9406
Langmuir 2005, 21, 9406-9408
Trichlorosilane Isocyanate as Coupling Agent for Mild Conditions Functionalization of Silica-Coated Surfaces Nicolas Arde`s-Guisot,† Jean-Olivier Durand,*,† Michel Granier,† Aurore Perzyna,† Yannick Coffinier,‡ Bruno Grandidier,‡ Xavier Wallart,‡ and Didier Stievenard‡ Chimie Mole´ culaire et Organization du Solide, UMR 5637 Case 007, Universite´ Montpellier 2, Place Euge` ne Bataillon, 34095 Montpellier Cedex 05, and Institut d’Electronique, de Microe´ lectronique et de Nanotechnologie, IEMN, UMR 8520 De´ partement ISEN, 41 Bd Vauban, 59046 Lille, Cedex France Received May 11, 2005 Despite the importance of the isocyanate group in chemistry, very few examples of isocyanate-modified silicas have been reported, and all of the strategies described so far led to partial or total hydrolysis or condensation of the isocyanate group. By synthesizing trichlorosilane isocyanate as the coupling reagent, we show that oxidized silicon wafers are successfully modified with the isocyanate group. Our method is achieved in mild conditions, at low temperature, without side-reactions and allows the formation of a self-assembled monolayer (SAM) of isocyanates. The isocyanate group then offers a flexible way to further functionalize silica substrates with different nucleophiles, due to its high and specific reactivity.
The isocyanate group has been used as an important coupling function in a wide area of chemistry from polymers and materials engineering to biomolecule science, due to its high and specific reactivity.1 Despite the importance of this group, the preparation of silica substrates with an isocyanate remains a difficult task. Although only a few attempts have been reported on glass slides,2 no spectroscopic proof of the presence of the isocyanate at the surface has been achieved so far. Furthermore, all of the strategies led to partial or total hydrolysis or side reactions of the isocyanate group when spectroscopic analyses were performed with silicas or solgel hybrid materials.3 Here, we relate the synthesis of 10-isocyanatodecyltrichlorosilane and show that this compound is particularly appropriate for the successful and easy preparation of an isocyanate-terminated monolayer on oxidized silicon wafers, without side reactions. The described surface modification overcomes the limitations of the previous methods and allows us to address the issue of the subsequent reactivity of the isocyanatefunctionalized surface. Although the synthesis of trichlorosilylpropylisocyanate was described 30 years ago,4 the further use of this compound, particularly for the attachment of the isocyanate group to a silica substrate has never been inves* To whom correspondence should be addressed. E-mail:
[email protected]. † Universite ´ Montpellier 2. ‡ IEMN. (1) Katritsky, A.; Rees, C.; Meth-Cohn, O. Comprehensive Organic Functional Group Transformation; Pergamon Press: Elmsford, NY, 1995; Vol. 5. (2) (a) Chun, Y. S.; Ha, K.; Lee, Y. J.; Lee, J. S.; Kim, H. S.; Park, Y. S.; Yoon, K. B. Chem Com. 2002, 1846. (b) Garimella, V. Patent 2003, WO 2003006676. (c) Olivier, C.; Hot, D.; Huot, L.; Ollivier, N.; El-Mahdi, O.; Gouyette, C.; Huynh-Dinh, T.; Gras-Masse, H.; Lemoine, Y.; Melnyk, O. Bioconj. Chem. 2003, 14, 430. (3) (a) Loy, D. A.; Baugher, C. R.; Schneider, C. R.; Duane, A.; Sanchez, A. F. Polym. Preprints 2001, 42, 180. (b) Timofte, R. S.; Woodward, S. Tetrahedron Lett. 2004, 45, 39. (c) Huh, S.; Wiench, J. W.; Yoo, J.-C.; Pruski, M.; Lin, V. S.-Y. Chem. Mater. 2003, 15, 4247. (d) Radu, D. R.; Lai, C. Y.; Jeftinija, K.; Rowe, E. W.; Jeftinija, S.; Lin, V. S.-Y. J. Am. Chem. Soc. 2004, 126, 13216. (d) Huesing, N.; Schubert, U.; Misof, K.; Fratzl, P. Chem. Mater. 1998, 10, 3024. (4) (a) Pepe, E. J. U.S. Patent, 3,511,866 (19700512), 1970. (b) Pepe, E. J. Patent 1971 DE 1938743 19710211, (c) Fan, Y. L.; Shaw, R. G. J. Org. Chem. 1973, 38, 2410. (d) Berger, A. U.S. Patent 4,097,511 (19780627), 1978.
