You can try to determine serial and shunt resistances of your cell from the I-V curves. This might help to understand something (or might not;)

Low fill factors are generally caused by high series resistance. For example, if the depletion region is increased by decreasing the doping density on the p-type side of the device, this will increase both the bulk resistance of the p-type side and the specific contact resistance of the electrical contact to the p-type side. This will increase the series resistance of the solar cell hence decreasing the fill factor.

Determining series and shunt resistances from IV curves will give you fruitful information. Series resistance and recombination processes in your device, mostly, kill your FF.

Fill factor shows you a degree of " square" or "rectangle" of your I-V curve. In ıdeal case FF=1. So as mentioned above you must eveluate your series and shunts resistace. Because each of them has its own mechanism. So finally you can improve your produtcion process.

A part from series and shunt resistances and of course the conatcts between organics and electrodes, the morphology of P3HT/PCBM heterojunction should play an important role in determining FF.

The fill factor , more commonly known by its abbreviation "FF", is a parameter which, in conjunction with Voc and Isc, determines the maximum power from a solar cell. The FF is defined as the ratio of the maximum power from the solar cell to the product of Voc and Isc. Graphically, the FF is a measure of the "squareness" of the solar cell and is also the area of the largest rectangle which will fit in the IV curve.

http://pveducation.org/pvcdrom/solar-cell-operation/fill-factor

To: Antonio Papagni.

Antonio, could you kindly explain, how morphology can change FF if it does not affect neither shunt nor serial resistance, please? I am trying to clarify this for my own sake for some time already, but no luck... May be, you have an example of such a morphology? All models I can imagine by myself, include some non-negligible resistance, at least one of the two...

No kidding, it will be highly appreciated.

In principle Isc is related also to the quality of heterojunction or how intimate is. In other words PCBM tends to segregate from P3HT polymer thus both charge percoling pathways (correlated to resistances) and of course the number of extitons that are able to reach the polymer/PCBM interface are affected with the result of reducing Isc and consequently the FF. Thermal annealings can stimulated PCBM segregation. The segregation speed is also related to P3HT/PCBM ratio, the solvent use for spincasting and annealing temperatures

The problem is that "everything" influences FF. Anything that changes charge transport or charge recombination will probably change the FF as well (mobility, domain sizes, geminate and non-geminate charge recombination, space charge near an electrode, etc. ). You say your "FF is low, but the rest is OK". What is your reference? (the MoO3-free device?) What do you see in the dark? (this often helps to differentiate between effects that include excitons and those that are purely determined by charge transport..).

@Vladimir: As far as I understand, Antonio did NOT say that the morphology will change the FF WITHOUT changing the extracted series or shunt resistance. I would expect the morphology to influence at least one of these.

Dear Marten, many thanks for your answer.

Poor fill factor is typically indicative of non-ideal morphology in the active layer. In other words, your P3HT:PCBM blend might be the problem.

It seems an issue of inefficient morphology. May I ask which type of treatment did you use, thermal annealing?

Three isssues! (1)ZnO/P3HT:PCBM interface is not Ohmic contact, resulting high interfacial resistance;(2) the bulk junction layer of P3HT:PCBM may not be in the ideal conductions, leading poor electron-hole seperation;(3) the layer of MoO3 is too thick.

It was proved that the thickness of MoO3 has no significant effect on efficiency, but to make it too thick is unnecessary also.

FF of any device is affected by its series and shunt resistances values. A major source of series resistance could be interface resistance between external contacts and active organic material responsible for absoption of light, exciton generation and transportation to front and rear contacts of the device.

Thanks everybody for your answers and discussion. Actually what made me confused is that I indeed achieved high FF (about 64% on average) in one batch of devices. Almost all devices had shown high FF; but later I could not reproduce it.

I tried three different thicknesses of MoO3: 7, 10 and 14 nm and found it didn't matter much in terms of FF.

Now there are two more possibilities as some of you mentioned above: ZnO/P3HT-PCBM interface and P3HT-PCBM morphology.

