Weak Shunt Identification in Solar Cells Production with Infrared Cameras
Illuminated Lock-in Thermography for Detecting and Eliminating Shunts in Solar Cells
Challenge
Shunts in silicon solar cells, caused by process defects or material irregularities, generate local heat and reduce efficiency. These failures are often invisible and difficult to detect with conventional steady-state thermography, making reliable identification in production lines a persistent challenge.
Solution
Using illuminated lock-in thermography, thermal responses of shunted areas are separated from intact regions by analyzing dynamic heating behavior under periodic light excitation. This method enables high-sensitivity detection of weak and otherwise undetectable defects in real production environments.
Benefits
- Detects even weak shunts that steady-state thermography often misses
- Enhances solar cell efficiency by preventing faulty units from reaching modules
- Reduces long-term module degradation through early defect identification
- Supports data-driven quality control using dynamic thermal evaluation
- Enables cost-effective defect screening suitable for mass production environments
Enhancing Efficiency by Detecting and Eliminating Shunts in Silicon Solar Cells
Silicon solar cells In terms of working principle photodiodes. They are conventionally made of bonded p-doped and n-doped wafers of thin thickness. The resulting p-n junction is covered by a silicon nitride layer that acts as an antireflection coating on the n side. The contacts to collect the current from the solar cell are made by a grid of silver lines on the front side and a full-area aluminum contact on the backside. To avoid a process-induced short circuit in solar cells, their edges must be isolated by a laser or chemical etching.
Any disruption of the p-n junction due to cracks and scratches that occur during the processing of a solar cell, as well as the edge, may produce failures. Many of these defects lead to unwanted leakage current. These types of failures are called process-induced shunts. Multicrystalline silicon may contain precipitates that can cause short circuits in solar cells, known as material-induced shunts. Shunts in solar cells are regions with increased dark current, which can significantly reduce the cell’s efficiency. It’s important to note that all types of shunts, regardless of their origin, have a detrimental effect on the efficiency of solar cells and should be avoided at all costs.
Shunts are often not visible, but manufacturers of solar cells. Still, manufacturers need to find and eliminate these to improve the efficiency of the cell. Additionally, shunts cause localized heating, creating hot spots that can damage the entire solar cell or module in the long term. This can lead to premature degradation or failure of the solar module. Identifying and eliminating shunts and delivering shunt-free solar cells is essential for reputation and market competitiveness.
Shunts create localized areas of increased current, which also generate local heat. Depending on the size of the shunt, they create thermal hotspots ranging from a few mK to several degrees Kelvin compared to the background. Since the noise limit of conventional infrared cameras is realistically in the range of 40 to 100 mK, this allows the detection of relatively strong local heat sources. This is sufficient to detect strong shunts.
Nevertheless, small solar cell failures might not be visible with conventional steady-state thermography. Additionally, due to lateral heat conduction in the solar cell, local heat sources might appear blurred in steady-state thermography.
Dynamic Shunt Imaging in Solar Cells with Illuminated Lock-in Techniques
Illuminated Lock-in techniques (ILIT) must be used to image very weak shunts in solar cells. The Illuminated lock-in thermography is a non-destructive characterization method fast enough for solar cell production lines.
The general idea is to exhibit the shunts’ dynamic, time-dependent thermal behavior. Shunts have different thermal material properties so that they will act thermally differently than other parts of the solar cell junction. When the input energy wave penetrates the object’s surface, it is absorbed and undergoes a phase shift. Upon encountering regions within the object where the thermophysical properties differ from the surrounding material, part of the wave is reflected. This reflected wave interferes with the incoming wave at the surface, creating an interference pattern in the local surface temperature that oscillates at the same frequency as the thermal wave. Therefore, a periodic excitation of fixed frequency is applied to the solar cell, and its time-dependent thermal behavior is observed over multiple iterations. The shunts will respond to thermal excitation with different amplitude and phase responses than other solar cell parts.
