Giant Defect-Induced Effects on Nanoscale Charge Separation in

Subscriber access provided by University of Winnipeg Library

Communication

Giant defect-induced effects on nanoscale charge separation in semiconductor photocatalysts Ruotian Chen, Shan Pang, Hongyu An, Thomas Dittrich, Fengtao Fan, and Can Li Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b04245 • Publication Date (Web): 26 Dec 2018 Downloaded from http://pubs.acs.org on December 26, 2018

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 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 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.

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 26 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

Giant defect-induced effects on nanoscale charge separation in semiconductor photocatalysts Ruotian Chen1,2, Shan Pang1, Hongyu An1, Thomas Dittrich3,*, Fengtao Fan1,*& Can Li1,*

1State

Key Laboratory of Catalysis, Dalian National Laboratory for Clean Energy, The

Collaborative Innovation Centre of Chemistry for Energy Materials (iChEM), Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Zhongshan Road 457, Dalian 116023, China 2

University of Chinese Academy of Sciences, Beijing 100049, China

3Helmholtz-Zentrum

Berlin für Materialien und Energie GmbH, Institut für Silizium-

Photovoltaik, Kekuléstr. 5, 12489 Berlin, Germany *Correspondence and requests for materials should be addressed to T.D., F.F. and C.L. (email: [email protected], [email protected], [email protected]).

1 ACS Paragon Plus Environment

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 26

Abstract Defects can markedly impact the performance of semiconductor-based photocatalysts, where the spatial separation of photogenerated charges is required for converting solar energy into fuels. However, understanding exactly how defects affect photogenerated charge separation at nanometer scale remains quite challenging. Here, using time- and space- resolved surface photovoltage approaches, we demonstrate that the distribution of surface photogenerated charges and the direction of photogenerated charge separation are determined by the defects distributed within a 100-nm surface region of a photocatalytic Cu2O particle. This is enabled by the defect-induced charge separation process, arising from the trapping of electrons at the near-surface defect states and the accumulation of holes at the surface states. More importantly, the driving force for defect-induced charge separation is greater than 4.2 kV/cm and can be used to drive photocatalytic reactions. These findings highlight the importance of near-surface defect engineering in promoting photogenerated charge separation and manipulating surface photogenerated charges; further, they open up a powerful avenue for improving photocatalytic charge separation and solar energy conversion efficiency.

Keywords: Defect, photocatalysis, solar energy conversion, charge separation, surface photovoltage


Page 3 of 26 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

The conversion of sunlight into chemical fuels on semiconductor-based photocatalysts provides a promising route to sustainable energy production1,2. Efficient solar fuel production relies on the effective separation and migration of photogenerated charges to the semiconductor surface for chemical reactions3,4. Therefore, understanding photogenerated charge separation and transport in photocatalytic semiconductors is of great importance5-8. Defects, which always exist in a semiconductor, significantly affect the charge separation and transport processes9-11. Defect-induced localized states, which generally lie energetically within the semiconductor band gap, can selectively capture the approaching charge carriers, leading to spatial charge separation and thus an improved photocatalytic activity12,13. Conversely, these defect states can also serve as recombination centers for charge carriers, as predicted by the Shockley-Read-Hall (SRH) statistics14; this is treated as an important loss mechanism in solar energy conversion15,16. Moreover, charge transport is also affected by the defect states, which can alter the acceleration vectors of free carriers17 and generate potential barriers5 or energetic canals18. Therefore, a clear understanding of the effect of defects on charge separation is the key to enabling the improvement of charge separation and solar fuel production efficiency. However, such understanding remains highly challenging due to the multifold roles of the defects and the complicated charge separation processes, which occur over a wide time range in micro- or nanometer-scale photocatalytic semiconductors5,19,20. Herein, using time- and space- resolved surface photovoltage (SPV) techniques, 3 ACS Paragon Plus Environment

