The meaning of PIN (Post-Intrinsic-Negative) is that a layer of semiconductor material with a very low doping concentration (such as Si) is inserted between P-type and N-type semiconductor materials. This layer is denoted as I (Intrinsic) and is called the intrinsic region. The structure of a PIN photodiode (PIN-PD) is shown in the left figure. In the figure, after incident light enters from the P* region, it is absorbed not only in the depletion region but also outside the depletion region. These absorptions form the diffusion component in the photocurrent. For example, electrons in the P* region first diffuse to the left boundary of the depletion region and then pass through the depletion region to reach the N* region. Similarly, holes in the N' region diffuse to the right boundary of the depletion region before passing through the depletion region to reach the P* region. The photocurrent in the depletion region is called the drift component, and its propagation time mainly depends on the width of the depletion region. Obviously, the propagation time of the diffusion current component is longer than that of the drift current component. As a result, the trailing edge of the output current pulse of the photodetector is lengthened, and the resulting time delay will affect the response speed of the photodetector.
If the depletion region is narrow, most photons will reach the N+ region before being absorbed by the depletion region. In this region, the electric field is very weak and cannot separate electrons and holes, resulting in a relatively low quantum efficiency.
A narrower depletion region width *w* results in a larger junction capacitance and a larger RC time constant, which is detrimental to high-speed data transmission.
Considering drift time and junction capacitance effects, the bandwidth of a photodiode can be expressed as：
In the formula, R1 is the load resistance.
The above analysis demonstrates that increasing the width of the depletion region is essential.
As shown in the figure above, the width of the I-region is much larger than that of the P+ and N+ regions. Therefore, more photons are absorbed in the I-region, increasing quantum efficiency while maintaining a small diffusion current. The reverse bias voltage of the PIN photodiode can be set to a smaller value because the thickness of its depletion region is essentially determined by the width of the I-region.
Of course, a wider I-region isn't always better. A larger width (w) results in a longer drift time for carriers in the depletion region, thus limiting the bandwidth. Therefore, a comprehensive consideration is necessary. Since different semiconductor materials have different absorption coefficients for different wavelengths of light, the width of the intrinsic region (I-region) varies. For example, the I-region width of a Si PIN photodiode is approximately 40 mm, while that of an InGaAs PIN photodiode is approximately 4 mm. This determines the different bandwidths and wavelength ranges of photodetectors made from these two different materials: Si PIN photodiodes are used in the 850 nm band, while InGaAs PIN photodiodes are used in the 1310 nm and 1550 nm bands.
(APD)Avalanche photodiode
An APD (Avalanche Photodiode) is a highly sensitive photodetector that utilizes the avalanche effect to multiply the photocurrent. The principle of the avalanche effect is as follows: An incident signal light generates initial electron-hole pairs in the APD. Due to the high reverse bias voltage applied to the APD, these electron-hole pairs accelerate under the influence of the electric field, gaining significant kinetic energy. When they collide with neutral atoms, electrons in the valence band of the neutral atoms gain energy and jump to the conduction band, thus generating new electron-hole pairs, called secondary electron-hole pairs. These secondary carriers can also collide with other neutral atoms under a strong electric field, generating new electron-hole pairs, thus inducing the avalanche process that produces new carriers. In other words, one photon ultimately generates many carriers, amplifying the optical signal within the APD. Structurally, the difference between an APD and a PIN photodiode lies in the addition of an additional P layer. The structure of an APD is shown in Figure 3-18. When reverse biased, a strong electric field exists in the PN junction sandwiched between the I layer and the N* layer. Once the incident signal light enters the I region from the left P* region, it is absorbed in the I region to generate electron-hole pairs. The electrons in the I region quickly drift to the PN junction region, and the strong electric field in the PN junction causes the electrons to produce an avalanche effect.
Structurally, the difference between an APD and a PIN photodiode lies in the addition of an extra layer, P. The structure of an APD is shown in the right figure. Under reverse bias, a strong electric field exists in the PN junction sandwiched between the I and N+ layers. Once the incident signal light enters the I region from the left P+ region, it is absorbed in the I region, generating electron-hole pairs. The electrons rapidly drift to the PN junction region, and the strong electric field in the PN junction causes an avalanche effect.
Compared to PIN photodiodes, the photocurrent is amplified internally by the APD, thus avoiding noise introduced by external circuitry. From a statistical average perspective, assuming one photon generates M carriers, this is equal to the ratio of the photocurrent I output after APD avalanche to the initial photocurrent I before multiplication.
In the formula, M is called the multiplication factor.
The multiplication factor is related to the ionization rate of the charge carriers, which refers to the average number of electron-hole pairs generated per unit distance of drift. The electron ionization rate and the hole ionization rate are different, denoted by α₀ and α₂, respectively. They are related to factors such as reverse bias voltage, depletion region width, and doping concentration, and are denoted as α₀.
In the formula, k is the ionization coefficient, which is a measure of the performance of a photodetector.
The effect of ionization rate on M can be given by the following formula:
When α = 0, only electrons participate in the avalanche process, M = e^(-ω), and the gain increases exponentially with ω. When αω = 1 and -1, according to equation (3-26), M → ∞, and avalanche breakdown occurs. Typically, the value of M ranges from 10 to 500. Avalanche breakdown in an APD occurs because the applied reverse bias voltage is too large. Considering the close relationship between M and the reverse bias voltage, an empirical formula is commonly used to describe their relationship, i.e：
In the formula, n is a temperature-dependent characteristic index, n = 2.5~7; Un is the avalanche breakdown voltage, which varies from 70 to 200V for different semiconductor materials; U is the reverse bias voltage, which is generally taken as 80% to 90% of UgR. When using an APD, it is essential to ensure that the operating voltage is kept below the avalanche breakdown voltage to avoid damaging the device.
