Flexible luminescent waveguiding photovoltaics exhibiting strong scattering effects from the dye aggregation

Flexible luminescent waveguiding photovoltaics exhibiting strong scattering effects from the dye aggregation

Nano Energy (2015) 15, 729–736 Available online at www.sciencedirect.com journal homepage: www.elsevier.com/locate/nanoenergy RAPID COMMUNICATION ...

3MB Sizes 1 Downloads 19 Views

Nano Energy (2015) 15, 729–736

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/nanoenergy

RAPID COMMUNICATION

Flexible luminescent waveguiding photovoltaics exhibiting strong scattering effects from the dye aggregation Chun-Hsien Choua,b, Min-Hung Hsub, Fung-Chung Chena,n a

Department of Photonics and Institute of Electro-Optical Engineering, National Chiao Tung University, Hsinchu 30010, Taiwan b Flexwave Company Limited, Hsinchu 30010, Taiwan Received 31 March 2015; received in revised form 18 May 2015; accepted 1 June 2015 Available online 10 June 2015

KEYWORDS

Abstract

Solar concentrator; Flexible Electronics; Waveguide; Luminescence; Photovoltaics

Luminescent solar concentrators have received renewed interest because they can harvest solar radiation without the need for expensive tacking systems. In this work, highly efficient luminescent waveguiding photovoltaic devices exhibiting mechanical flexibility and solar concentration ability have been prepared by integrating Si solar cells with soft polydimethylsiloxane (PDMS) waveguides. We observed that segregated dyes in the PDMS waveguides induce strong scattering effects in the long-wavelength range. The scattered photons were transported effectively to the solar cells, leading to high power conversion efficiencies (PCEs). A single module exhibited a remarkable power conversion efficiency of 4.6270.02%. After stacking two waveguides, we achieved PCEs as high as 5.2370.01%, with a projected PCE approaching 12%. & 2015 Elsevier Ltd. All rights reserved.

Introduction Photovoltaic (PV) technology is a promising approach toward supplying renewable energy and overcoming any potential worldwide energy crises. Wide acceptance of solar energy, n Corresponding author. Tel.: +886 3 5131 484; fax: +886 3 573 5601. E-mail address: [email protected] (F.-C. Chen).

http://dx.doi.org/10.1016/j.nanoen.2015.06.001 2211-2855/& 2015 Elsevier Ltd. All rights reserved.

however, has not occurred because of their high cost originated from the materials, the fabrication processes, and other soft expenses (installation, racking, permitting, etc.) [1]. One beneficial approach toward decreasing PV costs is to use solar concentrator systems to harvest more solar radiation over the limited area of a solar cell. Among the various types of PV concentrators, luminescent solar concentrators (LSCs) [2–8] have received renewed interest because they can harvest solar radiation without the need for expensive tacking systems. Typical LSCs feature a planar waveguide in which dye molecules