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 26

we study the effect of defects on charge separation on defect-engineered Cu2O single photocatalysts. We find that long-lived (μs-ms) charge separation and steady-state surface charge distribution are dependent on the dominating defects distributed within a 100-nm surface region. The near-surface defect states can initiate an inverted charge separation process in relation to conventional charge separation in the space charge region (SCR) by trapping of photogenerated charges. More importantly, the driving force for inverted charge separation is proven to be greater than the built-in electric field in the SCR (4.2 kV/cm) and can play an important role in solar energy conversion. Cu2O is a highly desirable semiconductor for applications in solar energy conversion21,22, but its performance is often limited by defects23,24. Well-defined cubic Cu2O single crystals were used as model photocatalysts due to their highly symmetrical structure (cubic phase, Fig. S1a) and identical {001} facet exposure of the single crystals (Fig. S2), which simplify the charge separation process probably affected by intrinsic asymmetric built-in electric fields25, crystal facets26,27 and grain boundaries5,18. Intrinsic defects in Cu2O are copper vacancies (VCu), which create accepter states and result in p-type conductivity28. Hole trap states were identified at energy levels 0.25 and 0.45 eV above the valence band edge of Cu2O and were assigned to VCu and VCusplit, respectively29. VCu can be partially passivated by hydrogen, whereas a (H – VCu) complex with a state near the middle of the band gap appears and acts as electron trap states30. In short, the oxidation state of copper is an indication of the type of copper defects. Apparent Cu2+ stands for VCu acting as a hole trap state, while Cu0 stands for (H – VCu) acting as an electron trap state31. Based on these observations, we prepared a 4 ACS Paragon Plus Environment

Page 5 of 26 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

series of cubic Cu2O single crystals via the electrochemical reduction of Cu2+ ions (producing hydrogen) and tuned the defects on the basis of the deposition current and reduction potential32,33. The corresponding samples are denoted as Cu2O-1, Cu2O-2, Cu2O-3 and Cu2O-4 according to their increasing deposition current density (Tab. S1). X-ray photo-electron spectroscopy (XPS) spectra show that the density of Cu2+ in the near-surface region of Cu2O crystals decreases from Cu2O-1 to Cu2O-4 (Fig. S3)34. Similar results could also be inferred from the variation in the Raman peak intensity caused by copper vacancy defects (Fig. S1b,c)35. The contributions of Cu0 and Cu2+ in Cu2O were separated by analyzing the Auger Cu LMM spectra36. A fitted peak in the range of 568.2 eV and 568.8 eV is assigned to the overlap of Cu0 (binding energy (BE) = 568.2 eV) and Cu2+ (BE = 568.8 eV) and shows a tendency towards lower BE from Cu2O-1 to Cu2O-4 (Fig. S4). The peak is divided into two contributions of Cu0 and Cu2+ as shown in Fig. 1a. From Cu2O-1 to Cu2O-4, the proportions of Cu0 and Cu2+ increase and decrease, respectively. The excess of Cu0 or Cu2+ was expressed by the ratio of the difference between the areas of the Auger Cu LMM peaks related to Cu0 (A(Cu0)) and Cu2+ (A(Cu2+)) and the area of the total Auger Cu LMM peak (A(Cu)). ∆𝑁𝑐𝑜𝑚𝑝 =

𝐴(𝐶𝑢0) ― 𝐴(𝐶𝑢2 + ) 𝐴(𝐶𝑢)

× 100%

(1)

The values of Ncomp (denoted as net defect density) describe the degree of compensation between the (H – VCu) defect states and VCu defect states, which amounted to -2.4%, -0.7%, 1.7%, and 3.3% for Cu2O-1 to Cu2O-4, respectively (Tab. S2). These quantitative results reveal an important feature of this sample series; the dominating defect states in the near-surface region are gradually changing from hole 5 ACS Paragon Plus Environment

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 26

trap states to electron trap states from Cu2O-1 to Cu2O-4. To quantitatively investigate the influence of these near-surface defects on charge separation, we mapped the single-particle SPV using surface photovoltage microscopy (SPVM)8,37. Figs. 1b – 1e show the SPVM images of representative Cu2O particles selected from the four samples. Although the morphologies (cube) and sizes (4 μm) of all particles are very similar (SEM images in Fig S2, AFM images in Fig S5), the steady-state SPV signals vary greatly for the four different particles. The photogenerated electrons (negative SPV, red color) are imaged on the surfaces of Cu2O1 and Cu2O-2 particles, while photogenerated holes (positive SPV, purple color) are imaged on the surfaces of Cu2O-3 and Cu2O-4 particles. These results were further confirmed by changes in contact potential difference (CPD) under chopped illumination (Fig. S6), which show a periodic variation in ΔCPD and are consistent with steady-state SPV signals in amplitude and sign. Fig. 1f compares the quantitative values of spatially resolved SPV signals with the net defect density (Ncomp). Interestingly, these two sets of data correlate very well with each other and hence suggest that more hole trap defects (VCu) in the near-surface region result in a greater photogenerated electron accumulation on the surface, whereas more electron trap defects (H-VCu) in the near-surface result in a higher accumulation of photogenerated holes on the surface. To further validate the relationship between defect and charge separation direction, we performed modulated SPV spectroscopy experiments (Fig. 1g)38,39. For absorption of photons within an energy range of 2.0 – 2.6 eV, the in-phase SPV signals 6 ACS Paragon Plus Environment

