Research Article www.acsami.org
Inhibition of Tafel Kinetics for Electrolytic Hydrogen Evolution on Isolated Micron Scale Electrocatalysts on Semiconductor Interfaces Robert H. Coridan,*,† Zebulon G. Schichtl,† Tao Sun,‡ and Kamel Fezzaa‡ †
Department of Chemistry and Biochemistry, University of Arkansas, CHEM 119, 1 University of Arkansas, Fayetteville, Arkansas 72701, United States ‡ X-ray Science Division, Advanced Photon Source, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, Illinois 60439, United States S Supporting Information *
ABSTRACT: Semiconductor−liquid junctions are ubiquitous in photoelectrochemical approaches to artificial photosynthesis. By analogy with the antennae and reaction centers in natural photosynthetic complexes, separating the light-absorbing semiconductor and electrocatalysts can improve catalytic efficiency. A catalytic layer can also impair the photovoltage-generating energetics of the electrode without appropriate microscopic organization of catalytically active area on the surface. Here, we have developed a method using high-speed X-ray phase contrast imaging to study in situ electrolytic bubble growth on semiconductor electrodes fabricated with isolated, micron-scale platinum electrocatalysts. X-rays are a nonperturbative probe by which gas evolution dynamics can be studied under conditions relevant to solar fuels applications. The self-limited growth of a bubble residing on the isolated electrocatalyst was measured by tracking the evolution of the gas−liquid boundary. Contrary to observations on macroscopic electrodes, bubble evolution on isolated, microscopic Pt pads on Si electrodes was insensitive to increasing overpotential. The persistence of the bubble causes mass transport limitations and inhibits the expected Tafel-like kinetics. The observed scaling of catalytic current densities with pad size implies that electrolysis is occurring predominantly on the perimeter of the active area. KEYWORDS: electrochemistry, catalysis, X-ray phase-contrast microscopy, solar fuels, bubble evolution, hydrogen evolution reaction
■
INTRODUCTION To become a competitive alternative to fossil fuels, solar energy requires a storage technology to account for the intermittency of sunlight and to enable on-demand power generation.1,2 Nature’s solution to this problem is to store solar energy in chemical bonds through photosynthesis.3 It is becoming increasingly viable to develop materials that mimic this process in integrated artificial photosynthetic systems that drive endothermic chemical reactions in solution.4,5 An example of this is the photoelectrochemical splitting of water into hydrogen and oxygen on semiconductor electrodes.6,7 Natural photosynthesis separates light absorption and catalysis into separate antennae and reaction center domains, respectively, in a hierarchical structure that operates more efficiently than a single domain performing both functions.8 In artificial systems, heterogeneous photoelectrochemical reactions at semiconductor−liquid junctions are commonly carried out with the addition of an electrocatalyst layer that exhibits a low catalytic overpotential for the reaction of interest.9 The semiconductor is chosen to efficiently convert photons into excited carriers, and the electrocatalyst improves the kinetics of the desired reaction compared to those of the semiconductor alone.10 A canonical example is the addition of a metal layer to a p-type, crystalline Si photocathode.11 Si has a relatively small band gap (Ebg = 1.1 © 2016 American Chemical Society
eV), which enables it to absorb a significant fraction of the solar spectrum. The photogenerated conduction band electrons have sufficient energy to reduce protons to molecular hydrogen.12 However, Si itself is a poor electrocatalyst for the hydrogen evolution reaction (HER) and requires a low overpotential electrocatalyst like Pt to achieve high efficiency photoelectrochemical performance.13,14 Natural photosynthetic complexes derive the ability to guide energy transfer from sunlight to storage from a sophisticated hierarchical structure that facilitates electron transfer through the reaction pathway.15 Similarly, the hierarchical structure and organization of the electrocatalyst layer on a semiconductor surface can have a significant effect on the potential energy gradient driving the reaction. Maximizing the electrochemically active surface area reduces catalytic overpotentials on a fixedarea electrode, but a continuous coverage layer of electrocatalyst on the semiconductor surface can impair the desired energetics of the semiconductor−liquid junction. In general, a good electrocatalyst will make an ohmic contact with the semiconductor to facilitate carrier transfer from the semiReceived: June 26, 2016 Accepted: August 30, 2016 Published: August 30, 2016 24612
DOI: 10.1021/acsami.6b07729 ACS Appl. Mater. Interfaces 2016, 8, 24612−24620
Research Article
ACS Applied Materials & Interfaces
dependent kinetics of HER as a function of applied potential and catalyst area. Rather than the Tafel-like kinetics observed on large scale, continuous electrodes, the bubbles generated on isolated electrocatalysts persist for long times and are pinned to the surface. The dynamics of the bubble growth shows that HER on these electrodes is mass-transport-limited. This is likely the effect of the limited mobility of reactant protons in the thin solvent layer between the bubble and the electrode. As a consequence, commonly used Tafel-like models based on current−overpotential measurements of bulk electrodes are inappropriate for describing catalysis on structured electrocatalysts in this regime. The organization of the electrocatalyst on the semiconductor interface is a critical consideration for evaluating materials and device designs for photo-electrochemical solar fuels applications.
