Azide Reactions for Controlling Clean Silicon Surface Chemistry: Benzylazide on Si(100)-2 × 1 Semyon Bocharov, Olga Dmitrenko, Lucila P. Me ´ ndez De Leo, and Andrew V. Teplyakov* Department of Chemistry and Biochemistry, UniVersity of Delaware, Newark, Delaware 19716 Received April 13, 2006; E-mail: andrewt@udel.edu With the recent advances in using molecular assemblies for electron transfer in lateral direction with respect to semiconductor surfaces, 1,2 the chemistry that can selectively and cleanly deliver aromatic compounds to the single crystal silicon surfaces has become of the utmost importance. Despite the fact that multiple surface features have already been produced by defect-induced or STM-induced chemistry, and despite the fact that formidable evidence exists for a set of very attractive properties based on self- assembly of aromatic molecules on Si(100)-2 × 1, the chemical toolbox that allows one to form such assemblies is incredibly limited. Traditional silicon surface attachment chemistry also has a limited potential because of the high reactivity of the silicon surface and difficulty controlling the chemistry of multiply func- tionalized molecules. At the same time, some of the most potent chemical transformations are based on oxygen-containing functional groups, which pose an additional burden because of the subsurface oxygen migration. 3-5 One of the most intriguing approaches to delivering aromatic compounds onto a semiconductor substrate can be based on the azide chemistry. 6 In fact, previous studies of hydrazoic acid, HN 3 , on Si(100)-2 × 1 proved that azide attachment followed by the nitrogen elimination can be conducted in ultrahigh vacuum cleanly, without oxidation of the surface. 7,8 Here we report the first confirmed investigation using a combination of experimental and computational methods of the azide chemistry on a Si(100) that delivers an organic group with an aromatic ring in a clean and selective manner, with a single attachment product. The elimination of a N 2 molecule produces a stable adduct on a Si(100)-2 × 1 surface. Similar chemistry can further be used with different alkyl and aryl substituents to produce the desired surface features based on assemblies of molecules. In fact, it is expected that phenylazide, PhN 3 , will provide the most robust surface features with tight overlap of π-orbitals, without oxygen incorporation. 6 However, one of the difficulties with handling phenylazide is the fact that it is explosive under distillation conditions. Thus, all the chemistry relevant for the future modification of the Si(100)-2 × 1 surface will be summarized here based on a less reactive benzylazide, PhCH 2 N 3 , and can be applied in the future to phenylazide and essentially any other aromatic hydrocarbons. This general approach also suggests that varying the linker group between the azide function and the aromatic system will allow for varying the overlap of the π-systems of the aromatic rings and adding one more handle to controlling the specific arrangement of surface adducts. The mechanism of azide reactions with unsaturated entities has been a subject of a debate for several decades. 9 It was suggested in a theoretical investigation that, on the Si(100)-2 × 1 surface, methylazide should react via a concerted 1,3-cycloaddition with no noticeable barriers and should ultimately lead to a stable surface adduct with a Si-Si-N cycle. 6 On the basis of the previous experimental studies of hydrazoic acid, 7,8 it can be concluded that the final product of an azide interaction with a Si(100)-2 × 1 surface is indeed similar to the one proposed theoretically. However, as the theoretical approach showed no barrier for the addition of methylazide to this surface and no intermediates leading to the product of 1,3-cycloaddition, practical importance of this reactive chemistry would be diminished since it would be difficult to control the reactions. Our detailed investigation of the benzylazide reaction pathways is summarized in Figure 1. Here, the B3LYP/6-311+G(d,p) computational method was used to simulate in detail the adsorption configurations of benzylazide on a silicon surface modeled by a single dimer Si 9 H 12 cluster. The silicon atoms representing the subsurface were not fixed at any particular positions in this investigation. However, since all the chemistry described here essentially involves only the top two atoms of the silicon dimer, it is not expected that fixing the subsurface silicon atoms in this model would affect the predicted binding energies or the transition states. In fact, a very detailed recent study of nitrobenzene on Si(100)-2 × 1 suggested that this effect would be negligibly small. 5 Instead of a barrierless addition leading to the 1,3-adduct proposed previously for methylazide, 6 benzylazide forms a stable, albeit a weakly bound adduct, Int1, with the Si(100)-2 × 1 surface. Multiple other reaction pathways following the weak adsorption have also been considered, and the exact assignments are a subject Figure 1. Surface reaction pathways for the interaction of benzylazide with a Si(100)-2 × 1 surface represented by a Si9H12 cluster. Computations were performed at the B3LYP/6-311+G(d,p) level of theory. Hydrogen atoms representing silicon cluster termination are omitted for clarity. Green, silicon; blue, nitrogen; gray, carbon; white, hydrogen. Published on Web 07/04/2006 9300 9 J. AM. CHEM. SOC. 2006, 128, 9300-9301 10.1021/ja0623663 CCC: $33.50 © 2006 American Chemical Society