Atomic Layer Deposition of Hafnium Oxide from Tetrakis

Apr 6, 2007 - Kejing Li , Shenggang Li , Ning Li , Tonya M. Klein , and David A. Dixon. The Journal of Physical Chemistry C 2011 115 (38), 18560-18571...
0 downloads 0 Views 229KB Size
Download PDF
J. Phys. Chem. C 2007, 111, 6495-6499

6495

Atomic Layer Deposition of Hafnium Oxide from Tetrakis(ethylmethylamino)hafnium and Water Precursors Wei Chen, Qing-Qing Sun, Min Xu, Shi-Jin Ding, David Wei Zhang,* and Li-Kang Wang Department of Microelectronics, Fudan UniVersity, Shanghai 200433, People’s Republic of China ReceiVed: January 16, 2007; In Final Form: February 10, 2007

We have calculated the atomistic mechanism for the HfO2 atomic layer deposition (ALD) using Hf(NEtMe)4 and H2O precursors using density functional theory. On hydroxylated Si surface, our results show overall Hf(NEtMe)4 half-reaction is exothermic by 1.65 eV with a small activation barrier of 0.10 eV. The activation barriers for water half-reaction are 0.24 and 0.20 eV. This indicates HfO2 ALD with Hf(NEtMe)4 and H2O has a faster deposition rate than that with HfCl4 and H2O and can be run at relatively low temperature. However, on H-terminated Si surface, the Hf(NEtMe)4 half-reaction is endothermic by 2.39 eV with high activation barriers. The H2O half-reaction is similar to that on hydroxylated surface which is kinetically favorable. The results suggest that long Hf(NEtMe)4 pulse time and high deposition temperature is required during the initial stage of ALD on H-terminated surface. Moreover, our calculations indicate the reaction byproduct HNEtMe is likely to be bound to the surface after each half-reaction because of high desorption energy of 0.7-0.9 eV. Thus, long precursor purge time should be used if HNEtMe is needed to be removed completely.

1. Introduction High-k oxides are being investigated extensively as alternative gate dielectrics to conventional SiO2 in silicon metal oxide semiconductor field effect transistor (MOSFET) devices to remain small gate leakage currents beyond the 65 nm technology node. Currently, HfO2 is the leading high-k dielectric candidate given by its large dielectric constant (∼25) and good thermodynamic stability.1,2 Among various film growth techniques, atomic layer deposition (ALD) possesses the advantages of forming ultrathin uniform and conformal films with accurate thickness control because the precursors are injected into the reactor alternately separated by inert gas purge, making ALD a completely surface reaction limited process.3-8 Therefore, ALD is very suitable for HfO2 deposition for next-generation MOSFET device integration. Hafnium halides (e.g., HfCl4) and water precursors were first implemented in HfO2 ALD processes.3,9-12 However, the deposited films from hafnium chloride suffer from residual chlorine content, which can be significant (1-3 atom %) at the common deposition temperature of 300 °C13 and even resistant to annealing.14 Chlorine contamination is a serious concern because it may increase charge concentration and introduce extra band gap and interface states in HfO2, leading to threshold voltage instability and increasing leakage current for MOSFET devices. Chlorine is also responsible for void defect generation during postdeposition annealing by etching the silicon substrate,15 which may deteriorate the dielectric performance. Moreover, because of the low volatility of HfCl4 as a solid precursor, the transport efficiency is limited, resulting in slow growth rates. Thus, chlorine-free liquid precursors are favored so as to improve the dielectric quality of ALD HfO2. As alternative Hf precursors, organometallic compounds, such as * To whom correspondence should be addressed. Phone/Fax: +86 2165642389. E-mail: [email protected].

