Canek Fuentes-Hernandez , Bernard Kippelen*
Center for Organic Photonics and Electronics (COPE), School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
Material Matters, 2017, 12.3
The drive toward device integration over the past 50 years has produced a staggering miniaturization of integrated circuits and increased computing power. In the age of mobile computing, a pocket-size smartphone contains a processor with billions of transistors, has a display and a camera with millions of pixels, and is capable of performing billions of calculations per second, even more computing power than IBM’s Deep Blue, the most powerful supercomputer in the world in 1996. Along with the internet, mobile computing has produced an information revolution over the last decade that has renewed interest in the development of lightweight, flexible, and stretchable electronics for wearable devices, robotics, and autonomous vehicles, helping to realize the vision of ubiquitous computing and the internet of things. In the same period of time, new technologies for the control of biological processes have revolutionized biological and medical sciences, creating an urgent need for stretchable soft electronics that bridge the gap between the biological world and that of stiff optoelectronic technologies. In this context, the new generation of stretchable semiconductors, elastic substrates, and optoelectronic device architectures are poised to provide form-factors and functionalities (i.e. sensing, power generation and storage, wireless communication) that are complementary to those provided by the enormous computing power of mobile devices. When combined, these technologies offer the potential to dramatically change the ways in which humans experience and interact with the natural and the digital world. Here, we provide a short review of advances in the area of soft optoelectronics, with a focus on the development of organic optoelectronic devices on shape memory polymers (SMP).
Interest in soft materials for optoelectronics has seen a dramatic increase in recent years, with impressive demonstrations of lightweight ultraflexible and stretchable devices such as thin-film transistors, photovoltaics, imaging arrays, light-emitting diodes, and many others.1 Despite great progress, these efforts also highlight the need for further innovations in semiconducting and elastic materials as well as in novel device architectures.
From a material perspective, novel materials must be engineered to display thermomechanical properties that better match those found in natural biological materials, such as cellular materials, elastomers, polymers and polymer composites, and ceramics. At the same time, these novel materials must preserve the optical and electrical properties required to ensure their functionality in device architectures. To understand these requirements, we must consider that in general, biological materials have a low density (<3 g/cm3) with a Young’s modulus that can vary in the range from kPa to ca. 100 GPa, while their (yield/fracture) strength varies in the range from tens of kPa up to a few GPa, depending on the type of material.2 For instance, natural elastomers such as skin, muscles, and arteries display a Young’s modulus smaller than 80 MPa and strength smaller than 10 MPa.2 These mechanical properties are in sharp contrast with those displayed by traditional stiff optoelectronic devices comprised of rigid materials like metals, metal-oxides, and inorganic semiconductors. This is because inorganic materials have a density in the range from 2 to 20 g/cm3, a Young’s modulus in the range from 10 to 1,000 GPa and a strength in the range from 8 to 2000 MPa.3,4 Furthermore, these materials generally require high degrees of crystallinity to enable efficient charge transport and consequently are processed at high temperatures that are typically incompatible with soft biological substrates.
