My former colleagues at Max Planck Institute of Microstructure Physics in Halle, Germany, are "shunt experts". Therefore I recommend to have a look at the following publications (especially the one from 2004, "Shunt types in crystalline silicon solar cells", and from 2007, "Electronic activity of SiC precipitates in multicrystalline solar silicon"):

https://www.researchgate.net/publication/3893688 (PDF file included)

https://www.researchgate.net/publication/222361822 (PDF file available: http://worldlasers.com/articles/research/sdarticlezz.pdf)

https://www.researchgate.net/publication/222563347 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2002/abs_pdf/1539_02.html)

https://www.researchgate.net/publication/227528398 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2004/abs_pdf/4554_04.html)

https://www.researchgate.net/publication/227675525 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2007/abs_pdf/7455_07.html)

https://www.researchgate.net/publication/258316780 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2008/abs_pdf/7920_08.html)

https://www.researchgate.net/publication/258323227 (PDF file included)

https://www.researchgate.net/publication/280914402 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2015/abs_pdf/12545_15.html)

Conference Paper Localization of shunts across the floating junction of DSBC ...

Article Shunts due to laser scribing of solar cells evaluated by hig...

Article Classification of shunting mechanisms in crystalline silicon...

Article Shunt types in crystalline silicon solar cells

Article Electronic activity of SiC precipitates in multicrystalline ...

Article On The Detection of Shunts in Silicon Solar Cells by Photo-a...

Article Influence of Defects on Solar Cell Characteristics

Article Nanoscopic studies of 2D-extended defects in silicon that ca...

Well, all the phenomenons that can shunt the main junction or make a path between the front and the rear side. For exemple, you can have aluminum spiking through the main junction, you can have bad edge isolation (if the cell is small) and so on

Shunt  resistance is due to non-idealities, mainly impurities near the p-n junction, which cause partial shorting of the junction. That is, if an impurity can join the front and back sides of the solar cell then it is responsible for shunt resistance.

hi, if you look at the equivalent circuit model of a solar cell, you will see that the shunt "resistor" is placed between the 2 poles... so, any effect that will reduce (or increase) such resistivity (as the ones mentioned above for the reduction) will have an impact on the overall shunt resistance of the solar cell. Clearly, the shunt resistance is a combination (integral) of such effects.

Dear Poornima

Shunt resistance related for dark current for device and this mean structure of device, and when it increase this mean give high value for open circuit voltage (Voc), also this mean the efficiency of solar cell will increase. One of the parameters which effect on shunt resistance can be pinholes which can be observed by SEM technique. 

With my best regards

My former colleagues at Max Planck Institute of Microstructure Physics in Halle, Germany, are "shunt experts". Therefore I recommend to have a look at the following publications (especially the one from 2004, "Shunt types in crystalline silicon solar cells", and from 2007, "Electronic activity of SiC precipitates in multicrystalline solar silicon"):

https://www.researchgate.net/publication/3893688 (PDF file included)

https://www.researchgate.net/publication/222361822 (PDF file available: http://worldlasers.com/articles/research/sdarticlezz.pdf)

https://www.researchgate.net/publication/222563347 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2002/abs_pdf/1539_02.html)

https://www.researchgate.net/publication/227528398 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2004/abs_pdf/4554_04.html)

https://www.researchgate.net/publication/227675525 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2007/abs_pdf/7455_07.html)

https://www.researchgate.net/publication/258316780 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2008/abs_pdf/7920_08.html)

https://www.researchgate.net/publication/258323227 (PDF file included)

https://www.researchgate.net/publication/280914402 (PDF file available at http://www-old.mpi-halle.mpg.de/mpi/publi/go/2015/abs_pdf/12545_15.html)

Conference Paper Localization of shunts across the floating junction of DSBC ...

Article Shunts due to laser scribing of solar cells evaluated by hig...

Article Classification of shunting mechanisms in crystalline silicon...

Article Shunt types in crystalline silicon solar cells

Article Electronic activity of SiC precipitates in multicrystalline ...

Article On The Detection of Shunts in Silicon Solar Cells by Photo-a...

Article Influence of Defects on Solar Cell Characteristics

Article Nanoscopic studies of 2D-extended defects in silicon that ca...

actually during manufacturing process small impurities come in between the bandgap such as iron piece etc which cause  trap assistant recombination so from it open circuit voltage also effected and second thing these impurieties give another path for output current thats why we represent it in figure as shunt resistant higher the shunt mean lower impurities between bandgap.