Page 7 of 26 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

are negative for Cu2O-1 and Cu2O-2 but positive for Cu2O-3 and Cu2O-4. This result reflects that photogenerated electrons are separated predominantly towards the external surfaces (Cu2O-1 and Cu2O-2) or internal bulk (Cu2O-3 and Cu2O-4) of the Cu2O crystals. The in-phase SPV signals at 2.2, 2.5 and 2.6 eV were compared with Ncomp (inset of Fig. 1g). The in-phase SPV signals changed sign at around Ncomp = 0, indicating that the variation in the dominating-defect type leads to an inversion of charge separation direction. The good correlations between SPV (steady-state SPV and modulated SPV) signals and Ncomp undoubtedly demonstrate the direct dependence of charge separation on defects in the near-surface region. Moreover, the modulated signals, which were phase-shifted by 90°, always display the opposite sign of the inphase signals (i.e., there is only one factor dominating charge separation), demonstrating that surface defects dominate the inverted charge separation process (Fig. S7)38. To gain an insight into the mechanism of defect-dominated charge separation process, we studied the dynamic behavior of photogenerated charges using transient SPV spectroscopy40. Figs. 2a-d show pseudocolor images of SPV as a function of decay time and photon energy for the four samples. Representative SPV transients at a specified super-band gap excitation (500 nm) and sub-band gap excitation (670 nm) are shown in Fig. 2e and 2f, respectively. At short times (time scale < 100 ns), SPV signals are negative (red color) at photon energies above 2 eV, i.e., above the band gap (Fig. S8) for all samples, independent of the surface defect types. As SPV signals at nanosecond time scales are dominated by fast charge separation in SCR41,42, the fast 7 ACS Paragon Plus Environment

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 26

negative SPV signals suggest that band bending behavior of all the samples is similar to that of a p-type-bulk semiconductor in depletion (downward band bending). At longer times (time scale ~ μs - ms), negative SPV signals at super-band gap excitation decrease for all samples due to carrier diffusion (or recombination) but remain negative for Cu2O-1 and Cu2O-2, whereas the SPV signals change into positive values (color from red to blue) for Cu2O-3 and Cu2O-4. The long-lived negative SPV signals for Cu2O-1 and Cu2O-2 can be interpreted as a relaxation limited by trapped charges at intrinsic defect states (VCu) and are demonstrated by a pronounced kink at about 10 μs when charge carriers diffuse (Fig. 2e). This effect indicates a general role of the intrinsic defect states, which can increase the lifetime of photogenerated charges separated by the built-in electric field in SCR to mediate the time gap between charge separation (ns) and surface reactions (ms-s)19,43. In contrast to the role of intrinsic defect states, more interestingly, the incorporation of artificial defects (H-VCu) in the near-surface regions of Cu2O-3 and Cu2O-4 leads to an inverted long-lived charge separation process with respect to charge separation in SCR. More importantly, this defect-induced charge separation process can be more dominant than that driven by the built-in electric field and thereby enables the inverted charge separation over long time periods. SPV at sub-band gap excitation is an indication of charge separation processes owing to the exchange of charges between the semiconductor bands and local defect states41. For Cu2O-1 and Cu2O-2, relatively low positive SPV signals appear between 1.7 and 1.9 eV at short times of about 100 ns, but they may disappear (Cu2O-2) or even change in sign (Cu2O-1) at longer times (Figs. 2a, b, and f). The positive signals are 8 ACS Paragon Plus Environment

