Germanium has a lower bandgap than Silicon. GaAs is a direct bandgap semiconductor unlike Silicon. Still Silicon is preferred over these two, for making solar cells. Why?
Dear Madam,
the commodity price for Si is currently about 1700 US-$ per ton. You need ~ 5 g of Si for 1 W of solar power. The worldwide solar power is abot 400 GW, so you have to pay about 3.4 billions US-$ for 2 million tons of Si.
The price for Ge is currently about 800 US-$ per kilogram (!!!). Assuming the same performance for Ge and Si, you have to pay 1600 billions US-$ for the raw material.
Against this economic background physical arguments seem to play a minor role ...
Sincerely, FM
Two other semiconductors, germanium and gallium arsenide, present
special problems while silicon has certain specific advantages not available
with the others. A major advantage of silicon, in addition to its abundant availability in the form of sand, is that it is possible to form a superior stable oxide, SiO2, which has superb insulating properties. Gallium arsenide crystals have a high density of crystal defects, which limit the performance of devices made from it. Compound semiconductors, such as GaAs (in contrast to elemental semiconductors such as Si and Ge) are much more difficult to grow in single crystal form. Both Si and Ge do not suffer, in the processing steps, from possible decomposition that may occur in compound semiconductors such as GaAs. Again, due to a higher range of temperature in Si, wider range of processing can be incorporated in Si as compared to Ge in a controlled environment.
Dear Madam,
the commodity price for Si is currently about 1700 US-$ per ton. You need ~ 5 g of Si for 1 W of solar power. The worldwide solar power is abot 400 GW, so you have to pay about 3.4 billions US-$ for 2 million tons of Si.
The price for Ge is currently about 800 US-$ per kilogram (!!!). Assuming the same performance for Ge and Si, you have to pay 1600 billions US-$ for the raw material.
Against this economic background physical arguments seem to play a minor role ...
Sincerely, FM
So far, the most decisive point is missing in the comparison of silicon and germanium: the band gap of the material. And this physical argument outweighs the economical aspect mentioned by Frank Müller. Why is that so? Because of the consequences of germanium having a significantly smaller band gap (about 0.7 eV at 300 K) than silicon (about 1.1 eV at 300 K). There are two points:
(i) Solar cells operate under sunlight, which has a characteristic spectrum. As a consequence, for a standard solar cell where each absorbed photon generates a single electron--hole pair, there is an optimum band gap to acheive the highest efficiency thermodynamically possible for this illumination (so-called Shockley--Queisser limit; cf. https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limit). Since this value is about 1.3 eV, the band gap of silicon is much closer to this optimum value than that of germanium.
(ii) Solar cells based on silicon or germanium are fabricated as large-area pn junction devices, i.e. they are semiconductor diodes. Since the forward diode current is a loss, the output voltage of a solar cell is limited by the forward voltage of the diode. Due to the lower band gap, this voltage is much lower for germanium than for silicon, so the losses are intrinsically higher in germanium (independent of the material quality!).
As an overall consequence, germanium-only solar cells would be performing so badly that nobody would want them, not even as a gift. (Remember that some years ago there were some more thin-film technologies used in photovoltaics, but they have died out already because of their poor performance.)
The use of Germanium as an absorber layer, reduces the open circuit voltage and the solar cell's efficiency. On the other hand a study from D’Souza et al. showed that a 60% reduction in costs is achievable if an 8-in. Si is used instead of a 4-in. Ge.
GaAs is significantly more expensive than silicon. Behind oxygen, silicon is the most frequent chemical in the earth’s crust. Sand has a high percentage of silicon. Silicon is cheap and plentiful.
Not only for solar cell , it is well established materials for microelectronics such as CMOS, BJT, UGT, MOSFET, IGBT, SCR, etc.
Silicon is one of the optimum semiconductors that is used for solar cell production because of its superior electronic properties, optical properties, thermal properties and mechanical as well as environmental properties. In addition to its availability, manufactureability, and cost. Moreover, there is still intensive research on new silicon structured solar cells that are efficient while consuming much less quantity of silicon.
I would like that you look at the link about the silicon as a semiconductor:https://www.researchgate.net/post/Why_and_how_is_silicon_prevailing_as_a_semiconductor
Best wishes
The silicon technology is the most controlled because it has been used for a long time by microelectronics.
Dear Shaona
According to Shockley--Queisser limit, GaAs and then Silicon (because of their band gap) are the most efficient semiconductor materials for solar cells. Moreover, Si is the most abundant semiconductor material on earth. Although C-Si is of the first generation but its overperformance comparing with other semiconductors makes Si the best. Solar cell designs in research try thin microsilicon layers to save materials for mass production while a thin layer of amorphous-Si can compensate and gives the required absorption. Best wishes
Maybe the question is related to its biggest prescence over the earth and also the easyness to produce the sylicon, against the minor amount of germanium and the larger difficult to produce the material.
