Rapid and Accurate C-V Measurements on the MFIA

Fast C-V measurements up to 3000 V/s

This blog post demonstrates the ability of the MFIA (also MFLI with MF-IA option) to measure capacitance on a short timescale (20 us) as a function of DC bias voltage, for example, for fast capacitance-voltage profiling measurements. Voltage sweeps of up to 3000 V/s are demonstrated.

C-V profiling involves measuring the capacitance of a semiconductor device or material as a function of the applied DC bias voltage. It is a commonly used technique to characterize myriad devices and materials such as MOSFETS, solar cells, OLEDs, Schottky diodes and TFTs, and serves to investigate the depletion region of the device.

Understanding the depletion region of a semiconductor junction is critical to understanding its behaviour. It consists of an area at the interface where the electrons and holes have been pushed out, meaning there are no free charge carriers. Although there is a lack charge carriers in the depletion region, it still contains electrically active defects (traps) and ionized sites which give rise to a capacitance. C-V profiling thus gives information on how the width of the depletion region (depletion width) changes with applied DC bias voltage. This yields important information on the trap density, doping concentration and doping profile of the device. As the charge carrier transfer is dominated by traps, it happens slowly within the depletion region giving rise to a hysteresis effect in C-V profiling. The traps can be frozen by cooling the device to low temperature thus giving access to the doping profile. However cooling is expensive and another method is to vary the speed of the voltage ramp used during C-V profiling.

 

Figure 1: MFIA with the MFITF fixture. A commercially available monocrystalline solar cell is mounted on an MFITF sample carrier inserted into the MFITF. The AUXIN1 connector receives the DC bias voltage generated at AUXOUT1 (requires MF-MD Option). Note: an external voltage generator can be substituted if no MF-MD option is available.

 

The MFIA can measure C-V straight from the box using the included sweeper module of the LabOne user interface. Figure 2 shows a DC bias voltage sweep taken on a commercially available solar cell in dark-state from -2 V to 1.8 V. The capacitance is measured as a function of offset DC bias voltage with an AC test signal of 100 mV at 1 MHz. The DC bias voltage was added to the test signal using the Aux Input 1 connector. Figure 2 shows that the capacitance (red trace) rises from 1.63 nF at zero bias, to 2.36 nF at +1 V bias. The sweep was taken at the leisurely speed of just 0.1 V/s.

 

Figure 2: LabOne Sweeper Module results showing capacitance (red trace) as a function of offset DC voltage. The trace comprises 200 points and took 10 seconds to acquire, it shows the capacitance of the cell rise from 1.63 nF to 2.36 nF at + 1 V DC offset bias. The voltage ramp speed was 0.1 V/s.

 

As the goal of this blog is to demonstrate fast reliable CV measurements with the MFIA, it was necessary to choose a sample whose C-V behaviour does not depend on bias voltage sweep speed. The monocrystalline solar cell sample suited this need nicely.

The MFIA can provide a DC bias voltage both internally and externally, via the AUXIN 1 connector on the front panel. For the measurements presented in this blog post, the sawtooth voltage pattern was generated by the MFIA (using the optional MF-MD option) and fed back into the AUXIN 1 as shown in the figure 1.

Starting with the LabOne plotter module, we can acquire the capacitance and DC bias voltage as two simultaneous traces and run the plotter continuously. Figure 3 shows the resulting traces, which show that the value of Capacitance (red trace) at zero offset is 1.63 nF. As the DC bias voltage (blue trace) is ramped to 1 V, C reaches 2.36 nF. Both of these capacitance values match nicely the slower sweep taken in the sweeper module shown in figure 2. This is to be expected as the voltage sweep rate is still only 0.5 V/s.

 

Figure 3: LabOne Plotter module showing capacitance (red trace) and corresponding DC bias voltage (blue trace). The voltage sweep rate is 0.5 V/s. The measured values of capacitance nicely match the measurements taken with the Sweeper module shown in figure 2.

The above sweep is still relatively slow, taking 2 seconds to complete the voltage ramp. The next step is to increase the speed of the ramp. For this, we turn to the LabOne DAQ module, which acquires a user defined number of data points and has an easily configurable trigger. The DC bias voltage ramp time is now at 2 ms, which corresponds to 500 V/s, shown in figure 4.

 

Figure 4: LabOne DAQ module showing capacitance (red trace) and corresponding DC bias voltage (blue trace). The voltage sweep rate is now 500 V/s. The measured values of capacitance nicely match the measurements taken with the Sweeper module shown in figure 2 and the Plotter module in figure 3.

Figure 4 shows the resulting capacitance behaviour for a ramp time of 2 ms (500 V/s). The values of capacitance at both zero offset and + 1 V matches the initial (slow CV) values; 1.63 nF and 2.36 nF. This demonstrates the high time resolution of the MFIA to measure capacitance. In fact, figure 5 shows a zoom into this data set at the point where the voltage is reset to zero shows that the value of capacitance is accurately measured before (2.36 nF) and after (1.63 nF) the reset. The time required for the MFIA to capture these accurate capacitance measurements is just 20 us!

 

Figure 5: LabOne DAQ module showing capacitance (red trace) and corresponding DC bias voltage (blue trace). The voltage sweep rate is now 500 V/s. The measured values of capacitance nicely match the measurements taken with the Sweeper module shown in figure 2 and the Plotter module in figure 3. The timebase has been zoomed to show the time required to measure the capacitance accurately before and after the reset of the DC bias voltage from 1 V back to 0 V. This time is just 20 us.

 

Take it to the bridge
The ability of the MFIA to measure capacitance on such a fast timescale is a very useful aspect when measuring fast C-V profiles. To push the limit, we increased the range of the DC bias voltage from -9.5 V to 1.5 V. A total of 11 V, swept in 3.3 ms, resulting in a voltage gradient of 3000 V/s. Figure 6 shows how the MFIA can reliably track the capacitance change even when the DC bias voltage is swept at 3000 V/s.

 

Figure 6: LabOne DAQ module showing capacitance (green trace) and corresponding DC bias voltage (blue trace). The voltage sweep rate is now 3000 V/s. The measured values of capacitance nicely match the values taken with slow voltage sweeps: 703 pF at -9.5V and 5.29 nF at +1.5 V.

Conclusion
we demonstrated fast C-V measurements of a monocrystalline solar cell using the MFIA. The quasistatic C-V measurements taken with the LabOne sweeper module show the unbiassed (zero offset) capacitance to be 1.63 nF rising to 2.36 nF at an offset voltage of +1 V. As the DC bias voltage ramp speed is increased, the MFIA is able to reliably measure and track the capacitance throughout the whole voltage sweep. Using the LabOne DAQ module, the fastest capacitance measurement is shown to be just 20 us, and could track capacitance during a DC bias voltage sweep rate of 3000 V/s.

 

Footnote:
A note on the MFIA voltage limits here; while offset bias voltages of up to +/-10 V are possible in a 2-terminal mode, in 4-terminal mode the offset bias voltage limit is +/- 3 V. As the solar cell has a relatively low impedance at 1 MHz of 100 Ohms, 4-terminal mode was used for the data presented in figures 2-5, and 2-terminal mode was used for the 3000 V/s sweep in Figure 6.