Page 9 of 26 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

ascribed to the excitation of electrons from surface states44. The disappearance of positive signals and appearance of additional negative signals can be ascribed to the quenching of electrons by the built-in electric field and excitation from defect states in the SCR. In the case of Cu2O-3 and Cu2O-4, remarkable positive SPV signals (red color) appear between about 1.7 eV and 1.9 eV for time periods of up to hundreds of µs (Figs. 2c, d, and f). This is because electrons excited from the surface states are captured and stabilized to a longer lifetime by high-density electron trap states (H-VCu), thus restricting quenching by the built-in electric field. The results not only demonstrate the existence of high-density electron trap states in Cu2O-3 and Cu2O-4, but also suggest an origin for the inverted charge separation process in Cu2O-3 and Cu2O-4 at superband gap excitation; it arises from the trapping of photogenerated electrons at electron trap states and the accumulation of photogenerated holes at the surface states. To determine the distribution of electron trap states, a part of a Cu2O-4 particle was partially removed by local scratching (Fig. 3a-c). Before scratching, the whole {001} facet of the particle shows a homogenous distribution of separated photogenerated holes (Fig. 3d, Fig. S9). Local scratching leads to the disappearance of photogenerated holes (Fig. 3e, color from purple to green) and the emergence of photogenerated electrons (Fig. 3f, color from purple to red) on the surface. The results directly demonstrate that the electron trap states, which are the origin states for inverted charge separation, are mainly distributed in the near-surface region. Also, the emergence of photogenerated electrons further gives evidence for the p-type nature of the bulk of the Cu2O particle. 9 ACS Paragon Plus Environment

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 26

Spatially resolved surface photovoltage spectroscopy (SRSPS)7 was used to explore the quantitative relationship between surface photogenerated charges and scratching depth that is determined by AFM height profile (Fig. S10). SRSPS spectra at the scratched region (p1) show that positive SPV signals continuously decrease and become even negative with an increase in scratching depth (Fig. 3g). In contrast, the SRSPS spectra at the unscratched region (p2) show only little changes (Fig. S11), verifying that the removal of the surface region containing the electron trap states is responsible for changes in the SPV signals. SPV signals at a photon energy of 2.7 eV can reflect the density of electron trap states, owing to a large absorption coefficient ( = 5 × 106, large photo-generation) and high density of electron trap states in the absorption region (-1 = 20 nm, Fig. S8c). Accordingly, the trap states are distributed within 100-nm surface region and decrease sharply with an increase in depth toward the bulk (blue line, Fig. 3h). In contrast, SPV signals at a photon energy of 2.1 eV (corresponding to an absorption length of 5 μm, which is greater than particle size), are a result of the combination of two opposite charge separation processes via trap states in the near-surface region and built-in electric fields in the SCR. The combination of both mechanisms results in a nearly linear change in the SPV from positive to negative values with an increase in scratching depth (red line, Fig. 3h). The results suggest that the driving force for charge separation via trap states can be stronger than the conventional built-in electric field in the SCR in the peripheral 50-nm region, where the density of trap states is so high that the charges trapped in the defect states overwhelm the charges at the surface states. 10 ACS Paragon Plus Environment

Page 11 of 26 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

To further evaluate charge separation by built-in electric fields and trap states, we investigated the dependence of SPV on light intensity (Fig. 4a). Data recorded on Cu2O-1 shows a logarithmic dependence of negative SPV on the light intensity for nearly the entire intensity range (red line in Fig. 4a), thus providing evidence for charge separation by the built-in electric field in the SCR42,45. In the case of Cu2O-4, the dependence can be separated into two parts with respect to light intensity. At light intensities below 10 mW/cm2, positive SPV increases with an increase in the light intensity and saturates gradually (blue dashed line in Fig. 4a). This process is attributed to charge separation via trap states until the trap states are filled up. More interestingly, the maximum positive SPV decreases logarithmically with a further increase in light intensity (blue line in Fig. 4a), indicating that the charge separation is simultaneously dominated by a built-in electric field yielding negative SPV and trap states yielding a constant positive SPV. The results were further confirmed by photoconductive AFM measurements, which showed two different photogenerated charge transport processes for light intensity below and above 10 mW/cm2 (Fig. S12). A comparison of the slope of the blue line and red line reveals that the built-in electric field of Cu2O-4 is the same in direction and larger in magnitude than the built-in electric field of Cu2O-1. Quantitative calculations show that the strength of built-in electric field increases from 1.3 (Supplementary Note) to 4.2 kV/cm (Fig. S13) owing to increased band bending, as validated by the increased dark surface potential (Fig. S14). We attribute the increased band bending to the pinning of Fermi energy at surface (H-VCu) states, as illustrated by energy-band scheme in dark conditions (Fig. S15). 11 ACS Paragon Plus Environment

