Evolution of c-Si PV Cell Technologies
Inside this Article
Since Bell Labs introduced the crystalline silicon solar cell to the world in 1954, the technology has enabled exploration in space and transformed electrical power systems back on earth.
A solar cell is an electronic device that directly converts sunlight into electricity. Light shining on the solar cell produces both a current and a voltage to generate electrical power. This process requires a material in which the absorption of light raises an electron to a higher-energy state so that it can break free from its atomic structure and move around. Certain metals and semiconductors exhibit this trait, known as the photoelectric effect. Once the higher-energy electron is free, it must be able to move from the solar cell into an external circuit to dissipate its energy. It then returns to the solar cell to complete the circuit.
Sunlight is a form of electromagnetic radiation, and the visible light that we see is a small subset of the total incident energy the sun emits. In 1905, while studying the photoelectric effect, Albert Einstein described light as packets or particles of energy, today known as photons. Even after Einstein explained the physics of the photoelectric effect, it took many years for a practical electrical-generation application of the technology to evolve.
In this article, I briefly explore the history, anatomy, physics and lexicon of crystalline silicon (c-Si) PV cells. I then consider the evolution of modern high-efficiency c-Si PV cells, which is a function of important manufacturing advances as well as innovations in solar cell technologies. Finally, I consider some of the most promising paths forward to higher-efficiency and lower-cost c-Si PV modules for terrestrial applications. As this story will tell, it can take a lot of time—and a measure of good luck—for a solar cell technology to journey from the research laboratory to a format that facilitates mass production and a cost structure that enables commercial market opportunities.
The solar industry widely recognizes Bell Labs as the inventor of the modern-day solar cell. As solar historian John Perlin details in From Space to Earth (see Resources), Bell Labs tasked a group of scientists with developing a source of freestanding power as an alternative to traditional dry-cell batteries, and they began experimenting with photosensitive materials in 1952. After their initial attempts to improve the power output of selenium-based solar cells fell short, researcher Gerald Pearson discovered that silicon-based semiconductors, which Bell was developing for use in telephone transistors, provided a much more efficient base material for PV cells.
In 1954, Bells Labs announced its development of the Bell Solar Battery, an n-type, rear-contact silicon solar cell, shown in Figure 1, with a conversion efficiency of 6%. In spite of media praise for the invention, the company struggled to find a serious market for this device outside of novelty items, such as toys or radios run on solar. The success of the semiconductor transistor, which rapidly achieved economies of scale, ultimately made the Bell Solar Battery obsolete for the telecommunications industry at that time.
The space race was a critical turning point for silicon-based solar cells, which would subsequently take off— literally. On March 17, 1958, the US Navy launched the Vanguard 1, the fourth-ever artificial earth orbital satellite and the first to include a PV power source. Whereas earlier satellites relied on battery power only and had a mission duration of days, the solar- and battery-powered Vanguard 1 remained in service for more than 6 years. The ability to extend the useful life of orbiting satellites by allowing the sun’s energy to recharge onboard batteries was critical to success in space-based applications.
Established in late 1958, NASA demonstrated an interest in photovoltaics, spurring technological advancements that would lead to the development of more-powerful and more-reliable PV cells. With further experimentation and refinement, researchers drastically improved solar cell efficiency to around 14% by 1960. Scientists discovered, for example, that antireflective (AR) coatings on the front surface of the PV cell helped improve the absorption of light compared to bare silicon, which otherwise has a surface reflectance of over 30%. Other advancements, such as applying electrical contacts to the front of the cell rather than the rear, improved manufacturing speed and cost.
In his history of c-Si cell technologies (see Resources), Martin Green notes that while the basic cell design for space applications, shown in Figure 2, remained largely unchanged for about a decade, a number of improvements came about in the 1970s. Researchers discovered, for example, that adding a thin aluminum layer to the back of the c-Si PV cell created a back-surface field that delivered a significant boost in performance. Not long afterward, COMSAT Laboratories boosted performance further by chemically etching the surface of the c-Si PV cell to produce pyramidal structures that reduced reflection. By 1974, terrestrial cell performance had achieved conversion efficiencies of more than 17%.
