Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Letter
  • Published:

Sub-particle reaction and photocurrent mapping to optimize catalyst-modified photoanodes

Abstract

The splitting of water photoelectrochemically into hydrogen and oxygen represents a promising technology for converting solar energy to fuel1,2. The main challenge is to ensure that photogenerated holes efficiently oxidize water, which generally requires modification of the photoanode with an oxygen evolution catalyst (OEC) to increase the photocurrent and reduce the onset potential3. However, because excess OEC material can hinder light absorption and decrease photoanode performance4, its deposition needs to be carefully controlled—yet it is unclear which semiconductor surface sites give optimal improvement if targeted for OEC deposition, and whether sites catalysing water oxidation also contribute to competing charge-carrier recombination with photogenerated electrons5. Surface heterogeneity6 exacerbates these uncertainties, especially for nanostructured photoanodes benefiting from small charge-carrier transport distances1,7,8. Here we use super-resolution imaging9,10,11,12,13, operated in a charge-carrier-selective manner and with a spatiotemporal resolution of approximately 30 nanometres and 15 milliseconds, to map both the electron- and hole-driven photoelectrocatalytic activities on single titanium oxide nanorods. We then map, with sub-particle resolution (about 390 nanometres), the photocurrent associated with water oxidation, and find that the most active sites for water oxidation are also the most important sites for charge-carrier recombination. Site-selective deposition of an OEC, guided by the activity maps, improves the overall performance of a given nanorod—even though more improvement in photocurrent efficiency correlates with less reduction in onset potential (and vice versa) at the sub-particle level. Moreover, the optimal catalyst deposition sites for photocurrent enhancement are the lower-activity sites, and for onset potential reduction the optimal sites are the sites with more positive onset potential, contrary to what is obtainable under typical deposition conditions. These findings allow us to suggest an activity-based strategy for rationally engineering catalyst-improved photoelectrodes, which should be widely applicable because our measurements can be performed for many different semiconductor and catalyst materials.

This is a preview of subscription content, access via your institution

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Super-resolution hole and electron surface reaction mapping.
Figure 2: Correlation between local iE responses, surface activities, and OEC effects.
Figure 3: η and Eon,GB changes of OEC-modified TiO2 nanorods.

Similar content being viewed by others

References

  1. Walter, M. G. et al. Solar water splitting cells. Chem. Rev. 110, 6446–6473 (2010)

    Article  CAS  PubMed  Google Scholar 

  2. Fujishima, A. & Honda, K. Electrochemical photolysis of water at a semiconductor electrode. Nature 238, 37–38 (1972)

    Article  ADS  CAS  PubMed  Google Scholar 

  3. Yang, J., Wang, D., Han, H. & Li, C. Roles of cocatalysts in photocatalysis and photoelectrocatalysis. Acc. Chem. Res. 46, 1900–1909 (2013)

    Article  CAS  PubMed  Google Scholar 

  4. Carroll, G., Zhong, D. & Gamelin, D. Mechanistic insights into solar water oxidation by cobalt-phosphate-modified α-Fe2O3 photoanodes. Energy Environ. Sci. 8, 577–584 (2015)

    Article  CAS  Google Scholar 

  5. Wilson, R. H. A model for the current-voltage curve of photoexcited semiconductor electrodes. J. Appl. Phys. 48, 4292–4297 (1977)

    Article  ADS  CAS  Google Scholar 

  6. Buurmans, I. L. C. & Weckhuysen, B. M. Heterogeneities of individual catalyst particles in space and time as monitored by spectroscopy. Nature Chem. 4, 873–886 (2012)

    Article  ADS  CAS  Google Scholar 

  7. Liu, C., Dasgupta, N. P. & Yang, P. Semiconductor nanowires for artificial photosynthesis. Chem. Mater. 26, 415–422 (2014)

    Article  CAS  Google Scholar 

  8. Warren, S. C. et al. Identifying champion nanostructures for solar water-splitting. Nature Mater. 12, 842–849 (2013)

    Article  ADS  CAS  Google Scholar 

  9. Roeffaers, M. B. J. et al. Super-resolution reactivity mapping of nanostructured catalyst particles. Angew. Chem. Int. Ed. 48, 9285–9289 (2009)

    Article  CAS  Google Scholar 

  10. Zhou, X. et al. Quantitative super-resolution imaging uncovers reactivity patterns on single nanocatalysts. Nature Nanotechnol. 7, 237–241 (2012)

    Article  ADS  CAS  Google Scholar 

  11. Zhang, Y. et al. Superresolution fluorescence mapping of single-nanoparticle catalysts reveals spatiotemporal variations in surface reactivity. Proc. Natl Acad. Sci. USA 112, 8959–8964 (2015)

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  12. Tachikawa, T., Yamashita, S. & Majima, T. Evidence for crystal-face-dependent TiO2 photocatalysis from single-molecule imaging and kinetic analysis. J. Am. Chem. Soc. 133, 7197–7204 (2011)

    Article  CAS  PubMed  Google Scholar 

  13. Xu, W. et al. Single-molecule electrocatalysis by single-walled carbon nanotubes. Nano Lett. 9, 3968–3973 (2009)

    Article  ADS  CAS  PubMed  Google Scholar 

  14. Liu, B. et al. Large-scale synthesis of transition-metal-doped TiO2 nanowires with controllable overpotential. J. Am. Chem. Soc. 135, 9995–9998 (2013)

