104 IEEE JOURNAL OF PHOTOVOLTAICS, VOL. 1, NO. 1, JULY 2011 Low-Temperature a-Si:H/GaAs Heterojunction Solar Cells Davood Shahrjerdi, Bahman Hekmatshoar, and Devendra K. Sadana Abstract—We propose a novel hydrogenated amorphous sili- con (a-Si:H)/GaAs heterostructure for photovoltaic solar cells. The structure has two key advantages: 1) low-temperature processing and 2) a relatively low cost of cell fabrication compared with con- ventional junction structures that require epitaxial growth. We investigate the impact of different hydrogen dilution levels used during a-Si:H deposition on the electrical characteristics of het- erojunction GaAs solar cells. It is interesting to note that epitaxial growth of silicon on GaAs occurred when relatively high hydro- gen dilution levels were used. The prospect of silicon epitaxy in improving the cell performance is discussed. Index Terms—Heterojunction, photovoltaic solar cells. I. INTRODUCTION D IRECT bandgap III–V materials are promising candidates for thin-film portable high-efficiency photovoltaic (PV) solar cell applications, due to their strong absorption properties. However, the high III-V substrate cost has always been the main impediment for their widespread use for terrestrial PV applica- tions. Therefore, development of various layer transfer schemes to produce thin sheets of single-crystalline III–V materials to re- duce substrate cost has been the active area of research [1]–[4]. Several approaches to fabricate thin-film single-crystal III–V solar cells have been published. One straightforward approach to obtain thin-film cells requires epitaxial growth of the III–V solar cell structures and its removal by a layer transfer tech- nique [3]–[5]. The high cost of epitaxy, however, is expected to increase the final cost of the solar cell. Alternative approaches, such as the use of Schottky-type junctions [6] and zinc diffu- sion [1], have been proposed to eliminate the need for epitaxial formation of p/n junctions to reduce cost. In this paper, we de- scribe an attractive approach that utilizes aSi:H/gallium arsenide (GaAs) heterojunction (HJ) solar cell for the first time. Not only does this approach use low processing temperatures (<200 ◦ C) for cell fabrication, but it also offers a path for low-cost, high- efficiency PV technology when implemented in conjunction with a layer transfer technique. HJs of a-Si:H on crystalline silicon (c-Si) was first proposed by Fuhs et al. [7]. This structure was later commercially devel- oped by Sanyo (commercially known as the heterojunction with Manuscript received April 10, 2011; accepted July 24, 2011. Date of publi- cation September 26, 2011; date of current version October 27, 2011. The authors are with the IBM T. J. Watson Research Center, Yorktown Heights, NY 10598 USA (e-mail: davood@us.ibm.com; hekmat@us.ibm.com; dksadana@us.ibm.com). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JPHOTOV.2011.2164391 Fig. 1. Cross-sectional scanning electron microscope image of an HJ GaAs PV solar cell, illustrating the device structure. intrinsic thin layer (HIT) structure), producing high-efficiency c-Si solar cells [8]. A thin layer of a-Si:H is believed to pro- vide exquisite surface passivation on c-Si [9]–[11]. In addition to c-Si, a-Si:H has been widely explored for passivating III–V materials. Published C–V and metal-oxide semiconductor field- effect transistor data from a-Si:H passivated GaAs show surface inversion indicating effective unpinning of the Fermi level at the GaAs surface [12], [13]. Effective surface passivation is equally crucial to improving the performance of PV solar cells to reduce the dark current. The reduction in surface recombination veloc- ity (SRV) with a-Si:H is attributed to 1) the hydrogen content in the a-Si:H, contributing to passivation of surface dangling bonds; and 2) the field-effect induced passivation due to con- duction and valence band offsets (ΔE c and ΔE v ) in a-Si:H/c-Si or a-Si:H/III–V heterostructures. Proper engineering of the en- ergy band structure of HJs is, therefore, required via a-Si:H deposition parameters. Hydrogen dilution is believed to be the key engineering parameter that allows altering energy band off- sets. Theoretical studies [14] suggest that the hydrogen content strongly influences the band offset and the resulting field-effect passivation. Therefore, this paper is also aimed at systematically investigating the effect of various hydrogen dilution ratios dur- ing a-Si:H deposition on the performance of HJ a-Si:H/GaAs solar cells. II. EXPERIMENTS The epitaxial solar cell structure was grown by metal–organic chemical vapor deposition (MOCVD) on (0 0 1) p-type Zn- doped GaAs substrates. Fig. 1 illustrates a cross-sectional sec- ondary electron microscopy image of the final HJ GaAs device. The structure consists of 0.2-μm-thick lattice-matched p-type (1 × 10 18 cm −3 ) InGaP for the back-surface field, followed by a 2.5-μm-thick p-type (1 × 10 17 cm −3 ) GaAs absorbing layer. 2156-3381/$26.00 © 2011 IEEE