In 1975, Spectrolab pioneered a screen-printing process for applying front metal contacts to the solar cell, using a metallic paste forced through a patterned template with a squeegee, a process similar to applying graphics to T-shirts. As crude as it may sound, this development ultimately led to dramatic manufacturing cost reductions and enabled PV technologies to become practical in terrestrial applications. Interestingly, oil and gas companies were some of the earliest adopters of terrestrial solar, deploying PV technology for offshore drilling wells. The process of screen-printing fingers on the front surface of a solar cell, such as the one shown in Figure 3, proved so effective that it is still in use today in large-scale c-Si PV manufacturing operations.
Anatomy of a Modern c-Si Cell
The structure of a modern c-Si solar cell is a stack of layers built up on either side of a silicon substrate known as the base layer. Silicon crystals are formed by growing a single continuous crystal ingot (monocrystalline silicon) or by producing a solid block of many different crystals (multi-crystalline silicon). The base layer of the solar cell begins as a thinly sliced wafer of c-Si that is charged either positively (p-type) or negatively (n-type). Manufacturers typically dope the p-type wafer with boron, resulting in a net positive charge. Conversely, introducing a negatively charged impurity such as phosphorous results in an n-type wafer with a net negative charge. (P-type solar cells are the most common commercial variety and have held the largest market share among all PV technologies for the last four decades.)
The front surface of the wafer, known as the emitter layer, is formed by injecting an extremely thin layer of dopants, which have a charge opposite that of the base layer. Since the near front surface absorbs a high percentage of light, the primary function of the emitter layer is to absorb incident photons, which in turn generates a pair of oppositely charged carriers. The intersection of the oppositely charged emitter layer and the base layer is known as the p-n junction. This is essentially two physically adjoined layers of silicon crystal with opposite charges, which help separate carriers based on their natural magnetic field.
Bare silicon is highly reflective. Therefore, to reduce the probability that a photon is reflected off the solar cell, manufacturers typically apply two additional layers to the front surface. First, they apply a prismatic texture, resembling miniature pyramids, to redirect photons toward the cell, thereby reducing the likelihood of reflecting light. Second, they typically apply a silicon-nitride AR coating, which helps prevent the unwanted reflection of light in the useful spectrum and acts as an effective front-surface passivation technique.
Finally, they apply metallic contacts to collect the positive and negative carriers. In a traditional c-Si solar cell, to form the rear contact, manufacturers screen-print a continuous solid layer of aluminum paste to the rear side of the solar cell. Also using a screen-printing process, they apply a silver paste to form a grid pattern on the front surface of the solar cell. The size, shape and placement of these grid lines is extremely important to cell performance. On the one hand, the grid lines block sunlight from reaching the surface of the cell, so they require sufficient spacing between them. On the other, if you space the grid lines too far apart, losses may increase due to the inherent resistance of the cell’s semiconductor material.
Physics of a Solar Cell
At its essence, the job of a solar cell is to generate a pair of oppositely charged light-generated carriers and transport these carriers so they can dissipate their power in an external load. As detailed on the PV Education Network website (see Resources), solar cells operate according to four basic steps: Generate light-generated carriers, collect light-generated carriers to generate a current, generate a large voltage across the solar cell, and dissipate power in the load and in parasitic resistances.
The generation of current in a solar cell, known as the light-generated current, involves two key processes. The first is the absorption of incident photons to create a pair of carriers known as an electron-hole pair. If an incident photon has a high enough energy when it impacts an atom in the silicon crystal, it can knock an electron loose from orbit around the silicon atom within the crystal. Once the electron is loose, it is free to move about the semiconductor and participate in conduction. However, when the photon knocks the electron loose from the orbit of the silicon atom, that leaves behind an empty space for another electron to fill. An electron from a neighboring atom can move into this empty space. When this electron moves, it leaves behind another space and so forth.
