Efficient Photoelectrochemical Hydrogen Evolution ... - ACS Publications

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Efficient Photoelectrochemical Hydrogen Evolution on Silicon Photocathodes Interfaced with Nanostructured NiP2 Cocatalyst Films Fengjiao Chen, Qishan Zhu, Yeyun Wang, Wei Cui, Xiaodong Su, and Yanguang Li ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b11197 • Publication Date (Web): 21 Oct 2016 Downloaded from http://pubs.acs.org on October 25, 2016

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.

ACS Applied Materials & Interfaces 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

ACS Applied Materials & Interfaces

Efficient Photoelectrochemical Hydrogen Evolution on Silicon Photocathodes Interfaced with Nanostructured NiP2 Cocatalyst Films Fengjiao Chen,1 Qishan Zhu,2 Yeyun Wang,1 Wei Cui,1 Xiaodong Su2* and Yanguang Li1* 1

Institute of Functional Nano & Soft Materials (FUNSOM), Jiangsu Key Laboratory for Carbon-

Based Functional Materials and Devices, Soochow University, Suzhou 215123, China 2

Department of Physics, Jiangsu Key Laboratory of Thin Films, Soochow University, Suzhou,

215006, China

ABSTRACT: Increasing attention has now been focused on the photoelectrochemical (PEC) hydrogen evolution as a promising route to transform solar energy into chemical fuels. Silicon is one of the most studied PEC electrode materials, but its performance is still limited by its inherent PEC instability and electrochemical inertness toward water splitting. To achieve significant PEC activities, silicon-based photoelectrodes usually have to be coupled with proper cocatalysts, and thus formed semiconductor/cocatalyst interface presents a critical structural parameter in the rational design of efficient PEC devices. In this study, we directly grow nanostructured pyrite-phase nickel phosphide (NiP2) cocatalyst films on textured pn+-Si photocathodes via on-surface reaction at high temperatures. The areal loading of the cocatalyst film can be tailored to achieve an optimal balance between its optical transparency and electrocatalytic activity. As a result, our pn+-Si/Ti/NiP2 photocathodes demonstrate a great PEC onset potential of 0.41 V versus reversible hydrogen electrode (RHE), decent photocurrent density of ~12 mA/cm2 at the thermodynamic potential of hydrogen evolution, and impressive

ACS Paragon Plus Environment

1


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

operation durability for at least 6 h in 0.5 M H2SO4. Comparable PEC performance is also observed in 1 M potassium borate buffer (pH = 9.5) using this device.

KEYWORDS: Photoelectrochemical hydrogen evolution, silicon photocathode, NiP2, onsurface synthesis; acidic and neutral electrolytes INTRODUCTION Photoelectrochemical (PEC) water splitting offers a promising path toward the sustainable production of hydrogen fuel using sunlight and water as sole inputs.1-3 Despite its great potential, several challenges have to be overcome prior to the realization of scalable PEC systems. The most critical one is the availability of suitable semiconductor/electrocatalyst couples for efficient light harvesting, electron-hole separation, and surface reactions (i.e. hydrogen or oxygen evolution).4-5 Among many semiconductor materials, Si is an attractive candidate by virtue of its excellent light absorption capability, high carrier mobility and earth abundance.6-8 It has been widely implemented in photovoltaic devices, but much less successfully for solar fuel generation. The major hurdles are its intrinsic PEC instability under aqueous conditions and electrochemical inertness toward water splitting.1 In recent years, tremendous efforts have been undertaken to address above problems with some encouraging progress.1, 4-5, 9-12 For example, in order to suppress the PEC corrosion of Si and increase its stability, several types of surface passivation layers, including transition metals and metal oxides have been explored.9, 13-14 Some of us have demonstrated that an ultrathin Ni layer from electron beam evaporation is sufficient to passivate the Si surface for sustained PEC water oxidation up to 80 h.9 On the other hand, integrating Si photoelectrodes with suitable cocatalysts for hydrogen or oxygen evolution reaction (HER/OER) can significantly accelerate surface