730 are dispersed to enhance the light absorption ability. After the thin plate receives direct and/or diffuse sunlight, the dyes converted the photonic energy to red-shifted photons; the emission is guided toward the solar cells attached to the sidewalls of the waveguide. As a result, sunlight is concentrated effectively in such a device configuration because the top surface area of the waveguide is usually much larger than that of the solar cells [2,7]. Weber and Lambe introduced the concept of LSCs in 1976 [2]; currently, the record efficiency has reached 7.1% for a device featuring four highly efficient GaAs solar cells attached to the edges of the concentrators [4]. The efficiencies of LSCs remain limited by several factors, including the limited absorption range, reabsorption of the luminophores, surface losses, and energy dissipation as heat during the red-shifting process [7,9–14]. Many approaches have been reported to overcome these obstacles and improve the efficiencies. For example, stacking dual waveguides containing different dyes has been suggested as a means to harvest broad band solar irradiation [15]. Patterning of the dye layer is also an effective method for minimizing reabsorption losses and increasing emission efficiencies [16,17]. Furthermore, selection and/or design of appropriate dyes is an important aspect of optimizing the performance of LSCs; for example, the harnessing of near-infrared (NIR) luminophores can aid the absorption of the NIR portion of the solar spectrum [18,19]. More recently, nanocluster phosphors exhibiting massive Stokes shifts have been adopted to harvest ultraviolet light selectively and decrease reabsorption losses simultaneously [20]. Moreover, the principle of aggregation-induced emission has been also used to design new fluorophores; their quantum efficiencies remain high in the solid state, thereby overcoming the problem of concentration quenching, regardless of whether the luminophores are dispersed in the polymer matrix or deposited as thin films [21]. Accordingly, it appears that further explorations of the fundamental properties of luminescent molecules will have great impact on our ability to prepare higher-efficiency LSCs. It is generally agreed that LSCs should complement Si cells, rather than become a competitive technology [22]. They have capability to be used in areas that obtain mostly diffuse light, where the efficiency of Si panels would drop significantly. Furthermore, the greater flexibility of LSCs, in terms of both color and shape, would potentially improve the esthetic appeal of solar cells, meaning that they could be positioned directly in a wider array of public areas. Nevertheless, the rigid substrates [e.g., poly(methyl methacrylate) (PMMA), high-density glass] adopted as waveguide components in conventional LSCs might complicate their fabrication and limit their applicable potentials. Previously, we reported high-performance flexible waveguiding photovoltaics (FWPVs) [23] that display outstanding flexibility because they use soft polydimethylsiloxane (PDMS) as the waveguide [24– 26]. The flexible module, which is equipped with a flexible bottom scattering reflector (BSR), can effectively concentrate diffuse light and direct the energy toward the cells. Moreover, other optical microstructures (e.g., microlenses) can be integrated readily into the waveguides to enhance their performance. In this present study, we incorporated organic dyes into a PDMS light guide to develop flexible luminescent waveguiding photovoltaics (FLWPVs, Figure 1). We have observed that segregated dyes in the PDMS waveguides induce strong scattering effects in the longwavelength range. Surprisingly, the scattered photons are

C.-H. Chou et al.

Figure 1 (a) Schematic representation of a FLWPV. The incident photons were absorbed and re-emitted, at a Stokesshifted wavelength, by the dye molecules; the luminescence and scattering were guided to the solar cells through total internal reflection in the waveguide. Scattering also occurred as a result of dye aggregation. (b) and (c) Photographs of FLWPV-2 modules doped with (b) C440 and (c) DSF. For the transparent module in (b), a high transparency could be obtained while retaining a PCE of approximately 1%. (d) Photograph of C440-doped transparent modules of various transparencies; from left to right, the concentrations of C440 were 87.12, 174.24, and 348.48 mg/L, respectively.

also transported effectively to the solar cells, leading to high power conversion efficiencies (PCEs). An FLWPV featuring a BSR delivered a PCE of 4.6270.02%. After stacking two waveguides, we achieved PCEs as high as 5.2370.01%, with a projected PCE approaching 12%. To the best of our knowledge, this efficiency is the highest ever reported for an LSC incorporating monocrystalline Si (m-Si) as its solar cells. We anticipate that this new scattering scheme might pave the way toward LSCs exhibiting even better performance.

Material and methods Figure 1a provides a schematic representation of the FLWPV module investigated in this study. We prepared the modules according to the integral molding method that we described previously [23]. The dimensions of the waveguide were fixed at 5.0 cm  5.0 cm  0.5 cm; we attached either two or four m-Si cells to the PDMS waveguide. For simplification, we name the modules incorporating four and two solar cells at the sidewalls

