Semiconductor technology

Negatively successful

Organic semiconducting polymers are promising electronic materials, but for full versatility they need to conduct negative as well as positive charge. A step towards that goal has now been taken.

Electronic devices based on organic materials are gradually becoming a mainstay of semiconductor technology, and are already used, for example, in flexible displays. However, a few hurdles remain before a breakthrough in practical applications can occur. One nagging problem is that organic transistors are so far only efficient at carrying electronic current via positive charge carriers (holes). To realize a complete electronic circuit based on organic polymers, analogous to conventional silicon chips, devices must also be able to carry negative charge (electrons). On page 194 of this issue, Chua et al.1 demonstrate a design principle that would turn any organic field-effect transistor (FET) into one that efficiently carries negative charge.

The past decade has seen many efforts to develop so-called ‘n-channel’ organic FETs, in which electrons, rather than holes, are the charge carriers2,3,4,5,6. This work is not aimed solely at immediate applications, but also at achieving a better understanding of the relation between materials and device performance.

The guiding principle in the design of semiconductor and electrode materials for n-channel FETs has been to optimize energy levels so as to achieve high electron mobility. But compared with ‘p-channel’ FETs (where the charge carriers are holes), such devices are generally found to be more sensitive to moisture and oxygen. To combat this problem, organic materials have been developed to have high electron affinity, which means that they strongly attract electrons and are not easily oxidized. However, electron affinity cannot be too high, or the material will become prone to unwanted doping, which is detrimental to the performance of a transistor.

An example of an organic semiconductor material based on this design principle of optimizing energy levels is a pair of materials known as F15NTCDI and H15NTCDI. Although the fluorine-free material has good electron mobility, transistors made with just this compound operate only under vacuum, where the effects of moisture and oxygen are much reduced. The fluorine-containing compound, on the other hand, has a higher electron affinity and can operate in air2.

Ultimately, materials are needed that are ‘ambipolar’ — that is, that can be used to create both n-channel and p-channel FETs. Until now, materials for making organic and polymer transistors have generally been lumped into one of the two categories, either n-channel or p-channel7. Traditional organic FET materials, such as pentacene, α-sexithiophene and polythiophenes, are known as p-channel materials and have been difficult to coerce into taking on n-channel behaviour.

However, experimental results suggest that the difficulty in observing efficient transport of electrons in an organic FET is an extrinsic effect caused by factors other than the organic semiconductor material itself. This conclusion would be in tune with the observation that, in purified single crystals of small organic semiconducting molecules, both electrons and holes move with roughly comparable mobilities.

For example, a traditional p-channel FET — based on pentacene — has been converted into an n-channel device by inserting calcium at the interface between pentacene and the thin insulating layer used in transistors8. The conclusion was that electrons from the calcium atoms fill up a large number of the ‘traps’ in the insulator that inhibit electron transport. Removal of these traps results in efficient n-channel behaviour.

Chua et al.1 make a convincing case that the trapping of electrons at the insulator–semiconductor interface is indeed the culprit, and they relate this trapping to electronegative hydroxyl (OH) groups in the insulator material. When they use materials that are free of hydroxyl groups, uninhibited electron transport is indeed observed. By using bisbenzocyclobutene as the insulator, for example, organic polymer FETs based on materials such as poly(fluorine-alt-bithiophene) and poly(fluorine-alt-benzo-thiadiazole) are shown to exhibit n-channel behaviour.

Chua et al. conclude that if the trapping of electrons by electronegative groups in the insulating layer could be avoided, then n-channel behaviour would be seen in a broad range of semiconductors. Their most convincing example is that of polythiophene, which is relatively prone to oxidization and therefore normally more difficult to use in n-channel FETs.

Of course, it remains to be seen whether, on prolonged exposure of the device to ambient conditions, the effects of atmospheric moisture and oxygen would negate those of a hydroxyl-free insulator material. But the finding is nevertheless a remarkable result and a major step forward in our understanding of the design principles of materials and devices for n-channel organic transistors.

Chua and colleagues' work also explains why some semiconductors transport both electrons and holes when used as an active layer in light-emitting diodes — where there are no insulating interfaces — but exhibit only n-channel behaviour in FETs9. The implication is that such materials, which readily exhibit n-channel behaviour but not p-channel behaviour, could also be made ambipolar by identifying what causes hole trapping at the semiconductor–insulator interface and eliminating it.

Another area that will benefit greatly from this study is the work on ambipolar organic FETs with a view to making efficient light-emitting transistors10. A wider choice of materials for this type of research will become available once electron trapping is eliminated or greatly reduced.


  1. 1

    Chua, L. -L. et al. Nature 434, 194–199 (2005).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Katz, H. E. et al. Nature 404, 478–481 (2000).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Facchetti, A. et al. J. Am. Chem. Soc. 126, 13859–13874 (2004).

    CAS  Article  Google Scholar 

  4. 4

    Bao, Z. Adv. Mater. 12, 227–230 (2000).

    CAS  Article  Google Scholar 

  5. 5

    Chesterfield, R. J. et al. J. Phys. Chem. B 108, 19281–19292 (2004).

    CAS  Article  Google Scholar 

  6. 6

    Crone, B. et al. Nature 403, 521–523 (2000).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Dimitrakopoulos, C. D. & Malenfant, P. R. L. Adv. Mater. 14, 99–117 (2002).

    CAS  Article  Google Scholar 

  8. 8

    Ahles, M., Schmechel, R. & Von Seggern, H. Appl. Phys. Lett. 85, 4499–4501 (2004).

    ADS  CAS  Article  Google Scholar 

  9. 9

    Ostrick, J. et al. J. Appl. Phys. 81, 6804–6808 (1997).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Ahles, M., Hepp, A., Schmechel, R. & Von Seggern, H. Appl. Phys. Lett. 84, 428–430 (2004).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Dodabalapur, A. Negatively successful. Nature 434, 151–152 (2005).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.