Despite challenges, rigid materials have been engineered to enable flexible and even stretchable optoelectronic devices. To understand this, it is important to note that the flexibility of a thin plate of a rigid material, quantified as the flexural rigidity D, depends cubically on the film’s thickness, is proportional to the Young’s modulus, and inversely proportional to one minus the square of the Poisson’s ratio of the material (i.e. the ratio of transverse contraction strain to longitudinal extension strain in the direction of the stretching force, typically in the range from 0.1, for stiff materials, to 0.5 for soft materials).5 Consequently, films of rigid materials with a large Young’s modulus in the hundreds of GPa range (e.g. glass, metals, silicon, metal-oxides, etc.) can become relatively flexible if processed into sufficiently thin films with a typical thickness of tens to hundreds of nanometers.4
Although optoelectronic devices comprised of thin films of rigid materials can be flexible, they are generally not very stretchable. To generate stretchable devices out of rigid materials, three main approaches shown in Figure 1 have been explored. In the first approach, small discrete devices deposited or transferred onto an elastomer substrate are connected with stretchable wirings. In the second and more common approach, flexible functional devices are fabricated or laminated onto a pre-stretched elastomer substrate, which is allowed to wrinkle when the strain is released. This strategy has been used to generate stretchable optoelectronic devices, comprising rigid materials, on soft elastic substrates such as poly(dimethylsiloxane) (PDMS).1,6 On the other hand, the emergence of nanostructured materials such as nanoparticles and nanowires has provided another route towards the realization of flexible and stretchable devices. In this approach, percolating networks of nanomaterials such as Ag nanowires (Cat. Nos. 778095, 739421 and 739448) are embedded in an elastomer matrix such as PDMS, by mixing nanomaterials and elastomer monomers prior to curing.7 Furthermore, stretchable sensors have also been fabricated using eutectic metal alloys (eutectic gallium–indium alloy, EuGaIn that is liquid at room temperature, Cat. No. 495425) embedded in PDMS using soft lithography and may provide an attractive route towards developing fully stretchable electrodes.8
Figure 1. Approaches towards stretchable optoelectronics. A) Discrete devices are connected with each other using stretchable wirings. B) Stretchable devices are fabricated by depositing the flexible layers of a device onto a pre-stretched elastomer (elastic) substrate, resulting in wrinkled devices upon releasing the strain on the substrate. C) Nanomaterials mixed with elastomer monomer followed by curing.
In contrast to inorganic semiconductors, organic semiconductors display thermomechanical properties (e.g. low density below 3 g/cm3 and Young’s modulus in the range 0.1 to 1 GPa)9 that are better matched to those of soft biological materials. In addition, organic semiconductor films are processed from solution at temperatures that are compatible with a wide range of soft elastic substrates. Consequently, the use of organic semiconductor films (polymers, small molecules, or blends) has led to the realization of ultra-flexible optoelectronic devices, such as thin-film transistors and photovoltaics10,11 that remain operational when flexed to a bending radius down to a few microns. Despite their inherent flexibility, the stretchability of organic semiconductors largely depends on their molecular weight, molecular packing in the solid state, and their molecular composition.9,12–14 Generally, amorphous polymer films display superior mechanical properties (i.e. smaller Young’s modulus), including stretchability, but also display inferior charge transport properties than films comprising more rigid materials such as crystalline polymers or small molecules.12 Because of this tradeoff, stretchable organic optoelectronic devices have largely relied on similar approaches to those used for inorganic semiconductors; fabrication of wrinkled devices on prestretched substrates or use of nanomaterials, such as carbon nanotubes, embedded on an elastic matrix.15 However, the ability to tailor the properties of organic semiconductors using synthetic chemistry and compositional engineering has, in recent years, produced important advances towards developing truly stretchable organic semiconductor layers. In one approach, polymer semiconductor nanofibers of high-mobility semiconducting polymers were formed through phase separation inside a soft elastic matrix (polystyrene-b-poly(ethyleneran- butylene)-b-polystyrene (SEBS), Cat. Nos. 200565 and 200557) leading to the devlopment of highly stretchable polymer semiconductor films and stretchable thin-film transistors that display charge mobility values in the range between 0.5 to 1 cm2/Vs even when subjected to a strain of 100%.16 In the second approach, akin to the connection of devices with stretchable wirings, molecular stretchability is engineered into conjugated polymers by adding chemical moieties to the backbone that, on one hand, do not hinder the formation of crystalline domains and, on the other, promote dynamic non-covalent crosslinking between flexible polymer chains in the amorphous regions. These dynamic bonds can easily break upon strain to dissipate energy through the amorphous regions while preserving the integrity of charge transport through the more ordered regions.14 Following this approach, semiconductor polymer films and thin-film transistors can be stretched up to 100% strain while displaying charge mobility values ca. 1 cm2/Vs.14 In addition to achieving high stretchability, this approach could also enable self-healing properties.14
Continued progress in the development of novel materials, material processing, and device engineering is producing significant advances in the performance and stability of organicbase optoelectronics devices such as organic thin-film transistors (OTFTs), organic light-emitting diodes (OLED), and organic photovoltaics (OPV). Reviewing the extensive amount of progress in each area is out of the scope of this contribution.