alkoxides and alkylamides, are promising candidates because of their high reactivity. Recently, tetrakis(ethylmethylamide)hafnium (TEMAH), that is, Hf[N(CH3)(C2H5)]4 or Hf(NEtMe)4, shows advantages in HfO2 ALD as the metal precursor, including lower deposition temperatures and higher quality than for the metal chloride process.5,16-20 It is, therefore, of importance to understand the chemical mechanisms of HfO2 ALD process using the TEMAH precursor. In the present study, we use ab initio computational methods to investigate the detailed atomistic mechanisms for the reactions of TEMAH and water with both hydroxylated and hydrogen-terminated Si surface. We first study the surface reaction between the TEMAH and Si surface as the TEMAH half-reaction, and then the reaction between the resulting surface and H2O is studied as the H2O half-reaction. 2. Computational Details The atomistic geometry structures and reaction kinetics are determined using the B3LYP density functional theory (DFT) with the Becke’s exchange functional and the Lee-Yang-Parr correlation functional, as is implemented in the Gaussian 03 suite of programs.21 To model the Si surface, we employ the Si(100)-2 × 1 reconstructed cluster, Si9H12, which consists of four layer atoms including two surface atoms, four second-layer atoms, two third-layer atoms, and one fourth-layer atom. The bulk Si atoms are terminated by hydrogen atoms to prevent unrealistic charge transfer. Then, hydroxyl groups and hydrogen atoms are bonded to the surface Si atoms to represent the hydroxylated and H-passivated Si surface, respectively. The electronic structure is expanded using atomic basis functions. All atoms are described by the LANL2DZ effective core potential (ECP) basis set, which consists of the Los Alamos LANL2 ECP and a valence double-ζ basis set.22-24 An enhanced mixed basis set scheme, in which Hf and the rest of the atoms are described by the LANL2DZ and 6-31+G(d,p) basis sets,25-29

10.1021/jp070362u CCC: $37.00 © 2007 American Chemical Society Published on Web 04/06/2007

6496 J. Phys. Chem. C, Vol. 111, No. 17, 2007

Chen et al.

Figure 1. Reaction pathway and optimized structures for Hf(NEtMe)4 half-reaction on Si9H12-(OH)2 cluster. (a) Separated reactants; (b) Hf(NEtMe)4 adsorption state; (c) HNEtMe desorption state; (d) transition state; (e) product complex Si-O-Hf-(NEtMe)2; (f) HNEtMe desorption state. The black, gray, white, red, blue, and green atoms represent silicon, carbon, hydrogen, oxygen, nitrogen, and hafnium atoms, respectively.

respectively, is used to calculate the single-point energies of some optimized structures. The energies reported in this paper are calculated using B3LYP/LANL2DZ by default. The clusters are fully relaxed without any constraint during optimizations. Frequency calculations are performed after geometry optimizations to verify whether a minimum or a first-order saddle point is reached. All energies reported are zero-point corrected. 3. Results and Discussions 3.1. HfO2 ALD on Hydroxylated Si Surface. Two consecutive reactions are involved in this HfO2 ALD process, one of which is Hf(NEtMe)4 half-reaction and the other is water halfreaction. In the first half-reaction, Hf(NEtMe)4 reacts with the surface hydroxyl group Si-OH* (eq 1), resulting in the formation of Hf-O bond and the elimination of HNEtMe.

Si-OH* + Hf(NEtMe)4 f Si-O-Hf(NEtMe)3* + HNEtMe Si-O-Hf(NEtMe)3* + Si-OH* f Si-O-Hf (NEtMe)2* + HNEtMe (1) During the water half-reaction, H2O reacts with the Hf-NEtMe surface site produced by eq 1 to form a new Hf-O bond (eq 2).

Si-O-Hf(NEtMe)2* + 2H2O f Si-O-Hf-(OH)2* + 2HNEtMe (2) The asterisks in eqs 1 and 2 denote surface species. The optimized structures and pathway of Hf(NEtMe)4 halfreaction (eq 1) on Si(100)-2 × 1 cluster are shown in Figure 1. It is clear from Figure 1 that the mixed basis set scheme results in higher total energies than the LANL2DZ basis set does. However, both basis sets predict qualitatively similar potential energy surface, and thus we use the results obtained

Figure 2. (a) Reaction pathway and optimized structures for H2O halfreaction on Si-O-Hf-(NEtMe)2 surface sites. (a) Separated reactants; (b) H2O adsorption state; (c) transition state; (d) product complex SiHf-NEtMe-OH; (e) HNEtMe desorption state. (b) Reaction pathway and optimized structures for H2O half-reaction on Si-O-Hf-NEtMeOH surface sites. (a) Separated reactants; (b) H2O adsorption state; (c) transition state; (d) product complex Si-O-Hf-(OH)2; (e) HNEtMe desorption state.

