J. Phys. Chem. C 2007, 111, 411-419

411

Thermal Control of Amide Product Distributions at the Ge(100)-2×1 Surface Albert J. Keung, Michael A. Filler, and Stacey F. Bent* Department of Chemical Engineering, Stanford UniVersity, Stanford, California 94305 ReceiVed: August 15, 2006; In Final Form: September 28, 2006

We have investigated the adsorption of a series of primary, secondary, and tertiary amides including formamide, N-methylformamide, N-methylacetamide, and N,N-dimethylacetamide on the Ge(100)-2×1 surface using multiple internal reflection Fourier transform infrared spectroscopy and density functional theory. At 310 K, primary and secondary amides were observed to form thermodynamically favored N-H dissociation structures with interdimer interactions as well as kinetically favored oxygen dative-bonded structures. The relative surface product distributions could be controlled thermally. Dative-bonded adducts were isolated by exposing the amides to the surface at 240 K, whereas N-H dissociation products were formed by annealing to 450 K. While the acetamides could potentially form “-ene” type products, such products were not observed and instead the acetamides exhibited the same reactivity pattern as the formamides in this study.

1. Introduction The fields of microelectronics and organic chemistry have merged in the creation of organic thin film semiconductor devices.1-5 It may thus be advantageous to develop a better fundamental understanding of the interfacial chemistry possible between organic and inorganic materials. For example, molecular adsorption of organics on an inorganic surface can provide an ordered, molecularly thin template layer on which further organic layers can be deposited with macroscale film growth methods such as spin-casting.6-8 In addition, while previously reported organic thin film microelectronic devices have feature sizes of at least several hundred nanometers,7,9 organic functionalization of semiconductor surfaces could lead to applications in chemical sensors, molecular electronics, and nanoscale lithography,wheremolecular-scalestructureswillbeimportant.10-13 Furthermore, there has been recent use of microelectronic devices in biological applications,14,15 motivating studies of biologically relevant compounds on inorganic surfaces. In the present work, we have chosen to investigate the adsorption of amides on the Ge(100)-2×1 surface. Amides, which contain both carbonyl and amine groups, are the linkages that form the backbone of proteins and are thus prevalent in biological systems. We have previously reported our findings on a series of tertiary amides adsorbed on the Ge(100)-2×1 surface.16 We found that tertiary amides dative-bond to the Ge(100)-2×1 surface through the oxygen atom and desorb over time near room temperature with a small fraction remaining stably on the surface. The present work is an extension of that investigation and includes a series of primary, secondary, and tertiary amides. Figure 1 shows the molecules studied, including N-methylformamide (NMF), N,N-dimethylformamide (DMF), N-methylacetamide (NMA), N,N-dimethylacetamide (DMA), and formamide. With N-H bonds available for reaction, we expect secondary and primary amides to have more complex reactivity than tertiary amides. Indeed, we observe the formation of dissociation products in addition to dative-bonded products * Corresponding author. 381 North South Mall, Stanford University, Stanford, CA 94305-5025. Phone: 650-723-0385. Fax: 650-723-9780. E-mail: [email protected].

for primary and secondary amides. Furthermore, experimental evidence points to the presence of bidentate-type structures for the dissociation products, similar to those reported for carboxylic acids17 and acyl halides18 on Ge(100)-2×1. Beyond identifying the surface products, we demonstrate the ability to thermally control their relative distribution. Gaining control of the homogeneity of surface products will likely be necessary in creating ordered and patterned interfaces. One way to shift product distributions is by thermal treatment. However, attempts to convert to more stable products by thermal annealing on Si(100)-2×1 are reported to result in surface species decomposition.11,19 Ge(100)-2×1 appears to be a better candidate for achieving selectivity because, according to the literature, control over thermodynamic and kinetic product distributions is easier to accomplish on Ge(100)-2×1 compared to Si(100)-2×1.20 Our results will show that dative bond and dissociation products can be individually isolated by controlling the temperature of the germanium surface. Specifically, experimental results, together with theoretical calculations, suggest that the temperature can be swept from a kinetically controlled regime (at low temperatures) to a thermodynamically controlled regime (at higher temperatures) to convert the dative-bonded product to the dissociation product. 2. Experimental and Computational Details All experiments were performed in an ultrahigh vacuum chamber with a crystal holder described previously.21 Trapezoidally shaped Ge crystals (1 × 14 × 19 mm, 45° beveled edges) were sputtered with Ar+ ions at room temperature (0.5 keV accelerating voltage, 20 mA emission current, ∼6-8 µA sample current) for 20 min followed by annealing to 900 K for 5 min to prepare the Ge(100)-2×1 surface. Low-energy electron diffraction (LEED) confirmed that the proper (100)-2×1 surface reconstruction was achieved and Auger electron spectroscopy (AES) verified that carbon, oxygen, and nitrogen surface concentrations were undetectable. For multiple internal reflection Fourier transform infrared (MIR-FTIR) spectroscopy experiments, the infrared beam from a Biorad FTS-60A FTIR spectrometer passed through two CaF2 windows on the vacuum chamber, limiting the observable

