A viable transfer printing-based methodology to introduce plasmonic metal nanostructures in solar cells is described. Using nanopillar poly(dimethylsiloxane) stamps, an Ag-based ordered nanodisk array was integrated with standard hydrogenated microcrystalline Si solar cells, which led to improved device performances due to plasmonic light trapping.
One of the potential applications of metal nanostructures is light trapping in solar cells, where unique optical properties of nanosized metals, commonly known as plasmonic effects, play an important role. Research in this field has, however, been impeded owing to the difficulty of fabricating devices containing the desired functional metal nanostructures. In order to provide a viable strategy to this issue, we herein show a transfer printing-based approach that allows the quick and low-cost integration of designed metal nanostructures with a variety of device architectures, including solar cells. Nanopillar poly(dimethylsiloxane) (PDMS) stamps were fabricated from a commercially available nanohole plastic film as a master mold. On this nanopatterned PDMS stamps, Ag films were deposited, which were then transfer-printed onto block copolymer (binding layer)-coated hydrogenated microcrystalline Si (µc-Si:H) surface to afford ordered Ag nanodisk structures. It was confirmed that the resulting Ag nanodisk-incorporated µc-Si:H solar cells show higher performances compared to a cell without the transfer-printed Ag nanodisks, thanks to plasmonic light trapping effect derived from the Ag nanodisks. Because of the simplicity and versatility, further device application would also be feasible thorough this approach.
There has been a long-standing demand for the application of functional nanostructures in a broad range of technological field. One of the expectations for this trend is to open new design of device architectures leading to improved or innovative performances. In the field of solar cells, for example, the use of metal nanostructures has been actively explored because of their intriguing optical (i.e., plasmonic) properties,1 potentially beneficial to construct effective light trapping systems.2,3 Indeed, some theoretical studies4-6 have suggested that such plasmonic light trapping could achieve effects exceeding the conventional ray optics (texturing)-based light trapping limit.7 As a result, developing strategies to integrate desired metal nanostructures with solar cells has become increasingly important in order to realize these theoretical predictions.
A number of strategies have been proposed to meet this challenge.8-24 These include, for instance, simple (low-cost) thermal annealing of metal films8,9 or dispersion of pre-synthesized metal nanoparticles,10,11 both of which resulted in successful demonstrations of plasmonic light trapping. However, it should be pointed out that the metal nanostructures fabricated by these approaches are usually challenging to match to the theoretical models. In contrast, the traditional nanofabrication techniques in semiconductor industries, such as photolithography and electron beam lithography,12,13 can control structures well below the sub-100 nm level, but they are often too expensive and time-consuming to apply to solar cells, where large-area capability with low cost is essential. In order to fulfill the low-cost, high-throughput, and large-area requirements with nanoscale controllability, methods such as nanoimprint lithography,14-16 soft lithography,17,18 nanosphere lithography,19-21 and hole-mask colloidal lithography22-24 would be promising. Among these choices, we have developed a soft lithographic, advanced transfer printing technique.25 Using a nanostructured poly(dimethylsiloxane) (PDMS) stamps and block copolymer-based adhesive layers, patterning of ordered metal nanostructures could be readily achieved on a number of technologically relevant materials, including the ones for solar cells.
The focus of this article is to describe the detailed procedure of our transfer printing approach to incorporate effective light trapping plasmonic nanostructures in existing solar cell structures. As a demonstrative case, Ag nanodisks and thin-film hydrogenated microcrystalline Si (µc-Si:H) solar cells were selected in this study (Figure 1),26 although other types of metals and solar cells are compatible with this approach. Together with its process simplicity, the approach would be of interest to diverse researchers as a handy tool to integrate functional metal nanostructures with devices.
1. Preparation of PDMS Stamps
2. Preparation of Block Copolymer Solutions
3. Preparation of µc-Si:H Substrates
4. Ag-coating of PDMS Stamps
5. Transfer Printing of Ag Nanodisks on Thin-film Si Surfaces
6. Completion of Thin-film Si Solar Cell Fabrication
7. Measurement of External Quantum Efficiency (EQE)
8. Measurement of Photovoltaic Current-Voltage (J-V) Characteristics
Figure 2 outlines the general process for the transfer printing of Ag nanodisks on the surface of µc-Si:H (n layer). Briefly, an Ag film (thickness: 10-80 nm) is first deposited on the surface of a nanopillar PDMS stamp by electron beam evaporation. In parallel, a PS-b-P2VP solution is spin-coated on the surface of a freshly prepared µc-Si:H n layer. Subsequently, a droplet of EtOH is placed on the PS-b-P2VP-coated surface, and the Ag-deposited PDMS stamp is placed on the EtOH-wet PS-b-P2VP surface. No pressing is necessary to the stamp, because an intimate contact between the stamp and substrate spontaneously forms due to the surface tension derived from the evaporation of the EtOH. After the EtOH is evaporated away (using a reduced pressure), the stamp is released from the substrate to complete the transfer of the Ag deposited on the raised region of the nanopillar PDMS stamp. Finally, an Ar plasma treatment is carried out to remove the PS-b-P2VP coating.