    Article  CAS  PubMed  Google Scholar 

  15. Salvador, P., Garcia Gonzalez, M. L. & Munoz, F. Catalytic role of lattice defects in the photoassisted oxidation of water at (001) n-TiO2 rutile. J. Phys. Chem. 96, 10349–10353 (1992)

    Article  CAS  Google Scholar 

  16. Choi, W., Termin, A. & Hoffmann, M. R. The role of metal ion dopants in quantum-sized TiO2: correlation between photoreactivity and charge carrier recombination dynamics. J. Phys. Chem. 98, 13669–13679 (1994)

    Article  Google Scholar 

  17. Giocondi, J., Salvador, P. & Rohrer, G. The origin of photochemical anisotropy in SrTiO3 . Top. Catal. 44, 529–533 (2007)

    Article  CAS  Google Scholar 

  18. Villarreal, T. L., Gómez, R., Neumann-Spallart, M., Alonso-Vante, N. & Salvador, P. Semiconductor photooxidation of pollutants dissolved in water: a kinetic model for distinguishing between direct and indirect interfacial hole transfer. I. Photoelectrochemical experiments with polycrystalline anatase electrodes under current doubling and absence of recombination. J. Phys. Chem. B 108, 15172–15181 (2004)

    Article  CAS  Google Scholar 

  19. Salvador, P. Kinetic approach to the photocurrent transients in water photoelectrolysis at n-TiO2 electrodes. 1. Analysis of the ratio of the instantaneous to steady-state photocurrent. J. Phys. Chem. 89, 3863–3869 (1985)

    Article  CAS  Google Scholar 

  20. Gärtner, W. W. Depletion-layer photoeffects in semiconductors. Phys. Rev. 116, 84–87 (1959)

    Article  ADS  Google Scholar 

  21. Butler, M. A. Photoelectrolysis and physical properties of the semiconducting electrode WO3 . J. Appl. Phys. 48, 1914–1920 (1977)

    Article  ADS  CAS  Google Scholar 

  22. Reichman, J. Collection efficiency of low-mobility solar cells. Appl. Phys. Lett. 38, 251–253 (1981)

    Article  ADS  CAS  Google Scholar 

  23. Dotan, H., Sivula, K., Gratzel, M., Rothschild, A. & Warren, S. C. Probing the photoelectrochemical properties of hematite (α-Fe2O3) electrodes using hydrogen peroxide as a hole scavenger. Energy Environ. Sci. 4, 958–964 (2011)

    Article  CAS  Google Scholar 

  24. Surendranath, Y., Lutterman, D. A., Liu, Y. & Nocera, D. G. Nucleation, growth, and repair of a cobalt-based oxygen evolving catalyst. J. Am. Chem. Soc. 134, 6326–6336 (2012)

    Article  CAS  PubMed  Google Scholar 

  25. Khnayzer, R. S. et al. Structure and activity of photochemically deposited “CoPi” oxygen evolving catalyst on titania. ACS Catal. 2, 2150–2160 (2012)

    Article  CAS  Google Scholar 

  26. Klahr, B., Gimenez, S., Fabregat-Santiago, F., Bisquert, J. & Hamann, T. W. Photoelectrochemical and impedance spectroscopic investigation of water oxidation with “Co–Pi”-coated hematite electrodes. J. Am. Chem. Soc. 134, 16693–16700 (2012)

    Article  CAS  PubMed  Google Scholar 

  27. Gamelin, D. R. Water splitting: catalyst or spectator? Nature Chem. 4, 965–967 (2012)

    Article  ADS  CAS  Google Scholar 

  28. Lin, F. & Boettcher, S. W. Adaptive semiconductor/electrocatalyst junctions in water-splitting photoanodes. Nature Mater. 13, 81–86 (2013)

    Article  ADS  CAS  Google Scholar 

  29. Peter, L. M., Li, J. & Peat, R. Surface recombination at semiconductor electrodes: Part I. Transient and steady-state photocurrents. J. Electroanal. Chem. Interfacial Electrochem. 165, 29–40 (1984)

    Article  CAS  Google Scholar 

  30. Steinmiller, E. M. P. & Choi, K. S. Photochemical deposition of cobalt-based oxygen evolving catalyst on a semiconductor photoanode for solar oxygen production. Proc. Natl Acad. Sci. USA 106, 20633–20636 (2009)

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  31. Li, R. et al. Spatial separation of photogenerated electrons and holes among {010} and {110} crystal facets of BiVO4 . Nature Commun. 4, 1432 (2013)

    Article  ADS  CAS  Google Scholar 

  32. Xu, W., Jain, P. K., Beberwyck, B. J. & Alivisatos, A. P. Probing redox photocatalysis of trapped electrons and holes on single Sb-doped titania nanorod surfaces. J. Am. Chem. Soc. 134, 3946–3949 (2012)

    Article  CAS  PubMed  Google Scholar 

  33. Chen, T.-Y. et al. Concentration and chromosome-organization dependent regulator unbinding from DNA for transcription regulation in living Escherichia coli cells. Nature Commun. 6, 7445 (2015)