This continual movement of the space, called a hole, marks the path of a positively charged particle through the crystalline structure. A moving hole is analogous to a bubble in a liquid, except that the crystalline structure does not move as a liquid does. Rather, it is the hole that moves. Because both the electron and the hole can participate in conduction, they are called carriers. Since carriers are moving through a solid crystal material by joining and rejoining neighboring silicon atoms, their path is not as direct as one might think. They tend to bob and weave as they make their way toward the oppositely charged surface of the solar cell.
Before a carrier is collected or swept across the p-n junction, it is not extremely stable. Therefore, the carrier will exist only for a limited period of time, referred to as the carrier lifetime, before recombining with a silicon atom. If the carrier recombines, then the light-generated electron-hole pair disappears without generating current or power.
The second key process for generating current in a solar cell is to prevent carrier recombination. This is accomplished by forming a p-n junction to collect the carriers, which helps to separate the electron from the hole spatially. An electric field that exists at the p-n junction separates the carriers. If the light-generated minority carrier reaches the p-n junction, the electric field sweeps it across the junction, and it becomes a majority carrier. Connect the emitter and base layers, and the light-generated carriers flow through the external circuit, putting the solar cell into operation.
Advances in c-Si PV
Cell architecture advancements in the 1980s and 1990s propelled c-Si PV technologies to conversion efficiencies that researchers had previously thought unimaginable. As early as the late 1970s, researchers were experimenting with a technique known as surface passivation, whereby they applied an oxide layer to the surface of the solar cell to reduce carrier recombination and improve open-circuit voltage levels. Green’s research group at the University of New South Wales (UNSW) achieved such success with surface passivation techniques that in 1985 it produced the first silicon cell to exceed 20% efficiency. This breakthrough was so extraordinary that Green has likened the accomplishment to breaking the 4-minute-mile mark in running.
While the theoretical maximum conversion efficiency for single-junction silicon solar cells is around 29%, industry experts such as Richard Swanson, SunPower’s retired founder, have concluded that the practical conversion efficiency limit in commercial mass production is likely in the 24%–25% range. (See the SunPower white paper in Resources.) To put the cell technology advances of the 1980s and ’90s into context, consider that researchers produced the first 24% efficient silicon cell in 1994 and extended this record to 24.5% by 1998. Fast-forward to January 2018, when Germany’s Institut für Solarenergieforschung announced that it had achieved a new international conversion efficiency record for a single-junction silicon cell of 26.1%. (See the PVTech article in Resources.)
The fact that it took two decades to improve single-junction cell efficiencies in the laboratory by 1.6% illustrates just how difficult it is to achieve additional incremental improvements in silicon cell performance. It also illustrates just how advanced silicon cell technologies were by the 1980s and ’90s. In fact, many of the high-efficiency cell technologies in mass production today date to this period of research and innovation. The most impactful advancements during this era were improvements in the manufacturing process, which drove PV module costs down from around $70 per watt in 1978, drawing a well-documented comparison to Moore’s Law.
Interdigitated back contact cells. Researchers at Stanford University developed the first high-effciency rear-contact silicon cell in the 1980s. Originally designed for concentrating PV applications, the Stanford cell was unusual in that it placed both the positive and negative contacts on the rear surface of the cell. On the face of it, this design is similar to the original Bell Solar Battery. However, the two cells operate very differently. By using extremely high-quality n-type silicon, Stanford’s rear-contact cell allowed light-generated carriers located near the top surface of the cell to travel to the rear contacts on the bottom surface of the cell.
To maximize the efficiency of a solar cell, the carriers must travel across the p-n junction as quickly and efficiently as possible. One simple but expensive way to improve carrier lifetime is to use ultra-purified silicon wafers, as this reduces the likelihood that a carrier will interact with impurities in the crystal structure. This is especially true for rear-contact solar cells, such as the Stanford cell, which need to transport carriers a further distance.