Flexible luminescent waveguiding photovoltaics as FLWPV-4 and FLWPV-2, respectively [23]. We blended two fluorescent dyes, coumarin 440 (C440) and disodium fluorescein (DSF), with the PDMS prepolymer and the curing agent during fabrication of the waveguides. Figure S1a and S1b (Supporting information) display the absorption and photoluminescence (PL) spectra of C440 and DSF, respectively, in EtOH. We observe that C440 absorbs mainly UV rays, while the absorption band of DSF is located in the region from 350 to 530 nm. Figure 1d reveals that the C440-doped modules exhibited high transparency at low levels of doping, with the transmittances decreasing as the concentration of C440 increased. Figure 2a and b displays the transmittance spectra of PDMS waveguides doped with C440 and DSF, respectively, at various doping levels. The average transmittance in the visible spectral region was much higher than 70% for the module prepared with 87.12 mg/L of C400; it remained at approximately 50% when the concentration increased to 348.48 mg/L. DSF absorbs photons strongly in the spectral range from 400 to 500 nm and emits orange photons; the emission spectrum matched quite well the wavelength region at which the m-Si solar cells exhibited higher efficiencies (Figure S1d, Supporting information). Upon increasing the concentration of DSF from 41.21 mg/L to 824.24 mg/L, the transparency decreased significantly (Figure S2a, Supporting information). Elastic polymer silicon (PDMS, Sylgrad 184, Dow Corning) was used to form the waveguides of the FLWPV modules. The organic dyes C440 and DSF were purchased from Exciton. The fabrication of FLWPV modules followed procedures reported previously [23]. In short, m-Si solar cells were first saw-cut into small pieces; the resulting cells exhibited PCEs of 15.7670.25%

731 (Figure S1c, Supporting information). The cells were fixed at the side walls of a 5 cm  5 cm acrylic mold. The dyes, PDMS prepolymer, and the curing agent were mixed well and then poured into the mold. The sample was cured at 90 1C for 20 min. The concentrations of the dyes were calculated from the weight ratio of the dyes and the volume of the PDMS prepolymer. For fabrication of the backside reflective layer, TiO2 powder was mixed with the PDMS pre-polymer at a weight concentration of 0.5 g mL  1; the resulting white mixture was poured onto the as-made PDMS waveguide layers. The thickness of the BSR was controlled at 0.10 cm. Finally, the solid FLWPVs were peeled from the mold. The surface of the resulting PDMS slab was characterized by using an Alpha-Step Profilometer. As shown in Figure S3 (Supporting information), the root-meansquare roughness of the PDMS surface was 14.587 nm. As a result, no apparent scattering effect was observed at the surface. The current–voltage (I–V) characteristics were measured under AM 1.5 G illumination (100 mW cm–2) using a Xe-based class-A solar simulator. Module certification (Supporting information) was performed at the Center for Measurement Standards of ITRI, certified (IEC 61215 CBTL) by the International Electrotechnical Commission System for Conformity Testing and Certification of Electrotechnical Equipment and Components. EQE spectra were recorded using an Enli Technology measuring system. The optical efficiency (ηopt) was calculated using the Equation [11, 23, 27]  PCE cell  Pout =0:1 W cm  2  Amodule PCE module Pout ¼ ¼ ¼ ηopt PCE cell PCE cell P in

ð1Þ

Figure 2 (a) and (b) Transmittance spectra of waveguides containing various concentrations of (a) C440 and (b) DSF. (c) and (d) EQE spectra of FLWPVs featuring (c) C440-and (d) DSF-doped waveguides. Note that both materials exhibited a scattering contribution in the NIR region.

732 where Pmodule and Pcell are the PCEs of the module and the solar cell, respectively; Pout and Pin are the output and input photon fluxes, respectively; 0.1 W/cm is the illumination of 1 sun under AM 1.5 G conditions; and Amodule is the area of the incident surface of the whole module. In this study, 5 cm  0.5 cm solar cells were employed to construct the FLWPVs with a fixed volume (5 cm  5 cm  0.5 cm); hence, the geometry factor (Cg) was constant throughout these measurements (Cg = 2.5 and 5 for FLWPV-4 and FLWPV-2, respectively).