conductor to the electrocatalyst. The semiconductor will equilibrate with an extended metal layer rather than with the solution. This disrupts the rectification of the semiconductor− liquid junction and limits any achievable photovoltage.16 Reducing the areal coverage and absolute size of metal catalysts has been shown to mitigate most of the negative energetic effects of the electrocatalyst−semiconductor contact.17,18 Additionally, the rate of light absorption in the semiconductor can be improved with a sparse distribution of electrocatalysts compared to a continuous metal layer by reducing reflection.19 Nanostructured metal particles can also be used to improve light absorption 20 or enhance catalysis via plasmonic excitations.21 An important difference between the reaction centers in natural photosynthetic complexes and the artificial electrocatalyst−semiconductor interfaces is the phase of the evolved products. NADPH is the liquid-phase proton reduction product in photosynthesis,22 and the rate of water oxidation is low enough to maintain the solubility of the oxygen produced.23 Most photoelectrochemical approaches to artificial photosynthesis produce gases like hydrogen and oxygen at rates high enough to saturate the local electrolyte and evolve bubbles under steady state operation. Gas evolution nucleates bubbles on the interface, forming a three-phase gas−liquid−solid boundary. The mass transport, electron transfer, and lightblocking issues resulting from evolving bubbles on photoelectrodes are an important consideration for any choice of hierarchical catalyst organization. Catalysts that perform efficiently with low overpotentials can evolve long-lived bubbles that block active area and limit catalytic current.24 This is generally not an issue for a continuous electrocatalyst layer but has an effect on localized, discrete catalysts. Significant increases to the overpotential required to maintain a given current density are observed on hydrogen bubble-evolving microelectrodes in acid.25 The relatively low current densities of solar-driven processes on semiconductors result in persistent bubble residence on the active surface, which can hinder systems engineered for high efficiency. The interaction between the bubble and the nonactive semiconductor surface is another potential complication. An appropriate description of bubble evolution dynamics on discrete electrocatalyst domains on semiconductors and the effects on catalytic kinetics is missing from descriptions of these interfaces, particularly for systemslevel approaches to artificial photosynthesis.26 In this work, we study the effects of bubble evolution for HER on artificial, microscopic electrocatalyst reaction centers patterned on semiconductor surfaces. Specifically, we use highspeed phase-contrast imaging to track the dynamics of hydrogen bubble evolution on isolated, micron-scale platinum electrocatalysts on degenerately doped n-type Si (a nonphotoactive cathode analogue of the p-type Si photocathode described above) in sulfuric acid. Synchrotron-based X-ray imaging techniques provide intense, penetrating radiation that allow for the direct observation of microsecond-scale dynamical processes on complex structures.27,28 More importantly, the penetration depth of X-rays allows for these structures to be studied in situ, without perturbing or impeding the targeted chemical processes, and under conditions identical to those of interest for photoelectrochemical solar fuels applications. The full-field phase-contrast imaging provides high-resolution identification of phase boundaries, like micrometer-sized bubbles in solution. By tracking the dynamics of the gas− liquid interface that defines the bubble, we measure the time-
■
MATERIALS AND METHODS
The 0.5 mm diameter platinum wire (99.9% trace metals basis) and methyl isobutyl ketone (Reagent grade, >98.5%) were used as received (Sigma-Aldrich). Concentrated sulfuric acid (VWR) was diluted to desired concentration with high-purity water (18 MΩ·cm, ELGA Labpure Ultra). Acetone, isopropanol, and methanol (ACS reagent grade, BDH Chemicals) were used as received. Electrode Fabrication. Degenerately doped n-type Si wafers (n+Si; ⟨100⟩-oriented, single-side polished, As-doped,