An integral component of any stretchable optoelectronic device is the selection of the elastic substrate. Elastic substrates must display characteristics such as (1) a Young’s modulus approaching that of the targeted biological material (e.g. Young’s modulus of the skin is ca. 0.7 MPa), (2) a low Poisson’s ratio to avoid the formation of cracks, (3) good thermomechanical stability to ensure repeatable deformation, (4) chemical stability and biocompatibility to allow the direct fabrication of optoelectronic devices, and (5) compatibility with the targeted biological materials. To date, PDMS remains the most widely used elastic substrate for stretchable devices. PDMS offers excellent stability and biocompatibility, but its stretchability is typically limited to 200%. Alternative materials that can achieve stretchability of ca. 700% are Ecoflex, a platinum-catalyzed silicone17 and poly[styrene-b-(ethylene-co-butylene)-b-styrene] resin and 3MTM VHBTM 4905 (transparent tape with a generalpurpose acrylic adhesive).18 Other elastic substrates include polyurethanes and polyacrylate elastomers.6
Among the wide range of potential soft substrates, shapememory polymers (SMPs) are an emerging class of active stimuli-responsive polymers with many properties that are attractive to the development of biomedical, robotic, and autonomous vehicle applications. SMPs display the ability of changing their shape in response to external stimuli. When this stimulus is temperature, SMPs are thermoresponsive. A thermoresponsive SMP presents three temperature-dependent states, a glassy state at low temperature characterized by a large Young’s modulus and low deformability. A region at a higher temperature where the material softens and exhibits pseudoelastic properties. This region is characterized by the glass transition temperature, Tg, around which the Young’s modulus experiences a drastic drop in value. At temperatures higher than Tg, the SMP enters into the rubbery regime, characterized by a small Young’s modulus and high deformability. In this region, the shape of the SMP can be modified and the modified shape can be preserved (or maintained) if the temperature is reduced below Tg. If the SMP is once more reheated above Tg, the transition into the rubbery regime causes the SMP to regain its original shape. To understand this reshaping mechanism, we need to consider that polymer networks in an SMP form in a stress-free, global, free-energy minimum that minimizes the entropy of the system. Upon heating, deformation, and cooling, configurational changes in the polymer chains acquire and store mechanical stress in the polymer, creating a higher-entropy stressed metastable equilibrium, which upon heating and in the absence of further mechanical deformations, relaxes back into the lower entropy, stress-free state characterized by the original polymer shape. This process is illustrated in Figure 2, where the SMP has been casted on a curved mold. The ability of an SMP to recover its shape is quantified as the recoverable strain. This ability to control the rigidity and shape of SMP substrates consequently could offer potential advantages over elastic substrates for the fabrication, handling, and deployment of soft optoelectronic devices.
Figure 2. Shape recovery of curved SMP.
Recently, thiolene chemistries have yielded SMPs that are biocompatible19–21 and suitable for use as active substrates for mechanically adaptable OTFTs.22–26 SMPs based on thiolene reactions (click reactions) display low-cure stresses due to the nature of the step-growth mechanism in this polymerization, resulting in highly uniform/dimensionally stable polymer networks with low shrinkage and surface roughness, as well as strong adhesion to metal layers. More significantly, this allows for various material properties (such as Tg, rubbery modulus, and hydrophobicity) to be altered by controlling the concentration of the constituting monomers. For example, varying the concentration of tricyclo[5.2.1.02,6]decanedimethanol diacrylate (TCMDA, Cat. No. 496669) in a blend of 1,3,5-triallyl-1,3,5- triazine-2,4,6(1H,3H,5H)-trione (TATATO, Cat. No. 114235), trimethylolpropane tris(3-mercaptopropionate) (TMTMP, Cat. No. 381489), and 2,2-dimethoxy-2-phenyl acetophenone (DMPA, Cat. No. 196118), a photocuring agent, changes the Tg of the SMP. We have recently taken this SMP composition to explore the realization of top-gate OTFTs with bilayer gate dielectrics and OLEDs.