While the light pulse is a step function, with illumination intensity going up and down, repeating over and over and introducing the heat with that frequency component into the part, a thermal camera then monitors the solar cell’s heat coming out. As a quality test, the solar cell flash ILIT involves applying high-intensity light to the solar panel. During a flash test, the PV module is exposed to a brief but intense flash of light from a xenon-filled arc lamp, metal halide lamps, or LEDs. The light spectrum emitted by these lamps should closely match the sun’s spectrum.
The flash light lamp produces a continuous spectral power distribution with a color temperature of about 5500K to 6200K, covering a wavelength range from 200 nm to 1200 nm. The solar cell temperature during testing is typically around room temperature, with the optimal temperature for solar panels being approximately 25 °C. Using a short-wavelength infrared camera in the same spectral range as the light emission may not be feasible due to potential crosstalk issues and the relatively low-temperature range. Therefore, a conventional infrared camera with a long-wavelength (LT) range is an optimal choice.
Light pulses are applied with the lock-in frequency, which triggers the light-emitting sun simulator. The lock-in frequency depends on the frame rate of the infrared camera. The maximum usable lock-in frequency is a quarter of the frame rate of the infrared camera due to the Shannon-Nyquist sampling theorem and the two-phase correlation between the input signal, the thermal excitation, and the thermal response. For an uncooled infrared camera, due to the bolometer detector’s limited response time, the bolometer’s thermal time constant is between 8 and 15 ms, which limits the physically relevant sampling frequency to about 125 Hz. Therefore, a lock-in frequency of less than 30Hz or exposure times down to 30 to 50 ms is possible. Nevertheless, IR cameras cannot be synchronized directly with the lock-in process, but the light pulse trigger can easily be derived from the infrared camera frame trigger signal.
After recording, a stack of thermal images shows the heat-up and cool-down phases with high synchronization to the excitation. Each image’s time series of each pixel is extracted and related to the input signal. Different mathematical methods can evaluate the transfer function between the signals. The power spectral density to quantify the peak heat-up amplitude and the cross-correlation of the two signals, which reveal the phase delay, are often utilized.
Improving the signal-to-noise ratio can be achieved through more iteration, higher excitation, or utilizing infrared cameras with lower thermal noise.
Utilizing Optris Xi400 for Solar Cell Analysis with Lock-in Thermography
Illuminated lock-in thermography is a highly sensitive tool for analyzing solar cells, offering detailed insights into local efficiency and failure mechanisms. Its ability to detect weak heat sources and perform quantitative evaluations makes it invaluable for improving solar cells by locating even very weak shunts.
Depending on the severity and type of shunt, manufacturers might attempt to repair the affected cell. This can involve laser-based techniques to isolate the shunted area or reprocessing the cell if the defect is surface-related. If a shunt cannot be effectively repaired, the defective cell is typically removed from the production line, preventing it from being integrated into a module.
Optris does not deliver evaluation software for the lock-in algorithm because every production process is unique. The data results must be referenced to healthy cells and require interpretation. A lock-in algorithm essentially identifies differences in the material but does not quantify the shunt. Usually, the lock-in algorithm must be tweaked and improved iteratively to gain the most insights into the material.
Cooled infrared cameras, which often offer a lower thermal noise and higher frame rate, might result in crispier lock-In images at a much higher price point. Still the utilized Xi400 infrared camera proves to be good enough and that the price performance ratio for test machinery in mass produced application is crucial.
The Optris Xi400 infrared camera high frame rate of 80 Hz and its ability to trigger the flashlights of the sun simulator was especially crucial in this application. Additionally, its extensive SDK and connectivity to Matlab or Python allow engineers to access thermal images and apply their functionality easily.
Recommended Products
Talk to us about your IR Temperature Measurement Requirements
There are over 300 different pyrometer variants to choose from in the Optris infrared pyrometer portfolio each optimized for material, spot size, distance from the target, and environmental conditions. Fortunately, there is a trained engineer to phone or chat with to guide you through the process of choosing the perfect infrared sensor for your application.
The same support is available for the extensive IR camera product line.
