pubs.acs.org/Langmuir © 2010 American Chemical Society
Characteristic IR CdC Stretch Enhancement in Monolayers by Nonconjugated, Noncumulated Unsaturated Bonds Michael V. Lee,*,† Dominik Enders,‡ Tadaaki Nagao,‡ and Katsuhiko Ariga‡ †
International Center for Young Scientists (ICYS), and ‡World Premier International (WPI) Research Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS), Japan Received January 12, 2010. Revised Manuscript Received February 2, 2010
Control over and understanding of single-molecule covalent coatings becomes increasingly important in tailoring surfaces during the fabrication of nanoscale electrical or optical elements, such as organic field-effect transistors and light-emitting devices as well as microelectromechanical systems as the relevant feature sizes decrease. In this work, we develop a model based on IR spectra from public databases and DFT calculations that can be used to semiquantitatively assess the level of double bonds in monolayer coatings. We use the model to show the enhancement of the CdC vibrational mode due to silicon substitution and also from additional unsaturated bonds. Simple models for other functional groups in organic monolayers could be produced similarly.
As feature sizes shrink in electrical, mechanical, and optical devices, unconventional materials are required and the interface between them becomes more important. Because of current leakage on that scale, silicon dioxide is already being replaced as the dielectric in nanoscale electronics; other higher-k dielectrics, whether organic, inorganic, or hybrid, are required to reduce this loss.1 A single monolayer with extremely low stiction is required to provide longevity to microscale and nanoscale electromechanical devices.2-4 Coatings on nanoscale porous silicon enhance and maintain electroluminescence.5-7 Nanometer-scale coatings with precise thicknesses are used to match the Fermi levels for charge injection in organic LEDs.8,9 At thicknesses approaching the molecular scale in these coatings, the covalent bond connecting the monolayer to the substrates becomes a significant fraction of the coating, which influences the structure of the layers and has a dramatic effect on the resulting properties. Success on the nanoscale requires a precise understanding of and control over binding between the substrate and a single monolayer. Many methods use complex instruments and processes to deposit films, yet even with the most common semiconductor, silicon, and some of the most simple monolayers, the structure is still not completely understood. In 1993, Linford and Chidsey reported the first monolayers using wet chemistry with carbon covalently bound directly to silicon rather than silicon dioxide;10 as such, it provided a direct connection to silicon and required only a typical organic laboratory. Initially, a peroxide was used *Corresponding author. E-mail:
[email protected]. (1) Ortiz, R. P.; Facchetti, A.; Marks, T. J. Chem. Rev. 2010, 110, 205–239. (2) Ashurst, W. R.; Yau, C.; Carraro, C.; Lee, C.; Kluth, G. J.; Howe, R. T.; Maboudian, R. Sens. Actuators, A 2001, 91, 239–248. (3) Lorenz, C.; Webb, E.; Stevens, M.; Chandross, M.; Grest, G. Tribol. Lett. 2005, 19, 93–98. (4) Choi, J.; Kawaguchi, M.; Kato, T. J. Appl. Phys. 2002, 91, 7574–7576. (5) Stewart, M. P.; Buriak, J. M. J. Am. Chem. Soc. 2001, 123, 7821–7830. (6) Boukherroub, R.; Morin, S.; Wayner, D.; Bensebaa, F.; Sproule, G.; Baribeau, J.; Lockwood, D. Chem. Mater. 2001, 13, 2002–2011. (7) Gelloz, B.; Sano, H.; Boukherroub, R.; Wayner, D. D. M.; Lockwood, D. J.; Koshida, N. Phys. Status Solidi C 2005, 2, 3273–3277 and references therein . (8) Matsushima, T.; Kinoshita, Y.; Murata, H. Appl. Phys. Lett. 2007, 91, 253504-3–. (9) Yang, C.; Chih, Y. J. Phys. Chem. B 2006, 110, 19412–19417. (10) Linford, M. R.; Chidsey, C. E. D. J. Am. Chem. Soc. 1993, 115, 12631– 12632.