Results and discussion Figure 3a and b displays the I–V characteristics of the modules containing various concentrations of C440 and DSF, respectively, under standard illumination conditions (AM 1.5, 100 mW cm-2). The short-circuit current (Isc) of the C440 FLWPV-4 module increased from 7.8170.03 to 15.6270.05 mA upon increasing the dopant concentration from 87.12 to 348.48 mg/L, presumably because of an increase in the amount of trapped photons in the waveguides. The PCE of the transparent C440 module (87.12 mg/L) was 0.4470.01%; its value of Isc, open-circuit voltage (Voc), and fill factor (FF) were 7.8170.03 mA, 1.9470.01 V, and 0.7170.01, respectively. The PCE increased to 0.9070.01% for the semi-transparent module in which the concentration was 348.48 mg/L. For the DSF FLWPV-4 modules, the PCE was highest when the concentration was 824.24 mg/L; the values of Isc, Voc, and FF were 38.7070.30 mA, 2.287 0.01 V, and 0.6570.01, respectively, resulting in a PCE of 2.3170.01%. The photocurrent of the DSF FLWPV-4 module

C.-H. Chou et al. improved notably, to 69.8470.50 mA, after integration with a TiO2-doped BSR to decrease the escaping efficiency of photons from the bottom surface of the waveguide [23]. The module exhibited a value of Voc of 2.4070.03 V and a FF of 0.6970.01, resulting in a very high PCE of 4.6270.02%. The calculated optical efficiency (ηopt) was approximately 30%. To further explore the optical behavior of the waveguides and investigate the mechanism behind the high efficiencies, we measured the EQE spectra of the FLWPVs prepared at various doping levels (Figures 2c and d). The C440 module exhibited its peak efficiency at 350 nm when the doping level was low; the quantum efficiency dropped steadily upon increasing the doping concentration presumably because of enhanced reabsorption losses. Notably, however, a broad band existed in the range from 700 to 1100 nm, and its intensity increased gradually upon increasing the doping concentration. Because C440 does not absorb in this spectral region, we suspect that the increased efficiencies in the NIR region were due to aggregation of the dopants [28–30]. We observed similar behavior for the DSF modules (Figure 2d). At a low doping concentration of 266.06 mg/L, small contribution of the photocurrent was still observed in the spectral range from 300 to 530 nm from the magnified EQE spectrum (Figure S4a, Supporting information), which was due to the absorption of the DSF molecules. However, the EQE values dropped almost to zero when the DSF concentration exceeded 412.12 mg/L in the same spectral region. Meanwhile, the EQE spectra featured a broad band in the NIR region (from 550 to 1100 nm) after higher amount of dyes were doped. Similarly, such spectral feature indicated that the dye molecules might start to aggregate, leading to significant scattering effects. Initially,

Figure 3 I–V curves of (a) and (b) FLWPV-4 modules prepared with C440 and DSF at various concentrations, (c) stacked FLWPVs (inset: configuration of the stacked module), and (d) the stacked FLWPV-4 module, exhibiting a maximum output power of 122.82 mW, corresponding to a PCE of 4.9270.09%.

733

0.4470.01 0.6970.01 0.9070.01 0.4670.01 1.0370.02 1.8870.02 2.3170.02 2.2870.02 4.6270.02 0.3770.02 1.1870.01 5.2370.02 (4.9270.09) 4.4870.02 0.7170.01 0.6770.01 0.6970.01 0.6770.01 0.6870.01 0.6870.01 0.6570.01 0.6570.01 0.6970.01 0.7270.02 0.6970.01 0.6870.01 (0.68770.013) 0.6970.01 7.8170.03 12.9070.10 15.6270.05 8.8770.04 17.5770.30 30.4670.30 38.7070.30 38.5470.30 69.8470.50 11.8270.05 38.0270.30 83.1770.50 (76.3071.4) 72.0470.50 ( ): ITRI certified numbers.