Shown in Figure 3 are scanning electron microscopy (SEM) images of the resulting Ag nanodisk array on (in) µc-Si:H surfaces (cells).26 Figure 3A and 3B are the top and tilted views of a same sample. The root mean square roughness (Rrms) of the underlying µc-Si:H surface was 6.6 nm; nevertheless, nearly complete transfer of the Ag nanodisks, whose diameter, center-to-center distance, and the thickness of the Ag nanodisks were 200, 460, and 40 nm, respectively, was achieved. Figure 3C is the cross-sectional view of a completed µc-Si:H cell structure; i.e., after the deposition (sputtering) of ZnO:Ga/Ag/ZnO:Ga layers on top of the Ag nanodisk array shown in Figure 3A and 3B. The embedded Ag nanodisks right on the photovoltaic µc-Si:Hp–i–n layers were clearly observed.
The EQE spectra of the fabricated cells are shown in Figure 4.26 Compared to a reference cell (a µc-Si:H cell fabricated simultaneously with skipping the transfer printing process), the EQE spectrum of the Ag nanodisk (thickness: 40 nm) incorporated cell showed higher signals in the long wavelength range (650-1,100 nm). Such wavelength-selective enhancement clearly indicated the preferential effect of the plasmonically-active Ag nanodisks for solar cell; namely, plasmonic light trapping. Quantification of the degree of the light trapping in the 650-1,100 nm range observed in Figure 3 was carried out by summing the EQE values of each cell and taking the ratio of them (Ag nanodisk-incorporated cell/reference cell). The value was 1.60; therefore, 60% EQE increase was achieved by the Ag nanodisk-mediated plasmonic light trapping.
Table 1 summarizes the photovoltaic characteristics of the Ag nanodisk-incorporated and reference cells.26 It was confirmed, as expected, that the short-circuit current density (Jsc) of the Ag nanodisk-incorporated cell increased compared to that of the reference cell (11.4 to 12.4 mA/cm2) because of the EQE enhancement described above. As for the open-circuit voltage (Voc) and fill factor (FF), those of the two cells were almost the same (Voc: ~0.52 V, FF: ~0.76). As a consequence, the photoconversion efficiency (η) of the Ag nanodisk-incorporated cell improved (4.5% to 5.0%).
Figure 1. Schematic Cross-Section of a Ag nanodisk-incorporated µc-Si:H solar cell. The Ag nanodisks locates at the rear side of the µc-Si:H solar cell. Please click here to view a larger version of this figure.
Figure 2. General Procedure for the Transfer printing of Ag Nanodisks. The Ag-coated nanopillar PDMS stamp is applied onto the PS-b-P2VP-coated thin-film Si substrate whose surface is wet with EtOH. Ar plasma treatment is applied to remove the PS-b-P2VP coating and to expose the thin-film Si layer. After this process, ZnO:Ga/Ag/ZnO:Ga layers need to be deposited on top of the transfer-printed Ag nanodisks to complete the entire solar cell structure shown in Figure 1. Please click here to view a larger version of this figure.
Figure 3. SEM Images of Transfer-Printed Ag Nanodisks (diameter = 200 nm, thickness =40 nm). (A) Top view of transfer-printed Ag nanodisks on a µc-Si:H surface. (B) Tilted view of transfer-printed Ag nanodisks on a µc-Si:H surface. (C) Cross-sectional view of transfer-printed Ag nanodisks (size: 200 nm) embedded in a µc-Si:H cell.26 Copyright 2014 The Japan Society of Applied Physics. Please click here to view a larger version of this figure.
Figure 4. EQE Spectra of µc-Si:H cells. (Blue line) Ag nanodisk (ND)-incorporated cell. (Red dashed line) Reference cell.26 Improved responses were observed with the blue line owing to the Ag ND-mediated light trapping. This figure has been modified from Ref. 26. Please click here to view a larger version of this figure.
Table 1. Summary of Photovoltaic J-V Characteristics of µc-Si:H cells.
Cell type | Jsc (mA/cm2) | Voc (V) | FF | η (%) |
Ag nanodisk-incorporated | 12.4 | 0.526 | 0.764 | 5 |
Reference | 11.4 | 0.521 | 0.763 | 4.5 |
Table 2. Deposition conditions.
