Qualitatively, phonons indeed scatter light, absorbing photon energy in the process. Quantitatively, several issues are to be considered: the energy range of the photons (i.e. the spectrum of the incident light --- any crystal acts like a filter on the spectrum of the light incident on it), the size of the band gap and the temperature. Phonon energies are typically of the order of 10-3 eV (generally, it is important that phonon spectrum in a crystal has a sharp high-energy cut-off, the temperature associated with this cut-off energy being of the order of the Debye temperature), while semi-conducting band gaps are typically of the order of 1 eV. This implies that low-order scattering events do not contribute to the scattering cross section of phonons; on the basis of the above-mentioned energetic considerations one needs of the order of 103 scattering events in order to connect an initial electronic state in the valence band to a final electronic state in the conduction band through electron-phonon coupling. This raises the question as to the length a photo should travel through a crystal in order to experience this number of scattering events. This is not the entire story, since in general the amplitude of the nth-order scattering event diminishes for increasing value of n. This is in particular relevant for light whose wavevector is small relative to the extent of the Brillouin zone of a typical insulating crystal, since the contribution of the nth-order scattering process to the scattering cross section scales with the nth power of the amplitude of this wavevector. For the relevant details I refer you to Appendix N of the book Solid State Physics, by Ashcroft and Mermin. Although this appendix concerns scattering of neutrons in crystals, the underlying mathematics is equally applicable to scattering of light in crystals.

The role of temperature is two-fold: in the case of interacting electrons it opens single-particle excitation channels that are not open at zero temperature (for interacting electrons gap in the single-particle spectrum is a zero-temperature characterization), and it changes the distribution of the phonons. For sufficiently large band gaps and sufficiently low temperatures, the contribution of temperature to light scattering in crystals is however exponentially small.

In the light of the above details, for answering the main question on this page one will have to know some characteristics of the problem of interest: the size of the band gap of the crystal used to manufacture the solar cell, the ambient temperature it is operating, the dimensions of the active part of the cell, etc.

PLEASE DEAR SHOW THE LINK

http://www2.lbl.gov/Science-Articles/Archive/MSD-perfect-solar-cell.html

Qualitatively, phonons indeed scatter light, absorbing photon energy in the process. Quantitatively, several issues are to be considered: the energy range of the photons (i.e. the spectrum of the incident light --- any crystal acts like a filter on the spectrum of the light incident on it), the size of the band gap and the temperature. Phonon energies are typically of the order of 10-3 eV (generally, it is important that phonon spectrum in a crystal has a sharp high-energy cut-off, the temperature associated with this cut-off energy being of the order of the Debye temperature), while semi-conducting band gaps are typically of the order of 1 eV. This implies that low-order scattering events do not contribute to the scattering cross section of phonons; on the basis of the above-mentioned energetic considerations one needs of the order of 103 scattering events in order to connect an initial electronic state in the valence band to a final electronic state in the conduction band through electron-phonon coupling. This raises the question as to the length a photo should travel through a crystal in order to experience this number of scattering events. This is not the entire story, since in general the amplitude of the nth-order scattering event diminishes for increasing value of n. This is in particular relevant for light whose wavevector is small relative to the extent of the Brillouin zone of a typical insulating crystal, since the contribution of the nth-order scattering process to the scattering cross section scales with the nth power of the amplitude of this wavevector. For the relevant details I refer you to Appendix N of the book Solid State Physics, by Ashcroft and Mermin. Although this appendix concerns scattering of neutrons in crystals, the underlying mathematics is equally applicable to scattering of light in crystals.

The role of temperature is two-fold: in the case of interacting electrons it opens single-particle excitation channels that are not open at zero temperature (for interacting electrons gap in the single-particle spectrum is a zero-temperature characterization), and it changes the distribution of the phonons. For sufficiently large band gaps and sufficiently low temperatures, the contribution of temperature to light scattering in crystals is however exponentially small.

In the light of the above details, for answering the main question on this page one will have to know some characteristics of the problem of interest: the size of the band gap of the crystal used to manufacture the solar cell, the ambient temperature it is operating, the dimensions of the active part of the cell, etc.

Thank you dr. Behnam for explaining the photon  phonon interaction. 

I would like to add to your answer as an engineer that any semiconductor material has a specific spectral repose to an incident photon flux. This response is defined by the energy band structure of the material. This response is measured by the so called absorption coefficient alpha as a function of the wavelength of the incident photons.

In order to be absorbed, the photon energy must be greater or equal to the minimum energy gap. In direct band gap material, the transition of the electrons from the valence band to the conduction band does not required participation from phonons since the electrons do not need to change their momentum as a direct consequence of direct band gap. In case of indirect band gap where there is a momentum shift between the maximum of the valence band and the minimum of the conduction band, The electron transitions from the valence band to the minimum of the conduction band needs the participation photons.so photons with energy smaller  than the photon energy by the photon energy may be able to affect the transition.If the photon energy is smaller than the the energy gap energy by twice the energy of a photon the probability of transition will be much smaller. This is because as the number of simultaneous quantum mechanical particles colliding with each other increases, the probability of of occurrence decreases appreciably. Accordingly, one considers that the material cuts off for Eph < =Eg. This means th at such photons practically pass the thickness of the wafer without absorption. It is said they will be transmitted and the material behaves consequently as low pass filter.

Thank you for your interesting questions 

wish you success

You are welcome Dr Zekry.

Omar, thanks for interesting question,and I agree 100% with Zerky:

" Accordingly, one considers that the material cuts off for Eph < =Eg. This means that such photons practically pass the thickness of the wafer without absorption"

There are few possible sources of your confusion: (1) band-gap materials- these are semiconductors nanoparticles that change their property with size of NP, so you can virtually tune the bandgap, (2) on the other hand you have a plasmonic phenomena which rise a new interaction of matter with light, and (3) ultimately you can have all kind of supermolecular phenomena in which any system, not necessarily molecular exhibit new features eg: superlattices, q.walls etc.

good luck

Although it is quite small, one might also consider the absorption by the band tails for Eph < Eg.   Google "Urbach tail."  The Urbach energy is related to the disorder in the semiconductor.

For photon energy less than the band gap, direct or indirect band gap point is not relevant. A photon has energy but practically no momentum; a phonon has momentum but practically no energy; so, interaction between photon and phonon is not possible, as both energy and momentum cannot be conserved. If the band gap has trap levels, then, interaction with the trapped electrons is possible, but, the interaction may be very slow, if phonons are required for the conservation of momentum. In device grade semiconductor, most of the incident sub-band-gap photons will be transmitted. 

The photon momentum is really very small and can be neglected in the momentum conservation. On the other side the energy of the  phonons can not be neglected as it is on the average in the order of the thermal energy kT. In room temperature, it amount s to about 26 mev. So, in indirect band semiconductor, where a transition of electron from the top of the valence band to the bottom of the conduction which has an offset in the momentum necessitates the share of phonons to satisfy the momentum conservation process. Accordingly, as the energy is conserved in this triple particle process, photons with energy smaller than Eg with the phonon energy  can share in this process. Because this process is a triple one , it will be less probable than the direct transition process. This explains the smaller absorption coefficient of indirect band gap material near the conduction band edge.For example Si and Ga As.