C440 FLWPV-2 DSF FLWPV-2 Stacked FLWPV-4

DSF FLWPV-4

87.12 (A) 174.24 (B) 348.48 41.21 206.06 412.12 824.24 2060.61 (C) 824.24 +BSR 348.48 824.24 (A) +(C) (B) +(C) C440 FLWPV-4

1.9470.01 2.0070.01 2.1170.01 1.9370.01 2.1570.02 2.2670.02 2.2870.01 2.2670.02 2.4070.03 1.0770.01 1.1270.01 2.3070.02 (2.343470.0087) 2.2670.02

Concentration (mg/L) Module

Figure 4 Optical microscopy images of PDMS waveguides containing DSF at concentrations of (a) 206.06 and (b) 824.24 mg/L.

Table 1

Performance characteristics of FLWPVs.

Voc (V)

Isc (mA)

FF

PCE (%)

ηopt (%)

the EQEs in the NIR region increased upon increasing the doping concentration, reaching the highest values when the concentration was 824.24 mg/L and decreasing thereafter. Further, we also noted that the EQE onset red-shifted when the DSF concentration was increased (Figure S4a, Supporting information). The onset was located at 500 nm at a low concentration of 266.06 mg/L; it shifted to ca. 530 nm at a higher doping concentration of 824.24 mg/L. Figure S4b, Supporting information displays the absorption and PL spectra of DSF. The spectral overlap, which can cause apparent reabsorption, existed in the wavelength range from 500 to 530 nm. Therefore, we infer that the red-shift of the EQE onset should be resulted from the increasing reabsorption losses. Figure 2d also shows the EQE spectrum of the module equipped with a back reflector. The quantum efficiencies increased further after incorporating the TiO2-doped back reflectors into the modules, due to the recovery of the photons directly transported through the waveguides; the maximum EQE reached as high as 29% at 626 nm. The absorption spectra of the PDMS slabs doped with various dye concentrations were also obtained (Figure S5, Supporting information). We observe that the absorbance values in the long wavelength region indeed increased with the increasing dye concentrations. Therefore, the results further support our previous hypothesis that dopants aggregated in the PDMS waveguides. To confirm that scattering contributed the greater efficiencies, we used optical microscopy to investigate the aggregation

2.79 4.38 5.71 2.92 6.54 11.93 14.66 14.47 29.31 2.35 7.49 33.19 (31.21) 28.43

Flexible luminescent waveguiding photovoltaics

734 behavior of the organic dyes. Figure 4 reveals that both the density and dimensions of the particles increased upon increasing the concentration of DSF; we estimated average particle sizes of 8.5 and 12.8 μm in the DSF modules prepared at concentrations of 206.06 and 824.24 mg/L, respectively. The results were similar for the waveguides doped with C440 (Figure S6, Supporting information). Because PDMS has very low surface energy, serious segregation of the organic dyes and PDMS occurred after solidification of the luminescent waveguides. Therefore, the luminescent molecules tended to aggregate in the solid PDMS waveguide, resulting in significant scattering effects. As illustrated in the inset to Figure 3c, we examined the effect of stacking a semi-transparent C440 FLWPV-4 module onto a DSF module equipped with a BSR. A small air gap existed between the two waveguides; the solar cells were connected in parallel to bypass the current-matching limitation. Figure 3c displays the I–V curves of stacked modules in which two representative concentrations of C440 were used in the front waveguide. The lower photocurrent was that of the stacked module containing C440 at 348.48 mg/L, possibly because of lower transmittances of the front waveguide blocked the solar radiation from reaching the back waveguide. The resulting PCE was 4.4870.02%, slightly lower than that of the single DSF module equipped with a BSR. When we decreased the concentration of C440, the photocurrent of the stacked module improved, presumably because of better optical matching between the modules. In this case, the PCE was high (5.2370.02%); the calculated value of ηopt was 33.19%. The EQE spectra of the stacked module revealed photocurrent contributions from the two different wavelength regimes (Figure S7, Supporting information). We had the stacked module certified at the Center for Measurement Standards, Industrial Technology Research Institute (ITRI) of Taiwan. As revealed in Figure 3d, the certified I–V curve exhibited a value of Voc of 2.343470.0087 V, a value of Isc of 76.3071.4 mA, and an FF of 0.68770.013, resulting in a PCE of 4.9270.09%. The complete certification report is included in Supporting information. To the best of our knowledge, this PCE is a new record for a luminescent concentrator system incorporating mono-Si cells. The predicted module efficiency would presumably reach close to 12% if we had used solar cells of higher efficiencies. Table S1 (Supporting information) lists the predicted PCEs for such modules prepared with different kinds of solar cells. Table 1 summarizes the performance characteristics of the various modules.

