The difference in thickness can be correlated to the mobility, lifetime and diffusion lengths of minority carriers. In silicon solar cell the minority carriers on p-side are electrons and on n-side these are holes. Since the electrons have a higher mobility, lifetime and diffusion lengths than holes, so the e-h pair generated in a thick p-layer after photon interactions will have greater possibility of collection across the load terminals. Of course the description given above is debatable, because efficient solar cells have been manufactured with thick n-layer also. Please refer to the attachments.

Due to general photovoltaic theory; light generated photocurrent (short circuit current) directly related created electron-hole pair number and it is well known that created e-h pairs directly related with captured number of photons. This process runs in the depletion width of the p/n junction. It is widely known that during the solar cell fabrication n-type material is chosen from large band –gap semiconductor materials (window layer) and p-type material is chosen from lower band–gap semiconductor materials (absorber layer). Because all photons should be collected into depletion layer. Therefore, window layer (n-type semiconductor) should be transparent to let all photons have to move thoughts the p layer. Strong absorption occurs in the absorption layer. On the other hand number of photons roughly equals to light intensity (remember that internal quantum efficiency). In addition, penetration depth of the light intensity in n or p type materials obeys the Beer–Lambert law. If you demand more photocurrent, you should keep your semiconductor material as thinner as possible Because of according to peer lamberts law penetration depth of the material is directly related to absorption coefficient and band gap of the materials.

The reason for one thin layer and one thick layer is that diffusion is performed upon the thick layer to fabricate a p-n junction. Shallow junction would help with carrier collection as carriers generated in the thin layer don't have to diffuse too long before they can be collected by the junction (I am talking about conventional front junction solar cells). Besides, shallow junctions are easier to achieve as they require less diffusion time, which is industrially more viable.

The reason for using p-type substrate is because phosphorous diffusion is easier compared with boron diffusion and the minority carrier in p-type base is electron, which features a higher mobility. However, n-type substrate is more tolerant to chemical and crystallographic defects, which implies high potential to achieve high efficiency solar cells such as Sunpower IBC (Lab) and Yingli Panda (Industrial).

I don't agree with Abdulkadir on "If you demand more photocurrent, you should keep your semiconductor material as thinner as possible Because of according to peer lamberts law penetration depth of the material is directly related to absorption coefficient and band gap of the materials." because too thin substrates would lead to the loss of long wavelength photon absorption, which reduces short-circuit current (bad QE at long wavelength).

I also don't agree with Abdelhalim on "The n+ top layer acts as an emitter layer. For optimum functioning , it must be heavily doped and thin to have a very small revere saturation current and decrease the absorption of the ultraviolet photons in it since it has a dead layer near its surface." as dead layer is caused by heavily doped emitter. Indeed, heavily doped emitter would reduce series resistance yet it causes the loss of short wavelength photon absorption (bad QE at short wavelength). A good idea to achieve both high QE at short wavelength and low series resistance is to use selective emitter, where you only doped the emitter underneath the metal contact heavily and leave other emitter regions lightly doped. This can be achieved using spin-on-dopant and subsequent laser doping, which UNSW has carried out a lot of research on.

First of all I would like to thank "Abdelhalim Zekry " for useful add-ons comments about the issue. And I certainly I agree with him. I also would like to stress that we have good background knowledge about historically improvement for conventional silicon solar cell structure. So we mention about the conventional experimental/theoretical observation about silicon solar cell device. Therefore all discussion is valid as long as we have “Si” based heterjunction solar cell.

Therefore, all questions should be considered valid as long as they meet the criteria.

The penetration depth of the short-wave-length photons is lower then long-wave-length photons so if you demand high penetration depth at short-wave-length radiation such as U.V then you have to tune your band gap from lower-band gap semiconductor material (absorber layer) Because it is well known that absorption spectrum is generally related with band-gap of the material. Α=sqrt(E-Egap)

This mathematical relation tells that there is not significant absorption below Egap value radiation. However above Egap value there will be significant absorption.

mostly the base is P-type, in order to acquire n-type layer, you have to add more impurities.

Thanks  for all

Since electron mobility is greater (roughly 3 times) than hole mobility, we prefer to have p layer thicker than n layer so that the equal number of charge carriers (electron and hole) could reach at the opposite electrodes in almost equal time without getting recombined. As this is highly important for obtaining maximum current.

All the above discussion is very useful and scientific. Here i want to add one more point...The purpose to make the width of  N-region is smaller is that, its width  would become less than the diffusion length of holes. It means that holes in  N-region (minority carriers) can easily diffuse to metal electrode before recombining. 

You can read the books

"Photovoltaics: Fundamentals, Technology and Practice

Konrad Mertens

ISBN: 978-1-118-63416-5" 

or 

"Crystalline silicon solar cells 

Martin Green "

for detailed answer. 

Your question is problematic. Either of the type might be thicker than the other. In the solar cell industry, most of them are p-type si wafer based technology, but the panda produced by YINGLI Solar is n-type wafer based technology, which has several benefit rather than p-type.

Just to add on to the answers.Normally the n-layer is heavily doped which creates a wide depletion region located most entirely in the p-layer. It is important to note that not just the n-layer is made thin but all the layers that come above the p-layer (i.e n-layer, transparent electrode, anti-reflection layer, SiO2 etc) .One reason is to ensure that the photons can penetrate the layers without getting absorbed much and that there is a large depletion region in which the photons can be absorbed and converted to electron/hole pairs which immediately get separated in the built-in electric field.

I have seen some n-type solar cells and they have a similar arrangement like the p-type based ones.(thicker n-layer and thinner,heavily doped p-layer)