A biomolecular light-emitting diode

Light-emitting diodes (LEDs) have become key components of numeric and alphanumeric display technology. The most common are p–n-junction diodes fabricated from inorganic semiconductors. In the past decades, nanotechnology advances gave rise to a new type of LED, the organic light-emitting diode (OLED), a prototype of which was first fabricated by Vincett et al.¹ OLED displays with small screens are now commercially available.

In the basic OLED design, a thin film of highly phosphorescent or fluorescent material is sandwiched between various types of metal electrode. These light-emitting materials are also especially designed for efficient charge injection from the electrodes. In this context, questions come to mind with no textbook to provide answers. What kind of organic materials can be used to fabricate OLEDs? Can we make them using common bio-functional materials?

Our work has allowed to provide answers in that we have recently succeeded in fabricating the first biomolecular light-emitting diode (BIODE).² We decided to use metalloporphyrins and related biomolecules—such as cytochrome c,^2–4 myoglobin,³ hemin,⁵ and chlorophyll a.⁶—for this purpose. Cytochrome c and myoglobin are known as heme proteins because their active site consists of an iron-porphyrin, called a heme (shown in Figure 1). In living systems, cytochromes are electron carriers, while myoglobin transports and stores oxygen. Chlorophyll a is the well-known photosynthetic pigment and hemin is the ferric oxidation product of heme. Except for chlorophyll a, these molecules exhibit no photoluminescence. We were, accordingly, greatly surprised to observe current flow and electroluminescence in all BIODE devices fabricated using them.

Figure 2 shows a schematic illustration of a BIODE. The device is a simple sandwich-type junction. We did not prepare carrier-transport layers, as done in most OLEDs, to balance the density of electrons and holes. This is because we wanted to exclude light emission from the transport layers to observe the intrinsic emission of our biomolecules. The observed electroluminescence spectra were broad but consistent with the absorption spectra. The quantum efficiencies of the fabricated devices were 4 × 10^-7 at most (in the case of a hemin BIODE ⁶), thus significantly lower than those of commercially available devices. However, one of the reasons for their small quantum efficiencies was their lack of transport layers.

As for observing electroluminescence in compounds that did not exhibit photoluminescence, our interpretation was to attribute the phenomenon to different excitation states in the electroluminescence and photoluminescence processes. In the case of electroluminescence, carriers are injected directly into d-orbitals whose energy levels are close to those of electrodes. On the other hand, photoexcitation produces electron-hole pairs in the π and π^* orbitals. Therefore, d electrons (or d holes) are hardly formed in photoluminescence, since the probability of d – π and d – π^* transitions are quite small.⁴

Figure 3 shows the drastic changes observed in the electroluminescence spectrum of hemin under applied voltages.⁶ On the basis of magnetic susceptibility and Raman measurements, we concluded that these changes resulted from the high-spin to low-spin transition of the iron. Since the spin state of the heme iron is associated with various bio-functionalities, this phenomenon is of significant interest. Although the transition is irreversible, reversibility may be induced by chemical modification of hemin. Such a transition could be applied to the design of new biosensors.

Biomolecular compounds have a wide range of functionalities. If we can learn to use them in electric devices, we may decrease environmental pollution and realize more eco-friendly technologies for the future. Our BIODE studies are a first step towards such a dream.