Highly conductive and stretchable poly(dimethylsiloxane):poly(3,4-ethylenedioxythiophene):poly(styrene sulfonic acid) blends for organic interconnects

Abstract

Naturally immiscible PEDOT:PSS and PDMS, which are a typical conducting polymer and an transparent elastomer, respectively, were blended by the support of PDMS-b-PEO. A block copolymer, PDMS-b-PEO, consisting of hydrophobic PDMS backbones and hydrophilic PEO side chains, significantly improved the miscibility of PEDOT:PSS and PDMS. At an optimal PDMS-b-PEO concentration of 30%, a cured PEDOT:PSS:PDMS film was found to be comprised of three-dimensional PDMS networks and a PEDOT:PSS phase filling in between the networks. The optimal blend film exhibited a conductivity comparable to a pure PEDOT:PSS film and a maximum strain to rupture of about 75%. It was also demonstrated that interconnects made of this blend film functioned well irrespective of the substrate and the pattern size, and could reproducibly operate under strains up to 50%. These results indicate that the PEDOT:PSS:PDMS blends could be a practical choice for organic interconnects for future stretchable electronics.

---

Introduction

Stretchable electronics are believed to gradually be applied in many areas, such as sensory skins, conductive fabric, wearable devices, health monitors, and bio-integrated devices.^1–4 To realize the stretchable electronics, both mechanical stretchability and electronic performance should be ensured at the same time. Although all-plastic electronics have shown successful demonstrations in various fields,^5,6 their performance falls short of well-developed conventional inorganic devices. Smart hybrid systems comprised of rigid or simply bendable active devices and stretchable interconnects have emerged as an alternative,^7,8 where the full systems become stretchable by the strain-absorbing capability of the interconnects. Two approaches have been pursued in embodying the stretchable interconnects: formation of wavy or mesh configuration using rigid materials^9–11 and direct formation of stretchable interconnects from elastomeric materials.^12,13 Of the two, the direct formation method of stretchable interconnects is more desirable since elaborate pattern-defining steps that are required for the first approach can be avoided. For this method, various polymer blends or polymers mixed with functional nanostructures have been explored more intensively than single conducting polymers that generally have poor mechanical properties.^13–16

Poly(3,4-ethylenedioxythiophene) (PEDOT) doped with poly(styrene sulfonic acid) (PSS), which is normally notated as PEDOT:PSS, is one of major conducting polymers that is already commercially applied to plastic electronics and organic light-emitting diodes (OLEDs).¹⁷ In the PEDOT:PSS blend, PEDOT oligomers are tightly bound to PSS chains by the electrostatic interaction, making it water-dispersible with high stability. In addition, it shows high electrical conductivity up to 500 S cm⁻¹ and excellent chemical stability.¹⁸ Despite these advantages, the PEDOT:PSS blend can not be engaged in stretchable interconnects unitarily because of its poor mechanical properties. To overcome this drawback, polymer blends such as PEDOT:polyurethane (PU)¹⁹ and PEDOT:PSS:polyvinyl alcohol (PVA)¹⁸ or crosslinking of the PEDOT:PSS¹⁷ have been studied. Attractive results have been reported: for instance, a high conductivity of 100 S cm⁻¹ was achieved at a strain of 100% for PEDOT:PU blends.¹⁹ However, the blends needed to undergo somewhat complicated polymerization process and their conductivities were still lower than pure PEDOT:PSS. Furthermore, the base resins that are responsible for reinforcing mechanical properties were not transparent, potentially limiting their use in transparent electronics and displays. In another approach, a resilient PEDOT:PSS film has been obtained by spin-coating PEDOT:PSS solution with a 1% Zonyl fluorosurfactant onto a ultraviolet/ozone-treated poly(dimethylsiloxane) (PDMS) substrate.²⁰ Although the PEDOT:PSS film showed a resistance increase by a factor of only 2 at a 50% strain, its stretching became irreversible beyond a 30% strain.

