Device physics

Enlightening solutions

White-light-emitting diodes are becoming increasingly important, but what is the best way to build compact devices possessing high efficiency? Bright prospects are offered by multi-layer organic devices grown from solution.

Sources of white light are found almost everywhere — in lighting and signage generally, but increasingly as backlights in all sorts of displays, for example in laptop computers or smart phones. The development of organic light-emitting diodes (OLEDs) promises further innovation in the field: such diodes are lightweight, provide high brightness at low power, and can be fabricated on flexible substrates to form thin devices at potentially low production cost1. Writing in Advanced Materials, Gong et al.2 demonstrate the use of semiconducting electrolytes to grow highly efficient, multi-layer, white-light OLEDs from solution.

The structure of an OLED is quite simple; it uses an organic material that either fluoresces or phosphoresces. (Both processes involve the re-emission of absorbed light at a longer wavelength; in phosphorescence, the quantum-mechanical processes that lead to re-emission are more complex, so the emission is delayed.) The light-emitting material is sandwiched as a thin film, typically 70–100 nanometres thick, between two electrodes. Of these, the anode is typically transparent and the cathode acts as a mirror that ideally reflects any incident photons back towards the transparent side.

When a voltage is applied to the electrodes, positive and negative charges (‘holes’ and electrons, respectively) are injected into the film and move towards each other, forming a body known as an exciton on meeting. This exciton can become de-excited by emitting a photon, which leaves the device through the transparent anode. Excitons are divided into two categories according to the alignment of the spins of the electrons involved relative to one another: if these are antiparallel, a ‘singlet’ with a total spin of zero is formed; if they are parallel, the state is a ‘triplet’ with a spin of one. In terms of quantum mechanics, three-quarters of excitons must be triplets and only one-quarter singlets. The prospect of achieving higher efficiency with OLEDs emitting from triplets has led to the investment of considerable effort2,3 in their development.

OLEDs are generally produced by one of two routes: the sublimation of small molecules in a vacuum, or the wet-chemical deposition of polymers onto a substrate. Which of these approaches will ultimately survive in a production environment is still a matter of debate. Although the small-molecule approach improved device performance, especially device lifetime, wet-chemical deposition is the more attractive technique for mass production. This is because it allows layers of polymers to be laid down cheaply in a roll-to-roll process using common techniques such as those used for screen and inkjet printing. It is therefore the method of choice for Gong et al.2 and many others.

Multi-layer OLEDs are generally more efficient than single-layer types. In the most efficient OLEDs, the emission layer is sandwiched between a hole-transport layer and an electron-transport layer (Fig. 1). Fabricating such multi-layer structures from solution is challenging (vacuum deposition of small molecules is relatively straightforward). It is crucial to ensure that layers already deposited from solution are totally resistant to the solvents used to deposit subsequent layers, to avoid intermixing.

Figure 1: Cross-section through the multi-layer OLED developed by Gong and colleagues2.

The staggered height of the layers indicates their different energies; electrons (blue) favour moving to lower energies, whereas holes (red) tend towards higher energies. The emissive layer (EML) is sandwiched between a hole-transport layer (HTL) and an electron-transport layer (ETL) consisting of transport agents (HTA/ETA) containing sulphonate groups (SO3). (M+ stands for a metal counter-ion.) This structure serves three purposes: first, to facilitate the injection of holes and electrons (A and C) into the emissive layer by reducing the energetic barriers to their passage; second, to enhance the recombination efficiency (formation of excitons) by blocking the passage of one type of carrier (B or D) from the emissive layer into the opposite transport layer through a large step in energy; and third, to avoid quenching reactions of excitons at the electrodes (+/−). The layers of the OLED are deposited alternately from water or ethanol (HTL and ETL) and from organic solvents (EML). The light emitted through recombination in the EML passes through the transparent anode to the left; the colour of the emission depends on the material or mixture of materials that forms the EML.