    Article  ADS  Google Scholar 

  34. Sambur, J. B. & Chen, P. Approaches to single-nanoparticle catalysis. Annu. Rev. Phys. Chem. 65, 395–422 (2014)

    Article  ADS  CAS  PubMed  Google Scholar 

  35. Byers, J. C., Güell, A. G. & Unwin, P. R. Nanoscale electrocatalysis: visualizing oxygen reduction at pristine, kinked, and oxidized sites on individual carbon nanotubes. J. Am. Chem. Soc. 136, 11252–11255 (2014)

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Chen, S. & Kucernak, A. Electrocatalysis under conditions of high mass transport rate: oxygen reduction on single submicrometer-sized Pt particles supported on carbon. J. Phys. Chem. B 108, 3262–3276 (2004)

    Article  CAS  Google Scholar 

  37. Meier, J., Friedrich, K. A. & Stimming, U. Novel method for the investigation of single nanoparticle reactivity. Faraday Discuss. 121, 365–372 (2002)

    Article  ADS  CAS  Google Scholar 

  38. Tel-Vered, R. & Bard, A. J. Generation and detection of single metal nanoparticles using scanning electrochemical microscopy techniques. J. Phys. Chem. B 110, 25279–25287 (2006)

    Article  CAS  PubMed  Google Scholar 

  39. Ebejer, N., Schnippering, M., Colburn, A. W., Edwards, M. A. & Unwin, P. R. Localized high resolution electrochemistry and multifunctional imaging: scanning electrochemical cell microscopy. Anal. Chem. 82, 9141–9145 (2010)

    Article  CAS  PubMed  Google Scholar 

  40. Li, Y., Cox, J. T. & Zhang, B. Electrochemical responses and electrocatalysis at single Au nanoparticles. J. Am. Chem. Soc. 132, 3047–3054 (2010)

    Article  CAS  PubMed  Google Scholar 

  41. Xiao, X. & Bard, A. J. Observing single nanoparticle collisions at an ultramicroelectrode by electrocatalytic amplification. J. Am. Chem. Soc. 129, 9610–9612 (2007)

    Article  CAS  PubMed  Google Scholar 

  42. Novo, C., Funston, A. M. & Mulvaney, P. Direct observation of chemical reactions on single gold nanocrystals using surface plasmon spectroscopy. Nature Nanotechnol. 3, 598–602 (2008)

    Article  CAS  Google Scholar 

  43. Shan, X. et al. Imaging the electrocatalytic activity of single nanoparticles. Nature Nanotechnol. 7, 668–672 (2012)

    Article  ADS  CAS  Google Scholar 

  44. Tang, M. L., Liu, N., Dionne, J. A. & Alivisatos, A. P. Observations of shape-dependent hydrogen uptake trajectories from single nanocrystals. J. Am. Chem. Soc. 133, 13220–13223 (2011)

    Article  CAS  PubMed  Google Scholar 

  45. Plessers, E. et al. Resolving interparticle heterogeneities in composition and hydrogenation performance between individual supported silver on silica catalysts. ACS Catal. 5, 6690–6695 (2015)

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Xing, J. et al. Stable isolated metal atoms as active sites for photocatalytic hydrogen evolution. Chem. Eur. J. 20, 2138–2144 (2014)

    Article  CAS  PubMed  Google Scholar 

  47. Ida, S., Kim, N., Ertekin, E., Takenaka, S. & Ishihara, T. Photocatalytic reaction centers in two-dimensional titanium oxide crystals. J. Am. Chem. Soc. 137, 239–244 (2015)

    Article  CAS  PubMed  Google Scholar 

  48. Sze, S. M. Semiconductor Devices, Physics and Technology 3rd edn (Wiley, 2011)

  49. Wang, Y., Warschkow, O. & Marks, L. D. Surface evolution of rutile TiO2 (100) in an oxidizing environment. Surf. Sci. 601, 63–67 (2007)

    Article  ADS  CAS  Google Scholar 

  50. Warschkow, O., Wang, Y., Subramanian, A., Asta, M. & Marks, L. D. Structure and local-equilibrium thermodynamics of the c(2 × 2) reconstruction of rutile TiO2 (100). Phys. Rev. Lett. 100, 086102 (2008)

    Article  ADS  CAS  PubMed  Google Scholar 

  51. Lindquist, S. E., Finnström, B. & Tegnér, L. Photoelectrochemical properties of polycrystalline TiO2 thin film electrodes on quartz substrates. J. Electrochem. Soc. 130, 351–358 (1983)

    Article  ADS  CAS  Google Scholar 

  52. Fàbrega, C. et al. Tuning the Fermi level and the kinetics of surface states of TiO2 nanorods by means of ammonia treatments. J. Phys. Chem. C 117, 20517–20524 (2013)

    Article  CAS  Google Scholar 

  53. Peterson, M. W., Turner, J. A. & Nozik, A. J. Mechanistic studies of the photocatalytic behavior of titania: particles in a photoelectrochemical slurry cell and the relevance to photodetoxification reactions. J. Phys. Chem. 95, 221–225 (1991)

    Article  CAS  Google Scholar 

  54. Shen, H., Zhou, X., Zou, N. & Chen, P. Single-molecule kinetics reveals a hidden surface reaction intermediate in single-nanoparticle catalysis. J. Phys. Chem. C 118, 26902–26911 (2014)