In this work, the PEDOT:PSS is combined with PDMS, which is widely used in transparent and flexible electronics.^21,22 Due to its high hydrophobicity, however, the PDMS is not mixed with most aqueous solvents.^23,24 For this reason, it has been ruled out in research of stretchable interconnects. In this work, to reduce the hydrophobicity of the PDMS and make possible blending with the PEDOT:PSS, poly(dimethylsiloxane-b-ethylene oxide) (PDMS-b-PEO), which is hydrophilic,²⁵ is added as a third component. The PEDOT:PSS:PDMS:PDMS-b-PEO blends are crosslinked using a curing agent. The miscibility, conductivity, and mechanical property of the blends are investigated, varying the relative compositions. Finally, the possibility of their use for stretchable interconnects is examined.

Experimental details

In this study, commercially available, highly conductive PEDOT:PSS (PEDOT:PSS = 1 : 2.5, Clevios PH1000, Heraeus) was used. The PDMS oligomer liquid (Sylgard 184) was purchased from Dow Corning Corporation in combination with a curing agent. The block copolymer, PDMS-b-PEO, which consists of hydrophobic PDMS backbone and hydrophilic ethylene oxide side chains, was purchased from Polysciences Inc. The PDMS-b-PEO is supposed to have a unit repetition ratio of PEO segment to PDMS segment of 3, a molecular weight of 600, and the viscosity of 20 cP.²⁵ All chemicals were used as purchased without further treatment. To prepare polymer blends, all four constituents (PDMS, curing agent, PDMS-b-PEO, PEDOT:PSS) were first dropped into a vial in a stepwise manner with tight weight control. The PEDOT:PSS with a bluish color was dropped at the last to watch if any phase separation takes place between that and the rest. On the basis of the PDMS weight, the relative weights of PEDOT:PSS and PDMS-b-PEO were varied: 200, 300% for PEDOT:PSS and 3, 10, 20, 30% for PDMS-b-PEO. The amount of the curing agent was controlled at 10% of the combined weight of PDMS and PDMS-b-PEO.^26,27 Here, the weights of all the components refer to the wet weights measured in their liquid states. They were vigorously stirred at 250 rpm for an hour and then the liquid mixture was poured onto a Petri-dish. It was kept in atmosphere for more than an hour to make the viscous mixture uniformly spread over the dish and vacuum-degassing step was followed. In the next step, it was cured at 70 °C for 2 hours and subsequently at 80 °C for another 1 hour. The thickness of the films prepared in this way ranged from 100 to 140 μm. Along with these polymer blend samples, pure PDMS and PEDOT:PSS films were also prepared as control samples.

Macroscopic phase separation was inspected above all in both the liquid state and the solid state. To further check the miscibility of the blends with PDMS, the blend dropplets were dropped onto the PDMS substrate and their spreadability was observed. The morphologies of the polymer blend films were examined using an optical microscope (Olympus BX 51). Microscopic blend structures and spatial element distribution were also analyzed employing field emission scanning electron microscopy (FE-SEM, Hitachi S4700). The acceleration voltage was 15 kV. The molecular motions of certain chemical groups were investigated using Fourier-transformed infrared (FT-IR) spectroscopy (Avatar 320 FT-IR, Nicolet). The basic mechanical properties of selected samples were evaluated using a universal testing machine (Instron 3366). The sheet resistances of the blend films were measured using a standard four-probe method and the conductivities were calculated from them. Finally, to evaluate its possibility of being used for stretchable interconnects, contact pads and interconnects were fabricated from a polymer blend employing a simple method depicted in Fig. 1. The current (I)–voltage (V) characteristics and the resistances of the interconnects were measured under no strain or a tensile strain, using a probe station that was connected to a high-resolution electrical characterization system (Keithley 4200-SCS).

Fig. 1 Fabrication procedures of polymer blend contact pads and interconnects for electrical testing under a tensile strain. Step 1: preparation of a PDMS sheet and a polymer blend liquid, step 2: selective masking of the PDMS sheet using Kapton tape and dropping the polymer blend onto the unmasked area, step 3: curing the polymer blend (2 hours at 70 °C then 1 hour at 80 °C), step 4: lift-off of the Kapton tape mask, step 5: I–V measurements under varying strains. The inset shows an actual pattern image fabricated from this method. The pattern was made of a polymer blend and formed on a PDMS sheet that is delineated with a yellow dotted line.