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 26

Fig. 4b schematically illustrate the effect of defects on the charge separation behavior in a Cu2O-4 particle. In the deep region (d > 50 nm), photogenerated charge carriers are dominantly separated by the built-in electric field (negative SPV) and then stabilized by intrinsic defects (VCu). This process is almost the same as that occurring in Cu2O-1 (Fig. S16). In the near-surface region (d < 50 nm), the high-density electron trap states enable a new charge separation process in a direction opposite to that in deeper regions, thus inverting the ultimate charge separation behavior. The new charge separation process is deduced to be the trapping of photogenerated electrons by electron trap states in the near-surface region and the trapping of photogenerated holes by the surface states. DFT calculations suggest that the incorporation of (H-VCu) at surface may modify surface states to be negatively charged (Fig. S17). The modified surface states, in favor of the accumulation of photogenerated holes at surface, collaborates with electron trap states in the near-surface region to allow the efficient charge separation process. These trap states stabilize the charge carriers over a long lifetime and lead to steady-state photogenerated hole accumulation at the surface (corresponding to 70 mV SPV). As the driving force of defect-induced charge separation process could be stronger than the 4.2-kV/cm built-in electric field, we exploited the potential application of this driving force in photoelectrochemical water splitting. Cu2O-4 and Cu2O-1 electrodes show typical anodic46 and cathodic21,22,24 photocurrent, respectively (Fig. 4c), yielding good agreement with surface photogenerated charges. The results demonstrate that the surface photogenerated holes separated via defect states can be used to drive 12 ACS Paragon Plus Environment

Page 13 of 26 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

photooxidation reactions and suggest that near-surface defect engineering can be an important strategy for providing giant driving force of photogenerated charge separation in solar energy conversion. In conclusion, using combined time- and space- resolved SPV approaches, we demonstrated the significant effect of defects on long-lived charge separation and steady-state charge distribution. The effect could be ascribed to the near-surface defect states that stabilized the photogenerated charges and initiated a new charge separation process by the trapping of photogenerated charges. The driving force for this new charge separation was estimated to be greater than the 4.2 - kV/cm built-in electric field in the SCR and was demonstrated to be potentially utilized in solar energy conversion. These findings highlighted the important role of defects in charge separation and will motivate research on defect engineering for enhancing solar energy conversion performance

in

photo-conversion

systems

such

as

photocatalysis

and

photoelectrocatalysis.


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 26

Figure 1. (a) De-convoluted Auger Cu LMM spectra of samples Cu2O-1 to Cu2O-4. The blue and red lines are the de-convoluted spectra of Cu0 and Cu2+. The characteristic binding energies of Cu0 and Cu2+ are marked by short dashed lines. (b-e) SPVM images of Cu2O particles selected from samples Cu2O-1 to Cu2O-4 (b to e, respectively). The images were measured under 450-nm laser illumination with light intensity of 10 mW/cm2. All scale bars are 2 μm. The colored SPV bar is applicable to all images. (f) Correlation between the degree of compensation between Cu0 and Cu2+ and the averaged single-particle SPV signals. Error bars of the averaged SPV values are obtained from several random selected particles of each sample. (g) Modulated in-phase SPV spectra of samples Cu2O-1 to Cu2O-4. The inset shows the dependence of in-phase SPV signals measured at 2.2, 2.5, and 2.6 eV on the defect density.


Page 15 of 26 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 2. (a-d) Pseudocolour images of the spectral and time- dependent distributions of SPV signals obtained for samples Cu2O-1 to Cu2O-4 (a to d, respectively). The colored SPV bar in d is applicable to all images. The red and blue colors correspond to negative and positive SPV signals, respectively. The dashed line corresponds to the band gap (Eg), which is about 1.92 eV. (e, f) SPV transients of four Cu2O samples collected at an excitation wavelength of 500 nm (e) and 670 nm (f), corresponding to photon energies of 2.48 eV (super-band gap excitation) and 1.85 eV (sub-band gap excitation), respectively.


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 26

Figure 3. (a-c) AFM images of one Cu2O-4 particle before (a) and after local scratching at depths of 48 and 77 nm (b and c, respectively). (d-f) Corresponding SPVM images of the Cu2O-4 particle at a photon energy of 2.4 eV before (d) and after local scratching at depths of 48 and 77 nm (e and f, respectively). The scratching depths are measured by using AFM height profiles (Fig. S10). Scale bars in a-f are 2 μm. (g) SRSPS spectra at the scratched regions (p1) with different scratching depths (d). The black, red, blue, pink, and green spectra correspond to scratching depths of 0, 16, 48, 77, and 108 nm, respectively. (h) SPV signals measured at photon energies of 2.7 eV (blue line) and 2.1 eV (red line) and plotted as functions of depth.