Steps | Deposition system | Materials | Conditions |
3.2 | Sputtering | ZnO:Ga | Ar flow rate = 200 sccm, pressure = 0.133 Pa, DC power = 200 W, sample rotation = 10 rpm, deposition rate = ~3.3 Å/sec. |
3.3 | CVD | µc-Si:H p | Flow rate of SiH4/H2/B2H6 = 3.5/450/2 sccm, substrate temperature = 140 °C, pressure = 1.5 torr, radio frequency (RF) power density = 80 mW/cm2, deposition time = 5 min 45 sec (rate = ~0.3 Å/sec). |
µc-Si:H i | Flow rate of SiH4/H2 = 10.5/380 sccm, substrate temperature = 180 °C, pressure = 200 Pa, RF power density = 40 mW/cm2, deposition time = 1 hr 2 min (rate = ~1.3 Å/sec). | ||
µc-Si:H n | Flow rate of SiH4/H2/PH3 = 3/148/12 sccm, substrate temperature = 195 °C, pressure = 40 Pa, RF power density = 80 mW/cm2, deposition time = 23 min (rate = ~0.3 Å/sec). | ||
6.2 | Sputtering | ZnO:Ga | Ar flow = 200 sccm, pressure = 0.133 Pa, DC power = 200 W, sample rotation = 10 rpm, deposition rate = ~3.3 Å/sec. |
Ag | Ar flow = 200 sccm, pressure = 0.133 Pa, DC power = 100 W, sample rotation = 10 rpm, deposition rate = ~6 Å/sec. |
In this article, a double-layered hard/soft PDMS composite was employed as stamp materials.27 This combination was found to be essential to precisely replicate the parent nanostructure in the mold, which was a hexagonally close-packed round-hole array whose diameter of 230 nm, depth of 500 nm, and hole center-to-center spacing of 460 nm. When only soft PDMS was used, the stamp always resulted in a poorly nanostructured surface (for example, no sharp edge in the inverted pillar structure) due to the low Young’s modulus;28 therefore, transfer printing of Ag nanodisks was never achievable.
The use of spin-coated block copolymer (PS-b-P2VP) thin-films as binding layers is another key for successful transfer printing on the µc-Si:H surfaces, which are not smooth (Rrms = ~6.6 nm). Although the transfer-printing of metal structures was originally developed using small organic molecules which forms self-assembled monolayers (SAMs) on surfaces,29 we found that the use of SAMs (3-mercaptopropyltrimethoxysilane) does not work for the (slightly) textured µc-Si:H surfaces. In addition, the formation of SAMs with good quality takes time (~a couple of hours), while our process requires less than 1 min (40 sec by spin coating). This point might be important for materials which need quick operations to avoid unfavorable surface events, such as the formation of excess oxides or contaminations.
The roles of EtOH in transfer printing should also be emphasized. The first role is, as already described, to assist the formation of the intimate contact between the stamp and target surface using the surface tension upon evaporation. The second role is to reconstruct the PS-b-P2VP thin film, which ensures the capturing of metals (Ag) on PDMS stamps through the formation of metal-pyridine coordination bonds.25 We believe that such dynamic event at the stamp/substrate interface is essential for the transfer printing especially on textured surfaces.
When the transfer printing by above procedure was unsuccessful, the reason was mostly in stamps employed. Since conformal contact between a stamp and a substrate is critical, the flatness of the stamp’s surface is very important. The flatness is determined by the status of an original mold; therefore, once transfer printing fails, it would be the time for changing the mold. According to our experience, the number of times a mold can be reused is five, but it would be increased by appropriate cleaning and storing of the mold. Flatness of a stamp would also have a significant impact on the area that can be uniformly transfer-printed. In this regards, we have demonstrated a 20 mm × 20 mm-scale patterning with a stamp prepared from a fresh mold.25
As for the pattern design of the mold/stamp, round-hole/pillar structures with the diameter of 230/200 nm were employed (the smaller diameter of the pillar is due to the tapered shape of the original hole). This choice was simply because such a mold (nanoimprinted plastic film) was commercially available, and it was not the one fabricated specifically to our solar cell application. This in turn means that there is plenty of room for pattern designs, which would lead to much superior light trapping ability compared to the result shown here. In this respect, the use of optical simulation would be helpful to search better patterns. Although the fabrication of actual molds (probably by electron beam lithography) can be expensive, once fabricated, the corresponding stamps can be replicated as many times as needed. Thus, the total process cost can be significantly suppressed, which is the great advantage of transfer printing approaches.