Results and discussion

Fig. 2(a) shows the natural immiscibility of PEDOT:PSS with PDMS. The initial PEDOT:PSS drops form a circular island on the surface of PDMS liquid. The phase separation is maintained even when the PEDOT:PSS fraction prevails over PDMS and the mixture is vigorously shook for 1 min (see the double-layer structure and the curved interface of Fig. 2(b)). The introduction of PDMS-b-PEO turned out to relieve the interfacial tension between the two phases. When a 10% PDMS-b-PEO relative to PDMS weight is added, the interfacial curvature is alleviated (Fig. 2(c)) and no significant phase separation is observed at a 30% PDMS-b-PEO (Fig. 2(d)). The efficacy of the PDMS-b-PEO is further supported by the shape and spreadability of PEDOT:PSS drops on the PDMS substrate. The PEDOT:PSS:PDMS blend with a 30% PDMS-b-PEO uniformly spreads over the substrate (Fig. 2(f)), while the pure PEDOT:PSS drops form an oblate spheroid-shaped agglomerate due to their high surface tension at interface with the PDMS substrate (Fig. 2(e)).

Fig. 2 (a) Photograph of a PEDOT:PSS island floating over PDMS liquid. (b)–(d) Photographs showing the relative miscibility of PEDOT:PSS and PDMS in a liquid state as a function of PDMS-b-PEO concentration. Photographs of (e) PEDOT:PSS drops and (f) drops of a polymer blend with a 30% PDMS-b-PEO on a PDMS substrate.

The effect of PDMS-b-PEO on the miscibility of PEDOT:PSS with PDMS is also demonstrated in the form of solid film. Fig. 2(a)–(e) show macroscale images of PEDOT:PSS:PDMS films with varying concentrations of PDMS-b-PEO relative to PDMS (0 to 30%). For all, the weight ratio of PEDOT:PSS to PDMS was 2. As expected from the observation in the liquid state (Fig. 2(a)), PEDOT:PSS and PDMS phases are completely separated without PDMS-b-PEO (Fig. 3(a)). Gradually increasing the concentration of PDMS-b-PEO, the two phases begin to be mixed. At 10%, the film is comprised of a PEDOT:PSS-PDMS mixed phase (primary phase) and an isolated PDMS phase (secondary phase), which is enclosed by the primary phase (Fig. 3(c)). The secondary phase becomes almost extinct at 20%, but the PDMS inclusions in the primary phase are still large (up to 3 mm, Fig 3(d)). Finally, the PDMS inclusion size is greatly reduced at a concentration of 30% and the film looks like a uniform film where PDMS is well-dispersed in the PEDOT:PSS matrix (Fig. 3(e)). From the point that the film is not as glossy as a pure PEDOT:PSS film and the total area fraction of the observable PDMS dots is far short of the expected level, it is inferred that the major portion of PDMS phase is present inside the film. The optical micrographs shown in Fig. 3(e)–(j) reveals a trend consistent with the observations at the macroscale. With an increase in the PDMS-b-PEO concentration, the size of isolated PDMS granules shrinks. Interestingly, the small PDMS granules seem to be connected beneath the film surface at a 30% PDMS-b-PEO, whereas the PEDOT:PSS phase forms a network structure meandering through the PDMS granules (Fig. 3(j)). The film uniformity appeared comparable to or even better than the 30% result when the PDMS-b-PEO concentration is increased further. However, the PDMS-b-PEO itself was found to be separated from the PEDOT:PSS:PDMS film at high concentrations (see Fig. S1(a)†).

Fig. 3 (a)–(e) Macroscopic images of PEDOT:PSS:PDMS mixtures with differing concentrations of PDMS-b-PEO (0 to 30%). The ratio of PEDOT:PSS to PDMS was 2 : 1 by weight for all samples. (f)–(j) Microscopic optical images of the selected areas of the corresponding macroscopic images ((a)–(e)). Scale bars 5 mm for (a)–(e) and 30 μm for (f)–(j).

