Making every photon count

Article metrics

More than 150 years after the invention of photography, the underlying solid-state photochemistry is still being investigated and improved. One aspect of continual importance is the efficiency of the image-recording stage, which strongly affects the sensitivity of the photographic emulsion. Photons falling on a silver halide microcrystal in the emulsion cause a halide ion (such as chloride or bromide) to lose an electron, which is then captured by a silver ion, converting it into a neutral silver atom. The build-up of silver atoms gives rise to silver clusters: the ‘latent image’, which can be developed after several more steps into a print1. In reality, the halide ions generate pairs of electrons and holes (the positively charged counterparts of electrons), and it is generally accepted that recombination of photogenerated electrons and holes is a major loss process2,3. On page 865 of this issue, Belloni et al.4 present an ingenious way of overcoming this recombination process by doping the silver halide with formate ions, which leads to an increase in the number of silver atoms generated per photon absorbed.

Image-forming reactions in silver halide microcrystals compete with various electron-loss processes, the dominant one being recombination. Overall, many of the photogenerated electrons are lost and the sensitivity of the silver halide emulsion is low. The sensitivity can be greatly improved by chemical sensitization, which is how all modern films and print materials are prepared. During chemical sensitization, reagents containing labile sulphur and gold atoms are used at low concentrations (parts per million or lower) to produce silver–gold sulphide clusters on the surface of the microcrystal. These clusters appear to deepen the existing traps for electrons, thereby reducing recombination5,6,7.

During film development the silver halide microcrystal is converted to metallic silver through an electrochemical reduction reaction catalysed by the latent image. A microcrystal can only be developed when the latent image contains three or more photogenerated silver atoms8. In the theoretical limit, one photon should produce one silver atom. So the most sensitive film imaginable will only require three photons per microcrystal to create a latent image. Despite the advantages of sulphur-plus-gold sensitization there is still some recombination, and three to ten times as many photons are needed9,10. So reaching this theoretical limit requires an additional mechanism to eliminate the remaining recombination. In the past this has been done by reduction sensitization, which chemically creates silver clusters on the surface of the microcrystal. These silver clusters have different properties from the photogenerated silver clusters in that they trap and destroy holes through an oxidation process at the silver cluster11. In model systems (but not in commercial films) reduction sensitization, in addition to the sulphur-plus-gold sensitization, allows one to reach the theoretical limit9,10.

Unfortunately a significant level of fog is often a byproduct of the combined sensitization processes. Fog is the result of latent-image formation in the absence of light exposure and, when present, reduces the dynamic range of the material for image capture. In addition, the stability of chemical reduction is often low, so it is not usually a viable option for photographic manufacturers. Placing the chemically produced silver clusters inside the microcrystal could eliminate the fog problem and presumably would stabilize the sensitization. But this may change the electronic properties of the silver clusters, because some experiments indicate that they have electron-trapping properties that would actually decrease the sensitivity12.

Efficiency improvements are important to photographic manufacturers because efficiency is a factor that determines the film speed. Light absorption also determines film speed and this aspect is controlled by the size of the microcrystal. This is the usual way in which different film speeds are obtained. Unfortunately, the noise in the final image also increases with microcrystal size13. So, higher and higher film speeds can be reached, but the image tends towards lower quality, particularly when enlargements are made. Increasing the efficiency of latent-image formation is a much better, although more difficult, way to achieve higher film speed.

The work of Belloni et al.4 shows one way to do this without incurring problems related to reduction sensitization. The use of formate (HCO21) as a dopant is a novel way to overcome the residual recombination in conventionally sensitized microcrystals. Not only does it appear to be efficient at destroying holes, but the resulting unstable CO21 radical is able to provide an extra electron for latent-image formation. Now the theoretical limit becomes two silver atoms per photon. Similar results have been observed with reduction sensitization, but never without some risk of fogging10,11,14,15.

Silver bromide emulsions, such as the one studied by Belloni et al., only absorb blue light. This is because quantum theory dictates that discrete (rather than continuous) energy levels are formed in the crystal lattice, creating a permitted ‘conduction band’ for free electrons with the right energy. To capture and display the full colour of the original scene it is necessary to use spectral sensitization in addition to chemical sensitization. In spectral sensitization, dyes are adsorbed on the microcrystal surface and provide absorption in the region of the spectrum where the microcrystal is unable to respond16. The absorption event places an electron in a higher energy state, from where it can be transferred to the conduction band of the microcrystal and then participate in the usual process of latent-image formation (Fig. 1). Unfortunately the dye also introduces an energy level for trapping holes, making it more difficult for the formate ions to destroy them.

