Background
Fig.1 Schematic image of hybrid optoelectronic interface concept.
Context and opportunity
The potential of active organic communications components, including organic semiconductor lasers (OSL’s) and amplifiers, has been very widely recognised, and the development of such devices is currently the subject of intensive worldwide interest. Significant and widespread progress has been made using optical pumping. To date though, despite such progress, there has been no demonstration of electrically driven lasing in organic semiconductors. Both charge-induced absorption and field quenching are currently recognised as major obstacles to organic injection lasing. The key challenge facing the community now, therefore, is how to develop high-performance, integrated and robust formats of organic optoelectronic devices suited to the host of potential applications that have been identified for this technology.
Here, we propose a specific and very timely solution to this challenge, one that makes a virtueof the optical pumping requirement and achieves electrical pumping indirectly. Our approach is to fully hybridise organic semiconductor optoelectronics with inorganic complementary-metal-oxide-silicon (CMOS) based electronic control circuitry by using intermediate gallium nitride optoelectronics. We recognise a ready compatibility between these hitherto largely disparate technologies:
- The ultraviolet to short-wavelength-visible outputs of gallium nitride light-emitters are directly compatible with the absorption characteristics of the major categories of organic semiconductor gain media.
- CMOS drive voltages, circuit layouts and drive formats can directly match the capabilities and requirements of gallium nitride optoelectronics and can address these in a ‘flip-chip’ format.
- CMOS on-chip photo-detectors show peak spectral sensitivity in the wavelength range of gallium nitride and organic light emitters and have time-resolution sufficient to fully-resolve pulsed and modulated outputs from these devices.
- Organic semiconductor materials are very versatile in their processing and can conform to a wide range of inorganic structures and surfaces. They show strong absorption and gain under pumping at normal incidence to the organic films.
- These technologies are effectively planar and hierarchical, offering clear routes to overall integratio
These factors combine with two further very important points to underpin our vision:
- High-performance organic lasers have recently been produced with threshold optical pump requirements accessible to gallium nitride light-emitting diodes (LED’s), thus obviating the many limitations (edge emission; limited custom-availability, - performance, - wavelengths, - format) that the forced use of gallium nitride lasers would have imposed on an approach such as ours.
- In the micro-pixellated GaN LED technology we have been developing for several years now (see http://bt-onethousand.photonics.ac.uk ), we have an ideal, planar format of GaN optoelectronics to facilitate interfacing between CMOS and the organics.
Our overall picture, then (Fig.1), is of a new family of highly-parallel, hybrid optoelectronic ‘information exchange’ interfaces. These are essentially arrays of smart pixels, consisting of CMOS electronics with on-chip photo-detection, intermediate gallium nitride micro-pixel LED’s in a range of wavelengths, layouts and formats, and organic lasers and amplifier elements in registry with these pump sources. Our approach facilitates not only integrated optical pumping in the soft materials, but also confers a number of benefits in the way the organic elements are fabricated, addressed and controlled. This potentially includes self-aligned photo-forming of the organic micro/nano-structures and pattern programmability of the emitter arrays via the optical interface. Furthermore, the outputs and performance of the GaN and organic devices can be monitored by high speed photo-detection on the CMOS chip, offering utilisation of a range of active feedback and control mechanisms. We see two major formats of organic devices based on this approach. The first are organic planar light-wave circuits (PLC’s), where linked organic components (lasers, amplifiers) are laid out above the gallium nitride interface. The second is where the organic elements are controlled discretely, and their output taken vertically from the chip for free-space optical projection and interconnects.
We believe that our approach is entirely complementary to all-inorganic optoelectronic interfaces operating in the near-infrared. We have a laser technology offering full coverage of the visible spectrum, very well suited to wavelength tuning and fast digital operation, which will be integrated with sub-ns and single-photon-sensitive on-chip detection. We forsee many opportunities for such a technology in areas including bio-instrumentation, bio-computing, optical radar, quantum information and quantum computation. We note that many of the atomic/ionic/molecular/colour centre/nanocrystal structures being explored for the latter have absorptions in the visible.
