-H.; Mayer, M. T.; Schreier, M.; Nazeeruddin, M. K.; Park, N.-G.; Tilley, S. D.; Fan, H. J.; Grätzel, M. Science 2014, 345, 1593-1596. 46. Trotochaud, L.; Young, S. L.; Ranney, J. K.; Boettcher, S. W. J. Am. Chem. Soc. 2014, 136, 6744-6753. 47. Louie, M. W.; Bell, A. T. J. Am. Chem. Soc. 2013, 135, 12329-12337.

Figure 1. Design principles of the photoelectrochemical (PEC) cells. Schematics of a photoelectrode architecture with (a), planar, and (c), three dimensionally (3D) structured cocatalysts. The simulated photo electrochemical current density-potential (j-V) curves of a


16

Page 17 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

photoelectrode with (b), planar, and (d), 3D-structured cocatalysts. Acats indicate the proportion of the metal cocatalysts on whole surface of a photoelectrode, respectively. A3D is the increased surface area of the 3D catalyst structure with the given Acat. The insets and dots in figure 1b and d exhibit the maximum half-cell solar-to-hydrogen efficiency (ηhalf-STH) of photoelectrodes, and the point which producing ηhalf-STH, respectively. In a conventional PEC cell, the opticalelectrochemical coupling imposes an optimum Acat for the maximum PEC performance. In contrast, a photoelectrode with the 3D cocatalysts can achieve the ultimate PEC performance of the semiconductor and cocatalysts materials by decoupling the photoresponse and electrochemical properties.

Figure 2. The simulated half-cell STH efficiency of a photoelectrode with the 3D-structured cocatalysts. (a) The simulated half-cell STH efficiency ηSTH contour plot as a function of the exchange current density jo and the footprint Acat for the 3D-structured cocatalysts with a 10-fold increased surface area A3D. (b) The simulated ηSTH contour plot as a function of jo and A3D with a


17

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 23

10% Acat. The black dot is the maximum ηSTH of a photoelectrode with the local 3D cocatalysts with a jo of 10-1 mA/cm2.

Figure 3. Schematic design and scanning electron microscopy (SEM) images of the oxidepassivated p+n-Si photoanode with micro-patterned Ni inverse opals (IOs). (a) and (b) Schematic design and tilted-view SEM image of the oxide-passivated p+n-Si photoanode with the micro-


18

Page 19 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

patterned Ni IOs. (c)-(f) Cross-sectional-view SEM images of the oxide-passivated p+n-Si photoanode with the micro-patterned Ni with various surface areas. c, planar Ni and Ni IO with (d), 3 (e), 7, and (f), 9 layers, respectively. (g)- (i) Plan-view SEM images of the oxidepassivated p+n-Si photoanode with the micro-patterned Ni IO with various Acats. The Acats of the Ni IOs are (g), 7%, (h), 14%, and (i), 32%.

Figure 4. PEC OER properties of the oxide-passivated p+n-Si photoanode with micro-patterned Ni IOs. (a) PEC j-V curves of the oxide-passivated p+n-Si photoanode with the micro-patterned Ni with various A3Ds at a given Acat of 11 ~ 13%. As the layer thickness of the Ni IOs was increased to 9 layers, the surface area increased about 12.45 times compared to the planar counterpart with the same Acat. The inset is the enlarged j-V curve of the redox peak associated with the oxidation of Ni2+ which appeared near 0.9 V vs. RHE. (b) The PEC j-V curves of the oxide-passivated p+n-Si photoanodes with the micro-patterned Ni IO with various Acats of the


19

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 23

cocatalysts. All the Ni IOs have 6 layers, and the Acats are 7% (black line), 14% (red line), and 32% (blue line). All PEC measurements were done at a 20 mV/s scan rate in 1M KOH under a 1 sun illumination.


20

Page 21 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 5. PEC performance and stability of the oxide-passivated p+n-Si photoanode with micropatterned NiFe IOs. (a) PEC j-V curves of the oxide-passivated p+n-Si photoanode with the micro-patterned planar Ni (black line), Ni IO (red line), and NiFe IO (blue line). The Ni and NiFe IOs have 6 and 7 layers, respectively. All Si photoanodes have an Acat of about 7%. Note that alloying Fe in Ni invokes about a 80 mV cathodic shift in the PEC j-V curves including the onset potential compared to the Ni IO counterpart. The simple surface area enhancement by the Ni IO only causes the j-V curve to be steeper without a significant shift in the onset potential. (b) Stability test of the oxide-passivated p+n-Si photoanode with the micro-patterned NiFe IOs with 7 layers and an Acat 7% in 1M KOH under a simulated 1 sun illumination. The current fluctuation is believed to originate from oxygen bubble formation and detachment during the PEC operation.


21

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 23

Table 1. Current-density-dependent overpotential reduction of the oxide-passivated p+n-Si with the micro-patterned Ni IO as a function of the Ni IO thicknesses in 1M KOH under a simulated 1 sun illumination.

At current density 2

5 mA/cm

Overpotential reduction, △η(Planar – IO) 3 layers

7 layers

9 layers

39 mV

56 mV

53 mV

2

10 mA/cm

48 mV

70 mV

62 mV

25 mA/cm2

96 mV

128 mV

97 mV


22

Page 23 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Insert Table of Contents Graphic and Synopsis Here


23

An Optically and Electrochemically Decoupled ... - ACS Publications

Recommend Documents