    Article  CAS  Google Scholar 

  55. Chen, H., Wei, Z., Yan, K., Bai, Y. & Yang, S. Unveiling two electron-transport modes in oxygen-deficient TiO2 nanowires and their influence on photoelectrochemical operation. J. Phys. Chem. Lett. 5, 2890–2896 (2014)

    Article  CAS  PubMed  Google Scholar 

  56. Surendranath, Y., Dinc, M. & Nocera, D. G. Electrolyte-dependent electrosynthesis and activity of cobalt-based water oxidation catalysts. J. Am. Chem. Soc. 131, 2615–2620 (2009)

    Article  CAS  PubMed  Google Scholar 

  57. Zhong, D. K., Cornuz, M., Sivula, K., Gratzel, M. & Gamelin, D. R. Photo-assisted electrodeposition of cobalt-phosphate (Co-Pi) catalyst on hematite photoanodes for solar water oxidation. Energy Environ. Sci. 4, 1759–1764 (2011)

    Article  CAS  Google Scholar 

  58. Ye, H., Park, H. S. & Bard, A. J. Screening of electrocatalysts for photoelectrochemical water oxidation on W-doped BiVO4 photocatalysts by scanning electrochemical microscopy. J. Phys. Chem. C 115, 12464–12470 (2011)

    Article  CAS  Google Scholar 

  59. Abdi, F. F. & van de Krol, R. Nature and light dependence of bulk recombination in Co-Pi-catalyzed BiVO4 photoanodes. J. Phys. Chem. C 116, 9398–9404 (2012)

    Article  CAS  Google Scholar 

  60. Pilli, S. K. et al. Efficient photoelectrochemical water oxidation over cobalt-phosphate (Co-Pi) catalyst modified BiVO4/1D-WO3 heterojunction electrodes. Phys. Chem. Chem. Phys. 15, 14723–14728 (2013)

    Article  CAS  PubMed  Google Scholar 

  61. Çakir, S. & Arslan, E. Voltammetry of resazurin at a mercury electrode. Chem. Pap. 64, 386–394 (2010)

    Article  CAS  Google Scholar 

  62. Cooper, G., Turner, J. A. & Nozik, A. J. Mott-Schottky plots and flatband potentials for single crystal rutile electrodes. J. Electrochem. Soc. 129, 1973–1977 (1982)

    Article  ADS  CAS  Google Scholar 

  63. Bolts, J. M. & Wrighton, M. S. Correlation of photocurrent-voltage curves with flat-band potential for stable photoelectrodes for the photoelectrolysis of water. J. Phys. Chem. 80, 2641–2645 (1976)

    Article  CAS  Google Scholar 

  64. Möllers, F., Tolle, H. J. & Memming, R. On the origin of the photocatalytic deposition of noble metals on TiO2 . J. Electrochem. Soc. 121, 1160–1167 (1974)

    Article  ADS  Google Scholar 

  65. Laser, D. & Gottesfeld, S. Photocurrents induced by subbandgap illumination in a Ti-oxide film electrode. J. Electrochem. Soc. 126, 475–478 (1979)

    Article  ADS  CAS  Google Scholar 

  66. Abrantes, L. M. & Peter, L. M. Transient photocurrents at passive iron electrodes. J. Electroanal. Chem. Interf. Electrochem. 150, 593–601 (1983)

    Article  CAS  Google Scholar 

  67. Peter, L. M. Dynamic aspects of semiconductor photoelectrochemistry. Chem. Rev. 90, 753–769 (1990)

    Article  CAS  Google Scholar 

  68. Zhang, Q. et al. Density of deep trap states in oriented TiO2 nanotube arrays. J. Phys. Chem. C 118, 18207–18213 (2014)

    Article  CAS  Google Scholar 

  69. Sagara, T. & Sukigara, M. Photogenerated surface states at n-TiO2 / aqueous solution interface. J. Electrochem. Soc. 135, 363–367 (1988)

    Article  ADS  CAS  Google Scholar 

  70. Tafalla, D., Pujadas, M. & Salvador, P. Direct measurements of flat-band potential shifts under illumination of the semiconductor-electrolyte interface by electrolyte electroreflectance. Surf. Sci. 215, 190–200 (1989)

    Article  ADS  CAS  Google Scholar 

Download references

Acknowledgements

The research is supported by the Department of Energy, Office of Science, Basic Energy Science (grant number DE-FG02-10ER16199), and in part by Army Research Office (grant numbers 63767-CH and 65814-CH), National Science Foundation (grant numbers CBET-1263736 and CHE-1137217 (to J.B.S.), and REU programme DMR-1063059 (to E.J.N. and E.M.T.)), and the Petroleum Research Foundation (grant number 54289-ND7). This work used the Cornell Center for Materials Research Shared Facilities, an NSF MRSEC programme (grant number DMR-1120296), and the Cornell NanoScale Facility, a member of the NSF National Nanotechnology Infrastructure Network (grant number ECCS-15420819). We thank D. Hill of the Cahoon Group at the University of North Carolina at Chapel Hill for the finite-difference frequency-domain simulations presented in Supplementary Fig. 24. We also thank J. McKone, S. Maldonado, B. A. Parkinson, H. D. Abruña and A. J. Nozik for discussions, J. Colson and W. Dichtel for X-ray diffraction measurements, J. Grazul for electron microscopy and H. D. Abruña for electrochemical instrumentation.