In 1988, the rear-contact cell achieved a conversion efficiency of 22%. Today, the industry refers to this silicon cell design as the interdigitated back contact (IBC) architecture due to the repeating pattern of positive and negative contacts on the backside of the cell. Incorporated in 1985 and publicly traded since 2005, SunPower is notable for having commercialized and improved upon this high-efficiency n-type silicon cell design, shown here in Figure 4. Trina Solar and Yingli Solar have both announced IBC cell efficiency records in recent years.
Passivated emitter and rear cell. For a period of time, it appeared as though rear-contact silicon cells had an insurmountable performance advantage over front-contact cell architectures. The latter, after all, must overcome the fact that metal contacts shade about 5% of the front cell surface. In the early 1990s, however, the research group at UNSW discovered a way to improve silicon cell efficiency. While experimenting with lasers to scribe and passivate the front cell surface, the Australian team discovered that cell performance improved when they applied higher concentrations of dopants precisely underneath the areas where they would later apply metal contacts.
The process of selectively doping cell areas effectively created a new solar cell category: selective emitter solar cells. The most common cell architecture in this category is the passivated emitter and rear cell (PERC) technology, which UNSW successfully demonstrated in 1989. The team started by combining, refining and building upon earlier technology advancements that improved the cell’s ability to capture light, such as front-surface texturing and using an AR coating. Its novel advancement was to use lasers to deposit much higher concentrations of phosphorous and boron selectively and accurately under the metal contact areas. The team’s 1989 PERC solar cell, shown in Figure 5, demonstrated a conversion efficiency of 22.8%.
The selective doping process improves performance by generating an additional electromagnetic pull in the crystalline structure that encourages carriers to travel to the exact locations where the circuit can use them. Today, the industry considers PERC architecture one of the most cost-effective ways to produce high-efficiency solar cells. Manufacturers with PERC modules in commercial mass production include Aleo, LONGi Solar, SolarWorld and Trina Solar.
Passivated emitter rear locally diffused cell. Not long after documenting the PERC architecture, the UNSW team announced that it had achieved a new world record for a single-junction silicon cell efficiency. The team had improved on its PERC design by laser-embedding tiny, highly doped dots on the rear surface of the cell rather than using a continuous metallic contact. In so doing, it was able to produce a 24% efficient solar cell in 1995 and extend the efficiency record for a single-junction c-Si solar cell to 24.5% by 1998— a record that stood until 2014.
The UNSW team dubbed its new architecture passivated emitter rear locally diffused (PERL), since it uses local diffusion at the rear-point contact to help collect light-induced carriers and reduce contact resistance. While the PERL architecture is extremely efficient, this cell technology and other specialized selective emitter techniques have yet to achieve commercialization due to the extremely high costs associated with the manufacturing process. However, the UNSW team made its mark on history, as PERC technology has become increasingly popular in both monocrystalline and multi-crystalline PV cell production.
What Lies Ahead?
It is very difficult to improve silicon cell efficiency while reducing costs. Consider that it has taken manufacturers more than 20 years to commercialize some high-efficiency cell architectures. This speaks largely to the difficulty of designing affordable, reliable and repeatable manufacturing techniques that utilize existing equipment and supply chains as much as possible. While IBC and PERL are good technologies for improving cell efficiency, they may not be the most promising options for commercial mass manufacturing because of the higher costs associated with the more-complex production processes. As a result, most manufacturers are looking to improve module output based on existing cell technologies, since these efforts require much lower investments.
Half-cell modules. A general manufacturing trend in the industry, going back several decades, is to slice as many silicon wafers as possible from a single ingot, resulting in increasingly thin wafers. Since cells are susceptible to breakage in the production process, there is an inherent tradeoff between saving material cost during production and incurring material cost associated with broken and scrapped cells. One clever method discovered for reusing broken silicon cells is to trim them in half for use in half-cell PV modules. Half-cut cells generate half the current of a standard cell, which reduces resistive losses in the interconnecting busbars within the module. Reducing internal resistance between the cells increases module power output. As a result, manufacturers today can increase module power output by 5 W–10 W, potentially at a lower cost per watt, by intentionally cutting cells in half. Manufacturers commercializing half-cell module designs include JinkoSolar, LONGi Solar, Mitsubishi, REC Solar and Trina Solar.