10.1021/jp065278d CCC: $37.00 © 2007 American Chemical Society Published on Web 12/02/2006

412 J. Phys. Chem. C, Vol. 111, No. 1, 2007

Figure 1. Amides investigated in this study with naming abbreviations used in this paper.

infrared spectrum to energies greater than 1100 cm-1. The infrared beam was focused on the beveled edge of the Ge crystal and underwent approximately 20 internal reflections within the crystal before exiting the chamber and focusing onto a narrowband HgCdTe detector. The beam path outside the chamber was purged with air filtered to remove water vapor and carbon dioxide in order to reduce the spectral features resulting from these gases. 1000 and 200 scans were co-added at 4 cm-1 resolution for chemisorption and multilayer experiments, respectively, and ratioed to a background scan taken immediately prior to dosing each compound. All spectra were corrected for baseline sloping and the presence of water vapor peaks. Formamide (Aldrich, 98%), NMF (Aldrich, 99%), DMF (Aldrich, 99.9%), NMA (Acros, 99+%), and DMA (Acros, 99%) are clear liquids at room temperature. All compounds were purified by several freeze-pump-thaw cycles prior to dosing. Amides were leaked into the chamber via a variable leak valve with a 0.5 in. diameter stainless steel directed doser positioned 0.5 in. from the sample surface. Room temperature dosing was adequate for DMF, NMA, and DMA whereas formamide and NMF required heating to 80 °C via a water bath to achieve ∼80 mtorr total pressure in the manifold (base pressure of 15 mtorr). In situ mass spectrometry confirmed the purity and identity of the compounds leaked into the chamber. Sample exposures, given in Langmuirs (1 L ) 10-6 Torr-s), were not corrected for ion gauge sensitivity. Density functional theory (DFT) calculations were also performed to obtain binding energies as well as theoretical spectra of adsorbed species. The Gaussian 03 software package22 using Becke3 Lee-Yang-Parr (B3LYP)23-25 three-parameter density functional theory26 was used. For single-dimer calculations, the Ge dimer was modeled using an unconstrained Ge2Si7H12 cluster. Previous results have shown that binding energies found with the single-dimer hybrid clusters are within 1 to 2 kcal/mol of those found with Ge9H12 clusters at the B3LYP/6-311++G(2df,pd) level of theory.27 In addition, we have performed two representative single-dimer cluster calculations for the NMF amide system using both Ge2Si7H12 and Ge9H12 clusters at the B3LYP/6-311++G(2df,pd) level of theory to validate our use of the computationally less intense Ge2Si7H12 cluster. As described in more detail in Section 3.2, these calculations showed good agreement between the two

Keung et al. clusters. For two-dimer across-trench calculations, an unconstrained Ge4Si19H24 cluster was used with two silicon atoms forming a fourth subsurface layer which restricted distortion of the two dimers relative to each other. In both cases, all subsurface silicon atoms were terminated by hydrogen to maintain the sp3 hybridization present in the bulk crystal diamond lattice. Only the top dimer atoms were modeled as Ge to keep calculations computationally manageable. Geometry optimizations and frequency calculations were performed with the 6-31G(d) basis set. For single-dimer clusters, frequency calculations were followed by single-point energy calculations performed with the higher 6-311++G(2df,pd) basis set. Singlepoint energy calculations at the 6-311++G(2df,pd) level were computationally too intensive for two-dimer clusters. Instead, only relatiVe energies and frequency calculations performed at the 6-31G(d) level are reported for two-dimer cluster calculations. All calculated frequencies were scaled by a factor of 0.96,28 and a Gaussian line-shape with a standard deviation of 10 cm-1 was used in constructing theoretical spectra. No negative frequencies were found for local minima structures, and all reported energies have been zero-point corrected. 3. Results and Discussion 3.1. NMF: Infrared Studies and Product Assignments. Figure 2a and 2b-d show typical infrared spectra of NMF from a low-temperature experiment and an annealing experiment, respectively. Figure 2a shows a spectrum taken after exposure of NMF to the Ge(100)-2×1 surface at 240 K. Figure 2b shows the chemisorption spectrum collected following a saturation exposure to the surface at 310 K in a separate experiment. After the 310 K exposure, the crystal surface was annealed to 450 K for 5 min and then cooled to 310 K at which time the infrared scan shown in Figure 2c was recorded. Figure 2d is the incremental spectrum of Figure 2c ratioed to Figure 2b, showing the change observed as a result of annealing to 450 K. It is evident that peaks at 1622 and near 1337 cm-1 dominate at low temperature (Figure 2a) while those at 1960 and 1557 cm-1 dominate at high temperature (Figure 2c). Furthermore, all four of these modes are observed at 310 K (Figure 2b), indicating that at least two distinct products are present at the surface at 310 K. The data suggest that one product is favored at lower temperatures (Figure 2a), while the other is favored at higher temperatures (Figure 2c). To understand what these two products are, we briefly review amine and tertiary amide reactivity in the context of identifying products that NMF would be likely to form on the Ge(100)-2×1 surface (Figure 3). NMF (Figure 1a) is similar in structure to the previously studied compound, DMF (Figure 1b).16 The only difference between the two compounds is an N-H bond in NMF instead of a methyl group on the nitrogen in DMF. Due to their structural and chemical similarity, we hypothesize that NMF can form a product dative-bonded through its oxygen atom, similar to that observed for DMF. Furthermore, the presence of an N-H bond in NMF allows this dative-bonded state to be stabilized by a Ge--H interaction with the other germanium atom in the dimer, similar to the interaction previously proposed for acetic acid;17 this structure is shown in Figure 3a. Our calculations show that the oxygen dative-bonded structure of NMF is 2.7 kcal/mol more stable with Ge--H interaction than without (EwithGe--H ) -19.4 kcal/mol and EwithoutGe--H ) -16.7 kcal/mol), resulting in an overall greater stability than dative-bonded DMF (-17.8 kcal/mol). With the presence of both oxygen and nitrogen in NMF, a double dativebonded structure with both O and N interacting with separate

