pubs.acs.org/cm Published on Web 05/17/2010 r 2010 American Chemical Society 3752 Chem. Mater. 2010, 22, 3752–3761 DOI:10.1021/cm100750z Synthesis of n-type CuInS 2 Particles Using N-methylimidazole, Characterization and Growth Mechanism Fabrice M. Courtel,* ,†,‡ Amer Hammami, § R egis Imbeault, § Gr egory Hersant, ^ Royston W. Paynter, † Benoıˆt Marsan, § and Mario Morin § † Centre Energie, Mat eriaux et T el ecommunications - Institut National de la Recherche Scientifique (INRS), 1650 boulevard Lionel-Boulet, Varennes, Qu ebec J3X 1S2, Canada, § D epartement de Chimie, Universit e du Qu ebec a Montr eal (UQ AM), C.P. 8888, Succursale Centre-ville, Montr eal, Qu ebec H3C 3P8, Canada, and ^ Department of Chemistry, Concordia University, 7141 Sherbrooke Street West, Montr eal, Qu ebec H4B 1R6, Canada. ‡ Current address: National Research Council Canada, 1200 Montreal Road, Ottawa, Ontario K1A 0R6, Canada Received March 14, 2010. Revised Manuscript Received May 1, 2010 We report on the growth of CuInS 2 n-type semiconductive particles, prepared using N-methyli- midazole, as a solvent and/or a complexing agent, as well as their chemical and electrochemical properties. XPS, EDX and ICP-AES have shown that an excess of indium was obtained, which was greater at the surface (CuIn 1.19 S 1.7 at 500 °C) than in the bulk (CuIn 1.07 S 1.9 at 500 °C). Solid state Raman spectroscopy revealed two crystalline phases: chalcopyrite and the so-called copper-gold phase, and by increasing the annealing temperature of the particles, the formation of the chalcopyrite phase is favored. UV-visible measurements showed that the n-type CuInS 2 possesses a direct bandgap energy of 1.55 eV. To perform the capacitance measurements on a CuInS 2 film by EIS, we used two organic redox couples in nonaqueous media: 5-mercapto-1-methyltetrazolate (T - )/di-5- (1-methyltetrazole) disulfide (T 2 ), and 5-trifluoromethyl-2-mercapto-1,3,4-thiadiazolate (G - )/5,5 0 - bis(2-trifluoromethyl-1,3,4-thiadiazole) disulfide (G 2 ). Using these redox couples, we determined Fermi levels of -4.51 eV and -4.53 eV, and majority charge carrier densities of 2.8 10 18 and 9.6 10 18 cm -3 , respectively. According to the energy level diagram of the CuInS 2 /electrolyte interface, the G - /G 2 redox couple is expected to lead to a more efficient device. The present work shows that the complexation of the metal ions and the negative charge on sulfur anions play a key role in the mechanism of formation of CuInS 2 particles. In situ Raman spectroscopy measurements showed that an indium-sulfur precursor is formed prior to the formation of CuInS 2 particles. Indeed, if an indium-sulfur precursor is formed prior to the reaction of sulfur with copper, a much better control of the n-type CuInS 2 properties is obtained. This explains the excess of indium at the surface of the CuInS 2 particles, as well as its n-type semiconductivity. Introduction Commercially available solar cells are based on a solid-solid junction between two semiconductors. n and p doped silicon are widely used in this kind of cells but they require an high silicon grade. Chalcopyrite semiconductors, such as CuInS 2 and CuInSe 2 , are also good candidates for photovoltaic applications because of their favorable optical properties. For instance, CuInS 2 has a high absorption coefficient of 1 10 5 cm -1 (at 730- 750 nm) that is 10 times greater than that of CdTe. 1 This ensures a complete absorption of photons in micro- metric range coatings. CuInS 2 exhibits a direct bandgap energy of 1.5 eV that matches closely the solar spectrum. 2 Moreover, in 1975, Meese et al. determined that CuInS 2 has a high theoretical conversion efficiency of 30%. 3 However, the control of stoichiometry is a major diffi- culty in the synthesis of ternary materials. Most of the CuInS 2 synthetic methods reported in the literature yield material with p-type semiconductivity. However, some colloidal methods yield n-type CuInS 2 4-6 that are In-rich 7 and might be used in photovoltaic and photoelectrochemi- cal cells. We recently reported on the growth mechanism of n-type CuInS 2 particles prepared using a modified Cze- kelius’s colloidal method, as well as their chemical and *Corresponding author. Phone: (613) 993-8573. Fax: (613) 949-4184. E-mail: Fabrice.Courtel@nrc-cnrc.gc.ca. (1) Mitchell, K.; Fahrenbruch, A. L.; Bube, R. H. J. Appl. Phys. 1977, 48(2), 829–830. (2) Moller, H. J., Semiconductors for Solar Cells. Artech House: Boston, 1993. (3) Meese, J. M.; Manthuruthil, J. C.; Locker, D. R. J. Am. Phys. Soc. 1975, 20, 696–703. (4) Czekelius, C.; Hilgendorff, M.; Spanhel, L.; Bedja, I.; Lerch, M.; M€ uller, G.; Bloeck, U.; Su, D.-S.; Giersig, M. Adv. Mater. 1999, 11 (8), 643–646. (5) Castro, S. L.; Bailey, S. G.; Raffaelle, R. P.; Banger, K. K.; Hepp, A. F. J. Phys. Chem. B 2004, 108(33), 12429–12435. (6) Arici, E.; Hoppe, H.; Reuning, A.; Sariciftci,N. S.; Meissner, D. In CIS Plastic Solar Cells, Proceedings of the 17th European Photo- voltaic Solar Energy Conference; Munich, Germany, October 22-26, 2001; James and James: London, 2001;pp 61-63. (7) Yoshino, K.; Nomoto, K.; Kinoshita, A.; Ikari, T.; Akaki, Y.; Yoshitake, T. J. Mater. Sci.: Mater. Electron. 2008, 19(4), 301–304.