Last time, we explained photoelectric conversion, which is the basis for understanding how CMOS image sensors work.
*Click here for the previous article → About photoelectric change, the entrance through which CMOS image sensors capture light


We explained that CMOS image sensors use the built-in photodetector to convert light into an electric charge using the internal photoelectric effect, and that the wavelength of light that can be captured is determined by the band gap of the semiconductor. For example, the band gap of silicon (Si) is 1.1 eV, and it can capture light with a wavelength of 1100 nm or less.

So, can any light with a wavelength of 1100 nm or less be converted into an electric charge? You may be wondering, but in reality this is not the case. CMOS image sensors with Si light receiving elements capture light with wavelengths between approximately 300 nm and 1100 nm, and it becomes difficult to capture light with wavelengths shorter than 300 nm. The reason for this is that substances have a fixed absorption coefficient. This time, we will explain the absorption coefficient.

The absorption coefficient is a constant that indicates how much light, electromagnetic waves, radiation, sound waves, etc. are absorbed per unit length when they travel through any material, including semiconductors. For the silicon used in the light receiving elements of CMOS image sensors, the characteristics are as follows, with the horizontal axis representing wavelength (nm) and the vertical axis representing absorption coefficient α (cm^-1).

This absorption coefficient is determined by the material; for example, the scintillators used in X-ray inspection equipment basically work on the same principle.

When light is incident on a material containing a semiconductor, the light intensity I at a depth x (cm) from the material surface can be expressed by the following equation.

Here, ​_I0 is the light intensity at the material surface x=0 (cm), and α is the absorption coefficient.

If we apply the absorption coefficient of Si discussed earlier to this formula and represent the light intensity (%) for each wavelength on the vertical axis and the depth to which the light penetrates (um) on the horizontal axis, we get the following characteristics.

This characteristic tells us that the shorter the wavelength of light, the more limited it is that it can reach the surface of a material, and the longer the wavelength of light, the further into the material it can reach.

The reason why a CMOS image sensor whose photodetector is made of silicon cannot capture a good image when exposed to blue light is that blue light, which has a short wavelength, does not penetrate deep into the silicon and sufficient photoelectric conversion does not take place.

In this way, the light-receiving elements of CMOS image sensors are designed to take advantage of the photoelectric conversion and absorption coefficient characteristics explained last time.