The photovoltaic (PV) industry is currently experiencing a greater than 20% annual growth that is expected to continue and possibly accelerate in the coming years. This market acceleration is being driven by continuing subsidies, a desire to reduce dependence on fossil fuels and carbon emissions and significant module price reductions. Solar cell prices were reduced by 30% in 2009 alone as a slower economy, tighter credit and increased manufacturing capacity combined to reduce demand relative to supply.
The key driver for commercial success for PV manufacturers is the overall cost to the end users, whether this be residential installations or grid-connected, utility-scale systems. To minimize the cost-per-watt, the conversion efficiency of cells must be maximized as designs move from the research lab to full-scale production, and quality must be maintained throughout the manufacturing cycle.
Multiple photovoltaic technologies compete for share in the growing solar market. Crystalline solar cells are the most mature technology and continue to dominate global production in total power-generating capacity. Amorphous thin-film silicon cells, which are generally lighter, more configurable, and have higher fill factors but also are less efficient, are gaining significant market share. Additional PV cell structures, including Cadmium Telluride (CdTe), CuInGa(Se)₂ or CIGS, and multiple junction (III/V) cells are also experiencing rapid growth. While there are advantages and disadvantages to each of these technologies, all share the need for precision surface metrology for quality control.
Surface metrology systems are used to optimize the design of cells, as well as to maintain efficiency and quality in production. Two-dimensional (2-D) stylus systems have long been deployed for stress, film thickness and basic roughness measurements. Recently, however, the added information provided by three-dimensional (3-D) optical systems, as well as their speed and versatility, are finding such systems complementing and, in many cases, displacing stylus technologies. With advanced surface metrology, solar cell manufacturers are able to increase yield and lower the overall production cost of solar cells through quantification, qualification or monitoring of various process steps.
Texture CharacterizationTexture has long been known as one of the critical surface parameters affecting solar cell efficiency. Roughening the surface allows reflected photons a chance to strike another portion of the PV material, be absorbed and generate power. Regular or random features are designed into many solar cells to increase their ability to trap light. Without such roughening, a silicon solar cell, for instance, reflects about 40% of the incident light, and most other PV materials would experience similar light losses.
On the other hand, if surfaces become too rough, the mean free path of the electron/hole pair becomes too long and efficiency can be reduced. In addition, the deep features of a rough surface can make cells less mechanically strong, leading to electrical shorts or other issues.
One of the more elusive aspects of texture has been determining the quantitative relationship between texture and solar cell efficiency. Creating the best texture has typically been done via trial and error, and directly relating process changes that affect texture to the final cell properties has not been possible. PV manufacturers are using both stylus and optical profilers to examine different surface metrics for the correlation of relative cell efficiency to texture.
Traditional 2-D surface parameters, such as average or root-mean-squared roughness (Ra, Rq) have been shown to have no correlation to cell performance. However, over a wide variety of textures, linear relations between different 3-D surface parameters (S parameters) and relative cell efficiency can be established. These S parameters are part of the current ISO effort to update surface metrology standards from traditional 2-D guidelines to the additional information possible through 3-D metrology techniques.
In an effort to quantify this correlation, there was an investigation of several sets of monocrystalline solar cells that utilized differing types of chemical etch processes. The surface skewness (Ssk) parameter, which measures the relative number and strength of peaks vs. valleys, was found to correlate over 95% to relative efficiency. Traditional methods of quantifying texture, such as average roughness (Ra) or feature counting did not correlate with efficiency. Other solar cells showed greater than 90% linear correlation to the surface bearing ratio index (Sbi) to relative efficiency. Thus far, more than 10 different cell types with optical profilers have been tested, all of which have shown a quantitative correlation between 3-D surface texture parameters and efficiency.
Trace and Line Width MeasurementsIn addition to excellent vertical resolution and rapid measurement times, noncontact optical profilers can segment data to evaluate critical properties on different levels of a sample surface. In PV applications, optical profilometry has become the preferred technique for large-scale trace and line width measurements.