Thermal Control of Amide Product Distributions

J. Phys. Chem. C, Vol. 111, No. 1, 2007 413

Figure 2. Spectra from temperature-dependent studies of NMF: (a) chemisorption spectrum taken after 2.0 L exposure at crystal temperature of 240 K, (b) chemisorption spectrum taken after 1.7 L saturation exposure at 310 K, (c) chemisorption spectrum taken at 310 K after a 450 K anneal of the surface shown in b, (d) incremental spectrum of c ratioed to b.

Figure 3. Possible surface adducts of NMF on Ge(100)-2×1.

dimers may also be possible. However, this product was not observed for DMF, and previous calculations found that dative bonds through amide nitrogens are exceedingly weak, with a binding energy of only 0.1 kcal/mol due to steric and electronic effects.16 Consequently, dative bonding through the N would not likely contribute more stability, especially when electron

density is already being used in forming the O dative bond, and we consider this product unlikely. N-H dissociation is also possible for NMF, as N-H dissociation has been observed for some primary and secondary amines on Ge(100)-2×1.29 The N-H dissociation product could bind through the N, as shown in Figure 3c, which is analogous to the amines, or through the O, as illustrated in Figure 3b. Furthermore, the dissociation products of carboxylic acids17 and acyl halides18 show a propensity toward forming interdimer structures. Therefore, it is likely that an N-H dissociation product for NMF would be further stabilized by interacting through either an unbound N or O atom, as shown in Figure 3d or 3e, respectively. To determine which structures correspond to the two observed surface products, detailed analysis of the experimental spectrum for NMF (Figure 4a) will be carried out, facilitated by simulated absorption spectra (Figure 4b-d) and by energetic arguments. Figure 4a is the experimental chemisorption spectrum previously shown in Figure 2b. Figure 4b is the theoretical spectrum of an O dative-bonded adduct on a Ge2Si7H12 cluster with Ge--H interaction (Figure 3a). Figure 4c and 4d are the theoretical spectra, calculated using two-dimer Ge4Si19H24 across-trench clusters, of N-H dissociation interdimer structures covalently bound to germanium through an oxygen (Figure 3d) and nitrogen (Figure 3e), respectively. In the calculations, both dissociation products interact through their other heteroatom with a germanium atom in the adjacent empty dimer of the Ge4Si19H24 cluster. These structures are shown in Figure 5.

414 J. Phys. Chem. C, Vol. 111, No. 1, 2007

Keung et al.

Figure 4. Infrared spectra of NMF on Ge(100)-2×1: (a) saturation exposure at 310 K, (b) one-dimer theoretical spectrum of NMF oxygen dativebonded with Ge--H interaction, (c) two-dimer theoretical spectrum of Ge-O bonded, N-H dissociation product with interdimer interaction, (d) two-dimer theoretical spectrum of Ge-N bonded, N-H dissociation product with interdimer interaction, (e) multilayers at 150 K (scaled).