Author information

Authors and Affiliations

Authors

Contributions

J.B.S. and P.C. designed research. J.B.S. performed research. T.-Y.C., E.C., G.C., E.J.N., E.M.T. and N.Z. contributed to research. J.B.S. and P.C. analysed data and wrote the manuscript.

Corresponding author

Correspondence to Peng Chen.

Ethics declarations

Competing interests

J.B.S. and P.C. are filing a provisional patent based on this study.

Extended data figures and tables

Extended Data Figure 1 Carrier-selective fluorogenic probe reactions, and electrochemistry and energy-level alignment of the probe molecule redox potentials relative to TiO2 band edges.

a, The oxidative N-deacetylation of amplex red to resorufin probes photogenerated holes (or consequent oxidizing species, such as surface-adsorbed hydroxyl radicals, OH radicals) and the reductive N-deoxygenation of resazurin to resorufin probes photogenerated electrons. b, Cyclic voltammetry of 0.5 mM resazurin (Rz, red trace) and 1.6 mM amplex red (AR, black and blue traces) in N2-purged electrolyte. Scan rate is 2 mV s−1 and the working electrode is glassy carbon. The irreversible reduction wave with a peak potential Ep,Rz = −0.21 V is due to the reduction of resazurin to resorufin (product P)13,61. The reversible redox waves with a half-wave potential E1/2 = −0.32 V are from the resorufin/dihydroresorufin redox couple (that is, resorufin/PH2)13,61. The irreversible oxidation wave was assigned to the irreversible oxidation of amplex red to resorufin, with a peak potential Ep,AR = +0.37 V. c, Energy-level diagram of the peak potentials of resazurin reduction (Ep,Rz) and amplex red (Ep,AR) oxidation relative to the conduction (ECB) and valence band (EVB) edges of bulk TiO2. The blue and red arrows schematically illustrate the energetically possible pathways of photogenerated holes and electrons to amplex red and resazurin, respectively. hv indicates incident light.

Extended Data Figure 2 TiO2 nanoparticles photocatalytically oxidize amplex red and reduce resazurin to generate resorufin.

Ensemble-level photocatalysis experiments were performed using P25 titanium dioxide nanoparticles, the precursor material for TiO2 nanorods, because they were easily dispersed in aqueous solution. The nanorods were easily dispersed in ethanol–water mixtures, but this solvent is not ideal for studying amplex red oxidation because ethanol could preferentially react with photogenerated holes. a and b, Fluorescence and ultraviolet–visible absorption spectroscopy confirm the generation of resorufin from amplex red oxidation (a) or resazurin reduction (b) photo-catalysed by TiO2 nanoparticles. a, Fluorescence spectra of a 10-μM amplex red solution in an aerated pH 8.3 electrolyte as a function of 365-nm LED illumination time (45 mW cm−2). b, Ultraviolet–visible absorption spectra of a 10-μM resazurin solution in N2-purged pH 8.3 electrolyte as a function of illumination time; no isosbestic point is observed in the data, suggesting that resorufin is further consumed in the photocatalytic reaction. ce, Photocatalytic kinetics of resazurin reduction (c, d) and amplex red oxidation (e) as a function of electrolyte composition. c, Fluorescence intensity at 585 nm versus time data used to determine the initial rate from a linear fit of the data after the 365 nm LED light is turned on (red line) for photocatalytic reduction of resazurin. The arbitrary fluorescence units were converted into concentration units by making a calibration curve using serial dilution of resorufin standards. NPs are nanoparticles. d, Initial rate of resazurin reduction versus bulk resazurin concentration at a fixed light intensity in aerated (red open circles) and N2-purged (blue open squares) electrolyte. The initial rate saturates with increasing resazurin concentration, [Rz], which suggests that the reaction occurs on the catalyst surface and the number of available surface sites on the catalyst limits the initial reaction rate at high concentrations of the reactant resazurin. e, Initial rate of amplex red oxidation versus bulk amplex red concentration, [AR], in aerated (red open circles) and N2-purged (blue open triangles) electrolyte. A control experiment without catalyst is also shown (black filled diamonds). In N2-purged electrolyte the initial amplex red rate also exhibits saturation kinetics with increasing [AR], but the saturation level is much lower than that in aerated solution. This lower-saturation oxidation rate may be due to the removal of the possible O2−•-induced amplex red oxidation channel or electron accumulation in TiO2 particles in the absence of electron acceptors in the solution (for example, dissolved O2), which will not occur in photoelectrocatalysis experiments because electrons will be collected by the working electrode. The autocatalytic amplex red conversion to resorufin under ultraviolet illumination does not contribute much to our single-molecule imaging results of photoelectrocatalytic oxidation of amplex red on TiO2 because the process occurs predominantly in the bulk solution rather than on the surface of TiO2 and would not depend on the electrode potential. f, Decomposition of resorufin by TiO2 nanoparticles in aerated (blue line) and N2-purged electrolyte (red line). This process could contribute to the short resorufin residence time on the TiO2 surface (Extended Data Fig. 7). The rate of resorufin decomposition is accelerated in aerated electrolyte, similar to amplex red, which suggests an oxidative pathway for the decomposition of resorufin. The black and grey lines are controls. The reaction volume in each experiment was 2 ml, into which 100 μl of a 1 mg ml−1 aqueous solution of P25 nanoparticles was added.