The conductive traces on a solar cell are both very costly and reduce efficiency. As such, the goal of PV cell manufacturers is to minimize the surface area covered and the material used, while maintaining excellent electrical properties. This requires fine control of the width, height and roughness of the traces.
Similarly, scribe lines, such as those used in thin-film processes, are later filled with conductive inks to create electrical connections between the various active areas. If the scribe lines are too shallow, conductivity may be too low, while lines that are too deep may create electrical shorts. The latter is particularly a problem near the end points of some laser scribe lines, where the scribe line generally deepens due to the scribing process. Misplaced lines may not connect to other features, thus leading to a failure to meet power generation requirements.
Today’s most advanced optical profiling analysis software automatically calculates line width, line spacing, depth, volume and roughness within the trace and on the substrate, as well as logs all parameters to a database with pass/fail capability for production control. These analyses can be performed on surfaces with varying features within the field of view, and many fields of view may be stitched together so that entire traces can be quantified and the process tightly controlled.
Advanced Materials ResearchAs solar cell designs continue to be refined and new materials and processing techniques are evaluated, the need to quantify surface properties under varying conditions increases. The rapid, area-based measurements of optical profilers also are well-suited to characterizing surface properties of materials and cataloguing the results of varying process conditions.
The Material Science and Engineering department of the University of Illinois, for example, utilizes an optical profiler to characterize grain boundary effects on the growth and optoelectronic efficiency in CIGS bi-crystals. Understanding what affected the growth of the CIGS material and how the device behaves on either side of the boundary are important.
By quickly quantifying these and other interactions at high resolution, optical profilers help researchers optimize solar cell performance. Image segmentation software even allows for counting and characterization of defects or contamination by volume, width, height and slope, so that the process may be better understood and improved.
Film ThicknessMaterial thickness, both transparent and opaque, needs proper characterization, particularly for CIGS devices. Contact stylus methods provide very rapid and accurate film thickness measurements, down to 10 nanometers, where there is an accessible film boundary.
The most advanced stylus profilers maintain very low contact forces, and are thus able to do this without damage, even on soft polymers. More importantly, because it is a contact technique, stylus profilometry is insensitive to the material property differences that can create offsets in optical techniques if the materials are too thin or have differing absorption. This information is obtainable in only a few seconds, so it becomes practical to perform frequent checks on the process quality.
If there is no film boundary, however, stylus profilers cannot measure the film thickness. For such cases, white light interferometric optical profilers are ideal for characterizing thicker, transparent films. Transparent films greater than about 2 microns in thickness can be measured across the entire sample surface. The optical system delivers faster area-based measurements than the stylus systems, and they are excellent at identifying defects, as well as regions of thicker or thinner material.
Additionally, the optical profilers can provide information on the surface roughness and defects for both the top and bottom surfaces of the films separately, so that the conformal properties of the film can be analyzed. If height offsets from the optical properties are present, however, a stylus profiler is more able to quickly calibrate the film and the offsets automatically applied to the optical data. Thus, as many solar manufacturers have learned, the two types of surface profiler systems (optical and stylus) work well together to ensure both film thickness and surface qualities are well-characterized to improve and maintain peak performance.
Meeting the Rapid AdvancesVarious PV technologies are competing to provide the lowest-cost solutions for a variety of deployment situations. Researchers and corporations are working hard to improve efficiency, lifetime and quality so that the total cost of ownership of PV systems is competitive with other power-generation technologies. Accurate surface metrology of key features is a critical part of this process. Optical and stylus profilers complement one another to provide the data necessary to improve solar cell development and production.
For the first time, the technology exists to enable texture to be quantitatively linked to cell efficiency. In addition, other metrics such as step height, trace width and height, scribe width and depth, film thickness and defect detection all help improve development and control the manufacturing process.
Meanwhile, researchers can use these surface profiling technologies to study material effects, environmental attack and fatigue, as well as perform sophisticated characterization to better understand the effect individual changes in the process equipment can have on the end products. Q