Bifacial modules. The global uptick in PERC module production is likely to lead to a future increase in the production of bifacial modules. As shown in Figure 6 (p. 28), cell manufacturers can produce bifacial solar cells by adding just one processing step to the standard PERC cell production line. Bifacial modules convert light captured on the backside of the module into electrical power, which could increase energy captured in the field by 10%–15% with only a modest increase in manufacturing and installation costs. (See “Bifacial PV Systems,” SolarPro, March/April 2017.)
Module manufacturers often use a glass-on-glass package for bifacial PV cells rather than the usual glass-on-film package. Bifacial PV systems also require specialty racking systems and unique mounting considerations to capture the maximum bifacial benefit. This bifacial ecosystem is emerging now. Over the last few years, PV module glass suppliers have begun offering ultra-thin PV glass (<2 mm thick), which reduces the weight of the resulting glass-on-glass module. Mounting system manufacturers are now offering specialty bifacial mounting systems, including single-axis trackers for large-scale PV power plants. Module manufacturers commercializing bifacial PV modules include LG, LONGi, SolarWorld, Sunpreme and Yingli Solar.
Multi-junction cells. Another way to improve cell performance is to stack multiple p-n junctions to selectively filter out light passing through the cell based on its energy level. For example, the manufacturer can tune a p-n junction near the top surface of the solar cell to absorb more light in the blue spectrum and a p-n junction toward the rear of the cell to absorb more red-spectrum light. Multi-junction cell designs have been around for decades and have carved out a niche in space applications and in concentrating PV. However, manufacturers have struggled to find commercial applications for this cell technology in conventional terrestrial applications due to prohibitively high manufacturing costs.
This is beginning to change, as it is increasingly common to see PV cell designs with additional p-n junctions built by depositing thin-film materials on a c-Si base layer. These so-called hybrid or heterojunction solar cells can take advantage of many of the benefits of thin film’s light-absorbing properties at only a fraction of the cost of building a pure c-Si multi-junction solar cell. The multi-junction cell trend will likely evolve as researchers and manufacturers learn more about perovskite materials, which some have dubbed a “wonder material.” Greentech Media reports (see Resources) that Oxford PV claims to have achieved 27.3% efficiency using a perovskite-silicon tandem junction cell technology and believes the technology is capable of breaking the 29% silicon cell efficiency limit. Since perovskites are affordable and can be tuned to low-energy wavelengths, these materials could begin to replace the top thin-film layer in heterojunction cells as they make their way into commercial production.
While multi-junction perovskite cells are admittedly complex, this is just one example of the exciting work under way to develop higher-performing silicon cells. As new cell technologies come to market, researchers can optimize PV module encapsulation materials to better match the light-absorption capabilities of these new cell designs. Each marginal increase in module output and efficiency is important because it ultimately serves to reduce the per-watt costs associated with a whole range of project variables, from transportation to land acquisition to the entire BOS ecosystem.
Blair Reynolds / SMA America / Rocklin, CA / sma-america.com
PV Education Network / pveducation.org
Deign, Jason, “New Efficiency Record for Perovskite Solar—Can Oxford PV Hit 30% by 2020?” Greentech Media, June 28, 2018
Martin, Green, “Developments in Crystalline Silicon Cells,” Solar Cell Manufacturing: Developing Technologies, edited by Gavin Conibeer and Arthur Willoughby, John Wiley & Sons, 2014
Osborne, Mark, “ISFH Pushes P-Type Mono Cell to Record 26.1% Conversion Efficiency,” PVTech, February 7, 2018
Perlin, John, From Space to Earth: The Story of Solar Electricity, Harvard University Press, 2002
SunPower, White Paper: SunPower Panels Generate the Highest Financial Return for Your Solar Investment, Summer 2008