Extended Data Figure 3 TiO2 nanorods photoelectrocatalytically oxidize amplex red and reduce resazurin to generate resorufin.

ad, vAR scales with E¼ for E ≥ −0.3 V, with I0½ at fixed positive E, and also scales with [AR] at fixed E and I0. a, A series of fluorescence spectra measured from a single TiO2 nanorod on ITO as a function of applied potential from −0.6 V to +0.0 V in N2-purged electrolyte with 50 nM amplex red under 375-nm laser illumination of 60 W cm−2 and 532-nm laser illumination of 630 W cm−2. Each spectrum was acquired for 5 s, and it shows the characteristic fluorescence spectrum of resorufin. b, Single-molecule amplex red oxidation rate (vAR) averaged over the 37 nanorods used in the study. The red line represents a fit to the data with equation (2) in the main text to illustrate the rate dependence on the applied potential (an ensemble-averaged Eon,GB = −0.64 V was obtained here). Only the data more positive than −0.3 V were fitted, for the reasons discussed in Extended Data Fig. 4b caption. The error bars represent s.e.m. The inset plots vAR4 versus E to show the linear behaviour at E more positive than −0.3 V. c, Amplex red oxidation rate per particle versus the square root of incident light power in the presence of 50-nM bulk [AR] and +0.2 V, demonstrating that vAR scales with I0½ at a fixed positive E. The red dots indicate data from individual particles and the black squares represent the average (error bars represent s.d.). d, At +0.2 V and 0.66-mW, 375-nm illumination (the highest light power in c) the reaction rate was observed to increase linearly with bulk [AR]. Two sets of data were included from two different flow cells, where the dots indicate individual particles and the solid squares represent the average (error bars are s.d.). Therefore, under our experimental single-molecule imaging conditions, the reaction rate is in the linear range in its dependence on [AR]. e, Resazurin is dominantly reduced by photogenerated conduction band electrons. Single-particle, single-molecule data of resazurin reduction averaged over a large number of TiO2 nanorods in N2-purged electrolyte contained 50 nM resazurin. The control (black circles) without 0.66-mW, 375-nm laser illumination is also shown. The difference here demonstrates that resazurin is dominantly reduced by photogenerated conduction band electrons, while resazurin reduction under dark is negligible.

Extended Data Figure 4 Photoelectrochemical water oxidation properties of TiO2 nanorods.

i scales with E½ for E ≥ −0.3 V and with I0 at fixed E, and determination of Eon,ss. a, i–E data of ~1,000 nanorods measured with lock-in detection method using 1-Hz chopped 20-mW, 532-nm light illumination (green triangles) and 12-mW, 375-nm laser illumination (blue squares) in the TIRF geometry, demonstrating that anodic photocurrent was only observed under band gap illumination with the 375-nm laser. The solid line is a fit to the data at E ≥ −0.3 V with equation (1) in the main text to yield Eon,GB = −0.65 V. This Eon,GB value agrees with the photocurrent onset potential determined for bulk rutile TiO2, which is also nearly equivalent to the flat band potential determined under dark conditions62,63. The inset shows that i2 scales linearly with E at E ≥ −0.3 V. For E < −0.3 V, the photocurrent deviates from the E½ dependence owing to photocurrent transient dynamics (see b), which have been previously observed and described in detail19,29,64,65,66,67,68. b, Representative photocurrent–time (i–t) responses over a range of potentials during a single on-off cycle at 10 mHz (top) or 1 Hz (bottom) light chopping (50% duty cycle) to mimic steady-state single-molecule imaging and single-nanorod photocurrent measurement conditions, respectively. The background dark current measured from the macroscopic ITO electrode was subtracted using a two-point linear background subtraction method to an average of five consecutive data points before and after illumination. Data was collected with 20-ms time resolution. Three distinct features are clear (denoted on the −0.4-V data; black trace): (1) an initial photocurrent spike (iinitial) when the light is turned on, (2) a decay of the photocurrent on a millisecond-to-second timescale while the light is on towards a steady-state current (iss), and (3) an initial negative, cathodic photocurrent spike (icath) when the illumination is turned off, which decays to the background current level also on a millisecond-to-second timescale. When the photocurrent transient behaviour dominates the i–t response at E < −0.3 V, the photocurrent signal measured by the lock-in amplifier is somewhere between the initial photocurrent spike (iinitial) and steady-state photocurrent (iss), which is expected because the lock-in detection is dependent on the signal shape versus time. Following references 19 and 69, we fit i–E data for E ≥ −0.3 V in a. c, Photocurrent versus 375-nm laser power I0 at +0.2 V for three different illumination conditions, demonstrating that at fixed positive potentials, i scales linearly with I0 over a broad range of power densities (milliwatts to megawatts per centimetre squared). The geometric spot size of the laser spot was (1) 1 × 1 cm2 for ensemble-averaged measurements (open green squares, spanning a power density range of 0.8–6.8 mW cm−2), (2) 80 × 95 μm2 to excite ~1,000 nanorods in the TIRF geometry (solid blue triangles, spanning a power density range of 32.9–151.3 W cm−2), and (3) 390 × 390 nm2 for single-nanorod measurements (solid red circles, spanning a power density range of 0.1–6.6 MW cm−2). d, Steady-state photocurrent issE data (black circles, same data as included in Fig. 1n) obtained from 10-mHz chopped-light illumination experiments. The steady-state photocurrent onset potential (Eon,ss) is defined as the intersection point of zero photocurrent and the tangent at maximum slope of photocurrent57. The solid black line indicates the linear fit to determine Eon,ss = −0.47 V. This Eon,ss value is ~200 mV more positive than the Eon,GB value (−0.65 V) in a; this difference has been previously observed and attributed to surface charging19,70. Despite the difference in their absolute values, the relative values and variations of Eon,GB among individual nanorod spots are good reflections of the relative values and variations of their Eon,ss, as Eon,GB is linearly correlated with Eon,ss (see Supplementary Fig. 20b in Supplementary Note 5.4).

