DOI: 10.1002/adma.200702304 Polymer Transfer Printing: Application to Layer Coating, Pattern Definition, and Diode Dark Current Blocking** By Lichun Chen, Patrick Degenaar, and Donal D. C. Bradley* The solution processability of polymer semiconductors offers attractive routes to high throughput, low temperature device fabrication using printing methods (e.g., ink-jet, gravure and screen). This opens new opportunities in many application sectors including displays, lighting, large area electronics, photodetection, imaging, solar energy conversion, and com- munication. [1,2] Vacuum deposited small molecule devices utilize multilayer structures to fine tune device performance, e.g., exciton blocking layers in phosphorescent light emitting diodes (LEDs). [3] It is not, however, straightforward to do the same from solution: So-called solvent orthogonality (incompat- ibility) needs to be engineered for the materials that have to be sequentially coated so that deposition of one layer doesn’t adversely affect the layers already in place. [4] This has also to be done in a way that ensures that the deposition of subsequent layers produces uniform coverage of the underlying material and an intimate physical contact thereto. One approach is to solution cast polymer films on water and to scoop them up as free-standing layers or, equivalently, to spin-coat on top of a water soluble substrate (e.g., KBr) and dissolve the substrate to leave a floating layer that can again be recovered onto a substrate. A related approach uses Langmuir–Blodgett (LB) monolayer deposition. Both float transfer and LB processes involve immersion of the substrate in water and require the structure to be carefully dried afterwards. [5] In addition, it is not straightforward to prepare flat, unwrinkled, 100 nm thickness films. Other approaches include precursor-route [6] or cross-linking [7] methods to generate insoluble layers on top of which others can then be coated. Drawbacks in the former case can include the requirement for conversion temperatures and/or environments (e.g., acidic) that are not conducive to the use of plastic substrates or a variety of electrodes (e.g., indium tin oxide). In the latter case there is a need for specific chemical functionalization and there may be issues over completion of the reaction (to mop up reactive moities) and/or removal of initiators. Shrinkage can also be a problem in both cases, though not always so: The Meerholz et al. oxetane route avoids this problem. [7] Here we report a new polymer transfer printing process that avoids the solvent orthogonality problem in fabricating polymer multilayers, that can be used to define patterns and that doesn’t require any chemical modification of the materials to be deposited. [8] The process uses a soft stamp to transfer a solid polymer layer of the appropriate thickness (10–100s nm) to a substrate and/or (as required) to pattern that layer during or after transfer. Poly(dimethylsiloxane) (PDMS) stamps are the bedrock of standard micro contact printing (mCP) techniques [9] and have also been used here. They readily make conformal contact with surfaces and (at least in principle) any type of layer deposited onto a PDMS stamp should be transferable to a substrate using mCP. In practice, however, successful transfer requires careful control over the relative adhesion between layer and substrate and layer and stamp and over the rigidity of the layer. The challenge has been to come up with a control process suitable for general use with the polymers found in typical devices. The benefits of this new process are illustrated for polymer/fullerene photodiodes through engineering a significant reduction in reverse bias dark current via adoption of a multilayer structure. The ability to sequentially deposit well defined functional layers enables the shunt resistance to be increased and thereby offers greater detection sensitivity. Our polymer transfer printing process generally comprises four steps (see Fig. 1a–d): Step 1: (N.B. if the substrate to be coated is soft, e.g., a typical polyester then this step can usually be omitted). For hard substrates such as glass or metal a thin glycerol film is deposited as an adhesion/rigidity control agent to aid conformal contact between the polymer layer to be transferred and the hard substrate. Step 2: The polymer layer to be transferred needs to be prepared on the PDMS stamp (see next section for discussion of pattern defi- nition). The polymer-coated stamp is then aligned and brought into contact with the substrate. Step 3: After the stamp and substrate are brought into contact they are briefly heated (typically 30–50 s) close to the glass transition temperature of the polymer to be transferred (i.e., in the range 120–140 8C for the polymers we have studied). This step facilitates adhesive contact between the polymer layer on the PDMS stamp COMMUNICATION [*] Prof. D. D. C. Bradley, Dr. L. Chen Experimental Solid State Physics Group, Blackett Laboratory Imperial College London Prince Consort Road, London SW7 2AZ (UK) E-mail: d.bradley@imperial.ac.uk Dr. P. Degenaar Institute for Biomedical Engineering, Imperial College London Exhibition Road, London SW7 2AZ (UK) [**] Based in part on a presentation at SID 2007, Long Beach, USA (May 24th 2007). We thank the Research Councils UK Basic Technology programme (GR/R87642 & EP/E045472) and the Royal Society Brian Mercer Bequest for Funding. Adv. Mater. 2008, 20, 1679–1683 ß 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1679