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


reactions and boost photocurrent density.9,

15

At present, Pt group metals are the best HER

electrocatalysts available, but their scarcity essentially renders them impractical for large-scale energy production.16 The search of low-cost alternatives with high HER activity and durability is an

ongoing

endeavor.17-18

Moreover,

for

PEC

water

splitting,

an

effective

semiconductor/cocatalyst interface is desired to facilitate the transfer of minority carriers.1, 9, 19 It thereby imposes additional requirements on how the semiconductor and cocatalyst should be integrated together. Learning from the past, it has become increasingly transparent to people that the most effective strategy always is to in-situ grow cocatalysts on semiconductor surfaces, via electrochemical deposition, photochemical deposition or on-surface synthesis.19-21 The future development of PEC devices demands continuous optimization of semiconductor/cocatalyst couples and their associated interfaces rather than merely individual components. Pyrite-phase nickel diphosphide (NiP2) was reported as a high-pressure phase about half a century ago.22 It is only since 2014 that its activity for HER electrocatalysis has been fully unveiled.23 Even though the integration of Si photocathodes with some other phosphide electrocatalysts (such as CoP and CoPS) has been explored,15, 24 the potential of NiP2 to couple with Si photocathode for PEC hydrogen evolution remains elusive. This is partly due to the lack of effective approach to integrate them together. To this end, we report in this article the onsurface synthesis and engineering of nanostructured NiP2 films atop textured pn+-Si photocathode for efficient and sustained PEC hydrogen evolution. Our devices demonstrate impressive PEC HER activity in both acidic and neutral condition.

RESULTS AND DISCUSSION


3


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

The stepwise synthetic procedure of NiP2-decorated pn+-Si photocathode is schematically illustrated in Figure 1. First, planar p-type Si wafers were etched by KOH to afford pyramidal surface texture of ~2 µm in depth. This step is aimed to increase the interfacial area and light trapping of the photoelectrodes.25 They were then doped with phosphorus using the drive-in diffusion method to form pn+-Si buried junction with an n+ layer of ~400 nm. Figure 2a shows the top-view scanning electron microscopy (SEM) image of the pyramidally textured pn+-Si electrode. The buried junction provides a large built-in potential to enhance the charge separation, and thereby decouples the band bending (hence photovoltage) from the energetic of the semiconductor-liquid interface. This strategy has been widely adopted in many PEC devices.26-27 Next, an optimal 10 nm Ti layer was deposited onto pn+-Si by electron beam evaporation (E-beam) to form pn+-Si/Ti. The metallic Ti layer is designed to form an ohmic contact with the n+-Si layer. At the same time, it improves the adhesion of HER cocatalyst and protects the underneath Si from contamination when on-surface synthesis of NiP2 is carried out.27-28 Following the successful Ti deposition, a 10 nm Ni layer was E-beam deposited on the top surface (pn+-Si/Ti/Ni), and subsequently converted to NiP2 by reacting with excessive red phosphorus powders at 450 oC for 2 h under Ar. The conversion is typically incomplete, and thus formed electrode is denoted as pn+-Si/Ti/NiP2+Ni. Figure 2b, c are the top-view SEM images of pn+-Si/Ti/NiP2+Ni at different magnifications. The most apparent modification of the electrode surface is the formation of a particulate film having an interesting lace-like pattern. Finally, metallic Ni residues in the top layer were removed by soaking the electrode in diluted H2SO4 solution for several hours (pn+-Si/Ti/NiP2). This acid treatment step is found to be vital to ensure good PEC performance since it not only fully exposes the active surface area of NiP2, but also removes the light-blocking Ni metal. After the acid treatment, the electrode surface is


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


considerably smoothened as shown from the top-view and cross-section SEM images (Figure 2d, e). The chemical composition of the top cocatalyst layer was confirmed by x-ray diffraction (XRD) and x-ray photoelectron spectroscopy (XPS). Since the cocatalyst layer converted from the standard 10 nm Ni was too thin to give reliable XRD signals, we decided to instead use electrodes converted from a much thicker Ni layer (~80 nm) for this particular study. The result is depicted in the Supporting Information Figure S1a, and is found well consistent with the standard XRD pattern of pyrite NiP2. It evidences that our on-surface reaction followed by the acid treatment can yield phase-pure NiP2. XPS measurement of the standard electrode discloses that P 2p3/2 peak of NiP2 is located at 129.1 eV (Figure S1b, c). This binding energy is lower than that of elemental phosphorus (129.9 ~ 130.5 eV), and closer to that of anionic phosphide (128.5 ~ 129.4 eV). The slight surface oxide of NiP2 — which seems to be inevitable for many phosphide compounds — is also reflected from the presence of the minor signal at ~134 eV (Figure S1c).29 Ni 2p3/2 peak of NiP2 has a binding energy of 854.8 eV, in a good agreement with divalent Ni (Figure S1d).30 Nevertheless, the absence of satellite peaks indicates that the electron structure of Ni in NiP2 is different from other divalent Ni compounds such as NiO or Ni(OH)2. In order to determine the thickness and dispersion of the NiP2 cocatalyst layer, we replicated the same experimental procedure on planar Si wafers, and examined their surface texture using atomic force microscopy (AFM). A typical image is depicted in Figure 2f. We can see that NiP2 nanoparticles uniformly coat the surface and form continuous islands. The profile of a random line scan (along the dotted white line) reveals that these nanoparticles have a height up to ~60 nm. Survey over the entire image indicates that the root-mean-square (rms) surface roughness is