(i) the unpassivated edges of the junction

(ii) unexpected diffusion of dopants or other impurities through the space charge region of the junction

(iii) extended crystallographic defects like dislocations, decorated twins and grain boundaries croosing the junction

(iv) poor doping level of the emitter

I cam late to this question but as this question is a basic one and other researchers may benefit from my answer , i would like to add to what has been said by the other colleagues above: The causes of the shunt resistance are deficient fabrication and production processes and surface leakage currents.

- The ideal junctions has only a current equal to the reverse saturation current which is extremely low.

- The presence of pinholes in the emitter layer between the substrate and the metallization layer leads to direct metal short of the junction by the reach through of the metal to the substrate. This is the most serious short.

This type of short is common in the thin film technology solar cells.

- Also shorts through the surface can be formed if the upper and lower metals meet or if the top and bottom heavily doped layers meet.

- If the material is poly crystalline and the material is doped by diffusion, the dopant atoms can diffuse with high speed along the grain boundaries causing low ohmic pates across the junction. This phenomenon is called preferential doping.

- Metal atoms can diffuse to the junction and cause recombination centers in the space charge region and higher current than it would be.

For the model of such effects please see the link: Article Capacitance and conductance of ZnxCd1-xS/ZnTe heterojunctions

Best wishes

In short, any mechanism that can bring a photo-excited electron from the conduction band back to the valence band without traveling through the external circuit is shunt. In this sense, radiative recombination has the effect of shunt (normally included in I0), and of course, and recombination through defects is shunt. Shunt resistance is typically introduced phenomenologically in a device model to take care of many contributions.

Well, although some people follow this "generalized shunt approach", there is a problem with such a notion. Look at the standard simple one-diode equivalent circuit of a solar cell: Shunts are represented by an ohmic resistor located (modulo series resistance) between the two terminals of the solar cell. So, the current flowing through the shunt (i) shows a linear current--voltage characteristic and (ii) consists of majority carriers only. Therefore, the recombination of minority carriers is not contained in the standard shunt resistance model. (Note that the starting question of this discussion reads "What are the causes of the shunt resistance formation in solar cells?")

Of course, in the "generalized shunt approach", the notion of a "nonlinear shunt" is introduced. But I'm not sure whether that makes things easier or clearer.

Dear Jan-Martin,

Thank you for your comments. I agree with that “generalized shunt approach” does not necessarily make things easier. However, I feel it offers a clearer physical picture of shunt.

The popular formula with (V+IRs)/Rsh term is basically a phenomenological 1-D model. The meaning of Rs and Rsh are vague, i.e., each may contains multiple contributions of very different in nature. It often can fit the I-V curve very well, but after all one is fitting a relatively simple curve with many parameters, not to mention more parameters are used for a two diode model.

People do realize the simple model is not ideal, so 3D models have been developed. It is probably impractical to formulate a realistic 3D model into the simple formula with an effective Rsh and Rs. Regarding shunt, I would say the effects of shunt are reflected in both I0 and Rsh. I0 is known to involve material properties such as diffusion length or carrier lifetime. Defects of course affect I0, but the effects are not fully captured in I0, because the diode equation is overly simplified for the real device. For instance, taking two solar cells made out of two normally identical materials but introducing some dislocations in one of them, one will find that the device with the defects yields a smaller shunt resistance when fitted to the diode formula. One would say these dislocations cause shunt. Similarly, a large number of point defects will have the same effect, so the point defects cause shunt. When one is measuring the I-V curve, one cannot really tell the underlying mechanism of the shunt. In this sense, the “generalized shunt approach” is meaningful, because it lumps the contributions of multiple mechanisms in it.

Dear Yong,

I fully agree with what you said; this gives the right picture! It's just that these considerations of yours are, so to speak, on the "next-higher level" (at least), while I had the lower level of sophistication in mind. But of course your considerations are highly relevant, because shunts (in the generalized sense, i.e. both linear and nonlinear ones) typically are localized phenomena, so using the average series resistance to describe the voltage relevant for the shunt current is not always correct; only if there are several shunts that experience different local series resistance values so that an average description is sensible, it might be correct to use the average series resistance.

Actually, I'm not aware of a paper where this is already treated systematically -- leaving full 3D models apart, of course; I'm just interested, so to speak, in an "intermediate-level description".

Dear Jan-Martin,

Thank you for your kind comments. Perhaps someone could use a somewhat simplified device structure to derive an I-V equation resembling the popular formula to reveal the physics more clearly.