@Tony Castillo-Calzadilla:
No, it's not about the lesser amount of germanium available and the larger difficulty to produce the material. Even if germanium were as abundant and as easy to produce as silicon, it would not be used for solar cells (in the same way as silicon is used, namely as single-junction cell) because of its lower band gap, making it perform really bad compared to silicon.
However, if it comes to multi-junction cells (i.e. tandem or triple or quadruple cells), then it can serve as bottom cell to absorb just the longest wavelengths. Yet this is a different story and has to be discussed elsewhere -- e.g. here: https://www.researchgate.net/post/Are_the_tandem_solar_cells_and_the_multi-junction_solar_cells_the_same_and_what_the_difference_between_them_and_the_hetero-junction
1. Since monocrystalline Ge is more expensive than mono Si , see the link below. The more ΔE of the semiconductor (ΔЕ= ћF = ћ (C / λ)), the more you can get the power P of the solar battery (in theory).
A possible future development direction is the use of SiC semiconductor.
2.Monocrystalline Ge is more expensive (and Ga As, ets. too) than mono Si , see link below
http://sovet-ingenera.com/eco-energy/sun/vidy-solnechnyx-batarej.html
Jan-Martin Wagner I see your point of view and I agree with you, but I also believe that the other two factors had been taken into in consideration as well at the time of making a descision about the point discussed in this question.
@Nick Proskurin: Well, it's not fully true that the larger the band gap (ΔЕ = Ec -- Ev), the more gain one has from a solar cell -- at least so if we consider just single-junction cells. For such cells there is an optimum band gap, resulting from a trade-off between minimizing non-absorption losses (of low-energy photons) and thermalization losses (of high-energy photons). For details, see https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limit
@Tony Castillo-Calzadilla: The other two factors you were talking about are the lesser amount of material available and the larger difficulty to produce the material. Well, these factors are much more relevant for GaAs than they are for Ge. (The starting question of this thread is about a comparison of Si against Ge and GaAs.)
Ge is not used because its band gap is too far away from the optimum (cf. Shockley--Queisser limit; see my preceding reply), and this fact is already sufficient to explain why Ge is not used.
On the other hand, the band gap of GaAs is rather close to the optimum, even closer than that of Si (if I remember it correctly), but still Si is preferred over GaAs for standard cells (single junction, no concentration of the sunlight). Thus, other factors come into play for GaAs, among them those mentioned here.
For solar cell application, the best bandgap range is 1.4eV to 1 eV (900nm-1240nm). If we use large bandgap material, most of the light wavelength will not absorb through the material. Only a few ranges of light wavelength will absorb for exciting the electrons. If we use lower bandgap material for the solar cell application, then high energy photon (lower wavelength) also gets absorbed for exciting electrons. These electrons reach the higher energy level of the conduction band. After a few time, these electrons get back to the conduction band's minima through the vibration with phonon by losing energy in the form of heat (Thermalization loss). So, if we use low bandgap material, then heat is produced in the solar cell. For that reason, Germanium (Band gap=0.67eV) is not preferable. Also, GaAs (Band gap=1.441eV) is one of the most efficient materials for solar cell applications, but this material is highly cost.
So, silicon is a preferable material for the solar cell over Germanium and GaAs.
Sabiar Rahaman -- yes, we had that already (look at the older contributions in this thread); and again, this is only true as long as one considers solar cells with a single absorber layer only. However, as soon as tandem or triple cells are concerned, things may change. At present, a lot of research and development is devoted to the "silicon plus x" direction, with x = perovskite, for example; Oxford PV is about to bring this technology to the market. However, other combinations might be feasible as well. In the ideal case, a tandem cell can reach 42 % efficiency (in the Shockley--Queisser sense, 1 sun, no concentration; cf. Article Detailed Balance Limit of the Efficiency of Tandem Solar Cells
), but who knows which technology will make it in the long run?
Dear Jan-Martin Wagner, Thank you for sharing your valuable response and suggesting me that article.
Pardon me, dear BENNACER Hamza
, but silicon doesn't have a high absorption (since it is an indirect-bandgap material). That is why silicon is fabricated into wafer-based solar cells (thickness approx. 180 µm), in contrast to thin-film technologies like CIGS or CdTe (thickness approx. 10 µm or less). Yes, once there were also thin-film modules based on silicon (employing tricky light trapping methods), but they didn't reach sufficiently high efficiencies to survive in the market.
As an Indian Institute of Technology Kharagpur student, I think you have learned about SQ limit of Solar cell. It is not only about cost issues it is rather a performance issue. Si microelectronic excellency is familiar even for low power and lower band < 30 GHZ RF design Si is an excellent material. Its carrier mobility and band edge are nearly matched with SQ thermalization and below band-edge recombination problem. It is still better than the suitable SQ band edge matching CdTe and costly GaAs materials.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Semiconductor