Spectra for the unidentate N-H dissociation products (Figures 3b and 3c) are not shown because these showed poor agreement with the experimental spectra. Finally, the experimental physisorbed multilayer spectrum at a crystal temperature of 150 K is shown in Figure 4e. Several peaks in the chemisorption spectrum (Figure 4a), i.e. those at 2967, 1622, 1327, and 1109 cm-1, match the theoretical O dative-bonded spectrum (Figure 4b) quite well. These same modes dominate the low temperature infrared absorption spectrum of Figure 2a, indicating that the O dative-bonded product is formed at both room temperature and low temperature. On the basis of the theoretical spectrum, we assign the peak at 1622 cm-1 in the chemisorption spectrum to the normal mode of an O dative-bonded structure involving asymmetric υas(N-C′)O) stretching coupled with aldehydic δal(C′-H) bending. This agrees well with the previous assignment of the 1640 cm-1 peak in a saturation spectrum of DMF.16 In addition, the 1622 cm-1 mode is red-shifted from that of the multilayer (1667 cm-1) which is expected for a dative-bonded product.16,20 The 1327 cm-1 mode is assigned to an aldehydic δal(C′-H) bending mode similarly observed at 1352 cm-1 for DMF.16 The 1109 cm-1 peak is assigned to a methyl wagging deformation, which is very weak in the theoretical dative-bonded spectrum (Figure 4b). Peaks in the 1350 to 1450 cm-1 region are not clearly resolved but are assigned generally to methyl δ(CH3) deformations. However, some peaks in that region may also arise from an N-H dissociation product which also possesses a methyl group. The final mode associated with the O dative-bonded structure is at 2967 cm-1. This mode supports the existence of a Ge--H interaction in the O dative-bonded adduct, as described in the following analysis. This region is typically associated with υ(C-H) stretching modes, while υ(N-H) stretching modes for

secondary amides usually appear much higher (between 3420 and 3440 cm-1).30 However, we do not believe the 2967 cm-1 peak arises from υ(C-H) stretching for several reasons. First, we did not observe peaks near that region in our previous studies on tertiary amides16 and other methyl-containing carbonyl molecules17,18 for which C-H bonds were also present. Second, this mode is not observed for the N-H dissociation products, as will be discussed below. Third, υ(C-H) stretches are generally weak in intensity in our infrared system, as evident in the multilayer spectrum in which they are not observed (Figure 4e). Consequently, we assign the 2967 cm-1 mode to an υ(N-H) stretching mode. The mode is significantly redshifted and intensified due to its interaction with the germanium atom in the O dative-bonded structure shown in Figure 3a. The theoretical spectrum of an O dative-bonded structure with Ge--H interaction strongly supports this assignment, with an unusually intense and red-shifted υ(N-H) stretching mode which appears at 2904 cm-1 (Figure 4b). The calculated relative intensity of the 2904 cm-1 υ(N-H) peak is three times that of the calculated 1669 cm-1 υas(N-C′)O) peak. We note that although present, the 2967 cm-1 peak is much weaker in the low-temperature spectrum (Figure 2a). We speculate that the peak attenuation relative to the other two modes at 1622 and 1337 cm-1 may be a temperature-dependent effect, in which additional dative-bonded configurations are stabilized at 240 K compared to at 310 K. For instance, some of these conformations may not involve Ge--H interactions, leading to a less intense and less red-shifted υ(N-H) stretching mode. This temperature-dependent effect may also explain the 10 cm-1 shift observed for the aldehydic δal(C′-H) bending mode of the dative-bonded product (from 1337 to 1327 cm-1) between the low temperature and room-temperature spectra (Figure 2a and 2b, respectively). Deuterated NMF may be

Thermal Control of Amide Product Distributions

Figure 5. Energy-minimized N-H dissociation structures for NMF on across-trench two-dimer Ge4Si19H24 clusters: (a) Ge-O bonded with Ge--N interaction and (b) Ge-N bonded with Ge--O interaction. Atom colors representations: red-germanium, green-oxygen, brown-carbon, blue-nitrogen, white-hydrogen, light brown-silicon.

helpful in further confirming our assignments, but we were unable to obtain deuterated NMF for these experiments. After assigning modes for the oxygen dative-bonded surface adduct, two clear spectral features, at 1557 and 1960 cm-1, remain to be assigned in Figure 4a. These are the same two modes that dominate the 450 K spectrum (Figure 2c). The 1960 cm-1 peak falls in the expected region for a υ(Ge-H) stretching mode and hence is assigned to a germanium hydride. The Ge-H species must arise from either N-H or C-H dissociation. However, C-H dissociation is not expected to occur in this system because it was not observed in the analogous molecule, DMF, and because N-H bonds are more prone to heterolytic/ acidic dissociation compared to C-H bonds.31 Consequently, we assign the υ(Ge-H) stretching mode to the product of an N-H dissociation reaction. The broadness of the peak may result from the Ge-H moieties being in several different surface environments. Notably, the other spectral feature, at 1557 cm-1, is consistent with an N-H dissociation product only if an interdimer interaction is present. The low-energy location of the carbonylrelated peak at 1557 cm-1 is significantly red-shifted from the “typical” location above 1600 cm-1 of carbonyl and asymmetric amide stretches even when they are dative-bonded.16,20,30,32,33 Theoretical calculations of the infrared spectra for two N-H

