Stable Perfluorosilane Self-Assembled Monolayers ... - ACS Publications

Nov 21, 2006 - spectrometer using Al KR1,2 monochromatic radiation at 15. kV and 150 W and operating ... performed in a range of 0-1100 eV binding ene...
0 downloads 0 Views 312KB Size
798

Chem. Mater. 2007, 19, 798-804

Stable Perfluorosilane Self-Assembled Monolayers on Copper Oxide Surfaces: Evidence of Siloxy-Copper Bond Formation E. Hoque,*,† J. A. DeRose,‡ R. Houriet,§ P. Hoffmann,| and H. J. Mathieu† LMCH, IMX; LTCM, ISE; LTP, IMX; LOA, IOA; Ecole Polytechnique Fe´ de´ rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland ReceiVed September 27, 2006. ReVised Manuscript ReceiVed NoVember 21, 2006

Formation of stable and hydrophobic self-assembled monolayers (SAMs) on an oxidized copper (Cu) surface has been accomplished via reaction with 1H,1H,2H,2H-perfluorodecyldimethylchlorosilane (PFMS). The perfluoroalkyl SAMs showed sessile drop static contact angles of more than 125° for pure water and stability against exposure to boiling water, boiling nitric acid solution, and warm sodium hydroxide solution for up to 30 min. X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared reflection/absorption spectroscopy (FT-IRRAS) data reveal a coordination of the PFMS silicon (Si) atom with a cuprate (CuO) molecule present on the oxidized copper substrate. The data give good evidence that the stability of the SAM film on the PFMS-modified oxidized Cu surface is largely due to the formation of a siloxy-copper (-Si-O-Cu-) bond via a condensation reaction between silanol (-Si-OH) and copper hydroxide (-CuOH). The extremely hydrophobic and stable SAMs on oxidized Cu could have useful applications in corrosion inhibition for micro/nanoelectronics and/or heat exchange surfaces exploiting dropwise condensation.

Introduction This report presents the results of self-assembled monolayers (SAMs) formed on oxidized copper (Cu) substrates via reaction with 1H,1H,2H,2H-perfluorodecyldimethylchlorosilane (PFMS). These SAMs are able to protect the Cu substrate from very harsh, corrosive conditions. Although SAMs formed by reaction of alkylsilanes with silicon oxides and alkanethiols with Cu and gold (Au) substrates have been extensively investigated,1-5 there is no report in the public domain on SAMs formed by perfluorosilanization of an oxidized copper (Cu) substrate (PFMS/Cu) to the best of the authors’ knowledge. A significant obstacle to the implementation of SAMs in everyday applications is the poor durability of these thin films under ambient or more extreme conditions. Formation of SAMs on Cu substrates has become an important field of both fundamental and applied research, for example, corrosion inhibition6-10 of Cu used in micro* Corresponding author. Fax: +41 21 693 3946. Tel: +41 693 2972. E-mail: [email protected]. † LMCH, IMX. ‡ LTCM, ISE. § LTP, IMX. | LOA, IOA.

(1) Pellerite, M. J.; Wood, E. J.; Jones, V. W. J. Phys. Chem. B 2002, 106, 4746. (2) Jennings, G. K.; Laibinis, P. E. Colloids Surf., Physicochem. Eng. Aspects 1996, 116, 105. (3) Laiho, T.; Leiro, J. A.; Heinonen, M. H.; Mattila, S. S.; Lukkari, J. J. Electron Spectrosc. Relat. Phenom. 2005, 142, 105. (4) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem. ReV. 2005, 105, 1103. (5) Sung, M. M.; Kim, Y. Bull. Korean Chem. Soc. 2001, 22, 748. (6) Whelan, C. M.; Kinsella, M.; Ho, H. M.; Maex, K. J. Electrochem. Soc. 2004, 151, B33. (7) Tan, Y. S.; Srinivasan, M. P.; Pehkonen, S. O.; Chooi, S. Y. M. J. Vac. Sci. Technol. A 2004, 22, 1917. (8) Tremont, R.; De Jesus-Cardona, H.; Garcia-Orozco, J.; Castro, R. J.; Cabrera, C. R. J. Appl. Electrochem. 2000, 30, 737.

