Arrays of Parallel Connected Coaxial Multiwall-Carbon- Nanotube–Amorphous-Silicon Solar Cells By Hang Zhou,* Alan Colli, Arman Ahnood, Yang Yang, Nalin Rupesinghe, Tim Butler, Ibraheem Haneef, Pritesh Hiralal, Arokia Nathan, and Gehan A. J. Amaratunga* Many nanotechnology-based enhancements for solar cells have been proposed. [1] On the one hand, inorganic nanoparticles, nanorods, and nanowires (NWs) have been incorporated into organic solar cells [2–7] to extend the available interface for charge separation and improve the overall mobility of the material. On the other hand, for conventional inorganic solar cells, it has also been theoretically shown that NW-like structures could potentially lead to a better collection efficiency of photogenerated carriers together with greater optical absorption from low-purity material. [8,9] Particular attention has been placed on the coaxial core/shell NW structure, where charge separation of photo- generated carriers occurs in the radial direction, that is, orthogonally to the optical absorption path. This idea can be implemented either with a core/shell NW p–n junction [10] or with a core/shell structure formed by two materials with type-II band alignment, which separate charge at the interface without the need for doping. [11,12] From a practical point of view, however, clear evidence of NW solar cells outperforming existing solutions based on planar thin films is still lacking. Silicon-based solar cells currently dominate the photovoltaic (PV) market. [13] Several attempts have been made to fabricate solar cells based on coaxial silicon nanowires (SiNWs). [14,15] The efficiency of SiNW p–n and p–i–n junctions as nanoelectronic power sources has been measured down to the individual-NW level. [10,16] The interpenetrated p–n junction geometry based on coaxial NWs, however, still relies on carrier-collection paths composed of semiconductor materials, which pose constraints on collection efficiency due to path resistance. Here, we extend this approach by fabricating solar cells with interpenetrated electrodes, that is, a multishell coaxial NW structure where a metallic inner core and a metallic outer shell act as proximity electrodes for the radial junction sandwiched in between. Specifically, our strategy is based on coating vertically aligned multiwall carbon nanotubes (MWNTs) with amorphous- silicon (a-Si:H) shells and indium tin oxide (ITO). MWNTs are known as quantum resistors, whose conductance is independent of the diameter and length of the tube (one unit of the conductance quantum G o ¼ 2e 2 =h ¼ 12:9kV ð Þ 1 ). [17] The use of MWNTs as core contacts can avoid electrical losses that might occur in other types of nanowires. Moreover, nanotube/nanowire (NT/NW) arrays form a natural light-trapping structure. [18] Indeed, a 25% increase of the short-circuit current is achieved for the NT/NW array compared to the planar a-Si cell used as reference. Most of this enhancement is measured to come from red photons. A similar experiment was recently carried out by Camacho et al., who used MWNTs as scaffold to support a CdTe heterojunction. [19] In contrast, our study concentrates on a-Si:H as a cheap and environment-friendly material, which at present is widely used in commercial solar-cell manufacturing. We also see that a-Si:H provides a more conformal coating on MWNTs compared to CdTe. Figure 1a outlines the fabrication process of the coaxial MWNT/a-Si:H solar-cell devices (I-III) and their final physical structure (IV). Details of the fabrication process can be found in the Experimental section. Scanning electron microscopy (SEM) images of vertically aligned MWNTs arrays and of coaxial MWNT–amorphous-silicon structures are shown in Figure 1b. MWNTs are conformally coated with a-Si:H with no evident tapering. The solar cells exhibit a periodic 3D nanostructure, leading to an amplification of the effective cell area. From top to bottom, our solar cells comprise layers of ITO/intrinsic a-Si:H/ n-type a-Si:H/MWNTs on a tungsten substrate. The solar cells are thus Schottky-type, as Schottky junctions are formed at the ITO/ a-Si:H interface. The built-in electric field across the absorbing layer – intrinsic a-Si:H – is formed through the alignment of the Fermi levels between the n þ doped a-Si:H and the ITO. The nominal band diagram for the structure is shown in Figure 1c. The work-function of the ITO is taken to be 4.9 eV. The band gap and electron affinity of a-Si:H are taken as 1.6 eV and 3.7 eV, respectively. The n-type amorphous Si is taken as having a Fermi level 0.2 eV below the effective conduction-band edge. The ITO and a-Si:H form a Schottky junction that is suitable for photogenerated hole flow. In order to assess the performance of the coaxial MWNT/ a-Si:H solar cells, a planar cell without MWNTs was prepared as a reference using the same fabrication process. The line-pattern COMMUNICATION www.advmat.de [*] Prof. G. A. J. Amaratunga, H. Zhou, Y. Yang, Dr. N. Rupesinghe, T. Butler, I. Haneef, P. Hiralal Electrical Engineering Division Department of Engineering University of Cambridge 9 JJ Thomson Avenue, Cambridge CB3 0FA (UK) E-mail: hz239@cam.ac.uk; gaja1@cam.ac.uk Dr. A. Colli Nokia Research Centre Cambridge U.K. C/O Nanoscience Centre Cambridge CB3 0FF (UK) A. Ahnood, Prof. A. Nathan London Centre for Nanotechnology University College London 17-19 Gordon Street, London WC1H 0AH (UK) DOI: 10.1002/adma.200901094 Adv. Mater. 2009, 21, 3919–3923 ß 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 3919