Page 17 of 26 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 4. (a) The dependence of SPV signals on light intensity in Cu2O-1 and Cu2O-4. The SPV signals were obtained by averaging the SPV values from an external Lock-In Amplifier with electronic noise (random errors) of 5 mV (error bar). (b) Schematic showing photogenerated charge separation within Cu2O-4. Charge separation near the surface (d < 50 nm) occurs between surface states and (H – VCu) states with surface states capturing photogenerated holes and (H – VCu) states trapping electrons. This process yields positive SPV, corresponding to a lower local vacuum level after illumination. Charge separation in the deep regions (d > 50 nm) is driven by the builtin electric field in the SCR and photogenerated holes are trapped by VCu, yielding negative SPV in this region. The potential distributions in depth is quantitatively determined by the comparison of calibrated surface potential of Cu2O-4 and Cu2O-1 (Fig. S14) and the calculation of bandbending of Cu2O-1 (Supplementary Note) and the measured saturated SPV values. (c) Current–potential characteristics in 0.5M Na2SO4 solution, using a Pt foil as the counter electrode, under chopped AM1.5 light illumination for Cu2O-1 and Cu2O-4 samples. 17 ACS Paragon Plus Environment

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 26

ASSOCIATED CONTENT Supporting Information. Experimental section, supplementary note, supplementary Tables S1-S2, and supplementary Figures S1-S17.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected], [email protected], [email protected] Notes The authors declare no competing financial interests.

ACKNOWLEDGMENT This work was supported by the National Natural Science Foundation of China (Grant No. 21633015, 21773228), the National Key Basic Research Program of China (973 Program, Grant No. 2014CB239403), and the Strategic Priority Research Program and Equipment Development Project of the Chinese Academy of Sciences, Grant No. XDB17000000, YJKYYQ20170002.

REFERENCES (1) Lewis, N. S. Science 2016, 351, aad1920. (2) Wang, Q.; Hisatomi, T.; Jia, Q.; Tokudome, H.; Zhong, M.; Wang, C.; Pan, Z.; Takata, T.; Nakabayashi, M.; Shibata, N.; Li, Y.; Sharp, I. D.; Kudo, A.; Yamada, T.; Domen, K. Nat. Mater. 2016, 15, 611. (3) Sambur, J. B.; Chen, T. Y.; Choudhary, E.; Chen, G.; Nissen, E. J.; Thomas, E. M.; Zou, N.; Chen, P. Nature 2016, 530, 77. (4) Wang, D.; Sheng, T.; Chen, J.; Wang, H.-F.; Hu, P. Nat. Catal. 2018, 1, 291. 18 ACS Paragon Plus Environment

Page 19 of 26 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

(5) Warren, S. C.; Voitchovsky, K.; Dotan, H.; Leroy, C. M.; Cornuz, M.; Stellacci, F.; Hebert, C.; Rothschild, A.; Gratzel, M. Nat. Mater. 2013, 12, 842. (6) Yang, Y.; Gu, J.; Young, J. L.; Miller, E. M.; Turner, J. A.; Neale, N. R.; Beard, M. C. Science 2015, 350, 1061. (7) Zhu, J.; Fan, F.; Chen, R.; An, H.; Feng, Z.; Li, C. Angew. Chem. Int. Edit. 2015, 54, 9111. (8) Chen, R.; Pang, S.; An, H.; Zhu, J.; Ye, S.; Gao, Y.; Fan, F.; Li, C. Nat. Energy 2018, 3, 655. (9) Queisser, H. J.; Haller, E. E. Science 1998, 281, 945. (10) Hoch, L. B.; Szymanski, P.; Ghuman, K. K.; He, L.; Liao, K.; Qiao, Q.; Reyes, L. M.; Zhu, Y.; El-Sayed, M. A.; Singh, C. V.; Ozin, G. A. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, E8011. (11) Chen, X. B.; Liu, L.; Yu, P. Y.; Mao, S. S. Science 2011, 331, 746. (12) Yan, X. D.; Tian, L. H.; Tan, X. Y.; Zhou, M. J.; Liu, L.; Chen, X. B. Mrs Commun. 2016, 6, 192. (13) Wang, H.; Yong, D.; Chen, S.; Jiang, S.; Zhang, X.; Shao, W.; Zhang, Q.; Yan, W.; Pan, B.; Xie, Y. J. Am. Chem. Soc. 2018, 140, 1760. (14) Shockley, W.; Read, W. T. Phys. Rev. 1952, 87, 835. (15) Yang, Y.; Yang, M. J.; Moore, D. T.; Yan, Y.; Miller, E. M.; Zhu, K.; Beard, M. C. Nat. Energy 2017, 2, 16207. (16) Bose, R.; Bera, A.; Parida, M. R.; Adhikari, A.; Shaheen, B. S.; Alarousu, E.; Sun, J.; Wu, T.; Bakr, O. M.; Mohammed, O. F. Nano Lett. 2016, 16, 4417. 19 ACS Paragon Plus Environment

