Outline. Photosensors in biology and in semiconductors. The visual pathway Bottom view. The visual pathway Side view

Transcription

1 Outline Photosensors in biology and in semiconductors CNS WS07-08 Class 1 Photosensors in biology The visual pathway and the retina Photoreceptors and the fovea Giacomo Indiveri Institute of Neuroinformatics University ETH Zurich Zurich, December Light, waves, and photons Frequency spectrum Photons 3 Photosensors in semiconductors The photoelectric effect Photodetectors The visual pathway Side view The visual pathway Bottom view

2 Human eye Cross section Human retina Cross section The retina The Outer Plexiform Layer Its the most remote and accessible part of the brain. Its a thin complex sensory tissue composed of six layers of cells. The ganglion cells (the output neurons of the retina) lie innermost in the retina closest to the lens and front of the eye, and the photosensors (the rods and cones) lie outermost in the retina The first area of neuropil is the outer plexiform layer (OPL) where connections between rod and cones, and vertically running bipolar cells and horizontally oriented horizontal cells occur.

3 The Inner Plexiform Layer Photoreceptors The second neuropil of the retina, is the inner plexiform layer (IPL), and it functions as a relay station for the bipolar cells to connect to ganglion cells. Different types of amacrine cells, interact to influence and integrate the ganglion cell signals. Photoreceptor cells in the rods and cones convert light first to chemical energy and then electrical energy. Rods function in dim light (night vision), and do not detect color. There are about 126 million rods in each eye. Cones have different spectral sensitivities (color vision). There are about 6 million cones in each eye. The Fovea Cones and Color Vision The center of the fovea is known as the foveal pit and is a highly specialized region of the retina. The foveal pit is an area where cone photoreceptors are concentrated at maximum density with exclusion of the rods, and arranged at their most efficient packing density which is in a hexagonal mosaic. Cones contain cone opsins as their visual pigments and, depending on the exact structure of the opsin molecule, are maximally sensitive to either long wavelengths of light (red light), medium wavelengths of light (green light) or short wavelengths of light (blue light).

4 Rods and Night Vision Light: waves or particles Rods are highly sensitive photoreceptors that contain the visual pigment - rhodopsin. They are sensitive to blue-green light with a peak sensitivity around 500 nm wavelength of light and are used for vision under dark-dim conditions at night. Rods are so sensitive that they can reliably signal the absorption of single photons of light. Light is electromagnetic radiation that has been typically described in terms of waves. In 1905 Einstein proposed that light propagated only in discrete irreducible packets or quanta (photons). This explained the non-classical features of the photoelectric effect. Photoreceptors catch photons, not waves. (R. W. Rodiek) Frequency scales Electromagnetic radiation comes in many flavours. The power company sends us electrons that oscillate at a rate of 50 times per second Radio stations send us photons that oscillate between 550 to 1600 khz in the AM spectral band, and from 88 to 108 MHz in the FM band. The sun sends us photons that lie mostly in the infrared region of the spectrum. But a quarter of them lie in the visible spectral range, from 430 to 750 THz. Photons A photon is created when it is emitted by an electron A photon moves at the speed of light. Photons are characterized by their oscillating frequency, and by their spin, or state of polarization. A photon s energy is directly proportional to its frequency. Each photon keeps the same energy from the time that it is emitted by an electron to the time it is absorbed (for all practical purpouses).

5 How rods and cones convert photons into current The photoelectric effect Whenever a photon is absorbed by a rod or cone visual pigment, the corresponding cromophore changes shape (photoisomerization). The photoisomerization of the cromophore triggers a biochemical cascade that causes a hyperpolarization of the cell s membrane potential and a decrease of glutamate release. The bipolar cells respond to the decrease in the concentration of the glutamate molecules. Rods and cones decrease their dark photocurrent in response to brief flashes of light. A process by which light releases electrons from metal surfaces, first described by Heinrich Hertz in Photons are absorbed by electrons. If energetic enough, the electron escapes from the material with a finite kinetic energy. One photon can only remove one electron. The electrons that are emitted are called photoelectrons. The photoelectric effect in silicon The photoelectric effect in silicon E ph = hν = hc λ h: Plank constant ν: Light frequency c: Speed of light λ : Light wavelength Intrinsic Interband transitions Generation of electron-hole pairs λ max = hc at room temperature E g (with 1eV = J,h = Js,c = m/s). Extrinsic = 1.24µm E g Transitions involve impurity energy levels in bandgap Generation of only one carrier type λ max depends on position of energy levels