electronic device fabrication and heat exchange surfaces which exploit dropwise condensation.11 Cu is destined to replace aluminum (Al) alloy as the principle VLSI/ULSI (very large- or ultra large-scale integration) interconnection material due to its lower electrical and thermal resistivity and higher resistance to electromigration. Despite these advantages, one of the major concerns when using it as an interconnection material is its susceptibility to corrosion and oxidation. SAMs and other thin films on Cu have been shown to inhibit corrosion,6,7,12 however, the durability of alkanethiol SAM/Cu films is too low for most practical applications. Heat transfer is most efficient via condensation of a fluid medium onto a heat exchange surface. Dropwise condensation has been known to produce heat transfer coefficients up to 20 times that of filmwise condensation; it is for this reason that, over the past few decades, considerable attention has been paid toward the development of suitable dropwise condensation promoters, but the main problem has been durability of these promoters.11,13 The structure, chemical composition, wetting properties, and stability of the PFMS/Cu surface have been analyzed by using a multiple technique approach: X-ray photoelectron spectroscopy (XPS), Fourier transform infrared reflection/ absorption spectroscopy (FT-IRRAS), atomic force microscopy (AFM), and contact angle measurement (CAM). The data show that the PFMS/Cu SAM is chemisorbed onto the (9) Skolnik, A. M.; Hughes, W. C.; Augustine, B. H. Chem. Educator 2000, 5, 8. (10) Laibinis, P. E.; Whitesides, G. M. J. Am. Chem. Soc. 1992, 114, 9022. (11) Rose, J. W. J. Power Energy: Proc. Inst. Mech. Engrs., Part A 2002, 216, 115. (12) Hoque, E.; DeRose, J. A.; Hoffmann, P.; Mathieu, H. J. Surf. Interface Anal. 2006, 38, 62. (13) Das, A. K.; Kitty, H. P.; Marto, P. J.; Andeen, G. B.; Kumar, K. Trans. ASME, J. Heat Transfer 2000, 122, 278.

10.1021/cm062318h CCC: $37.00 © 2007 American Chemical Society Published on Web 01/30/2007