And also sometimes, I find 'S' shape in the I-V curve. Any explanations on that?

Sanjib, do you mean S or N shape in an IV curve, when V is ascissa and I is orditate? S shape is something unusual in these coordinates IMHO...

The S-shape can be indicative of problems with one of the electrodes (reduced surface recombination -> space charge -> S-shape. Look for example at the paper of A.Wagenpfahl, "S-shaped current-voltage characteristics of organic solar devices").

I would suggest you swap each electrode (one by one) with a more standard one and observe if the S-shape vanishes. Standard would be ITO/PEDOT:PSS for the anode and Al (or Ca/Al) for the cathode. This should tell you which/whether your electrodes are the problem, or if it has something to do with the P3HT:PCBM.

@Marten: Wagenpfahl et al. are discussing the effect using the terms established previously and corresponding to dependences of voltage from current. Look at their curves, they are actually N-shaped as presented, in the coordinates more common now. Pity they do not mention the origin of the term... Or did I miss it? I only browsed through it fast.

@Sanjib: I'd suggest few nanometers of Alq3 on top of organics.

@Vladimir I guess it depends on the background which coordinates are more common? ;0) For me, it is common to put the voltage on the x-axis and current on the y-axis. I am not sure I ever saw it the other way around?

And in those coords with voltage on the x-axis, it does look a bit like an "S" if you only look at the quadrant with the maximum power point. You also have to start with a negative current and then going upwards to see the shape as an "S".

@Marten: Yes, coordinates change with the scope of interests. Bu I did not see recently the current on the horizontal axis with the voltage on the vertical one, only in very old books on radio-electronics... Though regularly I see all combinations for positive directions...

And yes, you are right, I can see the S-shaped part on their curves now. But the enlarged area of interest on Fig.2 and the description of the phenomenon (cite: "a device in conducting direction shows a local saturation and a later again increasing current around at a certain applied voltage region") is about the other part, at zero current, isn't that N?

I tried to find anything old-style -- nothing comes out. Am I too old with those died-out memories?! (I "new" what to look for on their curves...)

;-)

Thanks!

I think all explanations are good and more detail can also be find in Adv. Energy Mater., 2013, 3, 873–880, Adv. Energy Mater., 2011, 1, 352-355, and J. Appl. Phys. 2008, 104, 043107(1-6). This is also due to the misalignment of the energy levels of hole transport layer and your donor/acceptor materials. In general it is very difficult to get a clear explanation for lower FF or higher FF. In general we can say the morphology is not good for efficient charge transport/separation. For more detail you can look at the above articles.

@Marten: Thank you. Yes, I will try to make solar cells with regular structure, ITO/PEDOT:PSS/P3HT:PCBM/Ca/Al to see if it's because of the morphology.

@All: Actually, I am now more concerned about inverted structure. At first, I thought the low FF was because of MoO3 layer and should wait for some time after MoO3 deposition to start Ag deposition. But later, I couldn't reproduce what I got once. Now it looks like not because of MoO3 layer.

Dear Sanjib

You have to have a look at the dark characteristics of your device also and compare the magnitude of dark current with the short circuit current density. If your dark current magnitude after the open circuit voltage is lower than short circuit current density then you likely to get the sigmoidal kind of I-V characteristics. I hope you are getting some kind of curve which I have attached here. For better explanation on the curve you may refer to Prof. M. A. Alam Lecture

http://www.youtube.com/watch?v=HL4jpfeJNBE&list=PLQm9a6A7ufY48wRLBAE0BWAjbleboGD73&index=4

Are you using fresh MoO3 every time? Hot MoO3 will lose oxygen in a vacuum. I have found that MoO3 deposited by thermal evaporation is really MoOx where X is close to 3 for only about the first 5 nm. Additionally, if you try to reuse the source material for a second or third deposition you are actually starting with something like MoO2 or MoO instead of MoO3. The structure I use is Ito/ZnO(sol-gel)/P3HT:PCBM/MoO3(5nm)/Al(60nm) and get get fill factors just above 70%. Thicker layers of MoO3 have a detrimental effect.