In terms of the metals transfer-printable by the procedure described here, Au, Cu, and Ni were confirmed to be applicable. It should be mentioned that none of these metals provided better light trapping effects compared to the case of Ag. Another metal tested was Al, which is considered as a good candidate for plasmonic light trapping applications.30 It was found, however, that transfer printing of Al is unsuccessful, possibly due to the strong affinity with PDMS. Therefore, modifications of PDMS surfaces31 may be required to facilitate transfer printing by reducing the interaction between PDMS and deposited Al.
Other than µc-Si:H, the method can be used with a variety of materials, including highly rough (textured) surfaces (Rrms≥ 20 nm).25 In fact, we have previously addressed the possibility of synergistic texturing/plasmon-mediated light trapping by fabricating cells with textured glass/SnO2:F substrates.17 In addition, a similar light trapping effect has been documented using hydrogenated amorphous Si cells.32 Other technologically important materials, such as crystalline Si, GaAs, InP, and metal oxides, are also compatible with the method, and thus further device applications (not only for solar cells) would be expected.
The authors have nothing to disclose.
The authors thank New Energy and Industrial Technology Development Organization (NEDO) under Ministry of Economy, Trade, and Industry (METI), Japan, for the financial support.
Nanohole mold | Scivax http://www.scivax.com |
FLH230/500-120 | |
PTFE container | Eishin http://www.colbyeishin.com |
n/a | Custom made |
Hard-PDMS materials | Gelest http://www.gelest.com/gelest/forms/Home/home.aspx |
VDT-731 | Vinylmethylsiloxane-dimethylsiloxane copolymer |
SIP6831.1 | Pt-divinyltetramethyldisiloxane complex | ||
HMS-301 | Methylhydrosiloxane-dimethylsiloxane copolymer | ||
2,4,6,8-tetramethyltetra-vinylcyclotetrasiloxane | Sigma-Aldrich http://www.sigmaaldrich.com |
396281 | Additive for hard-PDMS |
Soft-PDMS materials | Dow Corning http://www.dowcorning.com |
Sylgard-184 | Silicone precursor |
PS-b-P2VP | Polymer Source http://polymersource.com |
P5742-S2VP | Mn × 103 = 133-b-132 |
Glass/SnO2:F substrates | Asahi Glass Co. Ltd. http://www.agc.com/english/company |
Type VU | Chemical mechanical polished by D-process Inc. (http://d-process.jp/index.html) to flatten the surfaces |
Detergent | Fruuchi Chemical Co. http://www.furuchi.co.jp/eng/main.htm |
Semico-clean 56 | Used for the cleaning of Glass/SnO2:F substrates |
ZnO:Ga supputtering target | AGC Ceramics Co. Ltd. http://www.agcc.jp/2005/en/index.html |
5.7GZO | |
Ag supputtering target | Mitsubishi Materials Co. http://www.mitsubishicarbide.com/mmc/en/index.html |
4NAg | |
Double-sided adhesive tape | Nisshin EM Co. http://nisshin-em.co.jp/information/carbontape.html |
732 | |
Polyimide tape | Dupont http://www.dupont.com/products-and-services/membranes-films/polyimide-films/brands/kapton-polyimide-film.html |
Kapton 650S#25 | |
Sn-Zn-based Solder | Kuroda Techno Co., Ltd. http://www.kuroda-techno.com/english/index.html |
Cerasolzer AL-200 | |
Digital micro pipette | Nichiryo http://www.nichiryo.co.jp/en/product/pipette/ex/index.html |
00-NPX2-20 00-NPX2-200 00-NPX2-1000 |
|
Heating chamber | Tokyo Rikakikai Co., Ltd. http://www.eyelaworld.com/product_view.php?id=120 |
VOS-201SD | |
Electron beam evaporator (two types) |
Canon-Anelva https://www.canon-anelva.co.jp/english/index.html |
n/a | Custom made |
Arios http://arios.com/ |
n/a | Custom made | |
Sputtering system | Ulvac http://www.ulvac.co.jp/en |
SBR-2306 | |
PECVD system | Shimadzu Emit Co. Ltd. http://www.shimadzu.co.jp/emit/en/ |
SLCM-13 | |
Ar plasma system | Diner Electric Gmbh http://www.plasma.de/index.html |
Femto | |
RIE system | Samco Inc. http://www.samcointl.com |
RIE-10NR | |
Ultrasonic soldering device | Colby-Eishin Enterprises, Inc. http://www.colbyeishin.com/sub_sunbonder.htm |
SUNBONDER | |
EQE measurement system | Bunkoukeiki Co. Ltd. http://www.bunkoukeiki.co.jp/ |
CEP-25BXS | |
J-V characteristics measurement system | OTENTOSUN-5S-I/V | ||
Amorphous Si reference cell | WPVS-NPB-S1 | For light intensity calibration | |
Digital multi-meter | Keithley Instruments Inc. http://www.keithley.com/ |
2400 |