by LANL2DZ basis set because the mixed basis set scheme results in longer computation time. We did not find a stable initial Hf(NEtMe)4 physisorbed complex in our calculations. The adsorption of Hf(NEtMe)4 on hydroxylated Si surface directly leads to a chemisorption state (b), which is located at 1.42 eV below the reactants (a). The chemisorption state involves the formation of two bonds; one is Hf-O bond formation through the lone-pair donation from the oxygen atom of surface hydroxyl group to the Hf atom of the precursor, and the other is formation of HNEtMe through the combination of one NEtMe ligand from precursor and a hydrogen atom from surface hydroxyl group, which subsequently desorbs. The desorption of HNEtMe is spontaneous as the desorption state (c) is -0.10 eV with respect to the chemisorption state (b). Following HNEtMe desorption, one NEtMe ligand from the chemisorbed precursor and a hydrogen atom from the surface hydroxyl group combine

Atomic Layer Deposition of Hafnium Oxide

J. Phys. Chem. C, Vol. 111, No. 17, 2007 6497

Figure 3. Reaction pathway and optimized structures for Hf(NEtMe)4 half-reaction on Si9H12-H2 cluster. (a) Separated reactants; (b) transition state; (c) product complex Si-Hf-(NEtMe)3; (d) HNEtMe desorption state; (e) transition state; (f) product complex Si-Hf-(NEtMe)2; (g) HNEtMe desorption state.

through a four-center transition state with an activation energy of 0.10 eV to form HNEtMe, which desorbs from the surface subsequently. The HNEtMe desorption state (f) is endothermic by 0.71 eV relative to the chemisorption state (e). The overall reaction is exothermic by 1.65 eV with respect to the reactants, which indicates that the first half-reaction is thermodynamically favorable. Widjaja and Musgrave have calculated the adsorption of HfCl4 on HfO2 which is modeled using Hf-OH* clusters.30 Their results show that the reaction proceeds through an activation barrier of 1.07 eV. More recently, Ren et al. have calculated the adsorption of HfCl4 precursor on hydroxylated Si surface and have found the reaction to be exothermic by 0.33 eV with an activation barrier of 0.75 eV.31 This suggests the adsorption of Hf(NEtMe)4 on hydroxylated Si surface is much more efficient than that of HfCl4. The additional energy required for the elimination of HNEtMe may result in incorporation of N into HfO2 if the purge time is not sufficient. The second half-reaction starts with the reaction of H2O and -Hf-(NEtMe)2 surface site (Figure 2a). H2O adsorbs molecularly on surface site by dative bonding to the Hf atom with an adsorption energy of 0.9 eV. HNEtMe then forms by transfer of a hydrogen from H2O to one NEtMe ligand with an energy barrier of 0.16 eV with respect to the complex intermediate (b). The HNEtMe desorbs subsequently with a desorption energy of 0.77 eV relative to the physisorption state (d). The overall reaction is exothermic by 1.13 eV. The remaining -Hf-NEtMe ligand can react with another H2O, which is shown in Figure 2b. Similar to the previous reaction, the first step of the oxygen-ligand exchange reaction involves the adsorption of H2O on the Hf atom to form a dativebonded water complex (b) with one of the oxygen lone pairs directed to Hf atom. The adsorption energy of the dative-bonded complex is 0.90 eV in our calculations. A hydrogen atom then transfers from the adsorbed water complex to the NEtMe ligand to form HNEtMe via a transition state (c) which is located at 0.20 eV with respect to the adsorbed complex (b). Finally, HNEtMe desorbs with a desorption energy of 0.95 eV to complete the conversion of the -NEtMe sites to surface hydroxyl groups. Thus, longer purge time is also essential in the water half-reaction for complete NEtMe removal. The