Figure 1: Colour photography requires three layers of silver halide emulsion sensitive to blue, red and green wavelengths of light.

The effect of adding a green- or red-absorbing dye to AgBr (which absorbs blue light) is shown here. Absorption involves movement of an electron from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO), leaving a hole in the HOMO level. The photoinduced electron transfer occurs between the LUMO and the AgBr conduction band, leaving behind an oxidized-dye radical. The LUMO levels for both the green-absorbing dye and the red-absorbing dye are positioned at the conduction-band minimum so that efficient electron transfer can occur. Thermal energy can be used to move the hole into the valence band, thereby preventing recombination with an electron. Belloni et al.4 show that formate doping increases the sensitivity of green-dyed AgBr. Improving the overall efficiency of colour films might be possible, if similar increases in sensitivity can be shown for red-dyed emulsions.

If formate doping is to be effective in increasing the sensitivity of dyed microcrystals it is necessary for the photogenerated hole to transfer to the valence band of the silver bromide crystal. Otherwise holes trapped by the dye act as recombination centres. Belloni et al. show that the hole will move into the valence band for a dye absorbing in the mid-green region, although the sensitivity increase is less than what they see with undyed microcrystals. One might suspect that this lower degree of efficiency is due to hole trapping by the dye so that holes are less likely to be destroyed by the formate dopant. Furthermore, one might expect the degree of sensitivity improvement to be even less for dyes absorbing in the red region, where holes are even more deeply trapped (Fig. 1). Studies with red dyes will be important for assessing the general value of formate doping as a way to remove residual inefficiencies in the photographic process and achieve the highest sensitivity possible.


  1. 1

    Tani, T. Photographic Sensitivity Ch. 4 (Oxford Univ. Press, 1995).

  2. 2

    Hamilton, J. F. in The Physics of Latent Image Formation in Silver Halides (eds Baldereschi, A., Czaja, W., Tosati, E. & Tosi, M.) 218 (World Scientific, Singapore, 1984).

  3. 3

    DiFrancesco, A. G., Tyne, M., Pryor, C. & Hailstone, R. K. J. Imaging Sci. Technol. 40, 576–581 (1996).

  4. 4

    Belloni, J., Treguer, M., Remita, H. & De Keyzer, R. Nature 402, 865–867 (1999).

  5. 5

    Hamilton, J. F., Harbison, J. M. & Jeanmaire, D. L. J. Imaging Sci. 32, 17–19 (1988).

  6. 6

    Zhang, D. & Hailstone, R. K. J. Imaging Sci. Technol. 37, 61–72 (1993).

  7. 7

    Morimura, K. & Mifune, H. J. Soc. Photogr. Sci. Technol. Japan 61, 175–180 (1998).

  8. 8

    Hailstone, R. K. & Hamilton, J. F. J. Imaging Sci. 29, 125–131 (1985).

  9. 9

    Babcock, T. A. & James, T. H. J. Photogr. Sci. 24, 19–23 (1976).

  10. 10

    Hailstone, R. K. et al. J. Imaging Sci. 32, 113–124 (1988).

  11. 11

    Hailstone, R. K., Liebert, N. B., Levy, M. & Hamilton, J. F. J. Imaging Sci. 35, 219–230 (1991).

  12. 12

    DiFrancesco, A. G., Tyne, M. & Hailstone, R. K. in Proc. Soc. Imaging Sci. Technol. 49th Ann. Conf. 222–223 (Soc. Imaging Sci. Technol., Springfield, Virginia, 1996).

  13. 13

    Dainty, J. C. & Shaw, R. Image Science 96 (Academic, New York, 1974).

  14. 14

    Tani, T. J. Imaging Sci. 29, 93–99 (1985).

  15. 15

    Tani, T. J. Imaging Sci. 30, 41–46 (1986).

  16. 16

    Tani, T. Photographic Sensitivity Ch. 5 (Oxford Univ. Press 1995).

Download references

Author information

Correspondence to Richard Hailstone.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hailstone, R. Making every photon count. Nature 402, 856–857 (1999) doi:10.1038/47198

Download citation

Further reading


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.