5


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

13.6 nm. Furthermore, we attempted to analyze the areal density of NiP2 by digesting the specimens in aqua regia for inductively coupled plasma optical emission spectroscopy (ICPOES) measurements. However, the dissolved nickel and phosphorus concentrations were too low to permit reliable ICP-OES detection. As an alternative approach, we estimated its areal density based on the reaction stoichiometry by assuming all Ni was converted to NiP2, and accordingly derived a low value of ~18 µg/cm2. This number clearly represents an upper-bound estimate because a significant amount of Ni was known to stay unreacted after the on-surface synthesis. Nevertheless, it is already one to two orders of magnitude lower than the typical areal loading (0.1~1 mg/cm2) used in the electrocatalytic studies of most non-precious metal based HER materials.18 For PEC applications, the cocatalyst layer atop the light-absorbing semiconductor photoelectrode has to simultaneously meet two criteria.1, 31 On one hand, it needs to be as thin as possible to reduce the optical loss, allowing more light to transmit through to the underneath photoelectrode; on the other hand, it cannot be too thin in order to have a reasonable electrocatalytic activity. Note that with their low costs, non-precious metal based HER electrocatalysts in principle can be used at a much higher loading to offset their lower intrinsic electrocatalytic activities compared to precious metals. It is, however, unfortunate that the need for optical transparency in PEC negates the possibility of higher loading even for them. Finding a subtle balance between the optical transparency and electrocatalytic activity of the cocatalyst material is therefore of decisive importance to the optimal PEC performance of the whole system. In what follows, we will separately discuss the optical and electrocatalytic properties of the NiP2 cocatalyst layer.


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


To study the optical property, quartz slides were used as the substrate for the Ti/Ni deposition as well as the subsequent conversion to NiP2 and acid leaching, following the same procedure outlined in Figure 1. The transmission spectra of specimens at different stages were collected in the range of 400 to 1200 nm. From the data summarized in Figure 3a, we can see that the deposition of 10 nm Ti alone causes little optical loss. It has ~90% transmission at a wavelength of 500 nm. However, further Ni deposition results in a substantial reduction of light transmission to ~52%. After Ni is partly converted to NiP2, the specimen becomes almost opaque (Figure 3b). Acid treatment removes the light-blocking Ni residue and reduces the surface roughness. The light transmission of the acid-treated specimen is recovered to ~66% at 500 nm. It indicates that when pn+-Si/Ti/NiP2 is used as the PEC photocathode, about two thirds of the light can transmit through the NiP2 cocatalyst layer and the Ti adhesion layer, and reach the np+-Si buried junction. The optical loss is still substantial but acceptable. Further, to study the electrocatalytic property of NiP2 at such a low areal loading, we used degenerately doped silicon wafer (p++-Si) as the conductive substrate to replicate the same cocatalyst layer. Electrodes were then evaluated for their HER performances. Measurements were conducted alongside with a reference sample prepared by sequentially depositing 10 nm Ti and 5 nm Pt on p++-Si for the benchmarking purpose. In 0.5 M H2SO4, p++-Si/Ti/NiP2 exhibits an onset overpotential of ~120 mV, beyond which the HER cathodic current density starts to take off and reaches 10 mA/cm2 at η = 270 mV (Figure 4a). These values are comparable to many non-precious metal based electrocatalysts such as double-gyroid MoS2 film or vertically aligned MoSe2 film.32-33 Nevertheless, it is worth noting that its apparent activity is not as attractive as other NiP2 or Ni2P catalysts reported in the literature.23, 34 This observation should not be taken as a sign that our NiP2 cocatalyst is inherently less active; it can be rationalized given that the