Fig. 4(a) shows a SEM image of PEDOT:PSS:PDMS film (2 : 1 by weight) with a 30% PDMS-b-PEO. Closely spaced, hairy ball-like domains (whitish part) are observed in the ruffled sea (the remaining part). Interestingly, the two parts are not completely separated from each other, but many whitish streamlines emanate from the ball-like domains, connecting a domain to its neighbors. Energy-dispersive X-ray spectroscopy (EDX) measurement was performed to analyze elemental distribution over the respective areas (dotted areas “1” and “2” in Fig. 4(a)). Four major elements (C, O, Si, S) were detected from both areas as shown in Fig. 4(b). The silicon concentration is comparable in both areas, but the sulfur concentration is slightly larger in area “2” while more oxygen atoms are detected from area “1”. This result indicates that both areas are comprised of PEDOT:PSS and PDMS, but more PEDOT:PSS is included in area “2” while the combined concentration of PDMS and PDMS-b-PEO is higher in area “1”. It is inferred that the main components of whitish streamlines and black matter between them in area “2” are PDMS and PEDOT:PSS, respectively. Furthermore, the relatively high oxygen concentration in area “1” reflects that PDMS-b-PEO mostly resides on the surface of PDMS domains, anchoring its backbones into the PDMS and holding PEDOT:PSS through its PEO side chains. Under these circumstances, the PDMS part is likely to form three-dimensionally networked nano- or micro-structures whereas the PEDOT:PSS part fills the empty space of the PDMS networks. The similar structures and elemental distribution were also obtained from other locations of the sample (see Fig. S3†). From this result in conjunction with the previous macroscale observations, it is inferred that relatively short PDMS segments of PDMS-b-PEO molecules are mixed and cured with PDMS component of a polymer blend, while long, hydrophilic PEO side chains are solubilized in PEDOT:PSS component of the blend. A combination of these roles of two segments of PDMS-b-PEO may make PEDOT:PSS molecules bound to the surface of nano- or micro-structures of PDMS, overcoming the natural hydrophobicity between PEDOT:PSS and PDMS.

Fig. 4 (a) SEM image of a polymer blend consisting of a 200% PEDOT:PSS, a 100% PDMS, and a 30% PDMS-b-PEO. (b) EDX spectrographs obtained from the two areas (“1” and “2”) marked in (a). The insets represent the relative atomic fractions of the four major elements.

Sheet resistances of various polymer blend films were measured using a four-probe method. Fig. 5(a) shows the sheet resistance distribution of PEDOT:PSS:PDMS films (2 : 1 by weight) with varying concentrations of PDMS-b-PEO (0 to 40%). Here, a 0% film represents PEDOT:PSS film and the sheet resistance of a 10% film was obtained only from PEDOT:PSS-PDMS mixed part (primary phase). PEDOT:PSS and PDMS are completely immiscible without PDMS-b-PEO and the sheet resistance of PDMS film is more than 10⁷ Ω □⁻¹. Comparing with the sheet resistance of PEDOT:PSS film that ranges from 0.9 to 1.1 kΩ □⁻¹, PEDOT:PSS:PDMS films generally exhibit a broader range of sheet resistances due to the inhomogeneous blending of the two components. The sheet resistance distribution becomes narrower as the PDMS-b-PEO concentration increases, and it is comparable with that of PEDOT:PSS at a concentration of 30% (200 to 300 Ω □⁻¹). Beyond this optimal concentration, the average sheet resistance and the sheet resistance distribution rapidly increase presumably due to the phase separation of surplus PDMS-b-PEO. In addition, the sheet resistance distribution is noticeably narrowed when the amount of PEDOT:PSS relative to PDMS is increased (Fig. S4†). Fig. 5(b) displays the distribution of conductivities converted from the sheet resistance. The conductivity was calculated using a simple equation, σ = 1/R_st, where σ, R_s, and t represent the conductivity, sheet resistance, and the thickness of a film. The relative distribution of conductivities resembles that of sheet resistances. Surprisingly, the conductivity (0.26–0.38 S cm⁻¹) of PEDOT:PSS:PDMS film with a 30% PDMS-b-PEO is almost the same as the value (0.3–0.37 S cm⁻¹) of PEDOT:PSS film, confirming the conductive PEDOT:PSS part well-functions in the blend film without performance degradation. The low conductivity of PEDOT:PSS film may be improved by adjusting the component ratio or adding a dopant.