J. Phys. Chem. C, Vol. 111, No. 1, 2007 415 dissociation products (Ge-O and Ge-N bound, Figure 3b and 3c, respectively) that lack interdimer interactions have high carbonyl- and imine-related stretching modes above 1700 cm-1 (data not shown), well above the observed value of 1557 cm-1. On the other hand, theoretical spectra of structures with interdimer interactions (Figure 4c-d) reveal red-shifts in the carbonyl- and imine-related asymmetric stretching modes and show better agreement with the 1557 cm-1 mode in the experimental spectrum of Figure 4a. Thus the mode at 1557 cm-1 is assigned to either an interdimer υas(N-C′)O) (for Ge-N bonded adduct) or υas(O-C′)N) (for Ge-O bonded adduct) asymmetric stretch. We have also carried out a comparison of the energies of the N-H dissociation product (Ge-N bound) with interdimer interaction versus the noninterdimer product, both on two-dimer Ge4Si19H24 across-trench clusters at the lower 6-31G(d) level. This comparison found that the interdimer interaction contributes 10 kcal/mol additional stability than without the interdimer interaction. This gain in stability, together with the red-shifted 1557 cm-1 location of the carbonyl/imine-containing normal mode in the 310 K chemisorption spectrum of NMF (Figure 4a), leads us to believe an interdimer N-H dissociation structure is present. 3.2. NMF: Theoretical Pathway and Energetics. Based on the surface products assigned above for NMF, a reaction pathway, shown in Figure 6, was calculated to check the energetic consistency of the proposed surface products with the low temperature and annealing experimental spectra (Figure 2). The reaction pathway of NMF was calculated using only onedimer clusters since calculating single point energies of twodimer clusters at the 6-311++G(2df,pd) level is computationally too expensive with our resources. The N-H dissociation products with interdimer interactions (Figure 3d,e) are therefore not shown on the pathway. However, we expect the interdimer products to be even more stable than the N-H dissociation products without interdimer interactions. In addition, to confirm the validity of using a Ge2Si7H12 cluster in place of a Ge9H12 cluster for these calculations, binding energies were calculated for the O dative-bonded structure with Ge--H interaction (Figure 3a) on Ge9H12 and were found to differ from the result on Ge2Si7H12 by only 0.7 kcal/mol (-20.1 and -19.4 kcal/mol, respectively). Binding energies calculated for the Ge-N bound N-H dissociation structure of NMF (Figure 3c) on Ge9H12 differed from the result on Ge2Si7H12 by only 2.8 kcal/mol (-31.8 and -34.6 kcal/mol, respectively). In addition, infrared vibrational frequencies calculated for these two structures on Ge9H12 and Ge2Si7H12 clusters, at the 6-31G(d) level of theory, showed less than a 1% difference between clusters. Starting with the reference energy as the energy level of the reactants (energy-minimized Ge2Si7H12 cluster and free NMF), the reaction pathway (Figure 6) shows the O dative-bonded structure with Ge--H interaction (Figure 3a) is formed with a binding energy of 19.4 kcal/mol. Passing over an activation barrier of 8.5 kcal/mol leads to an N-H dissociation product bound to Ge through the O with a binding energy of 28.4 kcal/mol (Figure 3b). However, there is a more stable product in which the amide fragment is bonded to Ge through the N instead of the O (Ebinding ) 34.6 kcal/mol) (Figure 3c). This product can be reached by passing through a four-membered ring transition state that is 19.3 kcal/mol below the energy of the reactants. According to the experimental assignments, the O dativebonded product corresponds with the low-temperature spectrum (Figure 2a). At room temperature, two principal products are observed, the dative-bonded product and one of the N-H

416 J. Phys. Chem. C, Vol. 111, No. 1, 2007

Keung et al.

Figure 6. Proposed theoretical reaction pathway for NMF. The reference energy (E ) 0 kcal/mol) is where a free NMF molecule and an unconstrained single-dimer Ge2Si7H12 cluster are minimized in energy. All geometry optimizations were performed at the 6-31G(d) level with subsequent singlepoint energies calculated at the 6-311++G(2df,pd) level.

dissociation products (Figure 3d or 3e). Upon annealing the substrate, only the N-H dissociation product is seen. The experimental observations are consistent with the theoretical reaction pathway. At low temperature, there is insufficient thermal energy to surmount the 8.5 kcal/mol calculated kinetic barrier to N-H dissociation, and therefore the dative-bonded product can be kinetically isolated. At room temperature, some products are trapped in the dative-bonded state while others proceed to the dissociation product. Annealing to 450 K apparently provides enough energy for the majority of absorbed molecules to surmount the kinetic barrier to N-H dissociation, allowing NMF to reach the more thermodynamically stable N-H dissociated state. Annealing may also provide enough energy to surmount the 9.1 kcal/mol kinetic barrier to form the Ge-N bound N-H dissociation product from the Ge-O bound product. However, it is important to note that distinguishing between the Ge-O and Ge-N bound N-H dissociation structures with infrared spectroscopy is difficult since they are so similar structurally and chemically. Further work and a different technique, such as X-ray photoelectron spectroscopy, will be required to distinguish these structures. Nevertheless, the analysis indicates that by varying the substrate temperature, one can convert from a kinetically controlled product distribution at low temperature to a thermodynamically controlled product distribution at high temperature. The annealing study also confirms that the two proposed surface products lie on a common reaction pathway since we are able to convert the kinetically favorable product (O dativebonded) into the thermodynamically favorable product (N-H dissociation) by raising the surface temperature. This is best seen in the annealing difference spectrum (Figure 2d). Without any additional exposure of NMF, the 1557 and 1960 cm-1 peaks associated with the thermodynamic product grow in while there are negative 2967, 1622, 1327, and 1109 cm-1 peaks indicating at least some conversion of the kinetic product on the surface into the thermodynamic product. The difference spectrum also helps associate the 2967 and 1109 cm-1 peaks with the kinetic product, which, as previously discussed, are not very prominent in Figure 2a possibly due to slightly different binding orientations at lower temperatures.