Extended Data Figure 5 Product localization accuracy, hotspot size, as well as position correlation of hole- and electron-induced reaction hotspots.

a, Localization precision of product molecule positions. Distribution of the error in x and y (that is, Errx or Erry) for all single molecules detected on the surface of a single nanorod, calculated using Supplementary equation (3). The average errors were determined with a one-dimensional Gaussian fit (solid lines). bd, Size of localized reaction hotspots, and distribution of distances between corresponding high-reaction-rate electron and hole hotspots. b, Two-dimensional histogram (15 × 15 nm2 bins) of all hole-induced reaction product positions over the potential range E = −0.6 V to +0.2 V of a single nanorod exhibiting a localized, hole activity hotspot. The red rectangular area represents the region where a one-dimensional histogram was plotted in c. c, One-dimensional histogram from the region in b. The solid line is a Gaussian fit. The full-width at half-maximum (FWHM), which is used as a measure of this hotspot size, is 52 nm, much smaller than the diffraction-limited resolution of ~300 nm. d, Distributions of distances between corresponding hole and electron reaction hotspots on individual nanorods. Each hotspot’s centre position was determined by two-dimensional Gaussian fitting to two-dimensional reaction product histograms. The average distance between localized hole and electron hot-spots is ~42 nm, and the minimum distance is ~13 nm.

Extended Data Figure 6 Elemental analysis and high-resolution electron microscopy for identifying possible active sites on these TiO2 nanorods.

See Methods for more details. a, Atomic concentration of impurity elements relative to Ti measured via ICP-AES in commercial P25 nanoparticles (black bars, precursor used in the synthesis) and the as-synthesized TiO2 nanorod sample (red bars). The dried powder samples were digested in 1 ml each of hot, concentrated HNO3 and HCl and submitted for elemental analysis at the Cornell Nutrient Analysis Laboratory. The total amount of TiO2 dissolved in the acid digestion process was difficult to quantify, so the concentration of impurity elements in each sample were calculated relative to the total amount of Ti detected. The ICP-AES data suggests that substantial concentration increases in elements such as Al, B, Ca, Fe, K and Mg, as well as the introduction of Zn, are due to impurity incorporation during the high-temperature synthetic method, probably from reagent impurities (for example, in reagent-grade NaCl and Na2HPO4). This hypothesis is further supported by ref. 14, which demonstrated that eight different transition metals (including Fe) could be substitutionally doped for Ti at an atomic concentration of 2%. b, High-angle annular dark field scanning TEM (HAADF-STEM) image of a single nanorod. The red box indicates a region of the image where energy-dispersive X-ray analysis was performed. cf, Energy-dispersive X-ray analysis mapping of Ti and other detected impurity elements revealed a heterogeneous distribution of impurities: Fe (d), Zn (e) and Mn (f). g, High-resolution bright field TEM image of a single nanorod shows structural irregularities along the side surfaces. h, Zoomed-in view from the dashed red box in g shows lattice fringes, evidence that each rod is a single crystal. Similar lattice fringes were also observed by ref. 14.

Extended Data Figure 7 The two structurally similar probe molecules amplex red and resazurin adsorb to all surface sites equivalently.

See Methods for more details. a, b, d, e, g, h, Scatter plots of all individual resorufin molecules generated from (a, d and g) hole-induced amplex red oxidation (orange dots) and (b, e and h) electron-induced resazurin reduction reactions (blue dots) on the same three TiO2 nanorods shown in Fig. 1. The red lines dissect each nanorod into 150-nm segments along its length, within each of which the average residence time is calculated. c, f and i, Average product residence time versus position along the nanorod for hole-induced amplex red oxidation (orange circles) and electron-induced resazurin reduction (blue circles) reactions. j, Average product residence time formed from hole-induced amplex red oxidation (red circles) and electron-induced resazurin reduction (blue diamonds) reactions plotted versus the corresponding kh and ke values for all 78 nanorod spots studied herein. The error bars represent s.d.