In the past, we have shown n- and p-channel top-gate OTFTs can be engineered to display excellent operational and environmental stability when a bilayer gate dielectric comprising a first CYTOP layer and a second metal-oxide layer processed by atomic layer deposition (ALD) is used.27,28 In this approach, the bilayer gate dielectric serves the role of an environmental barrier and introduces a second mechanism (e.g. dipolar orientation or charge injection) to compensate the threshold voltage shift induced by charge trapping. This device geometry produces OTFTs that sustain immersion in water for prolonged periods of time at temperatures near its boiling point,29 opening the door for stable chemical and biological sensors that operate in aqueous conditions.30,31 In addition, we have investigated the properties of top-gate OTFTs and circuits on thiolene based SMP substrates32 designed to yield a Tg of 43 °C. Figure 3, shows the properties of single devices and pseudo-complementary inverter circuits comprising solutionprocessed OTFTs, with a bilayer comprising a 35 nm layer of CYTOP and a 31 nm-thick dielectric nanolaminate29 fabricated by ALD on ca. 130 μm-thick SMP substrates. Compared to single-material ALD layers, the use of a nanolaminate was found to yield reduced cracking under mechanical deformation, possibly due to a reduction in the residual tensile stress produced during the ALD layers.33 OTFTs on SMPs yield mobility values of 0.9 ± 0.59 cm2/Vs with very small contact resistance due to the deposition of the p-type dopant molybdenum tris[1,2-bis(trifluoromethyl)ethane-1,2-dithiolene] (Mo(tfd)3 Cat. No. 795712)34–37 onto the source and drain electrodes. This level of performance is comparable to that displayed by top-gate OTFTs on glass or plastic substrates. In addition, devices exhibit good flexibility at room temperature, with no changes of their transfer characteristic before and after undergoing a compressive bending cycle down to a 7 mm radius of curvature (strain ca. 0.9%). When OTFTs are subjected to thermally-induced reshaping, devices maintain a high level of performance down to a bending radius of ca. 14 mm. However, beyond this point, the performance deteriorates due to cracking of the ALD nanolaminate layer, suggesting that the device is presumably subjected to large strains during the reshaping process. It should be noted that typical polycrystalline organic semiconducting layers9,12 as well as ALD layers38 are known to undergo significant cracking or even failure when reaching critical tensile strains in the range from ca. 1-2%. While the use of ALD layers may impose limitations to the development of fully stretchable devices, the benefits in device performance and stability may offset such limitations in applications that may not require extreme deformations. Alternatively, fabrication of OTFTs on pre-stretched SMPs may offer an alternative path towards improving the mechanical resilience of environmentally and operationally stable OTFTs.
Figure 3. A)Transfer characteristics of OTFTs on SMPs before and after flexing. B) Transfer characteristics of OTFTs on SMPs before and after thermally-induced reshaping at 60 oC. C) Output characteristics of pseudo-complementary inverted before and after thermally-induced reshaping. D) Photograph of curved pseudo-complementary inverter circuits on SMPs.
It should be noted that while other design strategies exist for improving the environmental stability of OTFTs through the passivation of defects using molecular additives such as 7,7,8,8-Tetracyanoquinodimethane (TCNQ, Cat. No. 157635), 2,3,5,6-Tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F4TCNQ, Cat. No. 376779) and 4-aminobenzonitrile (ABN),39 improved design strategies will be needed for the development of stretchable OTFTs that display both high operational and environmental stability at the same time.
Figure 4.Change in current efficacy of OLED on SMPs after reshaping and images of devices under tensile strength.