Fig. 5 Distributions of (a) sheet resistances and (b) conductivities of polymer blends with varying concentrations of PDMS-b-PEO (0 to 40%). The sheet resistance was measured by a four-probe method and the conductivity was calculated from the sheet resistance and the thickness of a film. The weight ratio of PEDOT:PSS to PDMS was fixed at 2 : 1. The data at 0% PDMS-b-PEO came from a PEDOT:PSS film with no PDMS. (c) Average sheet resistances of polymer blends with different compositions depending on the applied strain. Strains at rupture were marked with star symbols. The black star represents the rupturing point of a PEDOT:PSS film.

Sheet resistances of the blend films were also measured under elongation. The degree of elongation is characterized by strain, ε = (L − L₀)/L₀ × 100 (%), where L₀ and L are the original length and the elongated length, respectively. Fig. 5(c) shows the average sheet resistances of various blend films depending on strain. The sheet resistance increases at a slow rate in a low strain range (<50%) and at a relatively fast rate in a high strain range (>50%) with an increase in strain. The increase at low strains may be attributed to changes in conformation and length while the fast increase at high strains seems to originate from the delamination and breakdown of PEDOT:PSS part. From the point of film deformation, the low- and high-strain behaviors correspond to elastic and plastic deformation, respectively. Although blends with a 300% PEDOT:PSS with respect to PDMS weight show lower sheet resistances compared to their 200% counterparts, their rupturing points, which are marked with star symbols in Fig. 5(c), are smaller by about 20%. All blend films are revealed to rupture at sizably larger strains than that (12%, black star) of PEDOT:PSS film by the mechanical reinforcement effect of PDMS part. The maximum strain to rupture is approximately 75% obtained from a blend film consisting of a 200% PEDOT:PSS, a 100% PDMS, and a 30% PDMS-b-PEO. In a supplementary experiment, a 100% PDMS-b-PEO was blended with a 200% PEDOT:PSS without PDMS and annealed at a standard curing condition. However, the blend was not cured, justifying that the improved mechanical property of PEDOT:PSS:PDMS:PDMS-b-PEO blends resulted from a curing reaction between PDMS oligomers and a curing agent. Although the mechanical property of the blend films is greatly improved compared to a PEDOT:PSS film, it is expected to fall behind a PDMS film since the mechanically weak PEDOT:PSS occupies the largest fraction of the blend films. Tensile tests were performed on a blend film (0.18 mm thick) and a pure PDMS sheet (0.5 mm thick) to examine this. A blend with the composition that showed the largest rupturing strain (∼75%) in Fig. 5(c) was chosen as a sample. Indeed, the mechanical properties of the blend film appeared to be worse than PDMS sheet in terms of ultimate tensile strength, strain at rupture (ε_max), and Young's modulus (E), as shown in Fig. S5.† The Young's modulus of the blend film (395 kPa) was relatively close to that of PDMS sheet (605 kPa) while gaps in the other aspects were much larger for the two samples. A discrepancy in rupturing strains of Fig. 5(c) and Fig. S5† might be caused from the fact that sharp probing tips used for four-probe measurement are prone to locally tear the blend film, facilitating earlier fracture in the electrical measurement than in the tensile test.

