For many years, sunlight has been seen as a potential gold mine of useable energy for our global needs.  Having successfully used the sun to grow food to feed the world, people are now trying to harvest the sun’s energy and convert it into electric energy.  The principle way this has been accomplished is through the use of solar cells, also known as solar photovoltaics (PVs). Due to a variety of economic, scientific and safety reasons, silicon is the main material used to construct solar cells. Recent advances in technology have allowed engineers to make new types of silicon that may provide one avenue by which we can better harness untapped solar energy.

Although the sun produces ample energy, there are fundamental physical limits on how much energy from the sun can be converted from thermal radiation (heat that reaches Earth as sunlight) into the electric energy needed to supply power to people’s homes, businesses, and industrial buildings.  One of the most important restrictions on solar cell efficiency is the Shockley-Queisser Limit discovered in 1961.  It states that, with silicon, at most only about 34% of sunlight can be converted into electric energy [1].

To understand why 66% of sunlight cannot be converted into electric energy and to discuss how scientists are developing new materials to work around the Shockley-Queisser Limit, we must first understand some basics about the solar cell and semiconductors.  While conductors conduct electric energy very easily (like copper wires) and insulators prevent electric flow (like rubber coatings around wire), semiconductors are in-between, requiring a little sunlight to create electric current.  Just as an object is made up of billions of atoms, light is made up of billions of particles called photons. Photons are massless particles, each carrying a certain amount of energy. When light shines on a solar cell, the photons from the light interact with the semiconducting materials (e.g. silicon) in the solar cell. Specifically, the photons collide with electrons, very small particles that surround the nuclei (the core) of atoms.  Photons that have sufficient energy can transfer their energy to the electrons and “excite” them, allowing them to jump to a higher energy level. When this happens, it is said that the electrons have jumped the “band gap”, which is the difference in energy between something known as the valence band and the conduction band (Figure 1). The term “band” is used to refer to a range of energy levels, for instance in Figure 1, the valence band is the entire yellow ellipse which covers a range of energies (y-axis). In the valence band, electrons have a lower amount of energy and remain stuck to their atoms. However, when electrons jump the band gap to the conduction band, they can freely buzz around the pool of atoms in the silicon material. Electrons have a very small negative charge, and when many electrons move together, they create an electric current.  In solar panels, electrons normally are in the valence band, where they orbit an individual atomic nucleus, but when excited to the conduction band, the electrons flow freely through the solar cell and produce electricity.

Figure 1. The electrons sit in the valence band, but can be excited across the band gap into the conduction band by photons. The x-axis takes into account different orientations of the electron and the y-axis is the electron’s energy.  Electrons cannot sit in the band gap, and must have enough energy to get to the conduction band or they will sit in the valence band.

In terms of generating electricity, this all works wonderfully if a photon has enough energy to help an electron jump the band gap, but many photons do not have sufficient energy for this. What happens to the energy in photons that don’t carry enough energy? According to the Shockley-Queisser Limit, such energy cannot be converted to electric energy and is lost as “spectrum losses.”  The sun’s light spreads over a spectrum of energy values, which correlate to light of different colors. Some of this light is capable of exciting electrons to jump the gap, but much of it, especially light in the infrared ranges (which cannot be seen by human eyes) is lost.

Sunlight in the infrared region is made up of photons that do not have enough energy to excite electrons to jump the band gap in normal silicon. This means that infrared sunlight cannot be turned into electric energy. On the other hand, there is also a problem with photons that have too much energy. High-energy photons can excite electrons to the point that these electrons completely miss the conduction band and escape! This happens, for instance, with ultraviolet (UV) light. Of the 66% energy lost to silicon solar panels, most is lost as these spectrum losses with only 14% due to other minor physical phenomena. If solar cells could use this otherwise lost energy, they might overcome some of the current physical limits of typical PVs.

One material that might lead to improved PVs is black silicon.  Black silicon is simply a modified version of silicon, but unlike typical silicon, its surface can absorb both visible and infrared light. What makes it so special? It has a very rough surface consisting of microscopic peaks and valleys that aid in its ability to absorb a wider range of sunlight (Figure 2). This rougher variation of silicon was first produced during a manufacturing mishap, but people were quick to see that this type of surface could be very useful.  The varied surface structure causes more photons, including those in the infrared region to be physically trapped in the silicon by the many reflective surfaces. Normal silicon is comparably flat and allows the photons to pass through. To give you an idea of how small the peaks in Figure 3 are, 250,000 peaks would fit on the tip of your pencil [2]!

Figure 2. A standard electron microscope picture of black silicon.  The peaks have a height of approximately 10 microns and a diameter of one micron [3].

Eric Mazur, Dean of Applied Physics at Harvard University, heads one of the first laboratories to develop a process for producing these structures intentionally [4].  In 1999, his group used a femtosecond laser (a laser which fires a burst of high energy for 10-15 of a second) and sulfur hexafluoride gas to create surfaces similar to the one seen in Figure 2.  This process traps sulfur atoms in the silicon and turns it black, hence the name “black silicon”.

Recently, researchers at the Heinrich-Hertz Institute in Germany succeeded in doubling the efficiency of black silicon solar cells by modifying the pattern of the femtosecond laser.  They have also been working on a system of algorithms that will allow them to make the best black silicon solar cell through optimal laser use.  Black silicon has the potential to become the primary material for creating high-efficiency solar cells and could raise the overall efficiency of solar cells by 1%.

Beyond black silicon, scientists are developing many other methods to work around the fundamental limitations of current solar cell technology.  Other materials besides silicon have generated a lot of interest in the scientific community [5].  Some groups are stacking different types of solar cells in a pattern to absorb more light than achievable by a single cell.  Another method is to introduce impurities into the silicon structure so that lower-energy infrared photons (infra-red light) can push electrons to the mid-level of the band gap instead of not being used at all. For commercialization, there must be a fine balance between improving efficiencies and also keeping the cost of manufacturing low. Whatever the method, scientists are constantly looking for innovative ways to convert more energy from the sun to power the world’s electrical needs.

Michael Cox is a second year PhD student in engineering at Harvard University.


[1] Shockley, W, and Queisser, HJ. (1961) Detailed Balance Limit of Efficiency of p-n Junction Solar Cells. Journal of Applied Physics 32(3):510-520.

[2] “Velcro in Miniature” (June 21, 2006)

[3] “SEM of Black Silicon”

[4] Markoff, J. “Intuition + Money: An Aha Moment”, The New York Times (October 11, 2008)

[5] Currivan, JA. “The Promise of Organic Solar Cells: Flexible, Cheap, and Printable”, Science in the News Flash (March 2012)