The maximum power generated by a solar cell is dependent on the fill factor, FF. In case of a solar cell that behaves as an ideal diode only direct recombination occurs and the maximal FF is a function of Voc. In a practical solar cell the FF is lower than the ideal value due to following reasons:

• The voltage drop due to the series resistance Rs of a solar cell. The series resistance is introduced by the resistance of the main current path through which the photo-generated carriers arrive to the external circuit. The contributions to the series resistance come from the bulk resistance of the junction, the contact resistance between the junction and electrodes, the resistance of the electrodes.

• The voltage drop due to the leakage current and characterized by the shunt resistance Rp of a solar cell. The leakage current is caused by the current through local defects in the junction or due to the shunts at the edges of solar cells.

• The recombination in a non-ideal solar cell results in a decrease of the FF.

http://www.aerostudents.com/files/solarCells/CH5SolarCellConversionEfficiencyLimits.pdf

In the case of your cell there are many interfaces. Revealing each electrical contribution needs impedance analyses. Refer to Electrochimica Acta vol. 55 pp. 6849-6856 (2010)

http://www.sciencedirect.com/science/article/pii/S0013468610007838

Dear colleagues, Dear Das,

Please allow me to share in your discussion about the fill factor of the solar cells. The fill factor is defined by FF= Pm/ (Voc Isc), where Pm is the maximum output power, Voc is the open circuit voltage and Isc is the shortcircuit current. It is a measure of the squareness of the I-V characteristics of the solar cell. The factors affecting the fill factors are:

- The series resistance of the solar cell,Rs

-The parallel resistance of the solar cell.Rsh

-The recombination current in the space charge region of the cell.and

- The reverse saturation current of the junction I0

The series resistance is composed of the resistance of the different semiconductor layers of the cell in addition to the metal semiconductor contacts,the MS contacts.Because your structure is a thin film structure,its layer resistance is relatively high. But the more serious may be the MS contact resistance.

-The shunt resistance arises because of the leakage across the junction.

-If the recombination rate high and the carrier lifetime is relatively the recombination in the space charge region SCR may overweight the recombination in the regions outside the SCR reducing the ideality factor of the diode and consequently reducing the squareness of the I-V characteristics and the FF.

-AS the reverse saturation current decreases the squareness of the I-V curve is enhanced and consequently the FF increases.

So, you must measure these parameters in order to explore the observed value of the FF of your cells.

The maximum FF is achieved with:

Rs=zero,

Rsh= infinity,

and the ideality factor=1.

For more details see the paper: A. Zekry and G. El-Dllal, “effect of the MS contact on the electrical behavior of solar cells”, Journal of solid-state Electronics, Vol.31, No.1, 1988.,1.48 Impact Factor

Thank you

I have the solution for you, since I have the same problem, in fact, you must reduce the thickness of MoO3 (5 nm, by thermal evaporation, 0.5A ° / s), and also reduce the rate of deposition of Ag 1 nm / s (deposited by thermal evaporation)

Dears, I have a similar problem in organic solar cells. ITO/ ZnO cause either low short circuit or high leakage current. Do anybody knows why that could happen for the inverted stack using ZnO?

Even if the series resistance RS= zero and the Shunt Resistance Rsh is infinity, the fill factor will be still less than one because of the recombination of the photo generated electron hole pairs in the solar cell diode. ... As n increases the curve will be less square and the fill factor decreases.

The fill factor of organic solar cells can be limited by several factors: 1. Field-dependent geminate recombination, or in other words field-dependent photo-generation rate of free charge carriers. 2. Field-dependent extraction of free charge carriers. 3. Large series resistance. 4. Ideality coefficient of dark JV curve larger than unity. 5. Poor quality ohmic contacts with a potential barrier at the active layer/contact interface (so-called S-shape JV-curve).