In Situ Monitoring of the Accelerated Performance Degradation of Solar Cells and Modules: A Case Study for Cu(In,Ga)Se₂ Solar Cells

This method can help to answer key questions about photovoltaics'reliability, about the combined effect of illumination, electrical loads, humidity, and temperature. Compared to standard non-stress reliability tests, this method has the following advantages:Exposure to combined stresses, reduced testing time, and real-time performance monitoring of the device. The combination of stresses can be tuned to simulate the local climate, such as polar and desert conditions.

Electrical biases can be tuned to simulate effects such as personal shading. Real-time performance measurement allows faster and simpler testing. It learns us a lot about the process of degradation and it allows better prediction or prevention of the degradation mechanisms occurring in devices.

Demonstrating the procedure will be Hank Steijvers, Klaas Bakker, and Karolien Saliou, all from Solliance. To begin the procedure, deposit 0.5 microns of molybdenum on a soda lime glass substrate by direct current sputter coating in a bilayer stack sequence. Then, use a solution of one-molar sodium hydroxide and 0.3-molar potassium ferrocyanide to electrochemically etch a six-millimeter wide strip in the molybdenum at the edge of one long side.

Next, deposit a two-micron thick CIGS absorber layer by coevaporation in a vacuum chamber under a copper, indium, gallium, and selenium atmosphere. Then deposit 50 nanometers of cadmium sulfide on the sample by chemical bath deposition. Use radiofrequency sputtering to deposit 50 to 65 nanometers of intrinsic zinc oxide and 80 to 1, 000 nanometers of aluminum-doped zinc oxide on the sample.

Next, use a blade to remove the top four layers in a 14-millimeter wide strip on the non-etched long edge of the sample, leaving the molybdenum back contact intact. Center a mask on the sample and sputter coat the sample edges with 60 nanometers of gold to form the contacts, then use a glass cutter or diamond pen to cut the sample into seven-millimeter wide pieces, thereby forming samples with seven millimeter by five millimeter cell surfaces. You can test any device you want, any solar cell, any module, as long as you can contact them in a reliable way.

This requires contacts made of stable materials like gold. Otherwise you might be testing the stability of your contacts instead of of your devices. Next, measure the ex situ current voltage performance of the sample solar cells under standard conditions in a four-point probe configuration.

Then, place a sample under an illuminated lock-in thermography device with IR illumination and a 15-micron IR camera lens. Illuminate the sample and map the spatial differences in temperature. Use this to identify useful and bad samples.

After that, place the sample under a photoluminescence mapping instrument with a high-power LED and record a spatial photoluminescence image. You should also use several other analysis techniques including electroluminescence, spectrophotoluminescence, external quantum efficiency measurements, and microscopy. Based on these measurements before and after exposures, the degradation mechanisms can be determined and linked to the pre-exposure properties.

Evaluate each sample for visual and lateral defects in this way. Store at least two non-adjacent samples in an argon-filled glove box as references. Mount the non-reference solar cells in sample holders that cast no shadows on the cells.

Ensure that the measurement pins of the holders contact the front and back gold contacts of the samples. Place the sample holders on the sample rack of the CSI setup to enable electrical contact between the solar cells and the measurement tools. Connect thermocouples to the samples.

Position the sample rack for illumination by an air-mass 1.5 light source, then turn on the measurement equipment, the electrical loads, and the control computer. Open the measurement logging software to initialize the instruments, then design illumination profiles for the measurements. Next, fill in the sample information and select the linked thermocouple for each sample position.

Then, set the initial voltage, final voltage, and number of steps for the current voltage measurements. Set the autosave locations for the current voltage data. Define electrical biases for the samples if desired.

Then, create a measurement sequence and add the appropriate sample locations. Set the waiting time between sequences in the Automatic Measurement window, followed by a rapid increase to 85%relative humidity, then start the climate chamber sequence, turn on the illumination, switch to the monitoring window, and start recording current voltage measurements. Monitor the chamber and sample temperatures during the ramp to 85 degrees Celsius.

Confirm that the electrical parameters are being logged and current voltage curves are being generated. Once the chamber reaches 85 degrees Celsius, confirm that the chamber humidity increases to 85%Note this as the starting time of the degradation experiment. Leave the samples in the instrument for hundreds to thousands of hours, measuring the current voltage curves every five to 10 minutes.

Adjust the electrical biases applied to the samples during the experiment as desired. At the end of the experiment, allow the chamber to cool to room temperature over several hours before removing the samples. Plot the changes in electrical parameters as a function of exposure time.

Remove the cold samples from the chamber and promptly repeat the ex situ measurements. Execute all measurements used before exposure. Afterwards, characterize both the degraded and reference samples with X-ray diffraction, secondary ion mass spectroscopy, scanning electron microscopy, SEM, X-ray photoelectron spectroscopy, and other techniques to further investigate failure mechanisms.

In this example, data recorded during the temperature ramp prior to CIGS solar cell degradation experiments showed that the open circuit voltage varied as a function of temperature. These CIGS solar cells degraded in efficiency when simultaneously exposed to light, heat, and humidity. Minimal degradation was observed when the solar cells were exposed to dry heat and light.

Here a low negative-bias voltage had a more negative effect on CIGS solar cell stability than short circuit, open circuit, or maximum power point conditions in damp heat and light. A set of CIGS solar cells fabricated with high sodium and potassium contents initially showed high efficiencies when illuminated in damp heat, but they degraded faster than standard cells. However, cells fabricated with low alkali contents remained relatively stable under the same conditions.

Further analysis revealed a corresponding sharp decrease in the shunt resistance of alkali-rich cells, which was attributed to the migration of sodium. Following these results, the devices should be analyzed thoroughly again. Based on these results, the degradation mechanisms of the device can be determined.

This technique also allows the determination of degradation behavior of full-scale modules.