- Organic lasers and amplifiers: status and research challenges
A recent invited review by Samuel and Turnbull (Chem. Rev., 107, 1272 (2007)) summarises the state-of-the-art in organic semiconductor lasers (OSL’s) and amplifiers. As laser gain media, organic semiconductors have many distinctive features. They exhibit very strong absorption bands, with peak extinction coefficients ~10 5 cm -1, and large optical gain cross-sections (10 -15 cm 2), giving the possibility of substantial gain in very compact devices and enabling very simple optically side-pumped geometries. Typically, they have emission bandwidths of >100 nm; while through changes in chemical structure, the emission may be tuned from ~400 to 700 nm. Consequently, organic semiconductors are well suited to applications as tuneable lasers, ultrafast lasers or high-gain broadband amplifiers. Of further importance is their conformability and simple solution processing. Many different micro-resonator configurations have been demonstrated, including micro-cavities, micro-rings, distributed feedback (DFB) and photonic crystal lasers. Wavelength-scale structures for active photonic crystals can readily be moulded into polymer films using a variety of soft lithographic techniques.
The ‘indirect electrical pumping’ of OSL’s, central to this proposal, has been achieved to date using Q-switched Nd-microchip lasers and, very recently, gallium nitride diode lasers. To take full advantage of the opportunity for LED pumping, we must work to improve the gain materials themselves, to improve upon the fairly rudimentary laser structures that have hitherto been used, and to make these materials and structures compatible with the high-refractive-index semiconductor and sapphire substrates of our proposed device geometry. We envisage organic multi-layer structures on our substrates, to provide microstructure photo-forming capability, waveguiding, device interconnections, patterning and control of the device active regions. For these reasons, we augment the organic electronics and device development capability within the programme with a suitably targeted synthetic (Skabara) and polymer (Pethrick) ‘smart material design’ chemistry capability. We emphasise in this context that, for device development in soft materials, this custom chemistry serves the same ‘engineering’ function as conventional micro-fabrication techniques applied to inorganic devices.
- Gallium nitride light-emitting diodes: status and opportunity
Recent advances in gallium nitride UV/visible quantum well LED’s for illumination and displays favour ‘flip chip’ (sapphire substrate up) and ‘thin-film’ (sapphire substrate removed) formats of InGaN-GaN devices. This has resulted, at blue wavelengths for example, in continuous-wave (CW) output powers now exceeding 1W at currents of a few hundred mA, from chips of active area ~1mm 2. In addition, mW-power, semiconductor face up, so-called ‘conventional chip’ (CC) LED devices, of active area typically ~350x350µm 2, are available in mass-market quantities at very low cost. Industry’s single-minded focus on such volume markets has opened up an opportunity for higher functionality customised GaN devices that university laboratories have been quick to seize. One of the most promising of these new device technologies is the GaN micro-LED arrays, which we have recently extended into trial flip-chip format (Fig.2). From the first, we have been alert to the optical micro-systems potential of this technology.
Benefits of the flip-chip format (whether CC or micro-LED) include; better light extraction (down-going light is retro-reflected at the p-contact), better current injection uniformity through the (‘thick’) p-contact, suitability for control from a backplane (which can, of course, be silicon), and an optically flat and inert sapphire upper surface suitable for further integration steps (and into/onto which optical micro/nano-structures can be etched/assembled as needed).
Fig. 2 Flip-chip micro-LEDs: schematic; a 16 x 16 array; a 72µm diameter blue pixel from such an array; three adjacent 30µm diameter UV pixels from such an array.
1.3 Silicon CMOS: interfacing to optoelectronics
The interface between silicon micro/nano-electronics and optoelectronics, a subject of worldwide topicality and importance, is tending to follow two parallel tracks: 1.) In one, CMOS imaging detectors have made tremendous advances for use in video cameras (an area pioneered by UoE) and optical instrumentation and are widely in use in both commercial products and laboratories. The recent research focus in this area has been on high timing resolution, leading to developments such as time-correlated single photon counting (TCSPC) with CMOS single-photon avalanche diodes (SPAD’s), in which one of the project partners (UoE, together with close collaborators in EPFL, Lausanne) has leading expertise. These devices have been implemented in 0.35µm, 0.18µm and now 130nm CMOS technology. 2.) In the other, there is now a very substantial effort devoted to developing high-performance communications components for optoelectronic circuitry directly integrable with silicon CMOS. The target is to accomplish this in such a seamless manner that the fabs. “do not know they are doing optoelectronics” and this is the area becoming known as Silicon Photonics. It includes fast detectors, modulators, waveguides and lasers as components, all ideally to be embodied in silicon for what will otherwise probably follow fairly well established precepts for 0.85 -1.55µm optoelectronic integrated circuits. Our picture of silicon photonics in this programme differs from the above, in that we seek to use Si-CMOS as the basis of a hybrid communications technology that takes advantage of the driving/modulating/detecting capability of the CMOS linked through to organic devices.