4594 DOI: 10.1021/la1001418
Figure 1. Major products of reactions of 1-alkenes and 1-alkynes with hydrogen-terminated silicon. Double bonds are highlighted in red. The R group can include any organic molecule; in this letter, it represents either a saturated alkyl chain or an alkyl chain with a terminal double or triple bond.
to attack the hydrogen-terminated silicon (H-Si) surface and produce a reactive silicon-centered radical that mediates monolayer formation. Subsequently, the direct binding of 1-alkenes has been catalyzed by numerous methods.5,11-17 In the original reaction, an unsaturated bond in a hydrocarbon is attacked by a silicon-centered surface radical. On silicon surfaces including Si(100) and Si(111), this produces a carboncentered radical that can then abstract a hydrogen from a neighboring surface silicon atom, which regenerates the siliconcentered radical to repeat the process.11,18,19 The basic reactions for 1-alkenes and 1-alkynes generally produce products 1 and 4, (11) Linford, M. R.; Fenter, P.; Eisenberger, P. M.; Chidsey, C. E. D. J. Am. Chem. Soc. 1995, 117, 3145–3155. (12) Terry, J.; Linford, M. R.; Wigren, C.; Cao, R.; Pianetta, P.; Chidsey, C. E. D. Appl. Phys. Lett. 1997, 71, 1056–1058. (13) Lee, K.; Pan, F.; Carroll, G. T.; Turro, N. J.; Koberstein, J. T. Langmuir 2004, 20, 1812–1818. (14) Sun, Q.; de Smet, L. C. P. M.; van Lagen, B.; Giesbers, M.; Thune, P. C.; van Engelenburg, J.; de Wolf, F. A.; Zuilhof, H.; Sudholter, E. J. R. J. Am. Chem. Soc. 2005, 127, 2514–2523. (15) Scheres, L.; Arafat, A.; Zuilhof, H. Langmuir 2007, 23, 8343–8346. (16) Petit, A.; Delmotte, M.; Loupy, A.; Chazalviel, J.; Ozanam, F.; Boukherroub, R. J. Phys. Chem. C 2008, 112, 16622–16628. (17) Faucheux, A.; Gouget-Laemmel, A. C.; Henry de Villeneuve, C.; Boukherroub, R.; Ozanam, F.; Allongue, P.; Chazalviel, J. Langmuir 2006, 22, 153–162. (18) Lopinski, G. P.; Wayner, D. D. M.; Wolkow, R. A. Nature 2000, 406, 48–51. (19) Eves, B. J.; Sun, Q.; Lopinski, G. P.; Zuilhof, H. J. Am. Chem. Soc. 2004, 126, 14318–14319.
Published on Web 03/08/2010
Langmuir 2010, 26(7), 4594–4597
Lee et al.
Letter
Figure 2. (a) Plots of ACdC/AC-H versus the number of carbons in each molecule for the homologous series of 1-alkenes (circles) and 1-alkenylsilanes (triangles). The filled symbols represent spectra from NIST databases; the open symbols represent spectra produced by DFT calculations. The 1-alkene database spectra were fitted with a power series, which was then linearly scaled by 4.2 to fit the 1-alkenylsilane data points to produce the model. (b) Noise(pk-pk)CdC/AC-H from published monolayer spectra that showed no peak near 1640 cm-1. Green represents the IR spectra of monolayers from a 1-alkene reagent, and red represents an early IR spectrum of a monolayer prepared from a 1-alkyne. (c) ACdC/AC-H for monolayers prepared from alkynes plotted versus the 1-alkenylsilane model and also against the line for 40% of molecules with a double bond plus 60% without. (d) ACdC/AC-H for monolayers prepared from alkenes. The filled red diamond is a monolayer formed from a dialkene under the same conditions and from the same paper as one of the samples from b, represented by the green symbols. The open diamonds are DFT calculations of additional molecules as depicted. (e) ACdC/AC-H for monolayers from a dialkyne (red) and 1-alkyne (green) from the same paper, formed under the same conditions. The open diamonds are from DFT calculations for comparison.