3.3. Acetamides: Molecular Analogues to Formamides. To examine the generality of the reactivity pattern shown above for NMF (and DMF), we have investigated an analogous system of two acetamides, NMA and DMA (Figure 1c and 1d, respectively). These acetamides are similar to NMF and DMF in that one is a secondary amide and the other is a tertiary amide. The difference between these two systems is a methyl group in the acetamides substituted for the aldehydic hydrogen in the formamides. In addition to the reactions observed for NMF, the presence of the R-carbon methyl group opens the possibility of an “-ene” type reaction like that observed with acetone.20 Figure 7 shows two sets of spectra, with the formamides in the lower half and the acetamides in the top half. The theoretical spectra of DMF and DMA oxygen dative-bonded to Ge2Si7H12 clusters are shown in Figure 7a and 7d, respectively. The chemisorption spectra of DMF and DMA taken at 310 K are shown in Figure 7b and 7e, respectively. The chemisorption spectra for NMF and NMA, also taken at 310 K, are shown in Figure 7c and 7f, respectively. The two sets of spectra exhibit similarities between the formamides (Figure 7a-c) and the acetamides (Figure 7d-f), suggesting similar surface products. Detailed analysis of the vibrational spectra provides further information on specific adsorption species. We start with the tertiary amides (Figure 7b and 7e). The chemisorption spectra for both tertiary amides, i.e., that of DMF and DMA, match well with the respective theoretical spectra of the dative-bonded products. Neither DMF nor DMA exhibit any peaks in the 1900-2000 cm-1 ν(Ge-H) stretching region, consistent with the formation of a dativebonded product. In the case of DMA, the absence of a Ge-H mode further rules out the presence of any “-ene” type products. One difference between the chemisorption spectrum of DMA compared to that for DMF is that for the former, a doublet appears near 1600 cm-1, the region associated with the asymmetric υas(O)C′-N) stretch in DMF at 1640 cm-1. We postulate that this doublet may be due to different bonding environments created by the steric nature of the additional methyl group attached to the carbonyl carbon. In the secondary amides, a methyl group is replaced with an N-H bond. The spectrum of NMA (Figure 7f) exhibits a similar

Thermal Control of Amide Product Distributions

Figure 7. Infrared spectra for formamides and acetamides: (a) theoretical spectrum of DMF oxygen dative-bonded to a Ge2Si7H12 cluster,16 (b) chemisorption spectrum of DMF at 310 K,16 (c) chemisorption spectrum of NMF at 310 K, (d) theoretical spectrum of DMA oxygen dative-bonded to a Ge2Si7H12 cluster, (e) chemisorption spectrum of DMA at 310 K, (f) chemisorption spectrum of NMA at 310 K.

pattern to that of the related NMF (Figure 7c). In Figure 7f, a υ(Ge-H) stretching mode near 1937 cm-1 is observed along with a red-shifted carbonyl/imine-related mode at 1541 cm-1, similar to the red-shifted mode at 1557 cm-1 in the chemisorption spectrum of NMF (Figure 7c). By analogy to the NMF spectral analysis, these features are assigned to an interdimer N-H dissociation product at the surface. The peak at 1389 cm-1 for NMA is also assigned to an interdimer structure as a methyl δ(CH3) deformation mode. A second product is also apparent in the spectrum, with two small features in similar locations to the 1614 and 1591 cm-1 doublet in the chemisorption spectrum of DMA (Figure 7e). By analogy to DMA, these are assigned to the O dative-bonded product. The peaks associated with the dative-bonded structure near 1600 cm-1 are weaker in intensity for NMA than for NMF, however, suggesting lower relative coverage of the NMA dative-bonded product. The transition state between the dative-bonded minimum and the N-H dissociated minimum for NMA may involve charge transfer from the nitrogen to the carbonyl carbon, resulting in partial cationic character of the nitrogen atom. We therefore speculate that the additional methyl group, known to stabilize cationic and radical species in classical organic chemistry, may provide increased stabilization of this transition state, leading to a smaller kinetic barrier for N-H dissociation and consequently a lower surface coverage of the dative-bonded product for NMA. 3.4. Formamide. We have demonstrated a general pattern of reactivity for secondary amides, in which both dative bonding and interdimer N-H dissociation products are formed. However, one more level of complexity exists in the series of amides, namely primary amides which possess two N-H bonds. Figure 8 shows the chemisorption spectra of formamide taken during a coverage dependence study from 0.01 to 8.34 L (saturation). The chemisorption spectrum for NMF is included for compari-

J. Phys. Chem. C, Vol. 111, No. 1, 2007 417

Figure 8. Infrared spectra of (a) NMF chemisorbed at 310 K on Ge(100) and (b) coverage-dependent experiment with formamide on Ge(100) at 310 K.