overall reaction is exothermic by 1.04 eV. It is therefore concluded that the second water half-reaction is also thermodynamically favorable. The reaction of water with surface -HfClx group has an activation barrier of 0.80 eV,31 which is about 0.6 eV larger than that of the reaction of water with surface -Hf(NEtMe)x group. The results show the evidence that the atomic layer deposition using Hf(NEtMe)4 is much more efficient on hydroxylated Si surface than that using HfCl4 precursor and thus can be run under low temperature. Long precursor purge time is required to completely eliminate HNEtMe. On the other hand, nitrogen incorporation is known to be beneficial for HfO2 in some aspects. Nitrogen has been observed to suppress crystallization during high-temperature annealing and to reduce diffusion of impurities from the gate electrode. However, the improvement by nitrogen incorporation in HfO2 using Hf(NEtMe)4 ALD process still needs further investigation. 3.2. HfO2 ALD on H-Terminated Si Surface. Two corresponding half-reactions are involved in this ALD process. In the first half-reaction, Hf(NEtMe)4 reacts with surface Si-H sites and converts the surface sites to Hf(NEtMe)2 along with the elimination of HNEtMe.

Si-H* + Hf(NEtMe)4 f Si-Hf(NEtMe)3* + HNEtMe Si-Hf(NEtMe)3* + Si-H* f Si-Hf(NEtMe)2* + HNEtMe (3) In the second half-reaction, gaseous H2O molecules react with Hf(NEtMe)2 sites to form new Hf-O bonds, resulting in surface -Hf-OH groups.

Si-Hf(NEtMe)2* + 2H2O f Si-Hf-(OH)2* + 2HNEtMe (4) The asterisks in eqs 3 and 4 denote surface species. Figure 3 shows the optimized structures and reaction pathway of Hf(NEtMe)4 half-reaction on H-terminated Si(100)-2 × 1. We use B3LYP/LANL2DZ level of theory for the calculations. The reaction pathway proceeds through the combination of a NEtMe ligand from the metal precursor and one hydrogen atom

6498 J. Phys. Chem. C, Vol. 111, No. 17, 2007

Figure 4. (a) Reaction pathway and optimized structures for H2O halfreaction on Si-Hf-(NEtMe)2 surface sites. (a) Separated reactants; (b) H2O adsorption state; (c) transition state; (d) product complex SiO-Hf-NEtMe-OH; (e) HNEtMe desorption state. (b) Reaction pathway and optimized structures for H2O half-reaction on Si-HfNEtMe-OH surface sites. (a) Separated reactants; (b) H2O adsorption state; (c) transition state; (d) product complex Si-Hf-(OH)2; (e) HNEtMe desorption state.

from a neighboring Si-H* site via a four-center transition state (TS1) to form a Si-Hf(NEtMe)3 surface site without going through initial physisorption state. Then, a NEtMe ligand from the metal precursor abstracts one hydrogen atom from a neighboring Si-OH* site via a four-center transition state (TS1) to form a Si-Hf(NEtMe)3 surface site. The transition state (b) and resulting complex (c) is 0.77 and 0.36 eV higher in energy than the separated reactants, respectively. The gas-phase NEtMe then desorbs with a desorption energy of 0.08 eV. Next, hydrogen transfers from surface Si-H site to a NEtMe ligand from Hf(NEtMe)3 to form HNEtMe via a transition state (TS2) with an activation barrier of 1.54 eV. The energy of product complex is 1.49 eV relative to the reactants. Finally, NEtMe desorbs with a desorption energy of 0.9 eV, and the surface is converted to Si-Hf(NEtMe)2 sites. The overall reaction is endothermic by 2.39 eV, making it unfavorable compared to the adsorption on hydroxylated surface. The rate-limiting step of this half-reaction is the second NEtMe ligand-transfer step, which implies that the reaction might be trapped in product