There are three main ways to do this. The first is to use ‘orthogonal’ solvents for the individual layers — that is, the solvent used in one deposition does not dissolve any previous layer. For example, the conductive polymer poly(3,4-ethylenedioxy)thiophene (PEDOT), commonly used for OLED anodes, is deposited from an aqueous suspension. After drying, further organic layers can be deposited from typical organic solvents such as toluene without redissolving the PEDOT. A second method is to change the polarity or solubility of the deposited material. An example here is the first luminescent polymer ever discovered, poly(p-phenylenevinylene), or PPV (ref. 4), where a polar sulphonium precursor molecule is transformed by heating into a nonpolar polymer that is insoluble in all organic solvents.

A third, highly attractive approach is to introduce several reactive molecular groups into a semiconductor material. These can be polymerized after deposition to yield totally insoluble crosslinked layers, a process that can, in principle, be repeated indefinitely. In recent years, many materials that can be processed from solution and possess the ability to form multi-layers have been proposed, the most promising being oxetanes5, styrenes6, dienes7 and trifluorovinyl ethers8. The efficiency of OLEDs at incorporating such materials (most of which are hole-transporters) is in many cases greater than that for reference devices using just PEDOT as anode, or the transparent metal indium tin oxide (ITO).

Gong et al.2 propose an extension of the first, orthogonal-solvent approach. They developed derivatives of two commonly used organic semiconducting materials — the hole-conducting poly(N-vinylcarbazole), or PVK, well known from the early days of xerography, and the electron-conducting oxadiazole derivative PBD. The authors achieved this by incorporating into them ionic sulphonate groups, which make the derivatized material soluble in highly polar solvents such as water and ethanol but insoluble in organic solvents. This trick allowed them to create the layers using aqueous or ethanol solution, as the components of the emissive layer of the OLED were totally insoluble in either solvent. (The emissive layer contained a fluorescent polymer emitting green and blue light, doped with a phosphorescent molecule that sends out red light, yielding an overall white emission.)

By alternating deposition from hydrophilic and hydrophobic solvents, Gong and colleagues built a three-layer device (Fig. 1). Because the emissive layer acts as a barrier against the redissolution of the first deposited layer (the hole-transport layer) during the deposition of the third layer (the electron-transport layer), it was of crucial importance that all layers were free of pinholes. The resulting OLEDs produced a luminous intensity of around 10 candelas per ampere of supplied current — around 2.5 lumens per watt. The efficiency of the device is thus among the highest to date for solution-processed white-light-emitting devices1, and one-and-a-half to three times better than reference devices in which either of the two transport layers was missing.

It might be thought that the introduction of sulphonate groups and the consequent presence of mobile metal counter-ions might impair the device's performance. Certainly, devices using transport layers show slightly increased onset voltages — the minimum supply voltage at which emission will occur. According to Gong and colleagues, this is due to the greater thickness of the device compared with other OLEDs and could thus presumably be improved by adjusting the layer thicknesses, or by using different hole- and electron-transport materials containing sulphonate ions or other ionic groups.

The big question, however, is whether devices based on such sulphonate materials can reach the operating lifetimes necessary for practical applications (typically more than 10,000 hours). The presence of mobile metal ions could cause similar problems here to those seen with electrochemical emissive devices. Although such questions remain unresolved, Gong and colleagues' contribution2 is a step towards a more flexible, lower-cost source of white light.


  1. 1

    Andrade, B. W. & Forrest, S. R. Adv. Mater. 16, 1585–1595 (2004).

    Article  Google Scholar 

  2. 2

    Gong, X. et al. Adv. Mater. 17, 2053–2058 (2005).

    CAS  Article  Google Scholar 

  3. 3

    Baldo, M. A. et al. Nature 395, 151–154 (1998).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Burroughes, J. H. et al. Nature 347, 539–541 (1990).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Müller, C. D. et al. Nature 421, 829–833 (2003).

    ADS  Article  Google Scholar 

  6. 6

    Klärner, G. et al. Chem. Mater. 11, 1800–1807 (1999).

    Article  Google Scholar 

  7. 7

    Aldred, M. P. et al. Chem. Mater. 16, 4928–4936 (2004).

    CAS  Article  Google Scholar 

  8. 8

    Gong, X. et al. Appl. Phys. Lett. 83, 183–185 (2003).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Meerholz, K. Enlightening solutions. Nature 437, 327–328 (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.