C.-H. Chou et al. To test their ability to harvest photons incident at high angles, we investigated the performance of our FLWPVs at various angles of incidence [6,31,32]. Figure 5a displays the normalized optical efficiencies and photocurrents of FLWPV-4 at incident angles ranging from 0 to 901; Figure S8a (Supporting information) presents the corresponding I–V characteristics. The calculated value of ηopt remained greater than 70% when the incident angle reached at high as 701, indicating that the concentrator could indeed harvest sufficient photons with high incident angles without the need for a mechanical tracker. Because of the excellent mechanical property of PDMS, the FLWPV modules exhibited high flexibility. The four-cell modules could be bent only slightly because the soft PDMS somehow protected the Si solar cells embedded in the waveguides. In contrast, the two-cell FLWPVs exhibited much better mechanical bendability. Figure 5b displays the results of bending tests [23,33–35] of a DSF FLWPV-2 module. We used a bending apparatus that could fix the modules at bending radii of 20, 30, 40 and 50 mm; the original PCE of the flat DSF FLWPV-2 was 1.18% (Figure S8b and S8c, Supporting information). Unexpectedly, the PCEs increased under the convex bending conditions, presumably because, at larger curvature, the edge cells became tilted more to the direction of incidence, thereby receiving more direct solar radiation. For bending radii of 20, 30, 40, and 50 mm, the values of Isc were 51.1970.50, 50.6270.50, 49.3270.50, and 47.8070.45 mA, respectively; the corresponding module efficiencies were 1.6170.02, 1.5870.02, 1.5370.02, and 1.4970.02%, respectively (Figure S8d, Supporting information). Additionally, we also provided an illumination mask that is slightly smaller than the 5 cm  5 cm waveguide to prevent the direct illumination under the convex bending conditions; the measurement results are displayed in Figure S9 (Supporting information). We can see the values of Isc were indeed smaller than those obtained without using the mask, revealing that the photocurrent increment was due to somehow direct illumination under the bending conditions. Table S2 (Supporting information) further summarizes the results of the bending test.

Conclusions In summary, we have found a new scattering scheme in LSCs that led, unexpectedly, to high PCEs. In the waveguides, blended organic dye molecules that aggregated in the PDMS matrix behaved as strong scattering centers. Such a module

Figure 5 (a) Normalized optical efficiencies at various incident angles. (b) PCEs and values of Isc of the FLWPV-2 module when bent at various radii.

Flexible luminescent waveguiding photovoltaics featuring a single waveguide exhibited a very high PCE of 4.6270.02%. Moreover, a stacked FLWPV achieved a certified record efficiency of 4.9270.09% when the LSCs incorporated m-Si cells; projected PCEs reached as high as 12%. Furthermore, these FLWPV modules can be made highly transparent, with UV-blocking ability, yet exhibiting a PCE of close to 1%. In addition to use as building-integrated photovoltaics, these colorful flexible LSC modules open up great possibilities for other applications (e.g., integrated wearable devices [36] or clothing-integrated photovoltaics [37]).