To evaluate the possibility for use as stretchable interconnects, contact pads and interconnects were fabricated using a polymer blend (PEDOT:PSS (300%)/PDMS (100%)/PDMS-b-PEO (30%)). The basic fabrication method is previously described (Fig. 1). As a first step, they were fabricated on a glass substrate (see the inset of Fig 6(a) for their dimensions). For comparison, interconnects were made of both the polymer blend and a Ag paste (conductivity: 7.5 S cm⁻¹), and their widths were fixed at 2 and 0.5 mm. As shown in Fig. 6(a), the I–V curves are linear over the measured range for both types of interconnects, indicating the polymer blend strips function well as an interconnecting conductor. The resistances of the interconnects were calculated from the slopes of the respective I–V curves. The resistances of the polymer blend interconnects were about an order of magnitude larger than those of Ag paste interconnects at the same dimensions, which arises from the conductivity difference of the two interconnect materials (see Fig. S6†). In the next step, the polymer blend contact pads and interconnects were fabricated on the PDMS substrate. The width of interconnects was 0.5 mm and all other dimensions were same as those depicted in Fig. 6(a). The I–V characteristics were measured under tensile strains up to 50% and the interconnect resistances were calculated from them. For the first straining cycle, the interconnect resistance (456 kΩ) was very close to that of the interconnect of an identical width formed on the glass substrate, reflecting the reproducibility of the polymer blend and the interconnect fabrication method. Upon the application of a strain, the resistance changes, but its changing rate is in general very slow (see the black guide line in Fig. 6(b)). Furthermore, when measured again after strain relaxation, the strain-dependent interconnect resistances almost replicate the first cycle results (see the red guide line in Fig. 6(b)). These results suggest that the polymer blend can be considered for practical use as organic interconnect.

Fig. 6 (a) I–V curves obtained from contact pad and interconnects made of a polymer blend (a 200% PEDOT:PSS, a 100% PDMS, a 30% PDMS-b-PEO). For comparison, interconnects were also fabricated using a Ag paste. The inset shows the dimensions of the contact pads and interconnects. The interconnect width was either 2 or 0.5 mm. (b) Interconnect resistances measured under varying strains (0 to 50%). The 0.5 mm wide interconnects were fabricated on a PDMS substrate using the same polymer blend as in (a) (see the inset). The resistances were calculated from the respective I–V curves and the resistance measurement was repeated after completing the first cycle of straining.

Conclusions

Polymer blends of PEDOT:PSS and PDMS that are naturally insoluble in each other were prepared using a block copolymer, PDMS-b-PEO, as a mediator. The PDMS-b-PEO, in which hydrophilic PEO side chains are grafted on PDMS backbones, significantly improved the miscibility of PEDOT:PSS with PDMS. At a 30% PDMS-b-PEO to PDMS by weight, PEDOT:PSS and PDMS appeared to be mixed well without noticeable phase separation, and beyond this concentration, a PDMS-b-PEO phase began to be separated from the polymer mixture. The average sheet resistance and its distribution of the cured PEDOT:PSS:PDMS blend films became smaller with increasing the concentrations of PDMS-b-PEO and PEDOT:PSS. At an optimal combination (200% PEDOT:PSS, 100% PDMS, 30% PDMS-b-PEO), the conductivity of the blend film became comparable with a pure PEDOT:PSS film and its strain to rupture reached about 75%, which is more than 6-fold larger than the PEDOT:PSS film. Contact pads and interconnects made of this optimal blend showed ohmic behaviors regardless of the pattern size and the substrate used, and could be reproducibly stretched within a 50% strain with a slight resistance change. These results illuminate a new pathway to the realization of stretchable organic interconnects, on which PEDOT:PSS:PDMS blends are a viable choice.

Acknowledgements

This work was supported by the Gachon University research fund of 2013 (GCU-2013-R345). The author thanks professor Kwang S. Suh of Korea University for his assistance.