Figure 9. Infrared spectra from an annealing study of formamide: (a) formamide chemisorbed on Ge(100) at 310 K and (b) spectrum taken after a 450 K anneal.

son. Formamide and NMF exhibit similar spectral features at 1641 and 1622 cm-1, respectively, attributed to υas(O)C′-N) asymmetric stretches in oxygen dative-bonded surface adducts along with peaks in the 1300-1400 cm-1 region that could be assigned to either methyl δ(CH3) or δ(N-H) deformation modes. There is also a 1539 cm-1 peak similar to the 1557 cm-1 peak in NMF. It is assigned to an interdimer υas(O-C′)N) or υas(O)C′-N) asymmetric stretch from an N-H dissociation product. The υ(Ge-H) stretch observed is also assigned to an N-H dissociation product. Hence, formamide appears to form both an O dative-bonded product and an interdimer N-H dissociation product on Ge(100)-2×1 at room temperature. As with NMF, thermal annealing studies were carried out to probe the surface product distribution of formamide as a function of substrate temperature. Figure 9a shows a saturation chemisorption spectrum of formamide at 310 K. Figure 9b shows a spectrum taken after a 450 K anneal. Similar to NMF, the peaks associated with the dative-bonded structure attenuate at the higher temperature, while the peaks associated with the N-H dissociation product (υ(Ge-H) stretch and 1539 cm-1) peak

418 J. Phys. Chem. C, Vol. 111, No. 1, 2007 grow in. This behavior indicates that the dative-bonded structure converts to the N-H dissociation product on the germanium surface at higher temperatures. In addition to the modes at 1641 and 1539 cm-1, there is a large, broad signal in the 13001400 cm-1 region which we tentatively attribute to decomposition products. Formamide may be more susceptible to decomposition than NMF since it has one more reactive N-H functionality. While formamide and NMF share similar reactivities, they do have several differences as demonstrated in the coveragedependent spectrum (Figure 8b). The υas(O-C′)N) or υas(O)C′-N) asymmetric stretch in the 1641 cm-1 region is broader than that observed with NMF (Figure 8a). This could be due to hydrogen bonding interactions between adsorbates involving the free N-H groups not involved in dative bonding. In the structure shown in Figure 3a, with a Ge--H interaction, dative-bonded NMF lacks a free N-H group to participate in such intermolecular interactions. Formamide also exhibits strong coverage-dependent attenuation of the 1539 cm-1 peak (Figure 8b) while NMF exhibits barely noticeable attenuation (data not shown). We propose two possible explanations for the coverage-dependent attenuation of this peak, which has been assigned to the interdimer N-H dissociation product. The loss of this peak may be caused by reaction or decomposition of the N-H dissociation surface product.11,19 Alternatively, the N-H dissociation product may be retained, but the interdimer interaction lost. Addressing the first possibility, the N-H dissociation product for formamide still possesses one unreacted N-H group which may leave it more susceptible than NMF to further reaction or decomposition by impinging gas-phase molecules during high exposure doses. In contrast, the N-H dissociation product of NMF lacks a free N-H group and may therefore be less susceptible to further reaction/decomposition. The other possibility is that a higher coverage of dative-bonded adducts inhibits the interdimer interaction of the N-H dissociation product, similar to that observed with acetic acid.17 According to the theoretical calculations, a noninteracting N-H dissociation product (either Ge-O or Ge-N bound) should have a υas(O-C′)N) or υas(O)C′-N) asymmetric stretch above 1650 cm-1. There is a small peak that appears to grow in with coverage at 1695 cm-1 in Figure 8b; this peak may be a noninteracting asymmetric stretch, but further experiments are necessary to definitively make this assignment. 4. Conclusion Adsorption of a series of secondary and primary amides was investigated on Ge(100)-2×1 revealing similar oxygen dativebonded surface products to those observed with previously studied tertiary amides. In addition, N-H dissociation products were also observed, and on the basis of experimental and theoretical infrared spectra as well as energetic arguments, it was found these products formed interdimer structures similar to surface structures previously reported for acyl halides and acetic acid. The observation of interdimer products with amides is important in that it generalizes the interdimer structure to molecules with N-C-O backbones in addition to the O-C-O backbone in acetic acid and Ge-C-O backbone in the acyl halides. A reaction pathway was proposed for the model compound, NMF, showing adsorption into a dative-bonded state followed by N-H dissociation. Data from temperature and annealing experiments showed achievement of thermal control of the surface product distribution between dative-bonded (low tem-