Acknowledgments We thank the Ministry of Science and Technology (Grant nos. MOST 103–2218-E-009–034-MY3, MOST 102–2221-E-009–130MY3, and MOST 101–2628-E-009-008-MY3) and the Ministry of Education of Taiwan (through the ATU program) for financial support.

Appendix A.

Supporting information

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/ j.nanoen.2015.06.001.

References

735

[19] [20] [21] [22] [23] [24] [25]

[26] [27] [28] [29] [30] [31] [32]

[33]

[34] [35]

[1] S.B. Darling, F. You, T. Veselka, A. Velosa, Energy Environ. Sci 4 (2011) 3133–3139. [2] W.H. Weber, J. Lambe, Appl. Opt. 15 (1976) 2. [3] J.A. Levitt, W.H. Weber, Appl. Opt. 16 (1977) 2684–2689. [4] L.H. Slooff, E.E. Bende, A.R. Burgers, T. Budel, M. Pravettoni, R.P. Kenny, E.D. Dunlop, A. Büchtemann, Phys. Status Solidi – R 2 (2008) 257–259. [5] F. Purcell-Milton, Y.K. Gun'ko, J. Mater. Chem. 22 (2012) 16687–16697. [6] A. Kerrouche, D.A. Hardy, D. Ross, B.S. Richards, Sol. Energy Mater. Sol. Cells 122 (2014) 99–106. [7] A.F. Mansour, Polym. Test. 17 (1998) 333–343. [8] M.J. Currie, J.K. Mapel, T.D. Heidel, S. Goffri, M.A. Baldo, Science 321 (2008) 226–228. [9] H. Hernandez-Noyola, D.H. Potterveld, R.J. Holt, S.B. Darling, Energy Environ. Sci. 5 (2012) 5798–5802. [10] M.G. Debije, P.P.C. Verbunt, B.C. Rowan, B.S. Richards, T. L. Hoeks, Appl. Opt. 47 (2008) 6763–6768. [11] W.R.L. Thomas, J.M. Drake, M.L. Lesiecki, Appl. Opt. 22 (1983) 3440–3450. [12] L.R. Wilson, B.C. Rowan, N. Robertson, O. Moudam, A. C. Jones, B.S. Richards, Appl. Opt. 49 (2010) 1651–1661. [13] F. Meinardi, A. Colombo, K.A. Velizhanin, R. Simonutti, M. Lorenzon, L. Beverina, R. Viswanatha, V.I. Klimov, S. Brovelli, Nat. Photon 8 (2014) 392–399. [14] N.C. Giebink, G.P. Wiederrecht, M.R. Wasielewski, Nat. Photon 5 (2011) 694–701. [15] L. Desmet, A.J.M. Ras, D.K.G. de Boer, M.G. Debije, Opt. Lett. 37 (2012) 3087–3089. [16] S. Tsoi, D.J. Broer, C.W. Bastiaansen, M.G. Debije, Opt. Exp. 18 (2010) A536–A543. [17] S. Tsoi, D.J. Broer, C.W.M. Bastiaansen, M.G. Debije, Adv. Energy Mater. 3 (2013) 337–341. [18] A. Sanguineti, A. Monguzzi, G. Vaccaro, F. Meinardi, E. Ronchi, M. Moret, U. Cosentino, G. Moro, R. Simonutti, M. Mauri,

[36]

[37]