On the other hand, OLEDs also offer important opportunities to develop large-area light sources with low heat dissipation, attractive characteristics for optogenetic and other bioengineering applications, particularly when fabricated on biocompatible SMPs. OLEDs on SMP substrates were first demonstrated using a single-walled carbon nanotube/polymer composite electrode as an indium tin oxide (ITO) replacement. These devices produced a maximum current efficacy of 1.24 cd/A at 200 cd/m2 with a turn-on voltage of 4.8 V, and a maximum luminance of 300 cd/m2. More recently, we reported the development of inverted top-emitting green-phosphorescent OLEDs fabricated on biocompatible SMP substrates that produced current efficiencies of 33 cd/A at a high luminance of 1,000 cd/m2 with a low turn-on voltage of 3.4 V. These devices produce a maximum luminance of over 30,000 cd/m2, making them ideal candidates for bioengineering applications that require high irradiance levels.40 Moreover, these samples can be re-shaped and softened into a curved form factor with a bending radius of 5 mm (strain ca. 1.5%). Although these devices show a significant performance drop at low luminance levels (100 cd/m2), they perform equally as well when larger biases are applied and show great promise for applications that require high luminance values (Figure 4). All the devices measured could withstand the initial heating, reshaping, and return to their original form factors. Further opportunities to improve the performance and thermomechanical properties of OLEDs on biocompatible SMPs may exist with the use of thermally-activated delayed fluorescence (TADF) chromophores41 and high-Tg electron and hole transport materials.42 These tests demonstrate the ability to thermally evaporate inverted top-emitting OLEDs on SMP substrates for flexible and conformable devices, highlighting the potential for such sources for a variety of applications. However, the mechanical properties of these evaporated OLEDs are still limited and can lead to cracking of the metal layers and ultimately failure of the OLEDs upon repeated reshaping or under more severe deformation. Detailed studies are needed to assess the sources of failure and to address the best way to ensure the mechanical reliability of OLEDs on SMPs.
To reach the full potential of organic optoelectronics on soft substrates such as SMP’s, several problems still need to be addressed. One of the most pressing needs is ensuring adequate environmental and operational stability of devices fabricated on soft substrates. Environmental barriers such as those deposited by ALD on soft elastic substrates should be investigated to develop an in-depth understanding of the potential of using rigid barrier layers meant to improve environmental stability and reduce constraints pertaining to moisture penetration. From a material perspective, great progress has been made in developing organic semiconductors with stabilized frontier orbitals that are less prone to oxidation.30,43–45 Confinement of charge transport in nanoaggregates also appears to be a promising avenue for developing materials that are at the same time stretchable and may be less prone to interactions with oxygen and water in the environment. In addition, molecular additives have recently been demonstrated to improve the stability of OLEDs,46 OTFTs,39 and OPVs.47 On the other hand, advances in interfacial engineering show work function engineering is an effective route to improve the stability of devices. As an example, we have found that polyethylenimine ethoxylated (PEIE, Cat. No. 306185) and branched polyethylenimine (PEI, Cat. No. 482595)48 not only enable efficient electron injection and collection from a wide range of environmentally stable conductors, but this approach has also been shown to improve OPV stability.49 Aside from reducing the work function, these materials also lead to n-doping of organic semiconductors48 and consequently open the door for the creation of n-doped regions by phase segregation, shown to produce self-assembled electron-collecting interfaces in OPV devices.50 Analogously, the use of p-dopants is expected to provide a path towards creating environmentally robust interfaces that avoid use of reactive metals and also may lead to simplification of a device architecture. For example, we recently discovered organic semiconductor films are efficiently p-doped to a limited-depth by post-process immersion in a solution of phosphomolybdic acid (PMA, Cat. No. 431400) in nitromethane (Cat. No. 360554). The ability to electrically dope the interface removes constraints on the work function of the metal used, and when applied to a bulk-heterojunction film in which PEIE was mixed in solution prior to film formation, enables the realization of single-layer OPVs in which the electron-collection and holecollection functionalities are embedded within the photoactive layer.51 Hence, although the development of stretchable organic semiconductors and devices architectures may still be at its infancy, advances in the area of organic optoelectronics can be expected to enable the realization of devices that are not only mechanically compliant but that also achieve the level of performance required for next generation stretchable optoelectronic devices.
We would like to thank funding in part from the Office of Naval Research Awards N00014-
04-1-0313, N00014-14-1-0580, and N00014-16-1-2520 through the MURI Center CAOP.