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 26

(17) Kroeze, J. E.; Savenije, T. J.; Warman, J. M. J. Am. Chem. Soc. 2004, 126, 7608. (18) Luria, J.; Kutes, Y.; Moore, A.; Zhang, L. H.; Stach, E. A.; Huey, B. D. Nat. Energy 2016, 1. (19) Baxter, J. B.; Richter, C.; Schmuttenmaer, C. A. Annu. Rev. Phys. Chem. 2014, 65, 423. (20) Hesari, M.; Mao, X.; Chen, P. J. Am. Chem. Soc. 2018, 140, 6729. (21) Paracchino, A.; Laporte, V.; Sivula, K.; Gratzel, M.; Thimsen, E. Nat. Mater. 2011, 10, 456. (22) Pan, L.; Kim, J. H.; Mayer, M. T.; Son, M.-K.; Ummadisingu, A.; Lee, J. S.; Hagfeldt, A.; Luo, J.; Grätzel, M. Nat. Catal. 2018, 1, 412. (23) Lee, Y. S.; Chua, D.; Brandt, R. E.; Siah, S. C.; Li, J. V.; Mailoa, J. P.; Lee, S. W.; Gordon, R. G.; Buonassisi, T. Adv. Mater. 2014, 26, 4704. (24) Luo, J. S.; Steier, L.; Son, M. K.; Schreier, M.; Mayer, M. T.; Gratzel, M. Nano Lett. 2016, 16, 1848. (25) Li, L.; Salvador, P. A.; Rohrer, G. S. Nanoscale 2014, 6, 24. (26) Tan, C. S.; Hsu, S. C.; Ke, W. H.; Chen, L. J.; Huang, M. H. Nano Lett. 2015, 15, 2155. (27) Li, G.; Chen, X. B. ACS Appl. Energy Mater. 2018, 1, 4313. (28) Scanlon, D. O.; Morgan, B. J.; Watson, G. W.; Walsh, A. Phys. Rev. Lett. 2009, 103, 096405. (29) Scanlon, D. O.; Watson, G. W. J. Phys. Chem. Lett. 2010, 1, 2582. (30) Scanlon, D. O.; Watson, G. W. Phys. Rev. Lett. 2011, 106, 186403. 20 ACS Paragon Plus Environment

Page 21 of 26 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

(31) McShane, C. M.; Choi, K. S. Phys. Chem. Chem. Phys. 2012, 14, 6112. (32) Wang, S.; Chen, P.; Yun, J. H.; Hu, Y.; Wang, L. Angew. Chem. Int. Ed. Engl. 2017, 56, 8500. (33) Siegfried, M. J.; Choi, K. S. Adv. Mater. 2004, 16, 1743. (34) Tobin, J. P.; Hirschwald, W.; Cunningham, J. Appl. Surf. Sci. 1983, 16, 441. (35) Sander, T.; Reindl, C. T.; Giar, M.; Eifert, B.; Heinemann, M.; Heiliger, C.; Klar, P. J. Phys. Rev. B 2014, 90, 045203. (36) Poulston, S.; Parlett, P. M.; Stone, P.; Bowker, M. Surf. Interface Anal. 1996, 24, 811. (37) Zhu, J.; Pang, S.; Dittrich, T.; Gao, Y.; Nie, W.; Cui, J.; Chen, R.; An, H.; Fan, F.; Li, C. Nano Lett. 2017, 17, 6735. (38) Gross, D.; Mora-Sero, I.; Dittrich, T.; Belaidi, A.; Mauser, C.; Houtepen, A. J.; Da Como, E.; Rogach, A. L.; Feldmann, J. J. Am. Chem. Soc. 2010, 132, 5981. (39) Fengler, S.; Dittrich, T. J. Phys. Chem. C 2016, 120, 17777. (40) Dittrich, T.; Valle Rios, L. E.; Kapil, S.; Gurieva, G.; Rujisamphan, N.; Schorr, S. Appl. Phys. Lett. 2017, 110, 023901. (41) Kronik, L.; Shapira, Y. Surf. Sci. Rep. 1999, 37, 1. (42) Chen, R.; Fan, F.; Dittrich, T.; Li, C. Chem. Soc. Rev. 2018, 47, 8238. (43) Cowan, A. J.; Durrant, J. R. Chem. Soc. Rev. 2013, 42, 2281. (44) Lagowski, J.; Balestra, C. L.; Gatos, H. C. Surf. Sci. 1971, 27, 547. (45) Melo, M. A., Jr.; Wu, Z.; Nail, B. A.; De Denko, A. T.; Nogueira, A. F.; Osterloh, F. E. Nano Lett. 2018, 18, 805. 21 ACS Paragon Plus Environment

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 26

(46) McShane, C. M.; Choi, K. S. J. Am. Chem. Soc. 2009, 131, 2561.