respectively, shown in Figure 1. A recent near-edge X-ray absorption fine-structure spectroscopy (NEXAFS) study, which provides a low-level quantitative measurement of unsaturation in monolayers, discovered a previously unobserved 20% alkene content in monolayers produced by thermal hydrosilylation as either product 2 or 3 shown in Figure 1.20 It is likely that other methods of initiation for forming monolayers from 1-alkenes that are theorized to follow the same reaction pathways would also produce monolayers with some fraction of adsorbed molecules including a double bond. Apart from the recent NEXAFS study, monolayer examination has generally been qualitative with respect to double bonds, most commonly noting the presence or absence of the CdC stretch at 1700 cm-1. Other methods of reaction with 1-alkynes have also shown products 5 and 6.21,22 It would be useful to have a model that estimates the number of double bonds represented by an observed peak or the detection limit when a peak is not observed. The present effort creates a model for semiquantitatively evaluating double bond content from IR monolayer spectra with a CdC peak or for estimating the detection limit for double bonds in a particular spectrum when the peak is absent. The relative maximum intensity of the characteristic absorbance from the CdC stretching vibrational mode (∼1640 cm-1), ACdC, versus the maximum absorbance in the alkyl C-H region (28003100 cm-1), AC-H, in spectra of 1-alkenes from the publicly (20) Lee, M. V.; Lee, J. R. I.; Brehmer, D. E.; Linford, M. R.; Willey, T. M. Langmuir 2010, 26, 1512–1515. (21) Hurley, P. T.; Nemanick, E. J.; Brunschwig, B. S.; Lewis, N. S. J. Am. Chem. Soc. 2006, 128, 9990–9991. (22) Robins, E. G.; Stewart, M. P.; Buriak, J. M. Chem. Commun. 1999, 2479– 2480.
Langmuir 2010, 26(7), 4594–4597
available NIST Database of Infrared Spectra23,24 provided the basis for the model. An equation fit to ACdC/AC-H from the NIST 1-alkene spectra was scaled by a factor of 4.2, as shown in Figure 2a, to align with ACdC/AC-H obtained from the DFT vibrational analysis of 1-alkenylsilanes, thus forming our model.25 The model neglects intermolecular effects and assumes that molecules bound to the silicon surface will have a similar IR spectrum to those bound to silane. Although there are many methods available for calculating absorption spectra, NWChem (6-31þg*/B3LYP and 3-21g/B3LYP) is a standard publicly available package for DFT calculations.26-28 NWChem is free for research use and provides straightforward analysis for the (23) NIST Mass Spec Data Center; Stein, S. E., Director; In NIST Chemistry WebBook, NIST Standard Reference Database Number 69; Linstrom, P. J., Mallard, W. G., Eds.; National Institute of Standards and Technology: Gaithersburg, MD, 2008. (24) Coblentz, Inc. In NIST Chemistry WebBook, NIST Standard Reference Database Number 69; Linstrom, P. J., Mallard, W. G., Eds.; National Institute of Standards and Technology: Gaithersburg MD, 2008. (25) Bylaska, E.; de Jong, W.; Kowalski, K.; Straatsma, T.; Valiev, M.; Wang, D.; Apra, E.; Windus, T.; Hirata, S.; Hackler, M.; Zhao, Y.; Fan, P.; Harrison, R.; Dupuis, M.; Smith, D.; Nieplocha, J.; Tipparaju, V.; Krishnan, M.; Auer, A.; Nooijen, M.; Brown, E.; Cisneros, G.; Fann, G.; Fr€uchtl, H.; Garza, J.; Hirao, K.; Kendall, R.; Nichols, J.; Tsemekhman, K.; Wolinski, K.; Anchell, J.; Bernholdt, D.; Borowski, P.; Clark, T.; Clerc, D.; Dachsel, H.; Deegan, M.; Dyall, K.; Elwood, D.; Glendening, E.; Gutowski, M.; Hess, A.; Jaffe, J.; Johnson, B.; Ju, J.; Kobayashi, R.; Kutteh, R.; Lin, Z.; Littlefield, R.; Long, X.; Meng, B.; Nakajima, T.; Niu, S.; Pollack, L.; Rosing, M.; Sandrone, G.; Stave, M.; Taylor, H.; Thomas, G.; van Lenthe, J.; Wong, A.; Zhang, Z. NWChem, A Computational Chemistry Package for Parallel Computers, version 5.0; Pacific Northwest National Laboratory: Richland, WA; 2006. (26) Malloci, G.; Mulas, G.; Joblin, C. Astron. Astrophys. 2004, 426, 105. (27) Groenewold, G. S.; Gianotto, A. K.; Cossel, K. C.; Van Stipdonk, M. J.; Moore, D. T.; Polfer, N.; Oomens, J.; de Jong, W. A.; Visscher, L. J. Am. Chem. Soc. 2006, 128, 4802–4813. (28) Balabanov, N. B.; Peterson, K. A. J. Chem. Phys. 2003, 119, 12271–12278.