Scheme 1. Synthesis of 10-Isocyanatodecyltrichlorosilane; (i) HSiCl3, Karstedt Catalyst, RT, 94%
tigated since then. As it is better to graft silanes in a two-dimensional polycondensation manner to obtain a monolayer of well-organized molecules, the number n of carbon atoms should be 30 > n > 8 with trialkylchorosilanes (CnH2n+1SiCl3) to achieve this purpose.5 Thus, we synthesized 10-isocyanatodecyltrichlorosilane by hydrosilylation of 9-decenylisocyanate. Speier’s catalyst, which was efficient with the more reactive allylisocyanate,4 gave us an uncomplete reaction with byproducts. Indeed hydrosilylation of alkenes has been reported to compete with the reaction of the isocyanate group.6 In contrast, Karstedt’s catalyst allowed clean isolation of the chlorosilane isocyanate in 94% yield and no hydrosilylation of the isocyanate bonds was noticed (Scheme 1). The product was stable for months under an inert atmosphere at 0-4°C and easy to handle using standard techniques for moisture sensitive reagents. The grafting reaction was studied in situ by ATR-FTIR, on single silicon wafer crystal (100) with a silica layer of 1.8-2 nm thickness at the surface.7 The silane concentration was 10-2 M in CHCldCCl2 (Scheme 2). The first experiments were carried out without any base, and we observed an important decrease of the isocyanate band at 2274 cm-1 after 30 min with the appearance of two new bands at 1775 and 1730 cm-1. Since these bands are characteristic of the formation of carbamoyl chloride8 and urethane groups due to the release of HCl during the (5) Ulman, A. Chem. Rev. 1996, 96, 1533. (d) Ulman, A., Ed.; Organic Thin Films and Surfaces: Directions for the Nineties; Academic Press: New York, 1995. (6) Neumann, D.; Fisher, M.; Tran, L.; Matisons, J. G. J. Am. Chem. Soc. 2002, 124, 13998. (7) Granier, M.; Lanneau, G. F.; Moineau, J.; Girard, P.; Ramonda, M. Langmuir 2003, 19, 2691. (8) (a) Blacklock, T. J.; Shuman R. F.; Butcher J. W.; Shearin, Jr., W. E.; Budavari, J.; Grenda, V. J. J. Org. Chem. 1988, 53, 836. (b) Slocombe, R. J.; Hardy, E. E.; Saunders, J. H.; Jenkins, R. L. J. Am. Chem. Soc. 1950, 72, 1888.
10.1021/la051256r CCC: $30.25 © 2005 American Chemical Society Published on Web 09/07/2005
Letters
Langmuir, Vol. 21, No. 21, 2005 9407
Scheme 2. Grafting of 10-Isocyanatodecyltrichlorosilane; (i) (iPr)2EtN, CHCldCCl2, 5°C; (ii) NuH, 5°C
Table 1. Nucleophilic Addition (NuH 10-1 M in CHCldCCl2) on Isocyanate-Supported SAM NuH
IR (cm-1) CdO
q
thickness (Å)
roughness (Å)
p-toluidine p-anisidine p-methoxyphenola p-bromothiophenolb CH3(CH2)11SH b CH3(CH2)11NH2 NH2-NH-BOC NH2-NH-Fmoc
1645 1646 1707 1669 1650 1625 1725, 1660 c
83° 77° 78° 83° 89° 92° 80° 83°
23 23 21 22 25 25 18 22
1.4 1.7 1.7 1.9 1.6 2.2 1.7 2.3
a Et N was added with NuH (5 × 10-1 M CHCldCCl ). b Et N 3 2 3 (10-1 M CHCldCCl2). c DMF as solvent.
Figure 1. FTIR-ATR of grafted trichlorosilyldecylisocyanate. Background (wafer with 1.8-2 nm silica layer) was subtracted.
grafting procedure, its addition to the isocyanate group and further reaction with the silica surface, 10 equiv of diisopropylethylamine were added to trap HCl and avoid its side-reactions. The kinetic showed then a regular increase of the isocyanate band for 50 min. After 1 h, the isocyanate band at 2274 cm-1 was stable and did not increase any more what indicated the completion of the self-assembled monolayer (Figure 1). In addition, the νas CH2 and νs CH2 bands, observed at 2931 and 2858 cm-1 at the beginning of the process, progressively shifted to 2925 and 2854 cm-1 respectively, providing evidence for the organization of the aliphatic chains. A wide range of amines and thiols nucleophiles were then reacted onto the isocyanate-modified surface. Results are summarized in Table 1. The reaction was very fast with all of the amino nucleophiles and completed in less than half an hour. With aliphatic thiols, the reaction was slower and took 1 h for completion. Thiophenol and phenol reacted as fast as amino nucleophiles as they were deprotonated by triethylamine. In each case, FTIR-ATR spectra contained one or two CdO bands, which demonstrated the functionalization of the surface. Support for the formation of a monolayer was established by AFM
Figure 2. XPS C1s spectra of the NH2-NH-BOC modified surface, before (a) and after (b) the removal of the t-Boc group.