Perfluorosilane SAMs on Copper Oxide Surfaces

Chem. Mater., Vol. 19, No. 4, 2007 799

oxidized Cu surface via a siloxyscopper (-Si-O-Cu-) bond. As chlorosilanes are rapidly hydrolyzed by water and form silanols:14 -Si-Cl + H2O f -Si-OH + HCl the data collected lead to the conclusion that perfluorodecyldimethylsilanol (C8F17(CH2)2(CH3)2SiOH) undergoes a condensation reaction with copper hydroxide (CuOH) present on the oxidized Cu surface to form a siloxy-copper bond: C8F17(CH2)2(CH3)2Si-OH + OH-Cu- f C8F17(CH2)2(CH3)2Si-O-CuExperimental Section Circular Cu (99.98%) discs of 3-mm thickness and 12mm diameter were used for all experiments. Copper discs were mechanically polished with silicon carbide (SiC) paper of grades 500-4000 and a micropolishing pad until the copper surface showed a highly reflective “mirror” finish. All the polished samples were ultrasonicated in ethyl acetate, isopropyl alcohol, and deionized water (DI H2O) for 10 min each, blown dry, oxidized by exposure to a 5% aqueous hydrogen peroxide (H2O2) solution for 5-10 min at room temperature, and then reacted with 1H,1H,2H,2H-perfluorodecyldimethylchlorosilane (F3C(CF2)7(CH2)2(CH3)2SiCl; CAS #74612-30-9) [PFMS] (ABCR, Germany) by immersion for 50 min into a 0.1% (wt) PFMS solution using hexane (H3C(CH2)4CH3; Merck, Darmstatdt, Germany) as the solvent. Surface roughness of polished Cu and PFMS/Cu samples was measured from high-resolution topographic images obtained with a Park Scientific Instruments atomic force microscope (AFM) at room temperature and atmospheric pressure in contact mode. The polished, oxidized Cu and chemically modified Cu surface composition was analyzed by a Kratos Axis Ultra X-ray photoelectron spectrometer using Al KR1,2 monochromatic radiation at 15 kV and 150 W and operating in ultrahigh vacuum (UHV) with a base pressure below 10-9 mbar. Survey scans were performed in a range of 0-1100 eV binding energy with a pass energy of 80 eV and single peaks were measured with a pass energy of 20 eV. The core-level signals were obtained at a photoelectron takeoff angle of 90° (with respect to the sample surface). The binding energy (BE) values were referenced to the carbon C 1s photoelectron peak at 284.8 eV.15,16 After linear-type background subtraction, the deconvolution of high-resolution XPS peaks has been done by mixed Gaussian-Lorentzian (G/L) fit. Quantitative analysis of the species at the sample surface was performed using the XPS area peak intensities. XPS analysis was performed using the standard software CASA XPS.17 The estimated relative error for all XPS data used for elemental quantifica(14) Kallury, K. M. R.; Krull, U. J.; Thompson, M. Org. Mass Spectrom. 1991, 26, 81. (15) Briggs, D. D.; Seah, M. P. Practical Surface Analysis by Auger and X-ray Photoelec. Spec.; John Wiley & Sons: New York, 1983. (16) Hoque, E.; DeRose, J. A.; Hoffmann, P.; Mathieu, H. J.; Bhushan, B.; Cichomski, M. J. Chem. Phys. 2006, 124, 174710-1. (17) Fairley, N.; Carrick, A. The Casa CookBook: Recipes for XPS Data Processing; Acolyte Science: Chesire: UK, 2005.

Figure 1. Water sessile drop static contact angle measurement (S-CAM) data taken on (1) unmodified, oxidized Cu; (2) untreated PFMS/Cu; (3) PFMS/Cu treated with boiling HNO3 of pH 1.8 for 30 min); (4) PFMS/Cu treated with boiling H2O for 30 min; and (5) PFMS/Cu treated with NaOH of pH 12 at 60 °C for 30 min.

tion is (2%. Fourier transform infrared reflection/absorption spectroscopy (FT-IRRAS) was carried out with a PerkinElmer Spectrum optical bench using a Spectratech Grazing Angle FT-80 system equipped with a wire grid polarizer. The measurements were done with polarized light at grazing incidence (80° angle of incidence with respect to the surface normal). All spectra were recorded by collecting 256 scans over the 4000-400 cm-1 spectral region with a resolution of 4 cm-1. Measurements were performed with a 90° polarization angle (electric field vector, E, direction with respect to the sample surface; p-polarized). Static water sessile drop contact angle measurements (S-CAMs) and dynamic water sessile drop contact angle measurements (DCAMs) for determining the wetting properties of the unmodified Cu, PFMS-modified Cu, and stability-tested (treated) PFMS-modified Cu were performed with a GBX Digidrop contact angle system using ultrapure H2O (Fluka) at 75-90% relative humidity (RH) and 23-25 °C. Liquid drops of volume < 10 µL were used for S-CAMs and 5-10 µL for D-CAMs. All sessile drop contact angles have an absolute error or uncertainty of (3°. Stability tests were done by immersing the PFMS/Cu samples in boiling H2O, a boiling nitric (HNO3) acid aqueous solution with a pH of 1.8, and/or a sodium hydroxide (NaOH) solution at 60 °C with a pH of 12 for 30 min. Results and Discussion Characterization of PFMS/Cu SAM Film. The rms surface roughness of unmodified Cu and PFMS/Cu was measured over an area of 5 × 5 µm2 by AFM and found to be less than 15 nm. The AFM images for both the samples showed similar morphology. Figure 1 is a histogram showing the S-CAM data for pure H2O on the following: (1) unmodified Cu (