Capacitance Transients at Variable Temperature

October 5, 2023 by Tim Ashworth

This blog post presents capacitance transient data taken on a Schottky diode at variable temperatures from 70 ^oC to 150 ^oC. It demonstrates how the MFIA can act as part of a wider experimental set-up to acquire DLTS data.

DLTS is a powerful technique to study the defect landscape of semiconductor junctions. Developed in the 1970s [1], it has been tuned and improved so that today, turn-key DLTS systems are in regular use in R&D labs around the world. These systems follow well used procedures to measure DLTS and yet provide little flexibility on the measurement frequency and other parameters. An alternative to turnkey systems is a self-built system comprising temperature stage, capacitance meter, data acquisition, and voltage or light pulse generator. These set-ups allow the user to measure at lower frequencies and implement advanced types of DLTS to suit their needs. As part of such a set-up, the MFIA replaces the capacitance meter, data acquisition and voltage pulse generator. So with just a temperature stage and an MFIA you can take the transient data used for DLTS. The final DLTS analysis is done in an additional step of post-processing with third-party software or by using the API to control the MFIA and temperature stage, acquire the data and do the data processing.

Figure 1: Photo showing the MFIA connected to the temperature stage (Linkam HFS600) via a self-made cable to adapt from BNC to the chamber feedthrough. The third-party temperature controller (Linkam T96) is also shown on the right hand side.

Fig. 1 shows a photo of the MFIA connected to the temperature stage (Linkam HFS600) via self-made cables to adapt from BNC to the chamber feedthrough (BNC feedthroughs are also available). We use 2-terminal configuration in this case, but a 4-terminal configuration would be possible by adding two additional BNC cables to create a four point measurement to the sample.

The device under test is a Schottky diode (part number STPSC12H065CT), which shows capacitive transients with increasing temperature. The DUT was mounted inside the stage with spring contacts and two probes were engaged. The initial contact was confirmed by using the LabOne Plotter Module to ensure a good contact before sealing the chamber. The goal of this was not to investigate the DUT, but rather show how the MFIA can be set up to acquire capacitance transients for DLTS data and play a key role at the heart of a self-built DLTS system.

Figure 2: Screenshot of the LabOne DAQ Module showing a series of capacitance transients measured by the MFIA. The color coding is shown in the history tab on the right hand side. The voltage applied was +2 V for 4 ms. The area highlighted with the orange circle shows the area where we observe the transients. For a zoomed in view of the transients, see Fig. 3.

Stressing the DUT

We stressed the DUT with a rectangular voltage pulse produced by the MFIA. This DC voltage was fed into AUX IN 1 of the MFIA front panel, and then internally added to the AC test signal. The voltage pulse was 2 V for 4.2 ms. The resulting capacitance transients can be seen in the colored traces in Fig.2. The experiment was set up at room temperature where no capacitance dip can be seen. Note that the voltage pulse could alternatively be routed to a light controller to provide a light pulse in the case of optical-DLTS experiments but this is outside the scope of this blog post. We optimized the acquisition by selecting a suitable number of transients to average. Each transient acquisition is triggered either digitally or from an external trigger to ensure sharp averaging.

Changing the temperature and taking data

Once the DAQ module was set, we can start to raise the temperature. The temperature was set manually using the controller shown in figure one, but API scripting could also be used. We kept all measurement parameters fixed to avoid convoluting bandwidth changes with temperature related changes. Starting at 70 ^oC, we acquired a set of 100 averages and then increased the temperature in steps of 10 degrees to 150 ^oC. Fig. 3 shows the resulting data set of 9 transients. With colour coded traces according to the temperature in the history tab on the right hand side of the DAQ Module.

Figure 3: Screenshot of the LabOne DAQ Module showing nine traces of increasing temperatures from 70 ^oC to 150 ^oC. For all temperatures, the same measurement parameters were used to avoid any influence from the instrument measurement bandwidth.

Measuring the transients

Fig. 3 shows the transients to be decreasing in rise time as the temperature increases. In Fig. 4, we measure the rise time of the transient after measuring with the in-built cursor tools of LabOne. The rise time changes from 30 ms at 70 ^oC to 0.3 ms at 150 ^oC. These transients are the raw data for DLTS data processing which include deconvolution and the application of a rate window. This post-processing is out of the scope of this blog post.

Figure 4: Screenshot of LabOne DAQ Module showing how the risetime of the transient can be measured using the math tools. The rise-time of this transient is 0.3 ms at this temperature (140 ^oC), in contrast to 30 ms when the temperature is 70 ^oC (see figure 3 to see all transients).

Conclusion

This short blog post shows the ease of set-up of the MFIA to measure capacitive transients for DLTS. The temperature was controlled by a third-party temperature stage and controller (Linkam HFS600) , and the MFIA was connected in a 2-terminal configuration with a self-made BNC adaptor to suit the feed-through of the temperature stage. Using the LabOne DAQ module of the MFIA, we could acquire and average transients at each temperature and display them all together on the DAQ Module. The MFIA provides the functionality to measure the capacitance transients and also provide the voltage pulse to stress the DUT. The same voltage pulse functionality can be used to drive an external light source if the DUT is to be stressed by a light pulse.

The combination of MFIA plus temperature stage brings the possibility to tune the DLTS measurement parameters to suit your sample. Home-building your DLTS system with the MFIA gives you flexibility and the freedom to choose the test signal amplitude, frequency and even the mode of stressing pulse.

Acknowledgement

Many thanks to Dino Klotz for taking this data, and for his generous guidance and expertise in this field.

References

[1] Deep‐level transient spectroscopy: A new method to characterize traps in semiconductors. D. V. Lang, J. Appl. Phys. 45, 3023–3032 (1974)