Keung et al. perature) and N-H dissociation (high temperature) products. Control over surface product distribution will benefit the creation of homogeneous, ordered, and patterned surfaces. Also, the N-H dissociation products leave reactive functionalities (either an imine or carbonyl group) free for subsequent layer reactions. These discoveries may eventually have applications in thin film devices or molecular-based electronic elements. Acknowledgments. This work was supported by a grant from the National Science Foundation (CHE 0245260). A.J.K. would like to acknowledge support from the Stanford President’s Scholar Program and the Merck Award for Student Research. M.A.F. thanks the National Science Foundation for support in the form of a Graduate Research Fellowship. References and Notes (1) Chabinyc, M. L.; Loo, Y.-L. J. Macromol. Sci. - Polym. ReV. 2006, 46, 1. (2) Newman, C. R.; Frisbie, C. D.; da Silva, D. A.; Bredas, J. L.; Ewbank, P. C.; Mann, K. R. Chem. Mater. 2004, 16, 4436. (3) Chesterfield, R. J.; McKeen, J. C.; Newman, C. R.; Ewbank, P. C.; da Silva, D. A.; Bredas, J. L.; Miller, L. L.; Mann, K. R.; Frisbie, C. D. J. Phys. Chem. B 2004, 108, 19281. (4) Lee, T. W.; Zaumseil, J.; Bao, Z. N.; Hsu, J. W. P.; Rogers, J. A. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 429. (5) Reese, C.; Roberts, M.; Ling, M.-M.; Bao, Z. Mater. Today 2004, 7, 20. (6) Ling, M. M.; Bao, Z. N. Chem. Mater. 2004, 16, 4824. (7) Weidkamp, K. P.; Afzali, A.; Tromp, R. M.; Hamers, R. J. J. Am. Chem. Soc. 2004, 126, 12740. (8) Weidkamp, K. P.; Hacker, C. A.; Schwartz, M. P.; Cao, X. P.; Tromp, R. M.; Hamers, R. J. J. Phys. Chem. B 2003, 107, 11142. (9) Kwang Seok, L.; Blanchet, G. B.; Feng, G.; Yueh-Lin, L. Appl. Phys. Lett. 2005, 86, 74102. (10) Strother, T.: Hamers, R. J.; Smith, L. M. Nucleic Acids Res. 2000, 28, 3535. (11) Hamers, R. J.; Coulter, S. K.; Ellison, M. D.; Hovis, J. S.; Padowitz, D. F.; Schwartz, M. P.; Greenlief, C. M.; Russell, J. N. Acc. Chem. Res. 2000, 33, 617. (12) Hersam, M. C.; Guisinger, N. P.; Lyding, J. W. Nanotechnology 2000, 11, 70. (13) Lopinski, G. P.; Wayner, D. D. M.; Wolkow, R. A. Nature 2000, 406, 48. (14) Santini, J. T.; Cima, M. J.; Langer, R. Nature 1998, 397, 335. (15) Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J. J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005, 307, 538. (16) Keung, A. J.; Filler, M. A.; Porter, D. W.; Bent, S. F. Surf. Sci. 2005, 599, 41. (17) Filler, M. A.; Van Deventer, J. A.; Keung, A. J.; Bent, S. F. J. Am. Chem. Soc. 2006, 128, 770. (18) Filler, M. A.; Keung, A. J.; Porter, D. W.; Bent, S. F. J. Phys. Chem. B 2006,110, 4115. (19) Bent, S. F. J. Phys. Chem. B 2002, 106, 2830. (20) Wang, G. T.; Mui, C.; Musgrave, C. B.; Bent, S. F. J. Phys. Chem. B 2001, 105, 12559. (21) Kim, A.; Filler, M. A.; Kim, S.; Bent, S. F. J. Am. Chem. Soc. 2005, 127, 6123. (22) 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, Gaussian, Inc., Wallingford CT, 2004. (23) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.

Thermal Control of Amide Product Distributions (24) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (25) Becke, A. D. Journal of Chem. Phys. 1993, 98, 5648. (26) Kohn, W.; Sham, L. J. Phys. ReV. 1965, 140, 1133. (27) Mui, C.; Bent, S. F.; Musgrave, C. B. J. Phys. Chem. A 2000, 104, 2457. (28) Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502. (29) Wang, G. T.; Mui, C.; Tannaci, J. F.; Filler, M. A.; Musgrave, C. B.; Bent, S. F. J. Phys. Chem. B 2003, 107, 4982.

J. Phys. Chem. C, Vol. 111, No. 1, 2007 419 (30) Joseph H. F. S.; Lambert, B.; David, A.; Lightner, R.; Cooks, G. Organic structural spectroscopy; Prentice Hall: Upper Saddle River, NJ, 1998. (31) Vollhardt, K. P. C.; Schore, N. E. Organic Chemistry, Structure and Function; W.H. Freeman and Company: New York, 2003. (32) Durgaprasad, G.; Sathyanarayana, D. N.; Patel, C. C. Bull. Chem. Soc. Jpn. 1971, 44, 316. (33) Jao, T. C.; Scott, I. J. Mol. Spectrosc. 1982, 92, 1.