In physics, the Shockley–Queisser limit (published in 1961) refers to the maximum theoretical efficiency of a solar cell. Actually, the Wikipedia pagesays it all, as well as this Solar Central article. There are some standard assumptions (how strong is solar radiation etc.).
For ideal solar cells, the maximum theoretical efficiency is circa 34% when those assumptions apply. That's because only some photons have the right wavelength to free an electron. Shorter photons dissipate the excess energy and heat the material and longer photons just pass through the "trap". It means that from solar energy of 1000 W/m2 only less than 340 W/m2 converts into electricity. The rest 67% turns into heat, or the photons pass through without conversion.
In real cells, there are also some losses, because some light is reflected from the cell and some energy is lost due to resistance on contacts or due to impurities. In real solar panels is the best efficiency a bit above 20 % - not even close to the theoretical 34 %.
Well, but there are some solar cells with much higher efficiency, so how can it be? They use some workarounds! For example, you can stack another cell behind the first one and catch part of the photons that pass through the first layer. This is called tandem-cell. Every layer can be made from different semiconductor material, tuned for the best wavelength. This can improve the efficiency of practical solar cells well above 40%.
Another interesting technique is concentration of the solar radiation. More light makes more energy. And there are more tricks in the hat. You can guess that those special techniques have high cost and for the near future it won't give us more power from the same square meter of solar panel.