6 Optical Absorbption Quantum efficiency The number of absorbed photons at any given position in a semiconductor is given by αφ(x) x. Φ 0 : photon flux. dφ(x) = αφ(x) dx Φ(x) = Φ 0 e αx Φ 0 = 1 R A λ hc P opt with R=reflection A=Cross section area P opt =Incident optical power α=optical absorbption coefficient. Quantum efficiency (%) Responsivity (A/W) Si Ge InGaAsP Wavelength (μm) Quantum efficiency η: the number of electron-hole pairs contributing to the photocurrent, generated per incident photon. η hc qλ I ph P opt with I ph =photocurrent q=number of carriers in current. Photoconductor Steady-state Photoconductor Externally applied electric field In the presence of an electric field electrons and holes can separate and contribute to a photocurrent No. of photons arriving at surface per unit time is: P opt hν Generation rate of carriers: g = η P opt hν Recombination rate of carriers: r = n τ, where τ is the carrier lifetime. In steady-state generation is balanced by recombination: r = g n τ = η P opt hν n = τη P opt hν I p = ( ) qµn E WD ( = q η P ) ( ) opt µτ E hν L τ = I ph t r with µ=mobility L=length t r =carrier transit time (t r = L/µE).

7 Photodiodes Photodiode Photoconductors typically exhibit large dark currents (background current also present in the absence of optical stimulation), due to the large doping concentration (high conductivity) used in most modern semiconductor processes. p region Photon Depletion region n region A diode is a much more suitable photosensor: Depletion region with low conductivity Built-in electric field Current-Voltage characteristics I I ph V I Forward bias region. I = I 0 (e V U T 1) II Unused. III Reverse bias region. I independent of V, but proportional to photon flux. Hole diffusion Drift Electron diffusion IV Photovoltaic regime. A reverse current flows in the presence of a forward voltage: I P opt. I p = I ph = qη P opt hν reverse current Phototransistors Bipolar junction transistor with floating base Phototransistors Common-Emitter Current Gain 100 Current gain 10 1 h fe h FE I BE = I ph due to photogenerated holes I CB = I ph due to photogenerated electrons I C = I CB + I CE + I CE I BE I BE = I ph (1 + h FE ) h FE = I CE I BE common-emitter current gain I c (A) h fe : Small-signal common-emitter current gain h FE : Static common-emitter current gain.

8 Photodiode vs. phototransistor Photogate I p = I ph P opt Linear No gain Fast response Small size Low noise I p = I ph (1 + h FE ) Nonlinear: h FE (I CE ) Gain: h FE 10 2 Slower response Larger size Noise amplified Its a Metal-Insulator Semiconductor (MIS) structure where one type of photogenerated carriers is collected in a deplition region, underneath the conductor plate, and the other is collected by the substrate. It is the basis of Charged-Coupled Devices (CCD) imagers. V g inversion depletion V g n diffusion depletion p - substrate p - substrate E E Logarithmic photosensors Source-follower configuration Sources E M 1 I ph D 1 V b V out Typical irradiances under ambient lighting conditions elicit photocurrents in the pa or na range in photodiodes with areas of the order of (10 µm) 2 typical MOSFETs operate in the subthreshold domain: V out = κv g U T log dv out = U T di ph I ph ( Iph I 0 ) Slides from Tobi s lectures Pictures from Hubel & Wiesel, 1987 and from Biological photoreceptors material from The first steps in Seeing, by W.E. Rodiek VLSI material from class textbook Analog VLSI: circuits and principles