Chen et al. complex (c). Therefore, relatively high precursor vapor pressures or nonequilibrium flow conditions should be implemented to drive the reaction toward the product. Similar to reaction on hydroxylated Si surface, the large desorption energy of NEtMe suggests the residual byproduct may remain on the surface if short metal precursor purge time is applied. The reaction mechanism for second half-reaction of H2O with Hf(NEtMe)2 sites is similar to that of eq 2, which is depicted in Figure 4a. The adsorption of H2O on Hf(NEtMe)2 involves the lone-pair donation from H2O to the Hf atom. The adsorption energy is calculated to be 0.99 eV. Then, one hydrogen atom from H2O and one NEtMe ligand from surface sites combine through a four-center transition state with an activation barrier of 0.23 eV relative to the intermediate complex. The product complex has an energy of 2.07 eV lower than separated reactants. Then, an additional energy of 0.85 eV is needed for NEtMe to desorb from surface. The overall reaction is exothermic by 1.22 eV. The reaction pathway proceeds through the reaction of H2O and the remaining -Hf-NEtMe surface site, which is shown in Figure 4b. The H2O adsorption energy is 0.96 eV. The transition state is 0.64 eV below the reactants, leading to an activation barrier of 0.32 eV. The energy of product complex is -1.96 eV with respect to the reactants. Finally, the gas-phase HNEtMe desorbs with a desorption energy of 0.92 eV. Overall, the reaction is 1.04 eV exothermic, which indicates that the reaction is thermodynamically favorable. According to our calculations, it is suggested that during the initial stage of deposition, long hafnium precursor pulse time and high deposition temperature is required for good interface oxide quality on H-terminated Si because of the high activation barrier of the first Hf(NEtMe)4 half-reaction. If the H-terminated Si surface is exposed to water during the first half-reaction to convert the surface species to Si-OH*, the reaction still proceeds through a large barrier of 1.58 eV as is previously reported.32 Therefore, the initial deposition rate on Si-H* will be very slow no matter which precursor is pulsed onto the surface during the first half-reaction. During the second H2O half-reaction, short pulse time can be used thanks to the lower activation barrier and exothermic reaction. Long precursor purge time is required for both half-reactions if complete elimination of HNEtMe is needed. 4. Conclusions We have calculated the surface reaction of HfO2 ALD with Hf(NEtMe)4 and H2O precursors using density functional theory. On hydroxylated Si surface, our results show that both Hf(NEtMe)4 and water half-reactions are thermodynamically favorable. We find that the overall Hf(NEtMe)4 half-reaction is exothermic by 1.65 eV with a small activation barrier of 0.10 eV. The activation barriers for water half-reaction are 0.24 and 0.20 eV. This indicates HfO2 ALD with Hf(NEtMe)4 and H2O has a faster deposition rate than that with HfCl4 and H2O, and thus the ALD process can be run at relatively low temperature. However, on H-terminated Si surface, the Hf(NEtMe)4 halfreaction proceeds with very high activation barriers during ligand-transfer steps. The Hf(NEtMe)4 half-reaction on Hterminated Si is endothermic by 2.39 eV overall. The H2O halfreaction is similar to that on hydroxylated surface which is kinetically favorable. Therefore, our results suggest that long Hf(NEtMe)4 pulse time and high deposition temperature is required during the initial stage of ALD on H-terminated surface. Moreover, on both Si surfaces, the calculations indicate the reaction byproduct HNEtMe is likely to be bound to the surface