DOI: 10.1021/la1001418
4595
Letter
Figure 3. Graphical representations of DFT calculation results for (a) partial electrical charges on 1-octene, (b) the partial charges on 1-octenylsilane, (c) partial charges on the molecule from b with the terminal C-C bond replaced with a triple bond, and (d) the HOMO for the molecule from c. In a-c, red represents negative charge, blue color represents positive charge, and white represents neutral charge. The dark-blue atoms in b and c are silicon, and the rest are carbon and hydrogen and can be differentiated by size. In d, the light-blue atom is silicon, gray atoms are carbon, and white are hydrogen.
average scientist to replicate this type of model. The calculations are self-consistent and fit the published experimental data for free molecules, although they may underestimate the double bond content in monolayers (Supporting Information). The model is evaluated on the basis of published IR spectra and recent NEXAFS results and also provides a straightforward explanation for the enhancement seen with monolayers formed from difunctional molecules. A comparison of the ACdC/AC-H model from the 1-alkenes and 1-alkenylsilanes helps in assessing the signal-to-noise ratio. As shown in Figure 2b, by comparing the peak-to-peak noise (1600-1700 cm-1), Noise(pk-pk)CdC, relative to AC-H, we can determine whether the presence of double bonds can be conclusively dismissed. When we compare our model to published (29) Buriak, J. M.; Stewart, M. P.; Geders, T. W.; Allen, M. J.; Choi, H. C.; Smith, J.; Raftery, D.; Canham, L. T. J. Am. Chem. Soc. 1999, 121, 11491–11502.
4596 DOI: 10.1021/la1001418
Lee et al.
spectra for monolayers prepared from 1-alkenes that did not show a CdC peak in IR,6,16,17,29-32 we see that each of the samples could comprise at least 15% of molecules with a SiCdCR stretch among saturated adsorbates without showing a CdC peak in the IR. If thermal hydrosilylation monolayers include a combination of products 2 and 3 from Figure 1, then this level of double bond content is consistent with recent quantitative results by NEXAFS.20 A notable case of a low signal-to-noise ratio is an IR spectrum of a monolayer prepared from a 16-carbon 1-alkyne;33 as represented by the red in Figure 2b, the noise level is clearly high enough to mask a complete 1-hexadecenyl monolayer. Even though the X-ray reflectivity data of that same sample suggested the presence of a double bond, because the CdC peak at 1600 cm-1 was not observed the authors assumed only single bonds. In a subsequent paper (16-carbon in Figure 2b) on similar monolayers with an improved signal-to-noise ratio (Figure 2c, 16-carbons), the same group reported a CdC peak.15 In Figure 2c, we see that ACdC/AC-H in published spectra from 1-alkynes falls in a reasonable range of 20-70%,5,15,29,34 although without confirmation by NEXAFS absolute scaling is tentative. The data points exhibit significant scatter, suggesting real differences between the methods; a thorough study of combined IR and NEXAFS would be important in determining whether this is sample variation or intrinsic to each method. Figure 2d provides a surprising contrast to the straightforward presentation in Figure 2b,c. Even though a CdC peak was not observed in an IR spectrum of a monolayer formed from 1-decene (Figure 2b), in the same paper a spectrum from a monolayer from a dialkene shows ACdC/AC-H that is at least 50% greater than expected (i.e., a monolayer with SiCdCR bonds represented by the 0.15 ML curve in Figure 2b added to ACdC/AC-H for terminal alkenes sums to
