Measurement of Photon Statistics with Live Photoreceptor Cells

We analyzed the electrophysiological response of an isolated rod photoreceptor of Xenopus laevis under stimulation by coherent and pseudothermal light sources. Using the suction-electrode technique for single cell recordings and a fiber optics setup for light delivery allowed measurements of the major statistical characteristics of the rod response. The results indicate differences in average responses of rod cells to coherent and pseudothermal light of the same intensity and also differences in signal-to-noise ratios and second-order intensity correlation functions. These findings should be relevant for interdisciplinary studies seeking applications of quantum optics in biology.

Figure 1

(a) Optical layout. A laser beam is chopped by a shutter, attenuated by a neutral density filter (NDF), and split by a $50/50$ beam splitter (BS). Both of the beams are coupled into single-mode optical fibers (SMF) with lenses (L). One of the fibers is connected to an avalanche photodiode (APD) while the other stimulates the rod. The inset shows the pseudothermal light source: L is a lens, GD is a rotating ground glass disc, D is a diaphragm. (b) Microscope image (top view) of the fiber taper and a suction micropipette with the constrained rod cell.

Figure 2

Dependence of the normalized average amplitude of the rod photocurrent on the average number of impinging photons for coherent (red squares, solid line) and pseudothermal (black circles, dashed line) sources. Each data point is an average response to 80–100 light pulses. Six different cells, taken from 6 different animals were used. Lines show numerical calculations, and standard deviation is shown by error bars. The inset shows raw waveforms of the rod photocurrent in response to coherent pulses of different average intensity.

Figure 3

Probability histograms of normalized amplitudes measured for coherent (a),(b) and pseudothermal (c),(d) sources at different values of the average, shown with the corresponding SNR values. Each histogram was acquired for 80–100 light pulses. Standard deviation is shown by error bars.

Figure 4

Dependence of the SNR (a) and $g_0^{(2)}$ (b) on the average number of impinging photons for coherent (red squares, solid line) and pseudothermal source (black circles, dashed line). Results are averaged similarly to Fig. 2. Standard deviation is shown by error bars, which cannot be seen for the coherent source in (b). Lines show numerical calculations.