R. Tubino, L. Beverina, Phys. Chem. Chem. Phys. 14 (2012) 6452–6455. Y. Zhao, G.A. Meek, B.G. Levine, R.R. Lunt, Adv. Opt. Mater. 2 (2014) 606–611. Y. Zhao, R.R. Lunt, Adv. Energy Mater. 3 (2013) 1143–1148. J.L. Banal, J.M. White, K.P. Ghiggino, W.W.H. Wong, Sci. Rep. 4 (2014) 4635. M.G. Debije, P.P.C. Verbunt, Adv. Energy Mater. 2 (2012) 12–35. C.H. Chou, J.K. Chuang, F.C. Chen, Sci. Rep. 3 (2013) 2244. A. Mata, A. Fleischman, S. Roy, Biomed. Microdevices 7 (2005) 281–293. W.-M.P.F. Bosman, T.J. van der Steenhoven, D.R. Suárez, E. R. Valstar, A.C. de Vries, H.L.F. Brom, M.J. Jacobs, J. F. Hamming, J. Vasc. Surg. 52 (2010) 152–158. X. Zhao, L. Li, B. Li, J. Zhang, A. Wang, J. Mater. Chem. A 2 (2014) 18281–18287. J. Roncali, F. Garnier, Appl. Opt. 23 (1984) 2809–2817. F.Z. Haque, M. Husain, Optik 125 (2014) 2308–2312. S.L.L. Soltzberg, N. Okey-Igwe, R. Newman, Am. J. Anal. Chem. 3 (2012) 622–631. R.J. Clarke, A. Oprysa, J. Chem. Educ. 81 (2004) 705. J.M. Kim, P.S. Dutta, Sol. Energy Mater. Sol. Cells 103 (2012) 35–40. T. Uematsu, Y. Yazawa, Y. Miyamura, S. Muramatsu, H. Ohtsuka, K. Tsutsui, T. Warabisako, Sol. Energy Mater. Sol. Cells 67 (2001) 415–423. C. Roldan-Carmona, O. Malinkiewicz, A. Soriano, G. Minguez Espallargas, A. Garcia, P. Reinecke, T. Kroyer, M.I. Dar, M. K. Nazeeruddin, H.J. Bolink, Energy Environ. Sci. 7 (2014) 994–997. T.-H. Han, Y. Lee, M.-R. Choi, S.-H. Woo, S.-H. Bae, B.H. Hong, J.-H. Ahn, T.-W. Lee, Nat. Photon 6 (2012) 105–110. B.-J. Kim, d.H. Kim, Y.-Y. Lee, H.-W. Shin, G.S. Han, J.S. Hong, K. Mahmood, T. Ahn, Y.-C. Joo, K.S. Hong, N.-G. Park, S. Lee, H.S. Jung, Energy Environ. Sci. 8 (2015) 916–921. P.D. Mitcheson, Engineering in medicine and biology society, in: Proceedings of the Annual International Conference of the IEEE, 2010, pp. 3432–3436. M.B. Schuber, Y. Ishikawa, J.W. Kramer, C.E.M. Gemmer, J.J. H. Werner, Photovoltaic specialists conference, in: Proceedings of the Conference Record of the Thirty-first IEEE, 2005, pp. 1488–1491.

Chun-Hsien Chou received his M.S. degree from Institute of Physics in National Taiwan Normal University (2008). Prior to join National Chiao Tung University (NCTU) in 2011, he worked in PV industry as a R&D engineer for two years. He is currently a doctoral candidate in the group of Prof. F. C. Chen at Institute of Electro-Optical Engineering in NCTU and interests in the field of organic solar cells and flexible waveguiding photovoltaics (FWPV); he is also the founder of Flexwave Co., Ltd. for realizing the FWPV technique. Min-Hung Hsu received his B.S. and M.S. degrees in Physics from National Taiwan Normal University, Taiwan, in 2002 and 2007, respectively. He has two year working experience in a crystalline-Si solar cells company as a R&D engineer. In 2013, he joined the Ph.D. program in Heriot-Watt University, United Kingdom, focusing on the research of solid state dye-sensitized solar cells and Perovskites solar cells. Currently, he is a co-founder of Flexwave Co., Ltd.

736

C.-H. Chou et al. Fang-Chung Chen is a professor in Department of Photonics at National Chiao Tung University. He received his B.S. and M.S. degrees in Chemistry from National Taiwan University in 1996 and 1998, respectively, and his Ph.D. degree in Materials Science and Engineering from University of California, Los Angeles, USA, in 2003. His research interests include flexible solar cells, organic electronics, and low-dimentional nanomaterials.