DC to 65 GHz Wide Bandwidth InGaAs Photodiodes and Photoreceivers
- PDF / 2,824,428 Bytes
- 13 Pages / 415.8 x 637.2 pts Page_size
- 67 Downloads / 183 Views
POWER
InGaAs
PHOTODIODE
DESIGN
&
MODEL
Since 1995, Discovery Semiconductors has pioneered a high-speed detector technology known as the dual-depletion pin detector [7]. This design allows good responsivity, high speed, and high saturation currents from afront-illuminateddiode that is relatively easy to manufacture and integrate. The basic issues for high bandwidth detectors are the electrical capacitance of the detector and the transit time for carriers in the detector. For high optical input power, we must also consider the non-linear effects of space-charge effects, barrier trapping, and saturation. In the design process, one must balance all of these factors to achieve the optimum detector performance. This balance can be found by analyzing the underlying physics of carrier motion through a p-i-n structure. When the semiconductor is operating in the linear regime, the two major limits on device speed are the capacitive time-constant and the carrier transit-time. As analyzed in our previous work [7], there are two major capacitances to consider in optical detectors: the intrinsic 115 Mat. Res. Soc. Symp. Proc. Vol. 607 © 2000 Materials Research Society
capacitance of the diode itself, and the parasitic capacitance of the preamplifier circuitry. The intrinsic capacitance, C11 ,,, is given by the simple parallel plate capacitance of the p-i-n structure, and is proportional to the area and permittivity, A and -, and to the inverse of the depletion region thickness, d.
w
(1)
e6A
d
This capacitance, in association with the load resistance of the amplifier, R, determines the electrical time constant of detection, iRC.
(2)
T RC = RC Ilt.
Hence, the capacitance needs to be minimized and requires the depletion region (d) to be thick. The transit time, rTR, is the time required for the photogenerated carriers in the i region of the detector to move to the electrode surfaces. Given that the carriers are moving at their saturation velocities, v, the transit time is proportional to the intrinsic thickness, as given by Eq.
(3), TTR
d
(3)
-(
This time gives the limit of the duration of the impulse response of the detector. Thus, a thick detector will have a long transit time, and thus a slow response. Fortunately, the properties of InGaAs and InP are such that the electrons can travel as much as ten times faster than the holes at certain fields [7]. By fabricating the detector such that the carriers are generated only near the p contact, as shown in Fig. 1(a), a large speed increase can be realized. When the carriers are generated near the p contact, the slower holes travel a shorter distance than the faster electrons. This allows a thicker intrinsic region without the transit time penalty. Generation
n-type Fig. 1(a). A generic p-i-n detector structure, illustrating how selective carrier generation can accelerate the transit time effect because of the difference in hole and electron velocities. Our previous work involved making detectors of 10 ltm diameter and 1 itm thickness, with response speeds up to 50
Data Loading...
