Solar Energy Materials & Solar Cells 91 (2007) 847–850 Letter Assessment of a low-cost gold-free metallization for III–V high concentrator solar cells Ignacio Rey-Stolle à , Beatriz Galiana, Carlos Algora Instituto de Energı´a Solar, Universidad Polite´cnica de Madrid ETSI de Telecomunicacio´n, Avda, Complutense s/n, 28040 Madrid, Spain Received 21 September 2006; received in revised form 15 January 2007; accepted 15 January 2007 Available online 28 February 2007 Abstract A gold-free metallization is proposed to be used as the grid contact in III–V concentrator solar cells. This metallization is based on the Cu/Ge system which has been reported to attain very low specific contact resistances on n-GaAs. In this letter, we show that metal layers with low resistivity (13 mO cm) can be obtained if the copper content in the alloy is around 28% in weight for a wide range of annealing temperatures (400–450 1C). Finally, this metallization has been used to manufacture single-junction GaAs high concentrator solar cells. Efficiencies of 26.2% at 1000 suns have been reached. r 2007 Elsevier B.V. All rights reserved. Keywords: III–V solar cells; Front contact; Concentrator 1. Introduction As in many other electronic devices, a good ohmic contact is a key structural element for a high-efficiency solar cell. Solar cells are especially sensitive to the resistive behaviour of the top grid contact which greatly influences the fraction of the photogenerated power which will be dissipated internally through ohmic losses. This influence is especially striking in high concentrator solar cell devices which handle elevated light power densities and thus produce elevated current densities. In the case of solar cells made of III–V compound semiconductors, an effective front grid contact entails two independent requisites on the metallization system, namely: (1) a low sheet resistance of the metal layer itself and (2) a low specific contact resistance between the metal and the semiconductor (GaAs). Traditionally, the requisite of low contact resistance has been achieved using gold-based alloys namely, Au/Zn or Ti/Pt/Au for grids on p-type GaAs and Au/Ge/Ni or Ti/Pt/Au for grids on n-type GaAs, though some other approaches were explored [1,2]. On the other hand, the low metal sheet resistance is attained just by adding a gold cap layer of a suitable thickness— sometimes within the range of micrometres—either directly during the physical vapour deposition process or through a subsequent electrochemical growth (i.e. electrolysis of electroless) [3]. This approach fulfils the requirements in a non-optimum manner since it uses thick layers of very expensive gold material. In addition, in some cases, the chemical reactions between gold and the semiconductor can also form other by-products which influence the long- term reliability of the contact [1–3]. This is an issue in the solar cell industry where the device lifetime should exceed 2 10 5 h [4]. In this letter, we describe, for the first time to the best of our knowledge, the use of low-cost Cu/Ge metallizations as the top grid contact of III–V concentrator solar cells. For the typical metal quantities used in our solar cells (designed to be used at 1000 suns), the cost would drop by a factor of about 10 as compared to the gold-based metal systems. The Cu/Ge system has been proposed as an interesting substitute for the classic gold-based metallizations both for n-GaAs [5,6] and p-GaAs [7]. The Cu/Ge system demonstrated the following features [5,6]: (1) specific contact resistances as low as 6 10 7 O cm 2 for a wide range of Ge contents in the alloy; (2) high thermal stability; (3) planar and structurally abrupt metal/semiconductor ARTICLE IN PRESS www.elsevier.com/locate/solmat 0927-0248/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2007.01.014 à Corresponding author. Tel.: +34 91 5441060; fax: +34 91 5446341. E-mail address: irey@ies-def.upm.es (I. Rey-Stolle).