(low roughness in agreement with an homogeneous coverage7) and ellipsometry measurements which yield film thicknesses consistent with fully extended chains for the aromatic NuHs (theoretical values between 22 and 23 Å). With the dodecyl nucleophiles, the thicknesses (theoretical values between 29 and 30 Å) and contact angles (value of 103° with fully organized octadecyl chains perpendicular to the surface9a) were lower than expected. Thus aliphatic chains were not as dense as completely organized octadecyl chains, because of gauche conformations9b with decyl and dodecyl chains which result in slight disorder and tilting in the monolayer. The wavenumbers of 1626 cm-1 for amide I and 1575 cm-1 for amide II, observed with the dodecyamine nucleophile, are in agreement with strong associations of the urea groups.10 X-ray photoelectron spectroscopy further confirmed the modification of the surfaces by isocyanate and its further reactivity with the Fmoc (see the Supporting Information) and Boc-hydrazine nucleophiles (Figure 2a). Indeed, in this latter case, deconvolution11 of the C1s spectrum exhibits six peaks, consistent with the presence of the N-C*(dO)-N carbon atoms and all other Boc-hydrazine nucleophile constituents. To get more insight into the chemical transformation of the Fmoc and Boc terminated surfaces, standard deprotection of these two groups (DMF piperidine, and TFA CH2Cl2 respectively) were achieved. Only a 10% yield was obtained for the Fmoc terminated surface, whereas the Boc groups were completely removed as shown in Figure 2b. Such result as well as the ellipsometry result, where the expected film thickness for fully extended chains is respectively 23-24 and 21-22 Å for the Fmoc and Boc terminated surfaces, strongly supports the theoretical prediction for a lower conformational flexibility of the Fmoc compared to the Boc protecting group.12 A high density of Fmoc groups at the surface prevents piperidine from reacting following general base E1cB catalysis, whereas the Boc protecting group can adopt more disordered conformations at the surface allowing general acid catalysis E1 elimination. In conclusion, trichlorosilane isocyanate is a powerful precursor to terminate the surface of silica substrates with the isocyanate group in mild conditions, without sidereactions. Such group allows an easy tailoring of the surface chemistry, since a wide range of nucleophiles has (9) (a) Monsathaporn, S.; Effenberger, F. Langmuir 2004, 20, 10375. (b) Sieval, A. B.; Demirel, A. L.; Nissink, J. W. M.; Linford, M. R.; Van der Maas, J. H.; de Jeu, W. H.; Zuilhof, H.; Sudho¨lter, E. J. R. Langmuir 1998, 14, 1759. (10) Moreau, J. J. E.; Vellutini, L.; Wong Chi Man, M.; Bied, C. J. Am. Chem. Soc. 2001, 123, 1509. (11) Coffinier, Y.; Olivier, C.; Perzyna, A.; Grandidier, G.; Wallart, X.; Durand, J. O.; Melnyk, O.; Stie´venard, D. Langmuir 2005, 21, 1489. (12) Broda, M. A.; Mazur, L.; Koziol, A. E.; Rzeszotarska, B. J. Peptide Sci. 2004, 10, 448.
9408
Langmuir, Vol. 21, No. 21, 2005
been then successfully attached to the oxidized silicon surface. Finally, the use of a C10 chain is suitable for the preparation of dense well-ordered monolayers, which are required for a better understanding of the cleavage reactions of protecting groups frequently used in the attachment of bioactive moieties. Work is in progress to exploit this methodology for biochip applications and results will be reported in due course.
Letters
Supporting Information Available: XPS data for the Fmoc-terminated surface. Experimental details on the preparation and reactivity of the isocyanate-terminated surfaces. FTIRATR spectra of the nucleophilic addition on the isocyanatesupported SAM. This material is available free of charge via the Internet at http://pubs.acs.org. LA051256R