7


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

areal loading of NiP2 in our study is 1~2 orders of magnitude lower than all other studies. Tafel analysis indicates a slope value of ~61 mV/decade. Moreover, it is widely accepted that HER electrocatalysis in neutral electrolytes is substantially more challenging than that in acidic or alkaline electrolytes. There are few existing HER electrocatalysts that can operate in neutral solution with satisfactory activities.35-36 Here, we find that p++-Si/Ti/NiP2 has decent activity under the neutral condition. In 1 M potassium borate (KBi) buffer with pH = 9.5, its onset overpotential, overpotential at j = 10 mA/cm2 and Tafel slope are ~200 mV, ~430 mV, and 72 mV/decade, respectively (Figure 4b). In addition to impressive HER activities, the NiP2 cocatalyst also demonstrate excellent stability in both 0.5 M H2SO4 and 1 M KBi. Its chronopotentiometric (V-t) curves at a constant cathodic current density of 10 mA/cm2 fluctuate due to the periodic accumulation and release of hydrogen bubbles on the electrode surface, but their overall trajectory supports that the HER activity is mostly sustained during the course of evaluation for at least 6 h (Figure 4c,d). Taken together, above structural, optical and electrocatalytic characterizations corroborate that a thin and active NiP2 cocatalyst layer can form atop textured Si photocathode via the on-surface reaction between Ni metal and P powder. By controlling the thickness of the pre-deposited Ni layer, we are able to carefully tailor the areal loading of converted NiP2 on the top surface so that it is active and durable for HER electrocatalysis, while at the same time has a reasonable optical transparency. Photoelectrochemical hydrogen evolution was assessed using textured pn+-Si/Ti/NiP2 as the working electrode in a customized electrochemical cell. A diagram of the pn+-Si/Ti/NiP2 photocathode is shown in Figure 5a. Textured pn+-Si/Ti/Pt with 5 nm Pt cocatalyst layer was also included for comparison. No current is measured for both electrodes under the dark condition.


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


Upon the illumination of 100 mW/cm2 AM 1.5 G simulated solar light from the front side, pn+Si/Ti/NiP2 displays an obvious cathodic response with an onset potential of 0.41 V vs. reversible hydrogen electrode (RHE) in 0.5 M H2SO4 (Figure 5b). This number is close to that of pn+Si/Ti/Pt (~0.43 V). The positive gain of ~0.61 V relative to the onset of HER electrocatalysis on p++-Si/Ti/NiP2 corresponds to the photovoltage generated from the pn+-Si buried junction, well consistent with values reported for p-base/n+-emitter single crystalline Si photovoltaics. At the thermodynamic potential of hydrogen evolution (0 V vs. RHE), pn+-Si/Ti/NiP2 delivers a photocurrent density of ~12 mA/cm2. Its saturation photocurrent density reaches ~16 mA/cm2, and slightly declines along the cathodic sweep due to the gradual accumulation of hydrogen bubbles covering up the electrode surface. The half-cell solar-to-electricity conversion efficiency is calculated to 2.6%. Note that the saturation photocurrent density recorded here is lower than the ideal value of Si photoelectrodes (~35 mA/cm2). This is reasonably expected given the optical loss mainly associated with the presence of the Ti/NiP2 layer. In addition, the nonoptimized and relatively thick n+ layer also absorbs some light, but does not contribute much to the overall photocurrent density due to its rapid electron-hole recombination. Nevertheless, we note that the NiP2 cocatalyst layer is vital to the high PEC performance of our devices. Control experiments show that pn+-Si or pn+-Si/Ti alone has much inferior photocurrent onsets under similar conditions (Figure S2a). The PEC HER activity of pn+-Si/Ti/NiP2 is comparable or even superior to many previous Si-based photocathodes interfaced with non-precious metal based HER cocatalysts such as amorphous MoSx, Ni-Mo alloy, CoP or CoPS.15, 24, 27, 37 Moreover, our pn+-Si/Ti/NiP2 also exhibits reasonable PEC HER activity in neutral solution. In 1 M KBi, its onset potential and saturation current density are 0.35 V and 19 mA/cm2, respectively (Figure 5c). Compared to the light polarization curve in the acidic medium, the polarization curve in KBi


9


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

has a smaller fill factor. We attribute this difference to the lower activity of NiP2 cocatalyst in 1 M KBi relative to 0.5 M H2SO4 as well higher solution resistance. The great PEC HER activity of pn+-Si/Ti/NiP2 is believed to be the direct consequence of two structural factors: the pyramidal textured pn+-Si buried junction provides a large built-in potential to enhance the separation of photogenerated electrons and holes; the on-surface synthesis of particulate NiP2 cocatalyst layer on the photocathode affords a strongly coupled semiconductor/cocatalyst interface to facilitate the charge transfer process on the interface and accelerate HER. Importantly, we find that the thicknesses of pre-deposited Ti and Ni layers are also critical parameters to ensure good PEC activities. With no or too thin (

Efficient Photoelectrochemical Hydrogen Evolution ... - ACS Publications

Recommend Documents