Atomic Layer Deposition of Hafnium Oxide after each half-reaction because of high desorption energy of 0.7-0.9 eV. Thus, long precursor purge time should be used if HNEtMe is needed to be removed completely. References and Notes (1) Wallace, R. M.; Wilk, G. D. Crit. ReV. Solid State Mater. Sci. 2003, 28, 231. (2) Robertson, J. Rep. Prog. Phys. 2006, 69, 327. (3) Park, H. B.; Cho, M. J.; Park, J.; Lee, S. W.; Hwang, C. S.; Kim, J. P.; Lee, J. H.; Lee, N. I.; Kang, H. K.; Lee, J. C.; Oh, S. J. J. Appl. Phys. 2003, 94, 3641. (4) Javey, A.; Guo, J.; Farmer, D. B.; Wang, Q.; Wang, D. W.; Gordon, R. G.; Lundstrom, M.; Dai, H. J. Nano Lett. 2004, 4, 447. (5) Triyoso, D. H.; Hegde, R. I.; White, B. E.; Tobin, P. J. J. Appl. Phys. 2005, 97, 124107. (6) Kukli, K.; Pilvi, T.; Ritala, M.; Sajavaara, T.; Lu, J.; Leskela, M. Thin Solid Films 2005, 491, 328. (7) Kirsch, P. D.; Quevedo-Lopez, M. A.; Li, H. J.; Senzaki, Y.; Peterson, J. J.; Song, S. C.; Krishnan, S. A.; Moumen, N.; Barnett, J.; Bersuker, G.; Hung, P. Y.; Lee, B. H.; Lafford, T.; Wang, Q.; Gay, D.; Ekerdt, J. G. J. Appl. Phys. 2006, 99, 023508. (8) Zhong, L. J.; Daniel, W. L.; Zhang, Z. H.; Campbell, S. A.; Gladfelter, W. L. Chem. Vap. Deposition 2006, 12, 143. (9) Ritala, M.; Leskela, M.; Niinisto, L.; Prohaska, T.; Friedbacher, G.; Grasserbauer, M. Thin Solid Films 1994, 250, 72. (10) Kukli, K.; Ihanus, J.; Ritala, M.; Leskela, M. Appl. Phys. Lett. 1996, 68, 3737. (11) Aarik, J.; Aidla, A.; Mandar, H.; Uustare, T.; Kukli, K.; Schuisky, M. Appl. Surf. Sci. 2001, 173, 15. (12) Kukli, K.; Ritala, M.; Sajavaara, T.; Keinonen, J.; Leskela, M. Thin Solid Films 2002, 416, 72. (13) Aarik, J.; Aidla, A.; Kiisler, A. A.; Uustare, T.; Sammelselg, V. Thin Solid Films 1999, 340, 110. (14) Ferrari, S.; Scarel, G.; Wiemer, C.; Fanciulli, M. J. Appl. Phys. 2002, 92, 7675. (15) Lysaght, P. S.; Foran, B.; Bersuker, G.; Chen, P. J. J.; Murto, R. W.; Huff, H. R. Appl. Phys. Lett. 2003, 82, 1266. (16) Kukli, K.; Ritala, M.; Lu, J.; Harsta, A.; Leskela, M. J. Electrochem. Soc. 2004, 151, F189. (17) Ho, M. T.; Wang, Y.; Brewer, R. T.; Wielunski, L. S.; Chabal, Y. J.; Moumen, N.; Boleslawski, M. Appl. Phys. Lett. 2005, 87, 133103.

J. Phys. Chem. C, Vol. 111, No. 17, 2007 6499 (18) Senzaki, Y.; Park, S.; Chatham, H.; Bartholomew, L.; Nieveen, W. J. Vac. Sci. Technol., A 2004, 22, 1175. (19) Kukli, K.; Ritala, M.; Sajavaara, T.; Keinonen, J.; Leskela, M. Chem. Vap. Deposition 2002, 8, 199. (20) Liu, X. Y.; Ramanathan, S.; Longdergan, A.; Srivastava, A.; Lee, E.; Seidel, T. E.; Barton, J. T.; Pang, D.; Gordon, R. G. J. Electrochem. Soc. 2005, 152, G213. (21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A. Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision B.05; Gaussian, Inc.: Wallingford, CT, 2004. (22) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270. (23) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284. (24) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299. (25) Binkley, J. S.; Pople, J. A.; Hehre, W. J. J. Am. Chem. Soc. 1980, 102, 939. (26) Gordon, M. S.; Binkley, J. S.; Pople, J. A.; Pietro, W. J.; Hehre, W. J. J. Am. Chem. Soc. 1982, 104, 2797. (27) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon, M. S.; DeFrees, D. J.; Pople, J. A. J. Chem. Phys. 1982, 77, 3654. (28) Pietro, W. J.; Francl, M. M.; Hehre, W. J.; DeFrees, D. J.; Pople, J. A.; Binkley, J. S. J. Am. Chem. Soc. 1982, 104, 5039. (29) Frisch, M. J.; Pople, J. A.; Binkley, J. S. J. Chem. Phys. 1984, 80, 3265. (30) Widjaja, Y.; Musgrave, C. B. J. Chem. Phys. 2002, 117, 1931. (31) Ren, J.; Lu, H. L.; Chen, W.; Xu, M.; Zhang, D. W. Appl. Surf. Sci. 2006, 252, 8466. (32) Halls, M. D.; Raghavachari K. J. Chem. Phys. 2003, 118, 10221.