References

1. D. H. Kim and J. A. Rogers, Adv. Mater., 2008, 20, 4887–4892.
2. J. H. Ahn and J. H. Je, J. Phys. D: Appl. Phys., 2012, 45, 103001.
3. T. H. Seah and M. Pumera, Sens. Actuators, B, 2011, 156, 79–83.
4. T. Someya, T. Sekitani, S. Iba, Y. Kato, H. Kawaguchi and T. Sakurai, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 9966–9970.
5. G. H. Gelinck, H. Edzer, A. Huitema, E. Veenendaal, E. Cantatore, L. Schrijnemakers, J. B. P. H. Putten, T. C. T. Geuns, M. Beenhakkers, J. B. Giesbers, B. H. Huisman, E. J. Meijer, E. M. Benito, F. J. Touwslager, A. W. Marsman, B. J. E. Rens and D. M. Leeuw, Nat. Mater., 2004, 3, 106–110.
6. J. Ouyang, C. W. Chu, C. R. Szmanda, L. Ma and Y. Yang, Nat. Mater., 2004, 3, 918–922.
7. S. P. Lacour, J. Jones, S. Wagner, T. Li and Z. Suo, Proc. IEEE, 2005, 93, 1459–1467.
8. S. P. Lacour, C. Tsay and S. Wagner, IEEE Electron Device Lett., 2004, 25, 792–794.
9. D. Y. Khang, H. Jiang, Y. Huang and J. A. Rogers, Science, 2006, 311, 208–212.
10. H. C. Ko, M. P. Stoykovich, J. Song, V. Malyarchuk, W. M. Choi, C. J. Yu, J. B. Geddes III, J. Xiao, S. Wang, Y. Huang and J. A. Rogers, Nature, 2008, 454, 748–753.
11. T. Someya, Y. Kato, T. Sekitani, S. Iba, Y. Noguchi, Y. Murase, H. Kawaguchi and T. Sakurai, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 12321–12325.
12. Y. Li, L. Zhao and H. Shimizu, Macromol. Rapid Commun., 2011, 32, 289–294.
13. W. Hu, X. Niu, L. Li, S. Yun, Z. Yu and Q. Pei, Nanotechnology, 2012, 23, 344002.
14. J. A. Rogers, T. Someya and Y. Huang, Science, 2010, 327, 1603–1607.
15. Y. P. Fu, R. A. Weiss, P. P. Gan and M. D. Bessette, Polym. Eng. Sci., 1998, 38, 857–862.
16. T. Sekitani, Y. Noguchi, K. Hata, T. Fukushima, T. Aida and T. Someya, Science, 2008, 321, 1468–1472.
17. O. Fenwick, K. Oliver and F. Cacialli, Appl. Phys. Lett., 2012, 100, 053309.
18. C. H. Chen, J. C. LaRue, R. D. Nelson, L. Kulinsky and M. J. Madou, J. Appl. Polym. Sci., 2012, 125, 3134–3141.
19. T. S. Hansen, K. West, O. Hassager and N. B. Larsen, Adv. Funct. Mater., 2007, 17, 3069–3073.
20. D. J. Lipomi, J. A. Lee, M. Vosgueritchian, B. C. K. Tee, J. A. Bolander and Z. Bao, Chem. Mater., 2012, 24, 373–382.
21. S. C. B. Mannsfeld, B. C. K. Tee, R. M. Stoltenberg, C. V. H. H. Chen, S. Barman, B. V. O. Muir, A. N. Sokolov, C. Reese and Z. Bao, Nat. Mater., 2010, 9, 859–864.
22. C. Pang, G. Y. Lee, T. I. Kim, S. M. Kim, H. N. Kim, S. H. Ahn and K. Y. Suh, Nat. Mater., 2012, 11, 795–801.
23. M. Jin, X. Feng, J. Xi, J. Zhai, K. Cho, L. Feng and L. Jiang, Macromol. Rapid Commun., 2005, 26, 1085–1089.
24. J. Zhou, A. V. Ellis and N. H. Voelcker, Electrophoresis, 2010, 31, 2–16.
25. M. Yao and J. Fang, J. Micromech. Microeng., 2012, 22, 025012.
26. D. C. Duffy, J. C. McDonald, O. J. A. Schueller and G. M. Whitesides, Anal. Chem., 1998, 70, 4974–4984.
27. J. Lee, W. Shim, E. Lee, J. S. Noh and W. Lee, Angew. Chem., Int. Ed., 2011, 50, 5301–5305.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3ra46087h

---