Extended Data Figure 8 Photoelectrochemical properties of bare and (Co-Bi) modified TiO2 nanorods.

ac, Ensemble-level i–E and iI0 data of bare and (Co-Bi)-modified TiO2 nanorod thin films. a, Representative optical transmission image of a thin film of TiO2 nanorods on an ITO electrode. b, i2 is linearly proportional to E for the same bare (black circles) and OEC-modified thin film sample as in a at E ≥ −0.3 V. Data were obtained from chopped light (138 mW cm−2, 365 nm LED) linear sweep voltammograms (−10 mV s−1 from +0.3 V, see inset) for a bare ITO electrode (red trace), TiO2 nanorod-coated thin film electrode (black trace) and (Co-Bi)-TiO2 electrode in N2-purged electrolyte (blue trace). The data have been slightly offset for clarity. Co-Bi OEC was photoelectrochemically deposited from electrolyte containing 0.5 mM CoCl2 at 0.0 V for 30 min using 365-nm LED illumination at 45 mW cm−2. c, Photocurrent is linearly proportional to 365-nm light power at +0.3 V for the same TiO2 (red circles) and (Co-Bi)-TiO2 (blue circles) thin film sample in N2-purged electrolyte. d, Single-nanorod averaged i2 data are linearly proportional to E for bare (black circles) and Co-Bi OEC modified (red circles) TiO2 nanorods at E ≥ −0.3 V. The data represents the average photocurrent of 78 nanorod spots from 37 individual nanorods and the error bars represent the standard error of the mean. The inset shows i–E response obtained by lock-in detection under 1-Hz illumination at the highest power density (6.6 mW cm−2). The η and Eon,GB obtained from fitting the data (lines) with equation (1) in the main text are η = (5.72 ± 0.11) × 10−2% and Eon,GB = −0.66 ± 0.02 V for TiO2; and η = (9.62 ± 0.17) × 10−2% and Eon,GB = −0.72 ± 0.02 V for (Co-Bi)-TiO2. Together, these data demonstrate that the iE½ (at E ≥ −0.3 V) and i–I0 scaling laws hold for both bare and OEC-modified nanorods, and in all cases, the photocurrent for the OEC-modified TiO2 nanorod-coated electrode is higher than the bare TiO2 nanorod-coated electrode. The OEC-induced photocurrent enhancement is also not due to larger photocurrent transient dynamics (Supplementary Note 5.7).

Extended Data Figure 9 Developing a Co-Bi OEC deposition method on single nanorods that produces a saturated photocurrent enhancement.

a, Representative optical transmission image of a nanorod sample after localized photoelectrochemical deposition of Co-Bi catalysts after 20 s deposition time. The red arrows indicate the location on which the focused laser was positioned for guided catalyst deposition, where the deposited catalysts appear as dark spots. b, Photocurrent at +0.2 V versus photoelectrochemical deposition time of the catalyst for single nanorods (red circles) and nanorod-averaged results (black squares). This photocurrent shows a saturation behaviour versus OEC deposition time (and hence OEC deposition amount), which was also observed for planar Fe2O3 thin films modified with a Co-phosphate catalyst4,26. Error bars represent s.d. In all of our single-nanorod OEC deposition studies presented in the main text, we used an OEC deposition time of 5 s at which, according to b, the photocurrent enhancement is already saturated. Thus, the η parameter obtained from fitting the i–E data of OEC-modified nanorod spots represents the maximum achievable η per spot, as well as the maximum achievable change in η (that is, Δη) between the same bare and OEC-modified nanorod spot. c, Catalyst coverage versus kh (see Supplementary Fig. 2 for details on catalyst coverage quantification), demonstrating that more catalyst material is deposited onto higher-activity sites. The red circles represent data from individual nanorods and the black squares represent averaged data grouped by ten data points sorted by kh. The same panel is also presented as Fig. 3c in the main text. d, η for OEC-modified nanorod spots versus catalyst coverage. η is essentially independent of OEC coverage, consistent with the saturated photocurrent behaviour in b. Filled squares represent data grouped by ten data points sorted by their catalyst coverage values. The error bars for all individual spot values represent s.d. The error bars in c and d for all binned data represent s.e.m.

Extended Data Figure 10 Proposed block–deposit–remove strategy, which has the potential to selectively deposit a desired OEC on low-activity and positive onset potential sites.

First, a removable protective group can be photoelectrochemically deposited starting from negative potentials to block negative Eon,GB sites. To block high-activity sites with similar Eon,GB values, we propose that using short light pulses may cause more protecting group material to be deposited at higher-activity sites first. In this way, low η and positive Eon,GB sites will remain unblocked. The potential could then be stepped from negative E to the ensemble-averaged steady-state photocurrent onset value, for example, and thus all remaining sites with Eon,GB values more positive than the ensemble-averaged value will also remain unblocked. Second, the potential can be set to a positive value and a long light pulse can be used to deposit the desired OEC on all remaining sites: those with positive Eon,GB and low η. Finally, the protective group can be subsequently removed to yield a photoanode with optimally located OECs. Yellow dots represent the OEC.

Supplementary information

Supplementary Information

This file contains Supplementary Text, Supplementary Figures 1-25, Supplementary Table 1 and Supplementary References - see contents page for details. (PDF 7926 kb)

PowerPoint slides

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sambur, J., Chen, TY., Choudhary, E. et al. Sub-particle reaction and photocurrent mapping to optimize catalyst-modified photoanodes. Nature 530, 77–80 (2016). https://doi.org/10.1038/nature16534

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nature16534

This article is cited by

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing