Scanning-laser-microscope measurements of minority-carrier diffusion length and surface recombination velocity in polycrystalline silicon

1985 ◽  
Vol 63 (6) ◽  
pp. 870-875 ◽  
Author(s):  
S. Damaskinos ◽  
A. E. Dixon

A scanning laser microscope was used to study the electronic and recombination properties at grain boundaries of both n- and p-type Wacker polycrystalline silicon in a spatially resolved photoconductivity experiment. The light energy falling on the samples was varied over five orders of magnitude from 10−1 to 10−6 mW. For p-type material the measured L decreased with beam intensity from 150 to 60 μm, reaching a constant value at very low beam intensities. The small focal spot of the microscope allowed the measurements to be extended to include n-type samples. Forthese samples L was found to change from 90 to 18 μm with decreasing beam intensity. The surface recombination velocity SGB was evaluated for both samples. For p-type samples it decreased from 25 000 to 6000 cm/s and for n-type samples from 21 000 to 3000 cm/s with decreasing beam intensity. The quasi-Fermi level separation was determined as a function of the excess minority-carrier-concentration density at the grain boundary and found to increase linearly with beam intensity.


1988 ◽  
Vol 66 (3) ◽  
pp. 200-205
Author(s):  
S. Damaskinos ◽  
A. E. Dixon

A scanning laser microscope was used in a spatially resolved photoconductivity experiment to determine the minority-carrier diffusion length (L) and surface-recombination velocity at grain boundaries (SGB) in (i) n- and p-type Wacker polysilicon and (ii) neutron-transmutation-doped Metron polysilicon as a function of beam intensity. Different values of L were measured on opposite sides of the grain boundaries. For the Metron samples, L was measured at the same grain boundary using a series of samples doped to different levels. These samples had dopant concentrations between 1013 and 1017 atoms/cm3. L was found to decrease from 50 to 5 μm, with decreasing beam intensity, reaching a constant value at low beam intensities. L was also found to remain relatively unchanged for low dopant concentrations. The SGB values were found to increase with increasing beam intensity in both Wacker and Metron samples, ranging between 104 and 105 cm/s. L was also measured with a light-beam-induced-current technique (perpendicular geometry) and found to be in close agreement with values obtained using the photoconductivity technique.



Author(s):  
D.P. Malta ◽  
M.L. Timmons

Measurement of the minority carrier diffusion length (L) can be performed by measurement of the rate of decay of excess minority carriers with the distance (x) of an electron beam excitation source from a p-n junction or Schottky barrier junction perpendicular to the surface in an SEM. In an ideal case, the decay is exponential according to the equation, I = Ioexp(−x/L), where I is the current measured at x and Io is the maximum current measured at x=0. L can be obtained from the slope of the straight line when plotted on a semi-logarithmic scale. In reality, carriers recombine not only in the bulk but at the surface as well. The result is a non-exponential decay or a sublinear semi-logarithmic plot. The effective diffusion length (Leff) measured is shorter than the actual value. Some improvement in accuracy can be obtained by increasing the beam-energy, thereby increasing the penetration depth and reducing the percentage of carriers reaching the surface. For materials known to have a high surface recombination velocity s (cm/sec) such as GaAs and its alloys, increasing the beam energy is insufficient. Furthermore, one may find an upper limit on beam energy as the diameter of the signal generation volume approaches the device dimensions.



Sign in / Sign up

Export Citation Format

Share Document