Table of Contents artwork


Page 23 of 26 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 1. (a) De-convoluted Auger Cu LMM spectra of samples Cu2O-1 to Cu2O-4. The blue and red lines are the de-convoluted spectra of Cu0 and Cu2+. The characteristic binding energies of Cu0 and Cu2+ are marked by short dashed lines. (b-e) SPVM images of Cu2O particles selected from samples Cu2O-1 to Cu2O4 (b to e, respectively). The images were measured under 450-nm laser illumination with light intensity of 10 mW/cm2. All scale bars are 2 μm. The colored SPV bar is applicable to all images. (f) Correlation between the degree of compensation between Cu0 and Cu2+ and the averaged single-particle SPV signals. Error bars of the averaged SPV values are obtained from several random selected particles of each sample. (g) Modulated in-phase SPV spectra of samples Cu2O-1 to Cu2O-4. The inset shows the dependence of inphase SPV signals measured at 2.2, 2.5, and 2.6 eV on the defect density. 329x183mm (300 x 300 DPI)

ACS Paragon Plus Environment

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

Figure 2. (a-d) Pseudocolour images of the spectral and time- dependent distributions of SPV signals obtained for samples Cu2O-1 to Cu2O-4 (a to d, respectively). The colored SPV bar in d is applicable to all images. The red and blue colors correspond to negative and positive SPV signals, respectively. The dashed line corresponds to the band gap (Eg), which is about 1.92 eV. (e, f) SPV transients of four Cu2O samples collected at an excitation wavelength of 500 nm (e) and 670 nm (f), corresponding to photon energies of 2.48 eV (super-band gap excitation) and 1.85 eV (sub-band gap excitation), respectively. 319x181mm (300 x 300 DPI)


Page 24 of 26

Page 25 of 26 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 3. (a-c) AFM images of one Cu2O-4 particle before (a) and after local scratching at depths of 48 and 77 nm (b and c, respectively). (d-f) Corresponding SPVM images of the Cu2O-4 particle at a photon energy of 2.4 eV before (d) and after local scratching at depths of 48 and 77 nm (e and f, respectively). The scratching depths are measured by using AFM height profiles (Fig. S10). Scale bars in a-f are 2 μm. (g) SRSPS spectra at the scratched regions (p1) with different scratching depths (d). The black, red, blue, pink, and green spectra correspond to scratching depths of 0, 16, 48, 77, and 108 nm, respectively. (h) SPV signals measured at photon energies of 2.7 eV (blue line) and 2.1 eV (red line) and plotted as functions of depth. 205x176mm (300 x 300 DPI)


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

Figure 4. (a) The dependence of SPV signals on light intensity in Cu2O-1 and Cu2O-4. The SPV signals were obtained by averaging the SPV values from an external Lock-In Amplifier with electronic noise (random errors) of 5 mV (error bar). (b) Schematic showing photogenerated charge separation within Cu2O-4. Charge separation near the surface (d < 50 nm) occurs between surface states and (H – VCu) states with surface states capturing photogenerated holes and (H – VCu) states trapping electrons. This process yields positive SPV, corresponding to a lower local vacuum level after illumination. Charge separation in the deep regions (d > 50 nm) is driven by the built-in electric field in the SCR and photogenerated holes are trapped by VCu, yielding negative SPV in this region. The potential distributions in depth is quantitatively determined by the comparison of calibrated surface potential of Cu2O-4 and Cu2O-1 (Fig. S14) and the calculation of bandbending of Cu2O-1 (Supplementary Note) and the measured saturated SPV values. (c) Current– potential characteristics in 0.5M Na2SO4 solution, using a Pt foil as the counter electrode, under chopped AM1.5 light illumination for Cu2O-1 and Cu2O-4 samples. 406x200mm (300 x 300 DPI)


Page 26 of 26

Giant Defect-Induced Effects on Nanoscale Charge Separation in

Recommend Documents