One application which is frequently mentioned for nanometer-sized semiconductor crystals (quantum dots, QDs) is as transducers for the conversion between optical and electrical signals in so-called 'opto-electronic' devices. Of course, optical to electrical transduction requires that the absorption of light by QDs be detectable as an electrical signal. To our knowledge, there is just one published report of the electrical detection of optical processes in semiconductor QDs: Alivisatos and coworkers have incorporated CdSe QDs into polymer thin film photovoltaic devices. We have recently demonstrated the feasibility of electrically detecting the absorption of light by a submonolayer coverage of cadmium sulfide QDs supported on a graphite surface. A schematic diagram of the device used for this pupose is shown below:

These devices were constructed as follows: CdS nanocrystals (90 Å in radius) were synthesized on HOPG(0001) using the Electrochemical/Chemical Method as previously described. Then, a 10-100 nm thick layer of the insulating polymer, Formvar, was deposited by first dissolving the Formvar in chloroform, then floating a few drops of this solution onto the surface of Nanopure water in a beaker, and finally pulling the immersed graphite surface, on which CdS QDs had been deposited, through the floating polymer film. Following the evaporation of solvent from this film, a Scotch tape mask with a 5 mm diameter aperture was applied to the sample as a contact pad, then a thin semitransparent (10-15 nm) gold top electrode was evaporated onto the surface of the Scotch tape and the exposed Formvar film.

The Figure above shows room temperature transient photocurrent ('device') spectra, and low-temperature PL spectra, for a device based on CdS nanocrystals having a radius of 8 nm (estimated from the energy of maximum emission). These CdS particles were prepared by depositing Cd metal using a train of ten 20 ms voltage pulses - a tactic which we have shown improves the size monodispersity of CdS QDs. The resulting PL spectrum shows a slight blueshift of the emission maximum to 493 nm, corresponding to a particle radius of about 8 nm. This emission blueshift was found in about 60 % of all areas sampled with the fluorescence microprobe and can thus be expected to appear in the transient photocurrent spectrum. This is indeed the case: The absorption edge is blueshifted to 501 nm proving that the photocurrent measurement is sensitive to the nanoparticle band gap even at room temperature. In contrast, PL spectra are virtually impossible to obtain under these conditions of sample temperature and particle areal density (typically 10¹⁰ cm^-2). Control devices, with the construction shown above but without QDs (i.e., using clean graphite surfaces) exhibited no photocurrent peaks above background.